host interface + services

The separation of work in the p2p pkg is as follows:

- net implements the Swarm and connectivity
- protocol has muxer and header protocols
- host implements protocol muxing + services
- identify took over handshake completely! yay.
- p2p package works as a whole

p2p/host/basic/basic_host.go

+package basichost
+
+import (
+	context "github.com/jbenet/go-ipfs/Godeps/_workspace/src/code.google.com/p/go.net/context"
+
+	eventlog "github.com/jbenet/go-ipfs/util/eventlog"
+
+	inet "github.com/jbenet/go-ipfs/p2p/net2"
+	peer "github.com/jbenet/go-ipfs/p2p/peer"
+	protocol "github.com/jbenet/go-ipfs/p2p/protocol"
+	identify "github.com/jbenet/go-ipfs/p2p/protocol/identify"
+	relay "github.com/jbenet/go-ipfs/p2p/protocol/relay"
+)
+
+var log = eventlog.Logger("p2p/host/basic")
+
+type BasicHost struct {
+	network inet.Network
+	mux     protocol.Mux
+	ids     *identify.IDService
+	relay   *relay.RelayService
+}
+
+// New constructs and sets up a new *BasicHost with given Network
+func New(net inet.Network) *BasicHost {
+	h := &BasicHost{
+		network: net,
+		mux:     protocol.Mux{Handlers: protocol.StreamHandlerMap{}},
+	}
+
+	// setup host services
+	h.ids = identify.NewIDService(h)
+	h.relay = relay.NewRelayService(h, h.Mux().HandleSync)
+
+	net.SetConnHandler(h.newConnHandler)
+	net.SetStreamHandler(h.newStreamHandler)
+
+	return h
+}
+
+// newConnHandler is the remote-opened conn handler for inet.Network
+func (h *BasicHost) newConnHandler(c inet.Conn) {
+	h.ids.IdentifyConn(c)
+}
+
+// newStreamHandler is the remote-opened stream handler for inet.Network
+func (h *BasicHost) newStreamHandler(s inet.Stream) {
+	h.Mux().Handle(s)
+}
+
+// ID returns the (local) peer.ID associated with this Host
+func (h *BasicHost) ID() peer.ID {
+	return h.Network().LocalPeer()
+}
+
+// Peerstore returns the Host's repository of Peer Addresses and Keys.
+func (h *BasicHost) Peerstore() peer.Peerstore {
+	return h.Network().Peerstore()
+}
+
+// Networks returns the Network interface of the Host
+func (h *BasicHost) Network() inet.Network {
+	return h.network
+}
+
+// Mux returns the Mux multiplexing incoming streams to protocol handlers
+func (h *BasicHost) Mux() *protocol.Mux {
+	return &h.mux
+}
+
+func (h *BasicHost) IDService() *identify.IDService {
+	return h.ids
+}
+
+// SetStreamHandler sets the protocol handler on the Host's Mux.
+// This is equivalent to:
+//   host.Mux().SetHandler(proto, handler)
+// (Threadsafe)
+func (h *BasicHost) SetStreamHandler(pid protocol.ID, handler inet.StreamHandler) {
+	h.Mux().SetHandler(pid, handler)
+}
+
+// NewStream opens a new stream to given peer p, and writes a p2p/protocol
+// header with given protocol.ID. If there is no connection to p, attempts
+// to create one. If ProtocolID is "", writes no header.
+// (Threadsafe)
+func (h *BasicHost) NewStream(pid protocol.ID, p peer.ID) (inet.Stream, error) {
+	s, err := h.Network().NewStream(p)
+	if err != nil {
+		return nil, err
+	}
+
+	if err := protocol.WriteHeader(s, pid); err != nil {
+		s.Close()
+		return nil, err
+	}
+
+	return s, nil
+}
+
+// Connect ensures there is a connection between this host and the peer with
+// given peer.ID. Connect will absorb the addresses in pi into its internal
+// peerstore. If there is not an active connection, Connect will issue a
+// h.Network.Dial, and block until a connection is open, or an error is
+// returned. // TODO: Relay + NAT.
+func (h *BasicHost) Connect(ctx context.Context, pi peer.PeerInfo) error {
+
+	// absorb addresses into peerstore
+	h.Peerstore().AddPeerInfo(pi)
+
+	cs := h.Network().ConnsToPeer(pi.ID)
+	if len(cs) > 0 {
+		return nil
+	}
+
+	return h.dialPeer(ctx, pi.ID)
+}
+
+// dialPeer opens a connection to peer, and makes sure to identify
+// the connection once it has been opened.
+func (h *BasicHost) dialPeer(ctx context.Context, p peer.ID) error {
+	log.Debugf("host %s dialing %s", h.ID, p)
+	c, err := h.Network().DialPeer(ctx, p)
+	if err != nil {
+		return err
+	}
+
+	// identify the connection before returning.
+	done := make(chan struct{})
+	go func() {
+		h.ids.IdentifyConn(c)
+		close(done)
+	}()
+
+	// respect don contexteone
+	select {
+	case <-done:
+	case <-ctx.Done():
+		return ctx.Err()
+	}
+
+	log.Debugf("host %s finished dialing %s", h.ID, p)
+	return nil
+}
+
+// Close shuts down the Host's services (network, etc).
+func (h *BasicHost) Close() error {
+	return h.Network().Close()
+}

p2p/host/basic/basic_host_test.go

+package basichost_test
+
+import (
+	"bytes"
+	"io"
+	"testing"
+
+	inet "github.com/jbenet/go-ipfs/p2p/net2"
+	protocol "github.com/jbenet/go-ipfs/p2p/protocol"
+	testutil "github.com/jbenet/go-ipfs/p2p/test/util"
+
+	context "github.com/jbenet/go-ipfs/Godeps/_workspace/src/code.google.com/p/go.net/context"
+)
+
+func TestHostSimple(t *testing.T) {
+
+	ctx := context.Background()
+	h1 := testutil.GenHostSwarm(t, ctx)
+	h2 := testutil.GenHostSwarm(t, ctx)
+	defer h1.Close()
+	defer h2.Close()
+
+	h2pi := h2.Peerstore().PeerInfo(h2.ID())
+	if err := h1.Connect(ctx, h2pi); err != nil {
+		t.Fatal(err)
+	}
+
+	piper, pipew := io.Pipe()
+	h2.SetStreamHandler(protocol.TestingID, func(s inet.Stream) {
+		defer s.Close()
+		w := io.MultiWriter(s, pipew)
+		io.Copy(w, s) // mirror everything
+	})
+
+	s, err := h1.NewStream(protocol.TestingID, h2pi.ID)
+	if err != nil {
+		t.Fatal(err)
+	}
+
+	// write to the stream
+	buf1 := []byte("abcdefghijkl")
+	if _, err := s.Write(buf1); err != nil {
+		t.Fatal(err)
+	}
+
+	// get it from the stream (echoed)
+	buf2 := make([]byte, len(buf1))
+	if _, err := io.ReadFull(s, buf2); err != nil {
+		t.Fatal(err)
+	}
+	if !bytes.Equal(buf1, buf2) {
+		t.Fatal("buf1 != buf2 -- %x != %x", buf1, buf2)
+	}
+
+	// get it from the pipe (tee)
+	buf3 := make([]byte, len(buf1))
+	if _, err := io.ReadFull(piper, buf3); err != nil {
+		t.Fatal(err)
+	}
+	if !bytes.Equal(buf1, buf3) {
+		t.Fatal("buf1 != buf3 -- %x != %x", buf1, buf3)
+	}
+}

p2p/host/host.go

+package host
+
+import (
+	context "github.com/jbenet/go-ipfs/Godeps/_workspace/src/code.google.com/p/go.net/context"
+
+	eventlog "github.com/jbenet/go-ipfs/util/eventlog"
+
+	inet "github.com/jbenet/go-ipfs/p2p/net2"
+	peer "github.com/jbenet/go-ipfs/p2p/peer"
+	protocol "github.com/jbenet/go-ipfs/p2p/protocol"
+)
+
+var log = eventlog.Logger("p2p/host")
+
+// Host is an object participating in a p2p network, which
+// implements protocols or provides services. It handles
+// requests like a Server, and issues requests like a Client.
+// It is called Host because it is both Server and Client (and Peer
+// may be confusing).
+type Host interface {
+	// ID returns the (local) peer.ID associated with this Host
+	ID() peer.ID
+
+	// Peerstore returns the Host's repository of Peer Addresses and Keys.
+	Peerstore() peer.Peerstore
+
+	// Networks returns the Network interface of the Host
+	Network() inet.Network
+
+	// Mux returns the Mux multiplexing incoming streams to protocol handlers
+	Mux() *protocol.Mux
+
+	// Connect ensures there is a connection between this host and the peer with
+	// given peer.ID. Connect will absorb the addresses in pi into its internal
+	// peerstore. If there is not an active connection, Connect will issue a
+	// h.Network.Dial, and block until a connection is open, or an error is
+	// returned. // TODO: Relay + NAT.
+	Connect(ctx context.Context, pi peer.PeerInfo) error
+
+	// SetStreamHandler sets the protocol handler on the Host's Mux.
+	// This is equivalent to:
+	//   host.Mux().SetHandler(proto, handler)
+	// (Threadsafe)
+	SetStreamHandler(pid protocol.ID, handler inet.StreamHandler)
+
+	// NewStream opens a new stream to given peer p, and writes a p2p/protocol
+	// header with given protocol.ID. If there is no connection to p, attempts
+	// to create one. If ProtocolID is "", writes no header.
+	// (Threadsafe)
+	NewStream(pid protocol.ID, p peer.ID) (inet.Stream, error)
+
+	// Close shuts down the host, its Network, and services.
+	Close() error
+}

p2p/protocol/identify/id.go

+package identify
+
+import (
+	"fmt"
+	"sync"
+
+	context "github.com/jbenet/go-ipfs/Godeps/_workspace/src/code.google.com/p/go.net/context"
+	ggio "github.com/jbenet/go-ipfs/Godeps/_workspace/src/code.google.com/p/gogoprotobuf/io"
+	semver "github.com/jbenet/go-ipfs/Godeps/_workspace/src/github.com/coreos/go-semver/semver"
+	ma "github.com/jbenet/go-ipfs/Godeps/_workspace/src/github.com/jbenet/go-multiaddr"
+
+	config "github.com/jbenet/go-ipfs/config"
+	eventlog "github.com/jbenet/go-ipfs/util/eventlog"
+
+	host "github.com/jbenet/go-ipfs/p2p/host"
+	inet "github.com/jbenet/go-ipfs/p2p/net2"
+	protocol "github.com/jbenet/go-ipfs/p2p/protocol"
+
+	pb "github.com/jbenet/go-ipfs/p2p/protocol/identify/pb"
+)
+
+var log = eventlog.Logger("net/identify")
+
+// ID is the protocol.ID of the Identify Service.
+const ID protocol.ID = "/ipfs/identify"
+
+// IpfsVersion holds the current protocol version for a client running this code
+var IpfsVersion *semver.Version
+var ClientVersion = "go-ipfs/" + config.CurrentVersionNumber
+
+func init() {
+	var err error
+	IpfsVersion, err = semver.NewVersion("0.0.1")
+	if err != nil {
+		panic(fmt.Errorf("invalid protocol version: %v", err))
+	}
+}
+
+// IDService is a structure that implements ProtocolIdentify.
+// It is a trivial service that gives the other peer some
+// useful information about the local peer. A sort of hello.
+//
+// The IDService sends:
+//  * Our IPFS Protocol Version
+//  * Our IPFS Agent Version
+//  * Our public Listen Addresses
+type IDService struct {
+	Host host.Host
+
+	// connections undergoing identification
+	// for wait purposes
+	currid map[inet.Conn]chan struct{}
+	currmu sync.RWMutex
+}
+
+func NewIDService(h host.Host) *IDService {
+	s := &IDService{
+		Host:   h,
+		currid: make(map[inet.Conn]chan struct{}),
+	}
+	h.SetStreamHandler(ID, s.RequestHandler)
+	return s
+}
+
+func (ids *IDService) IdentifyConn(c inet.Conn) {
+	ids.currmu.Lock()
+	if wait, found := ids.currid[c]; found {
+		ids.currmu.Unlock()
+		log.Debugf("IdentifyConn called twice on: %s", c)
+		<-wait // already identifying it. wait for it.
+		return
+	}
+	ids.currid[c] = make(chan struct{})
+	ids.currmu.Unlock()
+
+	s, err := c.NewStream()
+	if err != nil {
+		log.Error("error opening initial stream for %s", ID)
+		log.Event(context.TODO(), "IdentifyOpenFailed", c.RemotePeer())
+	} else {
+
+		// ok give the response to our handler.
+		if err := protocol.WriteHeader(s, ID); err != nil {
+			log.Error("error writing stream header for %s", ID)
+			log.Event(context.TODO(), "IdentifyOpenFailed", c.RemotePeer())
+		}
+		ids.ResponseHandler(s)
+	}
+
+	ids.currmu.Lock()
+	ch, found := ids.currid[c]
+	delete(ids.currid, c)
+	ids.currmu.Unlock()
+
+	if !found {
+		log.Errorf("IdentifyConn failed to find channel (programmer error) for %s", c)
+		return
+	}
+
+	close(ch) // release everyone waiting.
+}
+
+func (ids *IDService) RequestHandler(s inet.Stream) {
+	defer s.Close()
+	c := s.Conn()
+
+	w := ggio.NewDelimitedWriter(s)
+	mes := pb.Identify{}
+	ids.populateMessage(&mes, s.Conn())
+	w.WriteMsg(&mes)
+
+	log.Debugf("%s sent message to %s %s", ID,
+		c.RemotePeer(), c.RemoteMultiaddr())
+}
+
+func (ids *IDService) ResponseHandler(s inet.Stream) {
+	defer s.Close()
+	c := s.Conn()
+
+	r := ggio.NewDelimitedReader(s, 2048)
+	mes := pb.Identify{}
+	if err := r.ReadMsg(&mes); err != nil {
+		log.Errorf("%s error receiving message from %s %s", ID,
+			c.RemotePeer(), c.RemoteMultiaddr())
+		return
+	}
+	ids.consumeMessage(&mes, c)
+
+	log.Debugf("%s received message from %s %s", ID,
+		c.RemotePeer(), c.RemoteMultiaddr())
+}
+
+func (ids *IDService) populateMessage(mes *pb.Identify, c inet.Conn) {
+
+	// set protocols this node is currently handling
+	protos := ids.Host.Mux().Protocols()
+	mes.Protocols = make([]string, len(protos))
+	for i, p := range protos {
+		mes.Protocols[i] = string(p)
+	}
+
+	// observed address so other side is informed of their
+	// "public" address, at least in relation to us.
+	mes.ObservedAddr = c.RemoteMultiaddr().Bytes()
+
+	// set listen addrs
+	laddrs, err := ids.Host.Network().InterfaceListenAddresses()
+	if err != nil {
+		log.Error(err)
+	} else {
+		mes.ListenAddrs = make([][]byte, len(laddrs))
+		for i, addr := range laddrs {
+			mes.ListenAddrs[i] = addr.Bytes()
+		}
+		log.Debugf("%s sent listen addrs to %s: %s", c.LocalPeer(), c.RemotePeer(), laddrs)
+	}
+
+	// set protocol versions
+	s := IpfsVersion.String()
+	mes.ProtocolVersion = &s
+	mes.AgentVersion = &ClientVersion
+}
+
+func (ids *IDService) consumeMessage(mes *pb.Identify, c inet.Conn) {
+	p := c.RemotePeer()
+
+	// mes.Protocols
+	// mes.ObservedAddr
+
+	// mes.ListenAddrs
+	laddrs := mes.GetListenAddrs()
+	lmaddrs := make([]ma.Multiaddr, 0, len(laddrs))
+	for _, addr := range laddrs {
+		maddr, err := ma.NewMultiaddrBytes(addr)
+		if err != nil {
+			log.Errorf("%s failed to parse multiaddr from %s %s", ID,
+				p, c.RemoteMultiaddr())
+			continue
+		}
+		lmaddrs = append(lmaddrs, maddr)
+	}
+
+	// update our peerstore with the addresses.
+	ids.Host.Peerstore().AddAddresses(p, lmaddrs)
+	log.Debugf("%s received listen addrs for %s: %s", c.LocalPeer(), c.RemotePeer(), lmaddrs)
+
+	// get protocol versions
+	pv := *mes.ProtocolVersion
+	av := *mes.AgentVersion
+	ids.Host.Peerstore().Put(p, "ProtocolVersion", pv)
+	ids.Host.Peerstore().Put(p, "AgentVersion", av)
+}
+
+// IdentifyWait returns a channel which will be closed once
+// "ProtocolIdentify" (handshake3) finishes on given conn.
+// This happens async so the connection can start to be used
+// even if handshake3 knowledge is not necesary.
+// Users **MUST** call IdentifyWait _after_ IdentifyConn
+func (ids *IDService) IdentifyWait(c inet.Conn) <-chan struct{} {
+	ids.currmu.Lock()
+	ch, found := ids.currid[c]
+	ids.currmu.Unlock()
+	if found {
+		return ch
+	}
+
+	// if not found, it means we are already done identifying it, or
+	// haven't even started. either way, return a new channel closed.
+	ch = make(chan struct{})
+	close(ch)
+	return ch
+}

p2p/protocol/identify/id_test.go

+package identify_test
+
+import (
+	"testing"
+	"time"
+
+	host "github.com/jbenet/go-ipfs/p2p/host"
+	peer "github.com/jbenet/go-ipfs/p2p/peer"
+	identify "github.com/jbenet/go-ipfs/p2p/protocol/identify"
+	testutil "github.com/jbenet/go-ipfs/p2p/test/util"
+
+	context "github.com/jbenet/go-ipfs/Godeps/_workspace/src/code.google.com/p/go.net/context"
+	ma "github.com/jbenet/go-ipfs/Godeps/_workspace/src/github.com/jbenet/go-multiaddr"
+)
+
+func subtestIDService(t *testing.T, postDialWait time.Duration) {
+
+	// the generated networks should have the id service wired in.
+	ctx := context.Background()
+	h1 := testutil.GenHostSwarm(t, ctx)
+	h2 := testutil.GenHostSwarm(t, ctx)
+
+	h1p := h1.ID()
+	h2p := h2.ID()
+
+	testKnowsAddrs(t, h1, h2p, []ma.Multiaddr{}) // nothing
+	testKnowsAddrs(t, h2, h1p, []ma.Multiaddr{}) // nothing
+
+	h2pi := h2.Peerstore().PeerInfo(h2p)
+	if err := h1.Connect(ctx, h2pi); err != nil {
+		t.Fatal(err)
+	}
+
+	// we need to wait here if Dial returns before ID service is finished.
+	if postDialWait > 0 {
+		<-time.After(postDialWait)
+	}
+
+	// the IDService should be opened automatically, by the network.
+	// what we should see now is that both peers know about each others listen addresses.
+	testKnowsAddrs(t, h1, h2p, h2.Peerstore().Addresses(h2p)) // has them
+	testHasProtocolVersions(t, h1, h2p)
+
+	// now, this wait we do have to do. it's the wait for the Listening side
+	// to be done identifying the connection.
+	c := h2.Network().ConnsToPeer(h1.ID())
+	if len(c) < 1 {
+		t.Fatal("should have connection by now at least.")
+	}
+	<-h2.IDService().IdentifyWait(c[0])
+
+	// and the protocol versions.
+	testKnowsAddrs(t, h2, h1p, h1.Peerstore().Addresses(h1p)) // has them
+	testHasProtocolVersions(t, h2, h1p)
+}
+
+func testKnowsAddrs(t *testing.T, h host.Host, p peer.ID, expected []ma.Multiaddr) {
+	actual := h.Peerstore().Addresses(p)
+
+	if len(actual) != len(expected) {
+		t.Error("dont have the same addresses")
+	}
+
+	have := map[string]struct{}{}
+	for _, addr := range actual {
+		have[addr.String()] = struct{}{}
+	}
+	for _, addr := range expected {
+		if _, found := have[addr.String()]; !found {
+			t.Errorf("%s did not have addr for %s: %s", h.ID(), p, addr)
+			// panic("ahhhhhhh")
+		}
+	}
+}
+
+func testHasProtocolVersions(t *testing.T, h host.Host, p peer.ID) {
+	v, err := h.Peerstore().Get(p, "ProtocolVersion")
+	if v == nil {
+		t.Error("no protocol version")
+		return
+	}
+	if v.(string) != identify.IpfsVersion.String() {
+		t.Error("protocol mismatch", err)
+	}
+	v, err = h.Peerstore().Get(p, "AgentVersion")
+	if v.(string) != identify.ClientVersion {
+		t.Error("agent version mismatch", err)
+	}
+}
+
+// TestIDServiceWait gives the ID service 100ms to finish after dialing
+// this is becasue it used to be concurrent. Now, Dial wait till the
+// id service is done.
+func TestIDServiceWait(t *testing.T) {
+	N := 3
+	for i := 0; i < N; i++ {
+		subtestIDService(t, 100*time.Millisecond)
+	}
+}
+
+func TestIDServiceNoWait(t *testing.T) {
+	N := 3
+	for i := 0; i < N; i++ {
+		subtestIDService(t, 0)
+	}
+}

p2p/protocol/identify/pb/Makefile

+
+PB = $(wildcard *.proto)
+GO = $(PB:.proto=.pb.go)
+
+all: $(GO)
+
+%.pb.go: %.proto
+	protoc --gogo_out=. --proto_path=../../../../../../:/usr/local/opt/protobuf/include:. $<
+
+clean:
+	rm *.pb.go

p2p/protocol/identify/pb/identify.pb.go

+// Code generated by protoc-gen-gogo.
+// source: identify.proto
+// DO NOT EDIT!
+
+/*
+Package identify_pb is a generated protocol buffer package.
+
+It is generated from these files:
+	identify.proto
+
+It has these top-level messages:
+	Identify
+*/
+package identify_pb
+
+import proto "code.google.com/p/gogoprotobuf/proto"
+import json "encoding/json"
+import math "math"
+
+// Reference proto, json, and math imports to suppress error if they are not otherwise used.
+var _ = proto.Marshal
+var _ = &json.SyntaxError{}
+var _ = math.Inf
+
+type Identify struct {
+	// protocolVersion determines compatibility between peers
+	ProtocolVersion *string `protobuf:"bytes,5,opt,name=protocolVersion" json:"protocolVersion,omitempty"`
+	// agentVersion is like a UserAgent string in browsers, or client version in bittorrent
+	// includes the client name and client.
+	AgentVersion *string `protobuf:"bytes,6,opt,name=agentVersion" json:"agentVersion,omitempty"`
+	// publicKey is this node's public key (which also gives its node.ID)
+	// - may not need to be sent, as secure channel implies it has been sent.
+	// - then again, if we change / disable secure channel, may still want it.
+	PublicKey []byte `protobuf:"bytes,1,opt,name=publicKey" json:"publicKey,omitempty"`
+	// listenAddrs are the multiaddrs the sender node listens for open connections on
+	ListenAddrs [][]byte `protobuf:"bytes,2,rep,name=listenAddrs" json:"listenAddrs,omitempty"`
+	// oservedAddr is the multiaddr of the remote endpoint that the sender node perceives
+	// this is useful information to convey to the other side, as it helps the remote endpoint
+	// determine whether its connection to the local peer goes through NAT.
+	ObservedAddr []byte `protobuf:"bytes,4,opt,name=observedAddr" json:"observedAddr,omitempty"`
+	// protocols are the services this node is running
+	Protocols        []string `protobuf:"bytes,3,rep,name=protocols" json:"protocols,omitempty"`
+	XXX_unrecognized []byte   `json:"-"`
+}
+
+func (m *Identify) Reset()         { *m = Identify{} }
+func (m *Identify) String() string { return proto.CompactTextString(m) }
+func (*Identify) ProtoMessage()    {}
+
+func (m *Identify) GetProtocolVersion() string {
+	if m != nil && m.ProtocolVersion != nil {
+		return *m.ProtocolVersion
+	}
+	return ""
+}
+
+func (m *Identify) GetAgentVersion() string {
+	if m != nil && m.AgentVersion != nil {
+		return *m.AgentVersion
+	}
+	return ""
+}
+
+func (m *Identify) GetPublicKey() []byte {
+	if m != nil {
+		return m.PublicKey
+	}
+	return nil
+}
+
+func (m *Identify) GetListenAddrs() [][]byte {
+	if m != nil {
+		return m.ListenAddrs
+	}
+	return nil
+}
+
+func (m *Identify) GetObservedAddr() []byte {
+	if m != nil {
+		return m.ObservedAddr
+	}
+	return nil
+}
+
+func (m *Identify) GetProtocols() []string {
+	if m != nil {
+		return m.Protocols
+	}
+	return nil
+}
+
+func init() {
+}

p2p/protocol/identify/pb/identify.proto

+package identify.pb;
+
+message Identify {
+
+  // protocolVersion determines compatibility between peers
+  optional string protocolVersion = 5; // e.g. ipfs/1.0.0
+
+  // agentVersion is like a UserAgent string in browsers, or client version in bittorrent
+  // includes the client name and client.
+  optional string agentVersion = 6; // e.g. go-ipfs/0.1.0
+
+  // publicKey is this node's public key (which also gives its node.ID)
+  // - may not need to be sent, as secure channel implies it has been sent.
+  // - then again, if we change / disable secure channel, may still want it.
+  optional bytes publicKey = 1;
+
+  // listenAddrs are the multiaddrs the sender node listens for open connections on
+  repeated bytes listenAddrs = 2;
+
+  // oservedAddr is the multiaddr of the remote endpoint that the sender node perceives
+  // this is useful information to convey to the other side, as it helps the remote endpoint
+  // determine whether its connection to the local peer goes through NAT.
+  optional bytes observedAddr = 4;
+
+  // protocols are the services this node is running
+  repeated string protocols = 3;
+}

p2p/protocol/relay/relay.go

+package relay
+
+import (
+	"fmt"
+	"io"
+
+	mh "github.com/jbenet/go-ipfs/Godeps/_workspace/src/github.com/jbenet/go-multihash"
+
+	host "github.com/jbenet/go-ipfs/p2p/host"
+	inet "github.com/jbenet/go-ipfs/p2p/net2"
+	peer "github.com/jbenet/go-ipfs/p2p/peer"
+	protocol "github.com/jbenet/go-ipfs/p2p/protocol"
+	eventlog "github.com/jbenet/go-ipfs/util/eventlog"
+)
+
+var log = eventlog.Logger("p2p/protocol/relay")
+
+// ID is the protocol.ID of the Relay Service.
+const ID protocol.ID = "/ipfs/relay"
+
+// Relay is a structure that implements ProtocolRelay.
+// It is a simple relay service which forwards traffic
+// between two directly connected peers.
+//
+// the protocol is very simple:
+//
+//   /ipfs/relay\n
+//   <multihash src id>
+//   <multihash dst id>
+//   <data stream>
+//
+type RelayService struct {
+	host    host.Host
+	handler inet.StreamHandler // for streams sent to us locally.
+}
+
+func NewRelayService(h host.Host, sh inet.StreamHandler) *RelayService {
+	s := &RelayService{
+		host:    h,
+		handler: sh,
+	}
+	h.SetStreamHandler(ID, s.requestHandler)
+	return s
+}
+
+// requestHandler is the function called by clients
+func (rs *RelayService) requestHandler(s inet.Stream) {
+	if err := rs.handleStream(s); err != nil {
+		log.Error("RelayService error:", err)
+	}
+}
+
+// handleStream is our own handler, which returns an error for simplicity.
+func (rs *RelayService) handleStream(s inet.Stream) error {
+	defer s.Close()
+
+	// read the header (src and dst peer.IDs)
+	src, dst, err := ReadHeader(s)
+	if err != nil {
+		return fmt.Errorf("stream with bad header: %s", err)
+	}
+
+	local := rs.host.ID()
+
+	switch {
+	case src == local:
+		return fmt.Errorf("relaying from self")
+	case dst == local: // it's for us! yaaay.
+		log.Debugf("%s consuming stream from %s", local, src)
+		return rs.consumeStream(s)
+	default: // src and dst are not local. relay it.
+		log.Debugf("%s relaying stream %s <--> %s", local, src, dst)
+		return rs.pipeStream(src, dst, s)
+	}
+}
+
+// consumeStream connects streams directed to the local peer
+// to our handler, with the header now stripped (read).
+func (rs *RelayService) consumeStream(s inet.Stream) error {
+	rs.handler(s) // boom.
+	return nil
+}
+
+// pipeStream relays over a stream to a remote peer. It's like `cat`
+func (rs *RelayService) pipeStream(src, dst peer.ID, s inet.Stream) error {
+	s2, err := rs.openStreamToPeer(dst)
+	if err != nil {
+		return fmt.Errorf("failed to open stream to peer: %s -- %s", dst, err)
+	}
+
+	if err := WriteHeader(s2, src, dst); err != nil {
+		return err
+	}
+
+	// connect the series of tubes.
+	done := make(chan retio, 2)
+	go func() {
+		n, err := io.Copy(s2, s)
+		done <- retio{n, err}
+	}()
+	go func() {
+		n, err := io.Copy(s, s2)
+		done <- retio{n, err}
+	}()
+
+	r1 := <-done
+	r2 := <-done
+	log.Infof("%s relayed %d/%d bytes between %s and %s", rs.host.ID(), r1.n, r2.n, src, dst)
+
+	if r1.err != nil {
+		return r1.err
+	}
+	return r2.err
+}
+
+// openStreamToPeer opens a pipe to a remote endpoint
+// for now, can only open streams to directly connected peers.
+// maybe we can do some routing later on.
+func (rs *RelayService) openStreamToPeer(p peer.ID) (inet.Stream, error) {
+	return rs.host.NewStream(ID, p)
+}
+
+func ReadHeader(r io.Reader) (src, dst peer.ID, err error) {
+
+	mhr := mh.NewReader(r)
+
+	s, err := mhr.ReadMultihash()
+	if err != nil {
+		return "", "", err
+	}
+
+	d, err := mhr.ReadMultihash()
+	if err != nil {
+		return "", "", err
+	}
+
+	return peer.ID(s), peer.ID(d), nil
+}
+
+func WriteHeader(w io.Writer, src, dst peer.ID) error {
+	// write header to w.
+	mhw := mh.NewWriter(w)
+	if err := mhw.WriteMultihash(mh.Multihash(src)); err != nil {
+		return fmt.Errorf("failed to write relay header: %s -- %s", dst, err)
+	}
+	if err := mhw.WriteMultihash(mh.Multihash(dst)); err != nil {
+		return fmt.Errorf("failed to write relay header: %s -- %s", dst, err)
+	}
+
+	return nil
+}
+
+type retio struct {
+	n   int64
+	err error
+}

p2p/protocol/relay/relay_test.go

+package relay_test
+
+import (
+	"io"
+	"testing"
+
+	inet "github.com/jbenet/go-ipfs/p2p/net2"
+	protocol "github.com/jbenet/go-ipfs/p2p/protocol"
+	relay "github.com/jbenet/go-ipfs/p2p/protocol/relay"
+	testutil "github.com/jbenet/go-ipfs/p2p/test/util"
+	eventlog "github.com/jbenet/go-ipfs/util/eventlog"
+
+	context "github.com/jbenet/go-ipfs/Godeps/_workspace/src/code.google.com/p/go.net/context"
+)
+
+var log = eventlog.Logger("relay_test")
+
+func TestRelaySimple(t *testing.T) {
+
+	ctx := context.Background()
+
+	// these networks have the relay service wired in already.
+	n1 := testutil.GenHostSwarm(t, ctx)
+	n2 := testutil.GenHostSwarm(t, ctx)
+	n3 := testutil.GenHostSwarm(t, ctx)
+
+	n1p := n1.ID()
+	n2p := n2.ID()
+	n3p := n3.ID()
+
+	n2pi := n2.Peerstore().PeerInfo(n2p)
+	if err := n1.Connect(ctx, n2pi); err != nil {
+		t.Fatal("Failed to connect:", err)
+	}
+	if err := n3.Connect(ctx, n2pi); err != nil {
+		t.Fatal("Failed to connect:", err)
+	}
+
+	// setup handler on n3 to copy everything over to the pipe.
+	piper, pipew := io.Pipe()
+	n3.SetStreamHandler(protocol.TestingID, func(s inet.Stream) {
+		log.Debug("relay stream opened to n3!")
+		log.Debug("piping and echoing everything")
+		w := io.MultiWriter(s, pipew)
+		io.Copy(w, s)
+		log.Debug("closing stream")
+		s.Close()
+	})
+
+	// ok, now we can try to relay n1--->n2--->n3.
+	log.Debug("open relay stream")
+	s, err := n1.NewStream(relay.ID, n2p)
+	if err != nil {
+		t.Fatal(err)
+	}
+
+	// ok first thing we write the relay header n1->n3
+	log.Debug("write relay header")
+	if err := relay.WriteHeader(s, n1p, n3p); err != nil {
+		t.Fatal(err)
+	}
+
+	// ok now the header's there, we can write the next protocol header.
+	log.Debug("write testing header")
+	if err := protocol.WriteHeader(s, protocol.TestingID); err != nil {
+		t.Fatal(err)
+	}
+
+	// okay, now we should be able to write text, and read it out.
+	buf1 := []byte("abcdefghij")
+	buf2 := make([]byte, 10)
+	buf3 := make([]byte, 10)
+	log.Debug("write in some text.")
+	if _, err := s.Write(buf1); err != nil {
+		t.Fatal(err)
+	}
+
+	// read it out from the pipe.
+	log.Debug("read it out from the pipe.")
+	if _, err := io.ReadFull(piper, buf2); err != nil {
+		t.Fatal(err)
+	}
+	if string(buf1) != string(buf2) {
+		t.Fatal("should've gotten that text out of the pipe")
+	}
+
+	// read it out from the stream (echoed)
+	log.Debug("read it out from the stream (echoed).")
+	if _, err := io.ReadFull(s, buf3); err != nil {
+		t.Fatal(err)
+	}
+	if string(buf1) != string(buf3) {
+		t.Fatal("should've gotten that text out of the stream")
+	}
+
+	// sweet. relay works.
+	log.Debug("sweet, relay works.")
+	s.Close()
+}
+
+func TestRelayAcrossFour(t *testing.T) {
+
+	ctx := context.Background()
+
+	// these networks have the relay service wired in already.
+	n1 := testutil.GenHostSwarm(t, ctx)
+	n2 := testutil.GenHostSwarm(t, ctx)
+	n3 := testutil.GenHostSwarm(t, ctx)
+	n4 := testutil.GenHostSwarm(t, ctx)
+	n5 := testutil.GenHostSwarm(t, ctx)
+
+	n1p := n1.ID()
+	n2p := n2.ID()
+	n3p := n3.ID()
+	n4p := n4.ID()
+	n5p := n5.ID()
+
+	n2pi := n2.Peerstore().PeerInfo(n2p)
+	n4pi := n4.Peerstore().PeerInfo(n4p)
+
+	if err := n1.Connect(ctx, n2pi); err != nil {
+		t.Fatalf("Failed to dial:", err)
+	}
+	if err := n3.Connect(ctx, n2pi); err != nil {
+		t.Fatalf("Failed to dial:", err)
+	}
+	if err := n3.Connect(ctx, n4pi); err != nil {
+		t.Fatalf("Failed to dial:", err)
+	}
+	if err := n5.Connect(ctx, n4pi); err != nil {
+		t.Fatalf("Failed to dial:", err)
+	}
+
+	// setup handler on n5 to copy everything over to the pipe.
+	piper, pipew := io.Pipe()
+	n5.SetStreamHandler(protocol.TestingID, func(s inet.Stream) {
+		log.Debug("relay stream opened to n5!")
+		log.Debug("piping and echoing everything")
+		w := io.MultiWriter(s, pipew)
+		io.Copy(w, s)
+		log.Debug("closing stream")
+		s.Close()
+	})
+
+	// ok, now we can try to relay n1--->n2--->n3--->n4--->n5
+	log.Debug("open relay stream")
+	s, err := n1.NewStream(relay.ID, n2p)
+	if err != nil {
+		t.Fatal(err)
+	}
+
+	log.Debugf("write relay header n1->n3 (%s -> %s)", n1p, n3p)
+	if err := relay.WriteHeader(s, n1p, n3p); err != nil {
+		t.Fatal(err)
+	}
+
+	log.Debugf("write relay header n1->n4 (%s -> %s)", n1p, n4p)
+	if err := protocol.WriteHeader(s, relay.ID); err != nil {
+		t.Fatal(err)
+	}
+	if err := relay.WriteHeader(s, n1p, n4p); err != nil {
+		t.Fatal(err)
+	}
+
+	log.Debugf("write relay header n1->n5 (%s -> %s)", n1p, n5p)
+	if err := protocol.WriteHeader(s, relay.ID); err != nil {
+		t.Fatal(err)
+	}
+	if err := relay.WriteHeader(s, n1p, n5p); err != nil {
+		t.Fatal(err)
+	}
+
+	// ok now the header's there, we can write the next protocol header.
+	log.Debug("write testing header")
+	if err := protocol.WriteHeader(s, protocol.TestingID); err != nil {
+		t.Fatal(err)
+	}
+
+	// okay, now we should be able to write text, and read it out.
+	buf1 := []byte("abcdefghij")
+	buf2 := make([]byte, 10)
+	buf3 := make([]byte, 10)
+	log.Debug("write in some text.")
+	if _, err := s.Write(buf1); err != nil {
+		t.Fatal(err)
+	}
+
+	// read it out from the pipe.
+	log.Debug("read it out from the pipe.")
+	if _, err := io.ReadFull(piper, buf2); err != nil {
+		t.Fatal(err)
+	}
+	if string(buf1) != string(buf2) {
+		t.Fatal("should've gotten that text out of the pipe")
+	}
+
+	// read it out from the stream (echoed)
+	log.Debug("read it out from the stream (echoed).")
+	if _, err := io.ReadFull(s, buf3); err != nil {
+		t.Fatal(err)
+	}
+	if string(buf1) != string(buf3) {
+		t.Fatal("should've gotten that text out of the stream")
+	}
+
+	// sweet. relay works.
+	log.Debug("sweet, relaying across 4 works.")
+	s.Close()
+}
+
+func TestRelayStress(t *testing.T) {
+	buflen := 1 << 18
+	iterations := 10
+
+	ctx := context.Background()
+
+	// these networks have the relay service wired in already.
+	n1 := testutil.GenHostSwarm(t, ctx)
+	n2 := testutil.GenHostSwarm(t, ctx)
+	n3 := testutil.GenHostSwarm(t, ctx)
+
+	n1p := n1.ID()
+	n2p := n2.ID()
+	n3p := n3.ID()
+
+	n2pi := n2.Peerstore().PeerInfo(n2p)
+	if err := n1.Connect(ctx, n2pi); err != nil {
+		t.Fatalf("Failed to dial:", err)
+	}
+	if err := n3.Connect(ctx, n2pi); err != nil {
+		t.Fatalf("Failed to dial:", err)
+	}
+
+	// setup handler on n3 to copy everything over to the pipe.
+	piper, pipew := io.Pipe()
+	n3.SetStreamHandler(protocol.TestingID, func(s inet.Stream) {
+		log.Debug("relay stream opened to n3!")
+		log.Debug("piping and echoing everything")
+		w := io.MultiWriter(s, pipew)
+		io.Copy(w, s)
+		log.Debug("closing stream")
+		s.Close()
+	})
+
+	// ok, now we can try to relay n1--->n2--->n3.
+	log.Debug("open relay stream")
+	s, err := n1.NewStream(relay.ID, n2p)
+	if err != nil {
+		t.Fatal(err)
+	}
+
+	// ok first thing we write the relay header n1->n3
+	log.Debug("write relay header")
+	if err := relay.WriteHeader(s, n1p, n3p); err != nil {
+		t.Fatal(err)
+	}
+
+	// ok now the header's there, we can write the next protocol header.
+	log.Debug("write testing header")
+	if err := protocol.WriteHeader(s, protocol.TestingID); err != nil {
+		t.Fatal(err)
+	}
+
+	// okay, now write lots of text and read it back out from both
+	// the pipe and the stream.
+	buf1 := make([]byte, buflen)
+	buf2 := make([]byte, len(buf1))
+	buf3 := make([]byte, len(buf1))
+
+	fillbuf := func(buf []byte, b byte) {
+		for i := range buf {
+			buf[i] = b
+		}
+	}
+
+	for i := 0; i < iterations; i++ {
+		fillbuf(buf1, byte(int('a')+i))
+		log.Debugf("writing %d bytes (%d/%d)", len(buf1), i, iterations)
+		if _, err := s.Write(buf1); err != nil {
+			t.Fatal(err)
+		}
+
+		log.Debug("read it out from the pipe.")
+		if _, err := io.ReadFull(piper, buf2); err != nil {
+			t.Fatal(err)
+		}
+		if string(buf1) != string(buf2) {
+			t.Fatal("should've gotten that text out of the pipe")
+		}
+
+		// read it out from the stream (echoed)
+		log.Debug("read it out from the stream (echoed).")
+		if _, err := io.ReadFull(s, buf3); err != nil {
+			t.Fatal(err)
+		}
+		if string(buf1) != string(buf3) {
+			t.Fatal("should've gotten that text out of the stream")
+		}
+	}
+
+	log.Debug("sweet, relay works under stress.")
+	s.Close()
+}

p2p/test/backpressure/backpressure.go

+package backpressure_tests

p2p/test/backpressure/backpressure_test.go

+package backpressure_tests
+
+import (
+	crand "crypto/rand"
+	"io"
+	"math/rand"
+	"testing"
+	"time"
+
+	host "github.com/jbenet/go-ipfs/p2p/host"
+	inet "github.com/jbenet/go-ipfs/p2p/net2"
+	peer "github.com/jbenet/go-ipfs/p2p/peer"
+	protocol "github.com/jbenet/go-ipfs/p2p/protocol"
+	testutil "github.com/jbenet/go-ipfs/p2p/test/util"
+	eventlog "github.com/jbenet/go-ipfs/util/eventlog"
+
+	context "github.com/jbenet/go-ipfs/Godeps/_workspace/src/code.google.com/p/go.net/context"
+)
+
+var log = eventlog.Logger("backpressure")
+
+// TestBackpressureStreamHandler tests whether mux handler
+// ratelimiting works. Meaning, since the handler is sequential
+// it should block senders.
+//
+// Important note: spdystream (which peerstream uses) has a set
+// of n workers (n=spdsystream.FRAME_WORKERS) which handle new
+// frames, including those starting new streams. So all of them
+// can be in the handler at one time. Also, the sending side
+// does not rate limit unless we call stream.Wait()
+//
+//
+// Note: right now, this happens muxer-wide. the muxer should
+// learn to flow control, so handlers cant block each other.
+func TestBackpressureStreamHandler(t *testing.T) {
+	t.Skip(`Sadly, as cool as this test is, it doesn't work
+Because spdystream doesnt handle stream open backpressure
+well IMO. I'll see about rewriting that part when it becomes
+a problem.
+`)
+
+	// a number of concurrent request handlers
+	limit := 10
+
+	// our way to signal that we're done with 1 request
+	requestHandled := make(chan struct{})
+
+	// handler rate limiting
+	receiverRatelimit := make(chan struct{}, limit)
+	for i := 0; i < limit; i++ {
+		receiverRatelimit <- struct{}{}
+	}
+
+	// sender counter of successfully opened streams
+	senderOpened := make(chan struct{}, limit*100)
+
+	// sender signals it's done (errored out)
+	senderDone := make(chan struct{})
+
+	// the receiver handles requests with some rate limiting
+	receiver := func(s inet.Stream) {
+		log.Debug("receiver received a stream")
+
+		<-receiverRatelimit // acquire
+		go func() {
+			// our request handler. can do stuff here. we
+			// simulate something taking time by waiting
+			// on requestHandled
+			log.Error("request worker handling...")
+			<-requestHandled
+			log.Error("request worker done!")
+			receiverRatelimit <- struct{}{} // release
+		}()
+	}
+
+	// the sender opens streams as fast as possible
+	sender := func(host host.Host, remote peer.ID) {
+		var s inet.Stream
+		var err error
+		defer func() {
+			t.Error(err)
+			log.Debug("sender error. exiting.")
+			senderDone <- struct{}{}
+		}()
+
+		for {
+			s, err = host.NewStream(protocol.TestingID, remote)
+			if err != nil {
+				return
+			}
+
+			_ = s
+			// if err = s.SwarmStream().Stream().Wait(); err != nil {
+			// 	return
+			// }
+
+			// "count" another successfully opened stream
+			// (large buffer so shouldn't block in normal operation)
+			log.Debug("sender opened another stream!")
+			senderOpened <- struct{}{}
+		}
+	}
+
+	// count our senderOpened events
+	countStreamsOpenedBySender := func(min int) int {
+		opened := 0
+		for opened < min {
+			log.Debugf("countStreamsOpenedBySender got %d (min %d)", opened, min)
+			select {
+			case <-senderOpened:
+				opened++
+			case <-time.After(10 * time.Millisecond):
+			}
+		}
+		return opened
+	}
+
+	// count our received events
+	// waitForNReceivedStreams := func(n int) {
+	// 	for n > 0 {
+	// 		log.Debugf("waiting for %d received streams...", n)
+	// 		select {
+	// 		case <-receiverRatelimit:
+	// 			n--
+	// 		}
+	// 	}
+	// }
+
+	testStreamsOpened := func(expected int) {
+		log.Debugf("testing rate limited to %d streams", expected)
+		if n := countStreamsOpenedBySender(expected); n != expected {
+			t.Fatalf("rate limiting did not work :( -- %d != %d", expected, n)
+		}
+	}
+
+	// ok that's enough setup. let's do it!
+
+	ctx := context.Background()
+	h1 := testutil.GenHostSwarm(t, ctx)
+	h2 := testutil.GenHostSwarm(t, ctx)
+
+	// setup receiver handler
+	h1.SetStreamHandler(protocol.TestingID, receiver)
+
+	h2pi := h2.Peerstore().PeerInfo(h2.ID())
+	log.Debugf("dialing %s", h2pi.Addrs)
+	if err := h1.Connect(ctx, h2pi); err != nil {
+		t.Fatalf("Failed to connect:", err)
+	}
+
+	// launch sender!
+	go sender(h2, h1.ID())
+
+	// ok, what do we expect to happen? the receiver should
+	// receive 10 requests and stop receiving, blocking the sender.
+	// we can test this by counting 10x senderOpened requests
+
+	<-senderOpened // wait for the sender to successfully open some.
+	testStreamsOpened(limit - 1)
+
+	// let's "handle" 3 requests.
+	<-requestHandled
+	<-requestHandled
+	<-requestHandled
+	// the sender should've now been able to open exactly 3 more.
+
+	testStreamsOpened(3)
+
+	// shouldn't have opened anything more
+	testStreamsOpened(0)
+
+	// let's "handle" 100 requests in batches of 5
+	for i := 0; i < 20; i++ {
+		<-requestHandled
+		<-requestHandled
+		<-requestHandled
+		<-requestHandled
+		<-requestHandled
+		testStreamsOpened(5)
+	}
+
+	// success!
+
+	// now for the sugar on top: let's tear down the receiver. it should
+	// exit the sender.
+	h1.Close()
+
+	// shouldn't have opened anything more
+	testStreamsOpened(0)
+
+	select {
+	case <-time.After(100 * time.Millisecond):
+		t.Error("receiver shutdown failed to exit sender")
+	case <-senderDone:
+		log.Info("handler backpressure works!")
+	}
+}
+
+// TestStBackpressureStreamWrite tests whether streams see proper
+// backpressure when writing data over the network streams.
+func TestStBackpressureStreamWrite(t *testing.T) {
+
+	// senderWrote signals that the sender wrote bytes to remote.
+	// the value is the count of bytes written.
+	senderWrote := make(chan int, 10000)
+
+	// sender signals it's done (errored out)
+	senderDone := make(chan struct{})
+
+	// writeStats lets us listen to all the writes and return
+	// how many happened and how much was written
+	writeStats := func() (int, int) {
+		writes := 0
+		bytes := 0
+		for {
+			select {
+			case n := <-senderWrote:
+				writes++
+				bytes = bytes + n
+			default:
+				log.Debugf("stats: sender wrote %d bytes, %d writes", bytes, writes)
+				return bytes, writes
+			}
+		}
+	}
+
+	// sender attempts to write as fast as possible, signaling on the
+	// completion of every write. This makes it possible to see how
+	// fast it's actually writing. We pair this with a receiver
+	// that waits for a signal to read.
+	sender := func(s inet.Stream) {
+		defer func() {
+			s.Close()
+			senderDone <- struct{}{}
+		}()
+
+		// ready a buffer of random data
+		buf := make([]byte, 65536)
+		crand.Read(buf)
+
+		for {
+			// send a randomly sized subchunk
+			from := rand.Intn(len(buf) / 2)
+			to := rand.Intn(len(buf) / 2)
+			sendbuf := buf[from : from+to]
+
+			n, err := s.Write(sendbuf)
+			if err != nil {
+				log.Debug("sender error. exiting:", err)
+				return
+			}
+
+			log.Debugf("sender wrote %d bytes", n)
+			senderWrote <- n
+		}
+	}
+
+	// receive a number of bytes from a stream.
+	// returns the number of bytes written.
+	receive := func(s inet.Stream, expect int) {
+		log.Debugf("receiver to read %d bytes", expect)
+		rbuf := make([]byte, expect)
+		n, err := io.ReadFull(s, rbuf)
+		if err != nil {
+			t.Error("read failed:", err)
+		}
+		if expect != n {
+			t.Error("read len differs: %d != %d", expect, n)
+		}
+	}
+
+	// ok let's do it!
+
+	// setup the networks
+	ctx := context.Background()
+	h1 := testutil.GenHostSwarm(t, ctx)
+	h2 := testutil.GenHostSwarm(t, ctx)
+
+	// setup sender handler on 1
+	h1.SetStreamHandler(protocol.TestingID, sender)
+
+	h2pi := h2.Peerstore().PeerInfo(h2.ID())
+	log.Debugf("dialing %s", h2pi.Addrs)
+	if err := h1.Connect(ctx, h2pi); err != nil {
+		t.Fatalf("Failed to connect:", err)
+	}
+
+	// open a stream, from 2->1, this is our reader
+	s, err := h2.NewStream(protocol.TestingID, h1.ID())
+	if err != nil {
+		t.Fatal(err)
+	}
+
+	// let's make sure r/w works.
+	testSenderWrote := func(bytesE int) {
+		bytesA, writesA := writeStats()
+		if bytesA != bytesE {
+			t.Errorf("numbers failed: %d =?= %d bytes, via %d writes", bytesA, bytesE, writesA)
+		}
+	}
+
+	// 500ms rounds of lockstep write + drain
+	roundsStart := time.Now()
+	roundsTotal := 0
+	for roundsTotal < (2 << 20) {
+		// let the sender fill its buffers, it will stop sending.
+		<-time.After(300 * time.Millisecond)
+		b, _ := writeStats()
+		testSenderWrote(0)
+		testSenderWrote(0)
+
+		// drain it all, wait again
+		receive(s, b)
+		roundsTotal = roundsTotal + b
+	}
+	roundsTime := time.Now().Sub(roundsStart)
+
+	// now read continously, while we measure stats.
+	stop := make(chan struct{})
+	contStart := time.Now()
+
+	go func() {
+		for {
+			select {
+			case <-stop:
+				return
+			default:
+				receive(s, 2<<15)
+			}
+		}
+	}()
+
+	contTotal := 0
+	for contTotal < (2 << 20) {
+		n := <-senderWrote
+		contTotal += n
+	}
+	stop <- struct{}{}
+	contTime := time.Now().Sub(contStart)
+
+	// now compare! continuous should've been faster AND larger
+	if roundsTime < contTime {
+		t.Error("continuous should have been faster")
+	}
+
+	if roundsTotal < contTotal {
+		t.Error("continuous should have been larger, too!")
+	}
+
+	// and a couple rounds more for good measure ;)
+	for i := 0; i < 3; i++ {
+		// let the sender fill its buffers, it will stop sending.
+		<-time.After(300 * time.Millisecond)
+		b, _ := writeStats()
+		testSenderWrote(0)
+		testSenderWrote(0)
+
+		// drain it all, wait again
+		receive(s, b)
+	}
+
+	// this doesn't work :(:
+	// // now for the sugar on top: let's tear down the receiver. it should
+	// // exit the sender.
+	// n1.Close()
+	// testSenderWrote(0)
+	// testSenderWrote(0)
+	// select {
+	// case <-time.After(2 * time.Second):
+	// 	t.Error("receiver shutdown failed to exit sender")
+	// case <-senderDone:
+	// 	log.Info("handler backpressure works!")
+	// }
+}

p2p/test/util/util.go

+package testutil
+
+import (
+	"testing"
+
+	bhost "github.com/jbenet/go-ipfs/p2p/host/basic"
+	inet "github.com/jbenet/go-ipfs/p2p/net2"
+	swarm "github.com/jbenet/go-ipfs/p2p/net2/swarm"
+	peer "github.com/jbenet/go-ipfs/p2p/peer"
+	tu "github.com/jbenet/go-ipfs/util/testutil"
+
+	context "github.com/jbenet/go-ipfs/Godeps/_workspace/src/code.google.com/p/go.net/context"
+)
+
+func GenSwarmNetwork(t *testing.T, ctx context.Context) *swarm.Network {
+	p := tu.RandPeerNetParamsOrFatal(t)
+	ps := peer.NewPeerstore()
+	ps.AddAddress(p.ID, p.Addr)
+	ps.AddPubKey(p.ID, p.PubKey)
+	ps.AddPrivKey(p.ID, p.PrivKey)
+	n, err := swarm.NewNetwork(ctx, ps.Addresses(p.ID), p.ID, ps)
+	if err != nil {
+		t.Fatal(err)
+	}
+	return n
+}
+
+func DivulgeAddresses(a, b inet.Network) {
+	id := a.LocalPeer()
+	addrs := a.Peerstore().Addresses(id)
+	b.Peerstore().AddAddresses(id, addrs)
+}
+
+func GenHostSwarm(t *testing.T, ctx context.Context) *bhost.BasicHost {
+	n := GenSwarmNetwork(t, ctx)
+	return bhost.New(n)
+}