[DCCP]: Initial implementation

Development to this point was done on a subversion repository at:

http://oops.ghostprotocols.net:81/cgi-bin/viewcvs.cgi/dccp-2.6/

This repository will be kept at this site for the foreseable future,
so that interested parties can see the history of this code,
attributions, etc.

If I ever decide to take this offline I'll provide the full history at
some other suitable place.

Signed-off-by: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Signed-off-by: David S. Miller <davem@davemloft.net>

include/linux/dccp.h

@@ -0,0 +1,432 @@
+#ifndef _LINUX_DCCP_H
+#define _LINUX_DCCP_H
+
+#include <linux/in.h>
+#include <linux/list.h>
+#include <linux/types.h>
+#include <linux/uio.h>
+#include <linux/workqueue.h>
+
+#include <net/inet_connection_sock.h>
+#include <net/sock.h>
+#include <net/tcp_states.h>
+#include <net/tcp.h>
+
+/* FIXME: this is utterly wrong */
+struct sockaddr_dccp {
+	struct sockaddr_in	in;
+	unsigned int		service;
+};
+
+enum dccp_state {
+	DCCP_OPEN	= TCP_ESTABLISHED,
+	DCCP_REQUESTING	= TCP_SYN_SENT,
+	DCCP_PARTOPEN	= TCP_FIN_WAIT1, /* FIXME:
+					    This mapping is horrible, but TCP has
+					    no matching state for DCCP_PARTOPEN,
+					    as TCP_SYN_RECV is already used by
+					    DCCP_RESPOND, why don't stop using TCP
+					    mapping of states? OK, now we don't use
+					    sk_stream_sendmsg anymore, so doesn't
+					    seem to exist any reason for us to
+					    do the TCP mapping here */
+	DCCP_LISTEN	= TCP_LISTEN,
+	DCCP_RESPOND	= TCP_SYN_RECV,
+	DCCP_CLOSING	= TCP_CLOSING,
+	DCCP_TIME_WAIT	= TCP_TIME_WAIT,
+	DCCP_CLOSED	= TCP_CLOSE,
+	DCCP_MAX_STATES = TCP_MAX_STATES,
+};
+
+#define DCCP_STATE_MASK 0xf
+#define DCCP_ACTION_FIN (1<<7)
+
+enum {
+	DCCPF_OPEN	 = TCPF_ESTABLISHED,
+	DCCPF_REQUESTING = TCPF_SYN_SENT,
+	DCCPF_PARTOPEN	 = TCPF_FIN_WAIT1,
+	DCCPF_LISTEN	 = TCPF_LISTEN,
+	DCCPF_RESPOND	 = TCPF_SYN_RECV,
+	DCCPF_CLOSING	 = TCPF_CLOSING,
+	DCCPF_TIME_WAIT	 = TCPF_TIME_WAIT,
+	DCCPF_CLOSED	 = TCPF_CLOSE,
+};
+
+/**
+ * struct dccp_hdr - generic part of DCCP packet header
+ *
+ * @dccph_sport - Relevant port on the endpoint that sent this packet
+ * @dccph_dport - Relevant port on the other endpoint
+ * @dccph_doff - Data Offset from the start of the DCCP header, in 32-bit words
+ * @dccph_ccval - Used by the HC-Sender CCID
+ * @dccph_cscov - Parts of the packet that are covered by the Checksum field
+ * @dccph_checksum - Internet checksum, depends on dccph_cscov
+ * @dccph_x - 0 = 24 bit sequence number, 1 = 48
+ * @dccph_type - packet type, see DCCP_PKT_ prefixed macros
+ * @dccph_seq - sequence number high or low order 24 bits, depends on dccph_x
+ */
+struct dccp_hdr {
+	__u16	dccph_sport,
+		dccph_dport;
+	__u8	dccph_doff;
+#if defined(__LITTLE_ENDIAN_BITFIELD)
+	__u8	dccph_cscov:4,
+		dccph_ccval:4;
+#elif defined(__BIG_ENDIAN_BITFIELD)
+	__u8	dccph_ccval:4,
+		dccph_cscov:4;
+#else
+#error  "Adjust your <asm/byteorder.h> defines"
+#endif
+	__u16	dccph_checksum;
+#if defined(__LITTLE_ENDIAN_BITFIELD)
+	__u32	dccph_x:1,
+		dccph_type:4,
+		dccph_reserved:3,
+		dccph_seq:24;
+#elif defined(__BIG_ENDIAN_BITFIELD)
+	__u32	dccph_reserved:3,
+		dccph_type:4,
+		dccph_x:1,
+		dccph_seq:24;
+#else
+#error  "Adjust your <asm/byteorder.h> defines"
+#endif
+};
+
+static inline struct dccp_hdr *dccp_hdr(const struct sk_buff *skb)
+{
+	return (struct dccp_hdr *)skb->h.raw;
+}
+
+/**
+ * struct dccp_hdr_ext - the low bits of a 48 bit seq packet
+ *
+ * @dccph_seq_low - low 24 bits of a 48 bit seq packet
+ */
+struct dccp_hdr_ext {
+	__u32	dccph_seq_low;
+};
+
+static inline struct dccp_hdr_ext *dccp_hdrx(const struct sk_buff *skb)
+{
+	return (struct dccp_hdr_ext *)(skb->h.raw + sizeof(struct dccp_hdr));
+}
+
+static inline unsigned int dccp_basic_hdr_len(const struct sk_buff *skb)
+{
+	const struct dccp_hdr *dh = dccp_hdr(skb);
+	return sizeof(*dh) + (dh->dccph_x ? sizeof(struct dccp_hdr_ext) : 0);
+}
+
+static inline __u64 dccp_hdr_seq(const struct sk_buff *skb)
+{
+	const struct dccp_hdr *dh = dccp_hdr(skb);
+#if defined(__LITTLE_ENDIAN_BITFIELD)
+	__u64 seq_nr = ntohl(dh->dccph_seq << 8);
+#elif defined(__BIG_ENDIAN_BITFIELD)
+	__u64 seq_nr = ntohl(dh->dccph_seq);
+#else
+#error  "Adjust your <asm/byteorder.h> defines"
+#endif
+
+	if (dh->dccph_x != 0)
+		seq_nr = (seq_nr << 32) + ntohl(dccp_hdrx(skb)->dccph_seq_low);
+
+	return seq_nr;
+}
+
+/**
+ * struct dccp_hdr_request - Conection initiation request header
+ *
+ * @dccph_req_service - Service to which the client app wants to connect
+ * @dccph_req_options - list of options (must be a multiple of 32 bits
+ */
+struct dccp_hdr_request {
+	__u32	dccph_req_service;
+};
+
+static inline struct dccp_hdr_request *dccp_hdr_request(struct sk_buff *skb)
+{
+	return (struct dccp_hdr_request *)(skb->h.raw + dccp_basic_hdr_len(skb));
+}
+
+/**
+ * struct dccp_hdr_ack_bits - acknowledgment bits common to most packets
+ *
+ * @dccph_resp_ack_nr_high - 48 bit ack number high order bits, contains GSR
+ * @dccph_resp_ack_nr_low - 48 bit ack number low order bits, contains GSR
+ */
+struct dccp_hdr_ack_bits {
+	__u32	dccph_reserved1:8,
+		dccph_ack_nr_high:24;
+	__u32	dccph_ack_nr_low;
+};
+
+static inline struct dccp_hdr_ack_bits *dccp_hdr_ack_bits(const struct sk_buff *skb)
+{
+	return (struct dccp_hdr_ack_bits *)(skb->h.raw + dccp_basic_hdr_len(skb));
+}
+
+static inline u64 dccp_hdr_ack_seq(const struct sk_buff *skb)
+{
+	const struct dccp_hdr_ack_bits *dhack = dccp_hdr_ack_bits(skb);
+#if defined(__LITTLE_ENDIAN_BITFIELD)
+	return (((u64)ntohl(dhack->dccph_ack_nr_high << 8)) << 32) + ntohl(dhack->dccph_ack_nr_low);
+#elif defined(__BIG_ENDIAN_BITFIELD)
+	return (((u64)ntohl(dhack->dccph_ack_nr_high)) << 32) + ntohl(dhack->dccph_ack_nr_low);
+#else
+#error  "Adjust your <asm/byteorder.h> defines"
+#endif
+}
+
+/**
+ * struct dccp_hdr_response - Conection initiation response header
+ *
+ * @dccph_resp_ack_nr_high - 48 bit ack number high order bits, contains GSR
+ * @dccph_resp_ack_nr_low - 48 bit ack number low order bits, contains GSR
+ * @dccph_resp_service - Echoes the Service Code on a received DCCP-Request
+ * @dccph_resp_options - list of options (must be a multiple of 32 bits
+ */
+struct dccp_hdr_response {
+	struct dccp_hdr_ack_bits	dccph_resp_ack;
+	__u32				dccph_resp_service;
+};
+
+static inline struct dccp_hdr_response *dccp_hdr_response(struct sk_buff *skb)
+{
+	return (struct dccp_hdr_response *)(skb->h.raw + dccp_basic_hdr_len(skb));
+}
+
+/**
+ * struct dccp_hdr_reset - Unconditionally shut down a connection
+ *
+ * @dccph_reset_service - Echoes the Service Code on a received DCCP-Request
+ * @dccph_reset_options - list of options (must be a multiple of 32 bits
+ */
+struct dccp_hdr_reset {
+	struct dccp_hdr_ack_bits	dccph_reset_ack;
+	__u8				dccph_reset_code,
+					dccph_reset_data[3];
+};
+
+static inline struct dccp_hdr_reset *dccp_hdr_reset(struct sk_buff *skb)
+{
+	return (struct dccp_hdr_reset *)(skb->h.raw + dccp_basic_hdr_len(skb));
+}
+
+enum dccp_pkt_type {
+	DCCP_PKT_REQUEST = 0,
+	DCCP_PKT_RESPONSE,
+	DCCP_PKT_DATA,
+	DCCP_PKT_ACK,
+	DCCP_PKT_DATAACK,
+	DCCP_PKT_CLOSEREQ,
+	DCCP_PKT_CLOSE,
+	DCCP_PKT_RESET,
+	DCCP_PKT_SYNC,
+	DCCP_PKT_SYNCACK,
+	DCCP_PKT_INVALID,
+};
+
+#define DCCP_NR_PKT_TYPES DCCP_PKT_INVALID
+
+static inline unsigned int dccp_packet_hdr_len(const __u8 type)
+{
+	if (type == DCCP_PKT_DATA)
+		return 0;
+	if (type == DCCP_PKT_DATAACK	||
+	    type == DCCP_PKT_ACK	||
+	    type == DCCP_PKT_SYNC	||
+	    type == DCCP_PKT_SYNCACK	||
+	    type == DCCP_PKT_CLOSE	||
+	    type == DCCP_PKT_CLOSEREQ)
+		return sizeof(struct dccp_hdr_ack_bits);
+	if (type == DCCP_PKT_REQUEST)
+		return sizeof(struct dccp_hdr_request);
+	if (type == DCCP_PKT_RESPONSE)
+		return sizeof(struct dccp_hdr_response);
+	return sizeof(struct dccp_hdr_reset);
+}
+
+static inline unsigned int dccp_hdr_len(const struct sk_buff *skb)
+{
+	return dccp_basic_hdr_len(skb) +
+	       dccp_packet_hdr_len(dccp_hdr(skb)->dccph_type);
+}
+
+enum dccp_reset_codes {
+	DCCP_RESET_CODE_UNSPECIFIED = 0,
+	DCCP_RESET_CODE_CLOSED,
+	DCCP_RESET_CODE_ABORTED,
+	DCCP_RESET_CODE_NO_CONNECTION,
+	DCCP_RESET_CODE_PACKET_ERROR,
+	DCCP_RESET_CODE_OPTION_ERROR,
+	DCCP_RESET_CODE_MANDATORY_ERROR,
+	DCCP_RESET_CODE_CONNECTION_REFUSED,
+	DCCP_RESET_CODE_BAD_SERVICE_CODE,
+	DCCP_RESET_CODE_TOO_BUSY,
+	DCCP_RESET_CODE_BAD_INIT_COOKIE,
+	DCCP_RESET_CODE_AGGRESSION_PENALTY,
+};
+
+/* DCCP options */
+enum {
+	DCCPO_PADDING = 0,
+	DCCPO_MANDATORY = 1,
+	DCCPO_MIN_RESERVED = 3,
+	DCCPO_MAX_RESERVED = 31,
+	DCCPO_NDP_COUNT = 37,
+	DCCPO_ACK_VECTOR_0 = 38,
+	DCCPO_ACK_VECTOR_1 = 39,
+	DCCPO_TIMESTAMP = 41,
+	DCCPO_TIMESTAMP_ECHO = 42,
+	DCCPO_ELAPSED_TIME = 43,
+	DCCPO_MAX = 45,
+	DCCPO_MIN_CCID_SPECIFIC = 128,
+	DCCPO_MAX_CCID_SPECIFIC = 255,
+};
+
+/* DCCP features */
+enum {
+	DCCPF_RESERVED = 0,
+	DCCPF_SEQUENCE_WINDOW = 3,
+	DCCPF_SEND_ACK_VECTOR = 6,
+	DCCPF_SEND_NDP_COUNT = 7,
+	/* 10-127 reserved */
+	DCCPF_MIN_CCID_SPECIFIC = 128,
+	DCCPF_MAX_CCID_SPECIFIC = 255,
+};
+
+/* initial values for each feature */
+#define DCCPF_INITIAL_SEQUENCE_WINDOW		100
+/* FIXME: for now we're using CCID 3 (TFRC) */
+#define DCCPF_INITIAL_CCID			3
+#define DCCPF_INITIAL_SEND_ACK_VECTOR		0
+/* FIXME: for now we're default to 1 but it should really be 0 */
+#define DCCPF_INITIAL_SEND_NDP_COUNT		1
+
+#define DCCP_NDP_LIMIT 0xFFFFFF
+
+/**
+  * struct dccp_options - option values for a DCCP connection
+  *	@dccpo_sequence_window - Sequence Window Feature (section 7.5.2)
+  *	@dccpo_ccid - Congestion Control Id (CCID) (section 10)
+  *	@dccpo_send_ack_vector - Send Ack Vector Feature (section 11.5)
+  *	@dccpo_send_ndp_count - Send NDP Count Feature (7.7.2)
+  */
+struct dccp_options {
+	__u64	dccpo_sequence_window;
+	__u8	dccpo_ccid;
+	__u8	dccpo_send_ack_vector;
+	__u8	dccpo_send_ndp_count;
+};
+
+extern void __dccp_options_init(struct dccp_options *dccpo);
+extern void dccp_options_init(struct dccp_options *dccpo);
+extern int dccp_parse_options(struct sock *sk, struct sk_buff *skb);
+
+struct dccp_request_sock {
+	struct inet_request_sock dreq_inet_rsk;
+	__u64			 dreq_iss;
+	__u64			 dreq_isr;
+	__u32			 dreq_service;
+};
+
+static inline struct dccp_request_sock *dccp_rsk(const struct request_sock *req)
+{
+	return (struct dccp_request_sock *)req;
+}
+
+/* Read about the ECN nonce to see why it is 253 */
+#define DCCP_MAX_ACK_VECTOR_LEN 253
+
+struct dccp_options_received {
+	u32	dccpor_ndp:24,
+		dccpor_ack_vector_len:8;
+	u32	dccpor_ack_vector_idx:10;
+	/* 22 bits hole, try to pack */
+	u32	dccpor_timestamp;
+	u32	dccpor_timestamp_echo;
+	u32	dccpor_elapsed_time;
+};
+
+struct ccid;
+
+enum dccp_role {
+	DCCP_ROLE_UNDEFINED,
+	DCCP_ROLE_LISTEN,
+	DCCP_ROLE_CLIENT,
+	DCCP_ROLE_SERVER,
+};
+
+/**
+ * struct dccp_sock - DCCP socket state
+ *
+ * @dccps_swl - sequence number window low
+ * @dccps_swh - sequence number window high
+ * @dccps_awl - acknowledgement number window low
+ * @dccps_awh - acknowledgement number window high
+ * @dccps_iss - initial sequence number sent
+ * @dccps_isr - initial sequence number received
+ * @dccps_osr - first OPEN sequence number received
+ * @dccps_gss - greatest sequence number sent
+ * @dccps_gsr - greatest valid sequence number received
+ * @dccps_gar - greatest valid ack number received on a non-Sync; initialized to %dccps_iss
+ * @dccps_timestamp_time - time of latest TIMESTAMP option
+ * @dccps_timestamp_echo - latest timestamp received on a TIMESTAMP option
+ * @dccps_ext_header_len - network protocol overhead (IP/IPv6 options)
+ * @dccps_pmtu_cookie - Last pmtu seen by socket
+ * @dccps_avg_packet_size - FIXME: has to be set by the app thru some setsockopt or ioctl, CCID3 uses it
+ * @dccps_role - Role of this sock, one of %dccp_role
+ * @dccps_ndp_count - number of Non Data Packets since last data packet
+ * @dccps_hc_rx_ackpkts - receiver half connection acked packets
+ */
+struct dccp_sock {
+	/* inet_connection_sock has to be the first member of dccp_sock */
+	struct inet_connection_sock	dccps_inet_connection;
+	__u64				dccps_swl;
+	__u64				dccps_swh;
+	__u64				dccps_awl;
+	__u64				dccps_awh;
+	__u64				dccps_iss;
+	__u64				dccps_isr;
+	__u64				dccps_osr;
+	__u64				dccps_gss;
+	__u64				dccps_gsr;
+	__u64				dccps_gar;
+	unsigned long			dccps_service;
+	unsigned long			dccps_timestamp_time;
+	__u32				dccps_timestamp_echo;
+	__u32				dccps_avg_packet_size;
+	unsigned long			dccps_ndp_count;
+	__u16				dccps_ext_header_len;
+	__u32				dccps_pmtu_cookie;
+	__u32				dccps_mss_cache;
+	struct dccp_options		dccps_options;
+	struct dccp_ackpkts		*dccps_hc_rx_ackpkts;
+	void				*dccps_hc_rx_ccid_private;
+	void				*dccps_hc_tx_ccid_private;
+	struct ccid			*dccps_hc_rx_ccid;
+	struct ccid			*dccps_hc_tx_ccid;
+	struct dccp_options_received	dccps_options_received;
+	enum dccp_role			dccps_role:2;
+};
+ 
+static inline struct dccp_sock *dccp_sk(const struct sock *sk)
+{
+	return (struct dccp_sock *)sk;
+}
+
+static inline const char *dccp_role(const struct sock *sk)
+{
+	switch (dccp_sk(sk)->dccps_role) {
+	case DCCP_ROLE_UNDEFINED: return "undefined";
+	case DCCP_ROLE_LISTEN:	  return "listen";
+	case DCCP_ROLE_SERVER:	  return "server";
+	case DCCP_ROLE_CLIENT:	  return "client";
+	}
+	return NULL;
+}
+
+#endif /* _LINUX_DCCP_H */

include/linux/in.h

@@ -32,6 +32,7 @@ enum {
   IPPROTO_PUP = 12,		/* PUP protocol				*/
   IPPROTO_UDP = 17,		/* User Datagram Protocol		*/
   IPPROTO_IDP = 22,		/* XNS IDP protocol			*/
+  IPPROTO_DCCP = 33,		/* Datagram Congestion Control Protocol */
   IPPROTO_RSVP = 46,		/* RSVP protocol			*/
   IPPROTO_GRE = 47,		/* Cisco GRE tunnels (rfc 1701,1702)	*/
 

include/linux/net.h

@@ -84,6 +84,7 @@ enum sock_type {
 	SOCK_RAW	= 3,
 	SOCK_RDM	= 4,
 	SOCK_SEQPACKET	= 5,
+	SOCK_DCCP	= 6,
 	SOCK_PACKET	= 10,
 };
 

include/linux/socket.h

@@ -271,6 +271,7 @@ struct ucred {
 #define SOL_IRDA        266
 #define SOL_NETBEUI	267
 #define SOL_LLC		268
+#define SOL_DCCP	269
 
 /* IPX options */
 #define IPX_TYPE	1

net/Kconfig

@@ -147,6 +147,7 @@ source "net/bridge/netfilter/Kconfig"
 
 endif
 
+source "net/dccp/Kconfig"
 source "net/sctp/Kconfig"
 source "net/atm/Kconfig"
 source "net/bridge/Kconfig"

net/Makefile

@@ -42,6 +42,7 @@ obj-$(CONFIG_ATM)		+= atm/
 obj-$(CONFIG_DECNET)		+= decnet/
 obj-$(CONFIG_ECONET)		+= econet/
 obj-$(CONFIG_VLAN_8021Q)	+= 8021q/
+obj-$(CONFIG_IP_DCCP)		+= dccp/
 obj-$(CONFIG_IP_SCTP)		+= sctp/
 
 ifeq ($(CONFIG_NET),y)

net/dccp/Kconfig

@@ -0,0 +1,24 @@
+menu "DCCP Configuration (EXPERIMENTAL)"
+	depends on INET && EXPERIMENTAL
+
+config IP_DCCP
+	tristate "The DCCP Protocol (EXPERIMENTAL)"
+	---help---
+	  Datagram Congestion Control Protocol
+
+	  From draft-ietf-dccp-spec-11 <http://www.icir.org/kohler/dcp/draft-ietf-dccp-spec-11.txt>.
+
+	  The Datagram Congestion Control Protocol (DCCP) is a transport
+	  protocol that implements bidirectional, unicast connections of
+	  congestion-controlled, unreliable datagrams. It should be suitable
+	  for use by applications such as streaming media, Internet telephony,
+	  and on-line games
+
+	  To compile this protocol support as a module, choose M here: the
+	  module will be called dccp.
+
+	  If in doubt, say N.
+
+source "net/dccp/ccids/Kconfig"
+
+endmenu

net/dccp/Makefile

@@ -0,0 +1,5 @@
+obj-$(CONFIG_IP_DCCP) += dccp.o
+
+dccp-y := ccid.o input.o ipv4.o minisocks.o options.o output.o proto.o timer.o
+
+obj-y += ccids/

net/dccp/ccid.c

@@ -0,0 +1,139 @@
+/*
+ *  net/dccp/ccid.c
+ *
+ *  An implementation of the DCCP protocol
+ *  Arnaldo Carvalho de Melo <acme@conectiva.com.br>
+ *
+ *  CCID infrastructure
+ *
+ *	This program is free software; you can redistribute it and/or modify it
+ *	under the terms of the GNU General Public License version 2 as
+ *	published by the Free Software Foundation.
+ */
+
+#include "ccid.h"
+
+static struct ccid *ccids[CCID_MAX];
+#if defined(CONFIG_SMP) || defined(CONFIG_PREEMPT)
+static atomic_t ccids_lockct = ATOMIC_INIT(0);
+static DEFINE_SPINLOCK(ccids_lock);
+
+/*
+ * The strategy is: modifications ccids vector are short, do not sleep and
+ * veeery rare, but read access should be free of any exclusive locks.
+ */
+static void ccids_write_lock(void)
+{
+	spin_lock(&ccids_lock);
+	while (atomic_read(&ccids_lockct) != 0) {
+		spin_unlock(&ccids_lock);
+		yield();
+		spin_lock(&ccids_lock);
+	}
+}
+
+static inline void ccids_write_unlock(void)
+{
+	spin_unlock(&ccids_lock);
+}
+
+static inline void ccids_read_lock(void)
+{
+	atomic_inc(&ccids_lockct);
+	spin_unlock_wait(&ccids_lock);
+}
+
+static inline void ccids_read_unlock(void)
+{
+	atomic_dec(&ccids_lockct);
+}
+
+#else
+#define ccids_write_lock() do { } while(0)
+#define ccids_write_unlock() do { } while(0)
+#define ccids_read_lock() do { } while(0)
+#define ccids_read_unlock() do { } while(0)
+#endif
+
+int ccid_register(struct ccid *ccid)
+{
+	int err;
+
+	if (ccid->ccid_init == NULL)
+		return -1;
+
+	ccids_write_lock();
+	err = -EEXIST;
+	if (ccids[ccid->ccid_id] == NULL) {
+		ccids[ccid->ccid_id] = ccid;
+		err = 0;
+	}
+	ccids_write_unlock();
+	if (err == 0)
+		pr_info("CCID: Registered CCID %d (%s)\n",
+			ccid->ccid_id, ccid->ccid_name);
+	return err;
+}
+
+EXPORT_SYMBOL_GPL(ccid_register);
+
+int ccid_unregister(struct ccid *ccid)
+{
+	ccids_write_lock();
+	ccids[ccid->ccid_id] = NULL;
+	ccids_write_unlock();
+	pr_info("CCID: Unregistered CCID %d (%s)\n",
+		ccid->ccid_id, ccid->ccid_name);
+	return 0;
+}
+
+EXPORT_SYMBOL_GPL(ccid_unregister);
+
+struct ccid *ccid_init(unsigned char id, struct sock *sk)
+{
+	struct ccid *ccid;
+
+#ifdef CONFIG_KMOD
+	if (ccids[id] == NULL)
+		request_module("net-dccp-ccid-%d", id);
+#endif
+	ccids_read_lock();
+
+	ccid = ccids[id];
+	if (ccid == NULL)
+		goto out;
+
+	if (!try_module_get(ccid->ccid_owner))
+		goto out_err;
+
+	if (ccid->ccid_init(sk) != 0)
+		goto out_module_put;
+out:
+	ccids_read_unlock();
+	return ccid;
+out_module_put:
+	module_put(ccid->ccid_owner);
+out_err:
+	ccid = NULL;
+	goto out;
+}
+
+EXPORT_SYMBOL_GPL(ccid_init);
+
+void ccid_exit(struct ccid *ccid, struct sock *sk)
+{
+	if (ccid == NULL)
+		return;
+
+	ccids_read_lock();
+
+	if (ccids[ccid->ccid_id] != NULL) {
+		if (ccid->ccid_exit != NULL)
+			ccid->ccid_exit(sk);
+		module_put(ccid->ccid_owner);
+	}
+
+	ccids_read_unlock();
+}
+
+EXPORT_SYMBOL_GPL(ccid_exit);

net/dccp/ccid.h

@@ -0,0 +1,156 @@
+#ifndef _CCID_H
+#define _CCID_H
+/*
+ *  net/dccp/ccid.h
+ *
+ *  An implementation of the DCCP protocol
+ *  Arnaldo Carvalho de Melo <acme@conectiva.com.br>
+ *
+ *  CCID infrastructure
+ *
+ *	This program is free software; you can redistribute it and/or modify it
+ *	under the terms of the GNU General Public License version 2 as
+ *	published by the Free Software Foundation.
+ */
+
+#include <net/sock.h>
+#include <linux/dccp.h>
+#include <linux/list.h>
+#include <linux/module.h>
+
+#define CCID_MAX 255
+
+struct ccid {
+	unsigned char	ccid_id;
+	const char	*ccid_name;
+	struct module	*ccid_owner;
+	int		(*ccid_init)(struct sock *sk);
+	void		(*ccid_exit)(struct sock *sk);
+	int		(*ccid_hc_rx_init)(struct sock *sk);
+	int		(*ccid_hc_tx_init)(struct sock *sk);
+	void		(*ccid_hc_rx_exit)(struct sock *sk);
+	void		(*ccid_hc_tx_exit)(struct sock *sk);
+	void		(*ccid_hc_rx_packet_recv)(struct sock *sk, struct sk_buff *skb);
+	int		(*ccid_hc_rx_parse_options)(struct sock *sk,
+						    unsigned char option,
+						    unsigned char len, u16 idx,
+						    unsigned char* value);
+	void		(*ccid_hc_rx_insert_options)(struct sock *sk, struct sk_buff *skb);
+	void		(*ccid_hc_tx_insert_options)(struct sock *sk, struct sk_buff *skb);
+	void		(*ccid_hc_tx_packet_recv)(struct sock *sk, struct sk_buff *skb);
+	int		(*ccid_hc_tx_parse_options)(struct sock *sk,
+						    unsigned char option,
+						    unsigned char len, u16 idx,
+						    unsigned char* value);
+	int		(*ccid_hc_tx_send_packet)(struct sock *sk,
+						  struct sk_buff *skb, int len,
+						  long *delay);
+	void		(*ccid_hc_tx_packet_sent)(struct sock *sk, int more, int len);
+};
+
+extern int	   ccid_register(struct ccid *ccid);
+extern int	   ccid_unregister(struct ccid *ccid);
+
+extern struct ccid *ccid_init(unsigned char id, struct sock *sk);
+extern void	   ccid_exit(struct ccid *ccid, struct sock *sk);
+
+static inline void __ccid_get(struct ccid *ccid)
+{
+	__module_get(ccid->ccid_owner);
+}
+
+static inline int ccid_hc_tx_send_packet(struct ccid *ccid, struct sock *sk,
+					 struct sk_buff *skb, int len,
+					 long *delay)
+{
+	int rc = 0;
+	if (ccid->ccid_hc_tx_send_packet != NULL)
+		rc = ccid->ccid_hc_tx_send_packet(sk, skb, len, delay);
+	return rc;
+}
+
+static inline void ccid_hc_tx_packet_sent(struct ccid *ccid, struct sock *sk,
+					  int more, int len)
+{
+	if (ccid->ccid_hc_tx_packet_sent != NULL)
+		ccid->ccid_hc_tx_packet_sent(sk, more, len);
+}
+
+static inline int ccid_hc_rx_init(struct ccid *ccid, struct sock *sk)
+{
+	int rc = 0;
+	if (ccid->ccid_hc_rx_init != NULL)
+		rc = ccid->ccid_hc_rx_init(sk);
+	return rc;
+}
+
+static inline int ccid_hc_tx_init(struct ccid *ccid, struct sock *sk)
+{
+	int rc = 0;
+	if (ccid->ccid_hc_tx_init != NULL)
+		rc = ccid->ccid_hc_tx_init(sk);
+	return rc;
+}
+
+static inline void ccid_hc_rx_exit(struct ccid *ccid, struct sock *sk)
+{
+	if (ccid->ccid_hc_rx_exit != NULL)
+		ccid->ccid_hc_rx_exit(sk);
+}
+
+static inline void ccid_hc_tx_exit(struct ccid *ccid, struct sock *sk)
+{
+	if (ccid->ccid_hc_tx_exit != NULL)
+		ccid->ccid_hc_tx_exit(sk);
+}
+
+static inline void ccid_hc_rx_packet_recv(struct ccid *ccid, struct sock *sk,
+					  struct sk_buff *skb)
+{
+	if (ccid->ccid_hc_rx_packet_recv != NULL)
+		ccid->ccid_hc_rx_packet_recv(sk, skb);
+}
+
+static inline void ccid_hc_tx_packet_recv(struct ccid *ccid, struct sock *sk,
+					  struct sk_buff *skb)
+{
+	if (ccid->ccid_hc_tx_packet_recv != NULL)
+		ccid->ccid_hc_tx_packet_recv(sk, skb);
+}
+
+static inline int ccid_hc_tx_parse_options(struct ccid *ccid, struct sock *sk,
+					   unsigned char option,
+					   unsigned char len, u16 idx,
+					   unsigned char* value)
+{
+	int rc = 0;
+	if (ccid->ccid_hc_tx_parse_options != NULL)
+		rc = ccid->ccid_hc_tx_parse_options(sk, option, len, idx, value);
+	return rc;
+}
+
+static inline int ccid_hc_rx_parse_options(struct ccid *ccid, struct sock *sk,
+					   unsigned char option,
+					   unsigned char len, u16 idx,
+					   unsigned char* value)
+{
+	int rc = 0;
+	if (ccid->ccid_hc_rx_parse_options != NULL)
+		rc = ccid->ccid_hc_rx_parse_options(sk, option, len, idx, value);
+	return rc;
+}
+
+static inline void ccid_hc_tx_insert_options(struct ccid *ccid, struct sock *sk,
+					     struct sk_buff *skb)
+{
+	if (ccid->ccid_hc_tx_insert_options != NULL)
+		ccid->ccid_hc_tx_insert_options(sk, skb);
+}
+
+static inline void ccid_hc_rx_insert_options(struct ccid *ccid, struct sock *sk,
+					     struct sk_buff *skb)
+{
+	if (ccid->ccid_hc_rx_insert_options != NULL)
+		ccid->ccid_hc_rx_insert_options(sk, skb);
+}
+#endif /* _CCID_H */

net/dccp/ccids/Kconfig

@@ -0,0 +1,25 @@
+menu "DCCP CCIDs Configuration (EXPERIMENTAL)"
+	depends on IP_DCCP && EXPERIMENTAL
+
+config IP_DCCP_CCID3
+	tristate "CCID3 (TFRC) (EXPERIMENTAL)"
+	depends on IP_DCCP
+	---help---
+	  CCID 3 denotes TCP-Friendly Rate Control (TFRC), an equation-based
+	  rate-controlled congestion control mechanism.  TFRC is designed to
+	  be reasonably fair when competing for bandwidth with TCP-like flows,
+	  where a flow is "reasonably fair" if its sending rate is generally
+	  within a factor of two of the sending rate of a TCP flow under the
+	  same conditions.  However, TFRC has a much lower variation of
+	  throughput over time compared with TCP, which makes CCID 3 more
+	  suitable than CCID 2 for applications such streaming media where a
+	  relatively smooth sending rate is of importance.
+
+	  CCID 3 is further described in [CCID 3 PROFILE]. The TFRC
+	  congestion control algorithms were initially described in RFC 3448.
+
+	  This text was extracted from draft-ietf-dccp-spec-11.txt.
+	  
+	  If in doubt, say M.
+
+endmenu

net/dccp/ccids/Makefile

@@ -0,0 +1,3 @@
+obj-$(CONFIG_IP_DCCP_CCID3) += dccp_ccid3.o
+
+dccp_ccid3-y := ccid3.o

net/dccp/ccids/ccid3.h

@@ -0,0 +1,137 @@
+/*
+ *  net/dccp/ccids/ccid3.h
+ *
+ *  Copyright (c) 2005 The University of Waikato, Hamilton, New Zealand.
+ *
+ *  An implementation of the DCCP protocol
+ *
+ *  This code has been developed by the University of Waikato WAND
+ *  research group. For further information please see http://www.wand.net.nz/
+ *  or e-mail Ian McDonald - iam4@cs.waikato.ac.nz
+ *
+ *  This code also uses code from Lulea University, rereleased as GPL by its
+ *  authors:
+ *  Copyright (c) 2003 Nils-Erik Mattsson, Joacim Haggmark, Magnus Erixzon
+ *
+ *  Changes to meet Linux coding standards, to make it meet latest ccid3 draft
+ *  and to make it work as a loadable module in the DCCP stack written by
+ *  Arnaldo Carvalho de Melo <acme@conectiva.com.br>.
+ *
+ *  Copyright (c) 2005 Arnaldo Carvalho de Melo <acme@conectiva.com.br>
+ *
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of the GNU General Public License as published by
+ *  the Free Software Foundation; either version 2 of the License, or
+ *  (at your option) any later version.
+ *
+ *  This program is distributed in the hope that it will be useful,
+ *  but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *  GNU General Public License for more details.
+ *
+ *  You should have received a copy of the GNU General Public License
+ *  along with this program; if not, write to the Free Software
+ *  Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
+ */
+#ifndef _DCCP_CCID3_H_
+#define _DCCP_CCID3_H_
+
+#include <linux/types.h>
+#include <linux/list.h>
+#include <linux/timer.h>
+
+struct ccid3_tx_hist_entry {
+	struct list_head	ccid3htx_node;
+	u64			ccid3htx_seqno:48,
+				ccid3htx_win_count:8,
+				ccid3htx_sent:1;
+	struct timeval		ccid3htx_tstamp;
+};
+
+struct ccid3_options_received {
+	u64 ccid3or_seqno:48,
+	    ccid3or_loss_intervals_idx:16;
+	u16 ccid3or_loss_intervals_len;
+	u32 ccid3or_loss_event_rate;
+	u32 ccid3or_receive_rate;
+};
+
+/** struct ccid3_hc_tx_sock - CCID3 sender half connection congestion control block
+ *
+  * @ccid3hctx_state - Sender state
+  * @ccid3hctx_x - Current sending rate
+  * @ccid3hctx_x_recv - Receive rate
+  * @ccid3hctx_x_calc - Calculated send (?) rate
+  * @ccid3hctx_s - Packet size
+  * @ccid3hctx_rtt - Estimate of current round trip time in usecs
+  * @@ccid3hctx_p - Current loss event rate (0-1) scaled by 1000000
+  * @ccid3hctx_last_win_count - Last window counter sent
+  * @ccid3hctx_t_last_win_count - Timestamp of earliest packet with last_win_count value sent
+  * @ccid3hctx_no_feedback_timer - Handle to no feedback timer
+  * @ccid3hctx_idle - FIXME
+  * @ccid3hctx_t_ld - Time last doubled during slow start
+  * @ccid3hctx_t_nom - Nominal send time of next packet
+  * @ccid3hctx_t_ipi - Interpacket (send) interval
+  * @ccid3hctx_delta - Send timer delta
+  * @ccid3hctx_hist - Packet history
+  */
+struct ccid3_hc_tx_sock {
+	u32				ccid3hctx_x;
+	u32				ccid3hctx_x_recv;
+	u32				ccid3hctx_x_calc;
+	u16				ccid3hctx_s;
+	u32				ccid3hctx_rtt;
+	u32				ccid3hctx_p;
+  	u8				ccid3hctx_state;
+	u8				ccid3hctx_last_win_count;
+	u8				ccid3hctx_idle;
+	struct timeval			ccid3hctx_t_last_win_count;
+	struct timer_list		ccid3hctx_no_feedback_timer;
+	struct timeval			ccid3hctx_t_ld;
+	struct timeval			ccid3hctx_t_nom;
+	u32				ccid3hctx_t_ipi;
+	u32				ccid3hctx_delta;
+	struct list_head		ccid3hctx_hist;
+	struct ccid3_options_received	ccid3hctx_options_received;
+};
+
+struct ccid3_loss_interval_hist_entry {
+	struct list_head	ccid3lih_node;
+	u64			ccid3lih_seqno:48,
+				ccid3lih_win_count:4;
+	u32			ccid3lih_interval;
+};
+
+struct ccid3_rx_hist_entry {
+	struct list_head	ccid3hrx_node;
+	u64			ccid3hrx_seqno:48,
+				ccid3hrx_win_count:4,
+				ccid3hrx_type:4;
+	u32			ccid3hrx_ndp; /* In fact it is from 8 to 24 bits */
+	struct timeval		ccid3hrx_tstamp;
+};
+
+struct ccid3_hc_rx_sock {
+  	u64			ccid3hcrx_seqno_last_counter:48,
+				ccid3hcrx_state:8,
+				ccid3hcrx_last_counter:4;
+	unsigned long		ccid3hcrx_rtt;
+  	u32			ccid3hcrx_p;
+  	u32			ccid3hcrx_bytes_recv;
+  	struct timeval		ccid3hcrx_tstamp_last_feedback;
+  	struct timeval		ccid3hcrx_tstamp_last_ack;
+	struct list_head	ccid3hcrx_hist;
+	struct list_head	ccid3hcrx_loss_interval_hist;
+  	u16			ccid3hcrx_s;
+  	u32			ccid3hcrx_pinv;
+  	u32			ccid3hcrx_elapsed_time;
+  	u32			ccid3hcrx_x_recv;
+};
+
+#define ccid3_hc_tx_field(s,field) (s->dccps_hc_tx_ccid_private == NULL ? 0 : \
+				    ((struct ccid3_hc_tx_sock *)s->dccps_hc_tx_ccid_private)->ccid3hctx_##field)
+
+#define ccid3_hc_rx_field(s,field) (s->dccps_hc_rx_ccid_private == NULL ? 0 : \
+				    ((struct ccid3_hc_rx_sock *)s->dccps_hc_rx_ccid_private)->ccid3hcrx_##field)
+
+#endif /* _DCCP_CCID3_H_ */

net/dccp/dccp.h

@@ -0,0 +1,422 @@
+#ifndef _DCCP_H
+#define _DCCP_H
+/*
+ *  net/dccp/dccp.h
+ *
+ *  An implementation of the DCCP protocol
+ *  Arnaldo Carvalho de Melo <acme@conectiva.com.br>
+ *
+ *	This program is free software; you can redistribute it and/or modify it
+ *	under the terms of the GNU General Public License version 2 as
+ *	published by the Free Software Foundation.
+ */
+
+#include <linux/dccp.h>
+#include <net/snmp.h>
+#include <net/sock.h>
+#include <net/tcp.h>
+
+#define DCCP_DEBUG
+
+#ifdef DCCP_DEBUG
+extern int dccp_debug;
+
+#define dccp_pr_debug(format, a...) \
+	do { if (dccp_debug) \
+		printk(KERN_DEBUG "%s: " format, __FUNCTION__ , ##a); \
+	} while (0)
+#define dccp_pr_debug_cat(format, a...) do { if (dccp_debug) printk(format, ##a); } while (0)
+#else
+#define dccp_pr_debug(format, a...)
+#define dccp_pr_debug_cat(format, a...)
+#endif
+
+extern struct inet_hashinfo dccp_hashinfo;
+
+extern atomic_t dccp_orphan_count;
+extern int dccp_tw_count;
+extern void dccp_tw_deschedule(struct inet_timewait_sock *tw);
+
+extern void dccp_time_wait(struct sock *sk, int state, int timeo);
+
+/* FIXME: Right size this */
+#define DCCP_MAX_OPT_LEN 128
+
+#define DCCP_MAX_PACKET_HDR 32
+
+#define MAX_DCCP_HEADER  (DCCP_MAX_PACKET_HDR + DCCP_MAX_OPT_LEN + MAX_HEADER)
+
+#define DCCP_TIMEWAIT_LEN (60 * HZ) /* how long to wait to destroy TIME-WAIT
+				     * state, about 60 seconds */
+
+/* draft-ietf-dccp-spec-11.txt initial RTO value */
+#define DCCP_TIMEOUT_INIT ((unsigned)(3 * HZ))
+
+/* Maximal interval between probes for local resources.  */
+#define DCCP_RESOURCE_PROBE_INTERVAL ((unsigned)(HZ / 2U))
+
+#define DCCP_RTO_MAX ((unsigned)(120 * HZ)) /* FIXME: using TCP value */
+
+extern struct proto dccp_v4_prot;
+
+/* is seq1 < seq2 ? */
+static inline const  int before48(const u64 seq1, const u64 seq2)
+{
+	return (const s64)((seq1 << 16) - (seq2 << 16)) < 0;
+}
+
+/* is seq1 > seq2 ? */
+static inline const  int after48(const u64 seq1, const u64 seq2)
+{
+	return (const s64)((seq2 << 16) - (seq1 << 16)) < 0;
+}
+
+/* is seq2 <= seq1 <= seq3 ? */
+static inline const int between48(const u64 seq1, const u64 seq2, const u64 seq3)
+{
+	return (seq3 << 16) - (seq2 << 16) >= (seq1 << 16) - (seq2 << 16);
+}
+
+static inline u64 max48(const u64 seq1, const u64 seq2)
+{
+	return after48(seq1, seq2) ? seq1 : seq2;
+}
+
+enum {
+	DCCP_MIB_NUM = 0,
+	DCCP_MIB_ACTIVEOPENS,			/* ActiveOpens */
+	DCCP_MIB_ESTABRESETS,			/* EstabResets */
+	DCCP_MIB_CURRESTAB,			/* CurrEstab */
+	DCCP_MIB_OUTSEGS,			/* OutSegs */ 
+	DCCP_MIB_OUTRSTS,
+	DCCP_MIB_ABORTONTIMEOUT,
+	DCCP_MIB_TIMEOUTS,
+	DCCP_MIB_ABORTFAILED,
+	DCCP_MIB_PASSIVEOPENS,
+	DCCP_MIB_ATTEMPTFAILS,
+	DCCP_MIB_OUTDATAGRAMS,
+	DCCP_MIB_INERRS,
+	DCCP_MIB_OPTMANDATORYERROR,
+	DCCP_MIB_INVALIDOPT,
+	__DCCP_MIB_MAX
+};
+
+#define DCCP_MIB_MAX	__DCCP_MIB_MAX
+struct dccp_mib {
+	unsigned long	mibs[DCCP_MIB_MAX];
+} __SNMP_MIB_ALIGN__;
+
+DECLARE_SNMP_STAT(struct dccp_mib, dccp_statistics);
+#define DCCP_INC_STATS(field)		SNMP_INC_STATS(dccp_statistics, field)
+#define DCCP_INC_STATS_BH(field)	SNMP_INC_STATS_BH(dccp_statistics, field)
+#define DCCP_INC_STATS_USER(field) 	SNMP_INC_STATS_USER(dccp_statistics, field)
+#define DCCP_DEC_STATS(field)		SNMP_DEC_STATS(dccp_statistics, field)
+#define DCCP_ADD_STATS_BH(field, val)	SNMP_ADD_STATS_BH(dccp_statistics, field, val)
+#define DCCP_ADD_STATS_USER(field, val)	SNMP_ADD_STATS_USER(dccp_statistics, field, val)
+
+extern int  dccp_transmit_skb(struct sock *sk, struct sk_buff *skb);
+extern int  dccp_retransmit_skb(struct sock *sk, struct sk_buff *skb);
+
+extern int dccp_send_response(struct sock *sk);
+extern void dccp_send_ack(struct sock *sk);
+extern void dccp_send_delayed_ack(struct sock *sk);
+extern void dccp_send_sync(struct sock *sk, u64 seq);
+
+extern void dccp_init_xmit_timers(struct sock *sk);
+static inline void dccp_clear_xmit_timers(struct sock *sk)
+{
+	inet_csk_clear_xmit_timers(sk);
+}
+
+extern unsigned int dccp_sync_mss(struct sock *sk, u32 pmtu);
+
+extern const char *dccp_packet_name(const int type);
+extern const char *dccp_state_name(const int state);
+
+static inline void dccp_set_state(struct sock *sk, const int state)
+{
+	const int oldstate = sk->sk_state;
+
+	dccp_pr_debug("%s(%p) %-10.10s -> %s\n",
+		      dccp_role(sk), sk,
+		      dccp_state_name(oldstate), dccp_state_name(state));
+	WARN_ON(state == oldstate);
+
+	switch (state) {
+	case DCCP_OPEN:
+		if (oldstate != DCCP_OPEN)
+			DCCP_INC_STATS(DCCP_MIB_CURRESTAB);
+		break;
+
+	case DCCP_CLOSED:
+		if (oldstate == DCCP_CLOSING || oldstate == DCCP_OPEN)
+			DCCP_INC_STATS(DCCP_MIB_ESTABRESETS);
+
+		sk->sk_prot->unhash(sk);
+		if (inet_csk(sk)->icsk_bind_hash != NULL &&
+		    !(sk->sk_userlocks & SOCK_BINDPORT_LOCK))
+			inet_put_port(&dccp_hashinfo, sk);
+		/* fall through */
+	default:
+		if (oldstate == DCCP_OPEN)
+			DCCP_DEC_STATS(DCCP_MIB_CURRESTAB);
+	}
+
+	/* Change state AFTER socket is unhashed to avoid closed
+	 * socket sitting in hash tables.
+	 */
+	sk->sk_state = state;
+}
+
+static inline void dccp_done(struct sock *sk)
+{
+	dccp_set_state(sk, DCCP_CLOSED);
+	dccp_clear_xmit_timers(sk);
+
+	sk->sk_shutdown = SHUTDOWN_MASK;
+
+	if (!sock_flag(sk, SOCK_DEAD))
+		sk->sk_state_change(sk);
+	else
+		inet_csk_destroy_sock(sk);
+}
+
+static inline void dccp_openreq_init(struct request_sock *req,
+				     struct dccp_sock *dp,
+				     struct sk_buff *skb)
+{
+	/*
+	 * FIXME: fill in the other req fields from the DCCP options
+	 * received
+	 */
+	inet_rsk(req)->rmt_port = dccp_hdr(skb)->dccph_sport;
+	inet_rsk(req)->acked	= 0;
+	req->rcv_wnd = 0;
+}
+
+extern void dccp_v4_send_check(struct sock *sk, struct dccp_hdr *dh, int len, 
+			       struct sk_buff *skb);
+extern int dccp_v4_conn_request(struct sock *sk, struct sk_buff *skb);
+
+extern struct sock *dccp_create_openreq_child(struct sock *sk,
+					      const struct request_sock *req,
+					      const struct sk_buff *skb);
+
+extern int dccp_v4_do_rcv(struct sock *sk, struct sk_buff *skb);
+
+extern void dccp_v4_err(struct sk_buff *skb, u32);
+
+extern int dccp_v4_rcv(struct sk_buff *skb);
+
+extern struct sock *dccp_v4_request_recv_sock(struct sock *sk,
+					      struct sk_buff *skb,
+					      struct request_sock *req,
+					      struct dst_entry *dst);
+extern struct sock *dccp_check_req(struct sock *sk, struct sk_buff *skb,
+				   struct request_sock *req,
+				   struct request_sock **prev);
+
+extern int dccp_child_process(struct sock *parent, struct sock *child,
+			      struct sk_buff *skb);
+extern int dccp_rcv_state_process(struct sock *sk, struct sk_buff *skb,
+				  struct dccp_hdr *dh, unsigned len);
+extern int dccp_rcv_established(struct sock *sk, struct sk_buff *skb,
+				const struct dccp_hdr *dh, const unsigned len);
+
+extern void		dccp_close(struct sock *sk, long timeout);
+extern struct sk_buff	*dccp_make_response(struct sock *sk,
+					    struct dst_entry *dst,
+					    struct request_sock *req);
+
+extern int	   dccp_connect(struct sock *sk);
+extern int	   dccp_disconnect(struct sock *sk, int flags);
+extern int	   dccp_getsockopt(struct sock *sk, int level, int optname,
+				   char *optval, int *optlen);
+extern int	   dccp_ioctl(struct sock *sk, int cmd, unsigned long arg);
+extern int	   dccp_sendmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg,
+				size_t size);
+extern int	   dccp_recvmsg(struct kiocb *iocb, struct sock *sk,
+				struct msghdr *msg, size_t len, int nonblock,
+				int flags, int *addr_len);
+extern int	   dccp_setsockopt(struct sock *sk, int level, int optname,
+				   char *optval, int optlen);
+extern void	   dccp_shutdown(struct sock *sk, int how);
+
+extern int	   dccp_v4_checksum(struct sk_buff *skb);
+
+extern int	   dccp_v4_send_reset(struct sock *sk, enum dccp_reset_codes code);
+extern void	   dccp_send_close(struct sock *sk);
+
+struct dccp_skb_cb {
+	__u8 dccpd_type;
+	__u8 dccpd_reset_code;
+	__u8 dccpd_service;
+	__u8 dccpd_ccval;
+	__u64 dccpd_seq;
+	__u64 dccpd_ack_seq;
+	int  dccpd_opt_len;
+};
+
+#define DCCP_SKB_CB(__skb) ((struct dccp_skb_cb *)&((__skb)->cb[0]))
+
+static inline int dccp_non_data_packet(const struct sk_buff *skb)
+{
+	const __u8 type = DCCP_SKB_CB(skb)->dccpd_type;
+
+	return type == DCCP_PKT_ACK	 ||
+	       type == DCCP_PKT_CLOSE	 ||
+	       type == DCCP_PKT_CLOSEREQ ||
+	       type == DCCP_PKT_RESET	 ||
+	       type == DCCP_PKT_SYNC	 ||
+	       type == DCCP_PKT_SYNCACK;
+}
+
+static inline int dccp_packet_without_ack(const struct sk_buff *skb)
+{
+	const __u8 type = DCCP_SKB_CB(skb)->dccpd_type;
+
+	return type == DCCP_PKT_DATA || type == DCCP_PKT_REQUEST;
+}
+
+#define DCCP_MAX_SEQNO ((((u64)1) << 48) - 1)
+#define DCCP_PKT_WITHOUT_ACK_SEQ (DCCP_MAX_SEQNO << 2)
+
+static inline void dccp_set_seqno(u64 *seqno, u64 value)
+{
+	if (value > DCCP_MAX_SEQNO)
+		value -= DCCP_MAX_SEQNO + 1;
+	*seqno = value;
+}
+
+static inline u64 dccp_delta_seqno(u64 seqno1, u64 seqno2)
+{
+	return ((seqno2 << 16) - (seqno1 << 16)) >> 16;
+}
+
+static inline void dccp_inc_seqno(u64 *seqno)
+{
+	if (++*seqno > DCCP_MAX_SEQNO)
+		*seqno = 0;
+}
+
+static inline void dccp_hdr_set_seq(struct dccp_hdr *dh, const u64 gss)
+{
+	struct dccp_hdr_ext *dhx = (struct dccp_hdr_ext *)((void *)dh + sizeof(*dh));
+
+#if defined(__LITTLE_ENDIAN_BITFIELD)
+	dh->dccph_seq	   = htonl((gss >> 32)) >> 8;
+#elif defined(__BIG_ENDIAN_BITFIELD)
+	dh->dccph_seq	   = htonl((gss >> 32));
+#else
+#error  "Adjust your <asm/byteorder.h> defines"
+#endif
+	dhx->dccph_seq_low = htonl(gss & 0xffffffff);
+}
+
+static inline void dccp_hdr_set_ack(struct dccp_hdr_ack_bits *dhack, const u64 gsr)
+{
+#if defined(__LITTLE_ENDIAN_BITFIELD)
+	dhack->dccph_ack_nr_high = htonl((gsr >> 32)) >> 8;
+#elif defined(__BIG_ENDIAN_BITFIELD)
+	dhack->dccph_ack_nr_high = htonl((gsr >> 32));
+#else
+#error  "Adjust your <asm/byteorder.h> defines"
+#endif
+	dhack->dccph_ack_nr_low  = htonl(gsr & 0xffffffff);
+}
+
+static inline void dccp_update_gsr(struct sock *sk, u64 seq)
+{
+	struct dccp_sock *dp = dccp_sk(sk);
+	u64 tmp_gsr;
+
+	dccp_set_seqno(&tmp_gsr, dp->dccps_gsr + 1 - (dp->dccps_options.dccpo_sequence_window / 4));
+	dp->dccps_gsr = seq;
+	dccp_set_seqno(&dp->dccps_swl, max48(tmp_gsr, dp->dccps_isr));
+	dccp_set_seqno(&dp->dccps_swh,
+		       dp->dccps_gsr + (3 * dp->dccps_options.dccpo_sequence_window) / 4);
+}
+
+static inline void dccp_update_gss(struct sock *sk, u64 seq)
+{
+	struct dccp_sock *dp = dccp_sk(sk);
+	u64 tmp_gss;
+
+	dccp_set_seqno(&tmp_gss, dp->dccps_gss - dp->dccps_options.dccpo_sequence_window + 1);
+	dp->dccps_awl = max48(tmp_gss, dp->dccps_iss);
+	dp->dccps_awh = dp->dccps_gss = seq;
+}
+
+extern void dccp_insert_options(struct sock *sk, struct sk_buff *skb);
+extern void dccp_insert_option_elapsed_time(struct sock *sk,
+					    struct sk_buff *skb,
+					    u32 elapsed_time);
+extern void dccp_insert_option(struct sock *sk, struct sk_buff *skb,
+			       unsigned char option,
+			       const void *value, unsigned char len);
+
+extern struct socket *dccp_ctl_socket;
+
+#define DCCP_ACKPKTS_STATE_RECEIVED	0
+#define DCCP_ACKPKTS_STATE_ECN_MARKED	(1 << 6)
+#define DCCP_ACKPKTS_STATE_NOT_RECEIVED	(3 << 6)
+
+#define DCCP_ACKPKTS_STATE_MASK		0xC0 /* 11000000 */
+#define DCCP_ACKPKTS_LEN_MASK		0x3F /* 00111111 */
+
+/** struct dccp_ackpkts - acknowledgeable packets
+ *
+ * This data structure is the one defined in the DCCP draft
+ * Appendix A.
+ *
+ * @dccpap_buf_head - circular buffer head
+ * @dccpap_buf_tail - circular buffer tail
+ * @dccpap_buf_ackno - ack # of the most recent packet acknoldgeable in the buffer (i.e. %dccpap_buf_head)
+ * @dccpap_buf_nonce - the one-bit sum of the ECN Nonces on all packets acked by the buffer with State 0
+ *
+ * Additionally, the HC-Receiver must keep some information about the
+ * Ack Vectors it has recently sent. For each packet sent carrying an
+ * Ack Vector, it remembers four variables:
+ *
+ * @dccpap_ack_seqno - the Sequence Number used for the packet (HC-Receiver seqno)
+ * @dccpap_ack_ptr - the value of buf_head at the time of acknowledgement.
+ * @dccpap_ack_ackno - the Acknowledgement Number used for the packet (HC-Sender seqno)
+ * @dccpap_ack_nonce - the one-bit sum of the ECN Nonces for all State 0.
+ *
+ * @dccpap_buf_len - circular buffer length
+ * @dccpap_buf - circular buffer of acknowledgeable packets
+ */
+struct dccp_ackpkts {
+	unsigned int		dccpap_buf_head;
+	unsigned int		dccpap_buf_tail;
+	u64			dccpap_buf_ackno;
+	u64			dccpap_ack_seqno;
+	u64			dccpap_ack_ackno;
+	unsigned int		dccpap_ack_ptr;
+	unsigned int		dccpap_buf_vector_len;
+	unsigned int		dccpap_ack_vector_len;
+	unsigned int		dccpap_buf_len;
+	unsigned long           dccpap_time;
+	u8			dccpap_buf_nonce;
+	u8			dccpap_ack_nonce;
+	u8			dccpap_buf[0];
+};
+
+extern struct dccp_ackpkts *dccp_ackpkts_alloc(unsigned int len, int priority);
+extern void dccp_ackpkts_free(struct dccp_ackpkts *ap);
+extern int dccp_ackpkts_add(struct dccp_ackpkts *ap, u64 ackno, u8 state);
+extern void dccp_ackpkts_check_rcv_ackno(struct dccp_ackpkts *ap,
+					 struct sock *sk, u64 ackno);
+
+#ifdef DCCP_DEBUG
+extern void dccp_ackvector_print(const u64 ackno,
+				 const unsigned char *vector, int len);
+extern void dccp_ackpkts_print(const struct dccp_ackpkts *ap);
+#else
+static inline void dccp_ackvector_print(const u64 ackno,
+					const unsigned char *vector,
+					int len) { }
+static inline void dccp_ackpkts_print(const struct dccp_ackpkts *ap) { }
+#endif
+
+#endif /* _DCCP_H */

net/dccp/input.c

@@ -0,0 +1,510 @@
+/*
+ *  net/dccp/input.c
+ * 
+ *  An implementation of the DCCP protocol
+ *  Arnaldo Carvalho de Melo <acme@conectiva.com.br>
+ *
+ *	This program is free software; you can redistribute it and/or
+ *	modify it under the terms of the GNU General Public License
+ *	as published by the Free Software Foundation; either version
+ *	2 of the License, or (at your option) any later version.
+ */
+
+#include <linux/config.h>
+#include <linux/dccp.h>
+#include <linux/skbuff.h>
+
+#include <net/sock.h>
+
+#include "ccid.h"
+#include "dccp.h"
+
+static void dccp_fin(struct sock *sk, struct sk_buff *skb)
+{
+	sk->sk_shutdown |= RCV_SHUTDOWN;
+	sock_set_flag(sk, SOCK_DONE);
+	__skb_pull(skb, dccp_hdr(skb)->dccph_doff * 4);
+	__skb_queue_tail(&sk->sk_receive_queue, skb);
+	skb_set_owner_r(skb, sk);
+	sk->sk_data_ready(sk, 0);
+}
+
+static void dccp_rcv_close(struct sock *sk, struct sk_buff *skb)
+{
+	switch (sk->sk_state) {
+	case DCCP_PARTOPEN:
+	case DCCP_OPEN:
+		dccp_v4_send_reset(sk, DCCP_RESET_CODE_CLOSED);
+		dccp_fin(sk, skb);
+		dccp_set_state(sk, DCCP_CLOSED);
+		break;
+	}
+}
+
+static void dccp_rcv_closereq(struct sock *sk, struct sk_buff *skb)
+{
+	/*
+	 *   Step 7: Check for unexpected packet types
+	 *      If (S.is_server and P.type == CloseReq)
+	 *	  Send Sync packet acknowledging P.seqno
+	 *	  Drop packet and return
+	 */
+	if (dccp_sk(sk)->dccps_role != DCCP_ROLE_CLIENT) {
+		dccp_send_sync(sk, DCCP_SKB_CB(skb)->dccpd_seq);
+		return;
+	}
+
+	switch (sk->sk_state) {
+	case DCCP_PARTOPEN:
+	case DCCP_OPEN:
+		dccp_set_state(sk, DCCP_CLOSING);
+		dccp_send_close(sk);
+		break;
+	}
+}
+
+static inline void dccp_event_ack_recv(struct sock *sk, struct sk_buff *skb)
+{
+	struct dccp_sock *dp = dccp_sk(sk);
+
+	if (dp->dccps_options.dccpo_send_ack_vector)
+		dccp_ackpkts_check_rcv_ackno(dp->dccps_hc_rx_ackpkts, sk,
+					     DCCP_SKB_CB(skb)->dccpd_ack_seq);
+}
+
+static int dccp_check_seqno(struct sock *sk, struct sk_buff *skb)
+{
+	const struct dccp_hdr *dh = dccp_hdr(skb);
+	struct dccp_sock *dp = dccp_sk(sk);
+	u64 lswl = dp->dccps_swl;
+	u64 lawl = dp->dccps_awl;
+
+	/*
+	 *   Step 5: Prepare sequence numbers for Sync
+	 *     If P.type == Sync or P.type == SyncAck,
+	 *	  If S.AWL <= P.ackno <= S.AWH and P.seqno >= S.SWL,
+	 *	     / * P is valid, so update sequence number variables
+	 *		 accordingly.  After this update, P will pass the tests
+	 *		 in Step 6.  A SyncAck is generated if necessary in
+	 *		 Step 15 * /
+	 *	     Update S.GSR, S.SWL, S.SWH
+	 *	  Otherwise,
+	 *	     Drop packet and return
+	 */
+	if (dh->dccph_type == DCCP_PKT_SYNC || 
+	    dh->dccph_type == DCCP_PKT_SYNCACK) {
+		if (between48(DCCP_SKB_CB(skb)->dccpd_ack_seq, dp->dccps_awl, dp->dccps_awh) &&
+		    !before48(DCCP_SKB_CB(skb)->dccpd_seq, dp->dccps_swl))
+			dccp_update_gsr(sk, DCCP_SKB_CB(skb)->dccpd_seq);
+		else
+			return -1;
+	/*
+	 *   Step 6: Check sequence numbers
+	 *      Let LSWL = S.SWL and LAWL = S.AWL
+	 *      If P.type == CloseReq or P.type == Close or P.type == Reset,
+	 *	  LSWL := S.GSR + 1, LAWL := S.GAR
+	 *      If LSWL <= P.seqno <= S.SWH
+	 *	     and (P.ackno does not exist or LAWL <= P.ackno <= S.AWH),
+	 *	  Update S.GSR, S.SWL, S.SWH
+	 *	  If P.type != Sync,
+	 *	     Update S.GAR
+	 *      Otherwise,
+	 *	  Send Sync packet acknowledging P.seqno
+	 *	  Drop packet and return
+	 */
+	} else if (dh->dccph_type == DCCP_PKT_CLOSEREQ ||
+		   dh->dccph_type == DCCP_PKT_CLOSE ||
+		   dh->dccph_type == DCCP_PKT_RESET) {
+		lswl = dp->dccps_gsr;
+		dccp_inc_seqno(&lswl);
+		lawl = dp->dccps_gar;
+	}
+
+	if (between48(DCCP_SKB_CB(skb)->dccpd_seq, lswl, dp->dccps_swh) &&
+	    (DCCP_SKB_CB(skb)->dccpd_ack_seq == DCCP_PKT_WITHOUT_ACK_SEQ ||
+	     between48(DCCP_SKB_CB(skb)->dccpd_ack_seq, lawl, dp->dccps_awh))) {
+		dccp_update_gsr(sk, DCCP_SKB_CB(skb)->dccpd_seq);
+
+		if (dh->dccph_type != DCCP_PKT_SYNC &&
+		    DCCP_SKB_CB(skb)->dccpd_ack_seq != DCCP_PKT_WITHOUT_ACK_SEQ)
+			dp->dccps_gar = DCCP_SKB_CB(skb)->dccpd_ack_seq;
+	} else {
+		dccp_pr_debug("Step 6 failed, sending SYNC...\n");
+		dccp_send_sync(sk, DCCP_SKB_CB(skb)->dccpd_seq);
+		return -1;
+	}
+
+	return 0;
+}
+
+int dccp_rcv_established(struct sock *sk, struct sk_buff *skb,
+			 const struct dccp_hdr *dh, const unsigned len)
+{
+	struct dccp_sock *dp = dccp_sk(sk);
+
+	if (dccp_check_seqno(sk, skb))
+		goto discard;
+
+	if (dccp_parse_options(sk, skb))
+		goto discard;
+
+	if (DCCP_SKB_CB(skb)->dccpd_ack_seq != DCCP_PKT_WITHOUT_ACK_SEQ)
+		dccp_event_ack_recv(sk, skb);
+
+	/*
+	 * FIXME: check ECN to see if we should use
+	 * DCCP_ACKPKTS_STATE_ECN_MARKED
+	 */
+	if (dp->dccps_options.dccpo_send_ack_vector) {
+		struct dccp_ackpkts *ap = dp->dccps_hc_rx_ackpkts;
+
+		if (dccp_ackpkts_add(dp->dccps_hc_rx_ackpkts,
+				     DCCP_SKB_CB(skb)->dccpd_seq,
+				     DCCP_ACKPKTS_STATE_RECEIVED)) {
+			LIMIT_NETDEBUG(pr_info("DCCP: acknowledgeable packets buffer full!\n"));
+			ap->dccpap_ack_seqno = DCCP_MAX_SEQNO + 1;
+			inet_csk_schedule_ack(sk);
+			inet_csk_reset_xmit_timer(sk, ICSK_TIME_DACK, TCP_DELACK_MIN, TCP_RTO_MAX);
+			goto discard;
+		}
+
+		/*
+		 * FIXME: this activation is probably wrong, have to study more
+		 * TCP delack machinery and how it fits into DCCP draft, but
+		 * for now it kinda "works" 8)
+		 */
+		if (!inet_csk_ack_scheduled(sk)) {
+			inet_csk_schedule_ack(sk);
+			inet_csk_reset_xmit_timer(sk, ICSK_TIME_DACK, 5 * HZ, TCP_RTO_MAX);
+		}
+	}
+
+	ccid_hc_rx_packet_recv(dp->dccps_hc_rx_ccid, sk, skb);
+	ccid_hc_tx_packet_recv(dp->dccps_hc_tx_ccid, sk, skb);
+
+	switch (dccp_hdr(skb)->dccph_type) {
+	case DCCP_PKT_DATAACK:
+	case DCCP_PKT_DATA:
+		/*
+		 * FIXME: check if sk_receive_queue is full, schedule DATA_DROPPED option
+		 * if it is.
+		 */
+		__skb_pull(skb, dh->dccph_doff * 4);
+		__skb_queue_tail(&sk->sk_receive_queue, skb);
+		skb_set_owner_r(skb, sk);
+		sk->sk_data_ready(sk, 0);
+		return 0;
+	case DCCP_PKT_ACK:
+		goto discard;
+	case DCCP_PKT_RESET:
+		/*
+		 *  Step 9: Process Reset
+		 *	If P.type == Reset,
+		 *		Tear down connection
+		 *		S.state := TIMEWAIT
+		 *		Set TIMEWAIT timer
+		 *		Drop packet and return
+		*/
+		dccp_fin(sk, skb);
+		dccp_time_wait(sk, DCCP_TIME_WAIT, 0);
+		return 0;
+	case DCCP_PKT_CLOSEREQ:
+		dccp_rcv_closereq(sk, skb);
+		goto discard;
+	case DCCP_PKT_CLOSE:
+		dccp_rcv_close(sk, skb);
+		return 0;
+	case DCCP_PKT_REQUEST:
+		/* Step 7 
+            	 *   or (S.is_server and P.type == Response)
+		 *   or (S.is_client and P.type == Request)
+		 *   or (S.state >= OPEN and P.type == Request
+		 *	and P.seqno >= S.OSR)
+		 *    or (S.state >= OPEN and P.type == Response
+		 *	and P.seqno >= S.OSR)
+		 *    or (S.state == RESPOND and P.type == Data),
+		 *  Send Sync packet acknowledging P.seqno
+		 *  Drop packet and return
+		 */
+		if (dp->dccps_role != DCCP_ROLE_LISTEN)
+			goto send_sync;
+		goto check_seq;
+	case DCCP_PKT_RESPONSE:
+		if (dp->dccps_role != DCCP_ROLE_CLIENT)
+			goto send_sync;
+check_seq:
+		if (!before48(DCCP_SKB_CB(skb)->dccpd_seq, dp->dccps_osr)) {
+send_sync:
+			dccp_send_sync(sk, DCCP_SKB_CB(skb)->dccpd_seq);
+		}
+		break;
+	}
+
+	DCCP_INC_STATS_BH(DCCP_MIB_INERRS);
+discard:
+	__kfree_skb(skb);
+	return 0;
+}
+
+static int dccp_rcv_request_sent_state_process(struct sock *sk,
+					       struct sk_buff *skb,
+					       const struct dccp_hdr *dh,
+					       const unsigned len)
+{
+	/* 
+	 *  Step 4: Prepare sequence numbers in REQUEST
+	 *     If S.state == REQUEST,
+	 *	  If (P.type == Response or P.type == Reset)
+	 *		and S.AWL <= P.ackno <= S.AWH,
+	 *	     / * Set sequence number variables corresponding to the
+	 *		other endpoint, so P will pass the tests in Step 6 * /
+	 *	     Set S.GSR, S.ISR, S.SWL, S.SWH
+	 *	     / * Response processing continues in Step 10; Reset
+	 *		processing continues in Step 9 * /
+	*/
+	if (dh->dccph_type == DCCP_PKT_RESPONSE) {
+		const struct inet_connection_sock *icsk = inet_csk(sk);
+		struct dccp_sock *dp = dccp_sk(sk);
+
+		/* Stop the REQUEST timer */
+		inet_csk_clear_xmit_timer(sk, ICSK_TIME_RETRANS);
+		BUG_TRAP(sk->sk_send_head != NULL);
+		__kfree_skb(sk->sk_send_head);
+		sk->sk_send_head = NULL;
+
+		if (!between48(DCCP_SKB_CB(skb)->dccpd_ack_seq, dp->dccps_awl, dp->dccps_awh)) {
+			dccp_pr_debug("invalid ackno: S.AWL=%llu, P.ackno=%llu, S.AWH=%llu \n",
+				      dp->dccps_awl, DCCP_SKB_CB(skb)->dccpd_ack_seq, dp->dccps_awh);
+			goto out_invalid_packet;
+		}
+
+		dp->dccps_isr = DCCP_SKB_CB(skb)->dccpd_seq;
+		dccp_update_gsr(sk, DCCP_SKB_CB(skb)->dccpd_seq);
+
+		if (ccid_hc_rx_init(dp->dccps_hc_rx_ccid, sk) != 0 ||
+		    ccid_hc_tx_init(dp->dccps_hc_tx_ccid, sk) != 0) {
+			ccid_hc_rx_exit(dp->dccps_hc_rx_ccid, sk);
+			ccid_hc_tx_exit(dp->dccps_hc_tx_ccid, sk);
+			/* FIXME: send appropriate RESET code */
+			goto out_invalid_packet;
+		}
+
+		dccp_sync_mss(sk, dp->dccps_pmtu_cookie);
+
+		/*
+		 *    Step 10: Process REQUEST state (second part)
+		 *       If S.state == REQUEST,
+		 *	  / * If we get here, P is a valid Response from the server (see
+		 *	     Step 4), and we should move to PARTOPEN state.  PARTOPEN
+		 *	     means send an Ack, don't send Data packets, retransmit
+		 *	     Acks periodically, and always include any Init Cookie from
+		 *	     the Response * /
+		 *	  S.state := PARTOPEN
+		 *	  Set PARTOPEN timer
+		 * 	  Continue with S.state == PARTOPEN
+		 *	  / * Step 12 will send the Ack completing the three-way
+		 *	     handshake * /
+		 */
+		dccp_set_state(sk, DCCP_PARTOPEN);
+
+		/* Make sure socket is routed, for correct metrics. */
+		inet_sk_rebuild_header(sk);
+
+		if (!sock_flag(sk, SOCK_DEAD)) {
+			sk->sk_state_change(sk);
+			sk_wake_async(sk, 0, POLL_OUT);
+		}
+
+		if (sk->sk_write_pending || icsk->icsk_ack.pingpong ||
+		    icsk->icsk_accept_queue.rskq_defer_accept) {
+			/* Save one ACK. Data will be ready after
+			 * several ticks, if write_pending is set.
+			 *
+			 * It may be deleted, but with this feature tcpdumps
+			 * look so _wonderfully_ clever, that I was not able
+			 * to stand against the temptation 8)     --ANK
+			 */
+			/*
+			 * OK, in DCCP we can as well do a similar trick, its
+			 * even in the draft, but there is no need for us to
+			 * schedule an ack here, as dccp_sendmsg does this for
+			 * us, also stated in the draft. -acme
+			 */
+			__kfree_skb(skb);
+			return 0;
+		} 
+		dccp_send_ack(sk);
+		return -1;
+	}
+
+out_invalid_packet:
+	return 1; /* dccp_v4_do_rcv will send a reset, but...
+		     FIXME: the reset code should be DCCP_RESET_CODE_PACKET_ERROR  */
+}
+
+static int dccp_rcv_respond_partopen_state_process(struct sock *sk,
+						   struct sk_buff *skb,
+						   const struct dccp_hdr *dh,
+						   const unsigned len)
+{
+	int queued = 0;
+
+	switch (dh->dccph_type) {
+	case DCCP_PKT_RESET:
+		inet_csk_clear_xmit_timer(sk, ICSK_TIME_DACK);
+		break;
+	case DCCP_PKT_DATAACK:
+	case DCCP_PKT_ACK:
+		/*
+		 * FIXME: we should be reseting the PARTOPEN (DELACK) timer here,
+		 * but only if we haven't used the DELACK timer for something else,
+		 * like sending a delayed ack for a TIMESTAMP echo, etc, for now
+		 * were not clearing it, sending an extra ACK when there is nothing
+		 * else to do in DELACK is not a big deal after all.
+		 */
+
+		/* Stop the PARTOPEN timer */
+		if (sk->sk_state == DCCP_PARTOPEN)
+			inet_csk_clear_xmit_timer(sk, ICSK_TIME_DACK);
+
+		dccp_sk(sk)->dccps_osr = DCCP_SKB_CB(skb)->dccpd_seq;
+		dccp_set_state(sk, DCCP_OPEN);
+
+		if (dh->dccph_type == DCCP_PKT_DATAACK) {
+			dccp_rcv_established(sk, skb, dh, len);
+			queued = 1; /* packet was queued (by dccp_rcv_established) */
+		}
+		break;
+	}
+
+	return queued;
+}
+
+int dccp_rcv_state_process(struct sock *sk, struct sk_buff *skb,
+			   struct dccp_hdr *dh, unsigned len)
+{
+	struct dccp_sock *dp = dccp_sk(sk);
+	const int old_state = sk->sk_state;
+	int queued = 0;
+
+	if (sk->sk_state != DCCP_LISTEN && sk->sk_state != DCCP_REQUESTING) {
+		if (dccp_check_seqno(sk, skb))
+			goto discard;
+
+		/*
+		 * Step 8: Process options and mark acknowledgeable
+		 */
+		if (dccp_parse_options(sk, skb))
+			goto discard;
+
+		if (DCCP_SKB_CB(skb)->dccpd_ack_seq != DCCP_PKT_WITHOUT_ACK_SEQ)
+			dccp_event_ack_recv(sk, skb);
+
+		ccid_hc_rx_packet_recv(dp->dccps_hc_rx_ccid, sk, skb);
+		ccid_hc_tx_packet_recv(dp->dccps_hc_tx_ccid, sk, skb);
+
+		/*
+		 * FIXME: check ECN to see if we should use
+		 * DCCP_ACKPKTS_STATE_ECN_MARKED
+		 */
+		if (dp->dccps_options.dccpo_send_ack_vector) {
+			if (dccp_ackpkts_add(dp->dccps_hc_rx_ackpkts,
+					     DCCP_SKB_CB(skb)->dccpd_seq,
+					     DCCP_ACKPKTS_STATE_RECEIVED))
+				goto discard;
+			/*
+			 * FIXME: this activation is probably wrong, have to study more
+			 * TCP delack machinery and how it fits into DCCP draft, but
+			 * for now it kinda "works" 8)
+			 */
+			if (dp->dccps_hc_rx_ackpkts->dccpap_ack_seqno == DCCP_MAX_SEQNO + 1 &&
+			    !inet_csk_ack_scheduled(sk)) {
+				inet_csk_schedule_ack(sk);
+				inet_csk_reset_xmit_timer(sk, ICSK_TIME_DACK, TCP_DELACK_MIN, TCP_RTO_MAX);
+			}
+		}
+	}
+
+	/*
+	 *  Step 9: Process Reset
+	 *	If P.type == Reset,
+	 *		Tear down connection
+	 *		S.state := TIMEWAIT
+	 *		Set TIMEWAIT timer
+	 *		Drop packet and return
+	*/
+	if (dh->dccph_type == DCCP_PKT_RESET) {
+		/* Queue the equivalent of TCP fin so that dccp_recvmsg exits the loop */
+		dccp_fin(sk, skb);
+		dccp_time_wait(sk, DCCP_TIME_WAIT, 0);
+		return 0;
+		/*
+		 *   Step 7: Check for unexpected packet types
+		 *      If (S.is_server and P.type == CloseReq)
+		 *	    or (S.is_server and P.type == Response)
+		 *	    or (S.is_client and P.type == Request)
+		 *	    or (S.state == RESPOND and P.type == Data),
+		 *	  Send Sync packet acknowledging P.seqno
+		 *	  Drop packet and return
+		 */
+	} else if ((dp->dccps_role != DCCP_ROLE_CLIENT &&
+		    (dh->dccph_type == DCCP_PKT_RESPONSE || dh->dccph_type == DCCP_PKT_CLOSEREQ)) ||
+		    (dp->dccps_role == DCCP_ROLE_CLIENT &&
+		     dh->dccph_type == DCCP_PKT_REQUEST) ||
+		    (sk->sk_state == DCCP_RESPOND && dh->dccph_type == DCCP_PKT_DATA)) {
+		dccp_send_sync(sk, DCCP_SKB_CB(skb)->dccpd_seq);
+		goto discard;
+	}
+
+	switch (sk->sk_state) {
+	case DCCP_CLOSED:
+		return 1;
+
+	case DCCP_LISTEN:
+		if (dh->dccph_type == DCCP_PKT_ACK ||
+		    dh->dccph_type == DCCP_PKT_DATAACK)
+			return 1;
+
+		if (dh->dccph_type == DCCP_PKT_RESET)
+			goto discard;
+
+		if (dh->dccph_type == DCCP_PKT_REQUEST) {
+			if (dccp_v4_conn_request(sk, skb) < 0)
+				return 1;
+
+			/* FIXME: do congestion control initialization */
+			goto discard;
+		}
+		goto discard;
+
+	case DCCP_REQUESTING:
+		/* FIXME: do congestion control initialization */
+
+		queued = dccp_rcv_request_sent_state_process(sk, skb, dh, len);
+		if (queued >= 0)
+			return queued;
+
+		__kfree_skb(skb);
+		return 0;
+
+	case DCCP_RESPOND:
+	case DCCP_PARTOPEN:
+		queued = dccp_rcv_respond_partopen_state_process(sk, skb, dh, len);
+		break;
+	}
+
+	if (dh->dccph_type == DCCP_PKT_ACK || dh->dccph_type == DCCP_PKT_DATAACK) {
+		switch (old_state) {
+		case DCCP_PARTOPEN:
+			sk->sk_state_change(sk);
+			sk_wake_async(sk, 0, POLL_OUT);
+			break;
+		}
+	}
+
+	if (!queued) { 
+discard:
+		__kfree_skb(skb);
+	}
+	return 0;
+}

net/dccp/minisocks.c

@@ -0,0 +1,199 @@
+/*
+ *  net/dccp/minisocks.c
+ *
+ *  An implementation of the DCCP protocol
+ *  Arnaldo Carvalho de Melo <acme@conectiva.com.br>
+ *
+ *	This program is free software; you can redistribute it and/or
+ *	modify it under the terms of the GNU General Public License
+ *	as published by the Free Software Foundation; either version
+ *	2 of the License, or (at your option) any later version.
+ */
+
+#include <linux/config.h>
+#include <linux/dccp.h>
+#include <linux/skbuff.h>
+#include <linux/timer.h>
+
+#include <net/sock.h>
+#include <net/xfrm.h>
+#include <net/inet_timewait_sock.h>
+
+#include "ccid.h"
+#include "dccp.h"
+
+void dccp_time_wait(struct sock *sk, int state, int timeo)
+{
+	/* FIXME: Implement */
+	dccp_pr_debug("Want to help? Start here\n");
+	dccp_set_state(sk, state);
+}
+
+/* This is for handling early-kills of TIME_WAIT sockets. */
+void dccp_tw_deschedule(struct inet_timewait_sock *tw)
+{
+	dccp_pr_debug("Want to help? Start here\n");
+	__inet_twsk_kill(tw, &dccp_hashinfo);
+}
+
+struct sock *dccp_create_openreq_child(struct sock *sk,
+				       const struct request_sock *req,
+				       const struct sk_buff *skb)
+{
+	/*
+	 * Step 3: Process LISTEN state
+	 *
+	 * // Generate a new socket and switch to that socket
+	 * Set S := new socket for this port pair
+	 */
+	struct sock *newsk = inet_csk_clone(sk, req, GFP_ATOMIC);
+
+	if (newsk != NULL) {
+		const struct dccp_request_sock *dreq = dccp_rsk(req);
+		struct inet_connection_sock *newicsk = inet_csk(sk);
+		struct dccp_sock *newdp = dccp_sk(newsk);
+
+		newdp->dccps_hc_rx_ackpkts = NULL;
+		newdp->dccps_role = DCCP_ROLE_SERVER;
+		newicsk->icsk_rto = TCP_TIMEOUT_INIT;
+
+		if (newdp->dccps_options.dccpo_send_ack_vector) {
+			newdp->dccps_hc_rx_ackpkts = dccp_ackpkts_alloc(DCCP_MAX_ACK_VECTOR_LEN,
+									GFP_ATOMIC);
+			/*
+			 * XXX: We're using the same CCIDs set on the parent, i.e. sk_clone
+			 * copied the master sock and left the CCID pointers for this child,
+			 * that is why we do the __ccid_get calls.
+			 */
+			if (unlikely(newdp->dccps_hc_rx_ackpkts == NULL))
+				goto out_free;
+		}
+
+		if (unlikely(ccid_hc_rx_init(newdp->dccps_hc_rx_ccid, newsk) != 0 ||
+			     ccid_hc_tx_init(newdp->dccps_hc_tx_ccid, newsk) != 0)) {
+			dccp_ackpkts_free(newdp->dccps_hc_rx_ackpkts);
+			ccid_hc_rx_exit(newdp->dccps_hc_rx_ccid, newsk);
+			ccid_hc_tx_exit(newdp->dccps_hc_tx_ccid, newsk);
+out_free:
+			/* It is still raw copy of parent, so invalidate
+			 * destructor and make plain sk_free() */
+			newsk->sk_destruct = NULL;
+			sk_free(newsk);
+			return NULL;
+		}
+
+		__ccid_get(newdp->dccps_hc_rx_ccid);
+		__ccid_get(newdp->dccps_hc_tx_ccid);
+
+		/*
+		 * Step 3: Process LISTEN state
+		 *
+		 *	Choose S.ISS (initial seqno) or set from Init Cookie
+		 *	Set S.ISR, S.GSR, S.SWL, S.SWH from packet or Init Cookie
+		 */
+
+		/* See dccp_v4_conn_request */
+		newdp->dccps_options.dccpo_sequence_window = req->rcv_wnd;
+
+		newdp->dccps_gar = newdp->dccps_isr = dreq->dreq_isr;
+		dccp_update_gsr(newsk, dreq->dreq_isr);
+
+		newdp->dccps_iss = dreq->dreq_iss;
+		dccp_update_gss(newsk, dreq->dreq_iss);
+
+		dccp_init_xmit_timers(newsk);
+
+		DCCP_INC_STATS_BH(DCCP_MIB_PASSIVEOPENS);
+	}
+	return newsk;
+}
+
+/* 
+ * Process an incoming packet for RESPOND sockets represented
+ * as an request_sock.
+ */
+struct sock *dccp_check_req(struct sock *sk, struct sk_buff *skb,
+			    struct request_sock *req,
+			    struct request_sock **prev)
+{
+	struct sock *child = NULL;
+
+	/* Check for retransmitted REQUEST */
+	if (dccp_hdr(skb)->dccph_type == DCCP_PKT_REQUEST) {
+		if (after48(DCCP_SKB_CB(skb)->dccpd_seq, dccp_rsk(req)->dreq_isr)) {
+			struct dccp_request_sock *dreq = dccp_rsk(req);
+
+			dccp_pr_debug("Retransmitted REQUEST\n");
+			/* Send another RESPONSE packet */
+			dccp_set_seqno(&dreq->dreq_iss, dreq->dreq_iss + 1);
+			dccp_set_seqno(&dreq->dreq_isr, DCCP_SKB_CB(skb)->dccpd_seq);
+			req->rsk_ops->rtx_syn_ack(sk, req, NULL);
+		}
+		/* Network Duplicate, discard packet */
+		return NULL;
+	}
+
+	DCCP_SKB_CB(skb)->dccpd_reset_code = DCCP_RESET_CODE_PACKET_ERROR;
+
+	if (dccp_hdr(skb)->dccph_type != DCCP_PKT_ACK &&
+	    dccp_hdr(skb)->dccph_type != DCCP_PKT_DATAACK)
+		goto drop;
+
+	/* Invalid ACK */
+	if (DCCP_SKB_CB(skb)->dccpd_ack_seq != dccp_rsk(req)->dreq_iss) {
+		dccp_pr_debug("Invalid ACK number: ack_seq=%llu, dreq_iss=%llu\n",
+			      DCCP_SKB_CB(skb)->dccpd_ack_seq, dccp_rsk(req)->dreq_iss);
+		goto drop;
+	}
+
+	child = dccp_v4_request_recv_sock(sk, skb, req, NULL);
+	if (child == NULL)
+		goto listen_overflow;
+
+	/* FIXME: deal with options */
+
+	inet_csk_reqsk_queue_unlink(sk, req, prev);
+	inet_csk_reqsk_queue_removed(sk, req);
+	inet_csk_reqsk_queue_add(sk, req, child);
+out:
+	return child;
+listen_overflow:
+	dccp_pr_debug("listen_overflow!\n");
+	DCCP_SKB_CB(skb)->dccpd_reset_code = DCCP_RESET_CODE_TOO_BUSY;
+drop:
+	if (dccp_hdr(skb)->dccph_type != DCCP_PKT_RESET)
+		req->rsk_ops->send_reset(skb);
+
+	inet_csk_reqsk_queue_drop(sk, req, prev);
+	goto out;
+}
+
+/*
+ *  Queue segment on the new socket if the new socket is active,
+ *  otherwise we just shortcircuit this and continue with
+ *  the new socket.
+ */
+int dccp_child_process(struct sock *parent, struct sock *child,
+		       struct sk_buff *skb)
+{
+	int ret = 0;
+	const int state = child->sk_state;
+
+	if (!sock_owned_by_user(child)) {
+		ret = dccp_rcv_state_process(child, skb, dccp_hdr(skb), skb->len);
+
+		/* Wakeup parent, send SIGIO */
+		if (state == DCCP_RESPOND && child->sk_state != state)
+			parent->sk_data_ready(parent, 0);
+	} else {
+		/* Alas, it is possible again, because we do lookup
+		 * in main socket hash table and lock on listening
+		 * socket does not protect us more.
+		 */
+		sk_add_backlog(child, skb);
+	}
+
+	bh_unlock_sock(child);
+	sock_put(child);
+	return ret;
+}

net/dccp/options.c

@@ -0,0 +1,763 @@
+/*
+ *  net/dccp/options.c
+ *
+ *  An implementation of the DCCP protocol
+ *  Aristeu Sergio Rozanski Filho <aris@cathedrallabs.org>
+ *  Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
+ *
+ *      This program is free software; you can redistribute it and/or
+ *      modify it under the terms of the GNU General Public License
+ *      as published by the Free Software Foundation; either version
+ *      2 of the License, or (at your option) any later version.
+ */
+#include <linux/config.h>
+#include <linux/dccp.h>
+#include <linux/module.h>
+#include <linux/types.h>
+#include <linux/kernel.h>
+#include <linux/skbuff.h>
+
+#include "ccid.h"
+#include "dccp.h"
+
+static void dccp_ackpkts_check_rcv_ackvector(struct dccp_ackpkts *ap,
+					     struct sock *sk,
+					     const u64 ackno,
+					     const unsigned char len,
+					     const unsigned char *vector);
+
+/* stores the default values for new connection. may be changed with sysctl */
+static const struct dccp_options dccpo_default_values = {
+	.dccpo_sequence_window	  = DCCPF_INITIAL_SEQUENCE_WINDOW,
+	.dccpo_ccid		  = DCCPF_INITIAL_CCID,
+	.dccpo_send_ack_vector	  = DCCPF_INITIAL_SEND_ACK_VECTOR,
+	.dccpo_send_ndp_count	  = DCCPF_INITIAL_SEND_NDP_COUNT,
+};
+
+void dccp_options_init(struct dccp_options *dccpo)
+{
+	memcpy(dccpo, &dccpo_default_values, sizeof(*dccpo));
+}
+
+static u32 dccp_decode_value_var(const unsigned char *bf, const u8 len)
+{
+	u32 value = 0;
+
+	if (len > 3)
+		value += *bf++ << 24;
+	if (len > 2)
+		value += *bf++ << 16;
+	if (len > 1)
+		value += *bf++ << 8;
+	if (len > 0)
+		value += *bf;
+
+	return value;
+}
+
+int dccp_parse_options(struct sock *sk, struct sk_buff *skb)
+{
+	struct dccp_sock *dp = dccp_sk(sk);
+#ifdef DCCP_DEBUG
+	const char *debug_prefix = dp->dccps_role == DCCP_ROLE_CLIENT ? "CLIENT rx opt: " :
+									"server rx opt: ";
+#endif
+	const struct dccp_hdr *dh = dccp_hdr(skb);
+	const u8 pkt_type = DCCP_SKB_CB(skb)->dccpd_type;
+	unsigned char *options = (unsigned char *)dh + dccp_hdr_len(skb);
+	unsigned char *opt_ptr = options;
+	const unsigned char *opt_end = (unsigned char *)dh + (dh->dccph_doff * 4);
+	struct dccp_options_received *opt_recv = &dp->dccps_options_received;
+	unsigned char opt, len;
+	unsigned char *value;
+
+	memset(opt_recv, 0, sizeof(*opt_recv));
+
+	while (opt_ptr != opt_end) {
+		opt   = *opt_ptr++;
+		len   = 0;
+		value = NULL;
+
+		/* Check if this isn't a single byte option */
+		if (opt > DCCPO_MAX_RESERVED) {
+			if (opt_ptr == opt_end)
+				goto out_invalid_option;
+
+			len = *opt_ptr++;
+			if (len < 3)
+				goto out_invalid_option;
+			/*
+			 * Remove the type and len fields, leaving
+			 * just the value size
+			 */
+			len	-= 2;
+			value	= opt_ptr;
+			opt_ptr += len;
+
+			if (opt_ptr > opt_end)
+				goto out_invalid_option;
+		}
+
+		switch (opt) {
+		case DCCPO_PADDING:
+			break;
+		case DCCPO_NDP_COUNT:
+			if (len > 3)
+				goto out_invalid_option;
+
+			opt_recv->dccpor_ndp = dccp_decode_value_var(value, len);
+			dccp_pr_debug("%sNDP count=%d\n", debug_prefix, opt_recv->dccpor_ndp);
+			break;
+		case DCCPO_ACK_VECTOR_0:
+			if (len > DCCP_MAX_ACK_VECTOR_LEN)
+				goto out_invalid_option;
+
+			if (pkt_type == DCCP_PKT_DATA)
+				continue;
+
+			opt_recv->dccpor_ack_vector_len = len;
+			opt_recv->dccpor_ack_vector_idx = value - options;
+
+			dccp_pr_debug("%sACK vector 0, len=%d, ack_ackno=%llu\n",
+				      debug_prefix, len, DCCP_SKB_CB(skb)->dccpd_ack_seq);
+			dccp_ackvector_print(DCCP_SKB_CB(skb)->dccpd_ack_seq,
+					     value, len);
+			dccp_ackpkts_check_rcv_ackvector(dp->dccps_hc_rx_ackpkts, sk,
+							 DCCP_SKB_CB(skb)->dccpd_ack_seq,
+							 len, value);
+			break;
+		case DCCPO_TIMESTAMP:
+			if (len != 4)
+				goto out_invalid_option;
+
+			opt_recv->dccpor_timestamp = ntohl(*(u32 *)value);
+
+			dp->dccps_timestamp_echo = opt_recv->dccpor_timestamp;
+			dp->dccps_timestamp_time = jiffies;
+
+			dccp_pr_debug("%sTIMESTAMP=%u, ackno=%llu\n",
+				      debug_prefix, opt_recv->dccpor_timestamp,
+				      DCCP_SKB_CB(skb)->dccpd_ack_seq);
+			break;
+		case DCCPO_TIMESTAMP_ECHO:
+			if (len < 4 || len > 8)
+				goto out_invalid_option;
+
+			opt_recv->dccpor_timestamp_echo = ntohl(*(u32 *)value);
+
+			dccp_pr_debug("%sTIMESTAMP_ECHO=%u, len=%d, ackno=%llu, diff=%u\n",
+				      debug_prefix, opt_recv->dccpor_timestamp_echo,
+				      len + 2, DCCP_SKB_CB(skb)->dccpd_ack_seq,
+				      tcp_time_stamp - opt_recv->dccpor_timestamp_echo);
+
+			opt_recv->dccpor_elapsed_time = dccp_decode_value_var(value + 4, len - 4);
+			dccp_pr_debug("%sTIMESTAMP_ECHO ELAPSED_TIME=%d\n", debug_prefix,
+				      opt_recv->dccpor_elapsed_time);
+			break;
+		case DCCPO_ELAPSED_TIME:
+			if (len > 4)
+				goto out_invalid_option;
+
+			if (pkt_type == DCCP_PKT_DATA)
+				continue;
+			opt_recv->dccpor_elapsed_time = dccp_decode_value_var(value, len);
+			dccp_pr_debug("%sELAPSED_TIME=%d\n", debug_prefix,
+				      opt_recv->dccpor_elapsed_time);
+			break;
+			/*
+			 * From draft-ietf-dccp-spec-11.txt:
+			 *
+			 *	Option numbers 128 through 191 are for options sent from the HC-
+			 *	Sender to the HC-Receiver; option numbers 192 through 255 are for
+			 *	options sent from the HC-Receiver to the HC-Sender.
+			 */
+		case 128 ... 191: {
+			const u16 idx = value - options;
+
+			if (ccid_hc_rx_parse_options(dp->dccps_hc_rx_ccid, sk, opt, len, idx, value) != 0)
+				goto out_invalid_option;
+		}
+			break;
+		case 192 ... 255: {
+			const u16 idx = value - options;
+
+			if (ccid_hc_tx_parse_options(dp->dccps_hc_tx_ccid, sk, opt, len, idx, value) != 0)
+				goto out_invalid_option;
+		}
+			break;
+		default:
+			pr_info("DCCP(%p): option %d(len=%d) not implemented, ignoring\n",
+				sk, opt, len);
+			break;
+	        }
+	}
+
+	return 0;
+
+out_invalid_option:
+	DCCP_INC_STATS_BH(DCCP_MIB_INVALIDOPT);
+	DCCP_SKB_CB(skb)->dccpd_reset_code = DCCP_RESET_CODE_OPTION_ERROR;
+	pr_info("DCCP(%p): invalid option %d, len=%d\n", sk, opt, len);
+	return -1;
+}
+
+static void dccp_encode_value_var(const u32 value, unsigned char *to,
+				  const unsigned int len)
+{
+	if (len > 3)
+		*to++ = (value & 0xFF000000) >> 24;
+	if (len > 2)
+		*to++ = (value & 0xFF0000) >> 16;
+	if (len > 1)
+		*to++ = (value & 0xFF00) >> 8;
+	if (len > 0)
+		*to++ = (value & 0xFF);
+}
+
+static inline int dccp_ndp_len(const int ndp)
+{
+	return likely(ndp <= 0xFF) ? 1 : ndp <= 0xFFFF ? 2 : 3;
+}
+
+void dccp_insert_option(struct sock *sk, struct sk_buff *skb,
+			const unsigned char option,
+			const void *value, const unsigned char len)
+{
+	unsigned char *to;
+
+	if (DCCP_SKB_CB(skb)->dccpd_opt_len + len + 2 > DCCP_MAX_OPT_LEN) {
+		LIMIT_NETDEBUG(pr_info("DCCP: packet too small to insert %d option!\n", option));
+		return;
+	}
+
+	DCCP_SKB_CB(skb)->dccpd_opt_len += len + 2;
+
+	to    = skb_push(skb, len + 2);
+	*to++ = option;
+	*to++ = len + 2;
+
+	memcpy(to, value, len);
+}
+
+EXPORT_SYMBOL_GPL(dccp_insert_option);
+
+static void dccp_insert_option_ndp(struct sock *sk, struct sk_buff *skb)
+{
+	struct dccp_sock *dp = dccp_sk(sk);
+	int ndp = dp->dccps_ndp_count;
+
+	if (dccp_non_data_packet(skb))
+		++dp->dccps_ndp_count;
+	else
+		dp->dccps_ndp_count = 0;
+
+	if (ndp > 0) {
+		unsigned char *ptr;
+		const int ndp_len = dccp_ndp_len(ndp);
+		const int len = ndp_len + 2;
+
+		if (DCCP_SKB_CB(skb)->dccpd_opt_len + len > DCCP_MAX_OPT_LEN)
+			return;
+
+		DCCP_SKB_CB(skb)->dccpd_opt_len += len;
+
+		ptr = skb_push(skb, len);
+		*ptr++ = DCCPO_NDP_COUNT;
+		*ptr++ = len;
+		dccp_encode_value_var(ndp, ptr, ndp_len);
+	}
+}
+
+static inline int dccp_elapsed_time_len(const u32 elapsed_time)
+{
+	return elapsed_time == 0 ? 0 :
+	       elapsed_time <= 0xFF ? 1 :
+	       elapsed_time <= 0xFFFF ? 2 :
+	       elapsed_time <= 0xFFFFFF ? 3 : 4;
+}
+
+void dccp_insert_option_elapsed_time(struct sock *sk,
+				     struct sk_buff *skb,
+				     u32 elapsed_time)
+{
+#ifdef DCCP_DEBUG
+	struct dccp_sock *dp = dccp_sk(sk);
+	const char *debug_prefix = dp->dccps_role == DCCP_ROLE_CLIENT ? "CLIENT TX opt: " :
+									"server TX opt: ";
+#endif
+	const int elapsed_time_len = dccp_elapsed_time_len(elapsed_time);
+	const int len = 2 + elapsed_time_len;
+	unsigned char *to;
+
+	/* If elapsed_time == 0... */
+	if (elapsed_time_len == 2)
+		return;
+
+	if (DCCP_SKB_CB(skb)->dccpd_opt_len + len > DCCP_MAX_OPT_LEN) {
+		LIMIT_NETDEBUG(pr_info("DCCP: packet too small to insert elapsed time!\n"));
+		return;
+	}
+
+	DCCP_SKB_CB(skb)->dccpd_opt_len += len;
+
+	to    = skb_push(skb, len);
+	*to++ = DCCPO_ELAPSED_TIME;
+	*to++ = len;
+
+	dccp_encode_value_var(elapsed_time, to, elapsed_time_len);
+
+	dccp_pr_debug("%sELAPSED_TIME=%u, len=%d, seqno=%llu\n",
+		      debug_prefix, elapsed_time,
+		      len, DCCP_SKB_CB(skb)->dccpd_seq);
+}
+
+EXPORT_SYMBOL(dccp_insert_option_elapsed_time);
+
+static void dccp_insert_option_ack_vector(struct sock *sk, struct sk_buff *skb)
+{
+	struct dccp_sock *dp = dccp_sk(sk);
+#ifdef DCCP_DEBUG
+	const char *debug_prefix = dp->dccps_role == DCCP_ROLE_CLIENT ? "CLIENT TX opt: " :
+									"server TX opt: ";
+#endif
+	struct dccp_ackpkts *ap = dp->dccps_hc_rx_ackpkts;
+	int len = ap->dccpap_buf_vector_len + 2;
+	const u32 elapsed_time = jiffies_to_usecs(jiffies - ap->dccpap_time) / 10;
+	unsigned char *to, *from;
+
+	if (elapsed_time != 0)
+		dccp_insert_option_elapsed_time(sk, skb, elapsed_time);
+
+	if (DCCP_SKB_CB(skb)->dccpd_opt_len + len > DCCP_MAX_OPT_LEN) {
+		LIMIT_NETDEBUG(pr_info("DCCP: packet too small to insert ACK Vector!\n"));
+		return;
+	}
+
+	/*
+	 * XXX: now we have just one ack vector sent record, so
+	 * we have to wait for it to be cleared.
+	 *
+	 * Of course this is not acceptable, but this is just for
+	 * basic testing now.
+	 */
+	if (ap->dccpap_ack_seqno != DCCP_MAX_SEQNO + 1)
+		return;
+
+	DCCP_SKB_CB(skb)->dccpd_opt_len += len;
+
+	to    = skb_push(skb, len);
+	*to++ = DCCPO_ACK_VECTOR_0;
+	*to++ = len;
+
+	len  = ap->dccpap_buf_vector_len;
+	from = ap->dccpap_buf + ap->dccpap_buf_head;
+
+	/* Check if buf_head wraps */
+	if (ap->dccpap_buf_head + len > ap->dccpap_buf_len) {
+		const unsigned int tailsize = ap->dccpap_buf_len - ap->dccpap_buf_head;
+
+		memcpy(to, from, tailsize);
+		to   += tailsize;
+		len  -= tailsize;
+		from = ap->dccpap_buf;
+	}
+
+	memcpy(to, from, len);
+	/*
+	 *	From draft-ietf-dccp-spec-11.txt:
+	 *
+	 *	For each acknowledgement it sends, the HC-Receiver will add an
+	 *	acknowledgement record.  ack_seqno will equal the HC-Receiver
+	 *	sequence number it used for the ack packet; ack_ptr will equal
+	 *	buf_head; ack_ackno will equal buf_ackno; and ack_nonce will equal
+	 *	buf_nonce.
+	 *
+	 * This implemention uses just one ack record for now.
+	 */
+	ap->dccpap_ack_seqno	  = DCCP_SKB_CB(skb)->dccpd_seq;
+	ap->dccpap_ack_ptr	  = ap->dccpap_buf_head;
+	ap->dccpap_ack_ackno	  = ap->dccpap_buf_ackno;
+	ap->dccpap_ack_nonce	  = ap->dccpap_buf_nonce;
+	ap->dccpap_ack_vector_len = ap->dccpap_buf_vector_len;
+
+	dccp_pr_debug("%sACK Vector 0, len=%d, ack_seqno=%llu, ack_ackno=%llu\n",
+		      debug_prefix, ap->dccpap_ack_vector_len,
+		      ap->dccpap_ack_seqno, ap->dccpap_ack_ackno);
+}
+
+static inline void dccp_insert_option_timestamp(struct sock *sk, struct sk_buff *skb)
+{
+	const u32 now = htonl(tcp_time_stamp);
+	dccp_insert_option(sk, skb, DCCPO_TIMESTAMP, &now, sizeof(now));
+}
+
+static void dccp_insert_option_timestamp_echo(struct sock *sk, struct sk_buff *skb)
+{
+	struct dccp_sock *dp = dccp_sk(sk);
+#ifdef DCCP_DEBUG
+	const char *debug_prefix = dp->dccps_role == DCCP_ROLE_CLIENT ? "CLIENT TX opt: " :
+									"server TX opt: ";
+#endif
+	u32 tstamp_echo;
+	const u32 elapsed_time = jiffies_to_usecs(jiffies - dp->dccps_timestamp_time) / 10;
+	const int elapsed_time_len = dccp_elapsed_time_len(elapsed_time);
+	const int len = 6 + elapsed_time_len;
+	unsigned char *to;
+
+	if (DCCP_SKB_CB(skb)->dccpd_opt_len + len > DCCP_MAX_OPT_LEN) {
+		LIMIT_NETDEBUG(pr_info("DCCP: packet too small to insert timestamp echo!\n"));
+		return;
+	}
+
+	DCCP_SKB_CB(skb)->dccpd_opt_len += len;
+
+	to    = skb_push(skb, len);
+	*to++ = DCCPO_TIMESTAMP_ECHO;
+	*to++ = len;
+
+	tstamp_echo = htonl(dp->dccps_timestamp_echo);
+	memcpy(to, &tstamp_echo, 4);
+	to += 4;
+	dccp_encode_value_var(elapsed_time, to, elapsed_time_len);
+
+	dccp_pr_debug("%sTIMESTAMP_ECHO=%u, len=%d, seqno=%llu\n",
+		      debug_prefix, dp->dccps_timestamp_echo,
+		      len, DCCP_SKB_CB(skb)->dccpd_seq);
+
+	dp->dccps_timestamp_echo = 0;
+	dp->dccps_timestamp_time = 0;
+}
+
+void dccp_insert_options(struct sock *sk, struct sk_buff *skb)
+{
+	struct dccp_sock *dp = dccp_sk(sk);
+
+	DCCP_SKB_CB(skb)->dccpd_opt_len = 0;
+
+	if (dp->dccps_options.dccpo_send_ndp_count)
+		dccp_insert_option_ndp(sk, skb);
+
+	if (!dccp_packet_without_ack(skb)) {
+		if (dp->dccps_options.dccpo_send_ack_vector &&
+		    dp->dccps_hc_rx_ackpkts->dccpap_buf_ackno != DCCP_MAX_SEQNO + 1)
+			dccp_insert_option_ack_vector(sk, skb);
+
+		dccp_insert_option_timestamp(sk, skb);
+		if (dp->dccps_timestamp_echo != 0)
+			dccp_insert_option_timestamp_echo(sk, skb);
+	}
+
+	ccid_hc_rx_insert_options(dp->dccps_hc_rx_ccid, sk, skb);
+	ccid_hc_tx_insert_options(dp->dccps_hc_tx_ccid, sk, skb);
+
+	/* XXX: insert other options when appropriate */
+
+	if (DCCP_SKB_CB(skb)->dccpd_opt_len != 0) {
+		/* The length of all options has to be a multiple of 4 */
+		int padding = DCCP_SKB_CB(skb)->dccpd_opt_len % 4;
+
+		if (padding != 0) {
+			padding = 4 - padding;
+			memset(skb_push(skb, padding), 0, padding);
+			DCCP_SKB_CB(skb)->dccpd_opt_len += padding;
+		}
+	}
+}
+
+struct dccp_ackpkts *dccp_ackpkts_alloc(unsigned int len, int priority)
+{
+	struct dccp_ackpkts *ap = kmalloc(sizeof(*ap) + len, priority);
+
+	if (ap != NULL) {
+#ifdef DCCP_DEBUG
+		memset(ap->dccpap_buf, 0xFF, len);
+#endif
+		ap->dccpap_buf_len	  = len;
+		ap->dccpap_buf_head	  = ap->dccpap_buf_tail = ap->dccpap_buf_len - 1;
+		ap->dccpap_buf_ackno	  = ap->dccpap_ack_ackno = ap->dccpap_ack_seqno = DCCP_MAX_SEQNO + 1;
+		ap->dccpap_buf_nonce	  = ap->dccpap_buf_nonce = 0;
+		ap->dccpap_ack_ptr	  = 0;
+		ap->dccpap_time		  = 0;
+		ap->dccpap_buf_vector_len = ap->dccpap_ack_vector_len = 0;
+	}
+
+	return ap;
+}
+
+void dccp_ackpkts_free(struct dccp_ackpkts *ap)
+{
+	if (ap != NULL) {
+#ifdef DCCP_DEBUG
+		memset(ap, 0xFF, sizeof(*ap) + ap->dccpap_buf_len);
+#endif
+		kfree(ap);
+	}
+}
+
+static inline u8 dccp_ackpkts_state(const struct dccp_ackpkts *ap,
+				    const unsigned int index)
+{
+	return ap->dccpap_buf[index] & DCCP_ACKPKTS_STATE_MASK;
+}
+
+static inline u8 dccp_ackpkts_len(const struct dccp_ackpkts *ap,
+				  const unsigned int index)
+{
+	return ap->dccpap_buf[index] & DCCP_ACKPKTS_LEN_MASK;
+}
+
+/*
+ * If several packets are missing, the HC-Receiver may prefer to enter multiple
+ * bytes with run length 0, rather than a single byte with a larger run length;
+ * this simplifies table updates if one of the missing packets arrives.
+ */
+static inline int dccp_ackpkts_set_buf_head_state(struct dccp_ackpkts *ap,
+						  const unsigned int packets,
+						  const unsigned char state)
+{
+	unsigned int gap;
+	signed long new_head;
+
+	if (ap->dccpap_buf_vector_len + packets > ap->dccpap_buf_len)
+		return -ENOBUFS;
+
+	gap	 = packets - 1;
+	new_head = ap->dccpap_buf_head - packets;
+
+	if (new_head < 0) {
+		if (gap > 0) {
+			memset(ap->dccpap_buf, DCCP_ACKPKTS_STATE_NOT_RECEIVED,
+			       gap + new_head + 1);
+			gap = -new_head;
+		}
+		new_head += ap->dccpap_buf_len;
+	} 
+
+	ap->dccpap_buf_head = new_head;
+
+	if (gap > 0)
+		memset(ap->dccpap_buf + ap->dccpap_buf_head + 1,
+		       DCCP_ACKPKTS_STATE_NOT_RECEIVED, gap);
+
+	ap->dccpap_buf[ap->dccpap_buf_head] = state;
+	ap->dccpap_buf_vector_len += packets;
+	return 0;
+}
+
+/*
+ * Implements the draft-ietf-dccp-spec-11.txt Appendix A
+ */
+int dccp_ackpkts_add(struct dccp_ackpkts *ap, u64 ackno, u8 state)
+{
+	/*
+	 * Check at the right places if the buffer is full, if it is, tell the
+	 * caller to start dropping packets till the HC-Sender acks our ACK
+	 * vectors, when we will free up space in dccpap_buf.
+	 *
+	 * We may well decide to do buffer compression, etc, but for now lets
+	 * just drop.
+	 *
+	 * From Appendix A:
+	 *
+	 *	Of course, the circular buffer may overflow, either when the HC-
+	 *	Sender is sending data at a very high rate, when the HC-Receiver's
+	 *	acknowledgements are not reaching the HC-Sender, or when the HC-
+	 *	Sender is forgetting to acknowledge those acks (so the HC-Receiver
+	 *	is unable to clean up old state).  In this case, the HC-Receiver
+	 *	should either compress the buffer (by increasing run lengths when
+	 *	possible), transfer its state to a larger buffer, or, as a last
+	 *	resort, drop all received packets, without processing them
+	 *	whatsoever, until its buffer shrinks again.
+	 */
+
+	/* See if this is the first ackno being inserted */
+	if (ap->dccpap_buf_vector_len == 0) {
+		ap->dccpap_buf[ap->dccpap_buf_head] = state;
+		ap->dccpap_buf_vector_len = 1;
+	} else if (after48(ackno, ap->dccpap_buf_ackno)) {
+		const u64 delta = dccp_delta_seqno(ap->dccpap_buf_ackno, ackno);
+
+		/*
+		 * Look if the state of this packet is the same as the previous ackno
+		 * and if so if we can bump the head len.
+		 */
+		if (delta == 1 &&
+		    dccp_ackpkts_state(ap, ap->dccpap_buf_head) == state &&
+		    dccp_ackpkts_len(ap, ap->dccpap_buf_head) < DCCP_ACKPKTS_LEN_MASK)
+			ap->dccpap_buf[ap->dccpap_buf_head]++;
+		else if (dccp_ackpkts_set_buf_head_state(ap, delta, state))
+			return -ENOBUFS;
+	} else {
+		/*
+		 * A.1.2.  Old Packets
+		 *
+		 *	When a packet with Sequence Number S arrives, and S <= buf_ackno,
+		 *	the HC-Receiver will scan the table for the byte corresponding to S.
+		 *	(Indexing structures could reduce the complexity of this scan.)
+		 */
+		u64 delta = dccp_delta_seqno(ackno, ap->dccpap_buf_ackno);
+		unsigned int index = ap->dccpap_buf_head;
+
+		while (1) {
+			const u8 len = dccp_ackpkts_len(ap, index);
+			const u8 state = dccp_ackpkts_state(ap, index);
+			/*
+			 * valid packets not yet in dccpap_buf have a reserved entry, with
+			 * a len equal to 0
+			 */
+			if (state == DCCP_ACKPKTS_STATE_NOT_RECEIVED &&
+			    len == 0 && delta == 0) { /* Found our reserved seat! */
+				dccp_pr_debug("Found %llu reserved seat!\n", ackno);
+				ap->dccpap_buf[index] = state;
+				goto out;
+			}
+			/* len == 0 means one packet */
+			if (delta < len + 1)
+				goto out_duplicate;
+
+			delta -= len + 1;
+			if (++index == ap->dccpap_buf_len)
+				index = 0;
+		}
+	}
+
+	ap->dccpap_buf_ackno = ackno;
+	ap->dccpap_time = jiffies;
+out:
+	dccp_pr_debug("");
+	dccp_ackpkts_print(ap);
+	return 0;
+
+out_duplicate:
+	/* Duplicate packet */
+	dccp_pr_debug("Received a dup or already considered lost packet: %llu\n", ackno);
+	return -EILSEQ;
+}
+
+#ifdef DCCP_DEBUG
+void dccp_ackvector_print(const u64 ackno, const unsigned char *vector, int len)
+{
+	if (!dccp_debug)
+		return;
+
+	printk("ACK vector len=%d, ackno=%llu |", len, ackno);
+
+	while (len--) {
+		const u8 state = (*vector & DCCP_ACKPKTS_STATE_MASK) >> 6;
+		const u8 rl = (*vector & DCCP_ACKPKTS_LEN_MASK);
+
+		printk("%d,%d|", state, rl);
+		++vector;
+	}
+
+	printk("\n");
+}
+
+void dccp_ackpkts_print(const struct dccp_ackpkts *ap)
+{
+	dccp_ackvector_print(ap->dccpap_buf_ackno,
+			     ap->dccpap_buf + ap->dccpap_buf_head,
+			     ap->dccpap_buf_vector_len);
+}
+#endif
+
+static void dccp_ackpkts_trow_away_ack_record(struct dccp_ackpkts *ap)
+{
+	/*
+	 * As we're keeping track of the ack vector size
+	 * (dccpap_buf_vector_len) and the sent ack vector size
+	 * (dccpap_ack_vector_len) we don't need dccpap_buf_tail at all, but
+	 * keep this code here as in the future we'll implement a vector of ack
+	 * records, as suggested in draft-ietf-dccp-spec-11.txt Appendix A. -acme
+	 */
+#if 0
+	ap->dccpap_buf_tail = ap->dccpap_ack_ptr + 1;
+	if (ap->dccpap_buf_tail >= ap->dccpap_buf_len)
+		ap->dccpap_buf_tail -= ap->dccpap_buf_len;
+#endif
+	ap->dccpap_buf_vector_len -= ap->dccpap_ack_vector_len;
+}
+
+void dccp_ackpkts_check_rcv_ackno(struct dccp_ackpkts *ap, struct sock *sk,
+				 u64 ackno)
+{
+	/* Check if we actually sent an ACK vector */
+	if (ap->dccpap_ack_seqno == DCCP_MAX_SEQNO + 1)
+		return;
+
+	if (ackno == ap->dccpap_ack_seqno) {
+#ifdef DCCP_DEBUG
+		struct dccp_sock *dp = dccp_sk(sk);
+		const char *debug_prefix = dp->dccps_role == DCCP_ROLE_CLIENT ? "CLIENT rx ack: " :
+										"server rx ack: ";
+#endif
+		dccp_pr_debug("%sACK packet 0, len=%d, ack_seqno=%llu, ack_ackno=%llu, ACKED!\n",
+			      debug_prefix, 1,
+			      ap->dccpap_ack_seqno, ap->dccpap_ack_ackno);
+		dccp_ackpkts_trow_away_ack_record(ap);
+		ap->dccpap_ack_seqno = DCCP_MAX_SEQNO + 1;
+	}
+}
+
+static void dccp_ackpkts_check_rcv_ackvector(struct dccp_ackpkts *ap,
+					     struct sock *sk, u64 ackno,
+					     const unsigned char len,
+					     const unsigned char *vector)
+{
+	unsigned char i;
+
+	/* Check if we actually sent an ACK vector */
+	if (ap->dccpap_ack_seqno == DCCP_MAX_SEQNO + 1)
+		return;
+	/*
+	 * We're in the receiver half connection, so if the received an ACK vector
+	 * ackno (e.g. 50) before dccpap_ack_seqno (e.g. 52), we're not interested.
+	 *
+	 * Extra explanation with example:
+	 * 
+	 * if we received an ACK vector with ackno 50, it can only be acking
+	 * 50, 49, 48, etc, not 52 (the seqno for the ACK vector we sent).
+	 */
+	// dccp_pr_debug("is %llu < %llu? ", ackno, ap->dccpap_ack_seqno);
+	if (before48(ackno, ap->dccpap_ack_seqno)) {
+		// dccp_pr_debug_cat("yes\n");
+		return;
+	}
+	// dccp_pr_debug_cat("no\n");
+
+	i = len;
+	while (i--) {
+		const u8 rl = (*vector & DCCP_ACKPKTS_LEN_MASK);
+		u64 ackno_end_rl;
+
+		dccp_set_seqno(&ackno_end_rl, ackno - rl);
+
+		// dccp_pr_debug("is %llu <= %llu <= %llu? ", ackno_end_rl, ap->dccpap_ack_seqno, ackno);
+		if (between48(ap->dccpap_ack_seqno, ackno_end_rl, ackno)) {
+			const u8 state = (*vector & DCCP_ACKPKTS_STATE_MASK) >> 6;
+			// dccp_pr_debug_cat("yes\n");
+
+			if (state != DCCP_ACKPKTS_STATE_NOT_RECEIVED) {
+#ifdef DCCP_DEBUG
+				struct dccp_sock *dp = dccp_sk(sk);
+				const char *debug_prefix = dp->dccps_role == DCCP_ROLE_CLIENT ? "CLIENT rx ack: " :
+												"server rx ack: ";
+#endif
+				dccp_pr_debug("%sACK vector 0, len=%d, ack_seqno=%llu, ack_ackno=%llu, ACKED!\n",
+					      debug_prefix, len,
+					      ap->dccpap_ack_seqno, ap->dccpap_ack_ackno);
+				dccp_ackpkts_trow_away_ack_record(ap);
+			}
+			/*
+			 * If dccpap_ack_seqno was not received, no problem we'll
+			 * send another ACK vector.
+			 */
+			ap->dccpap_ack_seqno = DCCP_MAX_SEQNO + 1;
+			break;
+		}
+		// dccp_pr_debug_cat("no\n");
+
+		dccp_set_seqno(&ackno, ackno_end_rl - 1);
+		++vector;
+	}
+}

net/dccp/output.c

@@ -0,0 +1,406 @@
+/*
+ *  net/dccp/output.c
+ * 
+ *  An implementation of the DCCP protocol
+ *  Arnaldo Carvalho de Melo <acme@conectiva.com.br>
+ *
+ *	This program is free software; you can redistribute it and/or
+ *	modify it under the terms of the GNU General Public License
+ *	as published by the Free Software Foundation; either version
+ *	2 of the License, or (at your option) any later version.
+ */
+
+#include <linux/config.h>
+#include <linux/dccp.h>
+#include <linux/skbuff.h>
+
+#include <net/sock.h>
+
+#include "ccid.h"
+#include "dccp.h"
+
+static inline void dccp_event_ack_sent(struct sock *sk)
+{
+	inet_csk_clear_xmit_timer(sk, ICSK_TIME_DACK);
+}
+
+/*
+ * All SKB's seen here are completely headerless. It is our
+ * job to build the DCCP header, and pass the packet down to
+ * IP so it can do the same plus pass the packet off to the
+ * device.
+ */
+int dccp_transmit_skb(struct sock *sk, struct sk_buff *skb)
+{
+	if (likely(skb != NULL)) {
+		const struct inet_sock *inet = inet_sk(sk);
+		struct dccp_sock *dp = dccp_sk(sk);
+		struct dccp_skb_cb *dcb = DCCP_SKB_CB(skb);
+		struct dccp_hdr *dh;
+		/* XXX For now we're using only 48 bits sequence numbers */
+		const int dccp_header_size = sizeof(*dh) +
+					     sizeof(struct dccp_hdr_ext) +
+					     dccp_packet_hdr_len(dcb->dccpd_type);
+		int err, set_ack = 1;
+		u64 ackno = dp->dccps_gsr;
+
+		/*
+		 * FIXME: study DCCP_PKT_SYNC[ACK] to see what is the right thing 
+		 * to do here...
+		 */
+		dccp_inc_seqno(&dp->dccps_gss);
+
+		dcb->dccpd_seq = dp->dccps_gss;
+		dccp_insert_options(sk, skb);
+
+		switch (dcb->dccpd_type) {
+		case DCCP_PKT_DATA:
+			set_ack = 0;
+			break;
+		case DCCP_PKT_SYNC:
+		case DCCP_PKT_SYNCACK:
+			ackno = dcb->dccpd_seq;
+			break;
+		}
+		
+		skb->h.raw = skb_push(skb, dccp_header_size);
+		dh = dccp_hdr(skb);
+		/* Data packets are not cloned as they are never retransmitted */
+		if (skb_cloned(skb))
+			skb_set_owner_w(skb, sk);
+
+		/* Build DCCP header and checksum it. */
+		memset(dh, 0, dccp_header_size);
+		dh->dccph_type	= dcb->dccpd_type;
+		dh->dccph_sport	= inet->sport;
+		dh->dccph_dport	= inet->dport;
+		dh->dccph_doff	= (dccp_header_size + dcb->dccpd_opt_len) / 4;
+		dh->dccph_ccval	= dcb->dccpd_ccval;
+		/* XXX For now we're using only 48 bits sequence numbers */
+		dh->dccph_x	= 1;
+
+		dp->dccps_awh = dp->dccps_gss;
+		dccp_hdr_set_seq(dh, dp->dccps_gss);
+		if (set_ack)
+			dccp_hdr_set_ack(dccp_hdr_ack_bits(skb), ackno);
+
+		switch (dcb->dccpd_type) {
+		case DCCP_PKT_REQUEST:
+			dccp_hdr_request(skb)->dccph_req_service = dcb->dccpd_service;
+			break;
+		case DCCP_PKT_RESET:
+			dccp_hdr_reset(skb)->dccph_reset_code = dcb->dccpd_reset_code;
+			break;
+		}
+
+		dh->dccph_checksum = dccp_v4_checksum(skb);
+
+		if (dcb->dccpd_type == DCCP_PKT_ACK ||
+		    dcb->dccpd_type == DCCP_PKT_DATAACK)
+			dccp_event_ack_sent(sk);
+
+		DCCP_INC_STATS(DCCP_MIB_OUTSEGS);
+
+		err = ip_queue_xmit(skb, 0);
+		if (err <= 0)
+			return err;
+
+		/* NET_XMIT_CN is special. It does not guarantee,
+		 * that this packet is lost. It tells that device
+		 * is about to start to drop packets or already
+		 * drops some packets of the same priority and
+		 * invokes us to send less aggressively.
+		 */
+		return err == NET_XMIT_CN ? 0 : err;
+	}
+	return -ENOBUFS;
+}
+
+unsigned int dccp_sync_mss(struct sock *sk, u32 pmtu)
+{
+	struct dccp_sock *dp = dccp_sk(sk);
+	int mss_now;
+
+	/*
+	 * FIXME: we really should be using the af_specific thing to support IPv6.
+	 * mss_now = pmtu - tp->af_specific->net_header_len - sizeof(struct dccp_hdr) - sizeof(struct dccp_hdr_ext);
+	 */
+	mss_now = pmtu - sizeof(struct iphdr) - sizeof(struct dccp_hdr) - sizeof(struct dccp_hdr_ext);
+
+	/* Now subtract optional transport overhead */
+	mss_now -= dp->dccps_ext_header_len;
+
+	/*
+	 * FIXME: this should come from the CCID infrastructure, where, say,
+	 * TFRC will say it wants TIMESTAMPS, ELAPSED time, etc, for now lets
+	 * put a rough estimate for NDP + TIMESTAMP + TIMESTAMP_ECHO + ELAPSED
+	 * TIME + TFRC_OPT_LOSS_EVENT_RATE + TFRC_OPT_RECEIVE_RATE + padding to
+	 * make it a multiple of 4
+	 */
+
+	mss_now -= ((5 + 6 + 10 + 6 + 6 + 6 + 3) / 4) * 4;
+
+	/* And store cached results */
+	dp->dccps_pmtu_cookie = pmtu;
+	dp->dccps_mss_cache = mss_now;
+
+	return mss_now;
+}
+
+int dccp_retransmit_skb(struct sock *sk, struct sk_buff *skb)
+{
+	if (inet_sk_rebuild_header(sk) != 0)
+		return -EHOSTUNREACH; /* Routing failure or similar. */
+
+	return dccp_transmit_skb(sk, (skb_cloned(skb) ?
+				      pskb_copy(skb, GFP_ATOMIC):
+				      skb_clone(skb, GFP_ATOMIC)));
+}
+
+struct sk_buff *dccp_make_response(struct sock *sk, struct dst_entry *dst,
+				   struct request_sock *req)
+{
+	struct dccp_hdr *dh;
+	const int dccp_header_size = sizeof(struct dccp_hdr) +
+				     sizeof(struct dccp_hdr_ext) +
+				     sizeof(struct dccp_hdr_response);
+	struct sk_buff *skb = sock_wmalloc(sk, MAX_HEADER + DCCP_MAX_OPT_LEN +
+					       dccp_header_size, 1,
+					   GFP_ATOMIC);
+	if (skb == NULL)
+		return NULL;
+
+	/* Reserve space for headers. */
+	skb_reserve(skb, MAX_HEADER + DCCP_MAX_OPT_LEN + dccp_header_size);
+
+	skb->dst = dst_clone(dst);
+	skb->csum = 0;
+
+	DCCP_SKB_CB(skb)->dccpd_type = DCCP_PKT_RESPONSE;
+	DCCP_SKB_CB(skb)->dccpd_seq  = dccp_rsk(req)->dreq_iss;
+	dccp_insert_options(sk, skb);
+
+	skb->h.raw = skb_push(skb, dccp_header_size);
+
+	dh = dccp_hdr(skb);
+	memset(dh, 0, dccp_header_size);
+
+	dh->dccph_sport	= inet_sk(sk)->sport;
+	dh->dccph_dport	= inet_rsk(req)->rmt_port;
+	dh->dccph_doff	= (dccp_header_size + DCCP_SKB_CB(skb)->dccpd_opt_len) / 4;
+	dh->dccph_type	= DCCP_PKT_RESPONSE;
+	dh->dccph_x	= 1;
+	dccp_hdr_set_seq(dh, dccp_rsk(req)->dreq_iss);
+	dccp_hdr_set_ack(dccp_hdr_ack_bits(skb), dccp_rsk(req)->dreq_isr);
+
+	dh->dccph_checksum = dccp_v4_checksum(skb);
+
+	DCCP_INC_STATS(DCCP_MIB_OUTSEGS);
+	return skb;
+}
+
+struct sk_buff *dccp_make_reset(struct sock *sk, struct dst_entry *dst,
+				const enum dccp_reset_codes code)
+				   
+{
+	struct dccp_hdr *dh;
+	struct dccp_sock *dp = dccp_sk(sk);
+	const int dccp_header_size = sizeof(struct dccp_hdr) +
+				     sizeof(struct dccp_hdr_ext) +
+				     sizeof(struct dccp_hdr_reset);
+	struct sk_buff *skb = sock_wmalloc(sk, MAX_HEADER + DCCP_MAX_OPT_LEN +
+					       dccp_header_size, 1,
+					   GFP_ATOMIC);
+	if (skb == NULL)
+		return NULL;
+
+	/* Reserve space for headers. */
+	skb_reserve(skb, MAX_HEADER + DCCP_MAX_OPT_LEN + dccp_header_size);
+
+	skb->dst = dst_clone(dst);
+	skb->csum = 0;
+
+	dccp_inc_seqno(&dp->dccps_gss);
+
+	DCCP_SKB_CB(skb)->dccpd_reset_code = code;
+	DCCP_SKB_CB(skb)->dccpd_type	   = DCCP_PKT_RESET;
+	DCCP_SKB_CB(skb)->dccpd_seq	   = dp->dccps_gss;
+	dccp_insert_options(sk, skb);
+
+	skb->h.raw = skb_push(skb, dccp_header_size);
+
+	dh = dccp_hdr(skb);
+	memset(dh, 0, dccp_header_size);
+
+	dh->dccph_sport	= inet_sk(sk)->sport;
+	dh->dccph_dport	= inet_sk(sk)->dport;
+	dh->dccph_doff	= (dccp_header_size + DCCP_SKB_CB(skb)->dccpd_opt_len) / 4;
+	dh->dccph_type	= DCCP_PKT_RESET;
+	dh->dccph_x	= 1;
+	dccp_hdr_set_seq(dh, dp->dccps_gss);
+	dccp_hdr_set_ack(dccp_hdr_ack_bits(skb), dp->dccps_gsr);
+
+	dccp_hdr_reset(skb)->dccph_reset_code = code;
+
+	dh->dccph_checksum = dccp_v4_checksum(skb);
+
+	DCCP_INC_STATS(DCCP_MIB_OUTSEGS);
+	return skb;
+}
+
+/*
+ * Do all connect socket setups that can be done AF independent.
+ */
+static inline void dccp_connect_init(struct sock *sk)
+{
+	struct dst_entry *dst = __sk_dst_get(sk);
+	struct inet_connection_sock *icsk = inet_csk(sk);
+
+	sk->sk_err = 0;
+	sock_reset_flag(sk, SOCK_DONE);
+	
+	dccp_sync_mss(sk, dst_mtu(dst));
+
+	/*
+	 * FIXME: set dp->{dccps_swh,dccps_swl}, with
+	 * something like dccp_inc_seq
+	 */
+
+	icsk->icsk_retransmits = 0;
+}
+
+int dccp_connect(struct sock *sk)
+{
+	struct sk_buff *skb;
+	struct inet_connection_sock *icsk = inet_csk(sk);
+
+	dccp_connect_init(sk);
+
+	skb = alloc_skb(MAX_DCCP_HEADER + 15, sk->sk_allocation);
+	if (unlikely(skb == NULL))
+		return -ENOBUFS;
+
+	/* Reserve space for headers. */
+	skb_reserve(skb, MAX_DCCP_HEADER);
+
+	DCCP_SKB_CB(skb)->dccpd_type = DCCP_PKT_REQUEST;
+	/* FIXME: set service to something meaningful, coming
+	 * from userspace*/
+	DCCP_SKB_CB(skb)->dccpd_service = 0;
+	skb->csum = 0;
+	skb_set_owner_w(skb, sk);
+
+	BUG_TRAP(sk->sk_send_head == NULL);
+	sk->sk_send_head = skb;
+	dccp_transmit_skb(sk, skb_clone(skb, GFP_KERNEL));
+	DCCP_INC_STATS(DCCP_MIB_ACTIVEOPENS);
+
+	/* Timer for repeating the REQUEST until an answer. */
+	inet_csk_reset_xmit_timer(sk, ICSK_TIME_RETRANS, icsk->icsk_rto, TCP_RTO_MAX);
+	return 0;
+}
+
+void dccp_send_ack(struct sock *sk)
+{
+	/* If we have been reset, we may not send again. */
+	if (sk->sk_state != DCCP_CLOSED) {
+		struct sk_buff *skb = alloc_skb(MAX_DCCP_HEADER, GFP_ATOMIC);
+
+		if (skb == NULL) {
+			inet_csk_schedule_ack(sk);
+			inet_csk(sk)->icsk_ack.ato = TCP_ATO_MIN;
+			inet_csk_reset_xmit_timer(sk, ICSK_TIME_DACK, TCP_DELACK_MAX, TCP_RTO_MAX);
+			return;
+		}
+
+		/* Reserve space for headers */
+		skb_reserve(skb, MAX_DCCP_HEADER);
+		skb->csum = 0;
+		DCCP_SKB_CB(skb)->dccpd_type = DCCP_PKT_ACK;
+		skb_set_owner_w(skb, sk);
+		dccp_transmit_skb(sk, skb);
+	}
+}
+
+EXPORT_SYMBOL_GPL(dccp_send_ack);
+
+void dccp_send_delayed_ack(struct sock *sk)
+{
+	struct inet_connection_sock *icsk = inet_csk(sk);
+	/*
+	 * FIXME: tune this timer. elapsed time fixes the skew, so no problem
+	 * with using 2s, and active senders also piggyback the ACK into a
+	 * DATAACK packet, so this is really for quiescent senders.
+	 */
+	unsigned long timeout = jiffies + 2 * HZ;
+
+	/* Use new timeout only if there wasn't a older one earlier. */
+	if (icsk->icsk_ack.pending & ICSK_ACK_TIMER) {
+		/* If delack timer was blocked or is about to expire,
+		 * send ACK now.
+		 *
+		 * FIXME: check the "about to expire" part
+		 */
+		if (icsk->icsk_ack.blocked) {
+			dccp_send_ack(sk);
+			return;
+		}
+
+		if (!time_before(timeout, icsk->icsk_ack.timeout))
+			timeout = icsk->icsk_ack.timeout;
+	}
+	icsk->icsk_ack.pending |= ICSK_ACK_SCHED | ICSK_ACK_TIMER;
+	icsk->icsk_ack.timeout = timeout;
+	sk_reset_timer(sk, &icsk->icsk_delack_timer, timeout);
+}
+
+void dccp_send_sync(struct sock *sk, u64 seq)
+{
+	/*
+	 * We are not putting this on the write queue, so
+	 * dccp_transmit_skb() will set the ownership to this
+	 * sock.
+	 */
+	struct sk_buff *skb = alloc_skb(MAX_DCCP_HEADER, GFP_ATOMIC);
+
+	if (skb == NULL)
+		/* FIXME: how to make sure the sync is sent? */
+		return;
+
+	/* Reserve space for headers and prepare control bits. */
+	skb_reserve(skb, MAX_DCCP_HEADER);
+	skb->csum = 0;
+	DCCP_SKB_CB(skb)->dccpd_type = DCCP_PKT_SYNC;
+	DCCP_SKB_CB(skb)->dccpd_seq = seq;
+
+	skb_set_owner_w(skb, sk);
+	dccp_transmit_skb(sk, skb);
+}
+
+/* Send a DCCP_PKT_CLOSE/CLOSEREQ.  The caller locks the socket for us.  This cannot be
+ * allowed to fail queueing a DCCP_PKT_CLOSE/CLOSEREQ frame under any circumstances.
+ */
+void dccp_send_close(struct sock *sk)
+{
+	struct dccp_sock *dp = dccp_sk(sk);
+	struct sk_buff *skb;
+
+	/* Socket is locked, keep trying until memory is available. */
+	for (;;) {
+		skb = alloc_skb(sk->sk_prot->max_header, GFP_KERNEL);
+		if (skb != NULL)
+			break;
+		yield();
+	}
+
+	/* Reserve space for headers and prepare control bits. */
+	skb_reserve(skb, sk->sk_prot->max_header);
+	skb->csum = 0;
+	DCCP_SKB_CB(skb)->dccpd_type = dp->dccps_role == DCCP_ROLE_CLIENT ? DCCP_PKT_CLOSE : DCCP_PKT_CLOSEREQ;
+
+	skb_set_owner_w(skb, sk);
+	dccp_transmit_skb(sk, skb);
+
+	ccid_hc_rx_exit(dp->dccps_hc_rx_ccid, sk);
+	ccid_hc_tx_exit(dp->dccps_hc_tx_ccid, sk);
+}

net/dccp/proto.c

@@ -0,0 +1,818 @@
+/*
+ *  net/dccp/proto.c
+ *
+ *  An implementation of the DCCP protocol
+ *  Arnaldo Carvalho de Melo <acme@conectiva.com.br>
+ *
+ *	This program is free software; you can redistribute it and/or modify it
+ *	under the terms of the GNU General Public License version 2 as
+ *	published by the Free Software Foundation.
+ */
+
+#include <linux/config.h>
+#include <linux/dccp.h>
+#include <linux/module.h>
+#include <linux/types.h>
+#include <linux/sched.h>
+#include <linux/kernel.h>
+#include <linux/skbuff.h>
+#include <linux/netdevice.h>
+#include <linux/in.h>
+#include <linux/if_arp.h>
+#include <linux/init.h>
+#include <linux/random.h>
+#include <net/checksum.h>
+
+#include <net/inet_common.h>
+#include <net/ip.h>
+#include <net/protocol.h>
+#include <net/sock.h>
+#include <net/xfrm.h>
+
+#include <asm/semaphore.h>
+#include <linux/spinlock.h>
+#include <linux/timer.h>
+#include <linux/delay.h>
+#include <linux/poll.h>
+#include <linux/dccp.h>
+
+#include "ccid.h"
+#include "dccp.h"
+
+DEFINE_SNMP_STAT(struct dccp_mib, dccp_statistics);
+
+atomic_t dccp_orphan_count = ATOMIC_INIT(0);
+
+static struct net_protocol dccp_protocol = {
+	.handler	= dccp_v4_rcv,
+	.err_handler	= dccp_v4_err,
+};
+
+const char *dccp_packet_name(const int type)
+{
+	static const char *dccp_packet_names[] = {
+		[DCCP_PKT_REQUEST]  = "REQUEST",
+		[DCCP_PKT_RESPONSE] = "RESPONSE",
+		[DCCP_PKT_DATA]	    = "DATA",
+		[DCCP_PKT_ACK]	    = "ACK",
+		[DCCP_PKT_DATAACK]  = "DATAACK",
+		[DCCP_PKT_CLOSEREQ] = "CLOSEREQ",
+		[DCCP_PKT_CLOSE]    = "CLOSE",
+		[DCCP_PKT_RESET]    = "RESET",
+		[DCCP_PKT_SYNC]	    = "SYNC",
+		[DCCP_PKT_SYNCACK]  = "SYNCACK",
+	};
+
+	if (type >= DCCP_NR_PKT_TYPES)
+		return "INVALID";
+	else
+		return dccp_packet_names[type];
+}
+
+EXPORT_SYMBOL_GPL(dccp_packet_name);
+
+const char *dccp_state_name(const int state)
+{
+	static char *dccp_state_names[] = {
+	[DCCP_OPEN]	  = "OPEN",
+	[DCCP_REQUESTING] = "REQUESTING",
+	[DCCP_PARTOPEN]	  = "PARTOPEN",
+	[DCCP_LISTEN]	  = "LISTEN",
+	[DCCP_RESPOND]	  = "RESPOND",
+	[DCCP_CLOSING]	  = "CLOSING",
+	[DCCP_TIME_WAIT]  = "TIME_WAIT",
+	[DCCP_CLOSED]	  = "CLOSED",
+	};
+
+	if (state >= DCCP_MAX_STATES)
+		return "INVALID STATE!";
+	else
+		return dccp_state_names[state];
+}
+
+EXPORT_SYMBOL_GPL(dccp_state_name);
+
+static inline int dccp_listen_start(struct sock *sk)
+{
+	dccp_sk(sk)->dccps_role = DCCP_ROLE_LISTEN;
+	return inet_csk_listen_start(sk, TCP_SYNQ_HSIZE);
+}
+
+int dccp_disconnect(struct sock *sk, int flags)
+{
+	struct inet_connection_sock *icsk = inet_csk(sk);
+	struct inet_sock *inet = inet_sk(sk);
+	int err = 0;
+	const int old_state = sk->sk_state;
+
+	if (old_state != DCCP_CLOSED)
+		dccp_set_state(sk, DCCP_CLOSED);
+
+	/* ABORT function of RFC793 */
+	if (old_state == DCCP_LISTEN) {
+		inet_csk_listen_stop(sk);
+	/* FIXME: do the active reset thing */
+	} else if (old_state == DCCP_REQUESTING)
+		sk->sk_err = ECONNRESET;
+
+	dccp_clear_xmit_timers(sk);
+	__skb_queue_purge(&sk->sk_receive_queue);
+	if (sk->sk_send_head != NULL) {
+		__kfree_skb(sk->sk_send_head);
+		sk->sk_send_head = NULL;
+	}
+
+	inet->dport = 0;
+
+	if (!(sk->sk_userlocks & SOCK_BINDADDR_LOCK))
+		inet_reset_saddr(sk);
+
+	sk->sk_shutdown = 0;
+	sock_reset_flag(sk, SOCK_DONE);
+
+	icsk->icsk_backoff = 0;
+	inet_csk_delack_init(sk);
+	__sk_dst_reset(sk);
+
+	BUG_TRAP(!inet->num || icsk->icsk_bind_hash);
+
+	sk->sk_error_report(sk);
+	return err;
+}
+
+int dccp_ioctl(struct sock *sk, int cmd, unsigned long arg)
+{
+	dccp_pr_debug("entry\n");
+	return -ENOIOCTLCMD;
+}
+
+int dccp_setsockopt(struct sock *sk, int level, int optname,
+		    char *optval, int optlen)
+{
+	dccp_pr_debug("entry\n");
+
+	if (level != SOL_DCCP)
+		return ip_setsockopt(sk, level, optname, optval, optlen);
+
+	return -EOPNOTSUPP;
+}
+
+int dccp_getsockopt(struct sock *sk, int level, int optname,
+		    char *optval, int *optlen)
+{
+	dccp_pr_debug("entry\n");
+
+	if (level != SOL_DCCP)
+		return ip_getsockopt(sk, level, optname, optval, optlen);
+
+	return -EOPNOTSUPP;
+}
+
+int dccp_sendmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg,
+		 size_t len)
+{
+	const struct dccp_sock *dp = dccp_sk(sk);
+	const int flags = msg->msg_flags;
+	const int noblock = flags & MSG_DONTWAIT;
+	struct sk_buff *skb;
+	int rc, size;
+	long timeo;
+
+	if (len > dp->dccps_mss_cache)
+		return -EMSGSIZE;
+
+	lock_sock(sk);
+
+	timeo = sock_sndtimeo(sk, flags & MSG_DONTWAIT);
+
+	/*
+	 * We have to use sk_stream_wait_connect here to set sk_write_pending,
+	 * so that the trick in dccp_rcv_request_sent_state_process.
+	 */
+	/* Wait for a connection to finish. */
+	if ((1 << sk->sk_state) & ~(DCCPF_OPEN | DCCPF_PARTOPEN | DCCPF_CLOSING))
+		if ((rc = sk_stream_wait_connect(sk, &timeo)) != 0)
+			goto out_err;
+
+	size = sk->sk_prot->max_header + len;
+	release_sock(sk);
+	skb = sock_alloc_send_skb(sk, size, noblock, &rc);
+	lock_sock(sk);
+
+	if (skb == NULL)
+		goto out_release;
+
+	skb_reserve(skb, sk->sk_prot->max_header);
+	rc = memcpy_fromiovec(skb_put(skb, len), msg->msg_iov, len);
+	if (rc == 0) {
+		struct dccp_skb_cb *dcb = DCCP_SKB_CB(skb);
+		const struct dccp_ackpkts *ap = dp->dccps_hc_rx_ackpkts;
+		long delay; 
+
+		/*
+		 * XXX: This is just to match the Waikato tree CA interaction
+		 * points, after the CCID3 code is stable and I have a better
+		 * understanding of behaviour I'll change this to look more like
+		 * TCP.
+		 */
+		while (1) {
+			rc = ccid_hc_tx_send_packet(dp->dccps_hc_tx_ccid, sk,
+						    skb, len, &delay);
+			if (rc == 0)
+				break;
+			if (rc != -EAGAIN)
+				goto out_discard;
+			if (delay > timeo)
+				goto out_discard;
+			release_sock(sk);
+			delay = schedule_timeout(delay);
+			lock_sock(sk);
+			timeo -= delay;
+			if (signal_pending(current))
+				goto out_interrupted;
+			rc = -EPIPE;
+			if (!(sk->sk_state == DCCP_PARTOPEN || sk->sk_state == DCCP_OPEN))
+				goto out_discard;
+		}
+
+		if (sk->sk_state == DCCP_PARTOPEN) {
+			/* See 8.1.5.  Handshake Completion */
+			inet_csk_schedule_ack(sk);
+			inet_csk_reset_xmit_timer(sk, ICSK_TIME_DACK, inet_csk(sk)->icsk_rto, TCP_RTO_MAX);
+			dcb->dccpd_type = DCCP_PKT_DATAACK;
+			/* FIXME: we really should have a dccps_ack_pending or use icsk */
+		} else if (inet_csk_ack_scheduled(sk) ||
+			   (dp->dccps_options.dccpo_send_ack_vector &&
+			    ap->dccpap_buf_ackno != DCCP_MAX_SEQNO + 1 &&
+			    ap->dccpap_ack_seqno == DCCP_MAX_SEQNO + 1))
+			dcb->dccpd_type = DCCP_PKT_DATAACK;
+		else
+			dcb->dccpd_type = DCCP_PKT_DATA;
+		dccp_transmit_skb(sk, skb);
+		ccid_hc_tx_packet_sent(dp->dccps_hc_tx_ccid, sk, 0, len);
+	} else {
+out_discard:
+		kfree_skb(skb);
+	}
+out_release:
+	release_sock(sk);
+	return rc ? : len;
+out_err:
+	rc = sk_stream_error(sk, flags, rc);
+	goto out_release;
+out_interrupted:
+	rc = sock_intr_errno(timeo);
+	goto out_discard;
+}
+
+EXPORT_SYMBOL(dccp_sendmsg);
+
+int dccp_recvmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg,
+		 size_t len, int nonblock, int flags, int *addr_len)
+{
+	const struct dccp_hdr *dh;
+	int copied = 0;
+	unsigned long used;
+	int err;
+	int target;		/* Read at least this many bytes */
+	long timeo;
+
+	lock_sock(sk);
+
+	err = -ENOTCONN;
+	if (sk->sk_state == DCCP_LISTEN)
+		goto out;
+
+	timeo = sock_rcvtimeo(sk, nonblock);
+
+	/* Urgent data needs to be handled specially. */
+	if (flags & MSG_OOB)
+		goto recv_urg;
+
+	/* FIXME */
+#if 0
+	seq = &tp->copied_seq;
+	if (flags & MSG_PEEK) {
+		peek_seq = tp->copied_seq;
+		seq = &peek_seq;
+	}
+#endif
+
+	target = sock_rcvlowat(sk, flags & MSG_WAITALL, len);
+
+	do {
+		struct sk_buff *skb;
+		u32 offset;
+
+	/* FIXME */
+#if 0
+		/* Are we at urgent data? Stop if we have read anything or have SIGURG pending. */
+		if (tp->urg_data && tp->urg_seq == *seq) {
+			if (copied)
+				break;
+			if (signal_pending(current)) {
+				copied = timeo ? sock_intr_errno(timeo) : -EAGAIN;
+				break;
+			}
+		}
+#endif
+
+		/* Next get a buffer. */
+
+		skb = skb_peek(&sk->sk_receive_queue);
+		do {
+			if (!skb)
+				break;
+
+			offset = 0;
+			dh = dccp_hdr(skb);
+
+			if (dh->dccph_type == DCCP_PKT_DATA ||
+			    dh->dccph_type == DCCP_PKT_DATAACK)
+				goto found_ok_skb;
+
+			if (dh->dccph_type == DCCP_PKT_RESET ||
+			    dh->dccph_type == DCCP_PKT_CLOSE) {
+				dccp_pr_debug("found fin ok!\n");
+				goto found_fin_ok;
+			}
+			dccp_pr_debug("packet_type=%s\n", dccp_packet_name(dh->dccph_type));
+			BUG_TRAP(flags & MSG_PEEK);
+			skb = skb->next;
+		} while (skb != (struct sk_buff *)&sk->sk_receive_queue);
+
+		/* Well, if we have backlog, try to process it now yet. */
+		if (copied >= target && !sk->sk_backlog.tail)
+			break;
+
+		if (copied) {
+			if (sk->sk_err ||
+			    sk->sk_state == DCCP_CLOSED ||
+			    (sk->sk_shutdown & RCV_SHUTDOWN) ||
+			    !timeo ||
+			    signal_pending(current) ||
+			    (flags & MSG_PEEK))
+				break;
+		} else {
+			if (sock_flag(sk, SOCK_DONE))
+				break;
+
+			if (sk->sk_err) {
+				copied = sock_error(sk);
+				break;
+			}
+
+			if (sk->sk_shutdown & RCV_SHUTDOWN)
+				break;
+
+			if (sk->sk_state == DCCP_CLOSED) {
+				if (!sock_flag(sk, SOCK_DONE)) {
+					/* This occurs when user tries to read
+					 * from never connected socket.
+					 */
+					copied = -ENOTCONN;
+					break;
+				}
+				break;
+			}
+
+			if (!timeo) {
+				copied = -EAGAIN;
+				break;
+			}
+
+			if (signal_pending(current)) {
+				copied = sock_intr_errno(timeo);
+				break;
+			}
+		}
+
+		/* FIXME: cleanup_rbuf(sk, copied); */
+
+		if (copied >= target) {
+			/* Do not sleep, just process backlog. */
+			release_sock(sk);
+			lock_sock(sk);
+		} else
+			sk_wait_data(sk, &timeo);
+
+		continue;
+
+	found_ok_skb:
+		/* Ok so how much can we use? */
+		used = skb->len - offset;
+		if (len < used)
+			used = len;
+
+		if (!(flags & MSG_TRUNC)) {
+			err = skb_copy_datagram_iovec(skb, offset,
+						      msg->msg_iov, used);
+			if (err) {
+				/* Exception. Bailout! */
+				if (!copied)
+					copied = -EFAULT;
+				break;
+			}
+		}
+
+		copied += used;
+		len -= used;
+
+		/* FIXME: tcp_rcv_space_adjust(sk); */
+
+//skip_copy:
+		if (used + offset < skb->len)
+			continue;
+
+		if (!(flags & MSG_PEEK))
+			sk_eat_skb(sk, skb);
+		continue;
+	found_fin_ok:
+		if (!(flags & MSG_PEEK))
+			sk_eat_skb(sk, skb);
+		break;
+		
+	} while (len > 0);
+
+	/* According to UNIX98, msg_name/msg_namelen are ignored
+	 * on connected socket. I was just happy when found this 8) --ANK
+	 */
+
+	/* Clean up data we have read: This will do ACK frames. */
+	/* FIXME: cleanup_rbuf(sk, copied); */
+
+	release_sock(sk);
+	return copied;
+
+out:
+	release_sock(sk);
+	return err;
+
+recv_urg:
+	/* FIXME: err = tcp_recv_urg(sk, timeo, msg, len, flags, addr_len); */
+	goto out;
+}
+
+static int inet_dccp_listen(struct socket *sock, int backlog)
+{
+	struct sock *sk = sock->sk;
+	unsigned char old_state;
+	int err;
+
+	lock_sock(sk);
+
+	err = -EINVAL;
+	if (sock->state != SS_UNCONNECTED || sock->type != SOCK_DCCP)
+		goto out;
+
+	old_state = sk->sk_state;
+	if (!((1 << old_state) & (DCCPF_CLOSED | DCCPF_LISTEN)))
+		goto out;
+
+	/* Really, if the socket is already in listen state
+	 * we can only allow the backlog to be adjusted.
+	 */
+	if (old_state != DCCP_LISTEN) {
+		/*
+		 * FIXME: here it probably should be sk->sk_prot->listen_start
+		 * see tcp_listen_start
+		 */
+		err = dccp_listen_start(sk);
+		if (err)
+			goto out;
+	}
+	sk->sk_max_ack_backlog = backlog;
+	err = 0;
+
+out:
+	release_sock(sk);
+	return err;
+}
+
+static const unsigned char dccp_new_state[] = {
+	/* current state:        new state:      action:	*/
+	[0]			= DCCP_CLOSED,
+	[DCCP_OPEN] 		= DCCP_CLOSING | DCCP_ACTION_FIN,
+	[DCCP_REQUESTING] 	= DCCP_CLOSED,
+	[DCCP_PARTOPEN]	= DCCP_CLOSING | DCCP_ACTION_FIN,
+	[DCCP_LISTEN]		= DCCP_CLOSED,
+	[DCCP_RESPOND] 	= DCCP_CLOSED,
+	[DCCP_CLOSING]	= DCCP_CLOSED,
+	[DCCP_TIME_WAIT] 	= DCCP_CLOSED,
+	[DCCP_CLOSED] 	= DCCP_CLOSED,
+};
+
+static int dccp_close_state(struct sock *sk)
+{
+	const int next = dccp_new_state[sk->sk_state];
+	const int ns = next & DCCP_STATE_MASK;
+
+	if (ns != sk->sk_state)
+		dccp_set_state(sk, ns);
+
+	return next & DCCP_ACTION_FIN;
+}
+
+void dccp_close(struct sock *sk, long timeout)
+{
+	struct sk_buff *skb;
+
+	lock_sock(sk);
+
+	sk->sk_shutdown = SHUTDOWN_MASK;
+
+	if (sk->sk_state == DCCP_LISTEN) {
+		dccp_set_state(sk, DCCP_CLOSED);
+
+		/* Special case. */
+		inet_csk_listen_stop(sk);
+
+		goto adjudge_to_death;
+	}
+
+	/*
+	 * We need to flush the recv. buffs.  We do this only on the
+	 * descriptor close, not protocol-sourced closes, because the
+	  *reader process may not have drained the data yet!
+	 */
+	/* FIXME: check for unread data */
+	while ((skb = __skb_dequeue(&sk->sk_receive_queue)) != NULL) {
+		__kfree_skb(skb);
+	}
+
+	if (sock_flag(sk, SOCK_LINGER) && !sk->sk_lingertime) {
+		/* Check zero linger _after_ checking for unread data. */
+		sk->sk_prot->disconnect(sk, 0);
+	} else if (dccp_close_state(sk)) {
+		dccp_send_close(sk);
+	}
+
+	sk_stream_wait_close(sk, timeout);
+
+adjudge_to_death:
+	release_sock(sk);
+	/*
+	 * Now socket is owned by kernel and we acquire BH lock
+	 * to finish close. No need to check for user refs.
+	 */
+	local_bh_disable();
+	bh_lock_sock(sk);
+	BUG_TRAP(!sock_owned_by_user(sk));
+
+	sock_hold(sk);
+	sock_orphan(sk);
+						
+	if (sk->sk_state != DCCP_CLOSED)
+		dccp_set_state(sk, DCCP_CLOSED);
+
+	atomic_inc(&dccp_orphan_count);
+	if (sk->sk_state == DCCP_CLOSED)
+		inet_csk_destroy_sock(sk);
+
+	/* Otherwise, socket is reprieved until protocol close. */
+
+	bh_unlock_sock(sk);
+	local_bh_enable();
+	sock_put(sk);
+}
+
+void dccp_shutdown(struct sock *sk, int how)
+{
+	dccp_pr_debug("entry\n");
+}
+
+struct proto_ops inet_dccp_ops = {
+	.family		= PF_INET,
+	.owner		= THIS_MODULE,
+	.release	= inet_release,
+	.bind		= inet_bind,
+	.connect	= inet_stream_connect,
+	.socketpair	= sock_no_socketpair,
+	.accept		= inet_accept,
+	.getname	= inet_getname,
+	.poll		= sock_no_poll,
+	.ioctl		= inet_ioctl,
+	.listen		= inet_dccp_listen, /* FIXME: work on inet_listen to rename it to sock_common_listen */
+	.shutdown	= inet_shutdown,
+	.setsockopt	= sock_common_setsockopt,
+	.getsockopt	= sock_common_getsockopt,
+	.sendmsg	= inet_sendmsg,
+	.recvmsg	= sock_common_recvmsg,
+	.mmap		= sock_no_mmap,
+	.sendpage	= sock_no_sendpage,
+};
+
+extern struct net_proto_family inet_family_ops;
+
+static struct inet_protosw dccp_v4_protosw = {
+	.type		= SOCK_DCCP,
+	.protocol	= IPPROTO_DCCP,
+	.prot		= &dccp_v4_prot,
+	.ops		= &inet_dccp_ops,
+	.capability	= -1,
+	.no_check	= 0,
+	.flags		= 0,
+};
+
+/*
+ * This is the global socket data structure used for responding to
+ * the Out-of-the-blue (OOTB) packets. A control sock will be created
+ * for this socket at the initialization time.
+ */
+struct socket *dccp_ctl_socket;
+
+static char dccp_ctl_socket_err_msg[] __initdata =
+	KERN_ERR "DCCP: Failed to create the control socket.\n";
+
+static int __init dccp_ctl_sock_init(void)
+{
+	int rc = sock_create_kern(PF_INET, SOCK_DCCP, IPPROTO_DCCP,
+				  &dccp_ctl_socket);
+	if (rc < 0)
+		printk(dccp_ctl_socket_err_msg);
+	else {
+		dccp_ctl_socket->sk->sk_allocation = GFP_ATOMIC;
+		inet_sk(dccp_ctl_socket->sk)->uc_ttl = -1;
+
+		/* Unhash it so that IP input processing does not even
+		 * see it, we do not wish this socket to see incoming
+		 * packets.
+		 */
+		dccp_ctl_socket->sk->sk_prot->unhash(dccp_ctl_socket->sk);
+	}
+
+	return rc;
+}
+
+static void __exit dccp_ctl_sock_exit(void)
+{
+	if (dccp_ctl_socket != NULL)
+		sock_release(dccp_ctl_socket);
+}
+
+static int __init init_dccp_v4_mibs(void)
+{
+	int rc = -ENOMEM;
+
+	dccp_statistics[0] = alloc_percpu(struct dccp_mib);
+	if (dccp_statistics[0] == NULL)
+		goto out;
+
+	dccp_statistics[1] = alloc_percpu(struct dccp_mib);
+	if (dccp_statistics[1] == NULL)
+		goto out_free_one;
+
+	rc = 0;
+out:
+	return rc;
+out_free_one:
+	free_percpu(dccp_statistics[0]);
+	dccp_statistics[0] = NULL;
+	goto out;
+
+}
+
+static int thash_entries;
+module_param(thash_entries, int, 0444);
+MODULE_PARM_DESC(thash_entries, "Number of ehash buckets");
+
+int dccp_debug;
+module_param(dccp_debug, int, 0444);
+MODULE_PARM_DESC(dccp_debug, "Enable debug messages");
+
+static int __init dccp_init(void)
+{
+	unsigned long goal;
+	int ehash_order, bhash_order, i;
+	int rc = proto_register(&dccp_v4_prot, 1);
+
+	if (rc)
+		goto out;
+
+	dccp_hashinfo.bind_bucket_cachep = kmem_cache_create("dccp_bind_bucket",
+					       sizeof(struct inet_bind_bucket),
+					       0, SLAB_HWCACHE_ALIGN,
+					       NULL, NULL);
+	if (!dccp_hashinfo.bind_bucket_cachep)
+		goto out_proto_unregister;
+
+	/*
+	 * Size and allocate the main established and bind bucket
+	 * hash tables.
+	 *
+	 * The methodology is similar to that of the buffer cache.
+	 */
+	if (num_physpages >= (128 * 1024))
+		goal = num_physpages >> (21 - PAGE_SHIFT);
+	else
+		goal = num_physpages >> (23 - PAGE_SHIFT);
+
+	if (thash_entries)
+		goal = (thash_entries * sizeof(struct inet_ehash_bucket)) >> PAGE_SHIFT;
+	for (ehash_order = 0; (1UL << ehash_order) < goal; ehash_order++)
+		;
+	do {
+		dccp_hashinfo.ehash_size = (1UL << ehash_order) * PAGE_SIZE /
+					sizeof(struct inet_ehash_bucket);
+		dccp_hashinfo.ehash_size >>= 1;
+		while (dccp_hashinfo.ehash_size & (dccp_hashinfo.ehash_size - 1))
+			dccp_hashinfo.ehash_size--;
+		dccp_hashinfo.ehash = (struct inet_ehash_bucket *)
+			__get_free_pages(GFP_ATOMIC, ehash_order);
+	} while (!dccp_hashinfo.ehash && --ehash_order > 0);
+
+	if (!dccp_hashinfo.ehash) {
+		printk(KERN_CRIT "Failed to allocate DCCP "
+				 "established hash table\n");
+		goto out_free_bind_bucket_cachep;
+	}
+
+	for (i = 0; i < (dccp_hashinfo.ehash_size << 1); i++) {
+		rwlock_init(&dccp_hashinfo.ehash[i].lock);
+		INIT_HLIST_HEAD(&dccp_hashinfo.ehash[i].chain);
+	}
+
+	bhash_order = ehash_order;
+
+	do {
+		dccp_hashinfo.bhash_size = (1UL << bhash_order) * PAGE_SIZE /
+					sizeof(struct inet_bind_hashbucket);
+		if ((dccp_hashinfo.bhash_size > (64 * 1024)) && bhash_order > 0)
+			continue;
+		dccp_hashinfo.bhash = (struct inet_bind_hashbucket *)
+			__get_free_pages(GFP_ATOMIC, bhash_order);
+	} while (!dccp_hashinfo.bhash && --bhash_order >= 0);
+
+	if (!dccp_hashinfo.bhash) {
+		printk(KERN_CRIT "Failed to allocate DCCP bind hash table\n");
+		goto out_free_dccp_ehash;
+	}
+
+	for (i = 0; i < dccp_hashinfo.bhash_size; i++) {
+		spin_lock_init(&dccp_hashinfo.bhash[i].lock);
+		INIT_HLIST_HEAD(&dccp_hashinfo.bhash[i].chain);
+	}
+
+	if (init_dccp_v4_mibs())
+		goto out_free_dccp_bhash;
+
+	rc = -EAGAIN;
+	if (inet_add_protocol(&dccp_protocol, IPPROTO_DCCP))
+		goto out_free_dccp_v4_mibs;
+
+	inet_register_protosw(&dccp_v4_protosw);
+
+	rc = dccp_ctl_sock_init();
+	if (rc)
+		goto out_unregister_protosw;
+out:
+	return rc;
+out_unregister_protosw:
+	inet_unregister_protosw(&dccp_v4_protosw);
+	inet_del_protocol(&dccp_protocol, IPPROTO_DCCP);
+out_free_dccp_v4_mibs:
+	free_percpu(dccp_statistics[0]);
+	free_percpu(dccp_statistics[1]);
+	dccp_statistics[0] = dccp_statistics[1] = NULL;
+out_free_dccp_bhash:
+	free_pages((unsigned long)dccp_hashinfo.bhash, bhash_order);
+	dccp_hashinfo.bhash = NULL;
+out_free_dccp_ehash:
+	free_pages((unsigned long)dccp_hashinfo.ehash, ehash_order);
+	dccp_hashinfo.ehash = NULL;
+out_free_bind_bucket_cachep:
+	kmem_cache_destroy(dccp_hashinfo.bind_bucket_cachep);
+	dccp_hashinfo.bind_bucket_cachep = NULL;
+out_proto_unregister:
+	proto_unregister(&dccp_v4_prot);
+	goto out;
+}
+
+static const char dccp_del_proto_err_msg[] __exitdata =
+	KERN_ERR "can't remove dccp net_protocol\n";
+
+static void __exit dccp_fini(void)
+{
+	dccp_ctl_sock_exit();
+
+	inet_unregister_protosw(&dccp_v4_protosw);
+
+	if (inet_del_protocol(&dccp_protocol, IPPROTO_DCCP) < 0)
+		printk(dccp_del_proto_err_msg);
+
+	/* Free the control endpoint.  */
+	sock_release(dccp_ctl_socket);
+
+	proto_unregister(&dccp_v4_prot);
+
+	kmem_cache_destroy(dccp_hashinfo.bind_bucket_cachep);
+}
+
+module_init(dccp_init);
+module_exit(dccp_fini);
+
+/* __stringify doesn't likes enums, so use SOCK_DCCP (6) value directly  */
+MODULE_ALIAS("net-pf-" __stringify(PF_INET) "-6");
+MODULE_LICENSE("GPL");
+MODULE_AUTHOR("Arnaldo Carvalho de Melo <acme@conectiva.com.br>");
+MODULE_DESCRIPTION("DCCP - Datagram Congestion Controlled Protocol");

net/dccp/timer.c

@@ -0,0 +1,249 @@
+/*
+ *  net/dccp/timer.c
+ * 
+ *  An implementation of the DCCP protocol
+ *  Arnaldo Carvalho de Melo <acme@conectiva.com.br>
+ *
+ *	This program is free software; you can redistribute it and/or
+ *	modify it under the terms of the GNU General Public License
+ *	as published by the Free Software Foundation; either version
+ *	2 of the License, or (at your option) any later version.
+ */
+
+#include <linux/config.h>
+#include <linux/dccp.h>
+#include <linux/skbuff.h>
+
+#include "dccp.h"
+
+static void dccp_write_timer(unsigned long data);
+static void dccp_keepalive_timer(unsigned long data);
+static void dccp_delack_timer(unsigned long data);
+
+void dccp_init_xmit_timers(struct sock *sk)
+{
+	inet_csk_init_xmit_timers(sk, &dccp_write_timer, &dccp_delack_timer,
+				  &dccp_keepalive_timer);
+}
+
+static void dccp_write_err(struct sock *sk)
+{
+	sk->sk_err = sk->sk_err_soft ? : ETIMEDOUT;
+	sk->sk_error_report(sk);
+
+	dccp_v4_send_reset(sk, DCCP_RESET_CODE_ABORTED);
+	dccp_done(sk);
+	DCCP_INC_STATS_BH(DCCP_MIB_ABORTONTIMEOUT);
+}
+
+/* A write timeout has occurred. Process the after effects. */
+static int dccp_write_timeout(struct sock *sk)
+{
+	const struct inet_connection_sock *icsk = inet_csk(sk);
+	int retry_until;
+
+	if (sk->sk_state == DCCP_REQUESTING || sk->sk_state == DCCP_PARTOPEN) {
+		if (icsk->icsk_retransmits != 0)
+			dst_negative_advice(&sk->sk_dst_cache);
+		retry_until = icsk->icsk_syn_retries ? :  /* FIXME! */ 3 /* FIXME! sysctl_tcp_syn_retries */;
+	} else {
+		if (icsk->icsk_retransmits >= /* FIXME! sysctl_tcp_retries1 */ 5 /* FIXME! */) {
+			/* NOTE. draft-ietf-tcpimpl-pmtud-01.txt requires pmtu black
+			   hole detection. :-(
+
+			   It is place to make it. It is not made. I do not want
+			   to make it. It is disguisting. It does not work in any
+			   case. Let me to cite the same draft, which requires for
+			   us to implement this:
+
+   "The one security concern raised by this memo is that ICMP black holes
+   are often caused by over-zealous security administrators who block
+   all ICMP messages.  It is vitally important that those who design and
+   deploy security systems understand the impact of strict filtering on
+   upper-layer protocols.  The safest web site in the world is worthless
+   if most TCP implementations cannot transfer data from it.  It would
+   be far nicer to have all of the black holes fixed rather than fixing
+   all of the TCP implementations."
+
+                           Golden words :-).
+		   */
+
+			dst_negative_advice(&sk->sk_dst_cache);
+		}
+
+		retry_until = /* FIXME! */ 15 /* FIXME! sysctl_tcp_retries2 */;
+		/*
+		 * FIXME: see tcp_write_timout and tcp_out_of_resources
+		 */
+	}
+
+	if (icsk->icsk_retransmits >= retry_until) {
+		/* Has it gone just too far? */
+		dccp_write_err(sk);
+		return 1;
+	}
+	return 0;
+}
+
+/* This is the same as tcp_delack_timer, sans prequeue & mem_reclaim stuff */
+static void dccp_delack_timer(unsigned long data)
+{
+	struct sock *sk = (struct sock *)data;
+	struct inet_connection_sock *icsk = inet_csk(sk);
+
+	bh_lock_sock(sk);
+	if (sock_owned_by_user(sk)) {
+		/* Try again later. */
+		icsk->icsk_ack.blocked = 1;
+		NET_INC_STATS_BH(LINUX_MIB_DELAYEDACKLOCKED);
+		sk_reset_timer(sk, &icsk->icsk_delack_timer, jiffies + TCP_DELACK_MIN);
+		goto out;
+	}
+
+	if (sk->sk_state == DCCP_CLOSED || !(icsk->icsk_ack.pending & ICSK_ACK_TIMER))
+		goto out;
+	if (time_after(icsk->icsk_ack.timeout, jiffies)) {
+		sk_reset_timer(sk, &icsk->icsk_delack_timer, icsk->icsk_ack.timeout);
+		goto out;
+	}
+
+	icsk->icsk_ack.pending &= ~ICSK_ACK_TIMER;
+
+	if (inet_csk_ack_scheduled(sk)) {
+		if (!icsk->icsk_ack.pingpong) {
+			/* Delayed ACK missed: inflate ATO. */
+			icsk->icsk_ack.ato = min(icsk->icsk_ack.ato << 1, icsk->icsk_rto);
+		} else {
+			/* Delayed ACK missed: leave pingpong mode and
+			 * deflate ATO.
+			 */
+			icsk->icsk_ack.pingpong = 0;
+			icsk->icsk_ack.ato = TCP_ATO_MIN;
+		}
+		dccp_send_ack(sk);
+		NET_INC_STATS_BH(LINUX_MIB_DELAYEDACKS);
+	}
+out:
+	bh_unlock_sock(sk);
+	sock_put(sk);
+}
+
+/*
+ *	The DCCP retransmit timer.
+ */
+static void dccp_retransmit_timer(struct sock *sk)
+{
+	struct inet_connection_sock *icsk = inet_csk(sk);
+
+	/*
+	 * sk->sk_send_head has to have one skb with
+	 * DCCP_SKB_CB(skb)->dccpd_type set to one of the retransmittable DCCP
+	 * packet types (REQUEST, RESPONSE, the ACK in the 3way hanshake
+	 * (PARTOPEN timer), etc).
+	 */
+	BUG_TRAP(sk->sk_send_head != NULL);
+
+	/* 
+	 * More than than 4MSL (8 minutes) has passed, a RESET(aborted) was
+	 * sent, no need to retransmit, this sock is dead.
+	 */
+	if (dccp_write_timeout(sk))
+		goto out;
+
+	/*
+	 * We want to know the number of packets retransmitted, not the
+	 * total number of retransmissions of clones of original packets.
+	 */
+	if (icsk->icsk_retransmits == 0)
+		DCCP_INC_STATS_BH(DCCP_MIB_TIMEOUTS);
+
+	if (dccp_retransmit_skb(sk, sk->sk_send_head) < 0) {
+		/*
+		 * Retransmission failed because of local congestion,
+		 * do not backoff.
+		 */
+		if (icsk->icsk_retransmits == 0)
+			icsk->icsk_retransmits = 1;
+		inet_csk_reset_xmit_timer(sk, ICSK_TIME_RETRANS,
+					  min(icsk->icsk_rto,
+					      TCP_RESOURCE_PROBE_INTERVAL),
+					  TCP_RTO_MAX);
+		goto out;
+	}
+
+	icsk->icsk_backoff++;
+	icsk->icsk_retransmits++;
+
+	icsk->icsk_rto = min(icsk->icsk_rto << 1, DCCP_RTO_MAX);
+	inet_csk_reset_xmit_timer(sk, ICSK_TIME_RETRANS, icsk->icsk_rto, TCP_RTO_MAX);
+	if (icsk->icsk_retransmits > 3 /* FIXME: sysctl_dccp_retries1 */)
+		__sk_dst_reset(sk);
+out:;
+}
+
+static void dccp_write_timer(unsigned long data)
+{
+	struct sock *sk = (struct sock *)data;
+	struct inet_connection_sock *icsk = inet_csk(sk);
+	int event = 0;
+
+	bh_lock_sock(sk);
+	if (sock_owned_by_user(sk)) {
+		/* Try again later */
+		sk_reset_timer(sk, &icsk->icsk_retransmit_timer, jiffies + (HZ / 20));
+		goto out;
+	}
+
+	if (sk->sk_state == DCCP_CLOSED || !icsk->icsk_pending)
+		goto out;
+
+	if (time_after(icsk->icsk_timeout, jiffies)) {
+		sk_reset_timer(sk, &icsk->icsk_retransmit_timer, icsk->icsk_timeout);
+		goto out;
+	}
+
+	event = icsk->icsk_pending;
+	icsk->icsk_pending = 0;
+
+	switch (event) {
+	case ICSK_TIME_RETRANS:
+		dccp_retransmit_timer(sk);
+		break;
+	}
+out:
+	bh_unlock_sock(sk);
+	sock_put(sk);
+}
+
+/*
+ *	Timer for listening sockets
+ */
+static void dccp_response_timer(struct sock *sk)
+{
+	struct inet_connection_sock *icsk = inet_csk(sk);
+	const int max_retries = icsk->icsk_syn_retries ? : TCP_SYNACK_RETRIES /* FIXME sysctl_tcp_synack_retries */;
+
+	reqsk_queue_prune(&icsk->icsk_accept_queue, sk, TCP_SYNQ_INTERVAL,
+			  DCCP_TIMEOUT_INIT, DCCP_RTO_MAX, max_retries);
+}
+
+static void dccp_keepalive_timer(unsigned long data)
+{
+	struct sock *sk = (struct sock *)data;
+
+	/* Only process if socket is not in use. */
+	bh_lock_sock(sk);
+	if (sock_owned_by_user(sk)) {
+		/* Try again later. */ 
+		inet_csk_reset_keepalive_timer(sk, HZ / 20);
+		goto out;
+	}
+
+	if (sk->sk_state == DCCP_LISTEN) {
+		dccp_response_timer(sk);
+		goto out;
+	}
+out:
+	bh_unlock_sock(sk);
+	sock_put(sk);
+}