FreeRDP/libfreerdp/core/transport.c

/**
 * FreeRDP: A Remote Desktop Protocol Implementation
 * Network Transport Layer
 *
 * Copyright 2011 Vic Lee
 *
 * Licensed under the Apache License, Version 2.0 (the "License");
 * you may not use this file except in compliance with the License.
 * You may obtain a copy of the License at
 *
 *     http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */

#ifdef HAVE_CONFIG_H
#include "config.h"
#endif

#include <assert.h>

#include <winpr/crt.h>
#include <winpr/synch.h>
#include <winpr/print.h>
#include <winpr/stream.h>
#include <winpr/winsock.h>

#include <freerdp/log.h>
#include <freerdp/error.h>
#include <freerdp/utils/ringbuffer.h>

#include <openssl/bio.h>
#include <time.h>
#include <errno.h>
#include <fcntl.h>

#ifndef _WIN32
#include <netdb.h>
#include <sys/socket.h>
#endif /* _WIN32 */

#ifdef HAVE_VALGRIND_MEMCHECK_H
#include <valgrind/memcheck.h>
#endif

#include "tpkt.h"
#include "fastpath.h"
#include "transport.h"
#include "rdp.h"

#define TAG FREERDP_TAG("core.transport")

#define BUFFER_SIZE 16384

static void* transport_client_thread(void* arg);

wStream* transport_send_stream_init(rdpTransport* transport, int size)
{
	wStream* s;
	s = StreamPool_Take(transport->ReceivePool, size);
	Stream_EnsureCapacity(s, size);
	Stream_SetPosition(s, 0);
	return s;
}

BOOL transport_attach(rdpTransport* transport, int sockfd)
{
	BIO* socketBio;
	BIO* bufferedBio;

	socketBio = BIO_new(BIO_s_simple_socket());

	if (!socketBio)
		return FALSE;

	BIO_set_fd(socketBio, sockfd, BIO_CLOSE);

	bufferedBio = BIO_new(BIO_s_buffered_socket());

	if (!bufferedBio)
		return FALSE;

	bufferedBio = BIO_push(bufferedBio, socketBio);

	transport->frontBio = bufferedBio;

	return TRUE;
}

BOOL transport_connect_rdp(rdpTransport* transport)
{
	/* RDP encryption */
	return TRUE;
}

BOOL transport_connect_tls(rdpTransport* transport)
{
	int tlsStatus;
	rdpTls* tls = NULL;
	rdpContext* context = transport->context;
	rdpSettings* settings = transport->settings;

	if (transport->GatewayEnabled)
	{
		tls = transport->tls = tls_new(settings);
		transport->layer = TRANSPORT_LAYER_TSG_TLS;
	}
	else
	{
		tls = transport->tls = tls_new(settings);
		transport->layer = TRANSPORT_LAYER_TLS;
	}

	transport->tls = tls;

	tls->hostname = settings->ServerHostname;
	tls->port = settings->ServerPort;

	if (tls->port == 0)
		tls->port = 3389;

	tls->isGatewayTransport = FALSE;
	tlsStatus = tls_connect(tls, transport->frontBio);

	if (tlsStatus < 1)
	{
		if (tlsStatus < 0)
		{
			if (!freerdp_get_last_error(context))
				freerdp_set_last_error(context, FREERDP_ERROR_TLS_CONNECT_FAILED);
		}
		else
		{
			if (!freerdp_get_last_error(context))
				freerdp_set_last_error(context, FREERDP_ERROR_CONNECT_CANCELLED);
		}

		return FALSE;
	}

	transport->frontBio = tls->bio;

	if (!transport->frontBio)
	{
		WLog_ERR(TAG, "unable to prepend a filtering TLS bio");
		return FALSE;
	}

	return TRUE;
}

BOOL transport_connect_nla(rdpTransport* transport)
{
	rdpContext* context = transport->context;
	rdpSettings* settings = context->settings;
	freerdp* instance = context->instance;
	rdpRdp* rdp = context->rdp;

	if (!transport_connect_tls(transport))
		return FALSE;

	if (!settings->Authentication)
		return TRUE;

	rdp->nla = nla_new(instance, transport, settings);

	if (!rdp->nla)
		return FALSE;

	transport_set_nla_mode(transport, TRUE);

	if (settings->AuthenticationServiceClass)
	{
		rdp->nla->ServicePrincipalName =
			nla_make_spn(settings->AuthenticationServiceClass, settings->ServerHostname);

		if (!rdp->nla->ServicePrincipalName)
			return FALSE;
	}

	if (nla_client_begin(rdp->nla) < 0)
	{
		if (!freerdp_get_last_error(context))
			freerdp_set_last_error(context, FREERDP_ERROR_AUTHENTICATION_FAILED);

		transport_set_nla_mode(transport, FALSE);

		return FALSE;
	}

	rdp_client_transition_to_state(rdp, CONNECTION_STATE_NLA);

	return TRUE;
}

BOOL transport_connect(rdpTransport* transport, const char* hostname, UINT16 port, int timeout)
{
	int sockfd;
	BOOL status = FALSE;
	rdpSettings* settings = transport->settings;

	transport->async = settings->AsyncTransport;

	if (transport->GatewayEnabled)
	{
		/* New RDP 8 gateway test. */
		transport->rdg = rdg_new(transport);
		if (!transport->rdg)
		{
			return FALSE;
		}
		status = rdg_connect(transport->rdg, hostname, port, timeout);
		if (status)
		{
			transport->frontBio = transport->rdg->frontBio;
			BIO_set_nonblock(transport->frontBio, 0);
			transport->layer = TRANSPORT_LAYER_TSG;

			status = TRUE;
		}
		else
		{
			if (transport->rdg->state != RDG_CLIENT_STATE_NOT_FOUND)
			{
				return FALSE;
			}

			transport->tsg = tsg_new(transport);

			if (!transport->tsg)
				return FALSE;

			if (!tsg_connect(transport->tsg, hostname, port, timeout))
				return FALSE;

			transport->frontBio = transport->tsg->bio;
			transport->layer = TRANSPORT_LAYER_TSG;

			status = TRUE;
		}
	}
	else
	{
		sockfd = freerdp_tcp_connect(settings, hostname, port, timeout);

		if (sockfd < 1)
			return FALSE;

		transport_attach(transport, sockfd);

		status = TRUE;
	}

	if (status)
	{
		if (transport->async)
		{
			transport->stopEvent = CreateEvent(NULL, TRUE, FALSE, NULL);
			transport->thread = CreateThread(NULL, 0,
				(LPTHREAD_START_ROUTINE) transport_client_thread, transport, 0, NULL);
		}
	}

	return status;
}

BOOL transport_accept_rdp(rdpTransport* transport)
{
	/* RDP encryption */
	return TRUE;
}

BOOL transport_accept_tls(rdpTransport* transport)
{
	rdpSettings* settings = transport->settings;

	if (!transport->tls)
		transport->tls = tls_new(transport->settings);

	transport->layer = TRANSPORT_LAYER_TLS;

	if (!tls_accept(transport->tls, transport->frontBio, settings->CertificateFile, settings->PrivateKeyFile))
		return FALSE;

	transport->frontBio = transport->tls->bio;

	return TRUE;
}

BOOL transport_accept_nla(rdpTransport* transport)
{
	rdpSettings* settings = transport->settings;
	freerdp* instance = (freerdp*) settings->instance;

	if (!transport->tls)
		transport->tls = tls_new(transport->settings);

	transport->layer = TRANSPORT_LAYER_TLS;

	if (!tls_accept(transport->tls, transport->frontBio, settings->CertificateFile, settings->PrivateKeyFile))
		return FALSE;

	transport->frontBio = transport->tls->bio;

	/* Network Level Authentication */

	if (!settings->Authentication)
		return TRUE;

	if (!transport->nla)
	{
		transport->nla = nla_new(instance, transport, settings);
		transport_set_nla_mode(transport, TRUE);
	}

	if (nla_authenticate(transport->nla) < 0)
	{
		WLog_ERR(TAG, "client authentication failure");
		transport_set_nla_mode(transport, FALSE);
		nla_free(transport->nla);
		transport->nla = NULL;
		tls_set_alert_code(transport->tls, TLS_ALERT_LEVEL_FATAL, TLS_ALERT_DESCRIPTION_ACCESS_DENIED);
		return FALSE;
	}

	/* don't free nla module yet, we need to copy the credentials from it first */
	transport_set_nla_mode(transport, FALSE);
	return TRUE;
}

int transport_read_layer(rdpTransport* transport, BYTE* data, int bytes)
{
	int read = 0;
	int status = -1;

	if (!transport->frontBio)
	{
		transport->layer = TRANSPORT_LAYER_CLOSED;
		return -1;
	}

	while (read < bytes)
	{
		status = BIO_read(transport->frontBio, data + read, bytes - read);

		if (status <= 0)
		{
			if (!transport->frontBio || !BIO_should_retry(transport->frontBio))
			{
				/* something unexpected happened, let's close */
				transport->layer = TRANSPORT_LAYER_CLOSED;
				return -1;
			}

			/* non blocking will survive a partial read */
			if (!transport->blocking)
				return read;

			/* blocking means that we can't continue until we have read the number of requested bytes */
			if (BIO_wait_read(transport->frontBio, 100) < 0)
			{
				WLog_ERR(TAG, "error when selecting for read");
				return -1;
			}

			continue;
		}

#ifdef HAVE_VALGRIND_MEMCHECK_H
		VALGRIND_MAKE_MEM_DEFINED(data + read, bytes - read);
#endif
		read += status;
	}

	return read;
}


/**
 * @brief Tries to read toRead bytes from the specified transport
 *
 * Try to read toRead bytes from the transport to the stream.
 * In case it was not possible to read toRead bytes 0 is returned. The stream is always advanced by the
 * number of bytes read.
 *
 * The function assumes that the stream has enough capacity to hold the data.
 *
 * @param[in] transport rdpTransport
 * @param[in] s wStream
 * @param[in] toRead number of bytes to read
 * @return < 0 on error; 0 if not enough data is available (non blocking mode); 1 toRead bytes read
 */
static int transport_read_layer_bytes(rdpTransport* transport, wStream* s, unsigned int toRead)
{
	int status;
	status = transport_read_layer(transport, Stream_Pointer(s), toRead);

	if (status <= 0)
		return status;

	Stream_Seek(s, status);
	return status == toRead ? 1 : 0;
}

/**
 * @brief Try to read a complete PDU (NLA, fast-path or tpkt) from the underlying transport.
 *
 * If possible a complete PDU is read, in case of non blocking transport this might not succeed.
 * Except in case of an error the passed stream will point to the last byte read (correct
 * position). When the pdu read is completed the stream is sealed and the pointer set to 0
 *
 * @param[in] transport rdpTransport
 * @param[in] s wStream
 * @return < 0 on error; 0 if not enough data is available (non blocking mode); > 0 number of
 * bytes of the *complete* pdu read
 */
int transport_read_pdu(rdpTransport* transport, wStream* s)
{
	int status;
	int position;
	int pduLength;
	BYTE* header;

	position = 0;
	pduLength = 0;

	if (!transport)
		return -1;

	if (!s)
		return -1;

	position = Stream_GetPosition(s);
	/* Make sure there is enough space for the longest header within the stream */
	Stream_EnsureCapacity(s, 4);

	/* Make sure at least two bytes are read for further processing */
	if (position < 2 && (status = transport_read_layer_bytes(transport, s, 2 - position)) != 1)
	{
		/* No data available at the moment */
		return status;
	}

	/* update position value for further checks */
	position = Stream_GetPosition(s);
	header = Stream_Buffer(s);

	if (transport->NlaMode)
	{
		/*
		 * In case NlaMode is set TSRequest package(s) are expected
		 * 0x30 = DER encoded data with these bits set:
		 * bit 6 P/C constructed
		 * bit 5 tag number - sequence
		 */
		if (header[0] == 0x30)
		{
			/* TSRequest (NLA) */
			if (header[1] & 0x80)
			{
				if ((header[1] & ~(0x80)) == 1)
				{
					/* check for header bytes already was readed in previous calls */
					if (position < 3
							&& (status = transport_read_layer_bytes(transport, s, 3 - position)) != 1)
						return status;

					pduLength = header[2];
					pduLength += 3;
				}
				else if ((header[1] & ~(0x80)) == 2)
				{
					/* check for header bytes already was readed in previous calls */
					if (position < 4
							&& (status = transport_read_layer_bytes(transport, s, 4 - position)) != 1)
						return status;

					pduLength = (header[2] << 8) | header[3];
					pduLength += 4;
				}
				else
				{
					WLog_ERR(TAG, "Error reading TSRequest!");
					return -1;
				}
			}
			else
			{
				pduLength = header[1];
				pduLength += 2;
			}
		}
	}
	else
	{
		if (header[0] == 0x03)
		{
			/* TPKT header */
			/* check for header bytes already was readed in previous calls */
			if (position < 4
					&& (status = transport_read_layer_bytes(transport, s, 4 - position)) != 1)
				return status;

			pduLength = (header[2] << 8) | header[3];

			/* min and max values according to ITU-T Rec. T.123 (01/2007) section 8 */
			if (pduLength < 7 || pduLength > 0xFFFF)
			{
				WLog_ERR(TAG, "tpkt - invalid pduLength: %d", pduLength);
				return -1;
			}
		}
		else
		{
			/* Fast-Path Header */
			if (header[1] & 0x80)
			{
				/* check for header bytes already was readed in previous calls */
				if (position < 3
						&& (status = transport_read_layer_bytes(transport, s, 3 - position)) != 1)
					return status;

				pduLength = ((header[1] & 0x7F) << 8) | header[2];
			}
			else
				pduLength = header[1];

			/*
			 * fast-path has 7 bits for length so the maximum size, including headers is 0x8000
			 * The theoretical minimum fast-path PDU consists only of two header bytes plus one
			 * byte for data (e.g. fast-path input synchronize pdu)
			 */
			if (pduLength < 3 || pduLength > 0x8000)
			{
				WLog_ERR(TAG, "fast path - invalid pduLength: %d", pduLength);
				return -1;
			}
		}
	}

	Stream_EnsureCapacity(s, Stream_GetPosition(s) + pduLength);
	status = transport_read_layer_bytes(transport, s, pduLength - Stream_GetPosition(s));

	if (status != 1)
		return status;

	if (Stream_GetPosition(s) >= pduLength)
		WLog_Packet(WLog_Get(TAG), WLOG_TRACE, Stream_Buffer(s), pduLength, WLOG_PACKET_INBOUND);

	Stream_SealLength(s);
	Stream_SetPosition(s, 0);
	return Stream_Length(s);
}

int transport_write(rdpTransport* transport, wStream* s)
{
	int length;
	int status = -1;

	EnterCriticalSection(&(transport->WriteLock));

	length = Stream_GetPosition(s);
	Stream_SetPosition(s, 0);

	if (length > 0)
	{
		WLog_Packet(WLog_Get(TAG), WLOG_TRACE, Stream_Buffer(s), length, WLOG_PACKET_OUTBOUND);
	}

	while (length > 0)
	{
		status = BIO_write(transport->frontBio, Stream_Pointer(s), length);

		if (status <= 0)
		{
			/* the buffered BIO that is at the end of the chain always says OK for writing,
			 * so a retry means that for any reason we need to read. The most probable
			 * is a SSL or TSG BIO in the chain.
			 */
			if (!BIO_should_retry(transport->frontBio))
				return status;

			/* non-blocking can live with blocked IOs */
			if (!transport->blocking)
				return status;

			if (BIO_wait_write(transport->frontBio, 100) < 0)
			{
				WLog_ERR(TAG, "error when selecting for write");
				return -1;
			}

			continue;
		}

		if (transport->blocking || transport->settings->WaitForOutputBufferFlush)
		{
			while (BIO_write_blocked(transport->frontBio))
			{
				if (BIO_wait_write(transport->frontBio, 100) < 0)
				{
					WLog_ERR(TAG, "error when selecting for write");
					return -1;
				}

				if (BIO_flush(transport->frontBio) < 1)
				{
					WLog_ERR(TAG, "error when flushing outputBuffer");
					return -1;
				}
			}
		}

		length -= status;
		Stream_Seek(s, status);
	}

	if (status < 0)
	{
		/* A write error indicates that the peer has dropped the connection */
		transport->layer = TRANSPORT_LAYER_CLOSED;
	}

	if (s->pool)
		Stream_Release(s);

	LeaveCriticalSection(&(transport->WriteLock));
	return status;
}

UINT32 transport_get_event_handles(rdpTransport* transport, HANDLE* events)
{
	UINT32 nCount = 0;

	if (!transport->GatewayEnabled)
	{
		if (events)
			BIO_get_event(transport->frontBio, &events[nCount]);
		nCount++;
	}
	else
	{
		if (transport->rdg)
		{
			nCount += rdg_get_event_handles(transport->rdg, events);
		}
		else if (transport->tsg)
		{
			nCount += tsg_get_event_handles(transport->tsg, events);
		}
	}

	return nCount;
}

void transport_get_fds(rdpTransport* transport, void** rfds, int* rcount)
{
	UINT32 index;
	UINT32 nCount;
	HANDLE events[64];

	nCount = transport_get_event_handles(transport, events);

	for (index = 0; index < nCount; index++)
	{
		rfds[*rcount] = GetEventWaitObject(events[index]);
		(*rcount)++;
	}
}

BOOL transport_is_write_blocked(rdpTransport* transport)
{
	return BIO_write_blocked(transport->frontBio);
}

int transport_drain_output_buffer(rdpTransport* transport)
{
	BOOL status = FALSE;

	if (BIO_write_blocked(transport->frontBio))
	{
		if (BIO_flush(transport->frontBio) < 1)
			return -1;

		status |= BIO_write_blocked(transport->frontBio);
	}

	return status;
}

int transport_check_fds(rdpTransport* transport)
{
	int status;
	int recv_status;
	wStream* received;

	if (!transport)
		return -1;

	/**
	 * Loop through and read all available PDUs.  Since multiple
	 * PDUs can exist, it's important to deliver them all before
	 * returning.  Otherwise we run the risk of having a thread
	 * wait for a socket to get signaled that data is available
	 * (which may never happen).
	 */
	for (;;)
	{
		/**
		 * Note: transport_read_pdu tries to read one PDU from
		 * the transport layer.
		 * The ReceiveBuffer might have a position > 0 in case of a non blocking
		 * transport. If transport_read_pdu returns 0 the pdu couldn't be read at
		 * this point.
		 * Note that transport->ReceiveBuffer is replaced after each iteration
		 * of this loop with a fresh stream instance from a pool.
		 */
		if ((status = transport_read_pdu(transport, transport->ReceiveBuffer)) <= 0)
		{
			return status;
		}

		received = transport->ReceiveBuffer;
		transport->ReceiveBuffer = StreamPool_Take(transport->ReceivePool, 0);
		/**
		 * status:
		 * 	-1: error
		 * 	 0: success
		 * 	 1: redirection
		 */
		recv_status = transport->ReceiveCallback(transport, received, transport->ReceiveExtra);
		Stream_Release(received);

		/* session redirection or activation */
		if (recv_status == 1 || recv_status == 2)
		{
			return recv_status;
		}

		if (recv_status < 0)
			return -1;
	}

	return 0;
}

BOOL transport_set_blocking_mode(rdpTransport* transport, BOOL blocking)
{
	transport->blocking = blocking;

	if (!BIO_set_nonblock(transport->frontBio, blocking ? FALSE : TRUE))
		return FALSE;

	return TRUE;
}

void transport_set_gateway_enabled(rdpTransport* transport, BOOL GatewayEnabled)
{
	transport->GatewayEnabled = GatewayEnabled;
}

void transport_set_nla_mode(rdpTransport* transport, BOOL NlaMode)
{
	transport->NlaMode = NlaMode;
}

void transport_stop(rdpTransport* transport)
{
	if (transport->async)
	{
		if (transport->stopEvent)
		{
			SetEvent(transport->stopEvent);
			WaitForSingleObject(transport->thread, INFINITE);
			CloseHandle(transport->thread);
			CloseHandle(transport->stopEvent);
			transport->thread = NULL;
			transport->stopEvent = NULL;
		}
	}
}

BOOL transport_disconnect(rdpTransport* transport)
{
	BOOL status = TRUE;

	if (!transport)
		return FALSE;

	transport_stop(transport);

	if (transport->tls)
	{
		tls_free(transport->tls);
		transport->tls = NULL;
	}
	else
	{
		if (transport->frontBio)
			BIO_free(transport->frontBio);
	}

	if (transport->tsg)
	{
		tsg_free(transport->tsg);
		transport->tsg = NULL;
	}

	if (transport->rdg)
	{
		rdg_free(transport->rdg);
		transport->rdg = NULL;
	}

	transport->frontBio = NULL;

	transport->layer = TRANSPORT_LAYER_TCP;

	return status;
}

static void* transport_client_thread(void* arg)
{
	DWORD status;
	DWORD nCount;
	HANDLE handles[64];
	rdpTransport* transport = (rdpTransport*) arg;
	rdpContext* context = transport->context;
	rdpRdp* rdp = context->rdp;

	WLog_DBG(TAG, "Starting transport thread");
	
	nCount = 0;
	handles[nCount++] = transport->stopEvent;
	handles[nCount++] = transport->connectedEvent;

	status = WaitForMultipleObjects(nCount, handles, FALSE, INFINITE);

	if (WaitForSingleObject(transport->stopEvent, 0) == WAIT_OBJECT_0)
	{
		WLog_DBG(TAG, "Terminating transport thread");
		ExitThread(0);
		return NULL;
	}

	WLog_DBG(TAG, "Asynchronous transport activated");

	while (1)
	{
		nCount = 0;
		handles[nCount++] = transport->stopEvent;
		nCount += freerdp_get_event_handles(context, &handles[nCount]);
		
		status = WaitForMultipleObjects(nCount, handles, FALSE, INFINITE);

		if (transport->layer == TRANSPORT_LAYER_CLOSED)
		{
			rdp_set_error_info(rdp, ERRINFO_PEER_DISCONNECTED);
			break;
		}

		if (WaitForSingleObject(transport->stopEvent, 0) == WAIT_OBJECT_0)
				break;

		if (WaitForMultipleObjects(nCount - 1, &handles[1], FALSE, 0) != WAIT_TIMEOUT)
		{
			if (!freerdp_check_event_handles(context))
			{
				rdp_set_error_info(rdp, ERRINFO_PEER_DISCONNECTED);
				break;
			}
		}
	}

	WLog_DBG(TAG, "Terminating transport thread");
	ExitThread(0);
	return NULL;
}

rdpTransport* transport_new(rdpContext* context)
{
	rdpTransport* transport;
	
	transport = (rdpTransport*) calloc(1, sizeof(rdpTransport));

	if (!transport)
		return NULL;

	transport->context = context;
	transport->settings = context->settings;

	transport->ReceivePool = StreamPool_New(TRUE, BUFFER_SIZE);

	if (!transport->ReceivePool)
		goto out_free_transport;

	/* receive buffer for non-blocking read. */
	transport->ReceiveBuffer = StreamPool_Take(transport->ReceivePool, 0);

	if (!transport->ReceiveBuffer)
		goto out_free_receivepool;

	transport->connectedEvent = CreateEvent(NULL, TRUE, FALSE, NULL);

	if (!transport->connectedEvent || transport->connectedEvent == INVALID_HANDLE_VALUE)
		goto out_free_receivebuffer;

	transport->blocking = TRUE;
	transport->GatewayEnabled = FALSE;
	transport->layer = TRANSPORT_LAYER_TCP;

	if (!InitializeCriticalSectionAndSpinCount(&(transport->ReadLock), 4000))
		goto out_free_connectedEvent;

	if (!InitializeCriticalSectionAndSpinCount(&(transport->WriteLock), 4000))
		goto out_free_readlock;

	return transport;
out_free_readlock:
	DeleteCriticalSection(&(transport->ReadLock));
out_free_connectedEvent:
	CloseHandle(transport->connectedEvent);
out_free_receivebuffer:
	StreamPool_Return(transport->ReceivePool, transport->ReceiveBuffer);
out_free_receivepool:
	StreamPool_Free(transport->ReceivePool);
out_free_transport:
	free(transport);
	return NULL;
}

void transport_free(rdpTransport* transport)
{
	if (!transport)
		return;

	transport_disconnect(transport);

	if (transport->ReceiveBuffer)
		Stream_Release(transport->ReceiveBuffer);

	StreamPool_Free(transport->ReceivePool);
	CloseHandle(transport->connectedEvent);
	DeleteCriticalSection(&(transport->ReadLock));
	DeleteCriticalSection(&(transport->WriteLock));

	free(transport);
}