diff --git a/Lib/asyncio/__init__.py b/Lib/asyncio/__init__.py
new file mode 100644
index 0000000..afc444d
--- /dev/null
+++ b/Lib/asyncio/__init__.py
@@ -0,0 +1,33 @@
+"""The asyncio package, tracking PEP 3156."""
+
+import sys
+
+# The selectors module is in the stdlib in Python 3.4 but not in 3.3.
+# Do this first, so the other submodules can use "from . import selectors".
+try:
+    import selectors  # Will also be exported.
+except ImportError:
+    from . import selectors
+
+# This relies on each of the submodules having an __all__ variable.
+from .futures import *
+from .events import *
+from .locks import *
+from .transports import *
+from .protocols import *
+from .streams import *
+from .tasks import *
+
+if sys.platform == 'win32':  # pragma: no cover
+    from .windows_events import *
+else:
+    from .unix_events import *  # pragma: no cover
+
+
+__all__ = (futures.__all__ +
+           events.__all__ +
+           locks.__all__ +
+           transports.__all__ +
+           protocols.__all__ +
+           streams.__all__ +
+           tasks.__all__)
diff --git a/Lib/asyncio/base_events.py b/Lib/asyncio/base_events.py
new file mode 100644
index 0000000..32457eb
--- /dev/null
+++ b/Lib/asyncio/base_events.py
@@ -0,0 +1,606 @@
+"""Base implementation of event loop.
+
+The event loop can be broken up into a multiplexer (the part
+responsible for notifying us of IO events) and the event loop proper,
+which wraps a multiplexer with functionality for scheduling callbacks,
+immediately or at a given time in the future.
+
+Whenever a public API takes a callback, subsequent positional
+arguments will be passed to the callback if/when it is called.  This
+avoids the proliferation of trivial lambdas implementing closures.
+Keyword arguments for the callback are not supported; this is a
+conscious design decision, leaving the door open for keyword arguments
+to modify the meaning of the API call itself.
+"""
+
+
+import collections
+import concurrent.futures
+import heapq
+import logging
+import socket
+import subprocess
+import time
+import os
+import sys
+
+from . import events
+from . import futures
+from . import tasks
+from .log import asyncio_log
+
+
+__all__ = ['BaseEventLoop', 'Server']
+
+
+# Argument for default thread pool executor creation.
+_MAX_WORKERS = 5
+
+
+class _StopError(BaseException):
+    """Raised to stop the event loop."""
+
+
+def _raise_stop_error(*args):
+    raise _StopError
+
+
+class Server(events.AbstractServer):
+
+    def __init__(self, loop, sockets):
+        self.loop = loop
+        self.sockets = sockets
+        self.active_count = 0
+        self.waiters = []
+
+    def attach(self, transport):
+        assert self.sockets is not None
+        self.active_count += 1
+
+    def detach(self, transport):
+        assert self.active_count > 0
+        self.active_count -= 1
+        if self.active_count == 0 and self.sockets is None:
+            self._wakeup()
+
+    def close(self):
+        sockets = self.sockets
+        if sockets is not None:
+            self.sockets = None
+            for sock in sockets:
+                self.loop._stop_serving(sock)
+            if self.active_count == 0:
+                self._wakeup()
+
+    def _wakeup(self):
+        waiters = self.waiters
+        self.waiters = None
+        for waiter in waiters:
+            if not waiter.done():
+                waiter.set_result(waiter)
+
+    @tasks.coroutine
+    def wait_closed(self):
+        if self.sockets is None or self.waiters is None:
+            return
+        waiter = futures.Future(loop=self.loop)
+        self.waiters.append(waiter)
+        yield from waiter
+
+
+class BaseEventLoop(events.AbstractEventLoop):
+
+    def __init__(self):
+        self._ready = collections.deque()
+        self._scheduled = []
+        self._default_executor = None
+        self._internal_fds = 0
+        self._running = False
+
+    def _make_socket_transport(self, sock, protocol, waiter=None, *,
+                               extra=None, server=None):
+        """Create socket transport."""
+        raise NotImplementedError
+
+    def _make_ssl_transport(self, rawsock, protocol, sslcontext, waiter, *,
+                            server_side=False, server_hostname=None,
+                            extra=None, server=None):
+        """Create SSL transport."""
+        raise NotImplementedError
+
+    def _make_datagram_transport(self, sock, protocol,
+                                 address=None, extra=None):
+        """Create datagram transport."""
+        raise NotImplementedError
+
+    def _make_read_pipe_transport(self, pipe, protocol, waiter=None,
+                                  extra=None):
+        """Create read pipe transport."""
+        raise NotImplementedError
+
+    def _make_write_pipe_transport(self, pipe, protocol, waiter=None,
+                                   extra=None):
+        """Create write pipe transport."""
+        raise NotImplementedError
+
+    @tasks.coroutine
+    def _make_subprocess_transport(self, protocol, args, shell,
+                                   stdin, stdout, stderr, bufsize,
+                                   extra=None, **kwargs):
+        """Create subprocess transport."""
+        raise NotImplementedError
+
+    def _read_from_self(self):
+        """XXX"""
+        raise NotImplementedError
+
+    def _write_to_self(self):
+        """XXX"""
+        raise NotImplementedError
+
+    def _process_events(self, event_list):
+        """Process selector events."""
+        raise NotImplementedError
+
+    def run_forever(self):
+        """Run until stop() is called."""
+        if self._running:
+            raise RuntimeError('Event loop is running.')
+        self._running = True
+        try:
+            while True:
+                try:
+                    self._run_once()
+                except _StopError:
+                    break
+        finally:
+            self._running = False
+
+    def run_until_complete(self, future):
+        """Run until the Future is done.
+
+        If the argument is a coroutine, it is wrapped in a Task.
+
+        XXX TBD: It would be disastrous to call run_until_complete()
+        with the same coroutine twice -- it would wrap it in two
+        different Tasks and that can't be good.
+
+        Return the Future's result, or raise its exception.
+        """
+        future = tasks.async(future, loop=self)
+        future.add_done_callback(_raise_stop_error)
+        self.run_forever()
+        future.remove_done_callback(_raise_stop_error)
+        if not future.done():
+            raise RuntimeError('Event loop stopped before Future completed.')
+
+        return future.result()
+
+    def stop(self):
+        """Stop running the event loop.
+
+        Every callback scheduled before stop() is called will run.
+        Callback scheduled after stop() is called won't.  However,
+        those callbacks will run if run() is called again later.
+        """
+        self.call_soon(_raise_stop_error)
+
+    def is_running(self):
+        """Returns running status of event loop."""
+        return self._running
+
+    def time(self):
+        """Return the time according to the event loop's clock."""
+        return time.monotonic()
+
+    def call_later(self, delay, callback, *args):
+        """Arrange for a callback to be called at a given time.
+
+        Return a Handle: an opaque object with a cancel() method that
+        can be used to cancel the call.
+
+        The delay can be an int or float, expressed in seconds.  It is
+        always a relative time.
+
+        Each callback will be called exactly once.  If two callbacks
+        are scheduled for exactly the same time, it undefined which
+        will be called first.
+
+        Any positional arguments after the callback will be passed to
+        the callback when it is called.
+        """
+        return self.call_at(self.time() + delay, callback, *args)
+
+    def call_at(self, when, callback, *args):
+        """Like call_later(), but uses an absolute time."""
+        timer = events.TimerHandle(when, callback, args)
+        heapq.heappush(self._scheduled, timer)
+        return timer
+
+    def call_soon(self, callback, *args):
+        """Arrange for a callback to be called as soon as possible.
+
+        This operates as a FIFO queue, callbacks are called in the
+        order in which they are registered.  Each callback will be
+        called exactly once.
+
+        Any positional arguments after the callback will be passed to
+        the callback when it is called.
+        """
+        handle = events.make_handle(callback, args)
+        self._ready.append(handle)
+        return handle
+
+    def call_soon_threadsafe(self, callback, *args):
+        """XXX"""
+        handle = self.call_soon(callback, *args)
+        self._write_to_self()
+        return handle
+
+    def run_in_executor(self, executor, callback, *args):
+        if isinstance(callback, events.Handle):
+            assert not args
+            assert not isinstance(callback, events.TimerHandle)
+            if callback._cancelled:
+                f = futures.Future(loop=self)
+                f.set_result(None)
+                return f
+            callback, args = callback._callback, callback._args
+        if executor is None:
+            executor = self._default_executor
+            if executor is None:
+                executor = concurrent.futures.ThreadPoolExecutor(_MAX_WORKERS)
+                self._default_executor = executor
+        return futures.wrap_future(executor.submit(callback, *args), loop=self)
+
+    def set_default_executor(self, executor):
+        self._default_executor = executor
+
+    def getaddrinfo(self, host, port, *,
+                    family=0, type=0, proto=0, flags=0):
+        return self.run_in_executor(None, socket.getaddrinfo,
+                                    host, port, family, type, proto, flags)
+
+    def getnameinfo(self, sockaddr, flags=0):
+        return self.run_in_executor(None, socket.getnameinfo, sockaddr, flags)
+
+    @tasks.coroutine
+    def create_connection(self, protocol_factory, host=None, port=None, *,
+                          ssl=None, family=0, proto=0, flags=0, sock=None,
+                          local_addr=None):
+        """XXX"""
+        if host is not None or port is not None:
+            if sock is not None:
+                raise ValueError(
+                    'host/port and sock can not be specified at the same time')
+
+            f1 = self.getaddrinfo(
+                host, port, family=family,
+                type=socket.SOCK_STREAM, proto=proto, flags=flags)
+            fs = [f1]
+            if local_addr is not None:
+                f2 = self.getaddrinfo(
+                    *local_addr, family=family,
+                    type=socket.SOCK_STREAM, proto=proto, flags=flags)
+                fs.append(f2)
+            else:
+                f2 = None
+
+            yield from tasks.wait(fs, loop=self)
+
+            infos = f1.result()
+            if not infos:
+                raise OSError('getaddrinfo() returned empty list')
+            if f2 is not None:
+                laddr_infos = f2.result()
+                if not laddr_infos:
+                    raise OSError('getaddrinfo() returned empty list')
+
+            exceptions = []
+            for family, type, proto, cname, address in infos:
+                try:
+                    sock = socket.socket(family=family, type=type, proto=proto)
+                    sock.setblocking(False)
+                    if f2 is not None:
+                        for _, _, _, _, laddr in laddr_infos:
+                            try:
+                                sock.bind(laddr)
+                                break
+                            except OSError as exc:
+                                exc = OSError(
+                                    exc.errno, 'error while '
+                                    'attempting to bind on address '
+                                    '{!r}: {}'.format(
+                                        laddr, exc.strerror.lower()))
+                                exceptions.append(exc)
+                        else:
+                            sock.close()
+                            sock = None
+                            continue
+                    yield from self.sock_connect(sock, address)
+                except OSError as exc:
+                    if sock is not None:
+                        sock.close()
+                    exceptions.append(exc)
+                else:
+                    break
+            else:
+                if len(exceptions) == 1:
+                    raise exceptions[0]
+                else:
+                    # If they all have the same str(), raise one.
+                    model = str(exceptions[0])
+                    if all(str(exc) == model for exc in exceptions):
+                        raise exceptions[0]
+                    # Raise a combined exception so the user can see all
+                    # the various error messages.
+                    raise OSError('Multiple exceptions: {}'.format(
+                        ', '.join(str(exc) for exc in exceptions)))
+
+        elif sock is None:
+            raise ValueError(
+                'host and port was not specified and no sock specified')
+
+        sock.setblocking(False)
+
+        protocol = protocol_factory()
+        waiter = futures.Future(loop=self)
+        if ssl:
+            sslcontext = None if isinstance(ssl, bool) else ssl
+            transport = self._make_ssl_transport(
+                sock, protocol, sslcontext, waiter,
+                server_side=False, server_hostname=host)
+        else:
+            transport = self._make_socket_transport(sock, protocol, waiter)
+
+        yield from waiter
+        return transport, protocol
+
+    @tasks.coroutine
+    def create_datagram_endpoint(self, protocol_factory,
+                                 local_addr=None, remote_addr=None, *,
+                                 family=0, proto=0, flags=0):
+        """Create datagram connection."""
+        if not (local_addr or remote_addr):
+            if family == 0:
+                raise ValueError('unexpected address family')
+            addr_pairs_info = (((family, proto), (None, None)),)
+        else:
+            # join addresss by (family, protocol)
+            addr_infos = collections.OrderedDict()
+            for idx, addr in ((0, local_addr), (1, remote_addr)):
+                if addr is not None:
+                    assert isinstance(addr, tuple) and len(addr) == 2, (
+                        '2-tuple is expected')
+
+                    infos = yield from self.getaddrinfo(
+                        *addr, family=family, type=socket.SOCK_DGRAM,
+                        proto=proto, flags=flags)
+                    if not infos:
+                        raise OSError('getaddrinfo() returned empty list')
+
+                    for fam, _, pro, _, address in infos:
+                        key = (fam, pro)
+                        if key not in addr_infos:
+                            addr_infos[key] = [None, None]
+                        addr_infos[key][idx] = address
+
+            # each addr has to have info for each (family, proto) pair
+            addr_pairs_info = [
+                (key, addr_pair) for key, addr_pair in addr_infos.items()
+                if not ((local_addr and addr_pair[0] is None) or
+                        (remote_addr and addr_pair[1] is None))]
+
+            if not addr_pairs_info:
+                raise ValueError('can not get address information')
+
+        exceptions = []
+
+        for ((family, proto),
+             (local_address, remote_address)) in addr_pairs_info:
+            sock = None
+            r_addr = None
+            try:
+                sock = socket.socket(
+                    family=family, type=socket.SOCK_DGRAM, proto=proto)
+                sock.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1)
+                sock.setblocking(False)
+
+                if local_addr:
+                    sock.bind(local_address)
+                if remote_addr:
+                    yield from self.sock_connect(sock, remote_address)
+                    r_addr = remote_address
+            except OSError as exc:
+                if sock is not None:
+                    sock.close()
+                exceptions.append(exc)
+            else:
+                break
+        else:
+            raise exceptions[0]
+
+        protocol = protocol_factory()
+        transport = self._make_datagram_transport(sock, protocol, r_addr)
+        return transport, protocol
+
+    @tasks.coroutine
+    def create_server(self, protocol_factory, host=None, port=None,
+                      *,
+                      family=socket.AF_UNSPEC,
+                      flags=socket.AI_PASSIVE,
+                      sock=None,
+                      backlog=100,
+                      ssl=None,
+                      reuse_address=None):
+        """XXX"""
+        if host is not None or port is not None:
+            if sock is not None:
+                raise ValueError(
+                    'host/port and sock can not be specified at the same time')
+
+            AF_INET6 = getattr(socket, 'AF_INET6', 0)
+            if reuse_address is None:
+                reuse_address = os.name == 'posix' and sys.platform != 'cygwin'
+            sockets = []
+            if host == '':
+                host = None
+
+            infos = yield from self.getaddrinfo(
+                host, port, family=family,
+                type=socket.SOCK_STREAM, proto=0, flags=flags)
+            if not infos:
+                raise OSError('getaddrinfo() returned empty list')
+
+            completed = False
+            try:
+                for res in infos:
+                    af, socktype, proto, canonname, sa = res
+                    sock = socket.socket(af, socktype, proto)
+                    sockets.append(sock)
+                    if reuse_address:
+                        sock.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR,
+                                        True)
+                    # Disable IPv4/IPv6 dual stack support (enabled by
+                    # default on Linux) which makes a single socket
+                    # listen on both address families.
+                    if af == AF_INET6 and hasattr(socket, 'IPPROTO_IPV6'):
+                        sock.setsockopt(socket.IPPROTO_IPV6,
+                                        socket.IPV6_V6ONLY,
+                                        True)
+                    try:
+                        sock.bind(sa)
+                    except OSError as err:
+                        raise OSError(err.errno, 'error while attempting '
+                                      'to bind on address %r: %s'
+                                      % (sa, err.strerror.lower()))
+                completed = True
+            finally:
+                if not completed:
+                    for sock in sockets:
+                        sock.close()
+        else:
+            if sock is None:
+                raise ValueError(
+                    'host and port was not specified and no sock specified')
+            sockets = [sock]
+
+        server = Server(self, sockets)
+        for sock in sockets:
+            sock.listen(backlog)
+            sock.setblocking(False)
+            self._start_serving(protocol_factory, sock, ssl, server)
+        return server
+
+    @tasks.coroutine
+    def connect_read_pipe(self, protocol_factory, pipe):
+        protocol = protocol_factory()
+        waiter = futures.Future(loop=self)
+        transport = self._make_read_pipe_transport(pipe, protocol, waiter)
+        yield from waiter
+        return transport, protocol
+
+    @tasks.coroutine
+    def connect_write_pipe(self, protocol_factory, pipe):
+        protocol = protocol_factory()
+        waiter = futures.Future(loop=self)
+        transport = self._make_write_pipe_transport(pipe, protocol, waiter)
+        yield from waiter
+        return transport, protocol
+
+    @tasks.coroutine
+    def subprocess_shell(self, protocol_factory, cmd, *, stdin=subprocess.PIPE,
+                         stdout=subprocess.PIPE, stderr=subprocess.PIPE,
+                         universal_newlines=False, shell=True, bufsize=0,
+                         **kwargs):
+        assert not universal_newlines, "universal_newlines must be False"
+        assert shell, "shell must be True"
+        assert isinstance(cmd, str), cmd
+        protocol = protocol_factory()
+        transport = yield from self._make_subprocess_transport(
+            protocol, cmd, True, stdin, stdout, stderr, bufsize, **kwargs)
+        return transport, protocol
+
+    @tasks.coroutine
+    def subprocess_exec(self, protocol_factory, *args, stdin=subprocess.PIPE,
+                        stdout=subprocess.PIPE, stderr=subprocess.PIPE,
+                        universal_newlines=False, shell=False, bufsize=0,
+                        **kwargs):
+        assert not universal_newlines, "universal_newlines must be False"
+        assert not shell, "shell must be False"
+        protocol = protocol_factory()
+        transport = yield from self._make_subprocess_transport(
+            protocol, args, False, stdin, stdout, stderr, bufsize, **kwargs)
+        return transport, protocol
+
+    def _add_callback(self, handle):
+        """Add a Handle to ready or scheduled."""
+        assert isinstance(handle, events.Handle), 'A Handle is required here'
+        if handle._cancelled:
+            return
+        if isinstance(handle, events.TimerHandle):
+            heapq.heappush(self._scheduled, handle)
+        else:
+            self._ready.append(handle)
+
+    def _add_callback_signalsafe(self, handle):
+        """Like _add_callback() but called from a signal handler."""
+        self._add_callback(handle)
+        self._write_to_self()
+
+    def _run_once(self):
+        """Run one full iteration of the event loop.
+
+        This calls all currently ready callbacks, polls for I/O,
+        schedules the resulting callbacks, and finally schedules
+        'call_later' callbacks.
+        """
+        # Remove delayed calls that were cancelled from head of queue.
+        while self._scheduled and self._scheduled[0]._cancelled:
+            heapq.heappop(self._scheduled)
+
+        timeout = None
+        if self._ready:
+            timeout = 0
+        elif self._scheduled:
+            # Compute the desired timeout.
+            when = self._scheduled[0]._when
+            deadline = max(0, when - self.time())
+            if timeout is None:
+                timeout = deadline
+            else:
+                timeout = min(timeout, deadline)
+
+        # TODO: Instrumentation only in debug mode?
+        t0 = self.time()
+        event_list = self._selector.select(timeout)
+        t1 = self.time()
+        argstr = '' if timeout is None else '{:.3f}'.format(timeout)
+        if t1-t0 >= 1:
+            level = logging.INFO
+        else:
+            level = logging.DEBUG
+        asyncio_log.log(level, 'poll%s took %.3f seconds', argstr, t1-t0)
+        self._process_events(event_list)
+
+        # Handle 'later' callbacks that are ready.
+        now = self.time()
+        while self._scheduled:
+            handle = self._scheduled[0]
+            if handle._when > now:
+                break
+            handle = heapq.heappop(self._scheduled)
+            self._ready.append(handle)
+
+        # This is the only place where callbacks are actually *called*.
+        # All other places just add them to ready.
+        # Note: We run all currently scheduled callbacks, but not any
+        # callbacks scheduled by callbacks run this time around --
+        # they will be run the next time (after another I/O poll).
+        # Use an idiom that is threadsafe without using locks.
+        ntodo = len(self._ready)
+        for i in range(ntodo):
+            handle = self._ready.popleft()
+            if not handle._cancelled:
+                handle._run()
+        handle = None  # Needed to break cycles when an exception occurs.
diff --git a/Lib/asyncio/constants.py b/Lib/asyncio/constants.py
new file mode 100644
index 0000000..79c3b93
--- /dev/null
+++ b/Lib/asyncio/constants.py
@@ -0,0 +1,4 @@
+"""Constants."""
+
+
+LOG_THRESHOLD_FOR_CONNLOST_WRITES = 5
diff --git a/Lib/asyncio/events.py b/Lib/asyncio/events.py
new file mode 100644
index 0000000..9724615
--- /dev/null
+++ b/Lib/asyncio/events.py
@@ -0,0 +1,395 @@
+"""Event loop and event loop policy."""
+
+__all__ = ['AbstractEventLoopPolicy', 'DefaultEventLoopPolicy',
+           'AbstractEventLoop', 'AbstractServer',
+           'Handle', 'TimerHandle',
+           'get_event_loop_policy', 'set_event_loop_policy',
+           'get_event_loop', 'set_event_loop', 'new_event_loop',
+           ]
+
+import subprocess
+import sys
+import threading
+import socket
+
+from .log import asyncio_log
+
+
+class Handle:
+    """Object returned by callback registration methods."""
+
+    def __init__(self, callback, args):
+        self._callback = callback
+        self._args = args
+        self._cancelled = False
+
+    def __repr__(self):
+        res = 'Handle({}, {})'.format(self._callback, self._args)
+        if self._cancelled:
+            res += '<cancelled>'
+        return res
+
+    def cancel(self):
+        self._cancelled = True
+
+    def _run(self):
+        try:
+            self._callback(*self._args)
+        except Exception:
+            asyncio_log.exception('Exception in callback %s %r',
+                                self._callback, self._args)
+        self = None  # Needed to break cycles when an exception occurs.
+
+
+def make_handle(callback, args):
+    # TODO: Inline this?
+    assert not isinstance(callback, Handle), 'A Handle is not a callback'
+    return Handle(callback, args)
+
+
+class TimerHandle(Handle):
+    """Object returned by timed callback registration methods."""
+
+    def __init__(self, when, callback, args):
+        assert when is not None
+        super().__init__(callback, args)
+
+        self._when = when
+
+    def __repr__(self):
+        res = 'TimerHandle({}, {}, {})'.format(self._when,
+                                               self._callback,
+                                               self._args)
+        if self._cancelled:
+            res += '<cancelled>'
+
+        return res
+
+    def __hash__(self):
+        return hash(self._when)
+
+    def __lt__(self, other):
+        return self._when < other._when
+
+    def __le__(self, other):
+        if self._when < other._when:
+            return True
+        return self.__eq__(other)
+
+    def __gt__(self, other):
+        return self._when > other._when
+
+    def __ge__(self, other):
+        if self._when > other._when:
+            return True
+        return self.__eq__(other)
+
+    def __eq__(self, other):
+        if isinstance(other, TimerHandle):
+            return (self._when == other._when and
+                    self._callback == other._callback and
+                    self._args == other._args and
+                    self._cancelled == other._cancelled)
+        return NotImplemented
+
+    def __ne__(self, other):
+        equal = self.__eq__(other)
+        return NotImplemented if equal is NotImplemented else not equal
+
+
+class AbstractServer:
+    """Abstract server returned by create_service()."""
+
+    def close(self):
+        """Stop serving.  This leaves existing connections open."""
+        return NotImplemented
+
+    def wait_closed(self):
+        """Coroutine to wait until service is closed."""
+        return NotImplemented
+
+
+class AbstractEventLoop:
+    """Abstract event loop."""
+
+    # Running and stopping the event loop.
+
+    def run_forever(self):
+        """Run the event loop until stop() is called."""
+        raise NotImplementedError
+
+    def run_until_complete(self, future):
+        """Run the event loop until a Future is done.
+
+        Return the Future's result, or raise its exception.
+        """
+        raise NotImplementedError
+
+    def stop(self):
+        """Stop the event loop as soon as reasonable.
+
+        Exactly how soon that is may depend on the implementation, but
+        no more I/O callbacks should be scheduled.
+        """
+        raise NotImplementedError
+
+    def is_running(self):
+        """Return whether the event loop is currently running."""
+        raise NotImplementedError
+
+    # Methods scheduling callbacks.  All these return Handles.
+
+    def call_soon(self, callback, *args):
+        return self.call_later(0, callback, *args)
+
+    def call_later(self, delay, callback, *args):
+        raise NotImplementedError
+
+    def call_at(self, when, callback, *args):
+        raise NotImplementedError
+
+    def time(self):
+        raise NotImplementedError
+
+    # Methods for interacting with threads.
+
+    def call_soon_threadsafe(self, callback, *args):
+        raise NotImplementedError
+
+    def run_in_executor(self, executor, callback, *args):
+        raise NotImplementedError
+
+    def set_default_executor(self, executor):
+        raise NotImplementedError
+
+    # Network I/O methods returning Futures.
+
+    def getaddrinfo(self, host, port, *, family=0, type=0, proto=0, flags=0):
+        raise NotImplementedError
+
+    def getnameinfo(self, sockaddr, flags=0):
+        raise NotImplementedError
+
+    def create_connection(self, protocol_factory, host=None, port=None, *,
+                          ssl=None, family=0, proto=0, flags=0, sock=None,
+                          local_addr=None):
+        raise NotImplementedError
+
+    def create_server(self, protocol_factory, host=None, port=None, *,
+                      family=socket.AF_UNSPEC, flags=socket.AI_PASSIVE,
+                      sock=None, backlog=100, ssl=None, reuse_address=None):
+        """A coroutine which creates a TCP server bound to host and port.
+
+        The return value is a Server object which can be used to stop
+        the service.
+
+        If host is an empty string or None all interfaces are assumed
+        and a list of multiple sockets will be returned (most likely
+        one for IPv4 and another one for IPv6).
+
+        family can be set to either AF_INET or AF_INET6 to force the
+        socket to use IPv4 or IPv6. If not set it will be determined
+        from host (defaults to AF_UNSPEC).
+
+        flags is a bitmask for getaddrinfo().
+
+        sock can optionally be specified in order to use a preexisting
+        socket object.
+
+        backlog is the maximum number of queued connections passed to
+        listen() (defaults to 100).
+
+        ssl can be set to an SSLContext to enable SSL over the
+        accepted connections.
+
+        reuse_address tells the kernel to reuse a local socket in
+        TIME_WAIT state, without waiting for its natural timeout to
+        expire. If not specified will automatically be set to True on
+        UNIX.
+        """
+        raise NotImplementedError
+
+    def create_datagram_endpoint(self, protocol_factory,
+                                 local_addr=None, remote_addr=None, *,
+                                 family=0, proto=0, flags=0):
+        raise NotImplementedError
+
+    def connect_read_pipe(self, protocol_factory, pipe):
+        """Register read pipe in eventloop.
+
+        protocol_factory should instantiate object with Protocol interface.
+        pipe is file-like object already switched to nonblocking.
+        Return pair (transport, protocol), where transport support
+        ReadTransport ABC"""
+        # The reason to accept file-like object instead of just file descriptor
+        # is: we need to own pipe and close it at transport finishing
+        # Can got complicated errors if pass f.fileno(),
+        # close fd in pipe transport then close f and vise versa.
+        raise NotImplementedError
+
+    def connect_write_pipe(self, protocol_factory, pipe):
+        """Register write pipe in eventloop.
+
+        protocol_factory should instantiate object with BaseProtocol interface.
+        Pipe is file-like object already switched to nonblocking.
+        Return pair (transport, protocol), where transport support
+        WriteTransport ABC"""
+        # The reason to accept file-like object instead of just file descriptor
+        # is: we need to own pipe and close it at transport finishing
+        # Can got complicated errors if pass f.fileno(),
+        # close fd in pipe transport then close f and vise versa.
+        raise NotImplementedError
+
+    def subprocess_shell(self, protocol_factory, cmd, *, stdin=subprocess.PIPE,
+                         stdout=subprocess.PIPE, stderr=subprocess.PIPE,
+                         **kwargs):
+        raise NotImplementedError
+
+    def subprocess_exec(self, protocol_factory, *args, stdin=subprocess.PIPE,
+                        stdout=subprocess.PIPE, stderr=subprocess.PIPE,
+                        **kwargs):
+        raise NotImplementedError
+
+    # Ready-based callback registration methods.
+    # The add_*() methods return None.
+    # The remove_*() methods return True if something was removed,
+    # False if there was nothing to delete.
+
+    def add_reader(self, fd, callback, *args):
+        raise NotImplementedError
+
+    def remove_reader(self, fd):
+        raise NotImplementedError
+
+    def add_writer(self, fd, callback, *args):
+        raise NotImplementedError
+
+    def remove_writer(self, fd):
+        raise NotImplementedError
+
+    # Completion based I/O methods returning Futures.
+
+    def sock_recv(self, sock, nbytes):
+        raise NotImplementedError
+
+    def sock_sendall(self, sock, data):
+        raise NotImplementedError
+
+    def sock_connect(self, sock, address):
+        raise NotImplementedError
+
+    def sock_accept(self, sock):
+        raise NotImplementedError
+
+    # Signal handling.
+
+    def add_signal_handler(self, sig, callback, *args):
+        raise NotImplementedError
+
+    def remove_signal_handler(self, sig):
+        raise NotImplementedError
+
+
+class AbstractEventLoopPolicy:
+    """Abstract policy for accessing the event loop."""
+
+    def get_event_loop(self):
+        """XXX"""
+        raise NotImplementedError
+
+    def set_event_loop(self, loop):
+        """XXX"""
+        raise NotImplementedError
+
+    def new_event_loop(self):
+        """XXX"""
+        raise NotImplementedError
+
+
+class DefaultEventLoopPolicy(threading.local, AbstractEventLoopPolicy):
+    """Default policy implementation for accessing the event loop.
+
+    In this policy, each thread has its own event loop.  However, we
+    only automatically create an event loop by default for the main
+    thread; other threads by default have no event loop.
+
+    Other policies may have different rules (e.g. a single global
+    event loop, or automatically creating an event loop per thread, or
+    using some other notion of context to which an event loop is
+    associated).
+    """
+
+    _loop = None
+    _set_called = False
+
+    def get_event_loop(self):
+        """Get the event loop.
+
+        This may be None or an instance of EventLoop.
+        """
+        if (self._loop is None and
+            not self._set_called and
+            isinstance(threading.current_thread(), threading._MainThread)):
+            self._loop = self.new_event_loop()
+        assert self._loop is not None, \
+               ('There is no current event loop in thread %r.' %
+                threading.current_thread().name)
+        return self._loop
+
+    def set_event_loop(self, loop):
+        """Set the event loop."""
+        # TODO: The isinstance() test violates the PEP.
+        self._set_called = True
+        assert loop is None or isinstance(loop, AbstractEventLoop)
+        self._loop = loop
+
+    def new_event_loop(self):
+        """Create a new event loop.
+
+        You must call set_event_loop() to make this the current event
+        loop.
+        """
+        if sys.platform == 'win32':  # pragma: no cover
+            from . import windows_events
+            return windows_events.SelectorEventLoop()
+        else:  # pragma: no cover
+            from . import unix_events
+            return unix_events.SelectorEventLoop()
+
+
+# Event loop policy.  The policy itself is always global, even if the
+# policy's rules say that there is an event loop per thread (or other
+# notion of context).  The default policy is installed by the first
+# call to get_event_loop_policy().
+_event_loop_policy = None
+
+
+def get_event_loop_policy():
+    """XXX"""
+    global _event_loop_policy
+    if _event_loop_policy is None:
+        _event_loop_policy = DefaultEventLoopPolicy()
+    return _event_loop_policy
+
+
+def set_event_loop_policy(policy):
+    """XXX"""
+    global _event_loop_policy
+    # TODO: The isinstance() test violates the PEP.
+    assert policy is None or isinstance(policy, AbstractEventLoopPolicy)
+    _event_loop_policy = policy
+
+
+def get_event_loop():
+    """XXX"""
+    return get_event_loop_policy().get_event_loop()
+
+
+def set_event_loop(loop):
+    """XXX"""
+    get_event_loop_policy().set_event_loop(loop)
+
+
+def new_event_loop():
+    """XXX"""
+    return get_event_loop_policy().new_event_loop()
diff --git a/Lib/asyncio/futures.py b/Lib/asyncio/futures.py
new file mode 100644
index 0000000..99a043b
--- /dev/null
+++ b/Lib/asyncio/futures.py
@@ -0,0 +1,338 @@
+"""A Future class similar to the one in PEP 3148."""
+
+__all__ = ['CancelledError', 'TimeoutError',
+           'InvalidStateError',
+           'Future', 'wrap_future',
+           ]
+
+import concurrent.futures._base
+import logging
+import traceback
+
+from . import events
+from .log import asyncio_log
+
+# States for Future.
+_PENDING = 'PENDING'
+_CANCELLED = 'CANCELLED'
+_FINISHED = 'FINISHED'
+
+# TODO: Do we really want to depend on concurrent.futures internals?
+Error = concurrent.futures._base.Error
+CancelledError = concurrent.futures.CancelledError
+TimeoutError = concurrent.futures.TimeoutError
+
+STACK_DEBUG = logging.DEBUG - 1  # heavy-duty debugging
+
+
+class InvalidStateError(Error):
+    """The operation is not allowed in this state."""
+    # TODO: Show the future, its state, the method, and the required state.
+
+
+class _TracebackLogger:
+    """Helper to log a traceback upon destruction if not cleared.
+
+    This solves a nasty problem with Futures and Tasks that have an
+    exception set: if nobody asks for the exception, the exception is
+    never logged.  This violates the Zen of Python: 'Errors should
+    never pass silently.  Unless explicitly silenced.'
+
+    However, we don't want to log the exception as soon as
+    set_exception() is called: if the calling code is written
+    properly, it will get the exception and handle it properly.  But
+    we *do* want to log it if result() or exception() was never called
+    -- otherwise developers waste a lot of time wondering why their
+    buggy code fails silently.
+
+    An earlier attempt added a __del__() method to the Future class
+    itself, but this backfired because the presence of __del__()
+    prevents garbage collection from breaking cycles.  A way out of
+    this catch-22 is to avoid having a __del__() method on the Future
+    class itself, but instead to have a reference to a helper object
+    with a __del__() method that logs the traceback, where we ensure
+    that the helper object doesn't participate in cycles, and only the
+    Future has a reference to it.
+
+    The helper object is added when set_exception() is called.  When
+    the Future is collected, and the helper is present, the helper
+    object is also collected, and its __del__() method will log the
+    traceback.  When the Future's result() or exception() method is
+    called (and a helper object is present), it removes the the helper
+    object, after calling its clear() method to prevent it from
+    logging.
+
+    One downside is that we do a fair amount of work to extract the
+    traceback from the exception, even when it is never logged.  It
+    would seem cheaper to just store the exception object, but that
+    references the traceback, which references stack frames, which may
+    reference the Future, which references the _TracebackLogger, and
+    then the _TracebackLogger would be included in a cycle, which is
+    what we're trying to avoid!  As an optimization, we don't
+    immediately format the exception; we only do the work when
+    activate() is called, which call is delayed until after all the
+    Future's callbacks have run.  Since usually a Future has at least
+    one callback (typically set by 'yield from') and usually that
+    callback extracts the callback, thereby removing the need to
+    format the exception.
+
+    PS. I don't claim credit for this solution.  I first heard of it
+    in a discussion about closing files when they are collected.
+    """
+
+    __slots__ = ['exc', 'tb']
+
+    def __init__(self, exc):
+        self.exc = exc
+        self.tb = None
+
+    def activate(self):
+        exc = self.exc
+        if exc is not None:
+            self.exc = None
+            self.tb = traceback.format_exception(exc.__class__, exc,
+                                                 exc.__traceback__)
+
+    def clear(self):
+        self.exc = None
+        self.tb = None
+
+    def __del__(self):
+        if self.tb:
+            asyncio_log.error('Future/Task exception was never retrieved:\n%s',
+                              ''.join(self.tb))
+
+
+class Future:
+    """This class is *almost* compatible with concurrent.futures.Future.
+
+    Differences:
+
+    - result() and exception() do not take a timeout argument and
+      raise an exception when the future isn't done yet.
+
+    - Callbacks registered with add_done_callback() are always called
+      via the event loop's call_soon_threadsafe().
+
+    - This class is not compatible with the wait() and as_completed()
+      methods in the concurrent.futures package.
+
+    (In Python 3.4 or later we may be able to unify the implementations.)
+    """
+
+    # Class variables serving as defaults for instance variables.
+    _state = _PENDING
+    _result = None
+    _exception = None
+    _loop = None
+
+    _blocking = False  # proper use of future (yield vs yield from)
+
+    _tb_logger = None
+
+    def __init__(self, *, loop=None):
+        """Initialize the future.
+
+        The optional event_loop argument allows to explicitly set the event
+        loop object used by the future. If it's not provided, the future uses
+        the default event loop.
+        """
+        if loop is None:
+            self._loop = events.get_event_loop()
+        else:
+            self._loop = loop
+        self._callbacks = []
+
+    def __repr__(self):
+        res = self.__class__.__name__
+        if self._state == _FINISHED:
+            if self._exception is not None:
+                res += '<exception={!r}>'.format(self._exception)
+            else:
+                res += '<result={!r}>'.format(self._result)
+        elif self._callbacks:
+            size = len(self._callbacks)
+            if size > 2:
+                res += '<{}, [{}, <{} more>, {}]>'.format(
+                    self._state, self._callbacks[0],
+                    size-2, self._callbacks[-1])
+            else:
+                res += '<{}, {}>'.format(self._state, self._callbacks)
+        else:
+            res += '<{}>'.format(self._state)
+        return res
+
+    def cancel(self):
+        """Cancel the future and schedule callbacks.
+
+        If the future is already done or cancelled, return False.  Otherwise,
+        change the future's state to cancelled, schedule the callbacks and
+        return True.
+        """
+        if self._state != _PENDING:
+            return False
+        self._state = _CANCELLED
+        self._schedule_callbacks()
+        return True
+
+    def _schedule_callbacks(self):
+        """Internal: Ask the event loop to call all callbacks.
+
+        The callbacks are scheduled to be called as soon as possible. Also
+        clears the callback list.
+        """
+        callbacks = self._callbacks[:]
+        if not callbacks:
+            return
+
+        self._callbacks[:] = []
+        for callback in callbacks:
+            self._loop.call_soon(callback, self)
+
+    def cancelled(self):
+        """Return True if the future was cancelled."""
+        return self._state == _CANCELLED
+
+    # Don't implement running(); see http://bugs.python.org/issue18699
+
+    def done(self):
+        """Return True if the future is done.
+
+        Done means either that a result / exception are available, or that the
+        future was cancelled.
+        """
+        return self._state != _PENDING
+
+    def result(self):
+        """Return the result this future represents.
+
+        If the future has been cancelled, raises CancelledError.  If the
+        future's result isn't yet available, raises InvalidStateError.  If
+        the future is done and has an exception set, this exception is raised.
+        """
+        if self._state == _CANCELLED:
+            raise CancelledError
+        if self._state != _FINISHED:
+            raise InvalidStateError('Result is not ready.')
+        if self._tb_logger is not None:
+            self._tb_logger.clear()
+            self._tb_logger = None
+        if self._exception is not None:
+            raise self._exception
+        return self._result
+
+    def exception(self):
+        """Return the exception that was set on this future.
+
+        The exception (or None if no exception was set) is returned only if
+        the future is done.  If the future has been cancelled, raises
+        CancelledError.  If the future isn't done yet, raises
+        InvalidStateError.
+        """
+        if self._state == _CANCELLED:
+            raise CancelledError
+        if self._state != _FINISHED:
+            raise InvalidStateError('Exception is not set.')
+        if self._tb_logger is not None:
+            self._tb_logger.clear()
+            self._tb_logger = None
+        return self._exception
+
+    def add_done_callback(self, fn):
+        """Add a callback to be run when the future becomes done.
+
+        The callback is called with a single argument - the future object. If
+        the future is already done when this is called, the callback is
+        scheduled with call_soon.
+        """
+        if self._state != _PENDING:
+            self._loop.call_soon(fn, self)
+        else:
+            self._callbacks.append(fn)
+
+    # New method not in PEP 3148.
+
+    def remove_done_callback(self, fn):
+        """Remove all instances of a callback from the "call when done" list.
+
+        Returns the number of callbacks removed.
+        """
+        filtered_callbacks = [f for f in self._callbacks if f != fn]
+        removed_count = len(self._callbacks) - len(filtered_callbacks)
+        if removed_count:
+            self._callbacks[:] = filtered_callbacks
+        return removed_count
+
+    # So-called internal methods (note: no set_running_or_notify_cancel()).
+
+    def set_result(self, result):
+        """Mark the future done and set its result.
+
+        If the future is already done when this method is called, raises
+        InvalidStateError.
+        """
+        if self._state != _PENDING:
+            raise InvalidStateError('{}: {!r}'.format(self._state, self))
+        self._result = result
+        self._state = _FINISHED
+        self._schedule_callbacks()
+
+    def set_exception(self, exception):
+        """Mark the future done and set an exception.
+
+        If the future is already done when this method is called, raises
+        InvalidStateError.
+        """
+        if self._state != _PENDING:
+            raise InvalidStateError('{}: {!r}'.format(self._state, self))
+        self._exception = exception
+        self._tb_logger = _TracebackLogger(exception)
+        self._state = _FINISHED
+        self._schedule_callbacks()
+        # Arrange for the logger to be activated after all callbacks
+        # have had a chance to call result() or exception().
+        self._loop.call_soon(self._tb_logger.activate)
+
+    # Truly internal methods.
+
+    def _copy_state(self, other):
+        """Internal helper to copy state from another Future.
+
+        The other Future may be a concurrent.futures.Future.
+        """
+        assert other.done()
+        assert not self.done()
+        if other.cancelled():
+            self.cancel()
+        else:
+            exception = other.exception()
+            if exception is not None:
+                self.set_exception(exception)
+            else:
+                result = other.result()
+                self.set_result(result)
+
+    def __iter__(self):
+        if not self.done():
+            self._blocking = True
+            yield self  # This tells Task to wait for completion.
+        assert self.done(), "yield from wasn't used with future"
+        return self.result()  # May raise too.
+
+
+def wrap_future(fut, *, loop=None):
+    """Wrap concurrent.futures.Future object."""
+    if isinstance(fut, Future):
+        return fut
+
+    assert isinstance(fut, concurrent.futures.Future), \
+        'concurrent.futures.Future is expected, got {!r}'.format(fut)
+
+    if loop is None:
+        loop = events.get_event_loop()
+
+    new_future = Future(loop=loop)
+    fut.add_done_callback(
+        lambda future: loop.call_soon_threadsafe(
+            new_future._copy_state, fut))
+    return new_future
diff --git a/Lib/asyncio/locks.py b/Lib/asyncio/locks.py
new file mode 100644
index 0000000..06edbbc
--- /dev/null
+++ b/Lib/asyncio/locks.py
@@ -0,0 +1,401 @@
+"""Synchronization primitives."""
+
+__all__ = ['Lock', 'Event', 'Condition', 'Semaphore']
+
+import collections
+
+from . import events
+from . import futures
+from . import tasks
+
+
+class Lock:
+    """Primitive lock objects.
+
+    A primitive lock is a synchronization primitive that is not owned
+    by a particular coroutine when locked.  A primitive lock is in one
+    of two states, 'locked' or 'unlocked'.
+
+    It is created in the unlocked state.  It has two basic methods,
+    acquire() and release().  When the state is unlocked, acquire()
+    changes the state to locked and returns immediately.  When the
+    state is locked, acquire() blocks until a call to release() in
+    another coroutine changes it to unlocked, then the acquire() call
+    resets it to locked and returns.  The release() method should only
+    be called in the locked state; it changes the state to unlocked
+    and returns immediately.  If an attempt is made to release an
+    unlocked lock, a RuntimeError will be raised.
+
+    When more than one coroutine is blocked in acquire() waiting for
+    the state to turn to unlocked, only one coroutine proceeds when a
+    release() call resets the state to unlocked; first coroutine which
+    is blocked in acquire() is being processed.
+
+    acquire() is a coroutine and should be called with 'yield from'.
+
+    Locks also support the context manager protocol.  '(yield from lock)'
+    should be used as context manager expression.
+
+    Usage:
+
+        lock = Lock()
+        ...
+        yield from lock
+        try:
+            ...
+        finally:
+            lock.release()
+
+    Context manager usage:
+
+        lock = Lock()
+        ...
+        with (yield from lock):
+             ...
+
+    Lock objects can be tested for locking state:
+
+        if not lock.locked():
+           yield from lock
+        else:
+           # lock is acquired
+           ...
+
+    """
+
+    def __init__(self, *, loop=None):
+        self._waiters = collections.deque()
+        self._locked = False
+        if loop is not None:
+            self._loop = loop
+        else:
+            self._loop = events.get_event_loop()
+
+    def __repr__(self):
+        res = super().__repr__()
+        extra = 'locked' if self._locked else 'unlocked'
+        if self._waiters:
+            extra = '{},waiters:{}'.format(extra, len(self._waiters))
+        return '<{} [{}]>'.format(res[1:-1], extra)
+
+    def locked(self):
+        """Return true if lock is acquired."""
+        return self._locked
+
+    @tasks.coroutine
+    def acquire(self):
+        """Acquire a lock.
+
+        This method blocks until the lock is unlocked, then sets it to
+        locked and returns True.
+        """
+        if not self._waiters and not self._locked:
+            self._locked = True
+            return True
+
+        fut = futures.Future(loop=self._loop)
+        self._waiters.append(fut)
+        try:
+            yield from fut
+            self._locked = True
+            return True
+        finally:
+            self._waiters.remove(fut)
+
+    def release(self):
+        """Release a lock.
+
+        When the lock is locked, reset it to unlocked, and return.
+        If any other coroutines are blocked waiting for the lock to become
+        unlocked, allow exactly one of them to proceed.
+
+        When invoked on an unlocked lock, a RuntimeError is raised.
+
+        There is no return value.
+        """
+        if self._locked:
+            self._locked = False
+            # Wake up the first waiter who isn't cancelled.
+            for fut in self._waiters:
+                if not fut.done():
+                    fut.set_result(True)
+                    break
+        else:
+            raise RuntimeError('Lock is not acquired.')
+
+    def __enter__(self):
+        if not self._locked:
+            raise RuntimeError(
+                '"yield from" should be used as context manager expression')
+        return True
+
+    def __exit__(self, *args):
+        self.release()
+
+    def __iter__(self):
+        yield from self.acquire()
+        return self
+
+
+class Event:
+    """An Event implementation, our equivalent to threading.Event.
+
+    Class implementing event objects. An event manages a flag that can be set
+    to true with the set() method and reset to false with the clear() method.
+    The wait() method blocks until the flag is true. The flag is initially
+    false.
+    """
+
+    def __init__(self, *, loop=None):
+        self._waiters = collections.deque()
+        self._value = False
+        if loop is not None:
+            self._loop = loop
+        else:
+            self._loop = events.get_event_loop()
+
+    def __repr__(self):
+        # TODO: add waiters:N if > 0.
+        res = super().__repr__()
+        return '<{} [{}]>'.format(res[1:-1], 'set' if self._value else 'unset')
+
+    def is_set(self):
+        """Return true if and only if the internal flag is true."""
+        return self._value
+
+    def set(self):
+        """Set the internal flag to true. All coroutines waiting for it to
+        become true are awakened. Coroutine that call wait() once the flag is
+        true will not block at all.
+        """
+        if not self._value:
+            self._value = True
+
+            for fut in self._waiters:
+                if not fut.done():
+                    fut.set_result(True)
+
+    def clear(self):
+        """Reset the internal flag to false. Subsequently, coroutines calling
+        wait() will block until set() is called to set the internal flag
+        to true again."""
+        self._value = False
+
+    @tasks.coroutine
+    def wait(self):
+        """Block until the internal flag is true.
+
+        If the internal flag is true on entry, return True
+        immediately.  Otherwise, block until another coroutine calls
+        set() to set the flag to true, then return True.
+        """
+        if self._value:
+            return True
+
+        fut = futures.Future(loop=self._loop)
+        self._waiters.append(fut)
+        try:
+            yield from fut
+            return True
+        finally:
+            self._waiters.remove(fut)
+
+
+# TODO: Why is this a Lock subclass?  threading.Condition *has* a lock.
+class Condition(Lock):
+    """A Condition implementation.
+
+    This class implements condition variable objects. A condition variable
+    allows one or more coroutines to wait until they are notified by another
+    coroutine.
+    """
+
+    def __init__(self, *, loop=None):
+        super().__init__(loop=loop)
+        self._condition_waiters = collections.deque()
+
+    # TODO: Add __repr__() with len(_condition_waiters).
+
+    @tasks.coroutine
+    def wait(self):
+        """Wait until notified.
+
+        If the calling coroutine has not acquired the lock when this
+        method is called, a RuntimeError is raised.
+
+        This method releases the underlying lock, and then blocks
+        until it is awakened by a notify() or notify_all() call for
+        the same condition variable in another coroutine.  Once
+        awakened, it re-acquires the lock and returns True.
+        """
+        if not self._locked:
+            raise RuntimeError('cannot wait on un-acquired lock')
+
+        keep_lock = True
+        self.release()
+        try:
+            fut = futures.Future(loop=self._loop)
+            self._condition_waiters.append(fut)
+            try:
+                yield from fut
+                return True
+            finally:
+                self._condition_waiters.remove(fut)
+
+        except GeneratorExit:
+            keep_lock = False  # Prevent yield in finally clause.
+            raise
+        finally:
+            if keep_lock:
+                yield from self.acquire()
+
+    @tasks.coroutine
+    def wait_for(self, predicate):
+        """Wait until a predicate becomes true.
+
+        The predicate should be a callable which result will be
+        interpreted as a boolean value.  The final predicate value is
+        the return value.
+        """
+        result = predicate()
+        while not result:
+            yield from self.wait()
+            result = predicate()
+        return result
+
+    def notify(self, n=1):
+        """By default, wake up one coroutine waiting on this condition, if any.
+        If the calling coroutine has not acquired the lock when this method
+        is called, a RuntimeError is raised.
+
+        This method wakes up at most n of the coroutines waiting for the
+        condition variable; it is a no-op if no coroutines are waiting.
+
+        Note: an awakened coroutine does not actually return from its
+        wait() call until it can reacquire the lock. Since notify() does
+        not release the lock, its caller should.
+        """
+        if not self._locked:
+            raise RuntimeError('cannot notify on un-acquired lock')
+
+        idx = 0
+        for fut in self._condition_waiters:
+            if idx >= n:
+                break
+
+            if not fut.done():
+                idx += 1
+                fut.set_result(False)
+
+    def notify_all(self):
+        """Wake up all threads waiting on this condition. This method acts
+        like notify(), but wakes up all waiting threads instead of one. If the
+        calling thread has not acquired the lock when this method is called,
+        a RuntimeError is raised.
+        """
+        self.notify(len(self._condition_waiters))
+
+
+class Semaphore:
+    """A Semaphore implementation.
+
+    A semaphore manages an internal counter which is decremented by each
+    acquire() call and incremented by each release() call. The counter
+    can never go below zero; when acquire() finds that it is zero, it blocks,
+    waiting until some other thread calls release().
+
+    Semaphores also support the context manager protocol.
+
+    The first optional argument gives the initial value for the internal
+    counter; it defaults to 1. If the value given is less than 0,
+    ValueError is raised.
+
+    The second optional argument determins can semophore be released more than
+    initial internal counter value; it defaults to False. If the value given
+    is True and number of release() is more than number of successfull
+    acquire() calls ValueError is raised.
+    """
+
+    def __init__(self, value=1, bound=False, *, loop=None):
+        if value < 0:
+            raise ValueError("Semaphore initial value must be > 0")
+        self._value = value
+        self._bound = bound
+        self._bound_value = value
+        self._waiters = collections.deque()
+        self._locked = False
+        if loop is not None:
+            self._loop = loop
+        else:
+            self._loop = events.get_event_loop()
+
+    def __repr__(self):
+        # TODO: add waiters:N if > 0.
+        res = super().__repr__()
+        return '<{} [{}]>'.format(
+            res[1:-1],
+            'locked' if self._locked else 'unlocked,value:{}'.format(
+                self._value))
+
+    def locked(self):
+        """Returns True if semaphore can not be acquired immediately."""
+        return self._locked
+
+    @tasks.coroutine
+    def acquire(self):
+        """Acquire a semaphore.
+
+        If the internal counter is larger than zero on entry,
+        decrement it by one and return True immediately.  If it is
+        zero on entry, block, waiting until some other coroutine has
+        called release() to make it larger than 0, and then return
+        True.
+        """
+        if not self._waiters and self._value > 0:
+            self._value -= 1
+            if self._value == 0:
+                self._locked = True
+            return True
+
+        fut = futures.Future(loop=self._loop)
+        self._waiters.append(fut)
+        try:
+            yield from fut
+            self._value -= 1
+            if self._value == 0:
+                self._locked = True
+            return True
+        finally:
+            self._waiters.remove(fut)
+
+    def release(self):
+        """Release a semaphore, incrementing the internal counter by one.
+        When it was zero on entry and another coroutine is waiting for it to
+        become larger than zero again, wake up that coroutine.
+
+        If Semaphore is create with "bound" paramter equals true, then
+        release() method checks to make sure its current value doesn't exceed
+        its initial value. If it does, ValueError is raised.
+        """
+        if self._bound and self._value >= self._bound_value:
+            raise ValueError('Semaphore released too many times')
+
+        self._value += 1
+        self._locked = False
+
+        for waiter in self._waiters:
+            if not waiter.done():
+                waiter.set_result(True)
+                break
+
+    def __enter__(self):
+        # TODO: This is questionable.  How do we know the user actually
+        # wrote "with (yield from sema)" instead of "with sema"?
+        return True
+
+    def __exit__(self, *args):
+        self.release()
+
+    def __iter__(self):
+        yield from self.acquire()
+        return self
diff --git a/Lib/asyncio/log.py b/Lib/asyncio/log.py
new file mode 100644
index 0000000..54dc784
--- /dev/null
+++ b/Lib/asyncio/log.py
@@ -0,0 +1,6 @@
+"""Logging configuration."""
+
+import logging
+
+
+asyncio_log = logging.getLogger("asyncio")
diff --git a/Lib/asyncio/proactor_events.py b/Lib/asyncio/proactor_events.py
new file mode 100644
index 0000000..348de03
--- /dev/null
+++ b/Lib/asyncio/proactor_events.py
@@ -0,0 +1,352 @@
+"""Event loop using a proactor and related classes.
+
+A proactor is a "notify-on-completion" multiplexer.  Currently a
+proactor is only implemented on Windows with IOCP.
+"""
+
+import socket
+
+from . import base_events
+from . import constants
+from . import futures
+from . import transports
+from .log import asyncio_log
+
+
+class _ProactorBasePipeTransport(transports.BaseTransport):
+    """Base class for pipe and socket transports."""
+
+    def __init__(self, loop, sock, protocol, waiter=None,
+                 extra=None, server=None):
+        super().__init__(extra)
+        self._set_extra(sock)
+        self._loop = loop
+        self._sock = sock
+        self._protocol = protocol
+        self._server = server
+        self._buffer = []
+        self._read_fut = None
+        self._write_fut = None
+        self._conn_lost = 0
+        self._closing = False  # Set when close() called.
+        self._eof_written = False
+        if self._server is not None:
+            self._server.attach(self)
+        self._loop.call_soon(self._protocol.connection_made, self)
+        if waiter is not None:
+            self._loop.call_soon(waiter.set_result, None)
+
+    def _set_extra(self, sock):
+        self._extra['pipe'] = sock
+
+    def close(self):
+        if self._closing:
+            return
+        self._closing = True
+        self._conn_lost += 1
+        if not self._buffer and self._write_fut is None:
+            self._loop.call_soon(self._call_connection_lost, None)
+        if self._read_fut is not None:
+            self._read_fut.cancel()
+
+    def _fatal_error(self, exc):
+        asyncio_log.exception('Fatal error for %s', self)
+        self._force_close(exc)
+
+    def _force_close(self, exc):
+        if self._closing:
+            return
+        self._closing = True
+        self._conn_lost += 1
+        if self._write_fut:
+            self._write_fut.cancel()
+        if self._read_fut:
+            self._read_fut.cancel()
+        self._write_fut = self._read_fut = None
+        self._buffer = []
+        self._loop.call_soon(self._call_connection_lost, exc)
+
+    def _call_connection_lost(self, exc):
+        try:
+            self._protocol.connection_lost(exc)
+        finally:
+            # XXX If there is a pending overlapped read on the other
+            # end then it may fail with ERROR_NETNAME_DELETED if we
+            # just close our end.  First calling shutdown() seems to
+            # cure it, but maybe using DisconnectEx() would be better.
+            if hasattr(self._sock, 'shutdown'):
+                self._sock.shutdown(socket.SHUT_RDWR)
+            self._sock.close()
+            server = self._server
+            if server is not None:
+                server.detach(self)
+                self._server = None
+
+
+class _ProactorReadPipeTransport(_ProactorBasePipeTransport,
+                                 transports.ReadTransport):
+    """Transport for read pipes."""
+
+    def __init__(self, loop, sock, protocol, waiter=None,
+                 extra=None, server=None):
+        super().__init__(loop, sock, protocol, waiter, extra, server)
+        self._read_fut = None
+        self._paused = False
+        self._loop.call_soon(self._loop_reading)
+
+    def pause(self):
+        assert not self._closing, 'Cannot pause() when closing'
+        assert not self._paused, 'Already paused'
+        self._paused = True
+
+    def resume(self):
+        assert self._paused, 'Not paused'
+        self._paused = False
+        if self._closing:
+            return
+        self._loop.call_soon(self._loop_reading, self._read_fut)
+
+    def _loop_reading(self, fut=None):
+        if self._paused:
+            return
+        data = None
+
+        try:
+            if fut is not None:
+                assert self._read_fut is fut or (self._read_fut is None and
+                                                 self._closing)
+                self._read_fut = None
+                data = fut.result()  # deliver data later in "finally" clause
+
+            if self._closing:
+                # since close() has been called we ignore any read data
+                data = None
+                return
+
+            if data == b'':
+                # we got end-of-file so no need to reschedule a new read
+                return
+
+            # reschedule a new read
+            self._read_fut = self._loop._proactor.recv(self._sock, 4096)
+        except ConnectionAbortedError as exc:
+            if not self._closing:
+                self._fatal_error(exc)
+        except ConnectionResetError as exc:
+            self._force_close(exc)
+        except OSError as exc:
+            self._fatal_error(exc)
+        except futures.CancelledError:
+            if not self._closing:
+                raise
+        else:
+            self._read_fut.add_done_callback(self._loop_reading)
+        finally:
+            if data:
+                self._protocol.data_received(data)
+            elif data is not None:
+                keep_open = self._protocol.eof_received()
+                if not keep_open:
+                    self.close()
+
+
+class _ProactorWritePipeTransport(_ProactorBasePipeTransport,
+                                  transports.WriteTransport):
+    """Transport for write pipes."""
+
+    def write(self, data):
+        assert isinstance(data, bytes), repr(data)
+        if self._eof_written:
+            raise IOError('write_eof() already called')
+
+        if not data:
+            return
+
+        if self._conn_lost:
+            if self._conn_lost >= constants.LOG_THRESHOLD_FOR_CONNLOST_WRITES:
+                asyncio_log.warning('socket.send() raised exception.')
+            self._conn_lost += 1
+            return
+        self._buffer.append(data)
+        if self._write_fut is None:
+            self._loop_writing()
+
+    def _loop_writing(self, f=None):
+        try:
+            assert f is self._write_fut
+            self._write_fut = None
+            if f:
+                f.result()
+            data = b''.join(self._buffer)
+            self._buffer = []
+            if not data:
+                if self._closing:
+                    self._loop.call_soon(self._call_connection_lost, None)
+                if self._eof_written:
+                    self._sock.shutdown(socket.SHUT_WR)
+                return
+            self._write_fut = self._loop._proactor.send(self._sock, data)
+            self._write_fut.add_done_callback(self._loop_writing)
+        except ConnectionResetError as exc:
+            self._force_close(exc)
+        except OSError as exc:
+            self._fatal_error(exc)
+
+    def can_write_eof(self):
+        return True
+
+    def write_eof(self):
+        self.close()
+
+    def abort(self):
+        self._force_close(None)
+
+
+class _ProactorDuplexPipeTransport(_ProactorReadPipeTransport,
+                                   _ProactorWritePipeTransport,
+                                   transports.Transport):
+    """Transport for duplex pipes."""
+
+    def can_write_eof(self):
+        return False
+
+    def write_eof(self):
+        raise NotImplementedError
+
+
+class _ProactorSocketTransport(_ProactorReadPipeTransport,
+                               _ProactorWritePipeTransport,
+                               transports.Transport):
+    """Transport for connected sockets."""
+
+    def _set_extra(self, sock):
+        self._extra['socket'] = sock
+        try:
+            self._extra['sockname'] = sock.getsockname()
+        except (socket.error, AttributeError):
+            pass
+        if 'peername' not in self._extra:
+            try:
+                self._extra['peername'] = sock.getpeername()
+            except (socket.error, AttributeError):
+                pass
+
+    def can_write_eof(self):
+        return True
+
+    def write_eof(self):
+        if self._closing or self._eof_written:
+            return
+        self._eof_written = True
+        if self._write_fut is None:
+            self._sock.shutdown(socket.SHUT_WR)
+
+
+class BaseProactorEventLoop(base_events.BaseEventLoop):
+
+    def __init__(self, proactor):
+        super().__init__()
+        asyncio_log.debug('Using proactor: %s', proactor.__class__.__name__)
+        self._proactor = proactor
+        self._selector = proactor   # convenient alias
+        proactor.set_loop(self)
+        self._make_self_pipe()
+
+    def _make_socket_transport(self, sock, protocol, waiter=None,
+                               extra=None, server=None):
+        return _ProactorSocketTransport(self, sock, protocol, waiter,
+                                        extra, server)
+
+    def _make_duplex_pipe_transport(self, sock, protocol, waiter=None,
+                                    extra=None):
+        return _ProactorDuplexPipeTransport(self,
+                                            sock, protocol, waiter, extra)
+
+    def _make_read_pipe_transport(self, sock, protocol, waiter=None,
+                                  extra=None):
+        return _ProactorReadPipeTransport(self, sock, protocol, waiter, extra)
+
+    def _make_write_pipe_transport(self, sock, protocol, waiter=None,
+                                   extra=None):
+        return _ProactorWritePipeTransport(self, sock, protocol, waiter, extra)
+
+    def close(self):
+        if self._proactor is not None:
+            self._close_self_pipe()
+            self._proactor.close()
+            self._proactor = None
+            self._selector = None
+
+    def sock_recv(self, sock, n):
+        return self._proactor.recv(sock, n)
+
+    def sock_sendall(self, sock, data):
+        return self._proactor.send(sock, data)
+
+    def sock_connect(self, sock, address):
+        return self._proactor.connect(sock, address)
+
+    def sock_accept(self, sock):
+        return self._proactor.accept(sock)
+
+    def _socketpair(self):
+        raise NotImplementedError
+
+    def _close_self_pipe(self):
+        self._ssock.close()
+        self._ssock = None
+        self._csock.close()
+        self._csock = None
+        self._internal_fds -= 1
+
+    def _make_self_pipe(self):
+        # A self-socket, really. :-)
+        self._ssock, self._csock = self._socketpair()
+        self._ssock.setblocking(False)
+        self._csock.setblocking(False)
+        self._internal_fds += 1
+        self.call_soon(self._loop_self_reading)
+
+    def _loop_self_reading(self, f=None):
+        try:
+            if f is not None:
+                f.result()  # may raise
+            f = self._proactor.recv(self._ssock, 4096)
+        except:
+            self.close()
+            raise
+        else:
+            f.add_done_callback(self._loop_self_reading)
+
+    def _write_to_self(self):
+        self._csock.send(b'x')
+
+    def _start_serving(self, protocol_factory, sock, ssl=None, server=None):
+        assert not ssl, 'IocpEventLoop is incompatible with SSL.'
+
+        def loop(f=None):
+            try:
+                if f is not None:
+                    conn, addr = f.result()
+                    protocol = protocol_factory()
+                    self._make_socket_transport(
+                        conn, protocol,
+                        extra={'peername': addr}, server=server)
+                f = self._proactor.accept(sock)
+            except OSError:
+                if sock.fileno() != -1:
+                    asyncio_log.exception('Accept failed')
+                    sock.close()
+            except futures.CancelledError:
+                sock.close()
+            else:
+                f.add_done_callback(loop)
+
+        self.call_soon(loop)
+
+    def _process_events(self, event_list):
+        pass    # XXX hard work currently done in poll
+
+    def _stop_serving(self, sock):
+        self._proactor._stop_serving(sock)
+        sock.close()
diff --git a/Lib/asyncio/protocols.py b/Lib/asyncio/protocols.py
new file mode 100644
index 0000000..a94abbe
--- /dev/null
+++ b/Lib/asyncio/protocols.py
@@ -0,0 +1,98 @@
+"""Abstract Protocol class."""
+
+__all__ = ['Protocol', 'DatagramProtocol']
+
+
+class BaseProtocol:
+    """ABC for base protocol class.
+
+    Usually user implements protocols that derived from BaseProtocol
+    like Protocol or ProcessProtocol.
+
+    The only case when BaseProtocol should be implemented directly is
+    write-only transport like write pipe
+    """
+
+    def connection_made(self, transport):
+        """Called when a connection is made.
+
+        The argument is the transport representing the pipe connection.
+        To receive data, wait for data_received() calls.
+        When the connection is closed, connection_lost() is called.
+        """
+
+    def connection_lost(self, exc):
+        """Called when the connection is lost or closed.
+
+        The argument is an exception object or None (the latter
+        meaning a regular EOF is received or the connection was
+        aborted or closed).
+        """
+
+
+class Protocol(BaseProtocol):
+    """ABC representing a protocol.
+
+    The user should implement this interface.  They can inherit from
+    this class but don't need to.  The implementations here do
+    nothing (they don't raise exceptions).
+
+    When the user wants to requests a transport, they pass a protocol
+    factory to a utility function (e.g., EventLoop.create_connection()).
+
+    When the connection is made successfully, connection_made() is
+    called with a suitable transport object.  Then data_received()
+    will be called 0 or more times with data (bytes) received from the
+    transport; finally, connection_lost() will be called exactly once
+    with either an exception object or None as an argument.
+
+    State machine of calls:
+
+      start -> CM [-> DR*] [-> ER?] -> CL -> end
+    """
+
+    def data_received(self, data):
+        """Called when some data is received.
+
+        The argument is a bytes object.
+        """
+
+    def eof_received(self):
+        """Called when the other end calls write_eof() or equivalent.
+
+        If this returns a false value (including None), the transport
+        will close itself.  If it returns a true value, closing the
+        transport is up to the protocol.
+        """
+
+
+class DatagramProtocol(BaseProtocol):
+    """ABC representing a datagram protocol."""
+
+    def datagram_received(self, data, addr):
+        """Called when some datagram is received."""
+
+    def connection_refused(self, exc):
+        """Connection is refused."""
+
+
+class SubprocessProtocol(BaseProtocol):
+    """ABC representing a protocol for subprocess calls."""
+
+    def pipe_data_received(self, fd, data):
+        """Called when subprocess write a data into stdout/stderr pipes.
+
+        fd is int file dascriptor.
+        data is bytes object.
+        """
+
+    def pipe_connection_lost(self, fd, exc):
+        """Called when a file descriptor associated with the child process is
+        closed.
+
+        fd is the int file descriptor that was closed.
+        """
+
+    def process_exited(self):
+        """Called when subprocess has exited.
+        """
diff --git a/Lib/asyncio/queues.py b/Lib/asyncio/queues.py
new file mode 100644
index 0000000..536de1c
--- /dev/null
+++ b/Lib/asyncio/queues.py
@@ -0,0 +1,284 @@
+"""Queues"""
+
+__all__ = ['Queue', 'PriorityQueue', 'LifoQueue', 'JoinableQueue',
+           'Full', 'Empty']
+
+import collections
+import heapq
+import queue
+
+from . import events
+from . import futures
+from . import locks
+from .tasks import coroutine
+
+
+# Re-export queue.Full and .Empty exceptions.
+Full = queue.Full
+Empty = queue.Empty
+
+
+class Queue:
+    """A queue, useful for coordinating producer and consumer coroutines.
+
+    If maxsize is less than or equal to zero, the queue size is infinite. If it
+    is an integer greater than 0, then "yield from put()" will block when the
+    queue reaches maxsize, until an item is removed by get().
+
+    Unlike the standard library Queue, you can reliably know this Queue's size
+    with qsize(), since your single-threaded Tulip application won't be
+    interrupted between calling qsize() and doing an operation on the Queue.
+    """
+
+    def __init__(self, maxsize=0, *, loop=None):
+        if loop is None:
+            self._loop = events.get_event_loop()
+        else:
+            self._loop = loop
+        self._maxsize = maxsize
+
+        # Futures.
+        self._getters = collections.deque()
+        # Pairs of (item, Future).
+        self._putters = collections.deque()
+        self._init(maxsize)
+
+    def _init(self, maxsize):
+        self._queue = collections.deque()
+
+    def _get(self):
+        return self._queue.popleft()
+
+    def _put(self, item):
+        self._queue.append(item)
+
+    def __repr__(self):
+        return '<{} at {:#x} {}>'.format(
+            type(self).__name__, id(self), self._format())
+
+    def __str__(self):
+        return '<{} {}>'.format(type(self).__name__, self._format())
+
+    def _format(self):
+        result = 'maxsize={!r}'.format(self._maxsize)
+        if getattr(self, '_queue', None):
+            result += ' _queue={!r}'.format(list(self._queue))
+        if self._getters:
+            result += ' _getters[{}]'.format(len(self._getters))
+        if self._putters:
+            result += ' _putters[{}]'.format(len(self._putters))
+        return result
+
+    def _consume_done_getters(self):
+        # Delete waiters at the head of the get() queue who've timed out.
+        while self._getters and self._getters[0].done():
+            self._getters.popleft()
+
+    def _consume_done_putters(self):
+        # Delete waiters at the head of the put() queue who've timed out.
+        while self._putters and self._putters[0][1].done():
+            self._putters.popleft()
+
+    def qsize(self):
+        """Number of items in the queue."""
+        return len(self._queue)
+
+    @property
+    def maxsize(self):
+        """Number of items allowed in the queue."""
+        return self._maxsize
+
+    def empty(self):
+        """Return True if the queue is empty, False otherwise."""
+        return not self._queue
+
+    def full(self):
+        """Return True if there are maxsize items in the queue.
+
+        Note: if the Queue was initialized with maxsize=0 (the default),
+        then full() is never True.
+        """
+        if self._maxsize <= 0:
+            return False
+        else:
+            return self.qsize() == self._maxsize
+
+    @coroutine
+    def put(self, item):
+        """Put an item into the queue.
+
+        If you yield from put(), wait until a free slot is available
+        before adding item.
+        """
+        self._consume_done_getters()
+        if self._getters:
+            assert not self._queue, (
+                'queue non-empty, why are getters waiting?')
+
+            getter = self._getters.popleft()
+
+            # Use _put and _get instead of passing item straight to getter, in
+            # case a subclass has logic that must run (e.g. JoinableQueue).
+            self._put(item)
+            getter.set_result(self._get())
+
+        elif self._maxsize > 0 and self._maxsize == self.qsize():
+            waiter = futures.Future(loop=self._loop)
+
+            self._putters.append((item, waiter))
+            yield from waiter
+
+        else:
+            self._put(item)
+
+    def put_nowait(self, item):
+        """Put an item into the queue without blocking.
+
+        If no free slot is immediately available, raise Full.
+        """
+        self._consume_done_getters()
+        if self._getters:
+            assert not self._queue, (
+                'queue non-empty, why are getters waiting?')
+
+            getter = self._getters.popleft()
+
+            # Use _put and _get instead of passing item straight to getter, in
+            # case a subclass has logic that must run (e.g. JoinableQueue).
+            self._put(item)
+            getter.set_result(self._get())
+
+        elif self._maxsize > 0 and self._maxsize == self.qsize():
+            raise Full
+        else:
+            self._put(item)
+
+    @coroutine
+    def get(self):
+        """Remove and return an item from the queue.
+
+        If you yield from get(), wait until a item is available.
+        """
+        self._consume_done_putters()
+        if self._putters:
+            assert self.full(), 'queue not full, why are putters waiting?'
+            item, putter = self._putters.popleft()
+            self._put(item)
+
+            # When a getter runs and frees up a slot so this putter can
+            # run, we need to defer the put for a tick to ensure that
+            # getters and putters alternate perfectly. See
+            # ChannelTest.test_wait.
+            self._loop.call_soon(putter.set_result, None)
+
+            return self._get()
+
+        elif self.qsize():
+            return self._get()
+        else:
+            waiter = futures.Future(loop=self._loop)
+
+            self._getters.append(waiter)
+            return (yield from waiter)
+
+    def get_nowait(self):
+        """Remove and return an item from the queue.
+
+        Return an item if one is immediately available, else raise Full.
+        """
+        self._consume_done_putters()
+        if self._putters:
+            assert self.full(), 'queue not full, why are putters waiting?'
+            item, putter = self._putters.popleft()
+            self._put(item)
+            # Wake putter on next tick.
+            putter.set_result(None)
+
+            return self._get()
+
+        elif self.qsize():
+            return self._get()
+        else:
+            raise Empty
+
+
+class PriorityQueue(Queue):
+    """A subclass of Queue; retrieves entries in priority order (lowest first).
+
+    Entries are typically tuples of the form: (priority number, data).
+    """
+
+    def _init(self, maxsize):
+        self._queue = []
+
+    def _put(self, item, heappush=heapq.heappush):
+        heappush(self._queue, item)
+
+    def _get(self, heappop=heapq.heappop):
+        return heappop(self._queue)
+
+
+class LifoQueue(Queue):
+    """A subclass of Queue that retrieves most recently added entries first."""
+
+    def _init(self, maxsize):
+        self._queue = []
+
+    def _put(self, item):
+        self._queue.append(item)
+
+    def _get(self):
+        return self._queue.pop()
+
+
+class JoinableQueue(Queue):
+    """A subclass of Queue with task_done() and join() methods."""
+
+    def __init__(self, maxsize=0, *, loop=None):
+        super().__init__(maxsize=maxsize, loop=loop)
+        self._unfinished_tasks = 0
+        self._finished = locks.Event(loop=self._loop)
+        self._finished.set()
+
+    def _format(self):
+        result = Queue._format(self)
+        if self._unfinished_tasks:
+            result += ' tasks={}'.format(self._unfinished_tasks)
+        return result
+
+    def _put(self, item):
+        super()._put(item)
+        self._unfinished_tasks += 1
+        self._finished.clear()
+
+    def task_done(self):
+        """Indicate that a formerly enqueued task is complete.
+
+        Used by queue consumers. For each get() used to fetch a task,
+        a subsequent call to task_done() tells the queue that the processing
+        on the task is complete.
+
+        If a join() is currently blocking, it will resume when all items have
+        been processed (meaning that a task_done() call was received for every
+        item that had been put() into the queue).
+
+        Raises ValueError if called more times than there were items placed in
+        the queue.
+        """
+        if self._unfinished_tasks <= 0:
+            raise ValueError('task_done() called too many times')
+        self._unfinished_tasks -= 1
+        if self._unfinished_tasks == 0:
+            self._finished.set()
+
+    @coroutine
+    def join(self):
+        """Block until all items in the queue have been gotten and processed.
+
+        The count of unfinished tasks goes up whenever an item is added to the
+        queue. The count goes down whenever a consumer thread calls task_done()
+        to indicate that the item was retrieved and all work on it is complete.
+        When the count of unfinished tasks drops to zero, join() unblocks.
+        """
+        if self._unfinished_tasks > 0:
+            yield from self._finished.wait()
diff --git a/Lib/asyncio/selector_events.py b/Lib/asyncio/selector_events.py
new file mode 100644
index 0000000..bae9a49
--- /dev/null
+++ b/Lib/asyncio/selector_events.py
@@ -0,0 +1,769 @@
+"""Event loop using a selector and related classes.
+
+A selector is a "notify-when-ready" multiplexer.  For a subclass which
+also includes support for signal handling, see the unix_events sub-module.
+"""
+
+import collections
+import socket
+try:
+    import ssl
+except ImportError:  # pragma: no cover
+    ssl = None
+
+from . import base_events
+from . import constants
+from . import events
+from . import futures
+from . import selectors
+from . import transports
+from .log import asyncio_log
+
+
+class BaseSelectorEventLoop(base_events.BaseEventLoop):
+    """Selector event loop.
+
+    See events.EventLoop for API specification.
+    """
+
+    def __init__(self, selector=None):
+        super().__init__()
+
+        if selector is None:
+            selector = selectors.DefaultSelector()
+        asyncio_log.debug('Using selector: %s', selector.__class__.__name__)
+        self._selector = selector
+        self._make_self_pipe()
+
+    def _make_socket_transport(self, sock, protocol, waiter=None, *,
+                               extra=None, server=None):
+        return _SelectorSocketTransport(self, sock, protocol, waiter,
+                                        extra, server)
+
+    def _make_ssl_transport(self, rawsock, protocol, sslcontext, waiter, *,
+                            server_side=False, server_hostname=None,
+                            extra=None, server=None):
+        return _SelectorSslTransport(
+            self, rawsock, protocol, sslcontext, waiter,
+            server_side, server_hostname, extra, server)
+
+    def _make_datagram_transport(self, sock, protocol,
+                                 address=None, extra=None):
+        return _SelectorDatagramTransport(self, sock, protocol, address, extra)
+
+    def close(self):
+        if self._selector is not None:
+            self._close_self_pipe()
+            self._selector.close()
+            self._selector = None
+
+    def _socketpair(self):
+        raise NotImplementedError
+
+    def _close_self_pipe(self):
+        self.remove_reader(self._ssock.fileno())
+        self._ssock.close()
+        self._ssock = None
+        self._csock.close()
+        self._csock = None
+        self._internal_fds -= 1
+
+    def _make_self_pipe(self):
+        # A self-socket, really. :-)
+        self._ssock, self._csock = self._socketpair()
+        self._ssock.setblocking(False)
+        self._csock.setblocking(False)
+        self._internal_fds += 1
+        self.add_reader(self._ssock.fileno(), self._read_from_self)
+
+    def _read_from_self(self):
+        try:
+            self._ssock.recv(1)
+        except (BlockingIOError, InterruptedError):
+            pass
+
+    def _write_to_self(self):
+        try:
+            self._csock.send(b'x')
+        except (BlockingIOError, InterruptedError):
+            pass
+
+    def _start_serving(self, protocol_factory, sock, ssl=None, server=None):
+        self.add_reader(sock.fileno(), self._accept_connection,
+                        protocol_factory, sock, ssl, server)
+
+    def _accept_connection(self, protocol_factory, sock, ssl=None,
+                           server=None):
+        try:
+            conn, addr = sock.accept()
+            conn.setblocking(False)
+        except (BlockingIOError, InterruptedError):
+            pass  # False alarm.
+        except Exception:
+            # Bad error. Stop serving.
+            self.remove_reader(sock.fileno())
+            sock.close()
+            # There's nowhere to send the error, so just log it.
+            # TODO: Someone will want an error handler for this.
+            asyncio_log.exception('Accept failed')
+        else:
+            if ssl:
+                self._make_ssl_transport(
+                    conn, protocol_factory(), ssl, None,
+                    server_side=True, extra={'peername': addr}, server=server)
+            else:
+                self._make_socket_transport(
+                    conn, protocol_factory(), extra={'peername': addr},
+                    server=server)
+        # It's now up to the protocol to handle the connection.
+
+    def add_reader(self, fd, callback, *args):
+        """Add a reader callback."""
+        handle = events.make_handle(callback, args)
+        try:
+            key = self._selector.get_key(fd)
+        except KeyError:
+            self._selector.register(fd, selectors.EVENT_READ,
+                                    (handle, None))
+        else:
+            mask, (reader, writer) = key.events, key.data
+            self._selector.modify(fd, mask | selectors.EVENT_READ,
+                                  (handle, writer))
+            if reader is not None:
+                reader.cancel()
+
+    def remove_reader(self, fd):
+        """Remove a reader callback."""
+        try:
+            key = self._selector.get_key(fd)
+        except KeyError:
+            return False
+        else:
+            mask, (reader, writer) = key.events, key.data
+            mask &= ~selectors.EVENT_READ
+            if not mask:
+                self._selector.unregister(fd)
+            else:
+                self._selector.modify(fd, mask, (None, writer))
+
+            if reader is not None:
+                reader.cancel()
+                return True
+            else:
+                return False
+
+    def add_writer(self, fd, callback, *args):
+        """Add a writer callback.."""
+        handle = events.make_handle(callback, args)
+        try:
+            key = self._selector.get_key(fd)
+        except KeyError:
+            self._selector.register(fd, selectors.EVENT_WRITE,
+                                    (None, handle))
+        else:
+            mask, (reader, writer) = key.events, key.data
+            self._selector.modify(fd, mask | selectors.EVENT_WRITE,
+                                  (reader, handle))
+            if writer is not None:
+                writer.cancel()
+
+    def remove_writer(self, fd):
+        """Remove a writer callback."""
+        try:
+            key = self._selector.get_key(fd)
+        except KeyError:
+            return False
+        else:
+            mask, (reader, writer) = key.events, key.data
+            # Remove both writer and connector.
+            mask &= ~selectors.EVENT_WRITE
+            if not mask:
+                self._selector.unregister(fd)
+            else:
+                self._selector.modify(fd, mask, (reader, None))
+
+            if writer is not None:
+                writer.cancel()
+                return True
+            else:
+                return False
+
+    def sock_recv(self, sock, n):
+        """XXX"""
+        fut = futures.Future(loop=self)
+        self._sock_recv(fut, False, sock, n)
+        return fut
+
+    def _sock_recv(self, fut, registered, sock, n):
+        fd = sock.fileno()
+        if registered:
+            # Remove the callback early.  It should be rare that the
+            # selector says the fd is ready but the call still returns
+            # EAGAIN, and I am willing to take a hit in that case in
+            # order to simplify the common case.
+            self.remove_reader(fd)
+        if fut.cancelled():
+            return
+        try:
+            data = sock.recv(n)
+        except (BlockingIOError, InterruptedError):
+            self.add_reader(fd, self._sock_recv, fut, True, sock, n)
+        except Exception as exc:
+            fut.set_exception(exc)
+        else:
+            fut.set_result(data)
+
+    def sock_sendall(self, sock, data):
+        """XXX"""
+        fut = futures.Future(loop=self)
+        if data:
+            self._sock_sendall(fut, False, sock, data)
+        else:
+            fut.set_result(None)
+        return fut
+
+    def _sock_sendall(self, fut, registered, sock, data):
+        fd = sock.fileno()
+
+        if registered:
+            self.remove_writer(fd)
+        if fut.cancelled():
+            return
+
+        try:
+            n = sock.send(data)
+        except (BlockingIOError, InterruptedError):
+            n = 0
+        except Exception as exc:
+            fut.set_exception(exc)
+            return
+
+        if n == len(data):
+            fut.set_result(None)
+        else:
+            if n:
+                data = data[n:]
+            self.add_writer(fd, self._sock_sendall, fut, True, sock, data)
+
+    def sock_connect(self, sock, address):
+        """XXX"""
+        # That address better not require a lookup!  We're not calling
+        # self.getaddrinfo() for you here.  But verifying this is
+        # complicated; the socket module doesn't have a pattern for
+        # IPv6 addresses (there are too many forms, apparently).
+        fut = futures.Future(loop=self)
+        self._sock_connect(fut, False, sock, address)
+        return fut
+
+    def _sock_connect(self, fut, registered, sock, address):
+        # TODO: Use getaddrinfo() to look up the address, to avoid the
+        # trap of hanging the entire event loop when the address
+        # requires doing a DNS lookup.  (OTOH, the caller should
+        # already have done this, so it would be nice if we could
+        # easily tell whether the address needs looking up or not.  I
+        # know how to do this for IPv4, but IPv6 addresses have many
+        # syntaxes.)
+        fd = sock.fileno()
+        if registered:
+            self.remove_writer(fd)
+        if fut.cancelled():
+            return
+        try:
+            if not registered:
+                # First time around.
+                sock.connect(address)
+            else:
+                err = sock.getsockopt(socket.SOL_SOCKET, socket.SO_ERROR)
+                if err != 0:
+                    # Jump to the except clause below.
+                    raise OSError(err, 'Connect call failed')
+        except (BlockingIOError, InterruptedError):
+            self.add_writer(fd, self._sock_connect, fut, True, sock, address)
+        except Exception as exc:
+            fut.set_exception(exc)
+        else:
+            fut.set_result(None)
+
+    def sock_accept(self, sock):
+        """XXX"""
+        fut = futures.Future(loop=self)
+        self._sock_accept(fut, False, sock)
+        return fut
+
+    def _sock_accept(self, fut, registered, sock):
+        fd = sock.fileno()
+        if registered:
+            self.remove_reader(fd)
+        if fut.cancelled():
+            return
+        try:
+            conn, address = sock.accept()
+            conn.setblocking(False)
+        except (BlockingIOError, InterruptedError):
+            self.add_reader(fd, self._sock_accept, fut, True, sock)
+        except Exception as exc:
+            fut.set_exception(exc)
+        else:
+            fut.set_result((conn, address))
+
+    def _process_events(self, event_list):
+        for key, mask in event_list:
+            fileobj, (reader, writer) = key.fileobj, key.data
+            if mask & selectors.EVENT_READ and reader is not None:
+                if reader._cancelled:
+                    self.remove_reader(fileobj)
+                else:
+                    self._add_callback(reader)
+            if mask & selectors.EVENT_WRITE and writer is not None:
+                if writer._cancelled:
+                    self.remove_writer(fileobj)
+                else:
+                    self._add_callback(writer)
+
+    def _stop_serving(self, sock):
+        self.remove_reader(sock.fileno())
+        sock.close()
+
+
+class _SelectorTransport(transports.Transport):
+
+    max_size = 256 * 1024  # Buffer size passed to recv().
+
+    def __init__(self, loop, sock, protocol, extra, server=None):
+        super().__init__(extra)
+        self._extra['socket'] = sock
+        self._extra['sockname'] = sock.getsockname()
+        if 'peername' not in self._extra:
+            try:
+                self._extra['peername'] = sock.getpeername()
+            except socket.error:
+                self._extra['peername'] = None
+        self._loop = loop
+        self._sock = sock
+        self._sock_fd = sock.fileno()
+        self._protocol = protocol
+        self._server = server
+        self._buffer = collections.deque()
+        self._conn_lost = 0
+        self._closing = False  # Set when close() called.
+        if server is not None:
+            server.attach(self)
+
+    def abort(self):
+        self._force_close(None)
+
+    def close(self):
+        if self._closing:
+            return
+        self._closing = True
+        self._conn_lost += 1
+        self._loop.remove_reader(self._sock_fd)
+        if not self._buffer:
+            self._loop.call_soon(self._call_connection_lost, None)
+
+    def _fatal_error(self, exc):
+        # should be called from exception handler only
+        asyncio_log.exception('Fatal error for %s', self)
+        self._force_close(exc)
+
+    def _force_close(self, exc):
+        if self._buffer:
+            self._buffer.clear()
+            self._loop.remove_writer(self._sock_fd)
+
+        if self._closing:
+            return
+
+        self._closing = True
+        self._conn_lost += 1
+        self._loop.remove_reader(self._sock_fd)
+        self._loop.call_soon(self._call_connection_lost, exc)
+
+    def _call_connection_lost(self, exc):
+        try:
+            self._protocol.connection_lost(exc)
+        finally:
+            self._sock.close()
+            self._sock = None
+            self._protocol = None
+            self._loop = None
+            server = self._server
+            if server is not None:
+                server.detach(self)
+                self._server = None
+
+
+class _SelectorSocketTransport(_SelectorTransport):
+
+    def __init__(self, loop, sock, protocol, waiter=None,
+                 extra=None, server=None):
+        super().__init__(loop, sock, protocol, extra, server)
+        self._eof = False
+        self._paused = False
+
+        self._loop.add_reader(self._sock_fd, self._read_ready)
+        self._loop.call_soon(self._protocol.connection_made, self)
+        if waiter is not None:
+            self._loop.call_soon(waiter.set_result, None)
+
+    def pause(self):
+        assert not self._closing, 'Cannot pause() when closing'
+        assert not self._paused, 'Already paused'
+        self._paused = True
+        self._loop.remove_reader(self._sock_fd)
+
+    def resume(self):
+        assert self._paused, 'Not paused'
+        self._paused = False
+        if self._closing:
+            return
+        self._loop.add_reader(self._sock_fd, self._read_ready)
+
+    def _read_ready(self):
+        try:
+            data = self._sock.recv(self.max_size)
+        except (BlockingIOError, InterruptedError):
+            pass
+        except ConnectionResetError as exc:
+            self._force_close(exc)
+        except Exception as exc:
+            self._fatal_error(exc)
+        else:
+            if data:
+                self._protocol.data_received(data)
+            else:
+                keep_open = self._protocol.eof_received()
+                if not keep_open:
+                    self.close()
+
+    def write(self, data):
+        assert isinstance(data, bytes), repr(type(data))
+        assert not self._eof, 'Cannot call write() after write_eof()'
+        if not data:
+            return
+
+        if self._conn_lost:
+            if self._conn_lost >= constants.LOG_THRESHOLD_FOR_CONNLOST_WRITES:
+                asyncio_log.warning('socket.send() raised exception.')
+            self._conn_lost += 1
+            return
+
+        if not self._buffer:
+            # Attempt to send it right away first.
+            try:
+                n = self._sock.send(data)
+            except (BlockingIOError, InterruptedError):
+                n = 0
+            except (BrokenPipeError, ConnectionResetError) as exc:
+                self._force_close(exc)
+                return
+            except OSError as exc:
+                self._fatal_error(exc)
+                return
+            else:
+                data = data[n:]
+                if not data:
+                    return
+            # Start async I/O.
+            self._loop.add_writer(self._sock_fd, self._write_ready)
+
+        self._buffer.append(data)
+
+    def _write_ready(self):
+        data = b''.join(self._buffer)
+        assert data, 'Data should not be empty'
+
+        self._buffer.clear()
+        try:
+            n = self._sock.send(data)
+        except (BlockingIOError, InterruptedError):
+            self._buffer.append(data)
+        except (BrokenPipeError, ConnectionResetError) as exc:
+            self._loop.remove_writer(self._sock_fd)
+            self._force_close(exc)
+        except Exception as exc:
+            self._loop.remove_writer(self._sock_fd)
+            self._fatal_error(exc)
+        else:
+            data = data[n:]
+            if not data:
+                self._loop.remove_writer(self._sock_fd)
+                if self._closing:
+                    self._call_connection_lost(None)
+                elif self._eof:
+                    self._sock.shutdown(socket.SHUT_WR)
+                return
+
+            self._buffer.append(data)  # Try again later.
+
+    def write_eof(self):
+        if self._eof:
+            return
+        self._eof = True
+        if not self._buffer:
+            self._sock.shutdown(socket.SHUT_WR)
+
+    def can_write_eof(self):
+        return True
+
+
+class _SelectorSslTransport(_SelectorTransport):
+
+    def __init__(self, loop, rawsock, protocol, sslcontext, waiter=None,
+                 server_side=False, server_hostname=None,
+                 extra=None, server=None):
+        if server_side:
+            assert isinstance(
+                sslcontext, ssl.SSLContext), 'Must pass an SSLContext'
+        else:
+            # Client-side may pass ssl=True to use a default context.
+            # The default is the same as used by urllib.
+            if sslcontext is None:
+                sslcontext = ssl.SSLContext(ssl.PROTOCOL_SSLv23)
+                sslcontext.options |= ssl.OP_NO_SSLv2
+                sslcontext.set_default_verify_paths()
+                sslcontext.verify_mode = ssl.CERT_REQUIRED
+        wrap_kwargs = {
+            'server_side': server_side,
+            'do_handshake_on_connect': False,
+        }
+        if server_hostname is not None and not server_side and ssl.HAS_SNI:
+            wrap_kwargs['server_hostname'] = server_hostname
+        sslsock = sslcontext.wrap_socket(rawsock, **wrap_kwargs)
+
+        super().__init__(loop, sslsock, protocol, extra, server)
+
+        self._server_hostname = server_hostname
+        self._waiter = waiter
+        self._rawsock = rawsock
+        self._sslcontext = sslcontext
+        self._paused = False
+
+        # SSL-specific extra info.  (peercert is set later)
+        self._extra.update(sslcontext=sslcontext)
+
+        self._on_handshake()
+
+    def _on_handshake(self):
+        try:
+            self._sock.do_handshake()
+        except ssl.SSLWantReadError:
+            self._loop.add_reader(self._sock_fd, self._on_handshake)
+            return
+        except ssl.SSLWantWriteError:
+            self._loop.add_writer(self._sock_fd, self._on_handshake)
+            return
+        except Exception as exc:
+            self._sock.close()
+            if self._waiter is not None:
+                self._waiter.set_exception(exc)
+            return
+        except BaseException as exc:
+            self._sock.close()
+            if self._waiter is not None:
+                self._waiter.set_exception(exc)
+            raise
+
+        # Verify hostname if requested.
+        peercert = self._sock.getpeercert()
+        if (self._server_hostname is not None and
+            self._sslcontext.verify_mode == ssl.CERT_REQUIRED):
+            try:
+                ssl.match_hostname(peercert, self._server_hostname)
+            except Exception as exc:
+                self._sock.close()
+                if self._waiter is not None:
+                    self._waiter.set_exception(exc)
+                return
+
+        # Add extra info that becomes available after handshake.
+        self._extra.update(peercert=peercert,
+                           cipher=self._sock.cipher(),
+                           compression=self._sock.compression(),
+                           )
+
+        self._loop.remove_reader(self._sock_fd)
+        self._loop.remove_writer(self._sock_fd)
+        self._loop.add_reader(self._sock_fd, self._on_ready)
+        self._loop.add_writer(self._sock_fd, self._on_ready)
+        self._loop.call_soon(self._protocol.connection_made, self)
+        if self._waiter is not None:
+            self._loop.call_soon(self._waiter.set_result, None)
+
+    def pause(self):
+        # XXX This is a bit icky, given the comment at the top of
+        # _on_ready().  Is it possible to evoke a deadlock?  I don't
+        # know, although it doesn't look like it; write() will still
+        # accept more data for the buffer and eventually the app will
+        # call resume() again, and things will flow again.
+
+        assert not self._closing, 'Cannot pause() when closing'
+        assert not self._paused, 'Already paused'
+        self._paused = True
+        self._loop.remove_reader(self._sock_fd)
+
+    def resume(self):
+        assert self._paused, 'Not paused'
+        self._paused = False
+        if self._closing:
+            return
+        self._loop.add_reader(self._sock_fd, self._on_ready)
+
+    def _on_ready(self):
+        # Because of renegotiations (?), there's no difference between
+        # readable and writable.  We just try both.  XXX This may be
+        # incorrect; we probably need to keep state about what we
+        # should do next.
+
+        # First try reading.
+        if not self._closing and not self._paused:
+            try:
+                data = self._sock.recv(self.max_size)
+            except (BlockingIOError, InterruptedError,
+                    ssl.SSLWantReadError, ssl.SSLWantWriteError):
+                pass
+            except ConnectionResetError as exc:
+                self._force_close(exc)
+            except Exception as exc:
+                self._fatal_error(exc)
+            else:
+                if data:
+                    self._protocol.data_received(data)
+                else:
+                    try:
+                        self._protocol.eof_received()
+                    finally:
+                        self.close()
+
+        # Now try writing, if there's anything to write.
+        if self._buffer:
+            data = b''.join(self._buffer)
+            self._buffer.clear()
+            try:
+                n = self._sock.send(data)
+            except (BlockingIOError, InterruptedError,
+                    ssl.SSLWantReadError, ssl.SSLWantWriteError):
+                n = 0
+            except (BrokenPipeError, ConnectionResetError) as exc:
+                self._loop.remove_writer(self._sock_fd)
+                self._force_close(exc)
+                return
+            except Exception as exc:
+                self._loop.remove_writer(self._sock_fd)
+                self._fatal_error(exc)
+                return
+
+            if n < len(data):
+                self._buffer.append(data[n:])
+
+        if self._closing and not self._buffer:
+            self._loop.remove_writer(self._sock_fd)
+            self._call_connection_lost(None)
+
+    def write(self, data):
+        assert isinstance(data, bytes), repr(type(data))
+        if not data:
+            return
+
+        if self._conn_lost:
+            if self._conn_lost >= constants.LOG_THRESHOLD_FOR_CONNLOST_WRITES:
+                asyncio_log.warning('socket.send() raised exception.')
+            self._conn_lost += 1
+            return
+
+        self._buffer.append(data)
+        # We could optimize, but the callback can do this for now.
+
+    def can_write_eof(self):
+        return False
+
+    def close(self):
+        if self._closing:
+            return
+        self._closing = True
+        self._conn_lost += 1
+        self._loop.remove_reader(self._sock_fd)
+
+
+class _SelectorDatagramTransport(_SelectorTransport):
+
+    def __init__(self, loop, sock, protocol, address=None, extra=None):
+        super().__init__(loop, sock, protocol, extra)
+
+        self._address = address
+        self._loop.add_reader(self._sock_fd, self._read_ready)
+        self._loop.call_soon(self._protocol.connection_made, self)
+
+    def _read_ready(self):
+        try:
+            data, addr = self._sock.recvfrom(self.max_size)
+        except (BlockingIOError, InterruptedError):
+            pass
+        except Exception as exc:
+            self._fatal_error(exc)
+        else:
+            self._protocol.datagram_received(data, addr)
+
+    def sendto(self, data, addr=None):
+        assert isinstance(data, bytes), repr(type(data))
+        if not data:
+            return
+
+        if self._address:
+            assert addr in (None, self._address)
+
+        if self._conn_lost and self._address:
+            if self._conn_lost >= constants.LOG_THRESHOLD_FOR_CONNLOST_WRITES:
+                asyncio_log.warning('socket.send() raised exception.')
+            self._conn_lost += 1
+            return
+
+        if not self._buffer:
+            # Attempt to send it right away first.
+            try:
+                if self._address:
+                    self._sock.send(data)
+                else:
+                    self._sock.sendto(data, addr)
+                return
+            except ConnectionRefusedError as exc:
+                if self._address:
+                    self._fatal_error(exc)
+                return
+            except (BlockingIOError, InterruptedError):
+                self._loop.add_writer(self._sock_fd, self._sendto_ready)
+            except Exception as exc:
+                self._fatal_error(exc)
+                return
+
+        self._buffer.append((data, addr))
+
+    def _sendto_ready(self):
+        while self._buffer:
+            data, addr = self._buffer.popleft()
+            try:
+                if self._address:
+                    self._sock.send(data)
+                else:
+                    self._sock.sendto(data, addr)
+            except ConnectionRefusedError as exc:
+                if self._address:
+                    self._fatal_error(exc)
+                return
+            except (BlockingIOError, InterruptedError):
+                self._buffer.appendleft((data, addr))  # Try again later.
+                break
+            except Exception as exc:
+                self._fatal_error(exc)
+                return
+
+        if not self._buffer:
+            self._loop.remove_writer(self._sock_fd)
+            if self._closing:
+                self._call_connection_lost(None)
+
+    def _force_close(self, exc):
+        if self._address and isinstance(exc, ConnectionRefusedError):
+            self._protocol.connection_refused(exc)
+
+        super()._force_close(exc)
diff --git a/Lib/asyncio/streams.py b/Lib/asyncio/streams.py
new file mode 100644
index 0000000..d0f12e8
--- /dev/null
+++ b/Lib/asyncio/streams.py
@@ -0,0 +1,257 @@
+"""Stream-related things."""
+
+__all__ = ['StreamReader', 'StreamReaderProtocol', 'open_connection']
+
+import collections
+
+from . import events
+from . import futures
+from . import protocols
+from . import tasks
+
+
+_DEFAULT_LIMIT = 2**16
+
+
+@tasks.coroutine
+def open_connection(host=None, port=None, *,
+                    loop=None, limit=_DEFAULT_LIMIT, **kwds):
+    """A wrapper for create_connection() returning a (reader, writer) pair.
+
+    The reader returned is a StreamReader instance; the writer is a
+    Transport.
+
+    The arguments are all the usual arguments to create_connection()
+    except protocol_factory; most common are positional host and port,
+    with various optional keyword arguments following.
+
+    Additional optional keyword arguments are loop (to set the event loop
+    instance to use) and limit (to set the buffer limit passed to the
+    StreamReader).
+
+    (If you want to customize the StreamReader and/or
+    StreamReaderProtocol classes, just copy the code -- there's
+    really nothing special here except some convenience.)
+    """
+    if loop is None:
+        loop = events.get_event_loop()
+    reader = StreamReader(limit=limit, loop=loop)
+    protocol = StreamReaderProtocol(reader)
+    transport, _ = yield from loop.create_connection(
+        lambda: protocol, host, port, **kwds)
+    return reader, transport  # (reader, writer)
+
+
+class StreamReaderProtocol(protocols.Protocol):
+    """Trivial helper class to adapt between Protocol and StreamReader.
+
+    (This is a helper class instead of making StreamReader itself a
+    Protocol subclass, because the StreamReader has other potential
+    uses, and to prevent the user of the StreamReader to accidentally
+    call inappropriate methods of the protocol.)
+    """
+
+    def __init__(self, stream_reader):
+        self.stream_reader = stream_reader
+
+    def connection_made(self, transport):
+        self.stream_reader.set_transport(transport)
+
+    def connection_lost(self, exc):
+        if exc is None:
+            self.stream_reader.feed_eof()
+        else:
+            self.stream_reader.set_exception(exc)
+
+    def data_received(self, data):
+        self.stream_reader.feed_data(data)
+
+    def eof_received(self):
+        self.stream_reader.feed_eof()
+
+
+class StreamReader:
+
+    def __init__(self, limit=_DEFAULT_LIMIT, loop=None):
+        # The line length limit is  a security feature;
+        # it also doubles as half the buffer limit.
+        self.limit = limit
+        if loop is None:
+            loop = events.get_event_loop()
+        self.loop = loop
+        self.buffer = collections.deque()  # Deque of bytes objects.
+        self.byte_count = 0  # Bytes in buffer.
+        self.eof = False  # Whether we're done.
+        self.waiter = None  # A future.
+        self._exception = None
+        self._transport = None
+        self._paused = False
+
+    def exception(self):
+        return self._exception
+
+    def set_exception(self, exc):
+        self._exception = exc
+
+        waiter = self.waiter
+        if waiter is not None:
+            self.waiter = None
+            if not waiter.cancelled():
+                waiter.set_exception(exc)
+
+    def set_transport(self, transport):
+        assert self._transport is None, 'Transport already set'
+        self._transport = transport
+
+    def _maybe_resume_transport(self):
+        if self._paused and self.byte_count <= self.limit:
+            self._paused = False
+            self._transport.resume()
+
+    def feed_eof(self):
+        self.eof = True
+        waiter = self.waiter
+        if waiter is not None:
+            self.waiter = None
+            if not waiter.cancelled():
+                waiter.set_result(True)
+
+    def feed_data(self, data):
+        if not data:
+            return
+
+        self.buffer.append(data)
+        self.byte_count += len(data)
+
+        waiter = self.waiter
+        if waiter is not None:
+            self.waiter = None
+            if not waiter.cancelled():
+                waiter.set_result(False)
+
+        if (self._transport is not None and
+            not self._paused and
+            self.byte_count > 2*self.limit):
+            try:
+                self._transport.pause()
+            except NotImplementedError:
+                # The transport can't be paused.
+                # We'll just have to buffer all data.
+                # Forget the transport so we don't keep trying.
+                self._transport = None
+            else:
+                self._paused = True
+
+    @tasks.coroutine
+    def readline(self):
+        if self._exception is not None:
+            raise self._exception
+
+        parts = []
+        parts_size = 0
+        not_enough = True
+
+        while not_enough:
+            while self.buffer and not_enough:
+                data = self.buffer.popleft()
+                ichar = data.find(b'\n')
+                if ichar < 0:
+                    parts.append(data)
+                    parts_size += len(data)
+                else:
+                    ichar += 1
+                    head, tail = data[:ichar], data[ichar:]
+                    if tail:
+                        self.buffer.appendleft(tail)
+                    not_enough = False
+                    parts.append(head)
+                    parts_size += len(head)
+
+                if parts_size > self.limit:
+                    self.byte_count -= parts_size
+                    self._maybe_resume_transport()
+                    raise ValueError('Line is too long')
+
+            if self.eof:
+                break
+
+            if not_enough:
+                assert self.waiter is None
+                self.waiter = futures.Future(loop=self.loop)
+                try:
+                    yield from self.waiter
+                finally:
+                    self.waiter = None
+
+        line = b''.join(parts)
+        self.byte_count -= parts_size
+        self._maybe_resume_transport()
+
+        return line
+
+    @tasks.coroutine
+    def read(self, n=-1):
+        if self._exception is not None:
+            raise self._exception
+
+        if not n:
+            return b''
+
+        if n < 0:
+            while not self.eof:
+                assert not self.waiter
+                self.waiter = futures.Future(loop=self.loop)
+                try:
+                    yield from self.waiter
+                finally:
+                    self.waiter = None
+        else:
+            if not self.byte_count and not self.eof:
+                assert not self.waiter
+                self.waiter = futures.Future(loop=self.loop)
+                try:
+                    yield from self.waiter
+                finally:
+                    self.waiter = None
+
+        if n < 0 or self.byte_count <= n:
+            data = b''.join(self.buffer)
+            self.buffer.clear()
+            self.byte_count = 0
+            self._maybe_resume_transport()
+            return data
+
+        parts = []
+        parts_bytes = 0
+        while self.buffer and parts_bytes < n:
+            data = self.buffer.popleft()
+            data_bytes = len(data)
+            if n < parts_bytes + data_bytes:
+                data_bytes = n - parts_bytes
+                data, rest = data[:data_bytes], data[data_bytes:]
+                self.buffer.appendleft(rest)
+
+            parts.append(data)
+            parts_bytes += data_bytes
+            self.byte_count -= data_bytes
+            self._maybe_resume_transport()
+
+        return b''.join(parts)
+
+    @tasks.coroutine
+    def readexactly(self, n):
+        if self._exception is not None:
+            raise self._exception
+
+        if n <= 0:
+            return b''
+
+        while self.byte_count < n and not self.eof:
+            assert not self.waiter
+            self.waiter = futures.Future(loop=self.loop)
+            try:
+                yield from self.waiter
+            finally:
+                self.waiter = None
+
+        return (yield from self.read(n))
diff --git a/Lib/asyncio/tasks.py b/Lib/asyncio/tasks.py
new file mode 100644
index 0000000..2c8579f
--- /dev/null
+++ b/Lib/asyncio/tasks.py
@@ -0,0 +1,636 @@
+"""Support for tasks, coroutines and the scheduler."""
+
+__all__ = ['coroutine', 'Task',
+           'FIRST_COMPLETED', 'FIRST_EXCEPTION', 'ALL_COMPLETED',
+           'wait', 'wait_for', 'as_completed', 'sleep', 'async',
+           'gather',
+           ]
+
+import collections
+import concurrent.futures
+import functools
+import inspect
+import linecache
+import traceback
+import weakref
+
+from . import events
+from . import futures
+from .log import asyncio_log
+
+# If you set _DEBUG to true, @coroutine will wrap the resulting
+# generator objects in a CoroWrapper instance (defined below).  That
+# instance will log a message when the generator is never iterated
+# over, which may happen when you forget to use "yield from" with a
+# coroutine call.  Note that the value of the _DEBUG flag is taken
+# when the decorator is used, so to be of any use it must be set
+# before you define your coroutines.  A downside of using this feature
+# is that tracebacks show entries for the CoroWrapper.__next__ method
+# when _DEBUG is true.
+_DEBUG = False
+
+
+class CoroWrapper:
+    """Wrapper for coroutine in _DEBUG mode."""
+
+    __slot__ = ['gen', 'func']
+
+    def __init__(self, gen, func):
+        assert inspect.isgenerator(gen), gen
+        self.gen = gen
+        self.func = func
+
+    def __iter__(self):
+        return self
+
+    def __next__(self):
+        return next(self.gen)
+
+    def send(self, value):
+        return self.gen.send(value)
+
+    def throw(self, exc):
+        return self.gen.throw(exc)
+
+    def close(self):
+        return self.gen.close()
+
+    def __del__(self):
+        frame = self.gen.gi_frame
+        if frame is not None and frame.f_lasti == -1:
+            func = self.func
+            code = func.__code__
+            filename = code.co_filename
+            lineno = code.co_firstlineno
+            asyncio_log.error('Coroutine %r defined at %s:%s was never yielded from',
+                              func.__name__, filename, lineno)
+
+
+def coroutine(func):
+    """Decorator to mark coroutines.
+
+    If the coroutine is not yielded from before it is destroyed,
+    an error message is logged.
+    """
+    if inspect.isgeneratorfunction(func):
+        coro = func
+    else:
+        @functools.wraps(func)
+        def coro(*args, **kw):
+            res = func(*args, **kw)
+            if isinstance(res, futures.Future) or inspect.isgenerator(res):
+                res = yield from res
+            return res
+
+    if not _DEBUG:
+        wrapper = coro
+    else:
+        @functools.wraps(func)
+        def wrapper(*args, **kwds):
+            w = CoroWrapper(coro(*args, **kwds), func)
+            w.__name__ = coro.__name__
+            w.__doc__ = coro.__doc__
+            return w
+
+    wrapper._is_coroutine = True  # For iscoroutinefunction().
+    return wrapper
+
+
+def iscoroutinefunction(func):
+    """Return True if func is a decorated coroutine function."""
+    return getattr(func, '_is_coroutine', False)
+
+
+def iscoroutine(obj):
+    """Return True if obj is a coroutine object."""
+    return isinstance(obj, CoroWrapper) or inspect.isgenerator(obj)
+
+
+class Task(futures.Future):
+    """A coroutine wrapped in a Future."""
+
+    # An important invariant maintained while a Task not done:
+    #
+    # - Either _fut_waiter is None, and _step() is scheduled;
+    # - or _fut_waiter is some Future, and _step() is *not* scheduled.
+    #
+    # The only transition from the latter to the former is through
+    # _wakeup().  When _fut_waiter is not None, one of its callbacks
+    # must be _wakeup().
+
+    # Weak set containing all tasks alive.
+    _all_tasks = weakref.WeakSet()
+
+    @classmethod
+    def all_tasks(cls, loop=None):
+        """Return a set of all tasks for an event loop.
+
+        By default all tasks for the current event loop are returned.
+        """
+        if loop is None:
+            loop = events.get_event_loop()
+        return {t for t in cls._all_tasks if t._loop is loop}
+
+    def __init__(self, coro, *, loop=None):
+        assert iscoroutine(coro), repr(coro)  # Not a coroutine function!
+        super().__init__(loop=loop)
+        self._coro = iter(coro)  # Use the iterator just in case.
+        self._fut_waiter = None
+        self._must_cancel = False
+        self._loop.call_soon(self._step)
+        self.__class__._all_tasks.add(self)
+
+    def __repr__(self):
+        res = super().__repr__()
+        if (self._must_cancel and
+            self._state == futures._PENDING and
+            '<PENDING' in res):
+            res = res.replace('<PENDING', '<CANCELLING', 1)
+        i = res.find('<')
+        if i < 0:
+            i = len(res)
+        res = res[:i] + '(<{}>)'.format(self._coro.__name__) + res[i:]
+        return res
+
+    def get_stack(self, *, limit=None):
+        """Return the list of stack frames for this task's coroutine.
+
+        If the coroutine is active, this returns the stack where it is
+        suspended.  If the coroutine has completed successfully or was
+        cancelled, this returns an empty list.  If the coroutine was
+        terminated by an exception, this returns the list of traceback
+        frames.
+
+        The frames are always ordered from oldest to newest.
+
+        The optional limit gives the maximum nummber of frames to
+        return; by default all available frames are returned.  Its
+        meaning differs depending on whether a stack or a traceback is
+        returned: the newest frames of a stack are returned, but the
+        oldest frames of a traceback are returned.  (This matches the
+        behavior of the traceback module.)
+
+        For reasons beyond our control, only one stack frame is
+        returned for a suspended coroutine.
+        """
+        frames = []
+        f = self._coro.gi_frame
+        if f is not None:
+            while f is not None:
+                if limit is not None:
+                    if limit <= 0:
+                        break
+                    limit -= 1
+                frames.append(f)
+                f = f.f_back
+            frames.reverse()
+        elif self._exception is not None:
+            tb = self._exception.__traceback__
+            while tb is not None:
+                if limit is not None:
+                    if limit <= 0:
+                        break
+                    limit -= 1
+                frames.append(tb.tb_frame)
+                tb = tb.tb_next
+        return frames
+
+    def print_stack(self, *, limit=None, file=None):
+        """Print the stack or traceback for this task's coroutine.
+
+        This produces output similar to that of the traceback module,
+        for the frames retrieved by get_stack().  The limit argument
+        is passed to get_stack().  The file argument is an I/O stream
+        to which the output goes; by default it goes to sys.stderr.
+        """
+        extracted_list = []
+        checked = set()
+        for f in self.get_stack(limit=limit):
+            lineno = f.f_lineno
+            co = f.f_code
+            filename = co.co_filename
+            name = co.co_name
+            if filename not in checked:
+                checked.add(filename)
+                linecache.checkcache(filename)
+            line = linecache.getline(filename, lineno, f.f_globals)
+            extracted_list.append((filename, lineno, name, line))
+        exc = self._exception
+        if not extracted_list:
+            print('No stack for %r' % self, file=file)
+        elif exc is not None:
+            print('Traceback for %r (most recent call last):' % self,
+                  file=file)
+        else:
+            print('Stack for %r (most recent call last):' % self,
+                  file=file)
+        traceback.print_list(extracted_list, file=file)
+        if exc is not None:
+            for line in traceback.format_exception_only(exc.__class__, exc):
+                print(line, file=file, end='')
+
+    def cancel(self):
+        if self.done():
+            return False
+        if self._fut_waiter is not None:
+            if self._fut_waiter.cancel():
+                # Leave self._fut_waiter; it may be a Task that
+                # catches and ignores the cancellation so we may have
+                # to cancel it again later.
+                return True
+        # It must be the case that self._step is already scheduled.
+        self._must_cancel = True
+        return True
+
+    def _step(self, value=None, exc=None):
+        assert not self.done(), \
+            '_step(): already done: {!r}, {!r}, {!r}'.format(self, value, exc)
+        if self._must_cancel:
+            if not isinstance(exc, futures.CancelledError):
+                exc = futures.CancelledError()
+            self._must_cancel = False
+        coro = self._coro
+        self._fut_waiter = None
+        # Call either coro.throw(exc) or coro.send(value).
+        try:
+            if exc is not None:
+                result = coro.throw(exc)
+            elif value is not None:
+                result = coro.send(value)
+            else:
+                result = next(coro)
+        except StopIteration as exc:
+            self.set_result(exc.value)
+        except futures.CancelledError as exc:
+            super().cancel()  # I.e., Future.cancel(self).
+        except Exception as exc:
+            self.set_exception(exc)
+        except BaseException as exc:
+            self.set_exception(exc)
+            raise
+        else:
+            if isinstance(result, futures.Future):
+                # Yielded Future must come from Future.__iter__().
+                if result._blocking:
+                    result._blocking = False
+                    result.add_done_callback(self._wakeup)
+                    self._fut_waiter = result
+                    if self._must_cancel:
+                        if self._fut_waiter.cancel():
+                            self._must_cancel = False
+                else:
+                    self._loop.call_soon(
+                        self._step, None,
+                        RuntimeError(
+                            'yield was used instead of yield from '
+                            'in task {!r} with {!r}'.format(self, result)))
+            elif result is None:
+                # Bare yield relinquishes control for one event loop iteration.
+                self._loop.call_soon(self._step)
+            elif inspect.isgenerator(result):
+                # Yielding a generator is just wrong.
+                self._loop.call_soon(
+                    self._step, None,
+                    RuntimeError(
+                        'yield was used instead of yield from for '
+                        'generator in task {!r} with {}'.format(
+                            self, result)))
+            else:
+                # Yielding something else is an error.
+                self._loop.call_soon(
+                    self._step, None,
+                    RuntimeError(
+                        'Task got bad yield: {!r}'.format(result)))
+        self = None
+
+    def _wakeup(self, future):
+        try:
+            value = future.result()
+        except Exception as exc:
+            # This may also be a cancellation.
+            self._step(None, exc)
+        else:
+            self._step(value, None)
+        self = None  # Needed to break cycles when an exception occurs.
+
+
+# wait() and as_completed() similar to those in PEP 3148.
+
+FIRST_COMPLETED = concurrent.futures.FIRST_COMPLETED
+FIRST_EXCEPTION = concurrent.futures.FIRST_EXCEPTION
+ALL_COMPLETED = concurrent.futures.ALL_COMPLETED
+
+
+@coroutine
+def wait(fs, *, loop=None, timeout=None, return_when=ALL_COMPLETED):
+    """Wait for the Futures and coroutines given by fs to complete.
+
+    Coroutines will be wrapped in Tasks.
+
+    Returns two sets of Future: (done, pending).
+
+    Usage:
+
+        done, pending = yield from asyncio.wait(fs)
+
+    Note: This does not raise TimeoutError! Futures that aren't done
+    when the timeout occurs are returned in the second set.
+    """
+    if not fs:
+        raise ValueError('Set of coroutines/Futures is empty.')
+
+    if loop is None:
+        loop = events.get_event_loop()
+
+    fs = set(async(f, loop=loop) for f in fs)
+
+    if return_when not in (FIRST_COMPLETED, FIRST_EXCEPTION, ALL_COMPLETED):
+        raise ValueError('Invalid return_when value: {}'.format(return_when))
+    return (yield from _wait(fs, timeout, return_when, loop))
+
+
+def _release_waiter(waiter, value=True, *args):
+    if not waiter.done():
+        waiter.set_result(value)
+
+
+@coroutine
+def wait_for(fut, timeout, *, loop=None):
+    """Wait for the single Future or coroutine to complete, with timeout.
+
+    Coroutine will be wrapped in Task.
+
+    Returns result of the Future or coroutine.  Raises TimeoutError when
+    timeout occurs.
+
+    Usage:
+
+        result = yield from asyncio.wait_for(fut, 10.0)
+
+    """
+    if loop is None:
+        loop = events.get_event_loop()
+
+    waiter = futures.Future(loop=loop)
+    timeout_handle = loop.call_later(timeout, _release_waiter, waiter, False)
+    cb = functools.partial(_release_waiter, waiter, True)
+
+    fut = async(fut, loop=loop)
+    fut.add_done_callback(cb)
+
+    try:
+        if (yield from waiter):
+            return fut.result()
+        else:
+            fut.remove_done_callback(cb)
+            raise futures.TimeoutError()
+    finally:
+        timeout_handle.cancel()
+
+
+@coroutine
+def _wait(fs, timeout, return_when, loop):
+    """Internal helper for wait() and _wait_for().
+
+    The fs argument must be a collection of Futures.
+    """
+    assert fs, 'Set of Futures is empty.'
+    waiter = futures.Future(loop=loop)
+    timeout_handle = None
+    if timeout is not None:
+        timeout_handle = loop.call_later(timeout, _release_waiter, waiter)
+    counter = len(fs)
+
+    def _on_completion(f):
+        nonlocal counter
+        counter -= 1
+        if (counter <= 0 or
+            return_when == FIRST_COMPLETED or
+            return_when == FIRST_EXCEPTION and (not f.cancelled() and
+                                                f.exception() is not None)):
+            if timeout_handle is not None:
+                timeout_handle.cancel()
+            if not waiter.done():
+                waiter.set_result(False)
+
+    for f in fs:
+        f.add_done_callback(_on_completion)
+
+    try:
+        yield from waiter
+    finally:
+        if timeout_handle is not None:
+            timeout_handle.cancel()
+
+    done, pending = set(), set()
+    for f in fs:
+        f.remove_done_callback(_on_completion)
+        if f.done():
+            done.add(f)
+        else:
+            pending.add(f)
+    return done, pending
+
+
+# This is *not* a @coroutine!  It is just an iterator (yielding Futures).
+def as_completed(fs, *, loop=None, timeout=None):
+    """Return an iterator whose values, when waited for, are Futures.
+
+    This differs from PEP 3148; the proper way to use this is:
+
+        for f in as_completed(fs):
+            result = yield from f  # The 'yield from' may raise.
+            # Use result.
+
+    Raises TimeoutError if the timeout occurs before all Futures are
+    done.
+
+    Note: The futures 'f' are not necessarily members of fs.
+    """
+    loop = loop if loop is not None else events.get_event_loop()
+    deadline = None if timeout is None else loop.time() + timeout
+    todo = set(async(f, loop=loop) for f in fs)
+    completed = collections.deque()
+
+    @coroutine
+    def _wait_for_one():
+        while not completed:
+            timeout = None
+            if deadline is not None:
+                timeout = deadline - loop.time()
+                if timeout < 0:
+                    raise futures.TimeoutError()
+            done, pending = yield from _wait(
+                todo, timeout, FIRST_COMPLETED, loop)
+            # Multiple callers might be waiting for the same events
+            # and getting the same outcome.  Dedupe by updating todo.
+            for f in done:
+                if f in todo:
+                    todo.remove(f)
+                    completed.append(f)
+        f = completed.popleft()
+        return f.result()  # May raise.
+
+    for _ in range(len(todo)):
+        yield _wait_for_one()
+
+
+@coroutine
+def sleep(delay, result=None, *, loop=None):
+    """Coroutine that completes after a given time (in seconds)."""
+    future = futures.Future(loop=loop)
+    h = future._loop.call_later(delay, future.set_result, result)
+    try:
+        return (yield from future)
+    finally:
+        h.cancel()
+
+
+def async(coro_or_future, *, loop=None):
+    """Wrap a coroutine in a future.
+
+    If the argument is a Future, it is returned directly.
+    """
+    if isinstance(coro_or_future, futures.Future):
+        if loop is not None and loop is not coro_or_future._loop:
+            raise ValueError('loop argument must agree with Future')
+        return coro_or_future
+    elif iscoroutine(coro_or_future):
+        return Task(coro_or_future, loop=loop)
+    else:
+        raise TypeError('A Future or coroutine is required')
+
+
+class _GatheringFuture(futures.Future):
+    """Helper for gather().
+
+    This overrides cancel() to cancel all the children and act more
+    like Task.cancel(), which doesn't immediately mark itself as
+    cancelled.
+    """
+
+    def __init__(self, children, *, loop=None):
+        super().__init__(loop=loop)
+        self._children = children
+
+    def cancel(self):
+        if self.done():
+            return False
+        for child in self._children:
+            child.cancel()
+        return True
+
+
+def gather(*coros_or_futures, loop=None, return_exceptions=False):
+    """Return a future aggregating results from the given coroutines
+    or futures.
+
+    All futures must share the same event loop.  If all the tasks are
+    done successfully, the returned future's result is the list of
+    results (in the order of the original sequence, not necessarily
+    the order of results arrival).  If *result_exception* is True,
+    exceptions in the tasks are treated the same as successful
+    results, and gathered in the result list; otherwise, the first
+    raised exception will be immediately propagated to the returned
+    future.
+
+    Cancellation: if the outer Future is cancelled, all children (that
+    have not completed yet) are also cancelled.  If any child is
+    cancelled, this is treated as if it raised CancelledError --
+    the outer Future is *not* cancelled in this case.  (This is to
+    prevent the cancellation of one child to cause other children to
+    be cancelled.)
+    """
+    children = [async(fut, loop=loop) for fut in coros_or_futures]
+    n = len(children)
+    if n == 0:
+        outer = futures.Future(loop=loop)
+        outer.set_result([])
+        return outer
+    if loop is None:
+        loop = children[0]._loop
+    for fut in children:
+        if fut._loop is not loop:
+            raise ValueError("futures are tied to different event loops")
+    outer = _GatheringFuture(children, loop=loop)
+    nfinished = 0
+    results = [None] * n
+
+    def _done_callback(i, fut):
+        nonlocal nfinished
+        if outer._state != futures._PENDING:
+            if fut._exception is not None:
+                # Mark exception retrieved.
+                fut.exception()
+            return
+        if fut._state == futures._CANCELLED:
+            res = futures.CancelledError()
+            if not return_exceptions:
+                outer.set_exception(res)
+                return
+        elif fut._exception is not None:
+            res = fut.exception()  # Mark exception retrieved.
+            if not return_exceptions:
+                outer.set_exception(res)
+                return
+        else:
+            res = fut._result
+        results[i] = res
+        nfinished += 1
+        if nfinished == n:
+            outer.set_result(results)
+
+    for i, fut in enumerate(children):
+        fut.add_done_callback(functools.partial(_done_callback, i))
+    return outer
+
+
+def shield(arg, *, loop=None):
+    """Wait for a future, shielding it from cancellation.
+
+    The statement
+
+        res = yield from shield(something())
+
+    is exactly equivalent to the statement
+
+        res = yield from something()
+
+    *except* that if the coroutine containing it is cancelled, the
+    task running in something() is not cancelled.  From the POV of
+    something(), the cancellation did not happen.  But its caller is
+    still cancelled, so the yield-from expression still raises
+    CancelledError.  Note: If something() is cancelled by other means
+    this will still cancel shield().
+
+    If you want to completely ignore cancellation (not recommended)
+    you can combine shield() with a try/except clause, as follows:
+
+        try:
+            res = yield from shield(something())
+        except CancelledError:
+            res = None
+    """
+    inner = async(arg, loop=loop)
+    if inner.done():
+        # Shortcut.
+        return inner
+    loop = inner._loop
+    outer = futures.Future(loop=loop)
+
+    def _done_callback(inner):
+        if outer.cancelled():
+            # Mark inner's result as retrieved.
+            inner.cancelled() or inner.exception()
+            return
+        if inner.cancelled():
+            outer.cancel()
+        else:
+            exc = inner.exception()
+            if exc is not None:
+                outer.set_exception(exc)
+            else:
+                outer.set_result(inner.result())
+
+    inner.add_done_callback(_done_callback)
+    return outer
diff --git a/Lib/asyncio/test_utils.py b/Lib/asyncio/test_utils.py
new file mode 100644
index 0000000..91bbedb
--- /dev/null
+++ b/Lib/asyncio/test_utils.py
@@ -0,0 +1,246 @@
+"""Utilities shared by tests."""
+
+import collections
+import contextlib
+import io
+import unittest.mock
+import os
+import sys
+import threading
+import unittest
+import unittest.mock
+from wsgiref.simple_server import make_server, WSGIRequestHandler, WSGIServer
+try:
+    import ssl
+except ImportError:  # pragma: no cover
+    ssl = None
+
+from . import tasks
+from . import base_events
+from . import events
+from . import selectors
+
+
+if sys.platform == 'win32':  # pragma: no cover
+    from .windows_utils import socketpair
+else:
+    from socket import socketpair  # pragma: no cover
+
+
+def dummy_ssl_context():
+    if ssl is None:
+        return None
+    else:
+        return ssl.SSLContext(ssl.PROTOCOL_SSLv23)
+
+
+def run_briefly(loop):
+    @tasks.coroutine
+    def once():
+        pass
+    gen = once()
+    t = tasks.Task(gen, loop=loop)
+    try:
+        loop.run_until_complete(t)
+    finally:
+        gen.close()
+
+
+def run_once(loop):
+    """loop.stop() schedules _raise_stop_error()
+    and run_forever() runs until _raise_stop_error() callback.
+    this wont work if test waits for some IO events, because
+    _raise_stop_error() runs before any of io events callbacks.
+    """
+    loop.stop()
+    loop.run_forever()
+
+
+@contextlib.contextmanager
+def run_test_server(*, host='127.0.0.1', port=0, use_ssl=False):
+
+    class SilentWSGIRequestHandler(WSGIRequestHandler):
+        def get_stderr(self):
+            return io.StringIO()
+
+        def log_message(self, format, *args):
+            pass
+
+    class SilentWSGIServer(WSGIServer):
+        def handle_error(self, request, client_address):
+            pass
+
+    class SSLWSGIServer(SilentWSGIServer):
+        def finish_request(self, request, client_address):
+            # The relative location of our test directory (which
+            # contains the sample key and certificate files) differs
+            # between the stdlib and stand-alone Tulip/asyncio.
+            # Prefer our own if we can find it.
+            here = os.path.join(os.path.dirname(__file__), '..', 'tests')
+            if not os.path.isdir(here):
+                here = os.path.join(os.path.dirname(os.__file__),
+                                    'test', 'test_asyncio')
+            keyfile = os.path.join(here, 'sample.key')
+            certfile = os.path.join(here, 'sample.crt')
+            ssock = ssl.wrap_socket(request,
+                                    keyfile=keyfile,
+                                    certfile=certfile,
+                                    server_side=True)
+            try:
+                self.RequestHandlerClass(ssock, client_address, self)
+                ssock.close()
+            except OSError:
+                # maybe socket has been closed by peer
+                pass
+
+    def app(environ, start_response):
+        status = '200 OK'
+        headers = [('Content-type', 'text/plain')]
+        start_response(status, headers)
+        return [b'Test message']
+
+    # Run the test WSGI server in a separate thread in order not to
+    # interfere with event handling in the main thread
+    server_class = SSLWSGIServer if use_ssl else SilentWSGIServer
+    httpd = make_server(host, port, app,
+                        server_class, SilentWSGIRequestHandler)
+    httpd.address = httpd.server_address
+    server_thread = threading.Thread(target=httpd.serve_forever)
+    server_thread.start()
+    try:
+        yield httpd
+    finally:
+        httpd.shutdown()
+        server_thread.join()
+
+
+def make_test_protocol(base):
+    dct = {}
+    for name in dir(base):
+        if name.startswith('__') and name.endswith('__'):
+            # skip magic names
+            continue
+        dct[name] = unittest.mock.Mock(return_value=None)
+    return type('TestProtocol', (base,) + base.__bases__, dct)()
+
+
+class TestSelector(selectors.BaseSelector):
+
+    def select(self, timeout):
+        return []
+
+
+class TestLoop(base_events.BaseEventLoop):
+    """Loop for unittests.
+
+    It manages self time directly.
+    If something scheduled to be executed later then
+    on next loop iteration after all ready handlers done
+    generator passed to __init__ is calling.
+
+    Generator should be like this:
+
+        def gen():
+            ...
+            when = yield ...
+            ... = yield time_advance
+
+    Value retuned by yield is absolute time of next scheduled handler.
+    Value passed to yield is time advance to move loop's time forward.
+    """
+
+    def __init__(self, gen=None):
+        super().__init__()
+
+        if gen is None:
+            def gen():
+                yield
+            self._check_on_close = False
+        else:
+            self._check_on_close = True
+
+        self._gen = gen()
+        next(self._gen)
+        self._time = 0
+        self._timers = []
+        self._selector = TestSelector()
+
+        self.readers = {}
+        self.writers = {}
+        self.reset_counters()
+
+    def time(self):
+        return self._time
+
+    def advance_time(self, advance):
+        """Move test time forward."""
+        if advance:
+            self._time += advance
+
+    def close(self):
+        if self._check_on_close:
+            try:
+                self._gen.send(0)
+            except StopIteration:
+                pass
+            else:  # pragma: no cover
+                raise AssertionError("Time generator is not finished")
+
+    def add_reader(self, fd, callback, *args):
+        self.readers[fd] = events.make_handle(callback, args)
+
+    def remove_reader(self, fd):
+        self.remove_reader_count[fd] += 1
+        if fd in self.readers:
+            del self.readers[fd]
+            return True
+        else:
+            return False
+
+    def assert_reader(self, fd, callback, *args):
+        assert fd in self.readers, 'fd {} is not registered'.format(fd)
+        handle = self.readers[fd]
+        assert handle._callback == callback, '{!r} != {!r}'.format(
+            handle._callback, callback)
+        assert handle._args == args, '{!r} != {!r}'.format(
+            handle._args, args)
+
+    def add_writer(self, fd, callback, *args):
+        self.writers[fd] = events.make_handle(callback, args)
+
+    def remove_writer(self, fd):
+        self.remove_writer_count[fd] += 1
+        if fd in self.writers:
+            del self.writers[fd]
+            return True
+        else:
+            return False
+
+    def assert_writer(self, fd, callback, *args):
+        assert fd in self.writers, 'fd {} is not registered'.format(fd)
+        handle = self.writers[fd]
+        assert handle._callback == callback, '{!r} != {!r}'.format(
+            handle._callback, callback)
+        assert handle._args == args, '{!r} != {!r}'.format(
+            handle._args, args)
+
+    def reset_counters(self):
+        self.remove_reader_count = collections.defaultdict(int)
+        self.remove_writer_count = collections.defaultdict(int)
+
+    def _run_once(self):
+        super()._run_once()
+        for when in self._timers:
+            advance = self._gen.send(when)
+            self.advance_time(advance)
+        self._timers = []
+
+    def call_at(self, when, callback, *args):
+        self._timers.append(when)
+        return super().call_at(when, callback, *args)
+
+    def _process_events(self, event_list):
+        return
+
+    def _write_to_self(self):
+        pass
diff --git a/Lib/asyncio/transports.py b/Lib/asyncio/transports.py
new file mode 100644
index 0000000..bf3adee
--- /dev/null
+++ b/Lib/asyncio/transports.py
@@ -0,0 +1,186 @@
+"""Abstract Transport class."""
+
+__all__ = ['ReadTransport', 'WriteTransport', 'Transport']
+
+
+class BaseTransport:
+    """Base ABC for transports."""
+
+    def __init__(self, extra=None):
+        if extra is None:
+            extra = {}
+        self._extra = extra
+
+    def get_extra_info(self, name, default=None):
+        """Get optional transport information."""
+        return self._extra.get(name, default)
+
+    def close(self):
+        """Closes the transport.
+
+        Buffered data will be flushed asynchronously.  No more data
+        will be received.  After all buffered data is flushed, the
+        protocol's connection_lost() method will (eventually) called
+        with None as its argument.
+        """
+        raise NotImplementedError
+
+
+class ReadTransport(BaseTransport):
+    """ABC for read-only transports."""
+
+    def pause(self):
+        """Pause the receiving end.
+
+        No data will be passed to the protocol's data_received()
+        method until resume() is called.
+        """
+        raise NotImplementedError
+
+    def resume(self):
+        """Resume the receiving end.
+
+        Data received will once again be passed to the protocol's
+        data_received() method.
+        """
+        raise NotImplementedError
+
+
+class WriteTransport(BaseTransport):
+    """ABC for write-only transports."""
+
+    def write(self, data):
+        """Write some data bytes to the transport.
+
+        This does not block; it buffers the data and arranges for it
+        to be sent out asynchronously.
+        """
+        raise NotImplementedError
+
+    def writelines(self, list_of_data):
+        """Write a list (or any iterable) of data bytes to the transport.
+
+        The default implementation just calls write() for each item in
+        the list/iterable.
+        """
+        for data in list_of_data:
+            self.write(data)
+
+    def write_eof(self):
+        """Closes the write end after flushing buffered data.
+
+        (This is like typing ^D into a UNIX program reading from stdin.)
+
+        Data may still be received.
+        """
+        raise NotImplementedError
+
+    def can_write_eof(self):
+        """Return True if this protocol supports write_eof(), False if not."""
+        raise NotImplementedError
+
+    def abort(self):
+        """Closes the transport immediately.
+
+        Buffered data will be lost.  No more data will be received.
+        The protocol's connection_lost() method will (eventually) be
+        called with None as its argument.
+        """
+        raise NotImplementedError
+
+
+class Transport(ReadTransport, WriteTransport):
+    """ABC representing a bidirectional transport.
+
+    There may be several implementations, but typically, the user does
+    not implement new transports; rather, the platform provides some
+    useful transports that are implemented using the platform's best
+    practices.
+
+    The user never instantiates a transport directly; they call a
+    utility function, passing it a protocol factory and other
+    information necessary to create the transport and protocol.  (E.g.
+    EventLoop.create_connection() or EventLoop.create_server().)
+
+    The utility function will asynchronously create a transport and a
+    protocol and hook them up by calling the protocol's
+    connection_made() method, passing it the transport.
+
+    The implementation here raises NotImplemented for every method
+    except writelines(), which calls write() in a loop.
+    """
+
+
+class DatagramTransport(BaseTransport):
+    """ABC for datagram (UDP) transports."""
+
+    def sendto(self, data, addr=None):
+        """Send data to the transport.
+
+        This does not block; it buffers the data and arranges for it
+        to be sent out asynchronously.
+        addr is target socket address.
+        If addr is None use target address pointed on transport creation.
+        """
+        raise NotImplementedError
+
+    def abort(self):
+        """Closes the transport immediately.
+
+        Buffered data will be lost.  No more data will be received.
+        The protocol's connection_lost() method will (eventually) be
+        called with None as its argument.
+        """
+        raise NotImplementedError
+
+
+class SubprocessTransport(BaseTransport):
+
+    def get_pid(self):
+        """Get subprocess id."""
+        raise NotImplementedError
+
+    def get_returncode(self):
+        """Get subprocess returncode.
+
+        See also
+        http://docs.python.org/3/library/subprocess#subprocess.Popen.returncode
+        """
+        raise NotImplementedError
+
+    def get_pipe_transport(self, fd):
+        """Get transport for pipe with number fd."""
+        raise NotImplementedError
+
+    def send_signal(self, signal):
+        """Send signal to subprocess.
+
+        See also:
+        docs.python.org/3/library/subprocess#subprocess.Popen.send_signal
+        """
+        raise NotImplementedError
+
+    def terminate(self):
+        """Stop the subprocess.
+
+        Alias for close() method.
+
+        On Posix OSs the method sends SIGTERM to the subprocess.
+        On Windows the Win32 API function TerminateProcess()
+         is called to stop the subprocess.
+
+        See also:
+        http://docs.python.org/3/library/subprocess#subprocess.Popen.terminate
+        """
+        raise NotImplementedError
+
+    def kill(self):
+        """Kill the subprocess.
+
+        On Posix OSs the function sends SIGKILL to the subprocess.
+        On Windows kill() is an alias for terminate().
+
+        See also:
+        http://docs.python.org/3/library/subprocess#subprocess.Popen.kill
+        """
+        raise NotImplementedError
diff --git a/Lib/asyncio/unix_events.py b/Lib/asyncio/unix_events.py
new file mode 100644
index 0000000..a3a8e11
--- /dev/null
+++ b/Lib/asyncio/unix_events.py
@@ -0,0 +1,541 @@
+"""Selector eventloop for Unix with signal handling."""
+
+import collections
+import errno
+import fcntl
+import functools
+import os
+import signal
+import socket
+import stat
+import subprocess
+import sys
+
+
+from . import constants
+from . import events
+from . import protocols
+from . import selector_events
+from . import tasks
+from . import transports
+from .log import asyncio_log
+
+
+__all__ = ['SelectorEventLoop', 'STDIN', 'STDOUT', 'STDERR']
+
+STDIN = 0
+STDOUT = 1
+STDERR = 2
+
+
+if sys.platform == 'win32':  # pragma: no cover
+    raise ImportError('Signals are not really supported on Windows')
+
+
+class SelectorEventLoop(selector_events.BaseSelectorEventLoop):
+    """Unix event loop
+
+    Adds signal handling to SelectorEventLoop
+    """
+
+    def __init__(self, selector=None):
+        super().__init__(selector)
+        self._signal_handlers = {}
+        self._subprocesses = {}
+
+    def _socketpair(self):
+        return socket.socketpair()
+
+    def close(self):
+        handler = self._signal_handlers.get(signal.SIGCHLD)
+        if handler is not None:
+            self.remove_signal_handler(signal.SIGCHLD)
+        super().close()
+
+    def add_signal_handler(self, sig, callback, *args):
+        """Add a handler for a signal.  UNIX only.
+
+        Raise ValueError if the signal number is invalid or uncatchable.
+        Raise RuntimeError if there is a problem setting up the handler.
+        """
+        self._check_signal(sig)
+        try:
+            # set_wakeup_fd() raises ValueError if this is not the
+            # main thread.  By calling it early we ensure that an
+            # event loop running in another thread cannot add a signal
+            # handler.
+            signal.set_wakeup_fd(self._csock.fileno())
+        except ValueError as exc:
+            raise RuntimeError(str(exc))
+
+        handle = events.make_handle(callback, args)
+        self._signal_handlers[sig] = handle
+
+        try:
+            signal.signal(sig, self._handle_signal)
+        except OSError as exc:
+            del self._signal_handlers[sig]
+            if not self._signal_handlers:
+                try:
+                    signal.set_wakeup_fd(-1)
+                except ValueError as nexc:
+                    asyncio_log.info('set_wakeup_fd(-1) failed: %s', nexc)
+
+            if exc.errno == errno.EINVAL:
+                raise RuntimeError('sig {} cannot be caught'.format(sig))
+            else:
+                raise
+
+    def _handle_signal(self, sig, arg):
+        """Internal helper that is the actual signal handler."""
+        handle = self._signal_handlers.get(sig)
+        if handle is None:
+            return  # Assume it's some race condition.
+        if handle._cancelled:
+            self.remove_signal_handler(sig)  # Remove it properly.
+        else:
+            self._add_callback_signalsafe(handle)
+
+    def remove_signal_handler(self, sig):
+        """Remove a handler for a signal.  UNIX only.
+
+        Return True if a signal handler was removed, False if not.
+        """
+        self._check_signal(sig)
+        try:
+            del self._signal_handlers[sig]
+        except KeyError:
+            return False
+
+        if sig == signal.SIGINT:
+            handler = signal.default_int_handler
+        else:
+            handler = signal.SIG_DFL
+
+        try:
+            signal.signal(sig, handler)
+        except OSError as exc:
+            if exc.errno == errno.EINVAL:
+                raise RuntimeError('sig {} cannot be caught'.format(sig))
+            else:
+                raise
+
+        if not self._signal_handlers:
+            try:
+                signal.set_wakeup_fd(-1)
+            except ValueError as exc:
+                asyncio_log.info('set_wakeup_fd(-1) failed: %s', exc)
+
+        return True
+
+    def _check_signal(self, sig):
+        """Internal helper to validate a signal.
+
+        Raise ValueError if the signal number is invalid or uncatchable.
+        Raise RuntimeError if there is a problem setting up the handler.
+        """
+        if not isinstance(sig, int):
+            raise TypeError('sig must be an int, not {!r}'.format(sig))
+
+        if not (1 <= sig < signal.NSIG):
+            raise ValueError(
+                'sig {} out of range(1, {})'.format(sig, signal.NSIG))
+
+    def _make_read_pipe_transport(self, pipe, protocol, waiter=None,
+                                  extra=None):
+        return _UnixReadPipeTransport(self, pipe, protocol, waiter, extra)
+
+    def _make_write_pipe_transport(self, pipe, protocol, waiter=None,
+                                   extra=None):
+        return _UnixWritePipeTransport(self, pipe, protocol, waiter, extra)
+
+    @tasks.coroutine
+    def _make_subprocess_transport(self, protocol, args, shell,
+                                   stdin, stdout, stderr, bufsize,
+                                   extra=None, **kwargs):
+        self._reg_sigchld()
+        transp = _UnixSubprocessTransport(self, protocol, args, shell,
+                                          stdin, stdout, stderr, bufsize,
+                                          extra=None, **kwargs)
+        self._subprocesses[transp.get_pid()] = transp
+        yield from transp._post_init()
+        return transp
+
+    def _reg_sigchld(self):
+        if signal.SIGCHLD not in self._signal_handlers:
+            self.add_signal_handler(signal.SIGCHLD, self._sig_chld)
+
+    def _sig_chld(self):
+        try:
+            try:
+                pid, status = os.waitpid(0, os.WNOHANG)
+            except ChildProcessError:
+                return
+            if pid == 0:
+                self.call_soon(self._sig_chld)
+                return
+            elif os.WIFSIGNALED(status):
+                returncode = -os.WTERMSIG(status)
+            elif os.WIFEXITED(status):
+                returncode = os.WEXITSTATUS(status)
+            else:
+                self.call_soon(self._sig_chld)
+                return
+            transp = self._subprocesses.get(pid)
+            if transp is not None:
+                transp._process_exited(returncode)
+        except Exception:
+            asyncio_log.exception('Unknown exception in SIGCHLD handler')
+
+    def _subprocess_closed(self, transport):
+        pid = transport.get_pid()
+        self._subprocesses.pop(pid, None)
+
+
+def _set_nonblocking(fd):
+    flags = fcntl.fcntl(fd, fcntl.F_GETFL)
+    flags = flags | os.O_NONBLOCK
+    fcntl.fcntl(fd, fcntl.F_SETFL, flags)
+
+
+class _UnixReadPipeTransport(transports.ReadTransport):
+
+    max_size = 256 * 1024  # max bytes we read in one eventloop iteration
+
+    def __init__(self, loop, pipe, protocol, waiter=None, extra=None):
+        super().__init__(extra)
+        self._extra['pipe'] = pipe
+        self._loop = loop
+        self._pipe = pipe
+        self._fileno = pipe.fileno()
+        _set_nonblocking(self._fileno)
+        self._protocol = protocol
+        self._closing = False
+        self._loop.add_reader(self._fileno, self._read_ready)
+        self._loop.call_soon(self._protocol.connection_made, self)
+        if waiter is not None:
+            self._loop.call_soon(waiter.set_result, None)
+
+    def _read_ready(self):
+        try:
+            data = os.read(self._fileno, self.max_size)
+        except (BlockingIOError, InterruptedError):
+            pass
+        except OSError as exc:
+            self._fatal_error(exc)
+        else:
+            if data:
+                self._protocol.data_received(data)
+            else:
+                self._closing = True
+                self._loop.remove_reader(self._fileno)
+                self._loop.call_soon(self._protocol.eof_received)
+                self._loop.call_soon(self._call_connection_lost, None)
+
+    def pause(self):
+        self._loop.remove_reader(self._fileno)
+
+    def resume(self):
+        self._loop.add_reader(self._fileno, self._read_ready)
+
+    def close(self):
+        if not self._closing:
+            self._close(None)
+
+    def _fatal_error(self, exc):
+        # should be called by exception handler only
+        asyncio_log.exception('Fatal error for %s', self)
+        self._close(exc)
+
+    def _close(self, exc):
+        self._closing = True
+        self._loop.remove_reader(self._fileno)
+        self._loop.call_soon(self._call_connection_lost, exc)
+
+    def _call_connection_lost(self, exc):
+        try:
+            self._protocol.connection_lost(exc)
+        finally:
+            self._pipe.close()
+            self._pipe = None
+            self._protocol = None
+            self._loop = None
+
+
+class _UnixWritePipeTransport(transports.WriteTransport):
+
+    def __init__(self, loop, pipe, protocol, waiter=None, extra=None):
+        super().__init__(extra)
+        self._extra['pipe'] = pipe
+        self._loop = loop
+        self._pipe = pipe
+        self._fileno = pipe.fileno()
+        if not stat.S_ISFIFO(os.fstat(self._fileno).st_mode):
+            raise ValueError("Pipe transport is for pipes only.")
+        _set_nonblocking(self._fileno)
+        self._protocol = protocol
+        self._buffer = []
+        self._conn_lost = 0
+        self._closing = False  # Set when close() or write_eof() called.
+        self._loop.add_reader(self._fileno, self._read_ready)
+
+        self._loop.call_soon(self._protocol.connection_made, self)
+        if waiter is not None:
+            self._loop.call_soon(waiter.set_result, None)
+
+    def _read_ready(self):
+        # pipe was closed by peer
+        self._close()
+
+    def write(self, data):
+        assert isinstance(data, bytes), repr(data)
+        if not data:
+            return
+
+        if self._conn_lost or self._closing:
+            if self._conn_lost >= constants.LOG_THRESHOLD_FOR_CONNLOST_WRITES:
+                asyncio_log.warning('pipe closed by peer or '
+                                    'os.write(pipe, data) raised exception.')
+            self._conn_lost += 1
+            return
+
+        if not self._buffer:
+            # Attempt to send it right away first.
+            try:
+                n = os.write(self._fileno, data)
+            except (BlockingIOError, InterruptedError):
+                n = 0
+            except Exception as exc:
+                self._conn_lost += 1
+                self._fatal_error(exc)
+                return
+            if n == len(data):
+                return
+            elif n > 0:
+                data = data[n:]
+            self._loop.add_writer(self._fileno, self._write_ready)
+
+        self._buffer.append(data)
+
+    def _write_ready(self):
+        data = b''.join(self._buffer)
+        assert data, 'Data should not be empty'
+
+        self._buffer.clear()
+        try:
+            n = os.write(self._fileno, data)
+        except (BlockingIOError, InterruptedError):
+            self._buffer.append(data)
+        except Exception as exc:
+            self._conn_lost += 1
+            # Remove writer here, _fatal_error() doesn't it
+            # because _buffer is empty.
+            self._loop.remove_writer(self._fileno)
+            self._fatal_error(exc)
+        else:
+            if n == len(data):
+                self._loop.remove_writer(self._fileno)
+                if self._closing:
+                    self._loop.remove_reader(self._fileno)
+                    self._call_connection_lost(None)
+                return
+            elif n > 0:
+                data = data[n:]
+
+            self._buffer.append(data)  # Try again later.
+
+    def can_write_eof(self):
+        return True
+
+    # TODO: Make the relationships between write_eof(), close(),
+    # abort(), _fatal_error() and _close() more straightforward.
+
+    def write_eof(self):
+        if self._closing:
+            return
+        assert self._pipe
+        self._closing = True
+        if not self._buffer:
+            self._loop.remove_reader(self._fileno)
+            self._loop.call_soon(self._call_connection_lost, None)
+
+    def close(self):
+        if not self._closing:
+            # write_eof is all what we needed to close the write pipe
+            self.write_eof()
+
+    def abort(self):
+        self._close(None)
+
+    def _fatal_error(self, exc):
+        # should be called by exception handler only
+        asyncio_log.exception('Fatal error for %s', self)
+        self._close(exc)
+
+    def _close(self, exc=None):
+        self._closing = True
+        if self._buffer:
+            self._loop.remove_writer(self._fileno)
+        self._buffer.clear()
+        self._loop.remove_reader(self._fileno)
+        self._loop.call_soon(self._call_connection_lost, exc)
+
+    def _call_connection_lost(self, exc):
+        try:
+            self._protocol.connection_lost(exc)
+        finally:
+            self._pipe.close()
+            self._pipe = None
+            self._protocol = None
+            self._loop = None
+
+
+class _UnixWriteSubprocessPipeProto(protocols.BaseProtocol):
+    pipe = None
+
+    def __init__(self, proc, fd):
+        self.proc = proc
+        self.fd = fd
+        self.connected = False
+        self.disconnected = False
+        proc._pipes[fd] = self
+
+    def connection_made(self, transport):
+        self.connected = True
+        self.pipe = transport
+        self.proc._try_connected()
+
+    def connection_lost(self, exc):
+        self.disconnected = True
+        self.proc._pipe_connection_lost(self.fd, exc)
+
+
+class _UnixReadSubprocessPipeProto(_UnixWriteSubprocessPipeProto,
+                                   protocols.Protocol):
+
+    def data_received(self, data):
+        self.proc._pipe_data_received(self.fd, data)
+
+    def eof_received(self):
+        pass
+
+
+class _UnixSubprocessTransport(transports.SubprocessTransport):
+
+    def __init__(self, loop, protocol, args, shell,
+                 stdin, stdout, stderr, bufsize,
+                 extra=None, **kwargs):
+        super().__init__(extra)
+        self._protocol = protocol
+        self._loop = loop
+
+        self._pipes = {}
+        if stdin == subprocess.PIPE:
+            self._pipes[STDIN] = None
+        if stdout == subprocess.PIPE:
+            self._pipes[STDOUT] = None
+        if stderr == subprocess.PIPE:
+            self._pipes[STDERR] = None
+        self._pending_calls = collections.deque()
+        self._finished = False
+        self._returncode = None
+
+        self._proc = subprocess.Popen(
+            args, shell=shell, stdin=stdin, stdout=stdout, stderr=stderr,
+            universal_newlines=False, bufsize=bufsize, **kwargs)
+        self._extra['subprocess'] = self._proc
+
+    def close(self):
+        for proto in self._pipes.values():
+            proto.pipe.close()
+        if self._returncode is None:
+            self.terminate()
+
+    def get_pid(self):
+        return self._proc.pid
+
+    def get_returncode(self):
+        return self._returncode
+
+    def get_pipe_transport(self, fd):
+        if fd in self._pipes:
+            return self._pipes[fd].pipe
+        else:
+            return None
+
+    def send_signal(self, signal):
+        self._proc.send_signal(signal)
+
+    def terminate(self):
+        self._proc.terminate()
+
+    def kill(self):
+        self._proc.kill()
+
+    @tasks.coroutine
+    def _post_init(self):
+        proc = self._proc
+        loop = self._loop
+        if proc.stdin is not None:
+            transp, proto = yield from loop.connect_write_pipe(
+                functools.partial(
+                    _UnixWriteSubprocessPipeProto, self, STDIN),
+                proc.stdin)
+        if proc.stdout is not None:
+            transp, proto = yield from loop.connect_read_pipe(
+                functools.partial(
+                    _UnixReadSubprocessPipeProto, self, STDOUT),
+                proc.stdout)
+        if proc.stderr is not None:
+            transp, proto = yield from loop.connect_read_pipe(
+                functools.partial(
+                    _UnixReadSubprocessPipeProto, self, STDERR),
+                proc.stderr)
+        if not self._pipes:
+            self._try_connected()
+
+    def _call(self, cb, *data):
+        if self._pending_calls is not None:
+            self._pending_calls.append((cb, data))
+        else:
+            self._loop.call_soon(cb, *data)
+
+    def _try_connected(self):
+        assert self._pending_calls is not None
+        if all(p is not None and p.connected for p in self._pipes.values()):
+            self._loop.call_soon(self._protocol.connection_made, self)
+            for callback, data in self._pending_calls:
+                self._loop.call_soon(callback, *data)
+            self._pending_calls = None
+
+    def _pipe_connection_lost(self, fd, exc):
+        self._call(self._protocol.pipe_connection_lost, fd, exc)
+        self._try_finish()
+
+    def _pipe_data_received(self, fd, data):
+        self._call(self._protocol.pipe_data_received, fd, data)
+
+    def _process_exited(self, returncode):
+        assert returncode is not None, returncode
+        assert self._returncode is None, self._returncode
+        self._returncode = returncode
+        self._loop._subprocess_closed(self)
+        self._call(self._protocol.process_exited)
+        self._try_finish()
+
+    def _try_finish(self):
+        assert not self._finished
+        if self._returncode is None:
+            return
+        if all(p is not None and p.disconnected
+               for p in self._pipes.values()):
+            self._finished = True
+            self._loop.call_soon(self._call_connection_lost, None)
+
+    def _call_connection_lost(self, exc):
+        try:
+            self._protocol.connection_lost(exc)
+        finally:
+            self._proc = None
+            self._protocol = None
+            self._loop = None
diff --git a/Lib/asyncio/windows_events.py b/Lib/asyncio/windows_events.py
new file mode 100644
index 0000000..1d0ad26
--- /dev/null
+++ b/Lib/asyncio/windows_events.py
@@ -0,0 +1,375 @@
+"""Selector and proactor eventloops for Windows."""
+
+import errno
+import socket
+import weakref
+import struct
+import _winapi
+
+from . import futures
+from . import proactor_events
+from . import selector_events
+from . import tasks
+from . import windows_utils
+from .log import asyncio_log
+
+try:
+    import _overlapped
+except ImportError:
+    from . import _overlapped
+
+
+__all__ = ['SelectorEventLoop', 'ProactorEventLoop', 'IocpProactor']
+
+
+NULL = 0
+INFINITE = 0xffffffff
+ERROR_CONNECTION_REFUSED = 1225
+ERROR_CONNECTION_ABORTED = 1236
+
+
+class _OverlappedFuture(futures.Future):
+    """Subclass of Future which represents an overlapped operation.
+
+    Cancelling it will immediately cancel the overlapped operation.
+    """
+
+    def __init__(self, ov, *, loop=None):
+        super().__init__(loop=loop)
+        self.ov = ov
+
+    def cancel(self):
+        try:
+            self.ov.cancel()
+        except OSError:
+            pass
+        return super().cancel()
+
+
+class PipeServer(object):
+    """Class representing a pipe server.
+
+    This is much like a bound, listening socket.
+    """
+    def __init__(self, address):
+        self._address = address
+        self._free_instances = weakref.WeakSet()
+        self._pipe = self._server_pipe_handle(True)
+
+    def _get_unconnected_pipe(self):
+        # Create new instance and return previous one.  This ensures
+        # that (until the server is closed) there is always at least
+        # one pipe handle for address.  Therefore if a client attempt
+        # to connect it will not fail with FileNotFoundError.
+        tmp, self._pipe = self._pipe, self._server_pipe_handle(False)
+        return tmp
+
+    def _server_pipe_handle(self, first):
+        # Return a wrapper for a new pipe handle.
+        if self._address is None:
+            return None
+        flags = _winapi.PIPE_ACCESS_DUPLEX | _winapi.FILE_FLAG_OVERLAPPED
+        if first:
+            flags |= _winapi.FILE_FLAG_FIRST_PIPE_INSTANCE
+        h = _winapi.CreateNamedPipe(
+            self._address, flags,
+            _winapi.PIPE_TYPE_MESSAGE | _winapi.PIPE_READMODE_MESSAGE |
+            _winapi.PIPE_WAIT,
+            _winapi.PIPE_UNLIMITED_INSTANCES,
+            windows_utils.BUFSIZE, windows_utils.BUFSIZE,
+            _winapi.NMPWAIT_WAIT_FOREVER, _winapi.NULL)
+        pipe = windows_utils.PipeHandle(h)
+        self._free_instances.add(pipe)
+        return pipe
+
+    def close(self):
+        # Close all instances which have not been connected to by a client.
+        if self._address is not None:
+            for pipe in self._free_instances:
+                pipe.close()
+            self._pipe = None
+            self._address = None
+            self._free_instances.clear()
+
+    __del__ = close
+
+
+class SelectorEventLoop(selector_events.BaseSelectorEventLoop):
+    """Windows version of selector event loop."""
+
+    def _socketpair(self):
+        return windows_utils.socketpair()
+
+
+class ProactorEventLoop(proactor_events.BaseProactorEventLoop):
+    """Windows version of proactor event loop using IOCP."""
+
+    def __init__(self, proactor=None):
+        if proactor is None:
+            proactor = IocpProactor()
+        super().__init__(proactor)
+
+    def _socketpair(self):
+        return windows_utils.socketpair()
+
+    @tasks.coroutine
+    def create_pipe_connection(self, protocol_factory, address):
+        f = self._proactor.connect_pipe(address)
+        pipe = yield from f
+        protocol = protocol_factory()
+        trans = self._make_duplex_pipe_transport(pipe, protocol,
+                                                 extra={'addr': address})
+        return trans, protocol
+
+    @tasks.coroutine
+    def start_serving_pipe(self, protocol_factory, address):
+        server = PipeServer(address)
+        def loop(f=None):
+            pipe = None
+            try:
+                if f:
+                    pipe = f.result()
+                    server._free_instances.discard(pipe)
+                    protocol = protocol_factory()
+                    self._make_duplex_pipe_transport(
+                        pipe, protocol, extra={'addr': address})
+                pipe = server._get_unconnected_pipe()
+                if pipe is None:
+                    return
+                f = self._proactor.accept_pipe(pipe)
+            except OSError:
+                if pipe and pipe.fileno() != -1:
+                    asyncio_log.exception('Pipe accept failed')
+                    pipe.close()
+            except futures.CancelledError:
+                if pipe:
+                    pipe.close()
+            else:
+                f.add_done_callback(loop)
+        self.call_soon(loop)
+        return [server]
+
+    def _stop_serving(self, server):
+        server.close()
+
+
+class IocpProactor:
+    """Proactor implementation using IOCP."""
+
+    def __init__(self, concurrency=0xffffffff):
+        self._loop = None
+        self._results = []
+        self._iocp = _overlapped.CreateIoCompletionPort(
+            _overlapped.INVALID_HANDLE_VALUE, NULL, 0, concurrency)
+        self._cache = {}
+        self._registered = weakref.WeakSet()
+        self._stopped_serving = weakref.WeakSet()
+
+    def set_loop(self, loop):
+        self._loop = loop
+
+    def select(self, timeout=None):
+        if not self._results:
+            self._poll(timeout)
+        tmp = self._results
+        self._results = []
+        return tmp
+
+    def recv(self, conn, nbytes, flags=0):
+        self._register_with_iocp(conn)
+        ov = _overlapped.Overlapped(NULL)
+        if isinstance(conn, socket.socket):
+            ov.WSARecv(conn.fileno(), nbytes, flags)
+        else:
+            ov.ReadFile(conn.fileno(), nbytes)
+        def finish(trans, key, ov):
+            try:
+                return ov.getresult()
+            except OSError as exc:
+                if exc.winerror == _overlapped.ERROR_NETNAME_DELETED:
+                    raise ConnectionResetError(*exc.args)
+                else:
+                    raise
+        return self._register(ov, conn, finish)
+
+    def send(self, conn, buf, flags=0):
+        self._register_with_iocp(conn)
+        ov = _overlapped.Overlapped(NULL)
+        if isinstance(conn, socket.socket):
+            ov.WSASend(conn.fileno(), buf, flags)
+        else:
+            ov.WriteFile(conn.fileno(), buf)
+        def finish(trans, key, ov):
+            try:
+                return ov.getresult()
+            except OSError as exc:
+                if exc.winerror == _overlapped.ERROR_NETNAME_DELETED:
+                    raise ConnectionResetError(*exc.args)
+                else:
+                    raise
+        return self._register(ov, conn, finish)
+
+    def accept(self, listener):
+        self._register_with_iocp(listener)
+        conn = self._get_accept_socket(listener.family)
+        ov = _overlapped.Overlapped(NULL)
+        ov.AcceptEx(listener.fileno(), conn.fileno())
+        def finish_accept(trans, key, ov):
+            ov.getresult()
+            # Use SO_UPDATE_ACCEPT_CONTEXT so getsockname() etc work.
+            buf = struct.pack('@P', listener.fileno())
+            conn.setsockopt(socket.SOL_SOCKET,
+                            _overlapped.SO_UPDATE_ACCEPT_CONTEXT, buf)
+            conn.settimeout(listener.gettimeout())
+            return conn, conn.getpeername()
+        return self._register(ov, listener, finish_accept)
+
+    def connect(self, conn, address):
+        self._register_with_iocp(conn)
+        # The socket needs to be locally bound before we call ConnectEx().
+        try:
+            _overlapped.BindLocal(conn.fileno(), conn.family)
+        except OSError as e:
+            if e.winerror != errno.WSAEINVAL:
+                raise
+            # Probably already locally bound; check using getsockname().
+            if conn.getsockname()[1] == 0:
+                raise
+        ov = _overlapped.Overlapped(NULL)
+        ov.ConnectEx(conn.fileno(), address)
+        def finish_connect(trans, key, ov):
+            ov.getresult()
+            # Use SO_UPDATE_CONNECT_CONTEXT so getsockname() etc work.
+            conn.setsockopt(socket.SOL_SOCKET,
+                            _overlapped.SO_UPDATE_CONNECT_CONTEXT, 0)
+            return conn
+        return self._register(ov, conn, finish_connect)
+
+    def accept_pipe(self, pipe):
+        self._register_with_iocp(pipe)
+        ov = _overlapped.Overlapped(NULL)
+        ov.ConnectNamedPipe(pipe.fileno())
+        def finish(trans, key, ov):
+            ov.getresult()
+            return pipe
+        return self._register(ov, pipe, finish)
+
+    def connect_pipe(self, address):
+        ov = _overlapped.Overlapped(NULL)
+        ov.WaitNamedPipeAndConnect(address, self._iocp, ov.address)
+        def finish(err, handle, ov):
+            # err, handle were arguments passed to PostQueuedCompletionStatus()
+            # in a function run in a thread pool.
+            if err == _overlapped.ERROR_SEM_TIMEOUT:
+                # Connection did not succeed within time limit.
+                msg = _overlapped.FormatMessage(err)
+                raise ConnectionRefusedError(0, msg, None, err)
+            elif err != 0:
+                msg = _overlapped.FormatMessage(err)
+                raise OSError(0, msg, None, err)
+            else:
+                return windows_utils.PipeHandle(handle)
+        return self._register(ov, None, finish, wait_for_post=True)
+
+    def _register_with_iocp(self, obj):
+        # To get notifications of finished ops on this objects sent to the
+        # completion port, were must register the handle.
+        if obj not in self._registered:
+            self._registered.add(obj)
+            _overlapped.CreateIoCompletionPort(obj.fileno(), self._iocp, 0, 0)
+            # XXX We could also use SetFileCompletionNotificationModes()
+            # to avoid sending notifications to completion port of ops
+            # that succeed immediately.
+
+    def _register(self, ov, obj, callback, wait_for_post=False):
+        # Return a future which will be set with the result of the
+        # operation when it completes.  The future's value is actually
+        # the value returned by callback().
+        f = _OverlappedFuture(ov, loop=self._loop)
+        if ov.pending or wait_for_post:
+            # Register the overlapped operation for later.  Note that
+            # we only store obj to prevent it from being garbage
+            # collected too early.
+            self._cache[ov.address] = (f, ov, obj, callback)
+        else:
+            # The operation has completed, so no need to postpone the
+            # work.  We cannot take this short cut if we need the
+            # NumberOfBytes, CompletionKey values returned by
+            # PostQueuedCompletionStatus().
+            try:
+                value = callback(None, None, ov)
+            except OSError as e:
+                f.set_exception(e)
+            else:
+                f.set_result(value)
+        return f
+
+    def _get_accept_socket(self, family):
+        s = socket.socket(family)
+        s.settimeout(0)
+        return s
+
+    def _poll(self, timeout=None):
+        if timeout is None:
+            ms = INFINITE
+        elif timeout < 0:
+            raise ValueError("negative timeout")
+        else:
+            ms = int(timeout * 1000 + 0.5)
+            if ms >= INFINITE:
+                raise ValueError("timeout too big")
+        while True:
+            status = _overlapped.GetQueuedCompletionStatus(self._iocp, ms)
+            if status is None:
+                return
+            err, transferred, key, address = status
+            try:
+                f, ov, obj, callback = self._cache.pop(address)
+            except KeyError:
+                # key is either zero, or it is used to return a pipe
+                # handle which should be closed to avoid a leak.
+                if key not in (0, _overlapped.INVALID_HANDLE_VALUE):
+                    _winapi.CloseHandle(key)
+                ms = 0
+                continue
+            if obj in self._stopped_serving:
+                f.cancel()
+            elif not f.cancelled():
+                try:
+                    value = callback(transferred, key, ov)
+                except OSError as e:
+                    f.set_exception(e)
+                    self._results.append(f)
+                else:
+                    f.set_result(value)
+                    self._results.append(f)
+            ms = 0
+
+    def _stop_serving(self, obj):
+        # obj is a socket or pipe handle.  It will be closed in
+        # BaseProactorEventLoop._stop_serving() which will make any
+        # pending operations fail quickly.
+        self._stopped_serving.add(obj)
+
+    def close(self):
+        # Cancel remaining registered operations.
+        for address, (f, ov, obj, callback) in list(self._cache.items()):
+            if obj is None:
+                # The operation was started with connect_pipe() which
+                # queues a task to Windows' thread pool.  This cannot
+                # be cancelled, so just forget it.
+                del self._cache[address]
+            else:
+                try:
+                    ov.cancel()
+                except OSError:
+                    pass
+
+        while self._cache:
+            if not self._poll(1):
+                asyncio_log.debug('taking long time to close proactor')
+
+        self._results = []
+        if self._iocp is not None:
+            _winapi.CloseHandle(self._iocp)
+            self._iocp = None
diff --git a/Lib/asyncio/windows_utils.py b/Lib/asyncio/windows_utils.py
new file mode 100644
index 0000000..04b43e9
--- /dev/null
+++ b/Lib/asyncio/windows_utils.py
@@ -0,0 +1,181 @@
+"""
+Various Windows specific bits and pieces
+"""
+
+import sys
+
+if sys.platform != 'win32':  # pragma: no cover
+    raise ImportError('win32 only')
+
+import socket
+import itertools
+import msvcrt
+import os
+import subprocess
+import tempfile
+import _winapi
+
+
+__all__ = ['socketpair', 'pipe', 'Popen', 'PIPE', 'PipeHandle']
+
+#
+# Constants/globals
+#
+
+BUFSIZE = 8192
+PIPE = subprocess.PIPE
+_mmap_counter = itertools.count()
+
+#
+# Replacement for socket.socketpair()
+#
+
+def socketpair(family=socket.AF_INET, type=socket.SOCK_STREAM, proto=0):
+    """A socket pair usable as a self-pipe, for Windows.
+
+    Origin: https://gist.github.com/4325783, by Geert Jansen.  Public domain.
+    """
+    # We create a connected TCP socket. Note the trick with setblocking(0)
+    # that prevents us from having to create a thread.
+    lsock = socket.socket(family, type, proto)
+    lsock.bind(('localhost', 0))
+    lsock.listen(1)
+    addr, port = lsock.getsockname()
+    csock = socket.socket(family, type, proto)
+    csock.setblocking(False)
+    try:
+        csock.connect((addr, port))
+    except (BlockingIOError, InterruptedError):
+        pass
+    except Exception:
+        lsock.close()
+        csock.close()
+        raise
+    ssock, _ = lsock.accept()
+    csock.setblocking(True)
+    lsock.close()
+    return (ssock, csock)
+
+#
+# Replacement for os.pipe() using handles instead of fds
+#
+
+def pipe(*, duplex=False, overlapped=(True, True), bufsize=BUFSIZE):
+    """Like os.pipe() but with overlapped support and using handles not fds."""
+    address = tempfile.mktemp(prefix=r'\\.\pipe\python-pipe-%d-%d-' %
+                              (os.getpid(), next(_mmap_counter)))
+
+    if duplex:
+        openmode = _winapi.PIPE_ACCESS_DUPLEX
+        access = _winapi.GENERIC_READ | _winapi.GENERIC_WRITE
+        obsize, ibsize = bufsize, bufsize
+    else:
+        openmode = _winapi.PIPE_ACCESS_INBOUND
+        access = _winapi.GENERIC_WRITE
+        obsize, ibsize = 0, bufsize
+
+    openmode |= _winapi.FILE_FLAG_FIRST_PIPE_INSTANCE
+
+    if overlapped[0]:
+        openmode |= _winapi.FILE_FLAG_OVERLAPPED
+
+    if overlapped[1]:
+        flags_and_attribs = _winapi.FILE_FLAG_OVERLAPPED
+    else:
+        flags_and_attribs = 0
+
+    h1 = h2 = None
+    try:
+        h1 = _winapi.CreateNamedPipe(
+            address, openmode, _winapi.PIPE_WAIT,
+            1, obsize, ibsize, _winapi.NMPWAIT_WAIT_FOREVER, _winapi.NULL)
+
+        h2 = _winapi.CreateFile(
+            address, access, 0, _winapi.NULL, _winapi.OPEN_EXISTING,
+            flags_and_attribs, _winapi.NULL)
+
+        ov = _winapi.ConnectNamedPipe(h1, overlapped=True)
+        ov.GetOverlappedResult(True)
+        return h1, h2
+    except:
+        if h1 is not None:
+            _winapi.CloseHandle(h1)
+        if h2 is not None:
+            _winapi.CloseHandle(h2)
+        raise
+
+#
+# Wrapper for a pipe handle
+#
+
+class PipeHandle:
+    """Wrapper for an overlapped pipe handle which is vaguely file-object like.
+
+    The IOCP event loop can use these instead of socket objects.
+    """
+    def __init__(self, handle):
+        self._handle = handle
+
+    @property
+    def handle(self):
+        return self._handle
+
+    def fileno(self):
+        return self._handle
+
+    def close(self, *, CloseHandle=_winapi.CloseHandle):
+        if self._handle != -1:
+            CloseHandle(self._handle)
+            self._handle = -1
+
+    __del__ = close
+
+    def __enter__(self):
+        return self
+
+    def __exit__(self, t, v, tb):
+        self.close()
+
+#
+# Replacement for subprocess.Popen using overlapped pipe handles
+#
+
+class Popen(subprocess.Popen):
+    """Replacement for subprocess.Popen using overlapped pipe handles.
+
+    The stdin, stdout, stderr are None or instances of PipeHandle.
+    """
+    def __init__(self, args, stdin=None, stdout=None, stderr=None, **kwds):
+        stdin_rfd = stdout_wfd = stderr_wfd = None
+        stdin_wh = stdout_rh = stderr_rh = None
+        if stdin == PIPE:
+            stdin_rh, stdin_wh = pipe(overlapped=(False, True))
+            stdin_rfd = msvcrt.open_osfhandle(stdin_rh, os.O_RDONLY)
+        if stdout == PIPE:
+            stdout_rh, stdout_wh = pipe(overlapped=(True, False))
+            stdout_wfd = msvcrt.open_osfhandle(stdout_wh, 0)
+        if stderr == PIPE:
+            stderr_rh, stderr_wh = pipe(overlapped=(True, False))
+            stderr_wfd = msvcrt.open_osfhandle(stderr_wh, 0)
+        try:
+            super().__init__(args, bufsize=0, universal_newlines=False,
+                             stdin=stdin_rfd, stdout=stdout_wfd,
+                             stderr=stderr_wfd, **kwds)
+        except:
+            for h in (stdin_wh, stdout_rh, stderr_rh):
+                _winapi.CloseHandle(h)
+            raise
+        else:
+            if stdin_wh is not None:
+                self.stdin = PipeHandle(stdin_wh)
+            if stdout_rh is not None:
+                self.stdout = PipeHandle(stdout_rh)
+            if stderr_rh is not None:
+                self.stderr = PipeHandle(stderr_rh)
+        finally:
+            if stdin == PIPE:
+                os.close(stdin_rfd)
+            if stdout == PIPE:
+                os.close(stdout_wfd)
+            if stderr == PIPE:
+                os.close(stderr_wfd)