path: root/agent/pseudotcp.c

/*
 * This file is part of the Nice GLib ICE library.
 *
 * (C) 2010, 2014 Collabora Ltd.
 *  Contact: Philip Withnall

 *
 * The contents of this file are subject to the Mozilla Public License Version
 * 1.1 (the "License"); you may not use this file except in compliance with
 * the License. You may obtain a copy of the License at
 * http://www.mozilla.org/MPL/
 *
 * Software distributed under the License is distributed on an "AS IS" basis,
 * WITHOUT WARRANTY OF ANY KIND, either express or implied. See the License
 * for the specific language governing rights and limitations under the
 * License.
 *
 * The Original Code is the Nice GLib ICE library.
 *
 * The Initial Developers of the Original Code are Collabora Ltd and Nokia
 * Corporation. All Rights Reserved.
 *
 * Contributors:
 *   Youness Alaoui, Collabora Ltd.
 *   Philip Withnall, Collabora Ltd.
 *
 * Alternatively, the contents of this file may be used under the terms of the
 * the GNU Lesser General Public License Version 2.1 (the "LGPL"), in which
 * case the provisions of LGPL are applicable instead of those above. If you
 * wish to allow use of your version of this file only under the terms of the
 * LGPL and not to allow others to use your version of this file under the
 * MPL, indicate your decision by deleting the provisions above and replace
 * them with the notice and other provisions required by the LGPL. If you do
 * not delete the provisions above, a recipient may use your version of this
 * file under either the MPL or the LGPL.
 */

/* Reproducing license from libjingle for copied code */

/*
 * libjingle
 * Copyright 2004--2005, Google Inc.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions are met:
 *
 *  1. Redistributions of source code must retain the above copyright notice,
 *     this list of conditions and the following disclaimer.
 *  2. Redistributions in binary form must reproduce the above copyright notice,
 *     this list of conditions and the following disclaimer in the documentation
 *     and/or other materials provided with the distribution.
 *  3. The name of the author may not be used to endorse or promote products
 *     derived from this software without specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR IMPLIED
 * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
 * MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO
 * EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
 * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
 * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
 * OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
 * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
 * OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
 * ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 */

#include <stdlib.h>
#include <errno.h>
#include <string.h>

#include <glib.h>

#ifndef G_OS_WIN32
#  include <arpa/inet.h>
#endif

#include "pseudotcp.h"
#include "agent-priv.h"

struct _PseudoTcpSocketClass {
    GIOStreamClass parent_class;
};

typedef struct _PseudoTcpSocketPrivate PseudoTcpSocketPrivate;


struct _PseudoTcpSocket {
    GIOStream parent;
    PseudoTcpSocketPrivate *priv;
};

G_DEFINE_TYPE (PseudoTcpSocket, pseudo_tcp_socket, G_TYPE_OBJECT);

//////////////////////////////////////////////////////////////////////
// Network Constants
//////////////////////////////////////////////////////////////////////

// Standard MTUs
const guint16 PACKET_MAXIMUMS[] = {
  65535,    // Theoretical maximum, Hyperchannel
  32000,    // Nothing
  17914,    // 16Mb IBM Token Ring
  8166,   // IEEE 802.4
  //4464,   // IEEE 802.5 (4Mb max)
  4352,   // FDDI
  //2048,   // Wideband Network
  2002,   // IEEE 802.5 (4Mb recommended)
  //1536,   // Expermental Ethernet Networks
  //1500,   // Ethernet, Point-to-Point (default)
  1492,   // IEEE 802.3
  1006,   // SLIP, ARPANET
  //576,    // X.25 Networks
  //544,    // DEC IP Portal
  //512,    // NETBIOS
  508,    // IEEE 802/Source-Rt Bridge, ARCNET
  296,    // Point-to-Point (low delay)
  //68,     // Official minimum
  0,      // End of list marker
};

// FIXME: This is a reasonable MTU, but we should get it from the lower layer
#define DEF_MTU 1400
#define MAX_PACKET 65532
// Note: we removed lowest level because packet overhead was larger!
#define MIN_PACKET 296

// (+ up to 40 bytes of options?)
#define IP_HEADER_SIZE 20
#define ICMP_HEADER_SIZE 8
#define UDP_HEADER_SIZE 8
// TODO: Make JINGLE_HEADER_SIZE transparent to this code?
// when relay framing is in use
#define JINGLE_HEADER_SIZE 64

//////////////////////////////////////////////////////////////////////
// Global Constants and Functions
//////////////////////////////////////////////////////////////////////
//
//    0                   1                   2                   3
//    0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
//    +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
//  0 |                      Conversation Number                      |
//    +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
//  4 |                        Sequence Number                        |
//    +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
//  8 |                     Acknowledgment Number                     |
//    +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
//    |               |   |U|A|P|R|S|F|                               |
// 12 |    Control    |   |R|C|S|S|Y|I|            Window             |
//    |               |   |G|K|H|T|N|N|                               |
//    +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
// 16 |                       Timestamp sending                       |
//    +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
// 20 |                      Timestamp receiving                      |
//    +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
// 24 |                             data                              |
//    +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
//
//////////////////////////////////////////////////////////////////////

#define MAX_SEQ 0xFFFFFFFF
#define HEADER_SIZE 24

#define PACKET_OVERHEAD (HEADER_SIZE + UDP_HEADER_SIZE + \
      IP_HEADER_SIZE + JINGLE_HEADER_SIZE)

// MIN_RTO = 250 ms (RFC1122, Sec 4.2.3.1 "fractions of a second")
#define MIN_RTO      250
#define DEF_RTO     1000 /* 1 seconds (RFC 6298 sect 2.1) */
#define MAX_RTO    60000 /* 60 seconds */
#define DEFAULT_ACK_DELAY    100 /* 100 milliseconds */
#define DEFAULT_NO_DELAY     FALSE

#define DEFAULT_RCV_BUF_SIZE (60 * 1024)
#define DEFAULT_SND_BUF_SIZE (90 * 1024)

/* NOTE: This must fit in 8 bits. This is used on the wire. */
typedef enum {
  /* Google-provided options: */
  TCP_OPT_EOL = 0,  /* end of list */
  TCP_OPT_NOOP = 1,  /* no-op */
  TCP_OPT_MSS = 2,  /* maximum segment size */
  TCP_OPT_WND_SCALE = 3,  /* window scale factor */
  /* libnice extensions: */
  TCP_OPT_FIN_ACK = 254,  /* FIN-ACK support */
} TcpOption;


/*
#define FLAG_SYN 0x02
#define FLAG_ACK 0x10
*/

/* NOTE: This must fit in 5 bits. This is used on the wire. */
typedef enum {
  FLAG_NONE = 0,
  FLAG_FIN = 1 << 0,
  FLAG_CTL = 1 << 1,
  FLAG_RST = 1 << 2,
} TcpFlags;

#define CTL_CONNECT  0
//#define CTL_REDIRECT  1
#define CTL_EXTRA 255


#define CTRL_BOUND 0x80000000

/* Maximum segment lifetime (1 minute).
 * RFC 793, §3.3 specifies 2 minutes; but Linux uses 1 minute, so let’s go with
 * that. */
#define TCP_MSL (60 * 1000)

// If there are no pending clocks, wake up every 4 seconds
#define DEFAULT_TIMEOUT 4000
// If the connection is closed, once per minute
#define CLOSED_TIMEOUT (60 * 1000)
/* Timeout after reaching the TIME_WAIT state, in milliseconds.
 * See: RFC 1122, §4.2.2.13.
 *
 * XXX: Since we can control the underlying layer’s channel ID, we can guarantee
 * delayed segments won’t affect subsequent connections, so can radically
 * shorten the TIME-WAIT timeout (to the extent that it basically doesn’t
 * exist). It would normally be (2 * TCP_MSL). */
#define TIME_WAIT_TIMEOUT 1

//////////////////////////////////////////////////////////////////////
// Helper Functions
//////////////////////////////////////////////////////////////////////
#ifndef G_OS_WIN32
#  define min(first, second) ((first) < (second) ? (first) : (second))
#  define max(first, second) ((first) > (second) ? (first) : (second))
#endif

static guint32
bound(guint32 lower, guint32 middle, guint32 upper)
{
   return min (max (lower, middle), upper);
}

static gboolean
time_is_between(guint32 later, guint32 middle, guint32 earlier)
{
  if (earlier <= later) {
    return ((earlier <= middle) && (middle <= later));
  } else {
    return !((later < middle) && (middle < earlier));
  }
}

static gint32
time_diff(guint32 later, guint32 earlier)
{
  guint32 LAST = 0xFFFFFFFF;
  guint32 HALF = 0x80000000;
  if (time_is_between(earlier + HALF, later, earlier)) {
    if (earlier <= later) {
      return (long)(later - earlier);
    } else {
      return (long)(later + (LAST - earlier) + 1);
    }
  } else {
    if (later <= earlier) {
      return -(long) (earlier - later);
    } else {
      return -(long)(earlier + (LAST - later) + 1);
    }
  }
}

////////////////////////////////////////////////////////
// PseudoTcpFifo works exactly like FifoBuffer in libjingle
////////////////////////////////////////////////////////


typedef struct {
  guint8 *buffer;
  gsize buffer_length;
  gsize data_length;
  gsize read_position;
} PseudoTcpFifo;


static void
pseudo_tcp_fifo_init (PseudoTcpFifo *b, gsize size)
{
  b->buffer = g_slice_alloc (size);
  b->buffer_length = size;
}

static void
pseudo_tcp_fifo_clear (PseudoTcpFifo *b)
{
  if (b->buffer)
    g_slice_free1 (b->buffer_length, b->buffer);
  b->buffer = NULL;
  b->buffer_length = 0;
}

static gsize
pseudo_tcp_fifo_get_buffered (PseudoTcpFifo *b)
{
  return b->data_length;
}

static gboolean
pseudo_tcp_fifo_set_capacity (PseudoTcpFifo *b, gsize size)
{
  if (b->data_length > size)
    return FALSE;

  if (size != b->data_length) {
    guint8 *buffer = g_slice_alloc (size);
    gsize copy = b->data_length;
    gsize tail_copy = min (copy, b->buffer_length - b->read_position);

    memcpy (buffer, &b->buffer[b->read_position], tail_copy);
    memcpy (buffer + tail_copy, &b->buffer[0], copy - tail_copy);
    g_slice_free1 (b->buffer_length, b->buffer);
    b->buffer = buffer;
    b->buffer_length = size;
    b->read_position = 0;
  }

  return TRUE;
}

static void
pseudo_tcp_fifo_consume_read_data (PseudoTcpFifo *b, gsize size)
{
  g_assert (size <= b->data_length);

  b->read_position = (b->read_position + size) % b->buffer_length;
  b->data_length -= size;
}

static void
pseudo_tcp_fifo_consume_write_buffer (PseudoTcpFifo *b, gsize size)
{
  g_assert (size <= b->buffer_length - b->data_length);

  b->data_length += size;
}

static gsize
pseudo_tcp_fifo_get_write_remaining (PseudoTcpFifo *b)
{
  return b->buffer_length - b->data_length;
}

static gsize
pseudo_tcp_fifo_read_offset (PseudoTcpFifo *b, guint8 *buffer, gsize bytes,
    gsize offset)
{
  gsize available = b->data_length - offset;
  gsize read_position = (b->read_position + offset) % b->buffer_length;
  gsize copy = min (bytes, available);
  gsize tail_copy = min(copy, b->buffer_length - read_position);

  /* EOS */
  if (offset >= b->data_length)
    return 0;

  memcpy(buffer, &b->buffer[read_position], tail_copy);
  memcpy(buffer + tail_copy, &b->buffer[0], copy - tail_copy);

  return copy;
}

static gsize
pseudo_tcp_fifo_write_offset (PseudoTcpFifo *b, const guint8 *buffer,
    gsize bytes, gsize offset)
{
  gsize available = b->buffer_length - b->data_length - offset;
  gsize write_position = (b->read_position + b->data_length + offset)
      % b->buffer_length;
  gsize copy = min (bytes, available);
  gsize tail_copy = min(copy, b->buffer_length - write_position);

  if (b->data_length + offset >= b->buffer_length) {
    return 0;
  }

  memcpy(&b->buffer[write_position], buffer, tail_copy);
  memcpy(&b->buffer[0], buffer + tail_copy, copy - tail_copy);

  return copy;
}

static gsize
pseudo_tcp_fifo_read (PseudoTcpFifo *b, guint8 *buffer, gsize bytes)
{
  gsize copy;

  copy = pseudo_tcp_fifo_read_offset (b, buffer, bytes, 0);

  b->read_position = (b->read_position + copy) % b->buffer_length;
  b->data_length -= copy;

  return copy;
}

static gsize
pseudo_tcp_fifo_write (PseudoTcpFifo *b, const guint8 *buffer, gsize bytes)
{
  gsize copy;

  copy = pseudo_tcp_fifo_write_offset (b, buffer, bytes, 0);
  b->data_length += copy;

  return copy;
}


//////////////////////////////////////////////////////////////////////
// PseudoTcp
//////////////////////////////////////////////////////////////////////

/* Only used if FIN-ACK support is disabled. */
typedef enum {
  SD_NONE,
  SD_GRACEFUL,
  SD_FORCEFUL
} Shutdown;

typedef enum {
  sfNone,
  sfDelayedAck,
  sfImmediateAck,
  sfFin,
  sfRst,
} SendFlags;

typedef struct {
  guint32 conv, seq, ack;
  TcpFlags flags;
  guint16 wnd;
  const gchar * data;
  guint32 len;
  guint32 tsval, tsecr;
} Segment;

typedef struct {
  guint32 seq, len;
  guint8 xmit;
  TcpFlags flags;
} SSegment;

typedef struct {
  guint32 seq, len;
} RSegment;

/**
 * ClosedownSource:
 * @CLOSEDOWN_LOCAL: Error detected locally, or connection forcefully closed
 * locally.
 * @CLOSEDOWN_REMOTE: RST segment received from the peer.
 *
 * Reasons for calling closedown().
 *
 * Since: 0.1.8
 */
typedef enum {
  CLOSEDOWN_LOCAL,
  CLOSEDOWN_REMOTE,
} ClosedownSource;


struct _PseudoTcpSocketPrivate {
  PseudoTcpCallbacks callbacks;

  Shutdown shutdown;  /* only used if !support_fin_ack */
  gboolean shutdown_reads;
  gint error;

  // TCB data
  PseudoTcpState state;
  guint32 conv;
  gboolean bReadEnable, bWriteEnable, bOutgoing;
  guint32 last_traffic;

  // Incoming data
  GList *rlist;
  guint32 rbuf_len, rcv_nxt, rcv_wnd, lastrecv;
  guint8 rwnd_scale; // Window scale factor
  PseudoTcpFifo rbuf;
  guint32 rcv_fin;  /* sequence number of the received FIN octet, or 0 */

  // Outgoing data
  GQueue slist;
  GQueue unsent_slist;
  guint32 sbuf_len, snd_nxt, snd_wnd, lastsend;
  guint32 snd_una;  /* oldest unacknowledged sequence number */
  guint8 swnd_scale; // Window scale factor
  PseudoTcpFifo sbuf;

  // Maximum segment size, estimated protocol level, largest segment sent
  guint32 mss, msslevel, largest, mtu_advise;
  // Retransmit timer
  guint32 rto_base;

  // Timestamp tracking
  guint32 ts_recent, ts_lastack;

  // Round-trip calculation
  guint32 rx_rttvar, rx_srtt, rx_rto;

  // Congestion avoidance, Fast retransmit/recovery, Delayed ACKs
  guint32 ssthresh, cwnd;
  guint8 dup_acks;
  guint32 recover;
  gboolean fast_recovery;
  guint32 t_ack;  /* time a delayed ack was scheduled; 0 if no acks scheduled */

  gboolean use_nagling;
  guint32 ack_delay;

  // This is used by unit tests to test backward compatibility of
  // PseudoTcp implementations that don't support window scaling.
  gboolean support_wnd_scale;

  /* Current time. Typically only used for testing, when non-zero. When zero,
   * the system monotonic clock is used. Units: monotonic milliseconds. */
  guint32 current_time;

  /* This is used by compatible implementations (with the TCP_OPT_FIN_ACK
   * option) to enable correct FIN-ACK connection termination. Defaults to
   * TRUE unless no compatible option is received. */
  gboolean support_fin_ack;
};

#define LARGER(a,b) (((a) - (b) - 1) < (G_MAXUINT32 >> 1))
#define LARGER_OR_EQUAL(a,b) (((a) - (b)) < (G_MAXUINT32 >> 1))
#define SMALLER(a,b) LARGER ((b),(a))
#define SMALLER_OR_EQUAL(a,b) LARGER_OR_EQUAL ((b),(a))

/* properties */
enum
{
  PROP_CONVERSATION = 1,
  PROP_CALLBACKS,
  PROP_STATE,
  PROP_ACK_DELAY,
  PROP_NO_DELAY,
  PROP_RCV_BUF,
  PROP_SND_BUF,
  PROP_SUPPORT_FIN_ACK,
  LAST_PROPERTY
};


static void pseudo_tcp_socket_get_property (GObject *object, guint property_id,
    GValue *value,  GParamSpec *pspec);
static void pseudo_tcp_socket_set_property (GObject *object, guint property_id,
    const GValue *value, GParamSpec *pspec);
static void pseudo_tcp_socket_finalize (GObject *object);


static void queue_connect_message (PseudoTcpSocket *self);
static guint32 queue (PseudoTcpSocket *self, const gchar *data,
    guint32 len, TcpFlags flags);
static PseudoTcpWriteResult packet(PseudoTcpSocket *self, guint32 seq,
    TcpFlags flags, guint32 offset, guint32 len, guint32 now);
static gboolean parse (PseudoTcpSocket *self,
    const guint8 *_header_buf, gsize header_buf_len,
    const guint8 *data_buf, gsize data_buf_len);
static gboolean process(PseudoTcpSocket *self, Segment *seg);
static int transmit(PseudoTcpSocket *self, SSegment *sseg, guint32 now);
static void attempt_send(PseudoTcpSocket *self, SendFlags sflags);
static void closedown (PseudoTcpSocket *self, guint32 err,
    ClosedownSource source);
static void adjustMTU(PseudoTcpSocket *self);
static void parse_options (PseudoTcpSocket *self, const guint8 *data,
    guint32 len);
static void resize_send_buffer (PseudoTcpSocket *self, guint32 new_size);
static void resize_receive_buffer (PseudoTcpSocket *self, guint32 new_size);
static void set_state (PseudoTcpSocket *self, PseudoTcpState new_state);
static void set_state_established (PseudoTcpSocket *self);
static void set_state_closed (PseudoTcpSocket *self, guint32 err);

static const gchar *pseudo_tcp_state_get_name (PseudoTcpState state);
static gboolean pseudo_tcp_state_has_sent_fin (PseudoTcpState state);
static gboolean pseudo_tcp_state_has_received_fin (PseudoTcpState state);
static gboolean pseudo_tcp_state_has_received_fin_ack (PseudoTcpState state);

// The following logging is for detailed (packet-level) pseudotcp analysis only.
static PseudoTcpDebugLevel debug_level = PSEUDO_TCP_DEBUG_NONE;

#define DEBUG(level, fmt, ...)                                          \
  if (debug_level >= level)                                             \
    g_log (level == PSEUDO_TCP_DEBUG_NORMAL ? "libnice-pseudotcp" : "libnice-pseudotcp-verbose", G_LOG_LEVEL_DEBUG, "PseudoTcpSocket %p %s: " fmt, \
        self, pseudo_tcp_state_get_name (self->priv->state), ## __VA_ARGS__)

void
pseudo_tcp_set_debug_level (PseudoTcpDebugLevel level)
{
  debug_level = level;
}

static guint32
get_current_time (PseudoTcpSocket *socket)
{
  if (G_UNLIKELY (socket->priv->current_time != 0))
    return socket->priv->current_time;

  return g_get_monotonic_time () / 1000;
}

void
pseudo_tcp_socket_set_time (PseudoTcpSocket *self, guint32 current_time)
{
  self->priv->current_time = current_time;
}

static void
pseudo_tcp_socket_class_init (PseudoTcpSocketClass *cls)
{
  GObjectClass *object_class = G_OBJECT_CLASS (cls);

  object_class->get_property = pseudo_tcp_socket_get_property;
  object_class->set_property = pseudo_tcp_socket_set_property;
  object_class->finalize = pseudo_tcp_socket_finalize;

  g_object_class_install_property (object_class, PROP_CONVERSATION,
      g_param_spec_uint ("conversation", "TCP Conversation ID",
          "The TCP Conversation ID",
          0, G_MAXUINT32, 0,
          G_PARAM_CONSTRUCT_ONLY | G_PARAM_READWRITE | G_PARAM_STATIC_STRINGS));

  g_object_class_install_property (object_class, PROP_CALLBACKS,
      g_param_spec_pointer ("callbacks", "PseudoTcp socket callbacks",
          "Structure with the callbacks to call when PseudoTcp events happen",
          G_PARAM_READWRITE | G_PARAM_STATIC_STRINGS));

  g_object_class_install_property (object_class, PROP_STATE,
      g_param_spec_uint ("state", "PseudoTcp State",
          "The current state (enum PseudoTcpState) of the PseudoTcp socket",
          TCP_LISTEN, TCP_CLOSED, TCP_LISTEN,
          G_PARAM_READABLE | G_PARAM_STATIC_STRINGS));

  g_object_class_install_property (object_class, PROP_ACK_DELAY,
      g_param_spec_uint ("ack-delay", "ACK Delay",
          "Delayed ACK timeout (in milliseconds)",
          0, G_MAXUINT, DEFAULT_ACK_DELAY,
          G_PARAM_READWRITE| G_PARAM_STATIC_STRINGS));

  g_object_class_install_property (object_class, PROP_NO_DELAY,
      g_param_spec_boolean ("no-delay", "No Delay",
          "Disable the Nagle algorithm (like the TCP_NODELAY option)",
          DEFAULT_NO_DELAY,
          G_PARAM_READWRITE| G_PARAM_STATIC_STRINGS));

  g_object_class_install_property (object_class, PROP_RCV_BUF,
      g_param_spec_uint ("rcv-buf", "Receive Buffer",
          "Receive Buffer size",
          1, G_MAXUINT, DEFAULT_RCV_BUF_SIZE,
          G_PARAM_READWRITE| G_PARAM_STATIC_STRINGS));

  g_object_class_install_property (object_class, PROP_SND_BUF,
      g_param_spec_uint ("snd-buf", "Send Buffer",
          "Send Buffer size",
          1, G_MAXUINT, DEFAULT_SND_BUF_SIZE,
          G_PARAM_READWRITE| G_PARAM_STATIC_STRINGS));

  /**
   * PseudoTcpSocket:support-fin-ack:
   *
   * Whether to support the FIN–ACK extension to the pseudo-TCP protocol for
   * this socket. The extension is only compatible with other libnice pseudo-TCP
   * stacks, and not with Jingle pseudo-TCP stacks. If enabled, support is
   * negotiatied on connection setup, so it is safe for a #PseudoTcpSocket with
   * support enabled to be used with one with it disabled, or with a Jingle
   * pseudo-TCP socket which doesn’t support it at all.
   *
   * Support is enabled by default.
   *
   * Since: 0.1.8
   */
  g_object_class_install_property (object_class, PROP_SUPPORT_FIN_ACK,
      g_param_spec_boolean ("support-fin-ack", "Support FIN–ACK",
          "Whether to enable the optional FIN–ACK support.",
          TRUE,
          G_PARAM_READWRITE | G_PARAM_CONSTRUCT_ONLY | G_PARAM_STATIC_STRINGS));
}


static void
pseudo_tcp_socket_get_property (GObject *object,
                                  guint property_id,
                                  GValue *value,
                                  GParamSpec *pspec)
{
  PseudoTcpSocket *self = PSEUDO_TCP_SOCKET (object);

  switch (property_id) {
    case PROP_CONVERSATION:
      g_value_set_uint (value, self->priv->conv);
      break;
    case PROP_CALLBACKS:
      g_value_set_pointer (value, (gpointer) &self->priv->callbacks);
      break;
    case PROP_STATE:
      g_value_set_uint (value, self->priv->state);
      break;
    case PROP_ACK_DELAY:
      g_value_set_uint (value, self->priv->ack_delay);
      break;
    case PROP_NO_DELAY:
      g_value_set_boolean (value, !self->priv->use_nagling);
      break;
    case PROP_RCV_BUF:
      g_value_set_uint (value, self->priv->rbuf_len);
      break;
    case PROP_SND_BUF:
      g_value_set_uint (value, self->priv->sbuf_len);
      break;
    case PROP_SUPPORT_FIN_ACK:
      g_value_set_boolean (value, self->priv->support_fin_ack);
      break;
    default:
      G_OBJECT_WARN_INVALID_PROPERTY_ID (object, property_id, pspec);
      break;
  }
}

static void
pseudo_tcp_socket_set_property (GObject *object,
                                  guint property_id,
                                  const GValue *value,
                                  GParamSpec *pspec)
{
  PseudoTcpSocket *self = PSEUDO_TCP_SOCKET (object);

  switch (property_id) {
    case PROP_CONVERSATION:
      self->priv->conv = g_value_get_uint (value);
      break;
    case PROP_CALLBACKS:
      {
        PseudoTcpCallbacks *c = g_value_get_pointer (value);
        self->priv->callbacks = *c;
      }
      break;
    case PROP_ACK_DELAY:
      self->priv->ack_delay = g_value_get_uint (value);
      break;
    case PROP_NO_DELAY:
      self->priv->use_nagling = !g_value_get_boolean (value);
      break;
    case PROP_RCV_BUF:
      g_return_if_fail (self->priv->state == TCP_LISTEN);
      resize_receive_buffer (self, g_value_get_uint (value));
      break;
    case PROP_SND_BUF:
      g_return_if_fail (self->priv->state == TCP_LISTEN);
      resize_send_buffer (self, g_value_get_uint (value));
      break;
    case PROP_SUPPORT_FIN_ACK:
      self->priv->support_fin_ack = g_value_get_boolean (value);
      break;
    default:
      G_OBJECT_WARN_INVALID_PROPERTY_ID (object, property_id, pspec);
      break;
  }
}

static void
pseudo_tcp_socket_finalize (GObject *object)
{
  PseudoTcpSocket *self = PSEUDO_TCP_SOCKET (object);
  PseudoTcpSocketPrivate *priv = self->priv;
  GList *i;
  SSegment *sseg;

  if (priv == NULL)
    return;

  while ((sseg = g_queue_pop_head (&priv->slist)))
    g_slice_free (SSegment, sseg);
  g_queue_clear (&priv->unsent_slist);
  for (i = priv->rlist; i; i = i->next) {
    RSegment *rseg = i->data;
    g_slice_free (RSegment, rseg);
  }
  g_list_free (priv->rlist);
  priv->rlist = NULL;

  pseudo_tcp_fifo_clear (&priv->rbuf);
  pseudo_tcp_fifo_clear (&priv->sbuf);

  g_free (priv);
  self->priv = NULL;

  if (G_OBJECT_CLASS (pseudo_tcp_socket_parent_class)->finalize)
    G_OBJECT_CLASS (pseudo_tcp_socket_parent_class)->finalize (object);
}


static void
pseudo_tcp_socket_init (PseudoTcpSocket *obj)
{
  /* Use g_new0, and do not use g_object_set_private because the size of
   * our private data is too big (150KB+) and the g_slice_allow cannot allocate
   * it. So we handle the private ourselves */
  PseudoTcpSocketPrivate *priv = g_new0 (PseudoTcpSocketPrivate, 1);

  obj->priv = priv;

  priv->shutdown = SD_NONE;
  priv->error = 0;

  priv->rbuf_len = DEFAULT_RCV_BUF_SIZE;
  pseudo_tcp_fifo_init (&priv->rbuf, priv->rbuf_len);
  priv->sbuf_len = DEFAULT_SND_BUF_SIZE;
  pseudo_tcp_fifo_init (&priv->sbuf, priv->sbuf_len);

  priv->state = TCP_LISTEN;
  priv->conv = 0;
  g_queue_init (&priv->slist);
  g_queue_init (&priv->unsent_slist);
  priv->rcv_wnd = priv->rbuf_len;
  priv->rwnd_scale = priv->swnd_scale = 0;
  priv->snd_nxt = 0;
  priv->snd_wnd = 1;
  priv->snd_una = priv->rcv_nxt = 0;
  priv->bReadEnable = TRUE;
  priv->bWriteEnable = FALSE;
  priv->rcv_fin = 0;

  priv->t_ack = 0;

  priv->msslevel = 0;
  priv->largest = 0;
  priv->mss = MIN_PACKET - PACKET_OVERHEAD;
  priv->mtu_advise = DEF_MTU;

  priv->rto_base = 0;

  priv->cwnd = 2 * priv->mss;
  priv->ssthresh = priv->rbuf_len;
  priv->lastrecv = priv->lastsend = priv->last_traffic = 0;
  priv->bOutgoing = FALSE;

  priv->dup_acks = 0;
  priv->recover = 0;

  priv->ts_recent = priv->ts_lastack = 0;

  priv->rx_rto = DEF_RTO;
  priv->rx_srtt = priv->rx_rttvar = 0;

  priv->ack_delay = DEFAULT_ACK_DELAY;
  priv->use_nagling = !DEFAULT_NO_DELAY;

  priv->support_wnd_scale = TRUE;
  priv->support_fin_ack = TRUE;
}

PseudoTcpSocket *pseudo_tcp_socket_new (guint32 conversation,
    PseudoTcpCallbacks *callbacks)
{

  return g_object_new (PSEUDO_TCP_SOCKET_TYPE,
      "conversation", conversation,
      "callbacks", callbacks,
      NULL);
}

static void
queue_connect_message (PseudoTcpSocket *self)
{
  PseudoTcpSocketPrivate *priv = self->priv;
  guint8 buf[8];
  gsize size = 0;

  buf[size++] = CTL_CONNECT;

  if (priv->support_wnd_scale) {
    buf[size++] = TCP_OPT_WND_SCALE;
    buf[size++] = 1;
    buf[size++] = priv->rwnd_scale;
  }

  if (priv->support_fin_ack) {
    buf[size++] = TCP_OPT_FIN_ACK;
    buf[size++] = 1;  /* option length; zero is invalid (RFC 1122, §4.2.2.5) */
    buf[size++] = 0;  /* currently unused */
  }

  priv->snd_wnd = size;

  queue (self, (char *) buf, size, FLAG_CTL);
}

static void
queue_fin_message (PseudoTcpSocket *self)
{
  g_assert (self->priv->support_fin_ack);

  /* FIN segments are always zero-length. */
  queue (self, "", 0, FLAG_FIN);
}

static void
queue_rst_message (PseudoTcpSocket *self)
{
  g_assert (self->priv->support_fin_ack);

  /* RST segments are always zero-length. */
  queue (self, "", 0, FLAG_RST);
}

gboolean
pseudo_tcp_socket_connect(PseudoTcpSocket *self)
{
  PseudoTcpSocketPrivate *priv = self->priv;

  if (priv->state != TCP_LISTEN) {
    priv->error = EINVAL;
    return FALSE;
  }

  set_state (self, TCP_SYN_SENT);

  queue_connect_message (self);
  attempt_send(self, sfNone);

  return TRUE;
}

void
pseudo_tcp_socket_notify_mtu(PseudoTcpSocket *self, guint16 mtu)
{
  PseudoTcpSocketPrivate *priv = self->priv;
  priv->mtu_advise = mtu;
  if (priv->state == TCP_ESTABLISHED) {
    adjustMTU(self);
  }
}

void
pseudo_tcp_socket_notify_clock(PseudoTcpSocket *self)
{
  PseudoTcpSocketPrivate *priv = self->priv;
  guint32 now = get_current_time (self);

  if (priv->state == TCP_CLOSED)
    return;

  /* If in the TIME-WAIT state, any delayed segments have passed and the
   * connection can be considered closed from both ends.
   * FIXME: This should probably actually compare a timestamp before
   * operating. */
  if (priv->support_fin_ack && priv->state == TCP_TIME_WAIT) {
    DEBUG (PSEUDO_TCP_DEBUG_NORMAL,
        "Notified clock in TIME-WAIT state; closing connection.");
    set_state_closed (self, 0);
  }

  /* If in the LAST-ACK state, resend the FIN because it hasn’t been ACKed yet.
   * FIXME: This should probably actually compare a timestamp before
   * operating. */
  if (priv->support_fin_ack && priv->state == TCP_LAST_ACK) {
    DEBUG (PSEUDO_TCP_DEBUG_NORMAL,
        "Notified clock in LAST-ACK state; resending FIN segment.");
    queue_fin_message (self);
    attempt_send (self, sfFin);
  }

  // Check if it's time to retransmit a segment
  if (priv->rto_base &&
      (time_diff(priv->rto_base + priv->rx_rto, now) <= 0)) {
    if (g_queue_get_length (&priv->slist) == 0) {
      g_assert_not_reached ();
    } else {
      // Note: (priv->slist.front().xmit == 0)) {
      // retransmit segments
      guint32 nInFlight;
      guint32 rto_limit;
      int transmit_status;

      DEBUG (PSEUDO_TCP_DEBUG_NORMAL, "timeout retransmit (rto: %u) "
          "(rto_base: %u) (now: %u) (dup_acks: %u)",
          priv->rx_rto, priv->rto_base, now, (guint) priv->dup_acks);

      transmit_status = transmit(self, g_queue_peek_head (&priv->slist), now);
      if (transmit_status != 0) {
        DEBUG (PSEUDO_TCP_DEBUG_NORMAL,
            "Error transmitting segment. Closing down.");
        closedown (self, transmit_status, CLOSEDOWN_LOCAL);
        return;
      }

      nInFlight = priv->snd_nxt - priv->snd_una;
      priv->ssthresh = max(nInFlight / 2, 2 * priv->mss);
      DEBUG (PSEUDO_TCP_DEBUG_NORMAL, "ssthresh: %u = (nInFlight: %u / 2) + "
          "2 * mss: %u", priv->ssthresh, nInFlight, priv->mss);
      //LOG(LS_INFO) << "priv->ssthresh: " << priv->ssthresh << "  nInFlight: " << nInFlight << "  priv->mss: " << priv->mss;
      priv->cwnd = priv->mss;

      // Back off retransmit timer.  Note: the limit is lower when connecting.
      rto_limit = (priv->state < TCP_ESTABLISHED) ? DEF_RTO : MAX_RTO;
      priv->rx_rto = min(rto_limit, priv->rx_rto * 2);
      priv->rto_base = now;

      priv->recover = priv->snd_nxt;
      if (priv->dup_acks >= 3) {
        priv->dup_acks = 0;
        priv->fast_recovery = FALSE;
        DEBUG (PSEUDO_TCP_DEBUG_NORMAL, "exit recovery on timeout");
      }
    }
  }

  // Check if it's time to probe closed windows
  if ((priv->snd_wnd == 0)
        && (time_diff(priv->lastsend + priv->rx_rto, now) <= 0)) {
    if (time_diff(now, priv->lastrecv) >= 15000) {
      DEBUG (PSEUDO_TCP_DEBUG_NORMAL, "Receive window closed. Closing down.");
      closedown (self, ECONNABORTED, CLOSEDOWN_LOCAL);
      return;
    }

    // probe the window
    packet(self, priv->snd_nxt - 1, 0, 0, 0, now);
    priv->lastsend = now;

    // back off retransmit timer
    priv->rx_rto = min(MAX_RTO, priv->rx_rto * 2);
  }

  // Check if it's time to send delayed acks
  if (priv->t_ack && (time_diff(priv->t_ack + priv->ack_delay, now) <= 0)) {
    packet(self, priv->snd_nxt, 0, 0, 0, now);
  }

}

gboolean
pseudo_tcp_socket_notify_packet(PseudoTcpSocket *self,
    const gchar * buffer, guint32 len)
{
  gboolean retval;

  if (len > MAX_PACKET) {
    //LOG_F(WARNING) << "packet too large";
    self->priv->error = EMSGSIZE;
    return FALSE;
  } else if (len < HEADER_SIZE) {
    //LOG_F(WARNING) << "packet too small";
    self->priv->error = EINVAL;
    return FALSE;
  }

  /* Hold a reference to the PseudoTcpSocket during parsing, since it may be
   * closed from within a callback. */
  g_object_ref (self);
  retval = parse (self, (guint8 *) buffer, HEADER_SIZE,
      (guint8 *) buffer + HEADER_SIZE, len - HEADER_SIZE);
  g_object_unref (self);

  return retval;
}

/* Assume there are two buffers in the given #NiceInputMessage: a 24-byte one
 * containing the header, and a bigger one for the data. */
gboolean
pseudo_tcp_socket_notify_message (PseudoTcpSocket *self,
    NiceInputMessage *message)
{
  gboolean retval;

  g_assert_cmpuint (message->n_buffers, >, 0);

  if (message->n_buffers == 1)
    return pseudo_tcp_socket_notify_packet (self, message->buffers[0].buffer,
        message->buffers[0].size);

  g_assert_cmpuint (message->n_buffers, ==, 2);
  g_assert_cmpuint (message->buffers[0].size, ==, HEADER_SIZE);

  if (message->length > MAX_PACKET) {
    //LOG_F(WARNING) << "packet too large";
    return FALSE;
  } else if (message->length < HEADER_SIZE) {
    //LOG_F(WARNING) << "packet too small";
    return FALSE;
  }

  /* Hold a reference to the PseudoTcpSocket during parsing, since it may be
   * closed from within a callback. */
  g_object_ref (self);
  retval = parse (self, message->buffers[0].buffer, message->buffers[0].size,
      message->buffers[1].buffer, message->length - message->buffers[0].size);
  g_object_unref (self);

  return retval;
}

gboolean
pseudo_tcp_socket_get_next_clock(PseudoTcpSocket *self, guint64 *timeout)
{
  PseudoTcpSocketPrivate *priv = self->priv;
  guint32 now = get_current_time (self);
  gsize snd_buffered;
  guint32 closed_timeout;

  if (priv->shutdown == SD_FORCEFUL) {
    if (priv->support_fin_ack) {
      DEBUG (PSEUDO_TCP_DEBUG_NORMAL,
          "‘Forceful’ shutdown used when FIN-ACK support is enabled");
    }

    /* Transition to the CLOSED state. */
    closedown (self, 0, CLOSEDOWN_REMOTE);

    return FALSE;
  }

  snd_buffered = pseudo_tcp_fifo_get_buffered (&priv->sbuf);
  if ((priv->shutdown == SD_GRACEFUL)
      && ((priv->state != TCP_ESTABLISHED)
          || ((snd_buffered == 0) && (priv->t_ack == 0)))) {
    if (priv->support_fin_ack) {
      DEBUG (PSEUDO_TCP_DEBUG_NORMAL,
          "‘Graceful’ shutdown used when FIN-ACK support is enabled");
    }

    /* Transition to the CLOSED state. */
    closedown (self, 0, CLOSEDOWN_REMOTE);

    return FALSE;
  }

  /* FIN-ACK support. The timeout for closing the socket if nothing is received
   * varies depending on whether the socket is waiting in the TIME-WAIT state
   * for delayed segments to pass.
   *
   * See: http://vincent.bernat.im/en/blog/2014-tcp-time-wait-state-linux.html
   */
  closed_timeout = CLOSED_TIMEOUT;
  if (priv->support_fin_ack && priv->state == TCP_TIME_WAIT)
    closed_timeout = TIME_WAIT_TIMEOUT;

  if (priv->support_fin_ack && priv->state == TCP_CLOSED) {
    return FALSE;
  }

  if (*timeout == 0 || *timeout < now)
    *timeout = now + closed_timeout;

  if (priv->support_fin_ack && priv->state == TCP_TIME_WAIT) {
    *timeout = min (*timeout, now + TIME_WAIT_TIMEOUT);
    return TRUE;
  }

  if (priv->state == TCP_CLOSED && !priv->support_fin_ack) {
    *timeout = min (*timeout, now + CLOSED_TIMEOUT);
    return TRUE;
  }

  *timeout = min (*timeout, now + DEFAULT_TIMEOUT);

  if (priv->t_ack) {
    *timeout = min(*timeout, priv->t_ack + priv->ack_delay);
  }
  if (priv->rto_base) {
    *timeout = min(*timeout, priv->rto_base + priv->rx_rto);
  }
  if (priv->snd_wnd == 0) {
    *timeout = min(*timeout, priv->lastsend + priv->rx_rto);
  }

  return TRUE;
}


gint
pseudo_tcp_socket_recv(PseudoTcpSocket *self, char * buffer, size_t len)
{
  PseudoTcpSocketPrivate *priv = self->priv;
  gsize bytesread;
  gsize available_space;

  /* Received a FIN from the peer, so return 0. RFC 793, §3.5, Case 2. */
  if (priv->support_fin_ack && priv->shutdown_reads) {
    return 0;
  }

  /* Return 0 if FIN-ACK is not supported but the socket has been closed. */
  if (!priv->support_fin_ack && pseudo_tcp_socket_is_closed (self)) {
    return 0;
  }

  /* Return ENOTCONN if FIN-ACK is not supported and the connection is not
   * ESTABLISHED. */
  if (!priv->support_fin_ack && priv->state != TCP_ESTABLISHED) {
    priv->error = ENOTCONN;
    return -1;
  }

  if (len == 0)
    return 0;

  bytesread = pseudo_tcp_fifo_read (&priv->rbuf, (guint8 *) buffer, len);

 // If there's no data in |m_rbuf|.
  if (bytesread == 0 &&
      !(pseudo_tcp_state_has_received_fin (priv->state) ||
        pseudo_tcp_state_has_received_fin_ack (priv->state))) {
    priv->bReadEnable = TRUE;
    priv->error = EWOULDBLOCK;
    return -1;
  }

  available_space = pseudo_tcp_fifo_get_write_remaining (&priv->rbuf);

  if (available_space - priv->rcv_wnd >=
      min (priv->rbuf_len / 2, priv->mss)) {
    // !?! Not sure about this was closed business
    gboolean bWasClosed = (priv->rcv_wnd == 0);

    priv->rcv_wnd = available_space;

    if (bWasClosed) {
      attempt_send(self, sfImmediateAck);
    }
  }

  return bytesread;
}

gint
pseudo_tcp_socket_send(PseudoTcpSocket *self, const char * buffer, guint32 len)
{
  PseudoTcpSocketPrivate *priv = self->priv;
  gint written;
  gsize available_space;

  if (priv->state != TCP_ESTABLISHED) {
    priv->error = pseudo_tcp_state_has_sent_fin (priv->state) ? EPIPE : ENOTCONN;
    return -1;
  }

  available_space = pseudo_tcp_fifo_get_write_remaining (&priv->sbuf);

  if (!available_space) {
    priv->bWriteEnable = TRUE;
    priv->error = EWOULDBLOCK;
    return -1;
  }

  written = queue (self, buffer, len, FLAG_NONE);
  attempt_send(self, sfNone);

  if (written > 0 && (guint32)written < len) {
    priv->bWriteEnable = TRUE;
  }

  return written;
}

void
pseudo_tcp_socket_close(PseudoTcpSocket *self, gboolean force)
{
  PseudoTcpSocketPrivate *priv = self->priv;

  DEBUG (PSEUDO_TCP_DEBUG_NORMAL, "Closing socket %p %s", self,
      force ? "forcefully" : "gracefully");

  /* Forced closure by sending an RST segment. RFC 1122, §4.2.2.13. */
  if (force && priv->state != TCP_CLOSED) {
    closedown (self, ECONNABORTED, CLOSEDOWN_LOCAL);
    return;
  }

  /* Fall back to shutdown(). */
  pseudo_tcp_socket_shutdown (self, PSEUDO_TCP_SHUTDOWN_RDWR);
}

void
pseudo_tcp_socket_shutdown (PseudoTcpSocket *self, PseudoTcpShutdown how)
{
  PseudoTcpSocketPrivate *priv = self->priv;

  DEBUG (PSEUDO_TCP_DEBUG_NORMAL, "Shutting down socket %p: %u", self, how);

  /* FIN-ACK--only stuff below here. */
  if (!priv->support_fin_ack) {
    if (priv->shutdown == SD_NONE)
      priv->shutdown = SD_GRACEFUL;
    return;
  }

  /* What needs shutting down? */
  switch (how) {
  case PSEUDO_TCP_SHUTDOWN_RD:
  case PSEUDO_TCP_SHUTDOWN_RDWR:
    priv->shutdown_reads = TRUE;
    break;
  case PSEUDO_TCP_SHUTDOWN_WR:
    /* Handled below. */
    break;
  default:
    DEBUG (PSEUDO_TCP_DEBUG_NORMAL, "Invalid shutdown method: %u.", how);
    break;
  }

  if (how == PSEUDO_TCP_SHUTDOWN_RD) {
    return;
  }

  /* Unforced write closure. */
  switch (priv->state) {
  case TCP_LISTEN:
  case TCP_SYN_SENT:
    /* Just abort the connection without completing the handshake. */
    set_state_closed (self, 0);
    break;
  case TCP_SYN_RECEIVED:
  case TCP_ESTABLISHED:
    /* Local user initiating the close: RFC 793, §3.5, Cases 1 and 3.
     * If there is pending receive data, send RST instead of FIN;
     * see RFC 1122, §4.2.2.13. */
    if (pseudo_tcp_socket_get_available_bytes (self) > 0) {
      closedown (self, ECONNABORTED, CLOSEDOWN_LOCAL);
    } else {
      queue_fin_message (self);
      attempt_send (self, sfFin);
      set_state (self, TCP_FIN_WAIT_1);
    }
    break;
  case TCP_CLOSE_WAIT:
    /* Remote user initiating the close: RFC 793, §3.5, Case 2.
     * We’ve previously received a FIN from the peer; now the user is closing
     * the local end of the connection. */
    queue_fin_message (self);
    attempt_send (self, sfFin);
    set_state (self, TCP_LAST_ACK);
    break;
  case TCP_CLOSING:
  case TCP_CLOSED:
    /* Already closed on both sides. */
    break;
  case TCP_FIN_WAIT_1:
  case TCP_FIN_WAIT_2:
  case TCP_TIME_WAIT:
  case TCP_LAST_ACK:
    /* Already closed locally. */
    break;
  default:
    /* Do nothing. */
    break;
  }
}

int
pseudo_tcp_socket_get_error(PseudoTcpSocket *self)
{
  PseudoTcpSocketPrivate *priv = self->priv;
  return priv->error;
}

//
// Internal Implementation
//

static guint32
queue (PseudoTcpSocket *self, const gchar * data, guint32 len, TcpFlags flags)
{
  PseudoTcpSocketPrivate *priv = self->priv;
  gsize available_space;

  available_space = pseudo_tcp_fifo_get_write_remaining (&priv->sbuf);
  if (len > available_space) {
    g_assert (flags == FLAG_NONE);
    len = available_space;
  }

  // We can concatenate data if the last segment is the same type
  // (control v. regular data), and has not been transmitted yet
  if (g_queue_get_length (&priv->slist) &&
      (((SSegment *)g_queue_peek_tail (&priv->slist))->flags == flags) &&
      (((SSegment *)g_queue_peek_tail (&priv->slist))->xmit == 0)) {
    ((SSegment *)g_queue_peek_tail (&priv->slist))->len += len;
  } else {
    SSegment *sseg = g_slice_new0 (SSegment);
    gsize snd_buffered = pseudo_tcp_fifo_get_buffered (&priv->sbuf);

    sseg->seq = priv->snd_una + snd_buffered;
    sseg->len = len;
    sseg->flags = flags;
    g_queue_push_tail (&priv->slist, sseg);
    g_queue_push_tail (&priv->unsent_slist, sseg);
  }

  //LOG(LS_INFO) << "PseudoTcp::queue - priv->slen = " << priv->slen;
  return pseudo_tcp_fifo_write (&priv->sbuf, (guint8*) data, len);;
}

// Creates a packet and submits it to the network. This method can either
// send payload or just an ACK packet.
//
// |seq| is the sequence number of this packet.
// |flags| is the flags for sending this packet.
// |offset| is the offset to read from |m_sbuf|.
// |len| is the number of bytes to read from |m_sbuf| as payload. If this
// value is 0 then this is an ACK packet, otherwise this packet has payload.

static PseudoTcpWriteResult
packet(PseudoTcpSocket *self, guint32 seq, TcpFlags flags,
    guint32 offset, guint32 len, guint32 now)
{
  PseudoTcpSocketPrivate *priv = self->priv;
  union {
    guint8 u8[MAX_PACKET];
    guint16 u16[MAX_PACKET / 2];
    guint32 u32[MAX_PACKET / 4];
  } buffer;
  PseudoTcpWriteResult wres = WR_SUCCESS;

  g_assert(HEADER_SIZE + len <= MAX_PACKET);

  *buffer.u32 = htonl(priv->conv);
  *(buffer.u32 + 1) = htonl(seq);
  *(buffer.u32 + 2) = htonl(priv->rcv_nxt);
  buffer.u8[12] = 0;
  buffer.u8[13] = flags;
  *(buffer.u16 + 7) = htons((guint16)(priv->rcv_wnd >> priv->rwnd_scale));

  // Timestamp computations
  *(buffer.u32 + 4) = htonl(now);
  *(buffer.u32 + 5) = htonl(priv->ts_recent);
  priv->ts_lastack = priv->rcv_nxt;

  if (len) {
    gsize bytes_read;

    bytes_read = pseudo_tcp_fifo_read_offset (&priv->sbuf, buffer.u8 + HEADER_SIZE,
        len, offset);
    g_assert (bytes_read == len);
  }

  DEBUG (PSEUDO_TCP_DEBUG_VERBOSE, "Sending <CONV=%u><FLG=%u><SEQ=%u:%u><ACK=%u>"
      "<WND=%u><TS=%u><TSR=%u><LEN=%u>",
      priv->conv, (unsigned)flags, seq, seq + len, priv->rcv_nxt, priv->rcv_wnd,
      now % 10000, priv->ts_recent % 10000, len);

  wres = priv->callbacks.WritePacket(self, (gchar *) buffer.u8, len + HEADER_SIZE,
                                     priv->callbacks.user_data);
  /* Note: When len is 0, this is an ACK packet.  We don't read the
     return value for those, and thus we won't retry.  So go ahead and treat
     the packet as a success (basically simulate as if it were dropped),
     which will prevent our timers from being messed up. */
  if ((wres != WR_SUCCESS) && (0 != len))
    return wres;

  priv->t_ack = 0;
  if (len > 0) {
    priv->lastsend = now;
  }
  priv->last_traffic = now;
  priv->bOutgoing = TRUE;

  return WR_SUCCESS;
}

static gboolean
parse (PseudoTcpSocket *self, const guint8 *_header_buf, gsize header_buf_len,
    const guint8 *data_buf, gsize data_buf_len)
{
  Segment seg;

  union {
    const guint8 *u8;
    const guint16 *u16;
    const guint32 *u32;
  } header_buf;

  header_buf.u8 = _header_buf;

  if (header_buf_len != 24)
    return FALSE;

  seg.conv = ntohl(*header_buf.u32);
  seg.seq = ntohl(*(header_buf.u32 + 1));
  seg.ack = ntohl(*(header_buf.u32 + 2));
  seg.flags = header_buf.u8[13];
  seg.wnd = ntohs(*(header_buf.u16 + 7));

  seg.tsval = ntohl(*(header_buf.u32 + 4));
  seg.tsecr = ntohl(*(header_buf.u32 + 5));

  seg.data = (const gchar *) data_buf;
  seg.len = data_buf_len;

  DEBUG (PSEUDO_TCP_DEBUG_VERBOSE,
      "Received <CONV=%u><FLG=%u><SEQ=%u:%u><ACK=%u>"
      "<WND=%u><TS=%u><TSR=%u><LEN=%u>",
      seg.conv, (unsigned)seg.flags, seg.seq, seg.seq + seg.len, seg.ack,
      seg.wnd, seg.tsval % 10000, seg.tsecr % 10000, seg.len);

  return process(self, &seg);
}

/* True iff the @state requires that a FIN has already been sent by this
 * host. */
static gboolean
pseudo_tcp_state_has_sent_fin (PseudoTcpState state)
{
  switch (state) {
  case TCP_LISTEN:
  case TCP_SYN_SENT:
  case TCP_SYN_RECEIVED:
  case TCP_ESTABLISHED:
  case TCP_CLOSE_WAIT:
    return FALSE;
  case TCP_CLOSED:
  case TCP_FIN_WAIT_1:
  case TCP_FIN_WAIT_2:
  case TCP_CLOSING:
  case TCP_TIME_WAIT:
  case TCP_LAST_ACK:
    return TRUE;
  default:
    return FALSE;
  }
}

/* True iff the @state requires that a FIN has already been received from the
 * peer. */
static gboolean
pseudo_tcp_state_has_received_fin (PseudoTcpState state)
{
  switch (state) {
  case TCP_LISTEN:
  case TCP_SYN_SENT:
  case TCP_SYN_RECEIVED:
  case TCP_ESTABLISHED:
  case TCP_FIN_WAIT_1:
  case TCP_FIN_WAIT_2:
    return FALSE;
  case TCP_CLOSED:
  case TCP_CLOSING:
  case TCP_TIME_WAIT:
  case TCP_CLOSE_WAIT:
  case TCP_LAST_ACK:
    return TRUE;
  default:
    return FALSE;
  }
}

/* True iff the @state requires that a FIN-ACK has already been received from
 * the peer. */
static gboolean
pseudo_tcp_state_has_received_fin_ack (PseudoTcpState state)
{
  switch (state) {
  case TCP_LISTEN:
  case TCP_SYN_SENT:
  case TCP_SYN_RECEIVED:
  case TCP_ESTABLISHED:
  case TCP_FIN_WAIT_1:
  case TCP_FIN_WAIT_2:
  case TCP_CLOSING:
  case TCP_CLOSE_WAIT:
  case TCP_LAST_ACK:
    return FALSE;
  case TCP_CLOSED:
  case TCP_TIME_WAIT:
    return TRUE;
  default:
    return FALSE;
  }
}

static gboolean
process(PseudoTcpSocket *self, Segment *seg)
{
  PseudoTcpSocketPrivate *priv = self->priv;
  guint32 now;
  SendFlags sflags = sfNone;
  gboolean bIgnoreData;
  gboolean bNewData;
  gboolean bConnect = FALSE;
  gsize snd_buffered;
  gsize available_space;
  guint32 kIdealRefillSize;
  gboolean is_valuable_ack, is_duplicate_ack, is_fin_ack = FALSE;
  gboolean received_fin = FALSE;

  /* If this is the wrong conversation, send a reset!?!
     (with the correct conversation?) */
  if (seg->conv != priv->conv) {
    //if ((seg->flags & FLAG_RST) == 0) {
    //  packet(sock, tcb, seg->ack, 0, FLAG_RST, 0, 0);
    //}
    DEBUG (PSEUDO_TCP_DEBUG_NORMAL, "wrong conversation");
    return FALSE;
  }

  now = get_current_time (self);
  priv->last_traffic = priv->lastrecv = now;
  priv->bOutgoing = FALSE;

  if (priv->state == TCP_CLOSED ||
      (pseudo_tcp_state_has_received_fin_ack (priv->state) && seg->len > 0)) {
    /* Send an RST segment. See: RFC 1122, §4.2.2.13; RFC 793, §3.4, point 3,
     * page 37. We can only send RST if we know the peer knows we’re closed;
     * otherwise this could be a timeout retransmit from them, due to our
     * packets from data through to FIN being dropped. */
    DEBUG (PSEUDO_TCP_DEBUG_NORMAL,
        "Segment received while closed; sending RST.");
    if ((seg->flags & FLAG_RST) == 0) {
      closedown (self, 0, CLOSEDOWN_LOCAL);
    }

    return FALSE;
  }

  // Check if this is a reset segment
  if (seg->flags & FLAG_RST) {
    DEBUG (PSEUDO_TCP_DEBUG_NORMAL, "Received RST segment; closing down.");
    closedown (self, ECONNRESET, CLOSEDOWN_REMOTE);
    return FALSE;
  }

  // Check for control data
  bConnect = FALSE;
  if (seg->flags & FLAG_CTL) {
    if (seg->len == 0) {
      DEBUG (PSEUDO_TCP_DEBUG_NORMAL, "Missing control code");
      return FALSE;
    } else if (seg->data[0] == CTL_CONNECT) {
      bConnect = TRUE;

      parse_options (self, (guint8 *) &seg->data[1], seg->len - 1);

      if (priv->state == TCP_LISTEN) {
        set_state (self, TCP_SYN_RECEIVED);
        queue_connect_message (self);
      } else if (priv->state == TCP_SYN_SENT) {
        set_state_established (self);
      }
    } else {
      DEBUG (PSEUDO_TCP_DEBUG_NORMAL, "Unknown control code: %u", seg->data[0]);
      return FALSE;
    }
  }

  // Update timestamp
  if (SMALLER_OR_EQUAL (seg->seq, priv->ts_lastack) &&
      SMALLER (priv->ts_lastack, seg->seq + seg->len)) {
    priv->ts_recent = seg->tsval;
  }

  // Check if this is a valuable ack
  is_valuable_ack = (LARGER(seg->ack, priv->snd_una) &&
      SMALLER_OR_EQUAL(seg->ack, priv->snd_nxt));
  is_duplicate_ack = (seg->ack == priv->snd_una);

  if (is_valuable_ack) {
    guint32 nAcked;
    guint32 nFree;

    // Calculate round-trip time
    if (seg->tsecr) {
      long rtt = time_diff(now, seg->tsecr);
      if (rtt >= 0) {
        if (priv->rx_srtt == 0) {
          priv->rx_srtt = rtt;
          priv->rx_rttvar = rtt / 2;
        } else {
          priv->rx_rttvar = (3 * priv->rx_rttvar +
              labs((long)(rtt - priv->rx_srtt))) / 4;
          priv->rx_srtt = (7 * priv->rx_srtt + rtt) / 8;
        }
        priv->rx_rto = bound(MIN_RTO,
            priv->rx_srtt + max(1LU, 4 * priv->rx_rttvar), MAX_RTO);

        DEBUG (PSEUDO_TCP_DEBUG_VERBOSE, "rtt: %ld srtt: %u rttvar: %u rto: %u",
            rtt, priv->rx_srtt, priv->rx_rttvar, priv->rx_rto);
      } else {
        DEBUG (PSEUDO_TCP_DEBUG_NORMAL, "Invalid RTT: %ld", rtt);
        return FALSE;
      }
    }

    priv->snd_wnd = seg->wnd << priv->swnd_scale;

    nAcked = seg->ack - priv->snd_una;
    priv->snd_una = seg->ack;

    priv->rto_base = (priv->snd_una == priv->snd_nxt) ? 0 : now;

    /* ACKs for FIN segments give an increment on nAcked, but there is no
     * corresponding byte to read because the FIN segment is empty (it just has
     * a sequence number). */
    if (nAcked == priv->sbuf.data_length + 1 &&
        pseudo_tcp_state_has_sent_fin (priv->state)) {
      is_fin_ack = TRUE;
      nAcked--;
    }

    pseudo_tcp_fifo_consume_read_data (&priv->sbuf, nAcked);

    for (nFree = nAcked; nFree > 0; ) {
      SSegment *data;

      g_assert(g_queue_get_length (&priv->slist) != 0);
      data = (SSegment *) g_queue_peek_head (&priv->slist);

      if (nFree < data->len) {
        data->len -= nFree;
        data->seq += nFree;
        nFree = 0;
      } else {
        if (data->len > priv->largest) {
          priv->largest = data->len;
        }
        nFree -= data->len;
        g_slice_free (SSegment, data);
        g_queue_pop_head (&priv->slist);
      }
    }

    if (priv->dup_acks >= 3) {
      if (LARGER_OR_EQUAL (priv->snd_una, priv->recover)) { // NewReno
        guint32 nInFlight = priv->snd_nxt - priv->snd_una;
        // (Fast Retransmit)
        priv->cwnd = min(priv->ssthresh,
            max (nInFlight, priv->mss) + priv->mss);
        DEBUG (PSEUDO_TCP_DEBUG_NORMAL, "exit recovery cwnd=%d ssthresh=%d nInFlight=%d mss: %d", priv->cwnd, priv->ssthresh, nInFlight, priv->mss);
        priv->fast_recovery = FALSE;
        priv->dup_acks = 0;
      } else {
        int transmit_status;

        DEBUG (PSEUDO_TCP_DEBUG_NORMAL, "recovery retransmit");
        transmit_status = transmit(self, g_queue_peek_head (&priv->slist), now);
        if (transmit_status != 0) {
          DEBUG (PSEUDO_TCP_DEBUG_NORMAL,
              "Error transmitting recovery retransmit segment. Closing down.");
          closedown (self, transmit_status, CLOSEDOWN_LOCAL);
          return FALSE;
        }
        priv->cwnd += (nAcked > priv->mss ? priv->mss : 0) -
            min(nAcked, priv->cwnd);
      }
    } else {
      priv->dup_acks = 0;
      // Slow start, congestion avoidance
      if (priv->cwnd < priv->ssthresh) {
        priv->cwnd += priv->mss;
      } else {
        priv->cwnd += max(1LU, priv->mss * priv->mss / priv->cwnd);
      }
    }
  } else if (is_duplicate_ack) {
    /* !?! Note, tcp says don't do this... but otherwise how does a
       closed window become open? */
    priv->snd_wnd = seg->wnd << priv->swnd_scale;

    // Check duplicate acks
    if (seg->len > 0) {
      // it's a dup ack, but with a data payload, so don't modify priv->dup_acks
    } else if (priv->snd_una != priv->snd_nxt) {
      guint32 nInFlight;

      priv->dup_acks += 1;
      DEBUG (PSEUDO_TCP_DEBUG_VERBOSE, "Received dup ack (dups: %u)",
          priv->dup_acks);
      if (priv->dup_acks == 3) { // (Fast Retransmit)
        int transmit_status;


        if (LARGER_OR_EQUAL (priv->snd_una, priv->recover)) { /* NewReno */
          /* Invoke fast retransmit  RFC3782 section 3 step 1A*/
          DEBUG (PSEUDO_TCP_DEBUG_NORMAL, "enter recovery");
          DEBUG (PSEUDO_TCP_DEBUG_NORMAL, "recovery retransmit");

          transmit_status = transmit(self, g_queue_peek_head (&priv->slist),
              now);
          if (transmit_status != 0) {
            DEBUG (PSEUDO_TCP_DEBUG_NORMAL,
                "Error transmitting recovery retransmit segment. Closing down.");

            closedown (self, transmit_status, CLOSEDOWN_LOCAL);
            return FALSE;
          }
          priv->recover = priv->snd_nxt;
          nInFlight = priv->snd_nxt - priv->snd_una;
          priv->ssthresh = max(nInFlight / 2, 2 * priv->mss);
          DEBUG (PSEUDO_TCP_DEBUG_NORMAL,
              "ssthresh: %u = max((nInFlight: %u / 2), 2 * mss: %u)",
              priv->ssthresh, nInFlight, priv->mss);
          priv->cwnd = priv->ssthresh + 3 * priv->mss;
          priv->fast_recovery = TRUE;
        } else {
          DEBUG (PSEUDO_TCP_DEBUG_VERBOSE,
              "Skipping fast recovery: recover: %u snd_una: %u", priv->recover,
              priv->snd_una);
        }
      } else if (priv->dup_acks > 3) {
        if (priv->fast_recovery)
          priv->cwnd += priv->mss;
      }
    } else {
      priv->dup_acks = 0;
    }
  }

  // !?! A bit hacky
  if ((priv->state == TCP_SYN_RECEIVED) && !bConnect) {
    set_state_established (self);
  }

  /* Check for connection closure. Only pay attention to FIN segments if they
   * are in sequence; otherwise we’ve missed a packet earlier in the stream and
   * need to request retransmission first. */
  if (priv->support_fin_ack) {
    /* @received_fin is set when, and only when, all segments preceding the FIN
     * have been acknowledged. This is to handle the case where the FIN arrives
     * out of order with a preceding data segment. */
    if (seg->flags & FLAG_FIN && priv->rcv_fin == 0) {
      priv->rcv_fin = seg->seq;
      DEBUG (PSEUDO_TCP_DEBUG_NORMAL, "Setting rcv_fin = %u", priv->rcv_fin);
    } else if (seg->flags & FLAG_FIN && seg->seq != priv->rcv_fin) {
      DEBUG (PSEUDO_TCP_DEBUG_NORMAL, "Second FIN segment received; ignored");
      return FALSE;
    }

    /* For the moment, FIN segments must not contain data. */
    if (seg->flags & FLAG_FIN && seg->len != 0) {
      DEBUG (PSEUDO_TCP_DEBUG_NORMAL, "FIN segment contained data; ignored");
      return FALSE;
    }

    received_fin = (priv->rcv_nxt != 0 && priv->rcv_nxt + seg->len == priv->rcv_fin);

    /* Update the state machine, implementing all transitions on ‘rcv FIN’ or
     * ‘rcv ACK of FIN’ from RFC 793, Figure 6; and RFC 1122, §4.2.2.8. */
    switch (priv->state) {
    case TCP_ESTABLISHED:
      if (received_fin) {
        /* Received a FIN from the network, RFC 793, §3.5, Case 2.
         * The code below will send an ACK for the FIN. */
         set_state (self, TCP_CLOSE_WAIT);
      }
      break;
    case TCP_CLOSING:
      if (is_fin_ack) {
        /* Handle the ACK of a locally-sent FIN flag. RFC 793, §3.5, Case 3. */
        set_state (self, TCP_TIME_WAIT);
      }
      break;
    case TCP_LAST_ACK:
      if (is_fin_ack) {
        /* Handle the ACK of a locally-sent FIN flag. RFC 793, §3.5, Case 2. */
        set_state_closed (self, 0);
      }
      break;
    case TCP_FIN_WAIT_1:
      if (is_fin_ack && received_fin) {
        /* Simultaneous close with an ACK for a FIN previously sent,
         * RFC 793, §3.5, Case 3. */
        set_state (self, TCP_TIME_WAIT);
      } else if (is_fin_ack) {
        /* Handle the ACK of a locally-sent FIN flag. RFC 793, §3.5, Case 1. */
        set_state (self, TCP_FIN_WAIT_2);
      } else if (received_fin) {
        /* Simultaneous close, RFC 793, §3.5, Case 3. */
        set_state (self, TCP_CLOSING);
      }
      break;
    case TCP_FIN_WAIT_2:
      if (received_fin) {
        /* Local user closed the connection, RFC 793, §3.5, Case 1. */
        set_state (self, TCP_TIME_WAIT);
      }
      break;
    case TCP_LISTEN:
    case TCP_SYN_SENT:
    case TCP_SYN_RECEIVED:
    case TCP_TIME_WAIT:
    case TCP_CLOSED:
    case TCP_CLOSE_WAIT:
      /* Shouldn’t ever hit these cases. */
      if (received_fin) {
        DEBUG (PSEUDO_TCP_DEBUG_NORMAL,
           "Unexpected state %u when FIN received", priv->state);
      } else if (is_fin_ack) {
        DEBUG (PSEUDO_TCP_DEBUG_NORMAL,
           "Unexpected state %u when FIN-ACK received", priv->state);
      }
      break;
    default:
      DEBUG (PSEUDO_TCP_DEBUG_NORMAL, "Invalid state %u when FIN received",
          priv->state);
      return FALSE;
    }
  } else if (seg->flags & FLAG_FIN) {
    DEBUG (PSEUDO_TCP_DEBUG_NORMAL,
        "Invalid FIN received when FIN-ACK support is disabled");
  } else if (is_fin_ack) {
    DEBUG (PSEUDO_TCP_DEBUG_NORMAL,
        "Invalid FIN-ACK received when FIN-ACK support is disabled");
  }

  // If we make room in the send queue, notify the user
  // The goal it to make sure we always have at least enough data to fill the
  // window.  We'd like to notify the app when we are halfway to that point.
  kIdealRefillSize = (priv->sbuf_len + priv->rbuf_len) / 2;

  snd_buffered = pseudo_tcp_fifo_get_buffered (&priv->sbuf);
  if (priv->bWriteEnable && snd_buffered < kIdealRefillSize) {
    priv->bWriteEnable = FALSE;
    if (priv->callbacks.PseudoTcpWritable)
      priv->callbacks.PseudoTcpWritable(self, priv->callbacks.user_data);
  }

  /* Conditions where acks must be sent:
   * 1) Segment is too old (they missed an ACK) (immediately)
   * 2) Segment is too new (we missed a segment) (immediately)
   * 3) Segment has data (so we need to ACK!) (delayed)
   * ... so the only time we don't need to ACK, is an empty segment
   * that points to rcv_nxt!
   * 4) Segment has the FIN flag set (immediately) — note that the FIN flag
   *    itself has to be included in the ACK as a numbered byte;
   *    see RFC 793, §3.3. Also see: RFC 793, §3.5.
   */
  if (seg->seq != priv->rcv_nxt) {
    sflags = sfImmediateAck; // (Fast Recovery)
  } else if (seg->len != 0) {
    if (priv->ack_delay == 0) {
      sflags = sfImmediateAck;
    } else {
      sflags = sfDelayedAck;
    }
  } else if (received_fin) {
    /* FIN flags have a sequence number. Only acknowledge them after all
     * preceding octets have been acknowledged. */
    sflags = sfImmediateAck;
  }

  if (sflags == sfImmediateAck) {
    if (seg->seq > priv->rcv_nxt) {
      DEBUG (PSEUDO_TCP_DEBUG_NORMAL, "too new");
    } else if (SMALLER_OR_EQUAL(seg->seq + seg->len, priv->rcv_nxt)) {
      DEBUG (PSEUDO_TCP_DEBUG_NORMAL, "too old");
    }
  }

  // Adjust the incoming segment to fit our receive buffer
  if (SMALLER(seg->seq, priv->rcv_nxt)) {
    guint32 nAdjust = priv->rcv_nxt - seg->seq;
    if (nAdjust < seg->len) {
      seg->seq += nAdjust;
      seg->data += nAdjust;
      seg->len -= nAdjust;
    } else {
      seg->len = 0;
    }
  }

  available_space = pseudo_tcp_fifo_get_write_remaining (&priv->rbuf);

  if ((seg->seq + seg->len - priv->rcv_nxt) > available_space) {
    guint32 nAdjust = seg->seq + seg->len - priv->rcv_nxt - available_space;
    if (nAdjust < seg->len) {
      seg->len -= nAdjust;
    } else {
      seg->len = 0;
    }
  }

  bIgnoreData = (seg->flags & FLAG_CTL);
  if (!priv->support_fin_ack)
    bIgnoreData |= (priv->shutdown != SD_NONE);

  bNewData = FALSE;

  if (seg->len > 0) {
    if (bIgnoreData) {
      if (seg->seq == priv->rcv_nxt) {
        priv->rcv_nxt += seg->len;
      }
    } else {
      guint32 nOffset = seg->seq - priv->rcv_nxt;
      gsize res;

      res = pseudo_tcp_fifo_write_offset (&priv->rbuf, (guint8 *) seg->data,
          seg->len, nOffset);
      g_assert (res == seg->len);

      if (seg->seq == priv->rcv_nxt) {
        GList *iter = NULL;

        pseudo_tcp_fifo_consume_write_buffer (&priv->rbuf, seg->len);
        priv->rcv_nxt += seg->len;
        priv->rcv_wnd -= seg->len;
        bNewData = TRUE;

        iter = priv->rlist;
        while (iter &&
            SMALLER_OR_EQUAL(((RSegment *)iter->data)->seq, priv->rcv_nxt)) {
          RSegment *data = (RSegment *)(iter->data);
          if (LARGER (data->seq + data->len, priv->rcv_nxt)) {
            guint32 nAdjust = (data->seq + data->len) - priv->rcv_nxt;
            sflags = sfImmediateAck; // (Fast Recovery)
            DEBUG (PSEUDO_TCP_DEBUG_NORMAL, "Recovered %u bytes (%u -> %u)",
                nAdjust, priv->rcv_nxt, priv->rcv_nxt + nAdjust);
            pseudo_tcp_fifo_consume_write_buffer (&priv->rbuf, nAdjust);
            priv->rcv_nxt += nAdjust;
            priv->rcv_wnd -= nAdjust;
          }
          g_slice_free (RSegment, priv->rlist->data);
          priv->rlist = g_list_delete_link (priv->rlist, priv->rlist);
          iter = priv->rlist;
        }
      } else {
        GList *iter = NULL;
        RSegment *rseg = g_slice_new0 (RSegment);

        DEBUG (PSEUDO_TCP_DEBUG_NORMAL, "Saving %u bytes (%u -> %u)",
            seg->len, seg->seq, seg->seq + seg->len);
        rseg->seq = seg->seq;
        rseg->len = seg->len;
        iter = priv->rlist;
        while (iter && SMALLER (((RSegment*)iter->data)->seq, rseg->seq)) {
          iter = g_list_next (iter);
        }
        priv->rlist = g_list_insert_before(priv->rlist, iter, rseg);
      }
    }
  }

  if (received_fin) {
    /* FIN flags have a sequence number. */
    priv->rcv_nxt++;
  }


  attempt_send(self, sflags);

  // If we have new data, notify the user
  if (bNewData && priv->bReadEnable) {
    /* priv->bReadEnable = FALSE; — removed so that we’re always notified of
     * incoming pseudo-TCP data, rather than having to read the entire buffer
     * on each readable() callback before the next callback is enabled.
     * (When client-provided buffers are small, this is not possible.) */
    if (priv->callbacks.PseudoTcpReadable)
      priv->callbacks.PseudoTcpReadable(self, priv->callbacks.user_data);
  }

  return TRUE;
}

static gboolean
transmit(PseudoTcpSocket *self, SSegment *segment, guint32 now)
{
  PseudoTcpSocketPrivate *priv = self->priv;
  guint32 nTransmit = min(segment->len, priv->mss);

  if (segment->xmit >= ((priv->state == TCP_ESTABLISHED) ? 15 : 30)) {
    DEBUG (PSEUDO_TCP_DEBUG_NORMAL, "too many retransmits");
    return ETIMEDOUT;
  }

  while (TRUE) {
    guint32 seq = segment->seq;
    guint8 flags = segment->flags;
    PseudoTcpWriteResult wres;

    /* The packet must not have already been acknowledged. */
    g_assert_cmpuint (segment->seq - priv->snd_una, <=, 1024 * 1024 * 64);

    /* Write out the packet. */
    wres = packet(self, seq, flags,
        segment->seq - priv->snd_una, nTransmit, now);

    if (wres == WR_SUCCESS)
      break;

    if (wres == WR_FAIL) {
      DEBUG (PSEUDO_TCP_DEBUG_NORMAL, "packet failed");
      return ECONNABORTED;  /* FIXME: This error code doesn’t quite seem right */
    }

    g_assert(wres == WR_TOO_LARGE);

    while (TRUE) {
      if (PACKET_MAXIMUMS[priv->msslevel + 1] == 0) {
        DEBUG (PSEUDO_TCP_DEBUG_NORMAL, "MTU too small");
        return EMSGSIZE;
      }
      /* !?! We need to break up all outstanding and pending packets
         and then retransmit!?! */

      priv->mss = PACKET_MAXIMUMS[++priv->msslevel] - PACKET_OVERHEAD;
      // I added this... haven't researched actual formula
      priv->cwnd = 2 * priv->mss;

      if (priv->mss < nTransmit) {
        nTransmit = priv->mss;
        break;
      }
    }
    DEBUG (PSEUDO_TCP_DEBUG_NORMAL, "Adjusting mss to %u bytes ", priv->mss);
  }

  if (nTransmit < segment->len) {
    SSegment *subseg = g_slice_new0 (SSegment);
    subseg->seq = segment->seq + nTransmit;
    subseg->len = segment->len - nTransmit;
    subseg->flags = segment->flags;
    subseg->xmit = segment->xmit;

    DEBUG (PSEUDO_TCP_DEBUG_NORMAL, "mss reduced to %u", priv->mss);

    segment->len = nTransmit;
    g_queue_insert_after (&priv->slist,
        g_queue_find (&priv->slist, segment), subseg);
    if (subseg->xmit == 0)
      g_queue_insert_after (&priv->unsent_slist,
          g_queue_find (&priv->unsent_slist, segment), subseg);
  }

  if (segment->xmit == 0) {
    g_assert (g_queue_peek_head (&priv->unsent_slist) == segment);
    g_queue_pop_head (&priv->unsent_slist);
    priv->snd_nxt += segment->len;

    /* FIN flags require acknowledgement. */
    if (segment->len == 0 && segment->flags & FLAG_FIN)
      priv->snd_nxt++;
  }
  segment->xmit += 1;

  if (priv->rto_base == 0) {
    priv->rto_base = now;
  }

  return 0;
}

static void
attempt_send(PseudoTcpSocket *self, SendFlags sflags)
{
  PseudoTcpSocketPrivate *priv = self->priv;
  guint32 now = get_current_time (self);
  gboolean bFirst = TRUE;

  DEBUG (PSEUDO_TCP_DEBUG_NORMAL, "Attempting send with flags %u.", sflags);

  if (time_diff(now, priv->lastsend) > (long) priv->rx_rto) {
    priv->cwnd = priv->mss;
  }


  while (TRUE) {
    guint32 cwnd;
    guint32 nWindow;
    guint32 nInFlight;
    guint32 nUseable;
    guint32 nAvailable;
    gsize snd_buffered;
    GList *iter;
    SSegment *sseg;
    int transmit_status;

    cwnd = priv->cwnd;
    if ((priv->dup_acks == 1) || (priv->dup_acks == 2)) { // Limited Transmit
      cwnd += priv->dup_acks * priv->mss;
    }
    nWindow = min(priv->snd_wnd, cwnd);
    nInFlight = priv->snd_nxt - priv->snd_una;
    nUseable = (nInFlight < nWindow) ? (nWindow - nInFlight) : 0;
    snd_buffered = pseudo_tcp_fifo_get_buffered (&priv->sbuf);
    if (snd_buffered < nInFlight)  /* iff a FIN has been sent */
      nAvailable = 0;
    else
      nAvailable = min(snd_buffered - nInFlight, priv->mss);

    if (nAvailable > nUseable) {
      if (nUseable * 4 < nWindow) {
        // RFC 813 - avoid SWS
        nAvailable = 0;
      } else {
        nAvailable = nUseable;
      }
    }

    if (bFirst) {
      gsize available_space = pseudo_tcp_fifo_get_write_remaining (&priv->sbuf);

      bFirst = FALSE;
      DEBUG (PSEUDO_TCP_DEBUG_VERBOSE, "[cwnd: %u  nWindow: %u  nInFlight: %u "
          "nAvailable: %u nQueued: %" G_GSIZE_FORMAT " nEmpty: %" G_GSIZE_FORMAT
          "  nWaiting: %zu ssthresh: %u]",
          priv->cwnd, nWindow, nInFlight, nAvailable, snd_buffered,
          available_space, snd_buffered - nInFlight, priv->ssthresh);
    }

    if (nAvailable == 0 && sflags != sfFin && sflags != sfRst) {
      if (sflags == sfNone)
        return;

      // If this is an immediate ack, or the second delayed ack
      if ((sflags == sfImmediateAck) || priv->t_ack) {
        packet(self, priv->snd_nxt, 0, 0, 0, now);
      } else {
        priv->t_ack = now;
      }
      return;
    }

    // Nagle algorithm
    // If there is data already in-flight, and we haven't a full segment of
    // data ready to send then hold off until we get more to send, or the
    // in-flight data is acknowledged.
    if (priv->use_nagling && sflags != sfFin && sflags != sfRst &&
        (priv->snd_nxt > priv->snd_una) &&
        (nAvailable < priv->mss))  {
      return;
    }

    // Find the next segment to transmit
    iter = g_queue_peek_head_link (&priv->unsent_slist);
    if (iter == NULL)
      return;
    sseg = iter->data;

    // If the segment is too large, break it into two
    if (sseg->len > nAvailable && sflags != sfFin && sflags != sfRst) {
      SSegment *subseg = g_slice_new0 (SSegment);
      subseg->seq = sseg->seq + nAvailable;
      subseg->len = sseg->len - nAvailable;
      subseg->flags = sseg->flags;

      sseg->len = nAvailable;
      g_queue_insert_after (&priv->unsent_slist, iter, subseg);
      g_queue_insert_after (&priv->slist, g_queue_find (&priv->slist, sseg),
          subseg);
    }

    transmit_status = transmit(self, sseg, now);
    if (transmit_status != 0) {
      DEBUG (PSEUDO_TCP_DEBUG_NORMAL, "transmit failed");

      // TODO: Is this the right thing ?
      closedown (self, transmit_status, CLOSEDOWN_REMOTE);
      return;
    }

    if (sflags == sfImmediateAck || sflags == sfDelayedAck)
      sflags = sfNone;
  }
}

/* If @source is %CLOSEDOWN_REMOTE, don’t send an RST packet, since closedown()
 * has been called as a result of an RST segment being received.
 * See: RFC 1122, §4.2.2.13. */
static void
closedown (PseudoTcpSocket *self, guint32 err, ClosedownSource source)
{
  PseudoTcpSocketPrivate *priv = self->priv;

  DEBUG (PSEUDO_TCP_DEBUG_NORMAL, "Closing down socket %p with %s error %u.",
      self, (source == CLOSEDOWN_LOCAL) ? "local" : "remote", err);

  if (source == CLOSEDOWN_LOCAL && priv->support_fin_ack) {
    queue_rst_message (self);
    attempt_send (self, sfRst);
  } else if (source == CLOSEDOWN_LOCAL) {
    priv->shutdown = SD_FORCEFUL;
  }

  /* ‘Cute’ little navigation through the state machine to avoid breaking the
   * invariant that CLOSED can only be reached from TIME-WAIT or LAST-ACK. */
  switch (priv->state) {
  case TCP_LISTEN:
  case TCP_SYN_SENT:
    break;
  case TCP_SYN_RECEIVED:
  case TCP_ESTABLISHED:
    set_state (self, TCP_FIN_WAIT_1);
    /* Fall through. */
  case TCP_FIN_WAIT_1:
    set_state (self, TCP_FIN_WAIT_2);
    /* Fall through. */
  case TCP_FIN_WAIT_2:
  case TCP_CLOSING:
    set_state (self, TCP_TIME_WAIT);
    break;
  case TCP_CLOSE_WAIT:
    set_state (self, TCP_LAST_ACK);
    break;
  case TCP_LAST_ACK:
  case TCP_TIME_WAIT:
  case TCP_CLOSED:
  default:
    break;
  }

  set_state_closed (self, err);
}

static void
adjustMTU(PseudoTcpSocket *self)
{
  PseudoTcpSocketPrivate *priv = self->priv;

  // Determine our current mss level, so that we can adjust appropriately later
  for (priv->msslevel = 0;
       PACKET_MAXIMUMS[priv->msslevel + 1] > 0;
       ++priv->msslevel) {
    if (((guint16)PACKET_MAXIMUMS[priv->msslevel]) <= priv->mtu_advise) {
      break;
    }
  }
  priv->mss = priv->mtu_advise - PACKET_OVERHEAD;
  // !?! Should we reset priv->largest here?
  DEBUG (PSEUDO_TCP_DEBUG_NORMAL, "Adjusting mss to %u bytes", priv->mss);
  // Enforce minimums on ssthresh and cwnd
  priv->ssthresh = max(priv->ssthresh, 2 * priv->mss);
  priv->cwnd = max(priv->cwnd, priv->mss);
}

static void
apply_window_scale_option (PseudoTcpSocket *self, guint8 scale_factor)
{
   PseudoTcpSocketPrivate *priv = self->priv;

   priv->swnd_scale = scale_factor;
   DEBUG (PSEUDO_TCP_DEBUG_NORMAL, "Setting scale factor to %u", scale_factor);
}

static void
apply_fin_ack_option (PseudoTcpSocket *self)
{
  PseudoTcpSocketPrivate *priv = self->priv;

  priv->support_fin_ack = TRUE;
}

static void
apply_option (PseudoTcpSocket *self, guint8 kind, const guint8 *data,
    guint32 len)
{
  switch (kind) {
  case TCP_OPT_MSS:
    DEBUG (PSEUDO_TCP_DEBUG_NORMAL,
        "Peer specified MSS option which is not supported.");
    // TODO: Implement.
    break;
  case TCP_OPT_WND_SCALE:
    // Window scale factor.
    // http://www.ietf.org/rfc/rfc1323.txt
    if (len != 1) {
      DEBUG (PSEUDO_TCP_DEBUG_NORMAL, "Invalid window scale option received.");
      return;
    }
    apply_window_scale_option(self, data[0]);
    break;
  case TCP_OPT_FIN_ACK:
    // FIN-ACK support.
    DEBUG (PSEUDO_TCP_DEBUG_NORMAL, "FIN-ACK support enabled.");
    apply_fin_ack_option (self);
    break;
  case TCP_OPT_EOL:
  case TCP_OPT_NOOP:
    /* Nothing to do. */
    break;
  default:
    DEBUG (PSEUDO_TCP_DEBUG_NORMAL, "Invalid TCP option %u", kind);
    break;
  }
}


static void
parse_options (PseudoTcpSocket *self, const guint8 *data, guint32 len)
{
  PseudoTcpSocketPrivate *priv = self->priv;
  gboolean has_window_scaling_option = FALSE;
  gboolean has_fin_ack_option = FALSE;
  guint32 pos = 0;

  // See http://www.freesoft.org/CIE/Course/Section4/8.htm for
  // parsing the options list.
  while (pos < len) {
    guint8 kind = TCP_OPT_EOL;
    guint8 opt_len;

    if (len < pos + 1)
      return;

    kind = data[pos];
    pos++;

    if (kind == TCP_OPT_EOL) {
      // End of option list.
      break;
    } else if (kind == TCP_OPT_NOOP) {
      // No op.
      continue;
    }

    if (len < pos + 1)
      return;

    // Length of this option.
    opt_len = data[pos];
    pos++;

    if (len < pos + opt_len)
      return;

    // Content of this option.
    if (opt_len <= len - pos) {
      apply_option (self, kind, data + pos, opt_len);
      pos += opt_len;
    } else {
      DEBUG (PSEUDO_TCP_DEBUG_NORMAL, "Invalid option length received.");
      return;
    }

    if (kind == TCP_OPT_WND_SCALE)
      has_window_scaling_option = TRUE;
    else if (kind == TCP_OPT_FIN_ACK)
      has_fin_ack_option = TRUE;
  }

  if (!has_window_scaling_option) {
    DEBUG (PSEUDO_TCP_DEBUG_NORMAL, "Peer doesn't support window scaling");
    if (priv->rwnd_scale > 0) {
      // Peer doesn't support TCP options and window scaling.
      // Revert receive buffer size to default value.
      resize_receive_buffer (self, DEFAULT_RCV_BUF_SIZE);
      priv->swnd_scale = 0;
    }
  }

  if (!has_fin_ack_option) {
    DEBUG (PSEUDO_TCP_DEBUG_NORMAL, "Peer doesn't support FIN-ACK");
    priv->support_fin_ack = FALSE;
  }
}

static void
resize_send_buffer (PseudoTcpSocket *self, guint32 new_size)
{
  PseudoTcpSocketPrivate *priv = self->priv;

  priv->sbuf_len = new_size;
  pseudo_tcp_fifo_set_capacity (&priv->sbuf, new_size);
}


static void
resize_receive_buffer (PseudoTcpSocket *self, guint32 new_size)
{
  PseudoTcpSocketPrivate *priv = self->priv;
  guint8 scale_factor = 0;
  gboolean result;
  gsize available_space;

  if (priv->rbuf_len == new_size)
    return;

  // Determine the scale factor such that the scaled window size can fit
  // in a 16-bit unsigned integer.
  while (new_size > 0xFFFF) {
    ++scale_factor;
    new_size >>= 1;
  }

  // Determine the proper size of the buffer.
  new_size <<= scale_factor;
  result = pseudo_tcp_fifo_set_capacity (&priv->rbuf, new_size);

  // Make sure the new buffer is large enough to contain data in the old
  // buffer. This should always be true because this method is called either
  // before connection is established or when peers are exchanging connect
  // messages.
  g_assert(result);
  priv->rbuf_len = new_size;
  priv->rwnd_scale = scale_factor;
  priv->ssthresh = new_size;

  available_space = pseudo_tcp_fifo_get_write_remaining (&priv->rbuf);
  priv->rcv_wnd = available_space;
}

gint
pseudo_tcp_socket_get_available_bytes (PseudoTcpSocket *self)
{
  PseudoTcpSocketPrivate *priv = self->priv;

  return pseudo_tcp_fifo_get_buffered (&priv->rbuf);
}

gboolean
pseudo_tcp_socket_can_send (PseudoTcpSocket *self)
{
  return (pseudo_tcp_socket_get_available_send_space (self) > 0);
}

gsize
pseudo_tcp_socket_get_available_send_space (PseudoTcpSocket *self)
{
  PseudoTcpSocketPrivate *priv = self->priv;
  gsize ret;

  if (!pseudo_tcp_state_has_sent_fin (priv->state)) {
    ret = pseudo_tcp_fifo_get_write_remaining (&priv->sbuf);
  } else {
    ret = 0;
  }

  if (ret == 0)
    priv->bWriteEnable = TRUE;

  return ret;
}

/* State names are capitalised and formatted as in RFC 793. */
static const gchar *
pseudo_tcp_state_get_name (PseudoTcpState state)
{
  switch (state) {
  case TCP_LISTEN: return "LISTEN";
  case TCP_SYN_SENT: return "SYN-SENT";
  case TCP_SYN_RECEIVED: return "SYN-RECEIVED";
  case TCP_ESTABLISHED: return "ESTABLISHED";
  case TCP_CLOSED: return "CLOSED";
  case TCP_FIN_WAIT_1: return "FIN-WAIT-1";
  case TCP_FIN_WAIT_2: return "FIN-WAIT-2";
  case TCP_CLOSING: return "CLOSING";
  case TCP_TIME_WAIT: return "TIME-WAIT";
  case TCP_CLOSE_WAIT: return "CLOSE-WAIT";
  case TCP_LAST_ACK: return "LAST-ACK";
  default: return "UNKNOWN";
  }
}

static void
set_state (PseudoTcpSocket *self, PseudoTcpState new_state)
{
  PseudoTcpSocketPrivate *priv = self->priv;
  PseudoTcpState old_state = priv->state;

  if (new_state == old_state)
    return;

  DEBUG (PSEUDO_TCP_DEBUG_NORMAL, "State %s → %s.",
      pseudo_tcp_state_get_name (old_state),
      pseudo_tcp_state_get_name (new_state));

  /* Check whether it’s a valid state transition. */
#define TRANSITION(OLD, NEW) \
    (old_state == TCP_##OLD && \
     new_state == TCP_##NEW)

  /* Valid transitions. See: RFC 793, p23; RFC 1122, §4.2.2.8. */
  g_assert (/* RFC 793, p23. */
            TRANSITION (CLOSED, SYN_SENT) ||
            TRANSITION (SYN_SENT, CLOSED) ||
            TRANSITION (CLOSED, LISTEN) ||
            TRANSITION (LISTEN, CLOSED) ||
            TRANSITION (LISTEN, SYN_SENT) ||
            TRANSITION (LISTEN, SYN_RECEIVED) ||
            TRANSITION (SYN_SENT, SYN_RECEIVED) ||
            TRANSITION (SYN_RECEIVED, ESTABLISHED) ||
            TRANSITION (SYN_SENT, ESTABLISHED) ||
            TRANSITION (SYN_RECEIVED, FIN_WAIT_1) ||
            TRANSITION (ESTABLISHED, FIN_WAIT_1) ||
            TRANSITION (ESTABLISHED, CLOSE_WAIT) ||
            TRANSITION (FIN_WAIT_1, FIN_WAIT_2) ||
            TRANSITION (FIN_WAIT_1, CLOSING) ||
            TRANSITION (CLOSE_WAIT, LAST_ACK) ||
            TRANSITION (FIN_WAIT_2, TIME_WAIT) ||
            TRANSITION (CLOSING, TIME_WAIT) ||
            TRANSITION (LAST_ACK, CLOSED) ||
            TRANSITION (TIME_WAIT, CLOSED) ||
            /* RFC 1122, §4.2.2.8. */
            TRANSITION (SYN_RECEIVED, LISTEN) ||
            TRANSITION (FIN_WAIT_1, TIME_WAIT));

#undef TRANSITION

  priv->state = new_state;
}

static void
set_state_established (PseudoTcpSocket *self)
{
  PseudoTcpSocketPrivate *priv = self->priv;

  set_state (self, TCP_ESTABLISHED);

  adjustMTU (self);
  if (priv->callbacks.PseudoTcpOpened)
    priv->callbacks.PseudoTcpOpened (self, priv->callbacks.user_data);
}

/* (err == 0) means no error. */
static void
set_state_closed (PseudoTcpSocket *self, guint32 err)
{
  PseudoTcpSocketPrivate *priv = self->priv;

  set_state (self, TCP_CLOSED);

  /* Only call the callback if there was an error. */
  if (priv->callbacks.PseudoTcpClosed && err != 0)
    priv->callbacks.PseudoTcpClosed (self, err, priv->callbacks.user_data);
}

gboolean
pseudo_tcp_socket_is_closed (PseudoTcpSocket *self)
{
  PseudoTcpSocketPrivate *priv = self->priv;

  return (priv->state == TCP_CLOSED);
}

gboolean
pseudo_tcp_socket_is_closed_remotely (PseudoTcpSocket *self)
{
  PseudoTcpSocketPrivate *priv = self->priv;

  return pseudo_tcp_state_has_received_fin (priv->state);
}