tcp.h 49.1 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8 9
/*
 * INET		An implementation of the TCP/IP protocol suite for the LINUX
 *		operating system.  INET is implemented using the  BSD Socket
 *		interface as the means of communication with the user level.
 *
 *		Definitions for the TCP module.
 *
 * Version:	@(#)tcp.h	1.0.5	05/23/93
 *
10
 * Authors:	Ross Biro
L
Linus Torvalds 已提交
11 12 13 14 15 16 17 18 19 20 21 22 23 24
 *		Fred N. van Kempen, <waltje@uWalt.NL.Mugnet.ORG>
 *
 *		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.
 */
#ifndef _TCP_H
#define _TCP_H

#define FASTRETRANS_DEBUG 1

#include <linux/list.h>
#include <linux/tcp.h>
25
#include <linux/bug.h>
L
Linus Torvalds 已提交
26 27 28
#include <linux/slab.h>
#include <linux/cache.h>
#include <linux/percpu.h>
29
#include <linux/skbuff.h>
30
#include <linux/dmaengine.h>
31
#include <linux/crypto.h>
32
#include <linux/cryptohash.h>
33
#include <linux/kref.h>
34 35

#include <net/inet_connection_sock.h>
36
#include <net/inet_timewait_sock.h>
37
#include <net/inet_hashtables.h>
L
Linus Torvalds 已提交
38
#include <net/checksum.h>
39
#include <net/request_sock.h>
L
Linus Torvalds 已提交
40 41 42
#include <net/sock.h>
#include <net/snmp.h>
#include <net/ip.h>
43
#include <net/tcp_states.h>
44
#include <net/inet_ecn.h>
45
#include <net/dst.h>
46

L
Linus Torvalds 已提交
47
#include <linux/seq_file.h>
48
#include <linux/memcontrol.h>
L
Linus Torvalds 已提交
49

50
extern struct inet_hashinfo tcp_hashinfo;
L
Linus Torvalds 已提交
51

52
extern struct percpu_counter tcp_orphan_count;
L
Linus Torvalds 已提交
53 54 55
extern void tcp_time_wait(struct sock *sk, int state, int timeo);

#define MAX_TCP_HEADER	(128 + MAX_HEADER)
A
Adam Langley 已提交
56
#define MAX_TCP_OPTION_SPACE 40
L
Linus Torvalds 已提交
57 58 59 60 61 62 63

/* 
 * Never offer a window over 32767 without using window scaling. Some
 * poor stacks do signed 16bit maths! 
 */
#define MAX_TCP_WINDOW		32767U

64 65 66
/* Offer an initial receive window of 10 mss. */
#define TCP_DEFAULT_INIT_RCVWND	10

L
Linus Torvalds 已提交
67 68 69
/* Minimal accepted MSS. It is (60+60+8) - (20+20). */
#define TCP_MIN_MSS		88U

J
John Heffner 已提交
70 71 72
/* The least MTU to use for probing */
#define TCP_BASE_MSS		512

L
Linus Torvalds 已提交
73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101
/* After receiving this amount of duplicate ACKs fast retransmit starts. */
#define TCP_FASTRETRANS_THRESH 3

/* Maximal reordering. */
#define TCP_MAX_REORDERING	127

/* Maximal number of ACKs sent quickly to accelerate slow-start. */
#define TCP_MAX_QUICKACKS	16U

/* urg_data states */
#define TCP_URG_VALID	0x0100
#define TCP_URG_NOTYET	0x0200
#define TCP_URG_READ	0x0400

#define TCP_RETR1	3	/*
				 * This is how many retries it does before it
				 * tries to figure out if the gateway is
				 * down. Minimal RFC value is 3; it corresponds
				 * to ~3sec-8min depending on RTO.
				 */

#define TCP_RETR2	15	/*
				 * This should take at least
				 * 90 minutes to time out.
				 * RFC1122 says that the limit is 100 sec.
				 * 15 is ~13-30min depending on RTO.
				 */

#define TCP_SYN_RETRIES	 5	/* number of times to retry active opening a
S
Stephen Hemminger 已提交
102
				 * connection: ~180sec is RFC minimum	*/
L
Linus Torvalds 已提交
103 104

#define TCP_SYNACK_RETRIES 5	/* number of times to retry passive opening a
S
Stephen Hemminger 已提交
105
				 * connection: ~180sec is RFC minimum	*/
L
Linus Torvalds 已提交
106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125

#define TCP_TIMEWAIT_LEN (60*HZ) /* how long to wait to destroy TIME-WAIT
				  * state, about 60 seconds	*/
#define TCP_FIN_TIMEOUT	TCP_TIMEWAIT_LEN
                                 /* BSD style FIN_WAIT2 deadlock breaker.
				  * It used to be 3min, new value is 60sec,
				  * to combine FIN-WAIT-2 timeout with
				  * TIME-WAIT timer.
				  */

#define TCP_DELACK_MAX	((unsigned)(HZ/5))	/* maximal time to delay before sending an ACK */
#if HZ >= 100
#define TCP_DELACK_MIN	((unsigned)(HZ/25))	/* minimal time to delay before sending an ACK */
#define TCP_ATO_MIN	((unsigned)(HZ/25))
#else
#define TCP_DELACK_MIN	4U
#define TCP_ATO_MIN	4U
#endif
#define TCP_RTO_MAX	((unsigned)(120*HZ))
#define TCP_RTO_MIN	((unsigned)(HZ/5))
126
#define TCP_TIMEOUT_INIT ((unsigned)(1*HZ))	/* RFC6298 2.1 initial RTO value	*/
127 128 129 130 131 132
#define TCP_TIMEOUT_FALLBACK ((unsigned)(3*HZ))	/* RFC 1122 initial RTO value, now
						 * used as a fallback RTO for the
						 * initial data transmission if no
						 * valid RTT sample has been acquired,
						 * most likely due to retrans in 3WHS.
						 */
L
Linus Torvalds 已提交
133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170

#define TCP_RESOURCE_PROBE_INTERVAL ((unsigned)(HZ/2U)) /* Maximal interval between probes
					                 * for local resources.
					                 */

#define TCP_KEEPALIVE_TIME	(120*60*HZ)	/* two hours */
#define TCP_KEEPALIVE_PROBES	9		/* Max of 9 keepalive probes	*/
#define TCP_KEEPALIVE_INTVL	(75*HZ)

#define MAX_TCP_KEEPIDLE	32767
#define MAX_TCP_KEEPINTVL	32767
#define MAX_TCP_KEEPCNT		127
#define MAX_TCP_SYNCNT		127

#define TCP_SYNQ_INTERVAL	(HZ/5)	/* Period of SYNACK timer */

#define TCP_PAWS_24DAYS	(60 * 60 * 24 * 24)
#define TCP_PAWS_MSL	60		/* Per-host timestamps are invalidated
					 * after this time. It should be equal
					 * (or greater than) TCP_TIMEWAIT_LEN
					 * to provide reliability equal to one
					 * provided by timewait state.
					 */
#define TCP_PAWS_WINDOW	1		/* Replay window for per-host
					 * timestamps. It must be less than
					 * minimal timewait lifetime.
					 */
/*
 *	TCP option
 */
 
#define TCPOPT_NOP		1	/* Padding */
#define TCPOPT_EOL		0	/* End of options */
#define TCPOPT_MSS		2	/* Segment size negotiating */
#define TCPOPT_WINDOW		3	/* Window scaling */
#define TCPOPT_SACK_PERM        4       /* SACK Permitted */
#define TCPOPT_SACK             5       /* SACK Block */
#define TCPOPT_TIMESTAMP	8	/* Better RTT estimations/PAWS */
171
#define TCPOPT_MD5SIG		19	/* MD5 Signature (RFC2385) */
172
#define TCPOPT_COOKIE		253	/* Cookie extension (experimental) */
L
Linus Torvalds 已提交
173 174 175 176 177 178 179 180 181

/*
 *     TCP option lengths
 */

#define TCPOLEN_MSS            4
#define TCPOLEN_WINDOW         3
#define TCPOLEN_SACK_PERM      2
#define TCPOLEN_TIMESTAMP      10
182
#define TCPOLEN_MD5SIG         18
183 184 185 186
#define TCPOLEN_COOKIE_BASE    2	/* Cookie-less header extension */
#define TCPOLEN_COOKIE_PAIR    3	/* Cookie pair header extension */
#define TCPOLEN_COOKIE_MIN     (TCPOLEN_COOKIE_BASE+TCP_COOKIE_MIN)
#define TCPOLEN_COOKIE_MAX     (TCPOLEN_COOKIE_BASE+TCP_COOKIE_MAX)
L
Linus Torvalds 已提交
187 188 189 190 191 192 193 194

/* But this is what stacks really send out. */
#define TCPOLEN_TSTAMP_ALIGNED		12
#define TCPOLEN_WSCALE_ALIGNED		4
#define TCPOLEN_SACKPERM_ALIGNED	4
#define TCPOLEN_SACK_BASE		2
#define TCPOLEN_SACK_BASE_ALIGNED	4
#define TCPOLEN_SACK_PERBLOCK		8
195
#define TCPOLEN_MD5SIG_ALIGNED		20
A
Adam Langley 已提交
196
#define TCPOLEN_MSS_ALIGNED		4
L
Linus Torvalds 已提交
197 198 199 200

/* Flags in tp->nonagle */
#define TCP_NAGLE_OFF		1	/* Nagle's algo is disabled */
#define TCP_NAGLE_CORK		2	/* Socket is corked	    */
S
Stephen Hemminger 已提交
201
#define TCP_NAGLE_PUSH		4	/* Cork is overridden for already queued data */
L
Linus Torvalds 已提交
202

A
Andreas Petlund 已提交
203 204 205
/* TCP thin-stream limits */
#define TCP_THIN_LINEAR_RETRIES 6       /* After 6 linear retries, do exp. backoff */

206
/* TCP initial congestion window as per draft-hkchu-tcpm-initcwnd-01 */
207 208
#define TCP_INIT_CWND		10

209 210
extern struct inet_timewait_death_row tcp_death_row;

L
Linus Torvalds 已提交
211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239
/* sysctl variables for tcp */
extern int sysctl_tcp_timestamps;
extern int sysctl_tcp_window_scaling;
extern int sysctl_tcp_sack;
extern int sysctl_tcp_fin_timeout;
extern int sysctl_tcp_keepalive_time;
extern int sysctl_tcp_keepalive_probes;
extern int sysctl_tcp_keepalive_intvl;
extern int sysctl_tcp_syn_retries;
extern int sysctl_tcp_synack_retries;
extern int sysctl_tcp_retries1;
extern int sysctl_tcp_retries2;
extern int sysctl_tcp_orphan_retries;
extern int sysctl_tcp_syncookies;
extern int sysctl_tcp_retrans_collapse;
extern int sysctl_tcp_stdurg;
extern int sysctl_tcp_rfc1337;
extern int sysctl_tcp_abort_on_overflow;
extern int sysctl_tcp_max_orphans;
extern int sysctl_tcp_fack;
extern int sysctl_tcp_reordering;
extern int sysctl_tcp_ecn;
extern int sysctl_tcp_dsack;
extern int sysctl_tcp_wmem[3];
extern int sysctl_tcp_rmem[3];
extern int sysctl_tcp_app_win;
extern int sysctl_tcp_adv_win_scale;
extern int sysctl_tcp_tw_reuse;
extern int sysctl_tcp_frto;
240
extern int sysctl_tcp_frto_response;
L
Linus Torvalds 已提交
241
extern int sysctl_tcp_low_latency;
242
extern int sysctl_tcp_dma_copybreak;
L
Linus Torvalds 已提交
243 244 245
extern int sysctl_tcp_nometrics_save;
extern int sysctl_tcp_moderate_rcvbuf;
extern int sysctl_tcp_tso_win_divisor;
246
extern int sysctl_tcp_abc;
J
John Heffner 已提交
247 248
extern int sysctl_tcp_mtu_probing;
extern int sysctl_tcp_base_mss;
249
extern int sysctl_tcp_workaround_signed_windows;
250
extern int sysctl_tcp_slow_start_after_idle;
251
extern int sysctl_tcp_max_ssthresh;
252
extern int sysctl_tcp_cookie_size;
A
Andreas Petlund 已提交
253
extern int sysctl_tcp_thin_linear_timeouts;
A
Andreas Petlund 已提交
254
extern int sysctl_tcp_thin_dupack;
Y
Yuchung Cheng 已提交
255
extern int sysctl_tcp_early_retrans;
L
Linus Torvalds 已提交
256

E
Eric Dumazet 已提交
257
extern atomic_long_t tcp_memory_allocated;
258
extern struct percpu_counter tcp_sockets_allocated;
L
Linus Torvalds 已提交
259 260 261 262 263 264 265
extern int tcp_memory_pressure;

/*
 * The next routines deal with comparing 32 bit unsigned ints
 * and worry about wraparound (automatic with unsigned arithmetic).
 */

E
Eric Dumazet 已提交
266
static inline bool before(__u32 seq1, __u32 seq2)
L
Linus Torvalds 已提交
267
{
268
        return (__s32)(seq1-seq2) < 0;
L
Linus Torvalds 已提交
269
}
270
#define after(seq2, seq1) 	before(seq1, seq2)
L
Linus Torvalds 已提交
271 272

/* is s2<=s1<=s3 ? */
E
Eric Dumazet 已提交
273
static inline bool between(__u32 seq1, __u32 seq2, __u32 seq3)
L
Linus Torvalds 已提交
274 275 276 277
{
	return seq3 - seq2 >= seq1 - seq2;
}

A
Arun Sharma 已提交
278 279 280 281 282 283 284 285
static inline bool tcp_out_of_memory(struct sock *sk)
{
	if (sk->sk_wmem_queued > SOCK_MIN_SNDBUF &&
	    sk_memory_allocated(sk) > sk_prot_mem_limits(sk, 2))
		return true;
	return false;
}

286
static inline bool tcp_too_many_orphans(struct sock *sk, int shift)
287
{
288 289 290 291 292 293 294 295 296
	struct percpu_counter *ocp = sk->sk_prot->orphan_count;
	int orphans = percpu_counter_read_positive(ocp);

	if (orphans << shift > sysctl_tcp_max_orphans) {
		orphans = percpu_counter_sum_positive(ocp);
		if (orphans << shift > sysctl_tcp_max_orphans)
			return true;
	}
	return false;
297
}
L
Linus Torvalds 已提交
298

A
Arun Sharma 已提交
299 300
extern bool tcp_check_oom(struct sock *sk, int shift);

301 302 303 304 305 306 307
/* syncookies: remember time of last synqueue overflow */
static inline void tcp_synq_overflow(struct sock *sk)
{
	tcp_sk(sk)->rx_opt.ts_recent_stamp = jiffies;
}

/* syncookies: no recent synqueue overflow on this listening socket? */
E
Eric Dumazet 已提交
308
static inline bool tcp_synq_no_recent_overflow(const struct sock *sk)
309 310
{
	unsigned long last_overflow = tcp_sk(sk)->rx_opt.ts_recent_stamp;
311
	return time_after(jiffies, last_overflow + TCP_TIMEOUT_FALLBACK);
312 313
}

L
Linus Torvalds 已提交
314 315
extern struct proto tcp_prot;

316 317 318 319
#define TCP_INC_STATS(net, field)	SNMP_INC_STATS((net)->mib.tcp_statistics, field)
#define TCP_INC_STATS_BH(net, field)	SNMP_INC_STATS_BH((net)->mib.tcp_statistics, field)
#define TCP_DEC_STATS(net, field)	SNMP_DEC_STATS((net)->mib.tcp_statistics, field)
#define TCP_ADD_STATS_USER(net, field, val) SNMP_ADD_STATS_USER((net)->mib.tcp_statistics, field, val)
320
#define TCP_ADD_STATS(net, field, val)	SNMP_ADD_STATS((net)->mib.tcp_statistics, field, val)
L
Linus Torvalds 已提交
321

322 323
extern void tcp_init_mem(struct net *net);

C
Changli Gao 已提交
324 325 326 327
extern void tcp_v4_err(struct sk_buff *skb, u32);

extern void tcp_shutdown (struct sock *sk, int how);

328
extern void tcp_v4_early_demux(struct sk_buff *skb);
C
Changli Gao 已提交
329 330
extern int tcp_v4_rcv(struct sk_buff *skb);

331
extern struct inet_peer *tcp_v4_get_peer(struct sock *sk);
C
Changli Gao 已提交
332
extern int tcp_v4_tw_remember_stamp(struct inet_timewait_sock *tw);
333 334 335 336
extern int tcp_sendmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg,
		       size_t size);
extern int tcp_sendpage(struct sock *sk, struct page *page, int offset,
			size_t size, int flags);
C
Changli Gao 已提交
337 338
extern int tcp_ioctl(struct sock *sk, int cmd, unsigned long arg);
extern int tcp_rcv_state_process(struct sock *sk, struct sk_buff *skb,
339
				 const struct tcphdr *th, unsigned int len);
C
Changli Gao 已提交
340
extern int tcp_rcv_established(struct sock *sk, struct sk_buff *skb,
341
			       const struct tcphdr *th, unsigned int len);
C
Changli Gao 已提交
342 343 344 345 346 347 348
extern void tcp_rcv_space_adjust(struct sock *sk);
extern void tcp_cleanup_rbuf(struct sock *sk, int copied);
extern int tcp_twsk_unique(struct sock *sk, struct sock *sktw, void *twp);
extern void tcp_twsk_destructor(struct sock *sk);
extern ssize_t tcp_splice_read(struct socket *sk, loff_t *ppos,
			       struct pipe_inode_info *pipe, size_t len,
			       unsigned int flags);
J
Jens Axboe 已提交
349

350 351
static inline void tcp_dec_quickack_mode(struct sock *sk,
					 const unsigned int pkts)
L
Linus Torvalds 已提交
352
{
353
	struct inet_connection_sock *icsk = inet_csk(sk);
354

355 356 357
	if (icsk->icsk_ack.quick) {
		if (pkts >= icsk->icsk_ack.quick) {
			icsk->icsk_ack.quick = 0;
358
			/* Leaving quickack mode we deflate ATO. */
359
			icsk->icsk_ack.ato   = TCP_ATO_MIN;
360
		} else
361
			icsk->icsk_ack.quick -= pkts;
L
Linus Torvalds 已提交
362 363 364
	}
}

365 366 367
#define	TCP_ECN_OK		1
#define	TCP_ECN_QUEUE_CWR	2
#define	TCP_ECN_DEMAND_CWR	4
368
#define	TCP_ECN_SEEN		8
369

E
Eric Dumazet 已提交
370
enum tcp_tw_status {
L
Linus Torvalds 已提交
371 372 373 374 375 376 377
	TCP_TW_SUCCESS = 0,
	TCP_TW_RST = 1,
	TCP_TW_ACK = 2,
	TCP_TW_SYN = 3
};


C
Changli Gao 已提交
378 379 380 381 382 383 384 385
extern enum tcp_tw_status tcp_timewait_state_process(struct inet_timewait_sock *tw,
						     struct sk_buff *skb,
						     const struct tcphdr *th);
extern struct sock * tcp_check_req(struct sock *sk,struct sk_buff *skb,
				   struct request_sock *req,
				   struct request_sock **prev);
extern int tcp_child_process(struct sock *parent, struct sock *child,
			     struct sk_buff *skb);
E
Eric Dumazet 已提交
386
extern bool tcp_use_frto(struct sock *sk);
C
Changli Gao 已提交
387 388 389 390
extern void tcp_enter_frto(struct sock *sk);
extern void tcp_enter_loss(struct sock *sk, int how);
extern void tcp_clear_retrans(struct tcp_sock *tp);
extern void tcp_update_metrics(struct sock *sk);
391
extern void tcp_init_metrics(struct sock *sk);
392
extern void tcp_metrics_init(void);
393 394 395 396
extern bool tcp_peer_is_proven(struct request_sock *req, struct dst_entry *dst, bool paws_check);
extern bool tcp_remember_stamp(struct sock *sk);
extern bool tcp_tw_remember_stamp(struct inet_timewait_sock *tw);
extern void tcp_fetch_timewait_stamp(struct sock *sk, struct dst_entry *dst);
397
extern void tcp_disable_fack(struct tcp_sock *tp);
C
Changli Gao 已提交
398
extern void tcp_close(struct sock *sk, long timeout);
399
extern void tcp_init_sock(struct sock *sk);
C
Changli Gao 已提交
400 401 402 403 404 405 406 407 408 409 410 411 412 413
extern unsigned int tcp_poll(struct file * file, struct socket *sock,
			     struct poll_table_struct *wait);
extern int tcp_getsockopt(struct sock *sk, int level, int optname,
			  char __user *optval, int __user *optlen);
extern int tcp_setsockopt(struct sock *sk, int level, int optname,
			  char __user *optval, unsigned int optlen);
extern int compat_tcp_getsockopt(struct sock *sk, int level, int optname,
				 char __user *optval, int __user *optlen);
extern int compat_tcp_setsockopt(struct sock *sk, int level, int optname,
				 char __user *optval, unsigned int optlen);
extern void tcp_set_keepalive(struct sock *sk, int val);
extern void tcp_syn_ack_timeout(struct sock *sk, struct request_sock *req);
extern int tcp_recvmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg,
		       size_t len, int nonblock, int flags, int *addr_len);
414 415
extern void tcp_parse_options(const struct sk_buff *skb,
			      struct tcp_options_received *opt_rx, const u8 **hvpp,
C
Changli Gao 已提交
416
			      int estab);
417
extern const u8 *tcp_parse_md5sig_option(const struct tcphdr *th);
418

L
Linus Torvalds 已提交
419 420 421 422
/*
 *	TCP v4 functions exported for the inet6 API
 */

C
Changli Gao 已提交
423 424 425 426
extern void tcp_v4_send_check(struct sock *sk, struct sk_buff *skb);
extern int tcp_v4_conn_request(struct sock *sk, struct sk_buff *skb);
extern struct sock * tcp_create_openreq_child(struct sock *sk,
					      struct request_sock *req,
L
Linus Torvalds 已提交
427
					      struct sk_buff *skb);
C
Changli Gao 已提交
428 429 430 431 432 433 434 435 436 437 438
extern struct sock * tcp_v4_syn_recv_sock(struct sock *sk, struct sk_buff *skb,
					  struct request_sock *req,
					  struct dst_entry *dst);
extern int tcp_v4_do_rcv(struct sock *sk, struct sk_buff *skb);
extern int tcp_v4_connect(struct sock *sk, struct sockaddr *uaddr,
			  int addr_len);
extern int tcp_connect(struct sock *sk);
extern struct sk_buff * tcp_make_synack(struct sock *sk, struct dst_entry *dst,
					struct request_sock *req,
					struct request_values *rvp);
extern int tcp_disconnect(struct sock *sk, int flags);
L
Linus Torvalds 已提交
439

P
Pavel Emelyanov 已提交
440 441
void tcp_connect_init(struct sock *sk);
void tcp_finish_connect(struct sock *sk, struct sk_buff *skb);
442
int tcp_send_rcvq(struct sock *sk, struct msghdr *msg, size_t size);
L
Linus Torvalds 已提交
443 444

/* From syncookies.c */
445
extern __u32 syncookie_secret[2][16-4+SHA_DIGEST_WORDS];
L
Linus Torvalds 已提交
446 447
extern struct sock *cookie_v4_check(struct sock *sk, struct sk_buff *skb, 
				    struct ip_options *opt);
448
#ifdef CONFIG_SYN_COOKIES
L
Linus Torvalds 已提交
449 450
extern __u32 cookie_v4_init_sequence(struct sock *sk, struct sk_buff *skb, 
				     __u16 *mss);
451 452 453 454 455 456 457 458
#else
static inline __u32 cookie_v4_init_sequence(struct sock *sk,
					    struct sk_buff *skb,
					    __u16 *mss)
{
	return 0;
}
#endif
L
Linus Torvalds 已提交
459

460
extern __u32 cookie_init_timestamp(struct request_sock *req);
461
extern bool cookie_check_timestamp(struct tcp_options_received *opt, bool *);
462

463 464
/* From net/ipv6/syncookies.c */
extern struct sock *cookie_v6_check(struct sock *sk, struct sk_buff *skb);
465
#ifdef CONFIG_SYN_COOKIES
466
extern __u32 cookie_v6_init_sequence(struct sock *sk, const struct sk_buff *skb,
467
				     __u16 *mss);
468 469 470 471 472 473 474 475
#else
static inline __u32 cookie_v6_init_sequence(struct sock *sk,
					    struct sk_buff *skb,
					    __u16 *mss)
{
	return 0;
}
#endif
L
Linus Torvalds 已提交
476 477
/* tcp_output.c */

478 479
extern void __tcp_push_pending_frames(struct sock *sk, unsigned int cur_mss,
				      int nonagle);
E
Eric Dumazet 已提交
480
extern bool tcp_may_send_now(struct sock *sk);
L
Linus Torvalds 已提交
481
extern int tcp_retransmit_skb(struct sock *, struct sk_buff *);
482
extern void tcp_retransmit_timer(struct sock *sk);
L
Linus Torvalds 已提交
483 484 485
extern void tcp_xmit_retransmit_queue(struct sock *);
extern void tcp_simple_retransmit(struct sock *);
extern int tcp_trim_head(struct sock *, struct sk_buff *, u32);
486
extern int tcp_fragment(struct sock *, struct sk_buff *, u32, unsigned int);
L
Linus Torvalds 已提交
487 488 489

extern void tcp_send_probe0(struct sock *);
extern void tcp_send_partial(struct sock *);
C
Changli Gao 已提交
490
extern int tcp_write_wakeup(struct sock *);
L
Linus Torvalds 已提交
491
extern void tcp_send_fin(struct sock *sk);
A
Al Viro 已提交
492
extern void tcp_send_active_reset(struct sock *sk, gfp_t priority);
C
Changli Gao 已提交
493
extern int tcp_send_synack(struct sock *);
E
Eric Dumazet 已提交
494 495 496
extern bool tcp_syn_flood_action(struct sock *sk,
				 const struct sk_buff *skb,
				 const char *proto);
497
extern void tcp_push_one(struct sock *, unsigned int mss_now);
L
Linus Torvalds 已提交
498 499 500
extern void tcp_send_ack(struct sock *sk);
extern void tcp_send_delayed_ack(struct sock *sk);

501 502
/* tcp_input.c */
extern void tcp_cwnd_application_limited(struct sock *sk);
503 504
extern void tcp_resume_early_retransmit(struct sock *sk);
extern void tcp_rearm_rto(struct sock *sk);
505

L
Linus Torvalds 已提交
506 507
/* tcp_timer.c */
extern void tcp_init_xmit_timers(struct sock *);
508 509 510 511
static inline void tcp_clear_xmit_timers(struct sock *sk)
{
	inet_csk_clear_xmit_timers(sk);
}
L
Linus Torvalds 已提交
512 513

extern unsigned int tcp_sync_mss(struct sock *sk, u32 pmtu);
I
Ilpo Järvinen 已提交
514 515 516 517 518
extern unsigned int tcp_current_mss(struct sock *sk);

/* Bound MSS / TSO packet size with the half of the window */
static inline int tcp_bound_to_half_wnd(struct tcp_sock *tp, int pktsize)
{
519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534
	int cutoff;

	/* When peer uses tiny windows, there is no use in packetizing
	 * to sub-MSS pieces for the sake of SWS or making sure there
	 * are enough packets in the pipe for fast recovery.
	 *
	 * On the other hand, for extremely large MSS devices, handling
	 * smaller than MSS windows in this way does make sense.
	 */
	if (tp->max_window >= 512)
		cutoff = (tp->max_window >> 1);
	else
		cutoff = tp->max_window;

	if (cutoff && pktsize > cutoff)
		return max_t(int, cutoff, 68U - tp->tcp_header_len);
I
Ilpo Järvinen 已提交
535 536 537
	else
		return pktsize;
}
L
Linus Torvalds 已提交
538

539
/* tcp.c */
540
extern void tcp_get_info(const struct sock *, struct tcp_info *);
L
Linus Torvalds 已提交
541 542 543 544 545 546 547

/* Read 'sendfile()'-style from a TCP socket */
typedef int (*sk_read_actor_t)(read_descriptor_t *, struct sk_buff *,
				unsigned int, size_t);
extern int tcp_read_sock(struct sock *sk, read_descriptor_t *desc,
			 sk_read_actor_t recv_actor);

S
Stephen Hemminger 已提交
548
extern void tcp_initialize_rcv_mss(struct sock *sk);
L
Linus Torvalds 已提交
549

550 551
extern int tcp_mtu_to_mss(struct sock *sk, int pmtu);
extern int tcp_mss_to_mtu(struct sock *sk, int mss);
J
John Heffner 已提交
552
extern void tcp_mtup_init(struct sock *sk);
553
extern void tcp_valid_rtt_meas(struct sock *sk, u32 seq_rtt);
J
John Heffner 已提交
554

555 556 557 558 559 560 561 562 563 564 565
static inline void tcp_bound_rto(const struct sock *sk)
{
	if (inet_csk(sk)->icsk_rto > TCP_RTO_MAX)
		inet_csk(sk)->icsk_rto = TCP_RTO_MAX;
}

static inline u32 __tcp_set_rto(const struct tcp_sock *tp)
{
	return (tp->srtt >> 3) + tp->rttvar;
}

566 567
extern void tcp_set_rto(struct sock *sk);

S
Stephen Hemminger 已提交
568
static inline void __tcp_fast_path_on(struct tcp_sock *tp, u32 snd_wnd)
L
Linus Torvalds 已提交
569 570 571 572 573 574
{
	tp->pred_flags = htonl((tp->tcp_header_len << 26) |
			       ntohl(TCP_FLAG_ACK) |
			       snd_wnd);
}

S
Stephen Hemminger 已提交
575
static inline void tcp_fast_path_on(struct tcp_sock *tp)
L
Linus Torvalds 已提交
576 577 578 579
{
	__tcp_fast_path_on(tp, tp->snd_wnd >> tp->rx_opt.snd_wscale);
}

580
static inline void tcp_fast_path_check(struct sock *sk)
L
Linus Torvalds 已提交
581
{
582 583
	struct tcp_sock *tp = tcp_sk(sk);

584
	if (skb_queue_empty(&tp->out_of_order_queue) &&
L
Linus Torvalds 已提交
585 586 587 588 589 590
	    tp->rcv_wnd &&
	    atomic_read(&sk->sk_rmem_alloc) < sk->sk_rcvbuf &&
	    !tp->urg_data)
		tcp_fast_path_on(tp);
}

591 592 593
/* Compute the actual rto_min value */
static inline u32 tcp_rto_min(struct sock *sk)
{
594
	const struct dst_entry *dst = __sk_dst_get(sk);
595 596 597 598 599 600 601
	u32 rto_min = TCP_RTO_MIN;

	if (dst && dst_metric_locked(dst, RTAX_RTO_MIN))
		rto_min = dst_metric_rtt(dst, RTAX_RTO_MIN);
	return rto_min;
}

L
Linus Torvalds 已提交
602 603 604 605
/* Compute the actual receive window we are currently advertising.
 * Rcv_nxt can be after the window if our peer push more data
 * than the offered window.
 */
S
Stephen Hemminger 已提交
606
static inline u32 tcp_receive_window(const struct tcp_sock *tp)
L
Linus Torvalds 已提交
607 608 609 610 611 612 613 614 615 616 617 618
{
	s32 win = tp->rcv_wup + tp->rcv_wnd - tp->rcv_nxt;

	if (win < 0)
		win = 0;
	return (u32) win;
}

/* Choose a new window, without checks for shrinking, and without
 * scaling applied to the result.  The caller does these things
 * if necessary.  This is a "raw" window selection.
 */
C
Changli Gao 已提交
619
extern u32 __tcp_select_window(struct sock *sk);
L
Linus Torvalds 已提交
620

P
Pavel Emelyanov 已提交
621 622
void tcp_send_window_probe(struct sock *sk);

L
Linus Torvalds 已提交
623 624
/* TCP timestamps are only 32-bits, this causes a slight
 * complication on 64-bit systems since we store a snapshot
S
Stephen Hemminger 已提交
625 626
 * of jiffies in the buffer control blocks below.  We decided
 * to use only the low 32-bits of jiffies and hide the ugly
L
Linus Torvalds 已提交
627 628 629 630
 * casts with the following macro.
 */
#define tcp_time_stamp		((__u32)(jiffies))

C
Changli Gao 已提交
631 632 633 634 635 636 637 638 639 640 641
#define tcp_flag_byte(th) (((u_int8_t *)th)[13])

#define TCPHDR_FIN 0x01
#define TCPHDR_SYN 0x02
#define TCPHDR_RST 0x04
#define TCPHDR_PSH 0x08
#define TCPHDR_ACK 0x10
#define TCPHDR_URG 0x20
#define TCPHDR_ECE 0x40
#define TCPHDR_CWR 0x80

S
Stephen Hemminger 已提交
642
/* This is what the send packet queuing engine uses to pass
643 644 645 646
 * TCP per-packet control information to the transmission code.
 * We also store the host-order sequence numbers in here too.
 * This is 44 bytes if IPV6 is enabled.
 * If this grows please adjust skbuff.h:skbuff->cb[xxx] size appropriately.
L
Linus Torvalds 已提交
647 648 649 650
 */
struct tcp_skb_cb {
	union {
		struct inet_skb_parm	h4;
E
Eric Dumazet 已提交
651
#if IS_ENABLED(CONFIG_IPV6)
L
Linus Torvalds 已提交
652 653 654 655 656 657
		struct inet6_skb_parm	h6;
#endif
	} header;	/* For incoming frames		*/
	__u32		seq;		/* Starting sequence number	*/
	__u32		end_seq;	/* SEQ + FIN + SYN + datalen	*/
	__u32		when;		/* used to compute rtt's	*/
E
Eric Dumazet 已提交
658
	__u8		tcp_flags;	/* TCP header flags. (tcp[13])	*/
659

L
Linus Torvalds 已提交
660 661 662 663 664 665 666 667
	__u8		sacked;		/* State flags for SACK/FACK.	*/
#define TCPCB_SACKED_ACKED	0x01	/* SKB ACK'd by a SACK block	*/
#define TCPCB_SACKED_RETRANS	0x02	/* SKB retransmitted		*/
#define TCPCB_LOST		0x04	/* SKB is lost			*/
#define TCPCB_TAGBITS		0x07	/* All tag bits			*/
#define TCPCB_EVER_RETRANS	0x80	/* Ever retransmitted frame	*/
#define TCPCB_RETRANS		(TCPCB_SACKED_RETRANS|TCPCB_EVER_RETRANS)

668 669
	__u8		ip_dsfield;	/* IPv4 tos or IPv6 dsfield	*/
	/* 1 byte hole */
L
Linus Torvalds 已提交
670 671 672 673 674
	__u32		ack_seq;	/* Sequence number ACK'd	*/
};

#define TCP_SKB_CB(__skb)	((struct tcp_skb_cb *)&((__skb)->cb[0]))

675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690
/* RFC3168 : 6.1.1 SYN packets must not have ECT/ECN bits set
 *
 * If we receive a SYN packet with these bits set, it means a network is
 * playing bad games with TOS bits. In order to avoid possible false congestion
 * notifications, we disable TCP ECN negociation.
 */
static inline void
TCP_ECN_create_request(struct request_sock *req, const struct sk_buff *skb)
{
	const struct tcphdr *th = tcp_hdr(skb);

	if (sysctl_tcp_ecn && th->ece && th->cwr &&
	    INET_ECN_is_not_ect(TCP_SKB_CB(skb)->ip_dsfield))
		inet_rsk(req)->ecn_ok = 1;
}

L
Linus Torvalds 已提交
691 692 693 694 695
/* Due to TSO, an SKB can be composed of multiple actual
 * packets.  To keep these tracked properly, we use this.
 */
static inline int tcp_skb_pcount(const struct sk_buff *skb)
{
696
	return skb_shinfo(skb)->gso_segs;
L
Linus Torvalds 已提交
697 698 699 700 701
}

/* This is valid iff tcp_skb_pcount() > 1. */
static inline int tcp_skb_mss(const struct sk_buff *skb)
{
702
	return skb_shinfo(skb)->gso_size;
L
Linus Torvalds 已提交
703 704
}

705 706 707 708 709 710 711 712 713 714 715 716 717 718 719
/* Events passed to congestion control interface */
enum tcp_ca_event {
	CA_EVENT_TX_START,	/* first transmit when no packets in flight */
	CA_EVENT_CWND_RESTART,	/* congestion window restart */
	CA_EVENT_COMPLETE_CWR,	/* end of congestion recovery */
	CA_EVENT_FRTO,		/* fast recovery timeout */
	CA_EVENT_LOSS,		/* loss timeout */
	CA_EVENT_FAST_ACK,	/* in sequence ack */
	CA_EVENT_SLOW_ACK,	/* other ack */
};

/*
 * Interface for adding new TCP congestion control handlers
 */
#define TCP_CA_NAME_MAX	16
720 721 722
#define TCP_CA_MAX	128
#define TCP_CA_BUF_MAX	(TCP_CA_NAME_MAX*TCP_CA_MAX)

723 724 725
#define TCP_CONG_NON_RESTRICTED 0x1
#define TCP_CONG_RTT_STAMP	0x2

726 727
struct tcp_congestion_ops {
	struct list_head	list;
728
	unsigned long flags;
729 730

	/* initialize private data (optional) */
731
	void (*init)(struct sock *sk);
732
	/* cleanup private data  (optional) */
733
	void (*release)(struct sock *sk);
734 735

	/* return slow start threshold (required) */
736
	u32 (*ssthresh)(struct sock *sk);
737
	/* lower bound for congestion window (optional) */
738
	u32 (*min_cwnd)(const struct sock *sk);
739
	/* do new cwnd calculation (required) */
740
	void (*cong_avoid)(struct sock *sk, u32 ack, u32 in_flight);
741
	/* call before changing ca_state (optional) */
742
	void (*set_state)(struct sock *sk, u8 new_state);
743
	/* call when cwnd event occurs (optional) */
744
	void (*cwnd_event)(struct sock *sk, enum tcp_ca_event ev);
745
	/* new value of cwnd after loss (optional) */
746
	u32  (*undo_cwnd)(struct sock *sk);
747
	/* hook for packet ack accounting (optional) */
748
	void (*pkts_acked)(struct sock *sk, u32 num_acked, s32 rtt_us);
749
	/* get info for inet_diag (optional) */
750
	void (*get_info)(struct sock *sk, u32 ext, struct sk_buff *skb);
751 752 753 754 755 756 757 758

	char 		name[TCP_CA_NAME_MAX];
	struct module 	*owner;
};

extern int tcp_register_congestion_control(struct tcp_congestion_ops *type);
extern void tcp_unregister_congestion_control(struct tcp_congestion_ops *type);

759 760
extern void tcp_init_congestion_control(struct sock *sk);
extern void tcp_cleanup_congestion_control(struct sock *sk);
761 762
extern int tcp_set_default_congestion_control(const char *name);
extern void tcp_get_default_congestion_control(char *name);
763
extern void tcp_get_available_congestion_control(char *buf, size_t len);
764 765
extern void tcp_get_allowed_congestion_control(char *buf, size_t len);
extern int tcp_set_allowed_congestion_control(char *allowed);
766
extern int tcp_set_congestion_control(struct sock *sk, const char *name);
S
Stephen Hemminger 已提交
767
extern void tcp_slow_start(struct tcp_sock *tp);
768
extern void tcp_cong_avoid_ai(struct tcp_sock *tp, u32 w);
769

770
extern struct tcp_congestion_ops tcp_init_congestion_ops;
771
extern u32 tcp_reno_ssthresh(struct sock *sk);
772
extern void tcp_reno_cong_avoid(struct sock *sk, u32 ack, u32 in_flight);
773
extern u32 tcp_reno_min_cwnd(const struct sock *sk);
774
extern struct tcp_congestion_ops tcp_reno;
775

776
static inline void tcp_set_ca_state(struct sock *sk, const u8 ca_state)
777
{
778 779 780 781 782
	struct inet_connection_sock *icsk = inet_csk(sk);

	if (icsk->icsk_ca_ops->set_state)
		icsk->icsk_ca_ops->set_state(sk, ca_state);
	icsk->icsk_ca_state = ca_state;
783 784
}

785
static inline void tcp_ca_event(struct sock *sk, const enum tcp_ca_event event)
786
{
787 788 789 790
	const struct inet_connection_sock *icsk = inet_csk(sk);

	if (icsk->icsk_ca_ops->cwnd_event)
		icsk->icsk_ca_ops->cwnd_event(sk, event);
791 792
}

793 794 795 796 797 798 799 800 801 802 803 804 805
/* These functions determine how the current flow behaves in respect of SACK
 * handling. SACK is negotiated with the peer, and therefore it can vary
 * between different flows.
 *
 * tcp_is_sack - SACK enabled
 * tcp_is_reno - No SACK
 * tcp_is_fack - FACK enabled, implies SACK enabled
 */
static inline int tcp_is_sack(const struct tcp_sock *tp)
{
	return tp->rx_opt.sack_ok;
}

E
Eric Dumazet 已提交
806
static inline bool tcp_is_reno(const struct tcp_sock *tp)
807 808 809 810
{
	return !tcp_is_sack(tp);
}

E
Eric Dumazet 已提交
811
static inline bool tcp_is_fack(const struct tcp_sock *tp)
812
{
813
	return tp->rx_opt.sack_ok & TCP_FACK_ENABLED;
814 815 816 817
}

static inline void tcp_enable_fack(struct tcp_sock *tp)
{
818
	tp->rx_opt.sack_ok |= TCP_FACK_ENABLED;
819 820
}

Y
Yuchung Cheng 已提交
821 822 823 824 825 826 827
/* TCP early-retransmit (ER) is similar to but more conservative than
 * the thin-dupack feature.  Enable ER only if thin-dupack is disabled.
 */
static inline void tcp_enable_early_retrans(struct tcp_sock *tp)
{
	tp->do_early_retrans = sysctl_tcp_early_retrans &&
		!sysctl_tcp_thin_dupack && sysctl_tcp_reordering == 3;
828
	tp->early_retrans_delayed = 0;
Y
Yuchung Cheng 已提交
829 830 831 832 833 834 835
}

static inline void tcp_disable_early_retrans(struct tcp_sock *tp)
{
	tp->do_early_retrans = 0;
}

836 837 838 839 840
static inline unsigned int tcp_left_out(const struct tcp_sock *tp)
{
	return tp->sacked_out + tp->lost_out;
}

L
Linus Torvalds 已提交
841 842 843 844 845 846 847 848 849 850 851 852 853 854
/* This determines how many packets are "in the network" to the best
 * of our knowledge.  In many cases it is conservative, but where
 * detailed information is available from the receiver (via SACK
 * blocks etc.) we can make more aggressive calculations.
 *
 * Use this for decisions involving congestion control, use just
 * tp->packets_out to determine if the send queue is empty or not.
 *
 * Read this equation as:
 *
 *	"Packets sent once on transmission queue" MINUS
 *	"Packets left network, but not honestly ACKed yet" PLUS
 *	"Packets fast retransmitted"
 */
S
Stephen Hemminger 已提交
855
static inline unsigned int tcp_packets_in_flight(const struct tcp_sock *tp)
L
Linus Torvalds 已提交
856
{
857
	return tp->packets_out - tcp_left_out(tp) + tp->retrans_out;
L
Linus Torvalds 已提交
858 859
}

I
Ilpo Järvinen 已提交
860 861 862 863 864 865 866
#define TCP_INFINITE_SSTHRESH	0x7fffffff

static inline bool tcp_in_initial_slowstart(const struct tcp_sock *tp)
{
	return tp->snd_ssthresh >= TCP_INFINITE_SSTHRESH;
}

L
Linus Torvalds 已提交
867 868 869 870
/* If cwnd > ssthresh, we may raise ssthresh to be half-way to cwnd.
 * The exception is rate halving phase, when cwnd is decreasing towards
 * ssthresh.
 */
871
static inline __u32 tcp_current_ssthresh(const struct sock *sk)
L
Linus Torvalds 已提交
872
{
873
	const struct tcp_sock *tp = tcp_sk(sk);
874

875
	if ((1 << inet_csk(sk)->icsk_ca_state) & (TCPF_CA_CWR | TCPF_CA_Recovery))
L
Linus Torvalds 已提交
876 877 878 879 880 881 882
		return tp->snd_ssthresh;
	else
		return max(tp->snd_ssthresh,
			   ((tp->snd_cwnd >> 1) +
			    (tp->snd_cwnd >> 2)));
}

883 884
/* Use define here intentionally to get WARN_ON location shown at the caller */
#define tcp_verify_left_out(tp)	WARN_ON(tcp_left_out(tp) > tp->packets_out)
L
Linus Torvalds 已提交
885

886
extern void tcp_enter_cwr(struct sock *sk, const int set_ssthresh);
887
extern __u32 tcp_init_cwnd(const struct tcp_sock *tp, const struct dst_entry *dst);
L
Linus Torvalds 已提交
888

889 890 891 892 893 894 895 896
/* The maximum number of MSS of available cwnd for which TSO defers
 * sending if not using sysctl_tcp_tso_win_divisor.
 */
static inline __u32 tcp_max_tso_deferred_mss(const struct tcp_sock *tp)
{
	return 3;
}

L
Linus Torvalds 已提交
897
/* Slow start with delack produces 3 packets of burst, so that
898 899 900 901
 * it is safe "de facto".  This will be the default - same as
 * the default reordering threshold - but if reordering increases,
 * we must be able to allow cwnd to burst at least this much in order
 * to not pull it back when holes are filled.
L
Linus Torvalds 已提交
902 903 904
 */
static __inline__ __u32 tcp_max_burst(const struct tcp_sock *tp)
{
905
	return tp->reordering;
L
Linus Torvalds 已提交
906 907
}

908 909 910 911 912
/* Returns end sequence number of the receiver's advertised window */
static inline u32 tcp_wnd_end(const struct tcp_sock *tp)
{
	return tp->snd_una + tp->snd_wnd;
}
E
Eric Dumazet 已提交
913
extern bool tcp_is_cwnd_limited(const struct sock *sk, u32 in_flight);
914

915
static inline void tcp_minshall_update(struct tcp_sock *tp, unsigned int mss,
S
Stephen Hemminger 已提交
916
				       const struct sk_buff *skb)
L
Linus Torvalds 已提交
917 918 919 920 921
{
	if (skb->len < mss)
		tp->snd_sml = TCP_SKB_CB(skb)->end_seq;
}

922
static inline void tcp_check_probe_timer(struct sock *sk)
L
Linus Torvalds 已提交
923
{
924
	const struct tcp_sock *tp = tcp_sk(sk);
925
	const struct inet_connection_sock *icsk = inet_csk(sk);
926

927
	if (!tp->packets_out && !icsk->icsk_pending)
928 929
		inet_csk_reset_xmit_timer(sk, ICSK_TIME_PROBE0,
					  icsk->icsk_rto, TCP_RTO_MAX);
L
Linus Torvalds 已提交
930 931
}

932
static inline void tcp_init_wl(struct tcp_sock *tp, u32 seq)
L
Linus Torvalds 已提交
933 934 935 936
{
	tp->snd_wl1 = seq;
}

937
static inline void tcp_update_wl(struct tcp_sock *tp, u32 seq)
L
Linus Torvalds 已提交
938 939 940 941 942 943 944
{
	tp->snd_wl1 = seq;
}

/*
 * Calculate(/check) TCP checksum
 */
945 946
static inline __sum16 tcp_v4_check(int len, __be32 saddr,
				   __be32 daddr, __wsum base)
L
Linus Torvalds 已提交
947 948 949 950
{
	return csum_tcpudp_magic(saddr,daddr,len,IPPROTO_TCP,base);
}

951
static inline __sum16 __tcp_checksum_complete(struct sk_buff *skb)
L
Linus Torvalds 已提交
952
{
953
	return __skb_checksum_complete(skb);
L
Linus Torvalds 已提交
954 955
}

E
Eric Dumazet 已提交
956
static inline bool tcp_checksum_complete(struct sk_buff *skb)
L
Linus Torvalds 已提交
957
{
958
	return !skb_csum_unnecessary(skb) &&
L
Linus Torvalds 已提交
959 960 961 962 963
		__tcp_checksum_complete(skb);
}

/* Prequeue for VJ style copy to user, combined with checksumming. */

S
Stephen Hemminger 已提交
964
static inline void tcp_prequeue_init(struct tcp_sock *tp)
L
Linus Torvalds 已提交
965 966 967 968 969
{
	tp->ucopy.task = NULL;
	tp->ucopy.len = 0;
	tp->ucopy.memory = 0;
	skb_queue_head_init(&tp->ucopy.prequeue);
970 971 972 973 974 975
#ifdef CONFIG_NET_DMA
	tp->ucopy.dma_chan = NULL;
	tp->ucopy.wakeup = 0;
	tp->ucopy.pinned_list = NULL;
	tp->ucopy.dma_cookie = 0;
#endif
L
Linus Torvalds 已提交
976 977 978 979 980 981 982 983 984 985
}

/* Packet is added to VJ-style prequeue for processing in process
 * context, if a reader task is waiting. Apparently, this exciting
 * idea (VJ's mail "Re: query about TCP header on tcp-ip" of 07 Sep 93)
 * failed somewhere. Latency? Burstiness? Well, at least now we will
 * see, why it failed. 8)8)				  --ANK
 *
 * NOTE: is this not too big to inline?
 */
E
Eric Dumazet 已提交
986
static inline bool tcp_prequeue(struct sock *sk, struct sk_buff *skb)
L
Linus Torvalds 已提交
987 988 989
{
	struct tcp_sock *tp = tcp_sk(sk);

E
Eric Dumazet 已提交
990
	if (sysctl_tcp_low_latency || !tp->ucopy.task)
E
Eric Dumazet 已提交
991
		return false;
E
Eric Dumazet 已提交
992 993 994 995 996 997 998 999 1000 1001 1002 1003

	__skb_queue_tail(&tp->ucopy.prequeue, skb);
	tp->ucopy.memory += skb->truesize;
	if (tp->ucopy.memory > sk->sk_rcvbuf) {
		struct sk_buff *skb1;

		BUG_ON(sock_owned_by_user(sk));

		while ((skb1 = __skb_dequeue(&tp->ucopy.prequeue)) != NULL) {
			sk_backlog_rcv(sk, skb1);
			NET_INC_STATS_BH(sock_net(sk),
					 LINUX_MIB_TCPPREQUEUEDROPPED);
L
Linus Torvalds 已提交
1004
		}
E
Eric Dumazet 已提交
1005 1006 1007

		tp->ucopy.memory = 0;
	} else if (skb_queue_len(&tp->ucopy.prequeue) == 1) {
E
Eric Dumazet 已提交
1008
		wake_up_interruptible_sync_poll(sk_sleep(sk),
1009
					   POLLIN | POLLRDNORM | POLLRDBAND);
E
Eric Dumazet 已提交
1010 1011
		if (!inet_csk_ack_scheduled(sk))
			inet_csk_reset_xmit_timer(sk, ICSK_TIME_DACK,
1012
						  (3 * tcp_rto_min(sk)) / 4,
E
Eric Dumazet 已提交
1013
						  TCP_RTO_MAX);
L
Linus Torvalds 已提交
1014
	}
E
Eric Dumazet 已提交
1015
	return true;
L
Linus Torvalds 已提交
1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027
}


#undef STATE_TRACE

#ifdef STATE_TRACE
static const char *statename[]={
	"Unused","Established","Syn Sent","Syn Recv",
	"Fin Wait 1","Fin Wait 2","Time Wait", "Close",
	"Close Wait","Last ACK","Listen","Closing"
};
#endif
I
Ilpo Järvinen 已提交
1028
extern void tcp_set_state(struct sock *sk, int state);
L
Linus Torvalds 已提交
1029

A
Andi Kleen 已提交
1030
extern void tcp_done(struct sock *sk);
L
Linus Torvalds 已提交
1031

S
Stephen Hemminger 已提交
1032
static inline void tcp_sack_reset(struct tcp_options_received *rx_opt)
L
Linus Torvalds 已提交
1033 1034 1035 1036 1037 1038 1039 1040
{
	rx_opt->dsack = 0;
	rx_opt->num_sacks = 0;
}

/* Determine a window scaling and initial window to offer. */
extern void tcp_select_initial_window(int __space, __u32 mss,
				      __u32 *rcv_wnd, __u32 *window_clamp,
1041 1042
				      int wscale_ok, __u8 *rcv_wscale,
				      __u32 init_rcv_wnd);
L
Linus Torvalds 已提交
1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062

static inline int tcp_win_from_space(int space)
{
	return sysctl_tcp_adv_win_scale<=0 ?
		(space>>(-sysctl_tcp_adv_win_scale)) :
		space - (space>>sysctl_tcp_adv_win_scale);
}

/* Note: caller must be prepared to deal with negative returns */ 
static inline int tcp_space(const struct sock *sk)
{
	return tcp_win_from_space(sk->sk_rcvbuf -
				  atomic_read(&sk->sk_rmem_alloc));
} 

static inline int tcp_full_space(const struct sock *sk)
{
	return tcp_win_from_space(sk->sk_rcvbuf); 
}

S
Stephen Hemminger 已提交
1063 1064 1065
static inline void tcp_openreq_init(struct request_sock *req,
				    struct tcp_options_received *rx_opt,
				    struct sk_buff *skb)
L
Linus Torvalds 已提交
1066
{
1067 1068
	struct inet_request_sock *ireq = inet_rsk(req);

L
Linus Torvalds 已提交
1069
	req->rcv_wnd = 0;		/* So that tcp_send_synack() knows! */
1070
	req->cookie_ts = 0;
1071
	tcp_rsk(req)->rcv_isn = TCP_SKB_CB(skb)->seq;
L
Linus Torvalds 已提交
1072 1073
	req->mss = rx_opt->mss_clamp;
	req->ts_recent = rx_opt->saw_tstamp ? rx_opt->rcv_tsval : 0;
1074 1075 1076 1077 1078 1079
	ireq->tstamp_ok = rx_opt->tstamp_ok;
	ireq->sack_ok = rx_opt->sack_ok;
	ireq->snd_wscale = rx_opt->snd_wscale;
	ireq->wscale_ok = rx_opt->wscale_ok;
	ireq->acked = 0;
	ireq->ecn_ok = 0;
1080
	ireq->rmt_port = tcp_hdr(skb)->source;
1081
	ireq->loc_port = tcp_hdr(skb)->dest;
L
Linus Torvalds 已提交
1082 1083
}

1084
extern void tcp_enter_memory_pressure(struct sock *sk);
L
Linus Torvalds 已提交
1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095

static inline int keepalive_intvl_when(const struct tcp_sock *tp)
{
	return tp->keepalive_intvl ? : sysctl_tcp_keepalive_intvl;
}

static inline int keepalive_time_when(const struct tcp_sock *tp)
{
	return tp->keepalive_time ? : sysctl_tcp_keepalive_time;
}

E
Eric Dumazet 已提交
1096 1097 1098 1099 1100
static inline int keepalive_probes(const struct tcp_sock *tp)
{
	return tp->keepalive_probes ? : sysctl_tcp_keepalive_probes;
}

1101 1102 1103 1104 1105 1106 1107 1108
static inline u32 keepalive_time_elapsed(const struct tcp_sock *tp)
{
	const struct inet_connection_sock *icsk = &tp->inet_conn;

	return min_t(u32, tcp_time_stamp - icsk->icsk_ack.lrcvtime,
			  tcp_time_stamp - tp->rcv_tstamp);
}

1109
static inline int tcp_fin_time(const struct sock *sk)
L
Linus Torvalds 已提交
1110
{
1111 1112
	int fin_timeout = tcp_sk(sk)->linger2 ? : sysctl_tcp_fin_timeout;
	const int rto = inet_csk(sk)->icsk_rto;
L
Linus Torvalds 已提交
1113

1114 1115
	if (fin_timeout < (rto << 2) - (rto >> 1))
		fin_timeout = (rto << 2) - (rto >> 1);
L
Linus Torvalds 已提交
1116 1117 1118 1119

	return fin_timeout;
}

E
Eric Dumazet 已提交
1120 1121
static inline bool tcp_paws_check(const struct tcp_options_received *rx_opt,
				  int paws_win)
L
Linus Torvalds 已提交
1122
{
I
Ilpo Järvinen 已提交
1123
	if ((s32)(rx_opt->ts_recent - rx_opt->rcv_tsval) <= paws_win)
E
Eric Dumazet 已提交
1124
		return true;
I
Ilpo Järvinen 已提交
1125
	if (unlikely(get_seconds() >= rx_opt->ts_recent_stamp + TCP_PAWS_24DAYS))
E
Eric Dumazet 已提交
1126
		return true;
E
Eric Dumazet 已提交
1127 1128 1129 1130 1131 1132
	/*
	 * Some OSes send SYN and SYNACK messages with tsval=0 tsecr=0,
	 * then following tcp messages have valid values. Ignore 0 value,
	 * or else 'negative' tsval might forbid us to accept their packets.
	 */
	if (!rx_opt->ts_recent)
E
Eric Dumazet 已提交
1133 1134
		return true;
	return false;
I
Ilpo Järvinen 已提交
1135 1136
}

E
Eric Dumazet 已提交
1137 1138
static inline bool tcp_paws_reject(const struct tcp_options_received *rx_opt,
				   int rst)
I
Ilpo Järvinen 已提交
1139 1140
{
	if (tcp_paws_check(rx_opt, 0))
E
Eric Dumazet 已提交
1141
		return false;
L
Linus Torvalds 已提交
1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154

	/* RST segments are not recommended to carry timestamp,
	   and, if they do, it is recommended to ignore PAWS because
	   "their cleanup function should take precedence over timestamps."
	   Certainly, it is mistake. It is necessary to understand the reasons
	   of this constraint to relax it: if peer reboots, clock may go
	   out-of-sync and half-open connections will not be reset.
	   Actually, the problem would be not existing if all
	   the implementations followed draft about maintaining clock
	   via reboots. Linux-2.2 DOES NOT!

	   However, we can relax time bounds for RST segments to MSL.
	 */
1155
	if (rst && get_seconds() >= rx_opt->ts_recent_stamp + TCP_PAWS_MSL)
E
Eric Dumazet 已提交
1156 1157
		return false;
	return true;
L
Linus Torvalds 已提交
1158 1159
}

P
Pavel Emelyanov 已提交
1160
static inline void tcp_mib_init(struct net *net)
L
Linus Torvalds 已提交
1161 1162
{
	/* See RFC 2012 */
1163 1164 1165 1166
	TCP_ADD_STATS_USER(net, TCP_MIB_RTOALGORITHM, 1);
	TCP_ADD_STATS_USER(net, TCP_MIB_RTOMIN, TCP_RTO_MIN*1000/HZ);
	TCP_ADD_STATS_USER(net, TCP_MIB_RTOMAX, TCP_RTO_MAX*1000/HZ);
	TCP_ADD_STATS_USER(net, TCP_MIB_MAXCONN, -1);
L
Linus Torvalds 已提交
1167 1168
}

1169
/* from STCP */
1170
static inline void tcp_clear_retrans_hints_partial(struct tcp_sock *tp)
D
David S. Miller 已提交
1171
{
1172 1173
	tp->lost_skb_hint = NULL;
	tp->scoreboard_skb_hint = NULL;
1174 1175 1176 1177 1178
}

static inline void tcp_clear_all_retrans_hints(struct tcp_sock *tp)
{
	tcp_clear_retrans_hints_partial(tp);
1179
	tp->retransmit_skb_hint = NULL;
1180 1181
}

1182 1183 1184
/* MD5 Signature */
struct crypto_hash;

E
Eric Dumazet 已提交
1185 1186 1187 1188 1189 1190 1191
union tcp_md5_addr {
	struct in_addr  a4;
#if IS_ENABLED(CONFIG_IPV6)
	struct in6_addr	a6;
#endif
};

1192 1193
/* - key database */
struct tcp_md5sig_key {
E
Eric Dumazet 已提交
1194
	struct hlist_node	node;
1195
	u8			keylen;
E
Eric Dumazet 已提交
1196 1197 1198 1199
	u8			family; /* AF_INET or AF_INET6 */
	union tcp_md5_addr	addr;
	u8			key[TCP_MD5SIG_MAXKEYLEN];
	struct rcu_head		rcu;
1200 1201 1202 1203
};

/* - sock block */
struct tcp_md5sig_info {
E
Eric Dumazet 已提交
1204
	struct hlist_head	head;
1205
	struct rcu_head		rcu;
1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225
};

/* - pseudo header */
struct tcp4_pseudohdr {
	__be32		saddr;
	__be32		daddr;
	__u8		pad;
	__u8		protocol;
	__be16		len;
};

struct tcp6_pseudohdr {
	struct in6_addr	saddr;
	struct in6_addr daddr;
	__be32		len;
	__be32		protocol;	/* including padding */
};

union tcp_md5sum_block {
	struct tcp4_pseudohdr ip4;
E
Eric Dumazet 已提交
1226
#if IS_ENABLED(CONFIG_IPV6)
1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237
	struct tcp6_pseudohdr ip6;
#endif
};

/* - pool: digest algorithm, hash description and scratch buffer */
struct tcp_md5sig_pool {
	struct hash_desc	md5_desc;
	union tcp_md5sum_block	md5_blk;
};

/* - functions */
C
Changli Gao 已提交
1238
extern int tcp_v4_md5_hash_skb(char *md5_hash, struct tcp_md5sig_key *key,
E
Eric Dumazet 已提交
1239 1240 1241
			       const struct sock *sk,
			       const struct request_sock *req,
			       const struct sk_buff *skb);
E
Eric Dumazet 已提交
1242 1243 1244 1245 1246 1247 1248
extern int tcp_md5_do_add(struct sock *sk, const union tcp_md5_addr *addr,
			  int family, const u8 *newkey,
			  u8 newkeylen, gfp_t gfp);
extern int tcp_md5_do_del(struct sock *sk, const union tcp_md5_addr *addr,
			  int family);
extern struct tcp_md5sig_key *tcp_v4_md5_lookup(struct sock *sk,
					 struct sock *addr_sk);
1249

1250
#ifdef CONFIG_TCP_MD5SIG
E
Eric Dumazet 已提交
1251 1252 1253
extern struct tcp_md5sig_key *tcp_md5_do_lookup(struct sock *sk,
			const union tcp_md5_addr *addr, int family);
#define tcp_twsk_md5_key(twsk)	((twsk)->tw_md5_key)
1254
#else
E
Eric Dumazet 已提交
1255 1256 1257 1258 1259 1260
static inline struct tcp_md5sig_key *tcp_md5_do_lookup(struct sock *sk,
					 const union tcp_md5_addr *addr,
					 int family)
{
	return NULL;
}
1261 1262 1263
#define tcp_twsk_md5_key(twsk)	NULL
#endif

1264
extern struct tcp_md5sig_pool __percpu *tcp_alloc_md5sig_pool(struct sock *);
C
Changli Gao 已提交
1265
extern void tcp_free_md5sig_pool(void);
1266

E
Eric Dumazet 已提交
1267
extern struct tcp_md5sig_pool	*tcp_get_md5sig_pool(void);
C
Changli Gao 已提交
1268
extern void tcp_put_md5sig_pool(void);
E
Eric Dumazet 已提交
1269

1270
extern int tcp_md5_hash_header(struct tcp_md5sig_pool *, const struct tcphdr *);
1271
extern int tcp_md5_hash_skb_data(struct tcp_md5sig_pool *, const struct sk_buff *,
1272
				 unsigned int header_len);
1273
extern int tcp_md5_hash_key(struct tcp_md5sig_pool *hp,
1274
			    const struct tcp_md5sig_key *key);
1275

1276 1277 1278 1279 1280 1281
/* write queue abstraction */
static inline void tcp_write_queue_purge(struct sock *sk)
{
	struct sk_buff *skb;

	while ((skb = __skb_dequeue(&sk->sk_write_queue)) != NULL)
1282 1283
		sk_wmem_free_skb(sk, skb);
	sk_mem_reclaim(sk);
1284
	tcp_clear_all_retrans_hints(tcp_sk(sk));
1285 1286
}

1287
static inline struct sk_buff *tcp_write_queue_head(const struct sock *sk)
1288
{
1289
	return skb_peek(&sk->sk_write_queue);
1290 1291
}

1292
static inline struct sk_buff *tcp_write_queue_tail(const struct sock *sk)
1293
{
1294
	return skb_peek_tail(&sk->sk_write_queue);
1295 1296
}

1297 1298
static inline struct sk_buff *tcp_write_queue_next(const struct sock *sk,
						   const struct sk_buff *skb)
1299
{
1300
	return skb_queue_next(&sk->sk_write_queue, skb);
1301 1302
}

1303 1304
static inline struct sk_buff *tcp_write_queue_prev(const struct sock *sk,
						   const struct sk_buff *skb)
1305 1306 1307 1308
{
	return skb_queue_prev(&sk->sk_write_queue, skb);
}

1309
#define tcp_for_write_queue(skb, sk)					\
1310
	skb_queue_walk(&(sk)->sk_write_queue, skb)
1311 1312

#define tcp_for_write_queue_from(skb, sk)				\
1313
	skb_queue_walk_from(&(sk)->sk_write_queue, skb)
1314

1315
#define tcp_for_write_queue_from_safe(skb, tmp, sk)			\
1316
	skb_queue_walk_from_safe(&(sk)->sk_write_queue, skb, tmp)
1317

1318
static inline struct sk_buff *tcp_send_head(const struct sock *sk)
1319 1320 1321 1322
{
	return sk->sk_send_head;
}

1323 1324 1325 1326 1327 1328
static inline bool tcp_skb_is_last(const struct sock *sk,
				   const struct sk_buff *skb)
{
	return skb_queue_is_last(&sk->sk_write_queue, skb);
}

1329
static inline void tcp_advance_send_head(struct sock *sk, const struct sk_buff *skb)
1330
{
1331
	if (tcp_skb_is_last(sk, skb))
1332
		sk->sk_send_head = NULL;
1333 1334
	else
		sk->sk_send_head = tcp_write_queue_next(sk, skb);
1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357
}

static inline void tcp_check_send_head(struct sock *sk, struct sk_buff *skb_unlinked)
{
	if (sk->sk_send_head == skb_unlinked)
		sk->sk_send_head = NULL;
}

static inline void tcp_init_send_head(struct sock *sk)
{
	sk->sk_send_head = NULL;
}

static inline void __tcp_add_write_queue_tail(struct sock *sk, struct sk_buff *skb)
{
	__skb_queue_tail(&sk->sk_write_queue, skb);
}

static inline void tcp_add_write_queue_tail(struct sock *sk, struct sk_buff *skb)
{
	__tcp_add_write_queue_tail(sk, skb);

	/* Queue it, remembering where we must start sending. */
1358
	if (sk->sk_send_head == NULL) {
1359
		sk->sk_send_head = skb;
1360 1361 1362 1363

		if (tcp_sk(sk)->highest_sack == NULL)
			tcp_sk(sk)->highest_sack = skb;
	}
1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375
}

static inline void __tcp_add_write_queue_head(struct sock *sk, struct sk_buff *skb)
{
	__skb_queue_head(&sk->sk_write_queue, skb);
}

/* Insert buff after skb on the write queue of sk.  */
static inline void tcp_insert_write_queue_after(struct sk_buff *skb,
						struct sk_buff *buff,
						struct sock *sk)
{
1376
	__skb_queue_after(&sk->sk_write_queue, skb, buff);
1377 1378
}

1379
/* Insert new before skb on the write queue of sk.  */
1380 1381 1382 1383
static inline void tcp_insert_write_queue_before(struct sk_buff *new,
						  struct sk_buff *skb,
						  struct sock *sk)
{
1384
	__skb_queue_before(&sk->sk_write_queue, skb, new);
1385 1386 1387

	if (sk->sk_send_head == skb)
		sk->sk_send_head = new;
1388 1389 1390 1391 1392 1393 1394
}

static inline void tcp_unlink_write_queue(struct sk_buff *skb, struct sock *sk)
{
	__skb_unlink(skb, &sk->sk_write_queue);
}

E
Eric Dumazet 已提交
1395
static inline bool tcp_write_queue_empty(struct sock *sk)
1396 1397 1398 1399
{
	return skb_queue_empty(&sk->sk_write_queue);
}

1400 1401 1402 1403 1404 1405 1406 1407 1408
static inline void tcp_push_pending_frames(struct sock *sk)
{
	if (tcp_send_head(sk)) {
		struct tcp_sock *tp = tcp_sk(sk);

		__tcp_push_pending_frames(sk, tcp_current_mss(sk), tp->nonagle);
	}
}

1409 1410 1411
/* Start sequence of the skb just after the highest skb with SACKed
 * bit, valid only if sacked_out > 0 or when the caller has ensured
 * validity by itself.
1412 1413 1414 1415 1416
 */
static inline u32 tcp_highest_sack_seq(struct tcp_sock *tp)
{
	if (!tp->sacked_out)
		return tp->snd_una;
1417 1418 1419 1420

	if (tp->highest_sack == NULL)
		return tp->snd_nxt;

1421 1422 1423
	return TCP_SKB_CB(tp->highest_sack)->seq;
}

1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448
static inline void tcp_advance_highest_sack(struct sock *sk, struct sk_buff *skb)
{
	tcp_sk(sk)->highest_sack = tcp_skb_is_last(sk, skb) ? NULL :
						tcp_write_queue_next(sk, skb);
}

static inline struct sk_buff *tcp_highest_sack(struct sock *sk)
{
	return tcp_sk(sk)->highest_sack;
}

static inline void tcp_highest_sack_reset(struct sock *sk)
{
	tcp_sk(sk)->highest_sack = tcp_write_queue_head(sk);
}

/* Called when old skb is about to be deleted (to be combined with new skb) */
static inline void tcp_highest_sack_combine(struct sock *sk,
					    struct sk_buff *old,
					    struct sk_buff *new)
{
	if (tcp_sk(sk)->sacked_out && (old == tcp_sk(sk)->highest_sack))
		tcp_sk(sk)->highest_sack = new;
}

A
Andreas Petlund 已提交
1449 1450 1451
/* Determines whether this is a thin stream (which may suffer from
 * increased latency). Used to trigger latency-reducing mechanisms.
 */
E
Eric Dumazet 已提交
1452
static inline bool tcp_stream_is_thin(struct tcp_sock *tp)
A
Andreas Petlund 已提交
1453 1454 1455 1456
{
	return tp->packets_out < 4 && !tcp_in_initial_slowstart(tp);
}

L
Linus Torvalds 已提交
1457 1458 1459 1460 1461 1462 1463 1464
/* /proc */
enum tcp_seq_states {
	TCP_SEQ_STATE_LISTENING,
	TCP_SEQ_STATE_OPENREQ,
	TCP_SEQ_STATE_ESTABLISHED,
	TCP_SEQ_STATE_TIME_WAIT,
};

1465 1466
int tcp_seq_open(struct inode *inode, struct file *file);

L
Linus Torvalds 已提交
1467
struct tcp_seq_afinfo {
1468 1469 1470 1471
	char				*name;
	sa_family_t			family;
	const struct file_operations	*seq_fops;
	struct seq_operations		seq_ops;
L
Linus Torvalds 已提交
1472 1473 1474
};

struct tcp_iter_state {
1475
	struct seq_net_private	p;
L
Linus Torvalds 已提交
1476 1477 1478
	sa_family_t		family;
	enum tcp_seq_states	state;
	struct sock		*syn_wait_sk;
1479 1480
	int			bucket, offset, sbucket, num, uid;
	loff_t			last_pos;
L
Linus Torvalds 已提交
1481 1482
};

1483 1484
extern int tcp_proc_register(struct net *net, struct tcp_seq_afinfo *afinfo);
extern void tcp_proc_unregister(struct net *net, struct tcp_seq_afinfo *afinfo);
L
Linus Torvalds 已提交
1485

1486
extern struct request_sock_ops tcp_request_sock_ops;
1487
extern struct request_sock_ops tcp6_request_sock_ops;
1488

1489
extern void tcp_v4_destroy_sock(struct sock *sk);
1490

1491
extern int tcp_v4_gso_send_check(struct sk_buff *skb);
1492 1493
extern struct sk_buff *tcp_tso_segment(struct sk_buff *skb,
				       netdev_features_t features);
H
Herbert Xu 已提交
1494 1495 1496 1497 1498 1499
extern struct sk_buff **tcp_gro_receive(struct sk_buff **head,
					struct sk_buff *skb);
extern struct sk_buff **tcp4_gro_receive(struct sk_buff **head,
					 struct sk_buff *skb);
extern int tcp_gro_complete(struct sk_buff *skb);
extern int tcp4_gro_complete(struct sk_buff *skb);
H
Herbert Xu 已提交
1500

1501
#ifdef CONFIG_PROC_FS
C
Changli Gao 已提交
1502
extern int tcp4_proc_init(void);
1503 1504 1505
extern void tcp4_proc_exit(void);
#endif

1506 1507 1508 1509 1510 1511 1512
/* TCP af-specific functions */
struct tcp_sock_af_ops {
#ifdef CONFIG_TCP_MD5SIG
	struct tcp_md5sig_key	*(*md5_lookup) (struct sock *sk,
						struct sock *addr_sk);
	int			(*calc_md5_hash) (char *location,
						  struct tcp_md5sig_key *md5,
E
Eric Dumazet 已提交
1513 1514 1515
						  const struct sock *sk,
						  const struct request_sock *req,
						  const struct sk_buff *skb);
1516 1517 1518 1519 1520 1521 1522 1523 1524 1525
	int			(*md5_parse) (struct sock *sk,
					      char __user *optval,
					      int optlen);
#endif
};

struct tcp_request_sock_ops {
#ifdef CONFIG_TCP_MD5SIG
	struct tcp_md5sig_key	*(*md5_lookup) (struct sock *sk,
						struct request_sock *req);
1526 1527
	int			(*calc_md5_hash) (char *location,
						  struct tcp_md5sig_key *md5,
E
Eric Dumazet 已提交
1528 1529 1530
						  const struct sock *sk,
						  const struct request_sock *req,
						  const struct sk_buff *skb);
1531 1532 1533
#endif
};

1534 1535 1536 1537 1538 1539 1540 1541
/* Using SHA1 for now, define some constants.
 */
#define COOKIE_DIGEST_WORDS (SHA_DIGEST_WORDS)
#define COOKIE_MESSAGE_WORDS (SHA_MESSAGE_BYTES / 4)
#define COOKIE_WORKSPACE_WORDS (COOKIE_DIGEST_WORDS + COOKIE_MESSAGE_WORDS)

extern int tcp_cookie_generator(u32 *bakery);

1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618
/**
 *	struct tcp_cookie_values - each socket needs extra space for the
 *	cookies, together with (optional) space for any SYN data.
 *
 *	A tcp_sock contains a pointer to the current value, and this is
 *	cloned to the tcp_timewait_sock.
 *
 * @cookie_pair:	variable data from the option exchange.
 *
 * @cookie_desired:	user specified tcpct_cookie_desired.  Zero
 *			indicates default (sysctl_tcp_cookie_size).
 *			After cookie sent, remembers size of cookie.
 *			Range 0, TCP_COOKIE_MIN to TCP_COOKIE_MAX.
 *
 * @s_data_desired:	user specified tcpct_s_data_desired.  When the
 *			constant payload is specified (@s_data_constant),
 *			holds its length instead.
 *			Range 0 to TCP_MSS_DESIRED.
 *
 * @s_data_payload:	constant data that is to be included in the
 *			payload of SYN or SYNACK segments when the
 *			cookie option is present.
 */
struct tcp_cookie_values {
	struct kref	kref;
	u8		cookie_pair[TCP_COOKIE_PAIR_SIZE];
	u8		cookie_pair_size;
	u8		cookie_desired;
	u16		s_data_desired:11,
			s_data_constant:1,
			s_data_in:1,
			s_data_out:1,
			s_data_unused:2;
	u8		s_data_payload[0];
};

static inline void tcp_cookie_values_release(struct kref *kref)
{
	kfree(container_of(kref, struct tcp_cookie_values, kref));
}

/* The length of constant payload data.  Note that s_data_desired is
 * overloaded, depending on s_data_constant: either the length of constant
 * data (returned here) or the limit on variable data.
 */
static inline int tcp_s_data_size(const struct tcp_sock *tp)
{
	return (tp->cookie_values != NULL && tp->cookie_values->s_data_constant)
		? tp->cookie_values->s_data_desired
		: 0;
}

/**
 *	struct tcp_extend_values - tcp_ipv?.c to tcp_output.c workspace.
 *
 *	As tcp_request_sock has already been extended in other places, the
 *	only remaining method is to pass stack values along as function
 *	parameters.  These parameters are not needed after sending SYNACK.
 *
 * @cookie_bakery:	cryptographic secret and message workspace.
 *
 * @cookie_plus:	bytes in authenticator/cookie option, copied from
 *			struct tcp_options_received (above).
 */
struct tcp_extend_values {
	struct request_values		rv;
	u32				cookie_bakery[COOKIE_WORKSPACE_WORDS];
	u8				cookie_plus:6,
					cookie_out_never:1,
					cookie_in_always:1;
};

static inline struct tcp_extend_values *tcp_xv(struct request_values *rvp)
{
	return (struct tcp_extend_values *)rvp;
}

1619
extern void tcp_v4_init(void);
1620 1621
extern void tcp_init(void);

L
Linus Torvalds 已提交
1622
#endif	/* _TCP_H */