premiere-libtorrent/src/utp_stream.cpp

/*

Copyright (c) 2009-2018, Arvid Norberg
All rights reserved.

Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions
are met:

    * Redistributions of source code must retain the above copyright
      notice, this list of conditions and the following disclaimer.
    * 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.
    * Neither the name of the author nor the names of its
      contributors may be used to endorse or promote products derived
      from this software without specific prior written permission.

THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "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 COPYRIGHT OWNER OR CONTRIBUTORS 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 "libtorrent/config.hpp"
#include "libtorrent/utp_stream.hpp"
#include "libtorrent/sliding_average.hpp"
#include "libtorrent/utp_socket_manager.hpp"
#include "libtorrent/aux_/alloca.hpp"
#include "libtorrent/timestamp_history.hpp"
#include "libtorrent/error.hpp"
#include "libtorrent/random.hpp"
#include "libtorrent/invariant_check.hpp"
#include "libtorrent/performance_counters.hpp"
#include "libtorrent/io_service.hpp"
#include <cstdint>
#include <limits>

#if TORRENT_USE_INVARIANT_CHECKS
#include <numeric> // for accumulate
#endif

#if TORRENT_UTP_LOG
#include <cstdarg>
#include <cinttypes> // for PRId64 et.al.
#include "libtorrent/socket_io.hpp"
#endif

namespace libtorrent {

#if TORRENT_UTP_LOG

static char const* packet_type_names[] = { "ST_DATA", "ST_FIN", "ST_STATE", "ST_RESET", "ST_SYN" };
static char const* socket_state_names[] = { "NONE", "SYN_SENT", "CONNECTED", "FIN_SENT", "ERROR", "DELETE" };

static struct utp_logger
{
	FILE* utp_log_file;
	std::mutex utp_log_mutex;

	utp_logger() : utp_log_file(nullptr) {}
	~utp_logger()
	{
		if (utp_log_file) fclose(utp_log_file);
	}
} log_file_holder;

TORRENT_FORMAT(1, 2)
void utp_log(char const* fmt, ...)
{
	if (log_file_holder.utp_log_file == nullptr) return;

	std::lock_guard<std::mutex> lock(log_file_holder.utp_log_mutex);
	static time_point start = clock_type::now();
	std::fprintf(log_file_holder.utp_log_file, "[%012" PRId64 "] ", total_microseconds(clock_type::now() - start));
	va_list l;
	va_start(l, fmt);
	vfprintf(log_file_holder.utp_log_file, fmt, l);
	va_end(l);
}

bool is_utp_stream_logging() {
	return log_file_holder.utp_log_file != nullptr;
}

void set_utp_stream_logging(bool enable) {
	if (enable)
	{
		if (log_file_holder.utp_log_file == nullptr)
		{
			log_file_holder.utp_log_file = fopen("utp.log", "w+");
		}
	}
	else
	{
		if (log_file_holder.utp_log_file != nullptr)
		{
			FILE* f = log_file_holder.utp_log_file;
			log_file_holder.utp_log_file = nullptr;
			fclose(f);
		}
	}
}

#define UTP_LOG utp_log
#if TORRENT_VERBOSE_UTP_LOG
#define UTP_LOGV utp_log
#else
#define UTP_LOGV TORRENT_WHILE_0 printf
#endif

#else

#define UTP_LOG(...) do {} while(false)
#define UTP_LOGV(...) do {} while(false)

#endif

enum
{
	ACK_MASK = 0xffff,

	// if a packet receives more than this number of
	// duplicate acks, we'll trigger a fast re-send
	dup_ack_limit = 3
};

// compare if lhs is less than rhs, taking wrapping
// into account. if lhs is close to UINT_MAX and rhs
// is close to 0, lhs is assumed to have wrapped and
// considered smaller
bool compare_less_wrap(std::uint32_t lhs
	, std::uint32_t rhs, std::uint32_t mask)
{
	// distance walking from lhs to rhs, downwards
	std::uint32_t dist_down = (lhs - rhs) & mask;
	// distance walking from lhs to rhs, upwards
	std::uint32_t dist_up = (rhs - lhs) & mask;

	// if the distance walking up is shorter, lhs
	// is less than rhs. If the distance walking down
	// is shorter, then rhs is less than lhs
	return dist_up < dist_down;
}

// since the uTP socket state may be needed after the
// utp_stream is closed, it's kept in a separate struct
// whose lifetime is not tied to the lifetime of utp_stream

// the utp socket is closely modelled after the asio async
// operations and handler model. For writing to the socket,
// the client provides a list of buffers (for gather/writev
// style of I/O) and whenever the socket can write another
// packet to the stream, it picks up data from these buffers.
// When all of the data has been written, or enough time has
// passed since we first started writing, the write handler
// is called and the write buffer is reset. This means that
// we're not writing anything at all while waiting for the
// client to re-issue a write request.

// reading is a little bit more complicated, since we must
// be able to receive data even when the user doesn't have
// an outstanding read operation on the socket. When the user
// does however, we want to receive data directly into the
// user's buffer instead of first copying it into our receive
// buffer. This is why the receive case is more complicated.
// There are two receive buffers. One provided by the user,
// which when present is always used. The other one is used
// when the user doesn't have an outstanding read request,
// and hence hasn't provided any buffer space to receive into.

// the user provided read buffer is called "m_read_buffer" and
// its size is "m_read_buffer_size". The buffer we spill over
// into when the user provided buffer is full or when there
// is none, is "m_receive_buffer" and "m_receive_buffer_size"
// respectively.

// in order to know when to trigger the read and write handlers
// there are two counters, m_read and m_written, which count
// the number of bytes we've stuffed into the user provided
// read buffer or written to the stream from the write buffer.
// These are used to trigger the handlers if we're written a
// large number of bytes. It's also triggered if we're filled
// the whole read buffer, or written the entire write buffer.
// The last way the handlers can be triggered is if we're read
// or written some, and enough time has elapsed since then.

// when we receive data into m_receive_buffer (i.e. the buffer
// used when there's no user provided one) is stored as a
// number of heap allocated packets. This is just because it's
// simple to reuse the data structured and it provides all the
// functionality needed for this buffer.

struct utp_socket_impl
{
	utp_socket_impl(std::uint16_t recv_id, std::uint16_t send_id
		, void* userdata, utp_socket_manager& sm)
		: m_sm(sm)
		, m_userdata(userdata)
		, m_timeout(clock_type::now() + milliseconds(m_sm.connect_timeout()))
		, m_send_id(send_id)
		, m_recv_id(recv_id)
		, m_delay_sample_idx(0)
		, m_state(UTP_STATE_NONE)
		, m_eof(false)
		, m_attached(true)
		, m_nagle(true)
		, m_slow_start(true)
		, m_cwnd_full(false)
		, m_null_buffers(false)
		, m_deferred_ack(false)
		, m_subscribe_drained(false)
		, m_stalled(false)
		, m_confirmed(false)
	{
		TORRENT_ASSERT((m_recv_id == ((m_send_id + 1) & 0xffff))
			|| (m_send_id == ((m_recv_id + 1) & 0xffff)));
		m_sm.inc_stats_counter(counters::num_utp_idle);
		TORRENT_ASSERT(m_userdata);
		m_delay_sample_hist.fill(std::numeric_limits<std::uint32_t>::max());
	}

	~utp_socket_impl();

	void tick(time_point now);
	void init_mtu(int link_mtu, int utp_mtu);
	bool incoming_packet(span<std::uint8_t const> buf
		, udp::endpoint const& ep, time_point receive_time);
	void writable();

	bool should_delete() const;
	tcp::endpoint remote_endpoint(error_code& ec) const
	{
		if (m_state == UTP_STATE_NONE)
			ec = boost::asio::error::not_connected;
		else
			TORRENT_ASSERT(m_remote_address != address_v4::any());
		return tcp::endpoint(m_remote_address, m_port);
	}
	std::size_t available() const;
	// returns true if there were handlers cancelled
	// if it returns false, we can detach immediately
	bool destroy();
	void set_close_reason(close_reason_t code);
	void detach();
	void send_syn();
	void send_fin();

	void subscribe_drained();
	void defer_ack();
	void remove_sack_header(packet* p);

	enum packet_flags_t { pkt_ack = 1, pkt_fin = 2 };
	bool send_pkt(int flags = 0);
	bool resend_packet(packet* p, bool fast_resend = false);
	void send_reset(utp_header const* ph);
	std::pair<std::uint32_t, int> parse_sack(std::uint16_t packet_ack, std::uint8_t const* ptr
		, int size, time_point now);
	void parse_close_reason(std::uint8_t const* ptr, int size);
	void write_payload(std::uint8_t* ptr, int size);
	void maybe_inc_acked_seq_nr();
	std::uint32_t ack_packet(packet_ptr p, time_point receive_time
		, std::uint16_t seq_nr);
	void write_sack(std::uint8_t* buf, int size) const;
	void incoming(std::uint8_t const* buf, int size, packet_ptr p, time_point now);
	void do_ledbat(int acked_bytes, int delay, int in_flight);
	int packet_timeout() const;
	bool test_socket_state();
	void maybe_trigger_receive_callback();
	void maybe_trigger_send_callback();
	bool cancel_handlers(error_code const& ec, bool shutdown);
	bool consume_incoming_data(
		utp_header const* ph, std::uint8_t const* ptr, int payload_size, time_point now);
	void update_mtu_limits();
	void experienced_loss(std::uint32_t seq_nr, time_point now);

	void set_state(int s);

	packet_ptr acquire_packet(int const allocate) { return m_sm.acquire_packet(allocate); }
	void release_packet(packet_ptr p) { m_sm.release_packet(std::move(p)); }

	// non-copyable
	utp_socket_impl(utp_socket_impl const&) = delete;
	utp_socket_impl const& operator=(utp_socket_impl const&) = delete;

	// TODO: 2 it would be nice if not everything would have to be public here

#if TORRENT_USE_INVARIANT_CHECKS
	void check_receive_buffers() const;
	void check_invariant() const;
#endif

	utp_socket_manager& m_sm;
	std::weak_ptr<utp_socket_interface> m_sock;

	// userdata pointer passed along
	// with any callback. This is initialized to 0
	// then set to point to the utp_stream when
	// hooked up, and then reset to 0 once the utp_stream
	// detaches. This is used to know whether or not
	// the socket impl is still attached to a utp_stream
	// object. When it isn't, we'll never be able to
	// signal anything back to the client, and in case
	// of errors, we just have to delete ourselves
	// i.e. transition to the UTP_STATE_DELETED state
	void* m_userdata;

	// This is a platform-independent replacement
	// for the regular iovec type in posix. Since
	// it's not used in any system call, we might as
	// well define our own type instead of wrapping
	// the system's type.
	struct iovec_t
	{
		iovec_t(void* b, std::size_t l): buf(b), len(l) {}
		void* buf;
		std::size_t len;
	};

	// if there's currently an async read or write
	// operation in progress, these buffers are initialized
	// and used, otherwise any bytes received are stuck in
	// m_receive_buffer until another read is made
	// as we flush from the write buffer, individual iovecs
	// are updated to only refer to unflushed portions of the
	// buffers. Buffers that empty are erased from the vector.
	std::vector<iovec_t> m_write_buffer;

	// if this is non nullptr, it's a packet. This packet was held off because
	// of NAGLE. We couldn't send it immediately. It's left
	// here to accrue more bytes before we send it.
	packet_ptr m_nagle_packet;

	// the user provided read buffer. If this has a size greater
	// than 0, we'll always prefer using it over putting received
	// data in the m_receive_buffer. As data is stored in the
	// read buffer, the iovec_t elements are adjusted to only
	// refer to the unwritten portions of the buffers, and the
	// ones that fill up are erased from the vector
	std::vector<iovec_t> m_read_buffer;

	// packets we've received without a read operation
	// active. Store them here until the client triggers
	// an async_read_some
	std::vector<packet_ptr> m_receive_buffer;

	// this is the error on this socket. If m_state is
	// set to UTP_STATE_ERROR_WAIT, this error should be
	// forwarded to the client as soon as we have a new
	// async operation initiated
	error_code m_error;

	// these indicate whether or not there is an outstanding read/write or
	// connect operation. i.e. is there upper layer subscribed to these events.
	bool m_read_handler = false;
	bool m_write_handler = false;
	bool m_connect_handler = false;

	// the address of the remote endpoint
	address m_remote_address;

	// the send and receive buffers
	// maps packet sequence numbers
	packet_buffer m_inbuf;
	packet_buffer m_outbuf;

	// the time when the last packet we sent times out. Including re-sends.
	// if we ever end up not having sent anything in one second (
	// or one mean rtt + 2 average deviations, whichever is greater)
	// we set our cwnd to 1 MSS. This condition can happen either because
	// a packet has timed out and needs to be resent or because our
	// cwnd is set to less than one MSS during congestion control.
	// it can also happen if the other end sends an advertised window
	// size less than one MSS.
	time_point m_timeout;

	// the last time we stepped the timestamp history
	time_point m_last_history_step = clock_type::now();

	// the next time we allow a lost packet to halve cwnd. We only do this once every
	// 100 ms
	time_point m_next_loss;

	// the max number of bytes in-flight. This is a fixed point
	// value, to get the true number of bytes, shift right 16 bits
	// the value is always >= 0, but the calculations performed on
	// it in do_ledbat() are signed.
	std::int64_t m_cwnd = TORRENT_ETHERNET_MTU << 16;

	timestamp_history m_delay_hist;
	timestamp_history m_their_delay_hist;

	// the slow-start threshold. This is the congestion window size (m_cwnd)
	// in bytes the last time we left slow-start mode. This is used as a
	// threshold to leave slow-start earlier next time, to avoid packet-loss
	std::int32_t m_ssthres = 0;

	// the number of bytes we have buffered in m_inbuf
	std::int32_t m_buffered_incoming_bytes = 0;

	// the timestamp diff in the last packet received
	// this is what we'll send back
	std::uint32_t m_reply_micro = 0;

	// this is the advertised receive window the other end sent
	// we'll never have more un-acked bytes in flight
	// if this ever gets set to zero, we'll try one packet every
	// second until the window opens up again
	std::uint32_t m_adv_wnd = TORRENT_ETHERNET_MTU;

	// the number of un-acked bytes we have sent
	std::int32_t m_bytes_in_flight = 0;

	// the number of bytes read into the user provided
	// buffer. If this grows too big, we'll trigger the
	// read handler.
	std::int32_t m_read = 0;

	// the sum of the lengths of all iovec in m_write_buffer
	std::int32_t m_write_buffer_size = 0;

	// the number of bytes already written to packets
	// from m_write_buffer
	std::int32_t m_written = 0;

	// the sum of all packets stored in m_receive_buffer
	std::int32_t m_receive_buffer_size = 0;

	// the sum of all buffers in m_read_buffer
	std::int32_t m_read_buffer_size = 0;

	// max number of bytes to allocate for receive buffer
	std::int32_t m_receive_buffer_capacity = 1024 * 1024;

	// this holds the 3 last delay measurements,
	// these are the actual corrected delay measurements.
	// the lowest of the 3 last ones is used in the congestion
	// controller. This is to not completely close the cwnd
	// by a single outlier.
	std::array<std::uint32_t, 3> m_delay_sample_hist;

	// counters
	std::uint32_t m_in_packets = 0;
	std::uint32_t m_out_packets = 0;

	// the last send delay sample
	std::int32_t m_send_delay = 0;
	// the last receive delay sample
	std::int32_t m_recv_delay = 0;

	// average RTT
	sliding_average<int, 16> m_rtt;

	// if this is != 0, it means the upper layer provided a reason for why
	// the connection is being closed. The reason is indicated by this
	// non-zero value which is included in a packet header extension
	close_reason_t m_close_reason = close_reason_t::none;

	// port of destination endpoint
	std::uint16_t m_port = 0;

	std::uint16_t m_send_id;
	std::uint16_t m_recv_id;

	// this is the ack we're sending back. We have
	// received all packets up to this sequence number
	std::uint16_t m_ack_nr = 0;

	// the sequence number of the next packet
	// we'll send
	std::uint16_t m_seq_nr = 0;

	// this is the sequence number of the packet that
	// everything has been ACKed up to. Everything we've
	// sent up to this point has been received by the other
	// end.
	std::uint16_t m_acked_seq_nr = 0;

	// each packet gets one chance of "fast resend". i.e.
	// if we have multiple duplicate acks, we may send a
	// packet immediately, if m_fast_resend_seq_nr is set
	// to that packet's sequence number
	std::uint16_t m_fast_resend_seq_nr = 0;

	// this is the sequence number of the FIN packet
	// we've received. This sequence number is only
	// valid if m_eof is true. We should not accept
	// any packets beyond this sequence number from the
	// other end
	std::uint16_t m_eof_seq_nr = 0;

	// this is the lowest sequence number that, when lost,
	// will cause the window size to be cut in half
	std::uint16_t m_loss_seq_nr = 0;

	// the max number of bytes we can send in a packet
	// including the header
	std::uint16_t m_mtu = TORRENT_ETHERNET_MTU - TORRENT_IPV4_HEADER - TORRENT_UDP_HEADER - 8 - 24 - 36;

	// the floor is the largest packet that we have
	// been able to get through without fragmentation
	std::uint16_t m_mtu_floor = TORRENT_INET_MIN_MTU - TORRENT_IPV4_HEADER - TORRENT_UDP_HEADER;

	// the ceiling is the largest packet that we might
	// be able to get through without fragmentation.
	// i.e. ceiling +1 is very likely to not get through
	// or we have in fact experienced a drop or ICMP
	// message indicating that it is
	std::uint16_t m_mtu_ceiling = TORRENT_ETHERNET_MTU - TORRENT_IPV4_HEADER - TORRENT_UDP_HEADER;

	// the sequence number of the probe in-flight
	// this is 0 if there is no probe in flight
	std::uint16_t m_mtu_seq = 0;

	// this is a counter of how many times the current m_acked_seq_nr
	// has been ACKed. If it's ACKed more than 3 times, we assume the
	// packet with the next sequence number has been lost, and we trigger
	// a re-send. Obviously an ACK only counts as a duplicate as long as
	// we have outstanding packets following it.
	std::uint8_t m_duplicate_acks = 0;

	// the number of packet timeouts we've seen in a row
	// this affects the packet timeout time
	std::uint8_t m_num_timeouts = 0;

	// it's important that these match the enums in performance_counters for
	// num_utp_idle etc.
	enum state_t {
		// not yet connected
		UTP_STATE_NONE,
		// sent a syn packet, not received any acks
		UTP_STATE_SYN_SENT,
		// syn-ack received and in normal operation
		// of sending and receiving data
		UTP_STATE_CONNECTED,
		// fin sent, but all packets up to the fin packet
		// have not yet been acked. We might still be waiting
		// for a FIN from the other end
		UTP_STATE_FIN_SENT,

		// ====== states beyond this point =====
		// === are considered closing states ===
		// === and will cause the socket to ====
		// ============ be deleted =============

		// the socket has been gracefully disconnected
		// and is waiting for the client to make a
		// socket call so that we can communicate this
		// fact and actually delete all the state, or
		// there is an error on this socket and we're
		// waiting to communicate this to the client in
		// a callback. The error in either case is stored
		// in m_error. If the socket has gracefully shut
		// down, the error is error::eof.
		UTP_STATE_ERROR_WAIT,

		// there are no more references to this socket
		// and we can delete it
		UTP_STATE_DELETE
	};

	// this is the cursor into m_delay_sample_hist
	std::uint8_t m_delay_sample_idx:2;

	// the state the socket is in
	std::uint8_t m_state:3;

	// this is set to true when we receive a fin
	bool m_eof:1;

	// is this socket state attached to a user space socket?
	bool m_attached:1;

	// this is true if nagle is enabled (which it is by default)
	bool m_nagle:1;

	// this is true while the socket is in slow start mode. It's
	// only in slow-start during the start-up phase. Slow start
	// (contrary to what its name suggest) means that we're growing
	// the congestion window (cwnd) exponentially rather than linearly.
	// this is done at startup of a socket in order to find its
	// link capacity faster. This behaves similar to TCP slow start
	bool m_slow_start:1;

	// this is true as long as we have as many packets in
	// flight as allowed by the congestion window (cwnd)
	bool m_cwnd_full:1;

	// this is set to one if the current read operation
	// has a null_buffer. i.e. we're not reading into a user-provided
	// buffer, we're just signalling when there's something
	// to read from our internal receive buffer
	bool m_null_buffers:1;

	// this is set to true when this socket has added itself to
	// the utp socket manager's list of deferred acks. Once the
	// burst of incoming UDP packets is all drained, the utp socket
	// manager will send acks for all sockets on this list.
	bool m_deferred_ack:1;

	// this is true if this socket has subscribed to be notified
	// when this receive round is done
	bool m_subscribe_drained:1;

	// if this socket tries to send a packet via the utp socket
	// manager, and it fails with EWOULDBLOCK, the socket
	// is stalled and this is set. It's also added to a list
	// of sockets in the utp_socket_manager to be notified of
	// the socket being writable again
	bool m_stalled:1;

	// this is false by default and set to true once we've received a non-SYN
	// packet for this connection with a correct ack_nr, confirming that the
	// other end is not spoofing its source IP
	bool m_confirmed:1;
};

utp_socket_impl* construct_utp_impl(std::uint16_t recv_id
	, std::uint16_t send_id, void* userdata
	, utp_socket_manager& sm)
{
	return new utp_socket_impl(recv_id, send_id, userdata, sm);
}

void detach_utp_impl(utp_socket_impl* s)
{
	s->detach();
}

void delete_utp_impl(utp_socket_impl* s)
{
	delete s;
}

void utp_abort(utp_socket_impl* s)
{
	s->m_error = boost::asio::error::connection_aborted;
	s->set_state(utp_socket_impl::UTP_STATE_ERROR_WAIT);
	s->test_socket_state();
}

bool should_delete(utp_socket_impl* s)
{
	return s->should_delete();
}

bool bound_to_udp_socket(utp_socket_impl* s, std::weak_ptr<utp_socket_interface> sock)
{
	return s->m_sock.lock() == sock.lock();
}

void tick_utp_impl(utp_socket_impl* s, time_point now)
{
	s->tick(now);
}

void utp_init_mtu(utp_socket_impl* s, int link_mtu, int utp_mtu)
{
	s->init_mtu(link_mtu, utp_mtu);
}

void utp_init_socket(utp_socket_impl* s, std::weak_ptr<utp_socket_interface> sock)
{
	s->m_sock = std::move(sock);
}

bool utp_incoming_packet(utp_socket_impl* s
	, span<char const> p
	, udp::endpoint const& ep, time_point const receive_time)
{
	return s->incoming_packet({reinterpret_cast<std::uint8_t const*>(p.data()), p.size()}
		, ep, receive_time);
}

bool utp_match(utp_socket_impl* s, udp::endpoint const& ep, std::uint16_t const id)
{
	return s->m_recv_id == id
		&& s->m_port == ep.port()
		&& s->m_remote_address == ep.address();
}

udp::endpoint utp_remote_endpoint(utp_socket_impl* s)
{
	return udp::endpoint(s->m_remote_address, s->m_port);
}

std::uint16_t utp_receive_id(utp_socket_impl* s)
{
	return s->m_recv_id;
}

void utp_writable(utp_socket_impl* s)
{
	TORRENT_ASSERT(s->m_stalled);
	s->m_stalled = false;
	s->writable();
}

void utp_send_ack(utp_socket_impl* s)
{
	TORRENT_ASSERT(s->m_deferred_ack);
	s->m_deferred_ack = false;
	s->send_pkt(utp_socket_impl::pkt_ack);
}

void utp_socket_drained(utp_socket_impl* s)
{
	s->m_subscribe_drained = false;

	// at this point, we know we won't receive any
	// more packets this round. So, we may want to
	// call the receive callback function to
	// let the user consume it

	s->maybe_trigger_receive_callback();
	s->maybe_trigger_send_callback();
}

void utp_socket_impl::update_mtu_limits()
{
	INVARIANT_CHECK;

	if (m_mtu_floor > m_mtu_ceiling) m_mtu_floor = m_mtu_ceiling;

	m_mtu = (m_mtu_floor + m_mtu_ceiling) / 2;

	if ((m_cwnd >> 16) < m_mtu) m_cwnd = std::int64_t(m_mtu) * (1 << 16);

	UTP_LOGV("%8p: updating MTU to: %d [%d, %d]\n"
		, static_cast<void*>(this), m_mtu, m_mtu_floor, m_mtu_ceiling);

	// clear the mtu probe sequence number since
	// it was either dropped or acked
	m_mtu_seq = 0;
}

int utp_socket_state(utp_socket_impl const* s)
{
	return s->m_state;
}

int utp_stream::send_delay() const
{
	return m_impl ? m_impl->m_send_delay : 0;
}

int utp_stream::recv_delay() const
{
	return m_impl ? m_impl->m_recv_delay : 0;
}

utp_stream::utp_stream(io_service& io_service)
	: m_io_service(io_service)
	, m_impl(nullptr)
	, m_open(false)
{
}

utp_socket_impl* utp_stream::get_impl()
{
	return m_impl;
}

void utp_stream::set_close_reason(close_reason_t code)
{
	if (!m_impl) return;
	m_impl->set_close_reason(code);
}

close_reason_t utp_stream::get_close_reason()
{
	return m_incoming_close_reason;
}

void utp_stream::close()
{
	if (!m_impl) return;
	if (!m_impl->destroy())
	{
		if (!m_impl) return;
		detach_utp_impl(m_impl);
		m_impl = nullptr;
	}
}

std::size_t utp_stream::available() const
{
	return m_impl ? m_impl->available() : 0;
}

utp_stream::endpoint_type utp_stream::remote_endpoint(error_code& ec) const
{
	if (!m_impl)
	{
		ec = boost::asio::error::not_connected;
		return endpoint_type();
	}
	return m_impl->remote_endpoint(ec);
}

utp_stream::endpoint_type utp_stream::local_endpoint(error_code& ec) const
{
	if (m_impl == nullptr)
	{
		ec = boost::asio::error::not_connected;
		return endpoint_type();
	}

	auto s = m_impl->m_sock.lock();
	if (!s)
	{
		ec = boost::asio::error::not_connected;
		return endpoint_type();
	}

	udp::endpoint ep = s->get_local_endpoint();
	return endpoint_type(ep.address(), ep.port());
}

utp_stream::~utp_stream()
{
	if (m_impl)
	{
		UTP_LOGV("%8p: utp_stream destructed\n", static_cast<void*>(m_impl));
		m_impl->destroy();
		detach_utp_impl(m_impl);
	}

	m_impl = nullptr;
}

void utp_stream::set_impl(utp_socket_impl* impl)
{
	TORRENT_ASSERT(m_impl == nullptr);
	TORRENT_ASSERT(!m_open);
	m_impl = impl;
	m_open = true;
}

int utp_stream::read_buffer_size() const
{
	TORRENT_ASSERT(m_impl);
	return m_impl->m_receive_buffer_size;
}

void utp_stream::on_close_reason(void* self, close_reason_t reason)
{
	auto* s = static_cast<utp_stream*>(self);

	// it's possible the socket has been unlinked already, in which case m_impl
	// will be nullptr
	if (s->m_impl)
		s->m_incoming_close_reason = reason;
}

void utp_stream::on_read(void* self, std::size_t const bytes_transferred
	, error_code const& ec, bool const shutdown)
{
	auto* s = static_cast<utp_stream*>(self);

	UTP_LOGV("%8p: calling read handler read:%d ec:%s shutdown:%d\n", static_cast<void*>(s->m_impl)
		, int(bytes_transferred), ec.message().c_str(), shutdown);

	TORRENT_ASSERT(s->m_read_handler);
	TORRENT_ASSERT(bytes_transferred > 0 || ec || s->m_impl->m_null_buffers);
	s->m_io_service.post(std::bind<void>(std::move(s->m_read_handler), ec, bytes_transferred));
	s->m_read_handler = nullptr;
	if (shutdown && s->m_impl)
	{
		TORRENT_ASSERT(ec);
		detach_utp_impl(s->m_impl);
		s->m_impl = nullptr;
	}
}

void utp_stream::on_write(void* self, std::size_t const bytes_transferred
	, error_code const& ec, bool const shutdown)
{
	auto* s = static_cast<utp_stream*>(self);

	UTP_LOGV("%8p: calling write handler written:%d ec:%s shutdown:%d\n"
		, static_cast<void*>(s->m_impl)
		, int(bytes_transferred), ec.message().c_str(), shutdown);

	TORRENT_ASSERT(s->m_write_handler);
	TORRENT_ASSERT(bytes_transferred > 0 || ec);
	s->m_io_service.post(std::bind<void>(std::move(s->m_write_handler), ec, bytes_transferred));
	s->m_write_handler = nullptr;
	if (shutdown && s->m_impl)
	{
		TORRENT_ASSERT(ec);
		detach_utp_impl(s->m_impl);
		s->m_impl = nullptr;
	}
}

void utp_stream::on_connect(void* self, error_code const& ec, bool const shutdown)
{
	auto* s = static_cast<utp_stream*>(self);
	TORRENT_ASSERT(s);

	UTP_LOGV("%8p: calling connect handler ec:%s shutdown:%d\n"
		, static_cast<void*>(s->m_impl), ec.message().c_str(), shutdown);

	TORRENT_ASSERT(s->m_connect_handler);
	s->m_io_service.post(std::bind<void>(std::move(s->m_connect_handler), ec));
	s->m_connect_handler = nullptr;
	if (shutdown && s->m_impl)
	{
		TORRENT_ASSERT(ec);
		detach_utp_impl(s->m_impl);
		s->m_impl = nullptr;
	}
}

void utp_stream::add_read_buffer(void* buf, std::size_t const len)
{
	TORRENT_ASSERT(m_impl);
	TORRENT_ASSERT(len < INT_MAX);
	TORRENT_ASSERT(len > 0);
	TORRENT_ASSERT(buf);
	m_impl->m_read_buffer.emplace_back(buf, len);
	m_impl->m_read_buffer_size += int(len);

	UTP_LOGV("%8p: add_read_buffer %d bytes\n", static_cast<void*>(m_impl), int(len));
}

// this is the wrapper to add a user provided write buffer to the
// utp_socket_impl. It makes sure the m_write_buffer_size is kept
// up to date
void utp_stream::add_write_buffer(void const* buf, std::size_t const len)
{
	TORRENT_ASSERT(m_impl);
	TORRENT_ASSERT(len < INT_MAX);
	TORRENT_ASSERT(len > 0);
	TORRENT_ASSERT(buf);

#if TORRENT_USE_ASSERTS
	int write_buffer_size = 0;
	for (auto const& i : m_impl->m_write_buffer)
	{
		TORRENT_ASSERT(std::numeric_limits<int>::max() - int(i.len) > write_buffer_size);
		write_buffer_size += int(i.len);
	}
	TORRENT_ASSERT(m_impl->m_write_buffer_size == write_buffer_size);
#endif

	m_impl->m_write_buffer.emplace_back(const_cast<void*>(buf), len);
	m_impl->m_write_buffer_size += int(len);

#if TORRENT_USE_ASSERTS
	write_buffer_size = 0;
	for (auto const& i : m_impl->m_write_buffer)
	{
		TORRENT_ASSERT(std::numeric_limits<int>::max() - int(i.len) > write_buffer_size);
		write_buffer_size += int(i.len);
	}
	TORRENT_ASSERT(m_impl->m_write_buffer_size == write_buffer_size);
#endif

	UTP_LOGV("%8p: add_write_buffer %d bytes\n", static_cast<void*>(m_impl), int(len));
}

// this is called when all user provided read buffers have been added
// and it's time to execute the async operation. The first thing we
// do is to copy any data stored in m_receive_buffer into the user
// provided buffer. This might be enough to in turn trigger the read
// handler immediately.
void utp_stream::issue_read()
{
	TORRENT_ASSERT(m_impl->m_userdata);
	TORRENT_ASSERT(!m_impl->m_read_handler);

	m_impl->m_null_buffers = m_impl->m_read_buffer_size == 0;

	m_impl->m_read_handler = true;
	if (m_impl->test_socket_state()) return;

	UTP_LOGV("%8p: new read handler. %d bytes in buffer\n"
		, static_cast<void*>(m_impl), m_impl->m_receive_buffer_size);

	// so, the client wants to read. If we already
	// have some data in the read buffer, move it into the
	// client's buffer right away

	m_impl->m_read += int(read_some(false));
	m_impl->maybe_trigger_receive_callback();
}

std::size_t utp_stream::read_some(bool const clear_buffers)
{
	if (m_impl->m_receive_buffer_size == 0)
	{
		if (clear_buffers)
		{
			m_impl->m_read_buffer_size = 0;
			m_impl->m_read_buffer.clear();
		}
		return 0;
	}

	auto target = m_impl->m_read_buffer.begin();

	std::size_t ret = 0;

	int pop_packets = 0;
	for (auto i = m_impl->m_receive_buffer.begin()
		, end(m_impl->m_receive_buffer.end()); i != end;)
	{
		if (target == m_impl->m_read_buffer.end())
		{
			UTP_LOGV("  No more target buffers: %d bytes left in buffer\n"
				, m_impl->m_receive_buffer_size);
			TORRENT_ASSERT(m_impl->m_read_buffer.empty());
			break;
		}

#if TORRENT_USE_INVARIANT_CHECKS
		m_impl->check_receive_buffers();
#endif

		packet* p = i->get();
		int to_copy = std::min(p->size - p->header_size, aux::numeric_cast<int>(target->len));
		TORRENT_ASSERT(to_copy >= 0);
		std::memcpy(target->buf, p->buf + p->header_size, std::size_t(to_copy));
		ret += std::size_t(to_copy);
		target->buf = static_cast<char*>(target->buf) + to_copy;
		TORRENT_ASSERT(target->len >= std::size_t(to_copy));
		target->len -= std::size_t(to_copy);
		m_impl->m_receive_buffer_size -= to_copy;
		TORRENT_ASSERT(m_impl->m_read_buffer_size >= to_copy);
		m_impl->m_read_buffer_size -= to_copy;
		p->header_size += std::uint16_t(to_copy);
		if (target->len == 0) target = m_impl->m_read_buffer.erase(target);

#if TORRENT_USE_INVARIANT_CHECKS
		m_impl->check_receive_buffers();
#endif

		TORRENT_ASSERT(m_impl->m_receive_buffer_size >= 0);

		// Consumed entire packet
		if (p->header_size == p->size)
		{
			m_impl->release_packet(std::move(*i));
			i->reset();
			++pop_packets;
			++i;
		}

		if (m_impl->m_receive_buffer_size == 0)
		{
			UTP_LOGV("%8p: Didn't fill entire target: %d bytes left in buffer\n"
				, static_cast<void*>(m_impl), m_impl->m_receive_buffer_size);
			break;
		}
	}
	// remove the packets from the receive_buffer that we already copied over
	// and freed
	m_impl->m_receive_buffer.erase(m_impl->m_receive_buffer.begin()
		, m_impl->m_receive_buffer.begin() + pop_packets);
	// we exited either because we ran out of bytes to copy
	// or because we ran out of space to copy the bytes to
	TORRENT_ASSERT(m_impl->m_receive_buffer_size == 0
		|| m_impl->m_read_buffer.empty());

	UTP_LOGV("%8p: %d packets moved from buffer to user space (%d bytes)\n"
		, static_cast<void*>(m_impl), pop_packets, int(ret));

	if (clear_buffers)
	{
		m_impl->m_read_buffer_size = 0;
		m_impl->m_read_buffer.clear();
	}
	TORRENT_ASSERT(ret > 0 || m_impl->m_null_buffers);
	return ret;
}

// this is called when all user provided write buffers have been
// added. Start trying to send packets with the payload immediately.
void utp_stream::issue_write()
{
	UTP_LOGV("%8p: new write handler. %d bytes to write\n"
		, static_cast<void*>(m_impl), m_impl->m_write_buffer_size);

	TORRENT_ASSERT(m_impl->m_write_buffer_size > 0);
	TORRENT_ASSERT(m_impl->m_write_handler == false);
	TORRENT_ASSERT(m_impl->m_userdata);

	m_impl->m_write_handler = true;
	m_impl->m_written = 0;
	if (m_impl->test_socket_state()) return;

	// try to write. send_pkt returns false if there's
	// no more payload to send or if the congestion window
	// is full and we can't send more packets right now
	while (m_impl->send_pkt());

	// if there was an error in send_pkt(), m_impl may be
	// 0 at this point
	if (m_impl) m_impl->maybe_trigger_send_callback();
}

void utp_stream::do_connect(tcp::endpoint const& ep)
{
	int link_mtu, utp_mtu;
	std::tie(link_mtu, utp_mtu) = m_impl->m_sm.mtu_for_dest(ep.address());
	m_impl->init_mtu(link_mtu, utp_mtu);
	TORRENT_ASSERT(m_impl->m_connect_handler == false);
	m_impl->m_remote_address = ep.address();
	m_impl->m_port = ep.port();

	m_impl->m_connect_handler = true;

	if (m_impl->test_socket_state()) return;
	m_impl->send_syn();
}

// =========== utp_socket_impl ============

utp_socket_impl::~utp_socket_impl()
{
	INVARIANT_CHECK;

	TORRENT_ASSERT(!m_attached);
	TORRENT_ASSERT(!m_deferred_ack);

	m_sm.inc_stats_counter(counters::num_utp_idle + m_state, -1);

	UTP_LOGV("%8p: destroying utp socket state\n", static_cast<void*>(this));

	// free any buffers we're holding
	for (std::uint16_t i = std::uint16_t(m_inbuf.cursor()), end((m_inbuf.cursor()
		+ m_inbuf.capacity()) & ACK_MASK);
		i != end; i = (i + 1) & ACK_MASK)
	{
		packet_ptr p = m_inbuf.remove(i);
		release_packet(std::move(p));
	}
	for (std::uint16_t i = std::uint16_t(m_outbuf.cursor()), end((m_outbuf.cursor()
		+ m_outbuf.capacity()) & ACK_MASK);
		i != end; i = (i + 1) & ACK_MASK)
	{
		packet_ptr p = m_outbuf.remove(i);
		release_packet(std::move(p));
	}

	for (auto& p : m_receive_buffer)
		release_packet(std::move(p));

	release_packet(std::move(m_nagle_packet));
	m_nagle_packet.reset();
}

bool utp_socket_impl::should_delete() const
{
	INVARIANT_CHECK;

	// if the socket state is not attached anymore we're free
	// to delete it from the client's point of view. The other
	// endpoint however might still need to be told that we're
	// closing the socket. Only delete the state if we're not
	// attached and we're in a state where the other end doesn't
	// expect the socket to still be alive
	// when m_stalled is true, it means the socket manager has a
	// pointer to this socket, waiting for the UDP socket to
	// become writable again. We have to wait for that, so that
	// the pointer is removed from that queue. Otherwise we would
	// leave a dangling pointer in the socket manager
	bool ret = (m_state >= UTP_STATE_ERROR_WAIT || m_state == UTP_STATE_NONE)
		&& !m_attached && !m_stalled;

	if (ret)
	{
		UTP_LOGV("%8p: should_delete() = true\n", static_cast<void const*>(this));
	}

	return ret;
}

void utp_socket_impl::maybe_trigger_receive_callback()
{
	INVARIANT_CHECK;

	if (m_read_handler == false) return;

	// nothing has been read or there's no outstanding read operation
	if (m_null_buffers && m_receive_buffer_size == 0) return;
	else if (!m_null_buffers && m_read == 0) return;

	UTP_LOGV("%8p: calling read handler read:%d\n", static_cast<void*>(this), m_read);
	m_read_handler = false;
	utp_stream::on_read(m_userdata, aux::numeric_cast<std::size_t>(m_read), m_error, false);
	m_read = 0;
	m_read_buffer_size = 0;
	m_read_buffer.clear();
}

void utp_socket_impl::maybe_trigger_send_callback()
{
	INVARIANT_CHECK;

	// nothing has been written or there's no outstanding write operation
	if (m_written == 0 || m_write_handler == false) return;

	UTP_LOGV("%8p: calling write handler written:%d\n", static_cast<void*>(this), m_written);

	m_write_handler = false;
	utp_stream::on_write(m_userdata, aux::numeric_cast<std::size_t>(m_written), m_error, false);
	m_written = 0;
	m_write_buffer_size = 0;
	m_write_buffer.clear();
}

void utp_socket_impl::set_close_reason(close_reason_t code)
{
#if TORRENT_UTP_LOG
	UTP_LOGV("%8p: set_close_reason: %d\n"
		, static_cast<void*>(this), static_cast<int>(m_close_reason));
#endif
	m_close_reason = code;
}

bool utp_socket_impl::destroy()
{
	INVARIANT_CHECK;

#if TORRENT_UTP_LOG
	UTP_LOGV("%8p: destroy state:%s (close-reason: %d)\n"
		, static_cast<void*>(this), socket_state_names[m_state], int(m_close_reason));
#endif

	if (m_userdata == nullptr) return false;

	if (m_state == UTP_STATE_CONNECTED)
		send_fin();

	bool cancelled = cancel_handlers(boost::asio::error::operation_aborted, true);

	m_userdata = nullptr;

	m_read_buffer.clear();
	m_read_buffer_size = 0;

	m_write_buffer.clear();
	m_write_buffer_size = 0;

	if ((m_state == UTP_STATE_ERROR_WAIT
		|| m_state == UTP_STATE_NONE
		|| m_state == UTP_STATE_SYN_SENT) && cancelled)
	{
		set_state(UTP_STATE_DELETE);
#if TORRENT_UTP_LOG
		UTP_LOGV("%8p: state:%s\n", static_cast<void*>(this), socket_state_names[m_state]);
#endif
	}

	return cancelled;

	// #error our end is closing. Wait for everything to be acked
}

void utp_socket_impl::detach()
{
	INVARIANT_CHECK;

	UTP_LOGV("%8p: detach()\n", static_cast<void*>(this));
	m_attached = false;
}

void utp_socket_impl::send_syn()
{
	INVARIANT_CHECK;

	m_seq_nr = std::uint16_t(random(0xffff));
	m_acked_seq_nr = (m_seq_nr - 1) & ACK_MASK;
	m_loss_seq_nr = m_acked_seq_nr;
	m_ack_nr = 0;
	m_fast_resend_seq_nr = m_seq_nr;

	packet_ptr p = acquire_packet(sizeof(utp_header));
	p->size = sizeof(utp_header);
	p->header_size = sizeof(utp_header);
	p->num_transmissions = 0;
	p->mtu_probe = false;
#if TORRENT_USE_ASSERTS
	p->num_fast_resend = 0;
#endif
	p->need_resend = false;
	auto* h = reinterpret_cast<utp_header*>(p->buf);
	h->type_ver = (ST_SYN << 4) | 1;
	h->extension = utp_no_extension;
	// using recv_id here is intentional! This is an odd
	// thing in uTP. The syn packet is sent with the connection
	// ID that it expects to receive the syn ack on. All
	// subsequent connection IDs will be this plus one.
	h->connection_id = m_recv_id;
	h->timestamp_difference_microseconds = m_reply_micro;
	h->wnd_size = 0;
	h->seq_nr = m_seq_nr;
	h->ack_nr = 0;

	time_point const now = clock_type::now();
	p->send_time = now;
	h->timestamp_microseconds = std::uint32_t(
		total_microseconds(now.time_since_epoch()) & 0xffffffff);

#if TORRENT_UTP_LOG
	UTP_LOGV("%8p: send_syn seq_nr:%d id:%d target:%s\n"
		, static_cast<void*>(this), int(m_seq_nr), int(m_recv_id)
		, print_endpoint(udp::endpoint(m_remote_address, m_port)).c_str());
#endif

	error_code ec;
	m_sm.send_packet(m_sock, udp::endpoint(m_remote_address, m_port)
		, reinterpret_cast<char const*>(h) , sizeof(utp_header), ec);

	if (ec == error::would_block || ec == error::try_again)
	{
#if TORRENT_UTP_LOG
		UTP_LOGV("%8p: socket stalled\n", static_cast<void*>(this));
#endif
		if (!m_stalled)
		{
			m_stalled = true;
			m_sm.subscribe_writable(this);
		}
	}
	else if (ec)
	{
		release_packet(std::move(p));
		m_error = ec;
		set_state(UTP_STATE_ERROR_WAIT);
		test_socket_state();
		return;
	}

	if (!m_stalled)
		++p->num_transmissions;

	TORRENT_ASSERT(!m_outbuf.at(m_seq_nr));
	TORRENT_ASSERT(h->seq_nr == m_seq_nr);
	TORRENT_ASSERT(p->buf == reinterpret_cast<std::uint8_t*>(h));
	m_outbuf.insert(m_seq_nr, std::move(p));

	m_seq_nr = (m_seq_nr + 1) & ACK_MASK;

	TORRENT_ASSERT(!m_error);
	set_state(UTP_STATE_SYN_SENT);
#if TORRENT_UTP_LOG
	UTP_LOGV("%8p: state:%s\n", static_cast<void*>(this), socket_state_names[m_state]);
#endif
}

// if a send ever failed with EWOULDBLOCK, we
// subscribe to the udp socket and will be
// signalled with this function.
void utp_socket_impl::writable()
{
#if TORRENT_UTP_LOG
	UTP_LOGV("%8p: writable\n", static_cast<void*>(this));
#endif
	if (should_delete()) return;

	while(send_pkt());

	maybe_trigger_send_callback();
}

void utp_socket_impl::send_fin()
{
	INVARIANT_CHECK;

	send_pkt(pkt_fin);
	// unless there was an error, we're now
	// in FIN-SENT state
	if (!m_error)
		set_state(UTP_STATE_FIN_SENT);

#if TORRENT_UTP_LOG
	UTP_LOGV("%8p: state:%s\n", static_cast<void*>(this), socket_state_names[m_state]);
#endif
}

void utp_socket_impl::send_reset(utp_header const* ph)
{
	INVARIANT_CHECK;

	utp_header h;
	h.type_ver = (ST_RESET << 4) | 1;
	h.extension = utp_no_extension;
	h.connection_id = m_send_id;
	h.timestamp_difference_microseconds = m_reply_micro;
	h.wnd_size = 0;
	h.seq_nr = std::uint16_t(random(0xffff));
	h.ack_nr = ph->seq_nr;
	time_point const now = clock_type::now();
	h.timestamp_microseconds = std::uint32_t(
		total_microseconds(now.time_since_epoch()) & 0xffffffff);

	UTP_LOGV("%8p: send_reset seq_nr:%d id:%d ack_nr:%d\n"
		, static_cast<void*>(this), int(h.seq_nr), int(m_send_id), int(ph->seq_nr));

	// ignore errors here
	error_code ec;
	m_sm.send_packet(m_sock, udp::endpoint(m_remote_address, m_port)
		, reinterpret_cast<char const*>(&h), sizeof(h), ec);
	if (ec)
	{
		UTP_LOGV("%8p: socket error: %s\n"
			, static_cast<void*>(this)
			, ec.message().c_str());
	}
}

std::size_t utp_socket_impl::available() const
{
	return aux::numeric_cast<std::size_t>(m_receive_buffer_size);
}

void utp_socket_impl::parse_close_reason(std::uint8_t const* ptr, int const size)
{
	if (size != 4) return;
	// skip reserved bytes
	ptr += 2;
	close_reason_t incoming_close_reason = static_cast<close_reason_t>(detail::read_uint16(ptr));

	UTP_LOGV("%8p: incoming close_reason: %d\n"
		, static_cast<void*>(this), int(incoming_close_reason));

	if (m_userdata == nullptr || !m_attached) return;

	utp_stream::on_close_reason(m_userdata, incoming_close_reason);
}

// returns (rtt, acked_bytes)
std::pair<std::uint32_t, int> utp_socket_impl::parse_sack(std::uint16_t const packet_ack
	, std::uint8_t const* ptr, int const size, time_point const now)
{
	INVARIANT_CHECK;

	if (size == 0) return { 0u, 0 };

	// this is the sequence number the current bit represents
	std::uint16_t ack_nr = (packet_ack + 2) & ACK_MASK;

#if TORRENT_VERBOSE_UTP_LOG
	std::string bitmask;
	bitmask.reserve(size);
	for (std::uint8_t const* b = ptr, *end = ptr + size; b != end; ++b)
	{
		unsigned char bitfield = unsigned(*b);
		unsigned char mask = 1;
		// for each bit
		for (int i = 0; i < 8; ++i)
		{
			bitmask += (mask & bitfield) ? "1" : "0";
			mask <<= 1;
		}
	}
	UTP_LOGV("%8p: got SACK first:%d %s our_seq_nr:%u fast_resend_seq_nr:%d\n"
		, static_cast<void*>(this), ack_nr, bitmask.c_str(), m_seq_nr, m_fast_resend_seq_nr);
#endif

	aux::array<std::uint16_t, 5> resend;
	int num_to_resend = 0;

	int acked_bytes = 0;
	std::uint32_t min_rtt = std::numeric_limits<std::uint32_t>::max();

	// this was implicitly lost
	if (!compare_less_wrap((packet_ack + 1) & ACK_MASK, m_fast_resend_seq_nr, ACK_MASK))
	{
		resend[num_to_resend++] = (packet_ack + 1) & ACK_MASK;
	}

	// for each byte
	std::uint8_t const* const start = ptr;
	std::uint8_t const* const end = ptr + size;
	for (; ptr != end; ++ptr)
	{
		std::uint8_t bitfield = *ptr;
		std::uint8_t mask = 1;
		// for each bit
		for (int i = 0; i < 8; ++i)
		{
			if (mask & bitfield)
			{
				// this bit was set, ack_nr was received
				packet_ptr p = m_outbuf.remove(aux::numeric_cast<packet_buffer::index_type>(ack_nr));
				if (p)
				{
					acked_bytes += p->size - p->header_size;
					// each ACKed packet counts as a duplicate ack
					UTP_LOGV("%8p: duplicate_acks:%u fast_resend_seq_nr:%u\n"
						, static_cast<void*>(this), m_duplicate_acks, m_fast_resend_seq_nr);
					min_rtt = std::min(min_rtt, ack_packet(std::move(p), now, std::uint16_t(ack_nr)));
				}
				else
				{
					// this packet might have been acked by a previous
					// selective ack
					maybe_inc_acked_seq_nr();
				}
			}
			else if (!compare_less_wrap(ack_nr, m_fast_resend_seq_nr, ACK_MASK)
				&& num_to_resend < int(resend.size()))
			{
				resend[num_to_resend++] = ack_nr;
			}

			mask <<= 1;
			ack_nr = (ack_nr + 1) & ACK_MASK;

			// we haven't sent packets past this point.
			// if there are any more bits set, we have to
			// ignore them anyway
			if (ack_nr == m_seq_nr) break;
		}
		if (ack_nr == m_seq_nr) break;
	}

	if (m_outbuf.empty()) m_duplicate_acks = 0;

	// now, scan the bits in reverse, and count the number of ACKed packets. Only
	// lost packets followed by 'dup_ack_limit' packets may be resent
	// start with the sequence number represented by the last bit in the SACK
	// bitmask
	std::uint16_t last_resend = (packet_ack + 1 + size * 8) & ACK_MASK;

	// the number of acked packets past the fast re-send sequence number
	// this is used to determine if we should trigger more fast re-sends
	int dups = 0;

	for (std::uint8_t const* i = end; i != start; --i)
	{
		std::uint8_t const bitfield = i[-1];
		std::uint8_t mask = 0x80;
		// for each bit
		for (int k = 0; k < 8; ++k)
		{
			if (mask & bitfield) ++dups;
			if (dups > dup_ack_limit) break;
			last_resend = (last_resend - 1) & ACK_MASK;
			mask >>= 1;
		}
		if (dups > dup_ack_limit) break;
	}

	// we did not get enough packets acked in this message to warrant a resend
	if (dups <= dup_ack_limit)
	{
		UTP_LOGV("%8p: only %d ACKs in SACK, requires more than %d to trigger fast retransmit\n"
			, static_cast<void*>(this), dups, dup_ack_limit);
		num_to_resend = 0;
	}

	// now we need to (likely) prune the tail of the resend list, since all
	// "unacked" packets that weren't followed by an acked one, don't count
	while (num_to_resend > 0 && !compare_less_wrap(resend[num_to_resend - 1], last_resend, ACK_MASK))
	{
		--num_to_resend;
	}

	TORRENT_ASSERT(m_outbuf.at((m_acked_seq_nr + 1) & ACK_MASK) || ((m_seq_nr - m_acked_seq_nr) & ACK_MASK) <= 1);

	bool cut_cwnd = true;

	// we received more than dup_ack_limit ACKs in this SACK message.
	// trigger fast re-send. This is not an equal check because 3 identical ACKS
	// are only 2 duplicates
	for (int i = 0; i < num_to_resend; ++i)
	{
		std::uint16_t const pkt_seq = resend[i];

		packet* p = m_outbuf.at(pkt_seq);
		UTP_LOGV("%8p: Packet %d lost. (fast_resend_seq_nr:%d trigger fast-resend)\n"
			, static_cast<void*>(this), pkt_seq, m_fast_resend_seq_nr);
		if (!p) continue;

		// don't cut cwnd if the packet we lost was the MTU probe
		// the logic to handle a lost MTU probe is in resend_packet()
		if (cut_cwnd && (pkt_seq != m_mtu_seq || m_mtu_seq == 0))
		{
			experienced_loss(pkt_seq, now);
			cut_cwnd = false;
		}

		if (resend_packet(p, true))
		{
			m_duplicate_acks = 0;
			m_fast_resend_seq_nr = (pkt_seq + 1) & ACK_MASK;
		}
	}

	return { min_rtt, acked_bytes };
}

// copies data from the write buffer into the packet
// pointed to by ptr
void utp_socket_impl::write_payload(std::uint8_t* ptr, int size)
{
	INVARIANT_CHECK;

#if TORRENT_USE_ASSERTS
	int write_buffer_size = 0;
	for (auto const& i : m_write_buffer)
	{
		TORRENT_ASSERT(std::numeric_limits<int>::max() - int(i.len) > write_buffer_size);
		write_buffer_size += int(i.len);
	}
	TORRENT_ASSERT(m_write_buffer_size == write_buffer_size);
#endif
	TORRENT_ASSERT(!m_write_buffer.empty() || size == 0);
	TORRENT_ASSERT(m_write_buffer_size >= size);
	auto i = m_write_buffer.begin();

	if (size == 0) return;

	int buffers_to_clear = 0;
	while (size > 0)
	{
		// i points to the iovec we'll start copying from
		int to_copy = std::min(size, int(i->len));
		TORRENT_ASSERT(to_copy >= 0);
		TORRENT_ASSERT(to_copy < INT_MAX / 2 && m_written < INT_MAX / 2);
		std::memcpy(ptr, static_cast<char const*>(i->buf), std::size_t(to_copy));
		size -= to_copy;
		m_written += to_copy;
		ptr += to_copy;
		i->len -= std::size_t(to_copy);
		TORRENT_ASSERT(m_write_buffer_size >= to_copy);
		m_write_buffer_size -= to_copy;
		i->buf = static_cast<char*>(i->buf) + to_copy;
		if (i->len == 0) ++buffers_to_clear;
		++i;
	}

	if (buffers_to_clear)
		m_write_buffer.erase(m_write_buffer.begin()
			, m_write_buffer.begin() + buffers_to_clear);

#if TORRENT_USE_ASSERTS
	write_buffer_size = 0;
	for (auto const& j : m_write_buffer)
	{
		TORRENT_ASSERT(std::numeric_limits<int>::max() - int(j.len) > write_buffer_size);
		write_buffer_size += int(j.len);
	}
	TORRENT_ASSERT(m_write_buffer_size == write_buffer_size);
#endif
}

void utp_socket_impl::subscribe_drained()
{
	INVARIANT_CHECK;

	if (m_subscribe_drained) return;

	UTP_LOGV("%8p: subscribe drained\n", static_cast<void*>(this));
	m_subscribe_drained = true;
	m_sm.subscribe_drained(this);
}

void utp_socket_impl::defer_ack()
{
	INVARIANT_CHECK;

	if (m_deferred_ack) return;

	UTP_LOGV("%8p: defer ack\n", static_cast<void*>(this));
	m_deferred_ack = true;
	m_sm.defer_ack(this);
}

void utp_socket_impl::remove_sack_header(packet* p)
{
	INVARIANT_CHECK;

	// remove the sack header
	std::uint8_t* ptr = p->buf + sizeof(utp_header);
	auto* h = reinterpret_cast<utp_header*>(p->buf);

	TORRENT_ASSERT(h->extension == utp_sack);

	h->extension = ptr[0];
	int sack_size = ptr[1];
	TORRENT_ASSERT(h->extension == utp_no_extension
		|| h->extension == utp_close_reason);

	UTP_LOGV("%8p: removing SACK header, %d bytes\n"
		, static_cast<void*>(this), sack_size + 2);

	TORRENT_ASSERT(p->size >= p->header_size);
	TORRENT_ASSERT(p->header_size >= sizeof(utp_header) + aux::numeric_cast<std::size_t>(sack_size) + 2);
	std::memmove(ptr, ptr + sack_size + 2, p->size - p->header_size);
	p->header_size -= std::uint16_t(sack_size + 2);
	p->size -= std::uint16_t(sack_size + 2);
}

// sends a packet, pulls data from the write buffer (if there's any)
// if ack is true, we need to send a packet regardless of if there's
// any data. Returns true if we could send more data (i.e. call
// send_pkt() again)
// returns true if there is more space for payload in our
// congestion window, false if there is no more space.
bool utp_socket_impl::send_pkt(int const flags)
{
#ifdef TORRENT_EXPENSIVE_INVARIANT_CHECKS
	INVARIANT_CHECK;
#endif

	bool const force = (flags & pkt_ack) || (flags & pkt_fin);

//	TORRENT_ASSERT(m_state != UTP_STATE_FIN_SENT || (flags & pkt_ack));

	// first see if we need to resend any packets

	// TODO: this loop is not very efficient. It could be fixed by having
	// a separate list of sequence numbers that need resending
	for (int i = (m_acked_seq_nr + 1) & ACK_MASK; i != m_seq_nr; i = (i + 1) & ACK_MASK)
	{
		packet* p = m_outbuf.at(aux::numeric_cast<packet_buffer::index_type>(i));
		if (!p) continue;
		if (!p->need_resend) continue;
		if (!resend_packet(p))
		{
			// we couldn't resend the packet. It probably doesn't
			// fit in our cwnd. If force is set, we need to continue
			// to send our packet anyway, if we don't have force set,
			// we might as well return
			if (!force) return false;
			// resend_packet might have failed
			if (m_state == UTP_STATE_ERROR_WAIT || m_state == UTP_STATE_DELETE) return false;
			break;
		}

		// don't fast-resend this packet
		if (m_fast_resend_seq_nr == i)
			m_fast_resend_seq_nr = (m_fast_resend_seq_nr + 1) & ACK_MASK;
	}

	// MTU DISCOVERY

	// under these conditions, the next packet we send should be an MTU probe.
	// MTU probes get to use the mid-point packet size, whereas other packets
	// use a conservative packet size of the largest known to work. The reason
	// for the cwnd condition is to make sure the probe is surrounded by non-
	// probes, to be able to distinguish a loss of the probe vs. just loss in
	// general.
	bool const mtu_probe = (m_mtu_seq == 0
		&& m_write_buffer_size >= m_mtu_floor * 3
		&& m_seq_nr != 0
		&& (m_cwnd >> 16) > m_mtu_floor * 3);
	// for non MTU-probes, use the conservative packet size
	int const effective_mtu = mtu_probe ? m_mtu : m_mtu_floor;

	std::uint32_t const close_reason = static_cast<std::uint32_t>(m_close_reason);

	int sack = 0;
	if (m_inbuf.size())
	{
		const int max_sack_size = effective_mtu
			- int(sizeof(utp_header))
			- 2 // for sack padding/header
			- (close_reason ? 6 : 0);

		// the SACK bitfield should ideally fit all
		// the pieces we have successfully received
		sack = (m_inbuf.span() + 7) / 8;
		if (sack > max_sack_size) sack = max_sack_size;
	}

	int const header_size = int(sizeof(utp_header))
		+ (sack ? sack + 2 : 0)
		+ (close_reason ? 6 : 0);

	// for non MTU-probes, use the conservative packet size
	int payload_size = std::min(m_write_buffer_size
		, effective_mtu - header_size);
	TORRENT_ASSERT(payload_size >= 0);

	// if we have one MSS worth of data, make sure it fits in our
	// congestion window and the advertised receive window from
	// the other end.
	if (m_bytes_in_flight + payload_size > std::min(int(m_cwnd >> 16)
		, int(m_adv_wnd)))
	{
		// this means there's not enough room in the send window for
		// another packet. We have to hold off sending this data.
		// we still need to send an ACK though
		// if we're trying to send a FIN, make an exception
		if ((flags & pkt_fin) == 0) payload_size = 0;

		// we're constrained by the window size
		m_cwnd_full = true;

		UTP_LOGV("%8p: no space in window send_buffer_size:%d cwnd:%d "
			"adv_wnd:%u in-flight:%d mtu:%u\n"
			, static_cast<void*>(this), m_write_buffer_size, int(m_cwnd >> 16)
			, m_adv_wnd, m_bytes_in_flight, m_mtu);

		if (!force)
		{
#if TORRENT_UTP_LOG
			UTP_LOGV("%8p: skipping send seq_nr:%d ack_nr:%d "
				"id:%d target:%s header_size:%d error:%s send_buffer_size:%d cwnd:%d "
				"adv_wnd:%d in-flight:%d mtu:%u effective-mtu:%d\n"
				, static_cast<void*>(this), int(m_seq_nr), int(m_ack_nr)
				, m_send_id, print_endpoint(udp::endpoint(m_remote_address, m_port)).c_str()
				, header_size, m_error.message().c_str(), m_write_buffer_size, int(m_cwnd >> 16)
				, m_adv_wnd, m_bytes_in_flight, m_mtu, effective_mtu);
#endif
			return false;
		}
	}

	// if we don't have any data to send, or can't send any data
	// and we don't have any data to force, don't send a packet
	if (payload_size == 0 && !force && !m_nagle_packet)
	{
#if TORRENT_UTP_LOG
		UTP_LOGV("%8p: skipping send (no payload and no force) seq_nr:%d ack_nr:%d "
			"id:%d target:%s header_size:%d error:%s send_buffer_size:%d cwnd:%d "
			"adv_wnd:%u in-flight:%d mtu:%u\n"
			, static_cast<void*>(this), int(m_seq_nr), int(m_ack_nr)
			, m_send_id, print_endpoint(udp::endpoint(m_remote_address, m_port)).c_str()
			, header_size, m_error.message().c_str(), m_write_buffer_size, int(m_cwnd >> 16)
			, m_adv_wnd, m_bytes_in_flight, m_mtu);
#endif
		return false;
	}

	packet_ptr p;
	std::uint8_t* ptr = nullptr;
	utp_header* h = nullptr;

	// payload size being zero means we're just sending
	// an force. We should not pick up the nagle packet
	if (!m_nagle_packet || (payload_size == 0 && force))
	{
		p = acquire_packet(effective_mtu);

		if (payload_size)
		{
			m_sm.inc_stats_counter(counters::utp_payload_pkts_out);
		}

		int const packet_size = header_size + payload_size;
		p->size = std::uint16_t(packet_size);
		p->header_size = std::uint16_t(packet_size - payload_size);
		p->num_transmissions = 0;
#if TORRENT_USE_ASSERTS
		p->num_fast_resend = 0;
#endif
		p->mtu_probe = false;
		p->need_resend = false;
		ptr = p->buf;
		h = reinterpret_cast<utp_header*>(ptr);
		ptr += sizeof(utp_header);

		h->extension = std::uint8_t(sack ? utp_sack
			: close_reason ? utp_close_reason : utp_no_extension);
		h->connection_id = m_send_id;
		// seq_nr is ignored for ST_STATE packets, so it doesn't
		// matter that we say this is a sequence number we haven't
		// actually sent yet
		h->seq_nr = m_seq_nr;
		h->type_ver = std::uint8_t(((payload_size ? ST_DATA : ST_STATE) << 4) | 1);

		write_payload(p->buf + p->header_size, payload_size);
	}
	else
	{
		// pick up the nagle packet and keep adding bytes to it
		p = std::move(m_nagle_packet);
		m_nagle_packet.reset();

		ptr = p->buf + sizeof(utp_header);
		h = reinterpret_cast<utp_header*>(p->buf);
		TORRENT_ASSERT(h->seq_nr == m_seq_nr);

		// if the packet has a selective force header, we'll need
		// to update it
		if (h->extension == utp_sack)
		{
			sack = ptr[1];
			// if we no longer have any out-of-order packets waiting
			// to be delivered, there's no selective ack to be sent.
			if (m_inbuf.size() == 0)
			{
				// we need to remove the sack header
				remove_sack_header(p.get());
				sack = 0;
			}
		}
		else
			sack = 0;

		std::int32_t const size_left = std::min(p->allocated - p->size
			, m_write_buffer_size);

		write_payload(p->buf + p->size, size_left);
		p->size += std::uint16_t(size_left);

		if (size_left > 0)
		{
			UTP_LOGV("%8p: NAGLE appending %d bytes to nagle packet. new size: %d allocated: %d\n"
				, static_cast<void*>(this), size_left, p->size, p->allocated);
		}

		// did we fill up the whole mtu?
		// if we didn't, we may still send it if there's
		// no bytes in flight
		if (m_bytes_in_flight > 0
			&& p->size < std::min(p->allocated, m_mtu_floor)
			&& !force
			&& m_nagle)
		{
			// the packet is still not a full MSS, so put it back into the nagle
			// packet
			m_nagle_packet = std::move(p);
			return false;
		}

		payload_size = p->size - p->header_size;
	}

	if (sack)
	{
		*ptr++ = std::uint8_t(close_reason ? utp_close_reason : utp_no_extension);
		*ptr++ = std::uint8_t(sack); // bytes for SACK bitfield
		write_sack(ptr, sack);
		ptr += sack;
		TORRENT_ASSERT(ptr <= p->buf + p->header_size);
	}

	if (close_reason)
	{
		*ptr++ = utp_no_extension;
		*ptr++ = 4;
		detail::write_uint32(close_reason, ptr);
	}

	if (m_bytes_in_flight > 0
		&& p->size < p->allocated
		&& !force
		&& m_nagle)
	{
		// this is nagle. If we don't have a full packet
		// worth of payload to send AND we have at least
		// one outstanding packet, hold off. Once the
		// outstanding packet is acked, we'll send this
		// payload
		UTP_LOGV("%8p: NAGLE not enough payload send_buffer_size:%d cwnd:%d "
			"adv_wnd:%u in-flight:%d mtu:%d effective_mtu:%d\n"
			, static_cast<void*>(this), m_write_buffer_size, int(m_cwnd >> 16)
			, m_adv_wnd, m_bytes_in_flight, m_mtu, effective_mtu);
		TORRENT_ASSERT(!m_nagle_packet);
		TORRENT_ASSERT(h->seq_nr == m_seq_nr);
		m_nagle_packet = std::move(p);
		return false;
	}

	// for ST_DATA packets, payload size is 0. Such packets do not have unique
	// sequence numbers and should never be used as mtu probes
	if ((mtu_probe || p->mtu_probe) && payload_size > m_mtu_floor)
	{
		p->mtu_probe = true;
		m_mtu_seq = m_seq_nr;
	}
	else
	{
		p->mtu_probe = false;
	}

	h->timestamp_difference_microseconds = m_reply_micro;
	h->wnd_size = static_cast<std::uint32_t>(std::max(
		m_receive_buffer_capacity - m_buffered_incoming_bytes
		- m_receive_buffer_size, 0));
	h->ack_nr = m_ack_nr;

	// if this is a FIN packet, override the type
	if (flags & pkt_fin)
		h->type_ver = (ST_FIN << 4) | 1;

	// fill in the timestamp as late as possible
	time_point const now = clock_type::now();
	p->send_time = now;
	h->timestamp_microseconds = std::uint32_t(
		total_microseconds(now.time_since_epoch()) & 0xffffffff);

#if TORRENT_UTP_LOG
	UTP_LOG("%8p: sending packet seq_nr:%d ack_nr:%d type:%s "
		"id:%d target:%s size:%d error:%s send_buffer_size:%d cwnd:%d "
		"adv_wnd:%d in-flight:%d mtu:%d timestamp:%u time_diff:%d "
		"mtu_probe:%d extension:%d\n"
		, static_cast<void*>(this), int(h->seq_nr), int(h->ack_nr), packet_type_names[h->get_type()]
		, m_send_id, print_endpoint(udp::endpoint(m_remote_address, m_port)).c_str()
		, p->size, m_error.message().c_str(), m_write_buffer_size, int(m_cwnd >> 16)
		, m_adv_wnd, m_bytes_in_flight, m_mtu, std::uint32_t(h->timestamp_microseconds)
		, std::uint32_t(h->timestamp_difference_microseconds), int(p->mtu_probe)
		, h->extension);
#endif

	error_code ec;
	m_sm.send_packet(m_sock, udp::endpoint(m_remote_address, m_port)
		, reinterpret_cast<char const*>(h), p->size, ec
		, p->mtu_probe ? udp_socket::dont_fragment : udp_send_flags_t{});

	++m_out_packets;
	m_sm.inc_stats_counter(counters::utp_packets_out);

	if (ec == error::message_size)
	{
#if TORRENT_UTP_LOG
		UTP_LOGV("%8p: error sending packet: %s\n"
			, static_cast<void*>(this)
			, ec.message().c_str());
#endif
		// if we fail even though this is not a probe, we're screwed
		// since we'd have to repacketize
		TORRENT_ASSERT(p->mtu_probe);
		m_mtu_ceiling = p->size - 1;
		if (m_mtu_floor > m_mtu_ceiling) m_mtu_floor = m_mtu_ceiling;
		update_mtu_limits();
		// resend the packet immediately without
		// it being an MTU probe
		p->mtu_probe = false;
		m_mtu_seq = 0;
		ec.clear();

#if TORRENT_UTP_LOG
		UTP_LOGV("%8p: re-sending\n", static_cast<void*>(this));
#endif
		m_sm.send_packet(m_sock, udp::endpoint(m_remote_address, m_port)
			, reinterpret_cast<char const*>(h), p->size, ec, {});
	}

	if (ec == error::would_block || ec == error::try_again)
	{
#if TORRENT_UTP_LOG
		UTP_LOGV("%8p: socket stalled\n", static_cast<void*>(this));
#endif
		if (!m_stalled)
		{
			m_stalled = true;
			m_sm.subscribe_writable(this);
		}
	}
	else if (ec)
	{
		m_error = ec;
		set_state(UTP_STATE_ERROR_WAIT);
		test_socket_state();
		release_packet(std::move(p));
		return false;
	}

	if (!m_stalled)
		++p->num_transmissions;

	// if we have payload, we need to save the packet until it's acked
	// and progress m_seq_nr
	if (p->size > p->header_size)
	{
		// if we're sending a payload packet, there should not
		// be a nagle packet waiting for more data
		TORRENT_ASSERT(!m_nagle_packet);

		TORRENT_ASSERT(h->seq_nr == m_seq_nr);

		// 0 is a special sequence number, since it's also used as "uninitialized".
		// we never send an mtu probe for sequence number 0
		TORRENT_ASSERT(p->mtu_probe == (m_seq_nr == m_mtu_seq)
			|| m_seq_nr == 0);

		// release the buffer, we're saving it in the circular
		// buffer of outgoing packets
		int const new_in_flight = p->size - p->header_size;
		packet_ptr old = m_outbuf.insert(m_seq_nr, std::move(p));
		if (old)
		{
//			TORRENT_ASSERT(reinterpret_cast<utp_header*>(old->buf)->seq_nr == m_seq_nr);
			if (!old->need_resend) m_bytes_in_flight -= old->size - old->header_size;
			release_packet(std::move(old));
		}
		TORRENT_ASSERT(h->seq_nr == m_seq_nr);
		m_seq_nr = (m_seq_nr + 1) & ACK_MASK;
		TORRENT_ASSERT(payload_size >= 0);
		m_bytes_in_flight += new_in_flight;
	}
	else
	{
		TORRENT_ASSERT(h->seq_nr == m_seq_nr);
	}

	// if the socket is stalled, always return false, don't
	// try to write more packets. We'll keep writing once
	// the underlying UDP socket becomes writable
	return m_write_buffer_size > 0 && !m_cwnd_full && !m_stalled;
}

// size is in bytes
void utp_socket_impl::write_sack(std::uint8_t* buf, int const size) const
{
	INVARIANT_CHECK;

	TORRENT_ASSERT(m_inbuf.size());
	int ack_nr = (m_ack_nr + 2) & ACK_MASK;
	std::uint8_t* end = buf + size;

	for (; buf != end; ++buf)
	{
		*buf = 0;
		int mask = 1;
		for (int i = 0; i < 8; ++i)
		{
			if (m_inbuf.at(aux::numeric_cast<packet_buffer::index_type>(ack_nr))) *buf |= mask;
			mask <<= 1;
			ack_nr = (ack_nr + 1) & ACK_MASK;
		}
	}
}

bool utp_socket_impl::resend_packet(packet* p, bool fast_resend)
{
	INVARIANT_CHECK;

	// for fast re-sends the packet hasn't been marked as needing resending
	TORRENT_ASSERT(p->need_resend || fast_resend);

	if (m_error) return false;

	if (((m_acked_seq_nr + 1) & ACK_MASK) == m_mtu_seq
		&& m_mtu_seq != 0)
	{
		m_mtu_seq = 0;
		p->mtu_probe = false;
		// we got multiple acks for the packet before our probe, assume
		// it was dropped because it was too big
		m_mtu_ceiling = p->size - 1;
		update_mtu_limits();
	}

	// we can only resend the packet if there's
	// enough space in our congestion window
	// since we can't re-packetize, some packets that are
	// larger than the congestion window must be allowed through
	// but only if we don't have any outstanding bytes
	int const window_size_left = std::min(int(m_cwnd >> 16), int(m_adv_wnd)) - m_bytes_in_flight;
	if (!fast_resend
		&& p->size - p->header_size > window_size_left
		&& m_bytes_in_flight > 0)
	{
		m_cwnd_full = true;
		return false;
	}

	// plus one since we have fast-resend as well, which doesn't
	// necessarily trigger by a timeout
	TORRENT_ASSERT(p->num_transmissions < m_sm.num_resends() + 1);

	TORRENT_ASSERT(p->size - p->header_size >= 0);
	if (p->need_resend) m_bytes_in_flight += p->size - p->header_size;

	m_sm.inc_stats_counter(counters::utp_packet_resend);
	if (fast_resend) m_sm.inc_stats_counter(counters::utp_fast_retransmit);

#if TORRENT_USE_ASSERTS
	if (fast_resend) ++p->num_fast_resend;
#endif
	p->need_resend = false;
	auto* h = reinterpret_cast<utp_header*>(p->buf);
	// update packet header
	h->timestamp_difference_microseconds = m_reply_micro;
	p->send_time = clock_type::now();
	h->timestamp_microseconds = std::uint32_t(
		total_microseconds(p->send_time.time_since_epoch()) & 0xffffffff);

	// if the packet has a selective ack header, we'll need
	// to update it
	if (h->extension == utp_sack && h->ack_nr != m_ack_nr)
	{
		std::uint8_t* ptr = p->buf + sizeof(utp_header);
		int sack_size = ptr[1];
		if (m_inbuf.size())
		{
			// update the sack header
			write_sack(ptr + 2, sack_size);
			TORRENT_ASSERT(ptr + sack_size + 2 <= p->buf + p->header_size);
		}
		else
		{
			remove_sack_header(p);
		}
	}

	h->ack_nr = m_ack_nr;

	error_code ec;
	m_sm.send_packet(m_sock, udp::endpoint(m_remote_address, m_port)
		, reinterpret_cast<char const*>(p->buf), p->size, ec);
	++m_out_packets;
	m_sm.inc_stats_counter(counters::utp_packets_out);


#if TORRENT_UTP_LOG
	UTP_LOGV("%8p: re-sending packet seq_nr:%d ack_nr:%d type:%s "
		"id:%d target:%s size:%d error:%s send_buffer_size:%d cwnd:%d "
		"adv_wnd:%d in-flight:%d mtu:%d timestamp:%u time_diff:%d\n"
		, static_cast<void*>(this), int(h->seq_nr), int(h->ack_nr), packet_type_names[h->get_type()]
		, m_send_id, print_endpoint(udp::endpoint(m_remote_address, m_port)).c_str()
		, p->size, ec.message().c_str(), m_write_buffer_size, int(m_cwnd >> 16)
		, m_adv_wnd, m_bytes_in_flight, m_mtu, std::uint32_t(h->timestamp_microseconds)
		, std::uint32_t(h->timestamp_difference_microseconds));
#endif

	if (ec == error::would_block || ec == error::try_again)
	{
#if TORRENT_UTP_LOG
		UTP_LOGV("%8p: socket stalled\n", static_cast<void*>(this));
#endif
		if (!m_stalled)
		{
			m_stalled = true;
			m_sm.subscribe_writable(this);
		}
	}
	else if (ec)
	{
		m_error = ec;
		set_state(UTP_STATE_ERROR_WAIT);
		test_socket_state();
		return false;
	}

	if (!m_stalled)
		++p->num_transmissions;

	return !m_stalled;
}

void utp_socket_impl::experienced_loss(std::uint32_t const seq_nr, time_point const now)
{
	INVARIANT_CHECK;

	// the window size could go below one MMS here, if it does,
	// we'll get a timeout in about one second

	m_sm.inc_stats_counter(counters::utp_packet_loss);

	// since loss often comes in bursts, we only cut the
	// window in half once per RTT. This is implemented
	// by limiting which packets can cause us to cut the
	// window size. The first packet that's lost will
	// update the limit to the last sequence number we sent.
	// i.e. only packet sent after this loss can cause another
	// window size cut. The +1 is to turn the comparison into
	// less than or equal to. If we experience loss of the
	// same packet again, ignore it.
	if (compare_less_wrap(seq_nr, m_loss_seq_nr + 1, ACK_MASK)) return;

	// don't reduce cwnd more than once every 100ms
	if (m_next_loss >= now) return;

	m_next_loss = now + milliseconds(m_sm.cwnd_reduce_timer());

	// cut window size in 2
	m_cwnd = std::max(m_cwnd * m_sm.loss_multiplier() / 100
		, std::int64_t(m_mtu) * (1 << 16));
	m_loss_seq_nr = m_seq_nr;
	UTP_LOGV("%8p: Lost packet %d caused cwnd cut. m_loss_seq_nr:%d\n"
		, static_cast<void*>(this), seq_nr, m_seq_nr);

	// if we happen to be in slow-start mode, we need to leave it
	// note that we set ssthres to the window size _after_ reducing it. Next slow
	// start should end before we over shoot.
	if (m_slow_start)
	{
		m_ssthres = std::int32_t(m_cwnd >> 16);
		m_slow_start = false;
		UTP_LOGV("%8p: experienced loss, slow_start -> 0 ssthres:%d\n"
			, static_cast<void*>(this), m_ssthres);
	}
}

void utp_socket_impl::set_state(int s)
{
	if (s == m_state) return;

	m_sm.inc_stats_counter(counters::num_utp_idle + m_state, -1);
	m_state = std::uint8_t(s);
	m_sm.inc_stats_counter(counters::num_utp_idle + m_state, 1);
}

void utp_socket_impl::maybe_inc_acked_seq_nr()
{
#ifdef TORRENT_EXPENSIVE_INVARIANT_CHECKS
	INVARIANT_CHECK;
#endif

	bool incremented = false;
	// don't pass m_seq_nr, since we move into sequence
	// numbers that haven't been sent yet, and aren't
	// supposed to be in m_outbuf
	// if the slot in m_outbuf is 0, it means the
	// packet has been ACKed and removed from the send buffer
	while (((m_acked_seq_nr + 1) & ACK_MASK) != m_seq_nr
		&& m_outbuf.at((m_acked_seq_nr + 1) & ACK_MASK) == nullptr)
	{
		// increment the fast resend sequence number
		if (m_fast_resend_seq_nr == m_acked_seq_nr)
			m_fast_resend_seq_nr = (m_fast_resend_seq_nr + 1) & ACK_MASK;

		m_acked_seq_nr = (m_acked_seq_nr + 1) & ACK_MASK;
		incremented = true;
	}

	if (!incremented) return;

	// update loss seq number if it's less than the packet
	// that was just acked. If loss seq nr is greater, it suggests
	// that we're still in a window that has experienced loss
	if (compare_less_wrap(m_loss_seq_nr, m_acked_seq_nr, ACK_MASK))
		m_loss_seq_nr = m_acked_seq_nr;
	m_duplicate_acks = 0;
}

// returns RTT
std::uint32_t utp_socket_impl::ack_packet(packet_ptr p, time_point const receive_time
	, std::uint16_t seq_nr)
{
#ifdef TORRENT_EXPENSIVE_INVARIANT_CHECKS
	INVARIANT_CHECK;
#endif

	TORRENT_ASSERT(p);

	// verify that the packet we're removing was in fact sent
	// with the sequence number we expect
//	TORRENT_ASSERT(reinterpret_cast<utp_header*>(p->buf)->seq_nr == seq_nr);

	if (!p->need_resend)
	{
		TORRENT_ASSERT(m_bytes_in_flight >= p->size - p->header_size);
		m_bytes_in_flight -= p->size - p->header_size;
	}

	if (seq_nr == m_mtu_seq && m_mtu_seq != 0)
	{
		TORRENT_ASSERT(p->mtu_probe);
		// our mtu probe was acked!
		m_mtu_floor = std::max(m_mtu_floor, p->size);
		update_mtu_limits();
	}

	// increment the acked sequence number counter
	maybe_inc_acked_seq_nr();

	std::uint32_t rtt = std::uint32_t(total_microseconds(receive_time - p->send_time));
	if (receive_time < p->send_time)
	{
		// this means our clock is not monotonic. Just assume the RTT was 100 ms
		rtt = 100000;

		// the clock for this platform is not monotonic!
		TORRENT_ASSERT_FAIL();
	}

	UTP_LOGV("%8p: acked packet %d (%d bytes) (rtt:%u)\n"
		, static_cast<void*>(this), seq_nr, p->size - p->header_size, rtt / 1000);

	m_rtt.add_sample(rtt / 1000);
	release_packet(std::move(p));
	return rtt;
}

void utp_socket_impl::incoming(std::uint8_t const* buf, int size, packet_ptr p
	, time_point /* now */)
{
#ifdef TORRENT_EXPENSIVE_INVARIANT_CHECKS
	INVARIANT_CHECK;
#endif

	while (!m_read_buffer.empty())
	{
		UTP_LOGV("%8p: incoming: have user buffer (%d)\n", static_cast<void*>(this), m_read_buffer_size);
		if (p)
		{
			buf = p->buf + p->header_size;
			TORRENT_ASSERT(p->size - p->header_size >= size);
		}
		iovec_t* target = &m_read_buffer.front();

		int const to_copy = std::min(size, aux::numeric_cast<int>(target->len));
		TORRENT_ASSERT(to_copy >= 0);
		std::memcpy(target->buf, buf, std::size_t(to_copy));
		m_read += to_copy;
		target->buf = reinterpret_cast<std::uint8_t*>(target->buf) + to_copy;
		target->len -= std::size_t(to_copy);
		buf += to_copy;
		UTP_LOGV("%8p: copied %d bytes into user receive buffer\n", static_cast<void*>(this), to_copy);
		TORRENT_ASSERT(m_read_buffer_size >= to_copy);
		m_read_buffer_size -= to_copy;
		size -= to_copy;
		if (target->len == 0) m_read_buffer.erase(m_read_buffer.begin());
		if (p)
		{
			p->header_size += std::uint16_t(to_copy);
			TORRENT_ASSERT(p->header_size <= p->size);
		}

		if (size == 0)
		{
			TORRENT_ASSERT(!p || p->header_size == p->size);
			release_packet(std::move(p));
			return;
		}
	}

	TORRENT_ASSERT(m_read_buffer_size == 0);

	if (!p)
	{
		TORRENT_ASSERT(buf);
		p = acquire_packet(size);
		p->size = std::uint16_t(size);
		p->header_size = 0;
		std::memcpy(p->buf, buf, aux::numeric_cast<std::size_t>(size));
	}
	// save this packet until the client issues another read
	m_receive_buffer_size += p->size - p->header_size;
	m_receive_buffer.emplace_back(std::move(p));

	UTP_LOGV("%8p: incoming: saving packet in receive buffer (%d)\n", static_cast<void*>(this), m_receive_buffer_size);

#if TORRENT_USE_INVARIANT_CHECKS
	check_receive_buffers();
#endif
}

bool utp_socket_impl::cancel_handlers(error_code const& ec, bool shutdown)
{
	INVARIANT_CHECK;

	TORRENT_ASSERT(ec);
	bool ret = m_read_handler || m_write_handler || m_connect_handler;

	// calling the callbacks with m_userdata being 0 will just crash
	TORRENT_ASSERT((ret && m_userdata != nullptr) || !ret);

	bool read = m_read_handler;
	bool write = m_write_handler;
	bool connect = m_connect_handler;
	m_read_handler = false;
	m_write_handler = false;
	m_connect_handler = false;

	if (read) utp_stream::on_read(m_userdata, 0, ec, shutdown);
	if (write) utp_stream::on_write(m_userdata, 0, ec, shutdown);
	if (connect) utp_stream::on_connect(m_userdata, ec, shutdown);
	return ret;
}

bool utp_socket_impl::consume_incoming_data(
	utp_header const* ph, std::uint8_t const* ptr, int const payload_size
	, time_point const now)
{
	INVARIANT_CHECK;

	if (ph->get_type() != ST_DATA) return false;

	if (m_eof && m_ack_nr == m_eof_seq_nr)
	{
		// What?! We've already received a FIN and everything up
		// to it has been acked. Ignore this packet
		UTP_LOG("%8p: ERROR: ignoring packet on shut down socket\n"
			, static_cast<void*>(this));
		return true;
	}

	if (m_read_buffer_size == 0
		&& m_receive_buffer_size >= m_receive_buffer_capacity - m_buffered_incoming_bytes)
	{
		// if we don't have a buffer from the upper layer, and the
		// number of queued up bytes, waiting for the upper layer,
		// exceeds the advertised receive window, start ignoring
		// more data packets
		UTP_LOG("%8p: ERROR: our advertized window is not honored. "
			"recv_buf: %d buffered_in: %d max_size: %d\n"
			, static_cast<void*>(this), m_receive_buffer_size, m_buffered_incoming_bytes, m_receive_buffer_capacity);
		return false;
	}

	if (ph->seq_nr == ((m_ack_nr + 1) & ACK_MASK))
	{
		TORRENT_ASSERT(m_inbuf.at(m_ack_nr) == nullptr);

		if (m_buffered_incoming_bytes + m_receive_buffer_size + payload_size > m_receive_buffer_capacity)
		{
			UTP_LOGV("%8p: other end is not honoring our advertised window, dropping packet\n"
				, static_cast<void*>(this));
			return true;
		}

		// we received a packet in order
		incoming(ptr, payload_size, packet_ptr(), now);
		m_ack_nr = (m_ack_nr + 1) & ACK_MASK;

		// If this packet was previously in the reorder buffer
		// it would have been acked when m_ack_nr-1 was acked.
		TORRENT_ASSERT(!m_inbuf.at(m_ack_nr));

		UTP_LOGV("%8p: remove inbuf: %d (%d)\n"
			, static_cast<void*>(this), m_ack_nr, int(m_inbuf.size()));

		for (;;)
		{
			int const next_ack_nr = (m_ack_nr + 1) & ACK_MASK;

			packet_ptr p = m_inbuf.remove(aux::numeric_cast<packet_buffer::index_type>(next_ack_nr));

			if (!p) break;

			TORRENT_ASSERT(p->size >= p->header_size);
			int const size = p->size - p->header_size;
			m_buffered_incoming_bytes -= size;
			incoming(nullptr, size, std::move(p), now);

			m_ack_nr = std::uint16_t(next_ack_nr);

			UTP_LOGV("%8p: reordered remove inbuf: %d (%d)\n"
				, static_cast<void*>(this), m_ack_nr, int(m_inbuf.size()));
		}
	}
	else
	{
		// this packet was received out of order. Stick it in the
		// reorder buffer until it can be delivered in order

		// have we already received this packet and passed it on
		// to the client?
		if (!compare_less_wrap(m_ack_nr, ph->seq_nr, ACK_MASK))
		{
			UTP_LOGV("%8p: already received seq_nr: %d\n"
				, static_cast<void*>(this), int(ph->seq_nr));
			return true;
		}

		// do we already have this packet? If so, just ignore it
		if (m_inbuf.at(ph->seq_nr))
		{
			UTP_LOGV("%8p: already received seq_nr: %d\n"
				, static_cast<void*>(this), int(ph->seq_nr));
			return true;
		}

		if (m_buffered_incoming_bytes + m_receive_buffer_size + payload_size > m_receive_buffer_capacity)
		{
			UTP_LOGV("%8p: other end is not honoring our advertised window, dropping packet %d\n"
				, static_cast<void*>(this), int(ph->seq_nr));
			return true;
		}

		// we don't need to save the packet header, just the payload
		packet_ptr p = acquire_packet(payload_size);
		p->size = std::uint16_t(payload_size);
		p->header_size = 0;
		p->num_transmissions = 0;
#if TORRENT_USE_ASSERTS
		p->num_fast_resend = 0;
#endif
		p->need_resend = false;
		std::memcpy(p->buf, ptr, aux::numeric_cast<std::size_t>(payload_size));
		m_buffered_incoming_bytes += p->size;
		m_inbuf.insert(ph->seq_nr, std::move(p));

		UTP_LOGV("%8p: out of order. insert inbuf: %d (%d) m_ack_nr: %d\n"
			, static_cast<void*>(this), int(ph->seq_nr), int(m_inbuf.size()), m_ack_nr);
	}

	return false;
}

// returns true of the socket was closed
bool utp_socket_impl::test_socket_state()
{
	INVARIANT_CHECK;

	// if the socket is in a state where it's dead, just waiting to
	// tell the client that it's closed. Do that and transition into
	// the deleted state, where it will be deleted
	// it might be possible to get here twice, in which we need to
	// cancel any new handlers as well, even though we're already
	// in the delete state
	if (!m_error) return false;
	TORRENT_ASSERT(m_state == UTP_STATE_ERROR_WAIT || m_state == UTP_STATE_DELETE);

#if TORRENT_UTP_LOG
	UTP_LOGV("%8p: state:%s error:%s\n"
		, static_cast<void*>(this), socket_state_names[m_state], m_error.message().c_str());
#endif

	if (cancel_handlers(m_error, true))
	{
		set_state(UTP_STATE_DELETE);
#if TORRENT_UTP_LOG
		UTP_LOGV("%8p: state:%s\n", static_cast<void*>(this), socket_state_names[m_state]);
#endif
		return true;
	}
	return false;
}

void utp_socket_impl::init_mtu(int const link_mtu, int utp_mtu)
{
	INVARIANT_CHECK;

	if (link_mtu > TORRENT_ETHERNET_MTU)
	{
		// we can't use larger packets than this since we're
		// not allocating any more memory for socket buffers
		int const decrease = link_mtu - TORRENT_ETHERNET_MTU;
		utp_mtu -= decrease;
	}

	// set the ceiling to what we found out from the interface
	m_mtu_ceiling = std::uint16_t(utp_mtu);

	// start in the middle of the PMTU search space
	m_mtu = (m_mtu_ceiling + m_mtu_floor) / 2;
	if (m_mtu > m_mtu_ceiling) m_mtu = m_mtu_ceiling;
	if (m_mtu_floor > utp_mtu) m_mtu_floor = std::uint16_t(utp_mtu);

	// if the window size is smaller than one packet size
	// set it to one
	if ((m_cwnd >> 16) < m_mtu) m_cwnd = std::int64_t(m_mtu) * (1 << 16);

	UTP_LOGV("%8p: initializing MTU to: %d [%d, %d]\n"
		, static_cast<void*>(this), m_mtu, m_mtu_floor, m_mtu_ceiling);
}

// return false if this is an invalid packet
bool utp_socket_impl::incoming_packet(span<std::uint8_t const> buf
	, udp::endpoint const& ep, time_point receive_time)
{
	INVARIANT_CHECK;

	auto const* ph = reinterpret_cast<utp_header const*>(buf.data());
	m_sm.inc_stats_counter(counters::utp_packets_in);

	if (buf.size() < int(sizeof(utp_header)))
	{
		UTP_LOG("%8p: ERROR: incoming packet size too small:%d (ignored)\n"
			, static_cast<void*>(this), int(buf.size()));
		m_sm.inc_stats_counter(counters::utp_invalid_pkts_in);
		return false;
	}

	if (ph->get_version() != 1)
	{
		UTP_LOG("%8p: ERROR: incoming packet version:%d (ignored)\n"
			, static_cast<void*>(this), int(ph->get_version()));
		m_sm.inc_stats_counter(counters::utp_invalid_pkts_in);
		return false;
	}

	// SYN packets have special (reverse) connection ids
	if (ph->get_type() != ST_SYN && ph->connection_id != m_recv_id)
	{
		UTP_LOG("%8p: ERROR: incoming packet id:%d expected:%d (ignored)\n"
			, static_cast<void*>(this), int(ph->connection_id), int(m_recv_id));
		m_sm.inc_stats_counter(counters::utp_invalid_pkts_in);
		return false;
	}

	if (ph->get_type() >= NUM_TYPES)
	{
		UTP_LOG("%8p: ERROR: incoming packet type:%d (ignored)\n"
			, static_cast<void*>(this), int(ph->get_type()));
		m_sm.inc_stats_counter(counters::utp_invalid_pkts_in);
		return false;
	}

	if (m_state == UTP_STATE_NONE && ph->get_type() == ST_SYN)
	{
		m_remote_address = ep.address();
		m_port = ep.port();
	}

	if (m_state != UTP_STATE_NONE && ph->get_type() == ST_SYN)
	{
		UTP_LOG("%8p: ERROR: incoming packet type:ST_SYN (ignored)\n"
			, static_cast<void*>(this));
		m_sm.inc_stats_counter(counters::utp_invalid_pkts_in);
		return true;
	}

	bool step = false;
	if (receive_time - m_last_history_step > minutes(1))
	{
		step = true;
		m_last_history_step = receive_time;
	}

	// this is the difference between their send time and our receive time
	// 0 means no sample yet
	std::uint32_t their_delay = 0;
	if (ph->timestamp_microseconds != 0)
	{
		std::uint32_t timestamp = std::uint32_t(total_microseconds(
			receive_time.time_since_epoch()) & 0xffffffff);
		m_reply_micro = timestamp - ph->timestamp_microseconds;
		std::uint32_t const prev_base = m_their_delay_hist.initialized() ? m_their_delay_hist.base() : 0;
		their_delay = m_their_delay_hist.add_sample(m_reply_micro, step);
		int const base_change = int(m_their_delay_hist.base() - prev_base);
		UTP_LOGV("%8p: their_delay::add_sample:%u prev_base:%u new_base:%u\n"
			, static_cast<void*>(this), m_reply_micro, prev_base, m_their_delay_hist.base());

		if (prev_base && base_change < 0 && base_change > -10000 && m_delay_hist.initialized())
		{
			// their base delay went down. This is caused by clock drift. To compensate,
			// adjust our base delay upwards
			// don't adjust more than 10 ms. If the change is that big, something is probably wrong
			m_delay_hist.adjust_base(-base_change);
		}

		UTP_LOGV("%8p: incoming packet reply_micro:%u base_change:%d\n"
			, static_cast<void*>(this), m_reply_micro, prev_base ? base_change : 0);
	}

	// is this ACK valid? If the other end is ACKing
	// a packet that hasn't been sent yet
	// just ignore it. A 3rd party could easily inject a packet
	// like this in a stream, don't sever it because of it.
	// since m_seq_nr is the sequence number of the next packet
	// we'll send (and m_seq_nr-1 was the last packet we sent),
	// if the ACK we got is greater than the last packet we sent
	// something is wrong.
	// If our state is state_none, this packet must be a syn packet
	// and the ack_nr should be ignored
	std::uint16_t const cmp_seq_nr =
		(m_state == UTP_STATE_SYN_SENT && ph->get_type() == ST_STATE)
		? m_seq_nr : (m_seq_nr - 1) & ACK_MASK;

	if ((m_state != UTP_STATE_NONE || ph->get_type() != ST_SYN)
		&& (compare_less_wrap(cmp_seq_nr, ph->ack_nr, ACK_MASK)
			|| compare_less_wrap(ph->ack_nr, m_acked_seq_nr
				- dup_ack_limit, ACK_MASK)))
	{
		UTP_LOG("%8p: ERROR: incoming packet ack_nr:%d our seq_nr:%d our "
			"acked_seq_nr:%d (ignored)\n"
			, static_cast<void*>(this), int(ph->ack_nr), m_seq_nr, m_acked_seq_nr);
		m_sm.inc_stats_counter(counters::utp_redundant_pkts_in);
		return true;
	}

	// check to make sure the sequence number of this packet
	// is reasonable. If it's a data packet and we've already
	// received it, ignore it. This is either a stray old packet
	// that finally made it here (after having been re-sent) or
	// an attempt to interfere with the connection from a 3rd party
	// in both cases, we can safely ignore the timestamp and ACK
	// information in this packet
/*
	// even if we've already received this packet, we need to
	// send another ack to it, since it may be a resend caused by
	// our ack getting dropped
	if (m_state != UTP_STATE_SYN_SENT
		&& ph->get_type() == ST_DATA
		&& !compare_less_wrap(m_ack_nr, ph->seq_nr, ACK_MASK))
	{
		// we've already received this packet
		UTP_LOGV("%8p: incoming packet seq_nr:%d our ack_nr:%d (ignored)\n"
			, static_cast<void*>(this), int(ph->seq_nr), m_ack_nr);
		m_sm.inc_stats_counter(counters::utp_redundant_pkts_in);
		return true;
	}
*/

	// if the socket is closing, always ignore any packet
	// with a higher sequence number than the FIN sequence number
	// ST_STATE messages always include the next seqnr.
	if (m_eof && (compare_less_wrap(m_eof_seq_nr, ph->seq_nr, ACK_MASK)
		|| (m_eof_seq_nr == ph->seq_nr && ph->get_type() != ST_STATE)))
	{
#if TORRENT_UTP_LOG
		UTP_LOG("%8p: ERROR: incoming packet type: %s seq_nr:%d eof_seq_nr:%d (ignored)\n"
			, static_cast<void*>(this), packet_type_names[ph->get_type()], int(ph->seq_nr), m_eof_seq_nr);
#endif
		return true;
	}

	if (ph->get_type() == ST_DATA)
		m_sm.inc_stats_counter(counters::utp_payload_pkts_in);

	// the number of packets that'll fit in the reorder buffer
	std::uint32_t const max_packets_reorder
		= static_cast<std::uint32_t>(std::max(16, m_receive_buffer_capacity / 1100));

	if (m_state != UTP_STATE_NONE
		&& m_state != UTP_STATE_SYN_SENT
		&& compare_less_wrap((m_ack_nr + max_packets_reorder) & ACK_MASK, ph->seq_nr, ACK_MASK))
	{
		// this is too far out to fit in our reorder buffer. Drop it
		// This is either an attack to try to break the connection
		// or a seriously damaged connection that lost a lot of
		// packets. Neither is very likely, and it should be OK
		// to drop the timestamp information.
		UTP_LOG("%8p: ERROR: incoming packet seq_nr:%d our ack_nr:%d (ignored)\n"
			, static_cast<void*>(this), int(ph->seq_nr), m_ack_nr);
		m_sm.inc_stats_counter(counters::utp_redundant_pkts_in);
		return true;
	}

	if (ph->get_type() == ST_RESET)
	{
		if (compare_less_wrap(cmp_seq_nr, ph->ack_nr, ACK_MASK))
		{
			UTP_LOG("%8p: ERROR: invalid RESET packet, ack_nr:%d our seq_nr:%d (ignored)\n"
				, static_cast<void*>(this), int(ph->ack_nr), m_seq_nr);
			return true;
		}
		UTP_LOGV("%8p: incoming packet type:RESET\n", static_cast<void*>(this));
		m_error = boost::asio::error::connection_reset;
		set_state(UTP_STATE_ERROR_WAIT);
		test_socket_state();
		return true;
	}

	++m_in_packets;

	// this is a valid incoming packet, update the timeout timer
	m_num_timeouts = 0;
	m_timeout = receive_time + milliseconds(packet_timeout());
	UTP_LOGV("%8p: updating timeout to: now + %d\n"
		, static_cast<void*>(this), packet_timeout());

	// the test for INT_MAX here is a work-around for a bug in uTorrent where
	// it's sometimes sent as INT_MAX when it is in fact uninitialized
	const std::uint32_t sample = ph->timestamp_difference_microseconds == INT_MAX
		? 0 : ph->timestamp_difference_microseconds;

	std::uint32_t delay = 0;
	if (sample != 0)
	{
		delay = m_delay_hist.add_sample(sample, step);
		m_delay_sample_hist[m_delay_sample_idx++] = delay;
		if (m_delay_sample_idx >= m_delay_sample_hist.size())
			m_delay_sample_idx = 0;
	}

	int acked_bytes = 0;

	TORRENT_ASSERT(m_bytes_in_flight >= 0);
	int prev_bytes_in_flight = m_bytes_in_flight;

	m_adv_wnd = ph->wnd_size;

	// if we get an ack for the same sequence number as
	// was last ACKed, and we have outstanding packets,
	// it counts as a duplicate ack. The reason to not count ST_DATA packets as
	// duplicate ACKs is because we may be receiving a stream of those
	// regardless of our outgoing traffic, which makes their ACK number not
	// indicative of a dropped packet
	if (ph->ack_nr == m_acked_seq_nr
		&& m_outbuf.size()
		&& ph->get_type() == ST_STATE)
	{
		++m_duplicate_acks;
	}

	TORRENT_ASSERT_VAL(m_outbuf.size() > 0 || m_duplicate_acks == 0, m_duplicate_acks);

	std::uint32_t min_rtt = std::numeric_limits<std::uint32_t>::max();

	TORRENT_ASSERT(m_outbuf.at((m_acked_seq_nr + 1) & ACK_MASK) || ((m_seq_nr - m_acked_seq_nr) & ACK_MASK) <= 1);

	// has this packet already been ACKed?
	// if the ACK we just got is less than the max ACKed
	// sequence number, it doesn't tell us anything.
	// So, only act on it if the ACK is greater than the last acked
	// sequence number
	if (m_state != UTP_STATE_NONE && compare_less_wrap(m_acked_seq_nr, ph->ack_nr, ACK_MASK))
	{
		int const next_ack_nr = ph->ack_nr;

		for (int ack_nr = (m_acked_seq_nr + 1) & ACK_MASK;
			ack_nr != ((next_ack_nr + 1) & ACK_MASK);
			ack_nr = (ack_nr + 1) & ACK_MASK)
		{
			if (m_fast_resend_seq_nr == ack_nr)
				m_fast_resend_seq_nr = (m_fast_resend_seq_nr + 1) & ACK_MASK;
			packet_ptr p = m_outbuf.remove(aux::numeric_cast<packet_buffer::index_type>(ack_nr));

			if (!p) continue;

			acked_bytes += p->size - p->header_size;
			std::uint32_t const rtt = ack_packet(std::move(p), receive_time, std::uint16_t(ack_nr));
			min_rtt = std::min(min_rtt, rtt);
		}

		maybe_inc_acked_seq_nr();
		if (m_outbuf.empty()) m_duplicate_acks = 0;
	}

	// look for extended headers
	std::uint8_t const* ptr = buf.data();
	int const size = int(buf.size());
	ptr += sizeof(utp_header);

	std::uint8_t extension = ph->extension;
	while (extension)
	{
		// invalid packet. It says it has an extension header
		// but the packet is too short
		if (ptr - buf.data() + 2 > size)
		{
			UTP_LOG("%8p: ERROR: invalid extension header\n", static_cast<void*>(this));
			m_sm.inc_stats_counter(counters::utp_invalid_pkts_in);
			return true;
		}
		std::uint8_t const next_extension = *ptr++;
		int const len = *ptr++;
		if (ptr - buf.data() + len > size)
		{
			UTP_LOG("%8p: ERROR: invalid extension header size:%d packet:%d\n"
				, static_cast<void*>(this), len, int(ptr - buf.data()));
			m_sm.inc_stats_counter(counters::utp_invalid_pkts_in);
			return true;
		}
		switch(extension)
		{
			case utp_sack: // selective ACKs
			{
				std::uint32_t rtt;
				std::tie(rtt, acked_bytes) = parse_sack(ph->ack_nr, ptr, len, receive_time);
				min_rtt = std::min(min_rtt, rtt);
				break;
			}
			case utp_close_reason:
				parse_close_reason(ptr, len);
				break;
		}
		ptr += len;
		extension = next_extension;
	}

	// the send operation in parse_sack() may have set the socket to an error
	// state, in which case we shouldn't continue
	if (m_state == UTP_STATE_ERROR_WAIT || m_state == UTP_STATE_DELETE) return true;

	if (m_duplicate_acks >= dup_ack_limit
		&& ((m_acked_seq_nr + 1) & ACK_MASK) == m_fast_resend_seq_nr)
	{
		// LOSS

		UTP_LOGV("%8p: Packet %d lost. (%d duplicate acks, trigger fast-resend)\n"
			, static_cast<void*>(this), m_fast_resend_seq_nr, m_duplicate_acks);

		// resend the lost packet
		packet* p = m_outbuf.at(m_fast_resend_seq_nr);
		TORRENT_ASSERT(p);

		// don't fast-resend this again
		m_fast_resend_seq_nr = (m_fast_resend_seq_nr + 1) & ACK_MASK;

		if (p)
		{
			// don't consider a lost probe as proper loss, it doesn't necessarily
			// signal congestion
			if (!p->mtu_probe) experienced_loss(m_fast_resend_seq_nr, receive_time);
			resend_packet(p, true);
			if (m_state == UTP_STATE_ERROR_WAIT || m_state == UTP_STATE_DELETE) return true;
		}
	}

	// ptr points to the payload of the packet
	// size is the packet size, payload is the
	// number of payload bytes are in this packet
	const int header_size = int(ptr - buf.data());
	const int payload_size = size - header_size;

#if TORRENT_UTP_LOG
	UTP_LOGV("%8p: incoming packet seq_nr:%d ack_nr:%d type:%s id:%d size:%d timestampdiff:%u timestamp:%u "
			"our ack_nr:%d our seq_nr:%d our acked_seq_nr:%d our state:%s\n"
		, static_cast<void*>(this), int(ph->seq_nr), int(ph->ack_nr), packet_type_names[ph->get_type()]
		, int(ph->connection_id), payload_size, std::uint32_t(ph->timestamp_difference_microseconds)
		, std::uint32_t(ph->timestamp_microseconds), m_ack_nr, m_seq_nr, m_acked_seq_nr, socket_state_names[m_state]);
#endif

	if (ph->get_type() == ST_FIN)
	{
		// We ignore duplicate FIN packets, but we still need to ACK them.
		if (ph->seq_nr == ((m_ack_nr + 1) & ACK_MASK)
			|| ph->seq_nr == m_ack_nr)
		{
			UTP_LOGV("%8p: FIN received in order\n", static_cast<void*>(this));

			// The FIN arrived in order, nothing else is in the
			// reorder buffer.

//			TORRENT_ASSERT(m_inbuf.size() == 0);
			m_ack_nr = ph->seq_nr;

			// Transition to UTP_STATE_FIN_SENT. The sent FIN is also an ack
			// to the FIN we received. Once we're in UTP_STATE_FIN_SENT we
			// just need to wait for our FIN to be acked.

			if (m_state == UTP_STATE_FIN_SENT)
			{
				send_pkt(pkt_ack);
				if (m_state == UTP_STATE_ERROR_WAIT || m_state == UTP_STATE_DELETE) return true;
			}
			else
			{
				send_fin();
				if (m_state == UTP_STATE_ERROR_WAIT || m_state == UTP_STATE_DELETE) return true;
			}
		}

		if (m_eof)
		{
			UTP_LOGV("%8p: duplicate FIN packet (ignoring)\n", static_cast<void*>(this));
			return true;
		}
		m_eof = true;
		m_eof_seq_nr = ph->seq_nr;

		// we will respond with a fin once we have received everything up to m_eof_seq_nr
	}

	switch (m_state)
	{
		case UTP_STATE_NONE:
		{
			if (ph->get_type() == ST_SYN)
			{
				// if we're in state_none, the only thing
				// we accept are SYN packets.
				set_state(UTP_STATE_CONNECTED);

				m_remote_address = ep.address();
				m_port = ep.port();

				m_ack_nr = ph->seq_nr;
				m_seq_nr = std::uint16_t(random(0xffff));
				m_acked_seq_nr = (m_seq_nr - 1) & ACK_MASK;
				m_loss_seq_nr = m_acked_seq_nr;
				m_fast_resend_seq_nr = m_seq_nr;

#if TORRENT_UTP_LOG
				UTP_LOGV("%8p: received ST_SYN state:%s seq_nr:%d ack_nr:%d\n"
					, static_cast<void*>(this), socket_state_names[m_state], m_seq_nr, m_ack_nr);
#endif
				if (m_send_id != ph->connection_id)
				{
#if TORRENT_UTP_LOG
					UTP_LOGV("%8p: received invalid connection_id:%d expected: %d\n"
						, static_cast<void*>(this), int(ph->connection_id), int(m_send_id));
#endif
					return false;
				}
				TORRENT_ASSERT(m_recv_id == ((m_send_id + 1) & 0xffff));

				defer_ack();
			}
			else
			{
#if TORRENT_UTP_LOG
				UTP_LOG("%8p: ERROR: type:%s state:%s (ignored)\n"
					, static_cast<void*>(this), packet_type_names[ph->get_type()], socket_state_names[m_state]);
#endif
			}
			break;
		}
		case UTP_STATE_SYN_SENT:
		{
			// just wait for an ack to our SYN, ignore everything else
			if (ph->ack_nr != ((m_seq_nr - 1) & ACK_MASK))
			{
#if TORRENT_UTP_LOG
				UTP_LOGV("%8p: incorrect ack_nr (%d) waiting for %d\n"
					, static_cast<void*>(this), int(ph->ack_nr), (m_seq_nr - 1) & ACK_MASK);
#endif
				break;
			}

			TORRENT_ASSERT(!m_error);
			set_state(UTP_STATE_CONNECTED);
#if TORRENT_UTP_LOG
			UTP_LOGV("%8p: state:%s\n", static_cast<void*>(this), socket_state_names[m_state]);
#endif

			// only progress our ack_nr on ST_DATA messages
			// since our m_ack_nr is uninitialized at this point
			// we still need to set it to something regardless
			if (ph->get_type() == ST_DATA)
				m_ack_nr = ph->seq_nr;
			else
				m_ack_nr = (ph->seq_nr - 1) & ACK_MASK;

			// notify the client that the socket connected
			if (m_connect_handler)
			{
				UTP_LOGV("%8p: calling connect handler\n", static_cast<void*>(this));
				m_connect_handler = false;
				utp_stream::on_connect(m_userdata, m_error, false);
			}
			BOOST_FALLTHROUGH;
		}
		// fall through
		case UTP_STATE_CONNECTED:
		{
			// the lowest seen RTT can be used to clamp the delay
			// within reasonable bounds. The one-way delay is never
			// higher than the round-trip time.

			if (sample && acked_bytes && prev_bytes_in_flight)
			{
				// only use the minimum from the last 3 delay measurements
				delay = *std::min_element(m_delay_sample_hist.begin()
					, m_delay_sample_hist.end());

				// it's impossible for delay to be more than the RTT, so make
				// sure to clamp it as a sanity check
				if (delay > min_rtt) delay = min_rtt;

				do_ledbat(acked_bytes, int(delay), prev_bytes_in_flight);
				m_send_delay = std::int32_t(delay);
			}

			m_recv_delay = std::int32_t(std::min(their_delay, min_rtt));

			consume_incoming_data(ph, ptr, payload_size, receive_time);

			// the parameter to send_pkt tells it if we're acking data
			// If we are, we'll send an ACK regardless of if we have any
			// space left in our send window or not. If we just got an ACK
			// (i.e. ST_STATE) we're not ACKing anything. If we just
			// received a FIN packet, we need to ack that as well
			bool has_ack = ph->get_type() == ST_DATA || ph->get_type() == ST_FIN || ph->get_type() == ST_SYN;
			std::uint32_t prev_out_packets = m_out_packets;

			// the connection is connected and this packet made it past all the
			// checks. We can now assume the other end is not spoofing it's IP.
			if (ph->get_type() != ST_SYN) m_confirmed = true;

			// try to send more data as long as we can
			// if send_pkt returns true
			while (send_pkt());

			if (has_ack && prev_out_packets == m_out_packets)
			{
				// we need to ack some data we received, and we didn't
				// end up sending any payload packets in the loop
				// above (because m_out_packets would have been incremented
				// in that case). This means we need to send an ack.
				// don't do it right away, because we may still receive
				// more packets. defer the ack to send as few acks as possible
				defer_ack();
			}

			// we may want to call the user callback function at the end
			// of this round. Subscribe to that event
			subscribe_drained();

			if (m_state == UTP_STATE_ERROR_WAIT || m_state == UTP_STATE_DELETE) return true;

			// Everything up to the FIN has been received, respond with a FIN
			// from our side.
			if (m_eof && m_ack_nr == ((m_eof_seq_nr - 1) & ACK_MASK))
			{
				UTP_LOGV("%8p: incoming stream consumed\n", static_cast<void*>(this));

				// This transitions to the UTP_STATE_FIN_SENT state.
				send_fin();
				if (m_state == UTP_STATE_ERROR_WAIT || m_state == UTP_STATE_DELETE) return true;
			}

			TORRENT_ASSERT(!compare_less_wrap(m_seq_nr, m_acked_seq_nr, ACK_MASK));

#if TORRENT_UTP_LOG
			if (sample && acked_bytes && prev_bytes_in_flight)
			{
				char their_delay_base[20];
				if (m_their_delay_hist.initialized())
					std::snprintf(their_delay_base, sizeof(their_delay_base), "%u", m_their_delay_hist.base());
				else
					strcpy(their_delay_base, "-");

				char our_delay_base[20];
				if (m_delay_hist.initialized())
					std::snprintf(our_delay_base, sizeof(our_delay_base), "%u", m_delay_hist.base());
				else
					strcpy(our_delay_base, "-");

				UTP_LOG("%8p: "
					"actual_delay:%u "
					"our_delay:%f "
					"their_delay:%f "
					"off_target:%f "
					"max_window:%u "
					"upload_rate:%d "
					"delay_base:%s "
					"delay_sum:%f "
					"target_delay:%d "
					"acked_bytes:%d "
					"cur_window:%d "
					"scaled_gain:%f "
					"rtt:%u "
					"rate:%d "
					"quota:%d "
					"wnduser:%u "
					"rto:%d "
					"timeout:%d "
					"get_microseconds:%u "
					"cur_window_packets:%u "
					"packet_size:%d "
					"their_delay_base:%s "
					"their_actual_delay:%u "
					"seq_nr:%u "
					"acked_seq_nr:%u "
					"reply_micro:%u "
					"min_rtt:%u "
					"send_buffer:%d "
					"recv_buffer:%d "
					"fast_resend_seq_nr:%d "
					"ssthres:%d "
					"\n"
					, static_cast<void*>(this)
					, sample
					, delay / 1000.0
					, their_delay / 1000.0
					, int(m_sm.target_delay() - delay) / 1000.0
					, std::uint32_t(m_cwnd >> 16)
					, 0
					, our_delay_base
					, (delay + their_delay) / 1000.0
					, m_sm.target_delay() / 1000
					, acked_bytes
					, m_bytes_in_flight
					, 0.0 // float(scaled_gain)
					, m_rtt.mean()
					, int((m_cwnd * 1000 / (m_rtt.mean()?m_rtt.mean():50)) >> 16)
					, 0
					, m_adv_wnd
					, packet_timeout()
					, int(total_milliseconds(m_timeout - receive_time))
					, int(total_microseconds(receive_time.time_since_epoch()))
					, m_outbuf.size()
					, m_mtu
					, their_delay_base
					, std::uint32_t(m_reply_micro)
					, m_seq_nr
					, m_acked_seq_nr
					, m_reply_micro
					, min_rtt / 1000
					, m_write_buffer_size
					, m_read_buffer_size
					, m_fast_resend_seq_nr
					, m_ssthres);
			}
#endif

			break;
		}
		case UTP_STATE_FIN_SENT:
		{
			// There are two ways we can end up in this state:
			//
			// 1. If the socket has been explicitly closed on our
			//    side, in which case m_eof is false.
			//
			// 2. If we received a FIN from the remote side, in which
			//    case m_eof is true. If this is the case, we don't
			//    come here until everything up to the FIN has been
			//    received.
			//
			//
			//

			// At this point m_seq_nr - 1 is the FIN sequence number.

			// We can receive both ST_DATA and ST_STATE here, because after
			// we have closed our end of the socket, the remote end might
			// have data in the pipeline. We don't really care about the
			// data, but we do have to ack it. Or rather, we have to ack
			// the FIN that will come after the data.

			// Case 1:
			// ---------------------------------------------------------------
			//
			// If we are here because the local endpoint was closed, we need
			// to first wait for all of our messages to be acked:
			//
			//   if (m_acked_seq_nr == ((m_seq_nr - 1) & ACK_MASK))
			//
			// `m_seq_nr - 1` is the ST_FIN message that we sent.
			//
			//                     ----------------------
			//
			// After that has happened we need to wait for the remote side
			// to send its ST_FIN message. When we receive that we send an
			// ST_STATE back to ack, and wait for a sufficient period.
			// During this wait we keep acking incoming ST_FIN's. This is
			// all handled at the top of this function.
			//
			// Note that the user handlers are all cancelled when the initial
			// close() call happens, so nothing will happen on the user side
			// after that.

			// Case 2:
			// ---------------------------------------------------------------
			//
			// If we are here because we received a ST_FIN message, and then
			// sent our own ST_FIN to ack that, we need to wait for our ST_FIN
			// to be acked:
			//
			//   if (m_acked_seq_nr == ((m_seq_nr - 1) & ACK_MASK))
			//
			// `m_seq_nr - 1` is the ST_FIN message that we sent.
			//
			// After that has happened we know the remote side has all our
			// data, and we can gracefully shut down.

			if (consume_incoming_data(ph, ptr, payload_size, receive_time))
			{
				break;
			}

			if (m_acked_seq_nr == ((m_seq_nr - 1) & ACK_MASK))
			{
				// When this happens we know that the remote side has
				// received all of our packets.

				UTP_LOGV("%8p: FIN acked\n", static_cast<void*>(this));

				if (!m_attached)
				{
					UTP_LOGV("%8p: close initiated here, delete socket\n"
						, static_cast<void*>(this));
					m_error = boost::asio::error::eof;
					set_state(UTP_STATE_DELETE);
					test_socket_state();
				}
				else
				{
					UTP_LOGV("%8p: closing socket\n", static_cast<void*>(this));
					m_error = boost::asio::error::eof;
					set_state(UTP_STATE_ERROR_WAIT);
					test_socket_state();
				}
			}

			break;
		}
		case UTP_STATE_DELETE:
		default:
		{
			// respond with a reset
			send_reset(ph);
			break;
		}
	}
	return true;
}

void utp_socket_impl::do_ledbat(const int acked_bytes, const int delay
	, const int in_flight)
{
	INVARIANT_CHECK;

	// the portion of the in-flight bytes that were acked. This is used to make
	// the gain factor be scaled by the rtt. The formula is applied once per
	// rtt, or on every ACK scaled by the number of ACKs per rtt
	TORRENT_ASSERT(in_flight > 0);
	TORRENT_ASSERT(acked_bytes > 0);

	const int target_delay = std::max(1, m_sm.target_delay());

	// true if the upper layer is pushing enough data down the socket to be
	// limited by the cwnd. If this is not the case, we should not adjust cwnd.
	const bool cwnd_saturated = (m_bytes_in_flight + acked_bytes + m_mtu > (m_cwnd >> 16));

	// all of these are fixed points with 16 bits fraction portion
	const std::int64_t window_factor = (std::int64_t(acked_bytes) * (1 << 16)) / in_flight;
	const std::int64_t delay_factor = (std::int64_t(target_delay - delay) * (1 << 16)) / target_delay;
	std::int64_t scaled_gain;

	if (delay >= target_delay)
	{
		if (m_slow_start)
		{
			UTP_LOGV("%8p: off_target: %d slow_start -> 0\n"
				, static_cast<void*>(this), target_delay - delay);
			m_ssthres = std::int32_t((m_cwnd >> 16) / 2);
			m_slow_start = false;
		}

		m_sm.inc_stats_counter(counters::utp_samples_above_target);
	}
	else
	{
		m_sm.inc_stats_counter(counters::utp_samples_below_target);
	}

	std::int64_t const linear_gain = ((window_factor * delay_factor) >> 16)
		* std::int64_t(m_sm.gain_factor());

	// if the user is not saturating the link (i.e. not filling the
	// congestion window), don't adjust it at all.
	if (cwnd_saturated)
	{
		std::int64_t const exponential_gain = std::int64_t(acked_bytes) * (1 << 16);
		if (m_slow_start)
		{
			// mimic TCP slow-start by adding the number of acked
			// bytes to cwnd
			if (m_ssthres != 0 && ((m_cwnd + exponential_gain) >> 16) > m_ssthres)
			{
				// if we would exceed the slow start threshold by growing the cwnd
				// exponentially, don't do it, and leave slow-start mode. This
				// make us avoid causing more delay and/or packet loss by being too
				// aggressive
				m_slow_start = false;
				scaled_gain = linear_gain;
				UTP_LOGV("%8p: cwnd > ssthres (%d) slow_start -> 0\n"
					, static_cast<void*>(this), m_ssthres);
			}
			else
			{
				scaled_gain = std::max(exponential_gain, linear_gain);
			}
		}
		else
		{
			scaled_gain = linear_gain;
		}
	}
	else
	{
		scaled_gain = 0;
	}

	// make sure we don't wrap the cwnd
	if (scaled_gain >= std::numeric_limits<std::int64_t>::max() - m_cwnd)
		scaled_gain = std::numeric_limits<std::int64_t>::max() - m_cwnd - 1;

	UTP_LOGV("%8p: do_ledbat delay:%d off_target: %d window_factor:%f target_factor:%f "
		"scaled_gain:%f cwnd:%d slow_start:%d\n"
		, static_cast<void*>(this), delay, target_delay - delay, window_factor / double(1 << 16)
		, delay_factor / double(1 << 16)
		, scaled_gain / double(1 << 16), int(m_cwnd >> 16)
		, int(m_slow_start));

	// if scaled_gain + m_cwnd <= 0, set m_cwnd to 0
	if (-scaled_gain >= m_cwnd)
	{
		m_cwnd = 0;
	}
	else
	{
		m_cwnd += scaled_gain;
		TORRENT_ASSERT(m_cwnd > 0);
	}

	TORRENT_ASSERT(m_cwnd >= 0);

	int const window_size_left = std::min(int(m_cwnd >> 16), int(m_adv_wnd)) - in_flight + acked_bytes;
	if (window_size_left >= m_mtu)
	{
		UTP_LOGV("%8p: mtu:%d in_flight:%d adv_wnd:%d cwnd:%d acked_bytes:%d cwnd_full -> 0\n"
			, static_cast<void*>(this), m_mtu, in_flight, int(m_adv_wnd), int(m_cwnd >> 16), acked_bytes);
		m_cwnd_full = false;
	}
/*
	if ((m_cwnd >> 16) >= m_adv_wnd)
	{
		m_slow_start = false;
		m_ssthres = (m_cwnd >> 16);
		UTP_LOGV("%8p: cwnd > advertized wnd (%u) slow_start -> 0\n"
			, static_cast<void*>(this), m_adv_wnd);
	}
*/
}

void utp_stream::bind(endpoint_type const&, error_code&) { }

void utp_stream::cancel_handlers(error_code const& ec)
{
	if (!m_impl) return;
	m_impl->cancel_handlers(ec, false);
}
// returns the number of milliseconds a packet would have before
// it would time-out if it was sent right now. Takes the RTT estimate
// into account
int utp_socket_impl::packet_timeout() const
{
	INVARIANT_CHECK;

	// SYN packets have a bit longer timeout, since we don't
	// have an RTT estimate yet, make a conservative guess
	if (m_state == UTP_STATE_NONE) return 3000;

	// avoid overflow by simply capping based on number of timeouts as well
	if (m_num_timeouts >= 7) return 60000;

	int timeout = std::max(m_sm.min_timeout(), m_rtt.mean() + m_rtt.avg_deviation() * 2);
	if (m_num_timeouts > 0) timeout += (1 << (int(m_num_timeouts) - 1)) * 1000;

	// timeouts over 1 minute are capped
	if (timeout > 60000) timeout = 60000;
	return timeout;
}

void utp_socket_impl::tick(time_point const now)
{
	INVARIANT_CHECK;

#if TORRENT_UTP_LOG
	UTP_LOGV("%8p: tick:%s r: %d (%s) w: %d (%s)\n"
		, static_cast<void*>(this), socket_state_names[m_state], m_read, m_read_handler ? "handler" : "no handler"
		, m_written, m_write_handler ? "handler" : "no handler");
#endif

	TORRENT_ASSERT(m_outbuf.at((m_acked_seq_nr + 1) & ACK_MASK) || ((m_seq_nr - m_acked_seq_nr) & ACK_MASK) <= 1);

	// if we're already in an error state, we're just waiting for the
	// client to perform an operation so that we can communicate the
	// error. No need to do anything else with this socket
	if (m_state == UTP_STATE_ERROR_WAIT || m_state == UTP_STATE_DELETE) return;

	if (now > m_timeout)
	{
		// TIMEOUT!

		bool ignore_loss = false;

		if (((m_acked_seq_nr + 1) & ACK_MASK) == m_mtu_seq
			&& ((m_seq_nr - 1) & ACK_MASK) == m_mtu_seq
			&& m_mtu_seq != 0)
		{
			// we timed out, and the only outstanding packet
			// we had was the probe. Assume it was dropped
			// because it was too big
			m_mtu_ceiling = m_mtu - 1;
			update_mtu_limits();
			ignore_loss = true;
		}

		// the close_reason here is a bit of a hack. When it's set, it indicates
		// that the upper layer intends to close the socket. However, it has been
		// observed that the SSL shutdown sometimes can hang in a state where
		// there's no outstanding data, and it won't receive any more from the
		// other end. This catches that case and let the socket time out.
		if (m_outbuf.size() || m_close_reason != close_reason_t::none)
		{
			// m_num_timeouts is used to update the connection timeout, and if we
			// lose this packet because it's an MTU-probe, don't change the timeout
			if (!ignore_loss) ++m_num_timeouts;
			m_sm.inc_stats_counter(counters::utp_timeout);
		}

		UTP_LOGV("%8p: timeout num-timeouts: %d max-resends: %d confirmed: %d "
			" acked-seq-num: %d mtu-seq: %d\n"
			, static_cast<void*>(this)
			, m_num_timeouts
			, m_sm.num_resends()
			, m_confirmed
			, m_acked_seq_nr
			, m_mtu_seq);

		// a socket that has not been confirmed to actually have a live remote end
		// (the IP may have been spoofed) fail on the first timeout. If we had
		// heard anything from this peer, it would have been confirmed.
		if (m_num_timeouts > m_sm.num_resends()
			|| (m_num_timeouts > 0 && !m_confirmed))
		{
			// the connection is dead
			m_error = boost::asio::error::timed_out;
			set_state(UTP_STATE_ERROR_WAIT);
			test_socket_state();
			return;
		}

		if (!ignore_loss)
		{
			// set cwnd to 1 MSS
			if (m_bytes_in_flight == 0 && (m_cwnd >> 16) >= m_mtu)
			{
				// this is just a timeout because this direction of
				// the stream is idle. Don't reset the cwnd, just decay it
				m_cwnd = std::max(m_cwnd * 2 / 3, std::int64_t(m_mtu) * (1 << 16));
			}
			else
			{
				// we timed out because a packet was not ACKed or because
				// the cwnd was made smaller than one packet
				m_cwnd = std::int64_t(m_mtu) * (1 << 16);
			}

			TORRENT_ASSERT(m_cwnd >= 0);

			m_timeout = now + milliseconds(packet_timeout());

			UTP_LOGV("%8p: resetting cwnd:%d\n"
				, static_cast<void*>(this), int(m_cwnd >> 16));

			// since we've already timed out now, don't count
			// loss that we might detect for packets that just
			// timed out
			m_loss_seq_nr = m_seq_nr;

			// when we time out, the cwnd is reset to 1 MSS, which means we
			// need to ramp it up quickly again. enter slow start mode. This time
			// we're very likely to have an ssthres set, which will make us leave
			// slow start before inducing more delay or loss.
			m_slow_start = true;
			UTP_LOGV("%8p: slow_start -> 1\n", static_cast<void*>(this));
		}

		// we dropped all packets, that includes the mtu probe
		m_mtu_seq = 0;

		// we need to go one past m_seq_nr to cover the case
		// where we just sent a SYN packet and then adjusted for
		// the uTorrent sequence number reuse
		for (int i = m_acked_seq_nr & ACK_MASK;
			i != ((m_seq_nr + 1) & ACK_MASK);
			i = (i + 1) & ACK_MASK)
		{
			packet* p = m_outbuf.at(aux::numeric_cast<packet_buffer::index_type>(i));
			if (!p) continue;
			if (p->need_resend) continue;
			p->need_resend = true;
			TORRENT_ASSERT(m_bytes_in_flight >= p->size - p->header_size);
			m_bytes_in_flight -= p->size - p->header_size;
			UTP_LOGV("%8p: Packet %d lost (timeout).\n", static_cast<void*>(this), i);
		}

		TORRENT_ASSERT(m_bytes_in_flight == 0);

		// if we have a packet that needs re-sending, resend it
		packet* p = m_outbuf.at((m_acked_seq_nr + 1) & ACK_MASK);
		if (p)
		{
			if (p->num_transmissions >= m_sm.num_resends()
				|| (m_state == UTP_STATE_SYN_SENT && p->num_transmissions >= m_sm.syn_resends())
				|| (m_state == UTP_STATE_FIN_SENT && p->num_transmissions >= m_sm.fin_resends()))
			{
#if TORRENT_UTP_LOG
				UTP_LOGV("%8p: %d failed sends in a row. Socket timed out. state:%s\n"
					, static_cast<void*>(this), p->num_transmissions, socket_state_names[m_state]);
#endif

				if (p->size > m_mtu_floor)
				{
					// the packet that caused the connection to fail was an mtu probe
					// (note that the mtu_probe field won't be set at this point because
					// it's cleared when the packet is re-sent). This suggests that
					// perhaps our network throws away oversized packets without
					// fragmenting them. Tell the socket manager to be more conservative
					// about mtu ceiling in the future
					m_sm.restrict_mtu(m_mtu);
				}
				// the connection is dead
				m_error = boost::asio::error::timed_out;
				set_state(UTP_STATE_ERROR_WAIT);
				test_socket_state();
				return;
			}

			// don't fast-resend this packet
			if (m_fast_resend_seq_nr == ((m_acked_seq_nr + 1) & ACK_MASK))
				m_fast_resend_seq_nr = (m_fast_resend_seq_nr + 1) & ACK_MASK;

			// the packet timed out, resend it
			resend_packet(p);
			if (m_state == UTP_STATE_ERROR_WAIT || m_state == UTP_STATE_DELETE) return;
		}
		else if (m_state < UTP_STATE_FIN_SENT)
		{
			send_pkt();
			if (m_state == UTP_STATE_ERROR_WAIT || m_state == UTP_STATE_DELETE) return;
		}
		else if (m_state == UTP_STATE_FIN_SENT)
		{
			// the connection is dead
			m_error = boost::asio::error::eof;
			set_state(UTP_STATE_ERROR_WAIT);
			test_socket_state();
			return;
		}
	}

	switch (m_state)
	{
		case UTP_STATE_NONE:
		case UTP_STATE_DELETE:
			return;
//		case UTP_STATE_SYN_SENT:
//
//			break;
	}
}

#if TORRENT_USE_INVARIANT_CHECKS
void utp_socket_impl::check_receive_buffers() const
{
	INVARIANT_CHECK;

	int const size = std::accumulate(m_receive_buffer.begin(), m_receive_buffer.end(), 0
		, [](int const acc, packet_ptr const& p) { return acc + (p ? p->size - p->header_size : 0); } );

	TORRENT_ASSERT(size == m_receive_buffer_size);
}
#endif

#if TORRENT_USE_INVARIANT_CHECKS
void utp_socket_impl::check_invariant() const
{
	for (packet_buffer::index_type i = m_outbuf.cursor();
		i != ((m_outbuf.cursor() + m_outbuf.span()) & ACK_MASK);
		i = (i + 1) & ACK_MASK)
	{
		packet* p = m_outbuf.at(i);
		if (!p) continue;
		if (m_mtu_seq == i && m_mtu_seq != 0)
		{
			TORRENT_ASSERT(p->mtu_probe);
		}
		TORRENT_ASSERT(reinterpret_cast<utp_header*>(p->buf)->seq_nr == i);
	}

	if (m_nagle_packet)
	{
		// if this packet is full, it should have been sent
		TORRENT_ASSERT(m_nagle_packet->size < m_nagle_packet->allocated);
	}
}
#endif
}