tcp.h 47.9 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;
53
void tcp_time_wait(struct sock *sk, int state, int timeo);
L
Linus Torvalds 已提交
54 55

#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 64 65 66

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

/* Minimal accepted MSS. It is (60+60+8) - (20+20). */
#define TCP_MIN_MSS		88U

J
John Heffner 已提交
67 68 69
/* The least MTU to use for probing */
#define TCP_BASE_MSS		512

L
Linus Torvalds 已提交
70 71 72 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
/* 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.
				 */

98 99 100 101 102 103 104 105
#define TCP_SYN_RETRIES	 6	/* This is how many retries are done
				 * when active opening a connection.
				 * RFC1122 says the minimum retry MUST
				 * be at least 180secs.  Nevertheless
				 * this value is corresponding to
				 * 63secs of retransmission with the
				 * current initial RTO.
				 */
L
Linus Torvalds 已提交
106

107 108 109 110 111 112
#define TCP_SYNACK_RETRIES 5	/* This is how may retries are done
				 * when passive opening a connection.
				 * This is corresponding to 31secs of
				 * retransmission with the current
				 * initial RTO.
				 */
L
Linus Torvalds 已提交
113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132

#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))
133
#define TCP_TIMEOUT_INIT ((unsigned)(1*HZ))	/* RFC6298 2.1 initial RTO value	*/
134 135 136 137 138 139
#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
140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177

#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 */
178
#define TCPOPT_MD5SIG		19	/* MD5 Signature (RFC2385) */
Y
Yuchung Cheng 已提交
179 180 181 182 183
#define TCPOPT_EXP		254	/* Experimental */
/* Magic number to be after the option value for sharing TCP
 * experimental options. See draft-ietf-tcpm-experimental-options-00.txt
 */
#define TCPOPT_FASTOPEN_MAGIC	0xF989
L
Linus Torvalds 已提交
184 185 186 187 188 189 190 191 192

/*
 *     TCP option lengths
 */

#define TCPOLEN_MSS            4
#define TCPOLEN_WINDOW         3
#define TCPOLEN_SACK_PERM      2
#define TCPOLEN_TIMESTAMP      10
193
#define TCPOLEN_MD5SIG         18
Y
Yuchung Cheng 已提交
194
#define TCPOLEN_EXP_FASTOPEN_BASE  4
L
Linus Torvalds 已提交
195 196 197 198 199 200 201 202

/* 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
203
#define TCPOLEN_MD5SIG_ALIGNED		20
A
Adam Langley 已提交
204
#define TCPOLEN_MSS_ALIGNED		4
L
Linus Torvalds 已提交
205 206 207 208

/* 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
209
#define TCP_NAGLE_PUSH		4	/* Cork is overridden for already queued data */
L
Linus Torvalds 已提交
210

A
Andreas Petlund 已提交
211 212 213
/* TCP thin-stream limits */
#define TCP_THIN_LINEAR_RETRIES 6       /* After 6 linear retries, do exp. backoff */

214
/* TCP initial congestion window as per draft-hkchu-tcpm-initcwnd-01 */
215 216
#define TCP_INIT_CWND		10

217 218
/* Bit Flags for sysctl_tcp_fastopen */
#define	TFO_CLIENT_ENABLE	1
219
#define	TFO_SERVER_ENABLE	2
220
#define	TFO_CLIENT_NO_COOKIE	4	/* Data in SYN w/o cookie option */
221

222 223 224 225 226 227 228 229 230 231 232 233 234 235 236
/* Process SYN data but skip cookie validation */
#define	TFO_SERVER_COOKIE_NOT_CHKED	0x100
/* Accept SYN data w/o any cookie option */
#define	TFO_SERVER_COOKIE_NOT_REQD	0x200

/* Force enable TFO on all listeners, i.e., not requiring the
 * TCP_FASTOPEN socket option. SOCKOPT1/2 determine how to set max_qlen.
 */
#define	TFO_SERVER_WO_SOCKOPT1	0x400
#define	TFO_SERVER_WO_SOCKOPT2	0x800
/* Always create TFO child sockets on a TFO listener even when
 * cookie/data not present. (For testing purpose!)
 */
#define	TFO_SERVER_ALWAYS	0x1000

237 238
extern struct inet_timewait_death_row tcp_death_row;

L
Linus Torvalds 已提交
239 240 241 242 243 244 245 246 247 248 249 250 251 252
/* 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;
Y
Yuchung Cheng 已提交
253
extern int sysctl_tcp_fastopen;
L
Linus Torvalds 已提交
254 255 256 257 258 259 260 261 262 263 264 265 266 267 268
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_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;
extern int sysctl_tcp_low_latency;
269
extern int sysctl_tcp_dma_copybreak;
L
Linus Torvalds 已提交
270 271 272
extern int sysctl_tcp_nometrics_save;
extern int sysctl_tcp_moderate_rcvbuf;
extern int sysctl_tcp_tso_win_divisor;
J
John Heffner 已提交
273 274
extern int sysctl_tcp_mtu_probing;
extern int sysctl_tcp_base_mss;
275
extern int sysctl_tcp_workaround_signed_windows;
276
extern int sysctl_tcp_slow_start_after_idle;
277
extern int sysctl_tcp_max_ssthresh;
A
Andreas Petlund 已提交
278
extern int sysctl_tcp_thin_linear_timeouts;
A
Andreas Petlund 已提交
279
extern int sysctl_tcp_thin_dupack;
Y
Yuchung Cheng 已提交
280
extern int sysctl_tcp_early_retrans;
E
Eric Dumazet 已提交
281
extern int sysctl_tcp_limit_output_bytes;
E
Eric Dumazet 已提交
282
extern int sysctl_tcp_challenge_ack_limit;
283
extern unsigned int sysctl_tcp_notsent_lowat;
E
Eric Dumazet 已提交
284
extern int sysctl_tcp_min_tso_segs;
L
Linus Torvalds 已提交
285

E
Eric Dumazet 已提交
286
extern atomic_long_t tcp_memory_allocated;
287
extern struct percpu_counter tcp_sockets_allocated;
L
Linus Torvalds 已提交
288 289 290 291 292 293 294
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
295
static inline bool before(__u32 seq1, __u32 seq2)
L
Linus Torvalds 已提交
296
{
297
        return (__s32)(seq1-seq2) < 0;
L
Linus Torvalds 已提交
298
}
299
#define after(seq2, seq1) 	before(seq1, seq2)
L
Linus Torvalds 已提交
300 301

/* is s2<=s1<=s3 ? */
E
Eric Dumazet 已提交
302
static inline bool between(__u32 seq1, __u32 seq2, __u32 seq3)
L
Linus Torvalds 已提交
303 304 305 306
{
	return seq3 - seq2 >= seq1 - seq2;
}

A
Arun Sharma 已提交
307 308 309 310 311 312 313 314
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;
}

315
static inline bool tcp_too_many_orphans(struct sock *sk, int shift)
316
{
317 318 319 320 321 322 323 324 325
	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;
326
}
L
Linus Torvalds 已提交
327

328
bool tcp_check_oom(struct sock *sk, int shift);
A
Arun Sharma 已提交
329

330 331 332 333 334 335 336
/* 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
337
static inline bool tcp_synq_no_recent_overflow(const struct sock *sk)
338 339
{
	unsigned long last_overflow = tcp_sk(sk)->rx_opt.ts_recent_stamp;
340
	return time_after(jiffies, last_overflow + TCP_TIMEOUT_FALLBACK);
341 342
}

L
Linus Torvalds 已提交
343 344
extern struct proto tcp_prot;

345 346 347 348
#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)
349
#define TCP_ADD_STATS(net, field, val)	SNMP_ADD_STATS((net)->mib.tcp_statistics, field, val)
L
Linus Torvalds 已提交
350

351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382
void tcp_init_mem(struct net *net);

void tcp_tasklet_init(void);

void tcp_v4_err(struct sk_buff *skb, u32);

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

void tcp_v4_early_demux(struct sk_buff *skb);
int tcp_v4_rcv(struct sk_buff *skb);

int tcp_v4_tw_remember_stamp(struct inet_timewait_sock *tw);
int tcp_sendmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg,
		size_t size);
int tcp_sendpage(struct sock *sk, struct page *page, int offset, size_t size,
		 int flags);
void tcp_release_cb(struct sock *sk);
void tcp_wfree(struct sk_buff *skb);
void tcp_write_timer_handler(struct sock *sk);
void tcp_delack_timer_handler(struct sock *sk);
int tcp_ioctl(struct sock *sk, int cmd, unsigned long arg);
int tcp_rcv_state_process(struct sock *sk, struct sk_buff *skb,
			  const struct tcphdr *th, unsigned int len);
void tcp_rcv_established(struct sock *sk, struct sk_buff *skb,
			 const struct tcphdr *th, unsigned int len);
void tcp_rcv_space_adjust(struct sock *sk);
void tcp_cleanup_rbuf(struct sock *sk, int copied);
int tcp_twsk_unique(struct sock *sk, struct sock *sktw, void *twp);
void tcp_twsk_destructor(struct sock *sk);
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
383

384 385
static inline void tcp_dec_quickack_mode(struct sock *sk,
					 const unsigned int pkts)
L
Linus Torvalds 已提交
386
{
387
	struct inet_connection_sock *icsk = inet_csk(sk);
388

389 390 391
	if (icsk->icsk_ack.quick) {
		if (pkts >= icsk->icsk_ack.quick) {
			icsk->icsk_ack.quick = 0;
392
			/* Leaving quickack mode we deflate ATO. */
393
			icsk->icsk_ack.ato   = TCP_ATO_MIN;
394
		} else
395
			icsk->icsk_ack.quick -= pkts;
L
Linus Torvalds 已提交
396 397 398
	}
}

399 400 401
#define	TCP_ECN_OK		1
#define	TCP_ECN_QUEUE_CWR	2
#define	TCP_ECN_DEMAND_CWR	4
402
#define	TCP_ECN_SEEN		8
403

E
Eric Dumazet 已提交
404
enum tcp_tw_status {
L
Linus Torvalds 已提交
405 406 407 408 409 410 411
	TCP_TW_SUCCESS = 0,
	TCP_TW_RST = 1,
	TCP_TW_ACK = 2,
	TCP_TW_SYN = 3
};


412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439
enum tcp_tw_status tcp_timewait_state_process(struct inet_timewait_sock *tw,
					      struct sk_buff *skb,
					      const struct tcphdr *th);
struct sock *tcp_check_req(struct sock *sk, struct sk_buff *skb,
			   struct request_sock *req, struct request_sock **prev,
			   bool fastopen);
int tcp_child_process(struct sock *parent, struct sock *child,
		      struct sk_buff *skb);
void tcp_enter_loss(struct sock *sk, int how);
void tcp_clear_retrans(struct tcp_sock *tp);
void tcp_update_metrics(struct sock *sk);
void tcp_init_metrics(struct sock *sk);
void tcp_metrics_init(void);
bool tcp_peer_is_proven(struct request_sock *req, struct dst_entry *dst,
			bool paws_check);
bool tcp_remember_stamp(struct sock *sk);
bool tcp_tw_remember_stamp(struct inet_timewait_sock *tw);
void tcp_fetch_timewait_stamp(struct sock *sk, struct dst_entry *dst);
void tcp_disable_fack(struct tcp_sock *tp);
void tcp_close(struct sock *sk, long timeout);
void tcp_init_sock(struct sock *sk);
unsigned int tcp_poll(struct file *file, struct socket *sock,
		      struct poll_table_struct *wait);
int tcp_getsockopt(struct sock *sk, int level, int optname,
		   char __user *optval, int __user *optlen);
int tcp_setsockopt(struct sock *sk, int level, int optname,
		   char __user *optval, unsigned int optlen);
int compat_tcp_getsockopt(struct sock *sk, int level, int optname,
C
Changli Gao 已提交
440
			  char __user *optval, int __user *optlen);
441
int compat_tcp_setsockopt(struct sock *sk, int level, int optname,
C
Changli Gao 已提交
442
			  char __user *optval, unsigned int optlen);
443 444 445 446 447 448 449 450
void tcp_set_keepalive(struct sock *sk, int val);
void tcp_syn_ack_timeout(struct sock *sk, struct request_sock *req);
int tcp_recvmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg,
		size_t len, int nonblock, int flags, int *addr_len);
void tcp_parse_options(const struct sk_buff *skb,
		       struct tcp_options_received *opt_rx,
		       int estab, struct tcp_fastopen_cookie *foc);
const u8 *tcp_parse_md5sig_option(const struct tcphdr *th);
451

L
Linus Torvalds 已提交
452 453 454 455
/*
 *	TCP v4 functions exported for the inet6 API
 */

456 457 458 459 460 461 462 463 464 465 466 467 468 469 470
void tcp_v4_send_check(struct sock *sk, struct sk_buff *skb);
int tcp_v4_conn_request(struct sock *sk, struct sk_buff *skb);
struct sock *tcp_create_openreq_child(struct sock *sk,
				      struct request_sock *req,
				      struct sk_buff *skb);
struct sock *tcp_v4_syn_recv_sock(struct sock *sk, struct sk_buff *skb,
				  struct request_sock *req,
				  struct dst_entry *dst);
int tcp_v4_do_rcv(struct sock *sk, struct sk_buff *skb);
int tcp_v4_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len);
int tcp_connect(struct sock *sk);
struct sk_buff *tcp_make_synack(struct sock *sk, struct dst_entry *dst,
				struct request_sock *req,
				struct tcp_fastopen_cookie *foc);
int tcp_disconnect(struct sock *sk, int flags);
L
Linus Torvalds 已提交
471

P
Pavel Emelyanov 已提交
472 473
void tcp_connect_init(struct sock *sk);
void tcp_finish_connect(struct sock *sk, struct sk_buff *skb);
474
int tcp_send_rcvq(struct sock *sk, struct msghdr *msg, size_t size);
475
void inet_sk_rx_dst_set(struct sock *sk, const struct sk_buff *skb);
L
Linus Torvalds 已提交
476 477

/* From syncookies.c */
478
extern __u32 syncookie_secret[2][16-4+SHA_DIGEST_WORDS];
479 480 481 482
int __cookie_v4_check(const struct iphdr *iph, const struct tcphdr *th,
		      u32 cookie);
struct sock *cookie_v4_check(struct sock *sk, struct sk_buff *skb,
			     struct ip_options *opt);
483
#ifdef CONFIG_SYN_COOKIES
484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501
#include <linux/ktime.h>

/* Syncookies use a monotonic timer which increments every 64 seconds.
 * This counter is used both as a hash input and partially encoded into
 * the cookie value.  A cookie is only validated further if the delta
 * between the current counter value and the encoded one is less than this,
 * i.e. a sent cookie is valid only at most for 128 seconds (or less if
 * the counter advances immediately after a cookie is generated).
 */
#define MAX_SYNCOOKIE_AGE 2

static inline u32 tcp_cookie_time(void)
{
	struct timespec now;
	getnstimeofday(&now);
	return now.tv_sec >> 6; /* 64 seconds granularity */
}

502 503 504
u32 __cookie_v4_init_sequence(const struct iphdr *iph, const struct tcphdr *th,
			      u16 *mssp);
__u32 cookie_v4_init_sequence(struct sock *sk, struct sk_buff *skb, __u16 *mss);
505 506 507 508 509 510 511 512
#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
513

514 515 516
__u32 cookie_init_timestamp(struct request_sock *req);
bool cookie_check_timestamp(struct tcp_options_received *opt, struct net *net,
			    bool *ecn_ok);
517

518
/* From net/ipv6/syncookies.c */
519 520 521
int __cookie_v6_check(const struct ipv6hdr *iph, const struct tcphdr *th,
		      u32 cookie);
struct sock *cookie_v6_check(struct sock *sk, struct sk_buff *skb);
522
#ifdef CONFIG_SYN_COOKIES
523 524 525 526
u32 __cookie_v6_init_sequence(const struct ipv6hdr *iph,
			      const struct tcphdr *th, u16 *mssp);
__u32 cookie_v6_init_sequence(struct sock *sk, const struct sk_buff *skb,
			      __u16 *mss);
527 528 529 530 531 532 533 534
#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
535 536
/* tcp_output.c */

537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560
void __tcp_push_pending_frames(struct sock *sk, unsigned int cur_mss,
			       int nonagle);
bool tcp_may_send_now(struct sock *sk);
int __tcp_retransmit_skb(struct sock *, struct sk_buff *);
int tcp_retransmit_skb(struct sock *, struct sk_buff *);
void tcp_retransmit_timer(struct sock *sk);
void tcp_xmit_retransmit_queue(struct sock *);
void tcp_simple_retransmit(struct sock *);
int tcp_trim_head(struct sock *, struct sk_buff *, u32);
int tcp_fragment(struct sock *, struct sk_buff *, u32, unsigned int);

void tcp_send_probe0(struct sock *);
void tcp_send_partial(struct sock *);
int tcp_write_wakeup(struct sock *);
void tcp_send_fin(struct sock *sk);
void tcp_send_active_reset(struct sock *sk, gfp_t priority);
int tcp_send_synack(struct sock *);
bool tcp_syn_flood_action(struct sock *sk, const struct sk_buff *skb,
			  const char *proto);
void tcp_push_one(struct sock *, unsigned int mss_now);
void tcp_send_ack(struct sock *sk);
void tcp_send_delayed_ack(struct sock *sk);
void tcp_send_loss_probe(struct sock *sk);
bool tcp_schedule_loss_probe(struct sock *sk);
L
Linus Torvalds 已提交
561

562
/* tcp_input.c */
563 564 565 566
void tcp_cwnd_application_limited(struct sock *sk);
void tcp_resume_early_retransmit(struct sock *sk);
void tcp_rearm_rto(struct sock *sk);
void tcp_reset(struct sock *sk);
567

L
Linus Torvalds 已提交
568
/* tcp_timer.c */
569
void tcp_init_xmit_timers(struct sock *);
570 571 572 573
static inline void tcp_clear_xmit_timers(struct sock *sk)
{
	inet_csk_clear_xmit_timers(sk);
}
L
Linus Torvalds 已提交
574

575 576
unsigned int tcp_sync_mss(struct sock *sk, u32 pmtu);
unsigned int tcp_current_mss(struct sock *sk);
I
Ilpo Järvinen 已提交
577 578 579 580

/* 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)
{
581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596
	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
597 598 599
	else
		return pktsize;
}
L
Linus Torvalds 已提交
600

601
/* tcp.c */
602
void tcp_get_info(const struct sock *, struct tcp_info *);
L
Linus Torvalds 已提交
603 604 605 606

/* Read 'sendfile()'-style from a TCP socket */
typedef int (*sk_read_actor_t)(read_descriptor_t *, struct sk_buff *,
				unsigned int, size_t);
607 608
int tcp_read_sock(struct sock *sk, read_descriptor_t *desc,
		  sk_read_actor_t recv_actor);
L
Linus Torvalds 已提交
609

610
void tcp_initialize_rcv_mss(struct sock *sk);
L
Linus Torvalds 已提交
611

612 613 614 615
int tcp_mtu_to_mss(struct sock *sk, int pmtu);
int tcp_mss_to_mtu(struct sock *sk, int mss);
void tcp_mtup_init(struct sock *sk);
void tcp_init_buffer_space(struct sock *sk);
J
John Heffner 已提交
616

617 618 619 620 621 622 623 624 625 626 627
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;
}

628
void tcp_set_rto(struct sock *sk);
629

S
Stephen Hemminger 已提交
630
static inline void __tcp_fast_path_on(struct tcp_sock *tp, u32 snd_wnd)
L
Linus Torvalds 已提交
631 632 633 634 635 636
{
	tp->pred_flags = htonl((tp->tcp_header_len << 26) |
			       ntohl(TCP_FLAG_ACK) |
			       snd_wnd);
}

S
Stephen Hemminger 已提交
637
static inline void tcp_fast_path_on(struct tcp_sock *tp)
L
Linus Torvalds 已提交
638 639 640 641
{
	__tcp_fast_path_on(tp, tp->snd_wnd >> tp->rx_opt.snd_wscale);
}

642
static inline void tcp_fast_path_check(struct sock *sk)
L
Linus Torvalds 已提交
643
{
644 645
	struct tcp_sock *tp = tcp_sk(sk);

646
	if (skb_queue_empty(&tp->out_of_order_queue) &&
L
Linus Torvalds 已提交
647 648 649 650 651 652
	    tp->rcv_wnd &&
	    atomic_read(&sk->sk_rmem_alloc) < sk->sk_rcvbuf &&
	    !tp->urg_data)
		tcp_fast_path_on(tp);
}

653 654 655
/* Compute the actual rto_min value */
static inline u32 tcp_rto_min(struct sock *sk)
{
656
	const struct dst_entry *dst = __sk_dst_get(sk);
657 658 659 660 661 662 663
	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
664 665 666 667
/* 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
668
static inline u32 tcp_receive_window(const struct tcp_sock *tp)
L
Linus Torvalds 已提交
669 670 671 672 673 674 675 676 677 678 679 680
{
	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.
 */
681
u32 __tcp_select_window(struct sock *sk);
L
Linus Torvalds 已提交
682

P
Pavel Emelyanov 已提交
683 684
void tcp_send_window_probe(struct sock *sk);

L
Linus Torvalds 已提交
685 686
/* 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
687 688
 * 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
689 690 691 692
 * casts with the following macro.
 */
#define tcp_time_stamp		((__u32)(jiffies))

C
Changli Gao 已提交
693 694 695 696 697 698 699 700 701 702 703
#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
704
/* This is what the send packet queuing engine uses to pass
705 706 707 708
 * 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
709 710 711 712
 */
struct tcp_skb_cb {
	union {
		struct inet_skb_parm	h4;
E
Eric Dumazet 已提交
713
#if IS_ENABLED(CONFIG_IPV6)
L
Linus Torvalds 已提交
714 715 716 717 718 719
		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
720
	__u8		tcp_flags;	/* TCP header flags. (tcp[13])	*/
721

L
Linus Torvalds 已提交
722 723 724 725 726 727 728 729
	__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)

730 731
	__u8		ip_dsfield;	/* IPv4 tos or IPv6 dsfield	*/
	/* 1 byte hole */
L
Linus Torvalds 已提交
732 733 734 735 736
	__u32		ack_seq;	/* Sequence number ACK'd	*/
};

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

737 738 739 740 741 742 743
/* 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
744 745
TCP_ECN_create_request(struct request_sock *req, const struct sk_buff *skb,
		struct net *net)
746 747 748
{
	const struct tcphdr *th = tcp_hdr(skb);

749
	if (net->ipv4.sysctl_tcp_ecn && th->ece && th->cwr &&
750 751 752 753
	    INET_ECN_is_not_ect(TCP_SKB_CB(skb)->ip_dsfield))
		inet_rsk(req)->ecn_ok = 1;
}

L
Linus Torvalds 已提交
754 755 756 757 758
/* 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)
{
759
	return skb_shinfo(skb)->gso_segs;
L
Linus Torvalds 已提交
760 761 762 763 764
}

/* This is valid iff tcp_skb_pcount() > 1. */
static inline int tcp_skb_mss(const struct sk_buff *skb)
{
765
	return skb_shinfo(skb)->gso_size;
L
Linus Torvalds 已提交
766 767
}

768 769 770 771 772 773 774 775 776 777 778 779 780 781
/* 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_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
782 783 784
#define TCP_CA_MAX	128
#define TCP_CA_BUF_MAX	(TCP_CA_NAME_MAX*TCP_CA_MAX)

785 786 787
#define TCP_CONG_NON_RESTRICTED 0x1
#define TCP_CONG_RTT_STAMP	0x2

788 789
struct tcp_congestion_ops {
	struct list_head	list;
790
	unsigned long flags;
791 792

	/* initialize private data (optional) */
793
	void (*init)(struct sock *sk);
794
	/* cleanup private data  (optional) */
795
	void (*release)(struct sock *sk);
796 797

	/* return slow start threshold (required) */
798
	u32 (*ssthresh)(struct sock *sk);
799
	/* lower bound for congestion window (optional) */
800
	u32 (*min_cwnd)(const struct sock *sk);
801
	/* do new cwnd calculation (required) */
802
	void (*cong_avoid)(struct sock *sk, u32 ack, u32 in_flight);
803
	/* call before changing ca_state (optional) */
804
	void (*set_state)(struct sock *sk, u8 new_state);
805
	/* call when cwnd event occurs (optional) */
806
	void (*cwnd_event)(struct sock *sk, enum tcp_ca_event ev);
807
	/* new value of cwnd after loss (optional) */
808
	u32  (*undo_cwnd)(struct sock *sk);
809
	/* hook for packet ack accounting (optional) */
810
	void (*pkts_acked)(struct sock *sk, u32 num_acked, s32 rtt_us);
811
	/* get info for inet_diag (optional) */
812
	void (*get_info)(struct sock *sk, u32 ext, struct sk_buff *skb);
813 814 815 816 817

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

818 819
int tcp_register_congestion_control(struct tcp_congestion_ops *type);
void tcp_unregister_congestion_control(struct tcp_congestion_ops *type);
820

821 822 823 824 825 826 827 828 829 830
void tcp_init_congestion_control(struct sock *sk);
void tcp_cleanup_congestion_control(struct sock *sk);
int tcp_set_default_congestion_control(const char *name);
void tcp_get_default_congestion_control(char *name);
void tcp_get_available_congestion_control(char *buf, size_t len);
void tcp_get_allowed_congestion_control(char *buf, size_t len);
int tcp_set_allowed_congestion_control(char *allowed);
int tcp_set_congestion_control(struct sock *sk, const char *name);
void tcp_slow_start(struct tcp_sock *tp);
void tcp_cong_avoid_ai(struct tcp_sock *tp, u32 w);
831

832
extern struct tcp_congestion_ops tcp_init_congestion_ops;
833 834 835
u32 tcp_reno_ssthresh(struct sock *sk);
void tcp_reno_cong_avoid(struct sock *sk, u32 ack, u32 in_flight);
u32 tcp_reno_min_cwnd(const struct sock *sk);
836
extern struct tcp_congestion_ops tcp_reno;
837

838
static inline void tcp_set_ca_state(struct sock *sk, const u8 ca_state)
839
{
840 841 842 843 844
	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;
845 846
}

847
static inline void tcp_ca_event(struct sock *sk, const enum tcp_ca_event event)
848
{
849 850 851 852
	const struct inet_connection_sock *icsk = inet_csk(sk);

	if (icsk->icsk_ca_ops->cwnd_event)
		icsk->icsk_ca_ops->cwnd_event(sk, event);
853 854
}

855 856 857 858 859 860 861 862 863 864 865 866 867
/* 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
868
static inline bool tcp_is_reno(const struct tcp_sock *tp)
869 870 871 872
{
	return !tcp_is_sack(tp);
}

E
Eric Dumazet 已提交
873
static inline bool tcp_is_fack(const struct tcp_sock *tp)
874
{
875
	return tp->rx_opt.sack_ok & TCP_FACK_ENABLED;
876 877 878 879
}

static inline void tcp_enable_fack(struct tcp_sock *tp)
{
880
	tp->rx_opt.sack_ok |= TCP_FACK_ENABLED;
881 882
}

Y
Yuchung Cheng 已提交
883 884 885 886 887 888
/* 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 &&
N
Nandita Dukkipati 已提交
889 890
		sysctl_tcp_early_retrans < 4 && !sysctl_tcp_thin_dupack &&
		sysctl_tcp_reordering == 3;
Y
Yuchung Cheng 已提交
891 892 893 894 895 896 897
}

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

898 899 900 901 902
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
903 904 905 906 907 908 909 910 911 912 913 914 915 916
/* 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
917
static inline unsigned int tcp_packets_in_flight(const struct tcp_sock *tp)
L
Linus Torvalds 已提交
918
{
919
	return tp->packets_out - tcp_left_out(tp) + tp->retrans_out;
L
Linus Torvalds 已提交
920 921
}

I
Ilpo Järvinen 已提交
922 923 924 925 926 927 928
#define TCP_INFINITE_SSTHRESH	0x7fffffff

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

929 930 931 932 933 934
static inline bool tcp_in_cwnd_reduction(const struct sock *sk)
{
	return (TCPF_CA_CWR | TCPF_CA_Recovery) &
	       (1 << inet_csk(sk)->icsk_ca_state);
}

L
Linus Torvalds 已提交
935
/* If cwnd > ssthresh, we may raise ssthresh to be half-way to cwnd.
936
 * The exception is cwnd reduction phase, when cwnd is decreasing towards
L
Linus Torvalds 已提交
937 938
 * ssthresh.
 */
939
static inline __u32 tcp_current_ssthresh(const struct sock *sk)
L
Linus Torvalds 已提交
940
{
941
	const struct tcp_sock *tp = tcp_sk(sk);
942

943
	if (tcp_in_cwnd_reduction(sk))
L
Linus Torvalds 已提交
944 945 946 947 948 949 950
		return tp->snd_ssthresh;
	else
		return max(tp->snd_ssthresh,
			   ((tp->snd_cwnd >> 1) +
			    (tp->snd_cwnd >> 2)));
}

951 952
/* 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
953

954 955
void tcp_enter_cwr(struct sock *sk, const int set_ssthresh);
__u32 tcp_init_cwnd(const struct tcp_sock *tp, const struct dst_entry *dst);
L
Linus Torvalds 已提交
956

957 958 959 960 961 962 963 964
/* 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
965
/* Slow start with delack produces 3 packets of burst, so that
966 967 968 969
 * 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
970 971 972
 */
static __inline__ __u32 tcp_max_burst(const struct tcp_sock *tp)
{
973
	return tp->reordering;
L
Linus Torvalds 已提交
974 975
}

976 977 978 979 980
/* 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;
}
981
bool tcp_is_cwnd_limited(const struct sock *sk, u32 in_flight);
982

983
static inline void tcp_minshall_update(struct tcp_sock *tp, unsigned int mss,
S
Stephen Hemminger 已提交
984
				       const struct sk_buff *skb)
L
Linus Torvalds 已提交
985 986 987 988 989
{
	if (skb->len < mss)
		tp->snd_sml = TCP_SKB_CB(skb)->end_seq;
}

990
static inline void tcp_check_probe_timer(struct sock *sk)
L
Linus Torvalds 已提交
991
{
992
	const struct tcp_sock *tp = tcp_sk(sk);
993
	const struct inet_connection_sock *icsk = inet_csk(sk);
994

995
	if (!tp->packets_out && !icsk->icsk_pending)
996 997
		inet_csk_reset_xmit_timer(sk, ICSK_TIME_PROBE0,
					  icsk->icsk_rto, TCP_RTO_MAX);
L
Linus Torvalds 已提交
998 999
}

1000
static inline void tcp_init_wl(struct tcp_sock *tp, u32 seq)
L
Linus Torvalds 已提交
1001 1002 1003 1004
{
	tp->snd_wl1 = seq;
}

1005
static inline void tcp_update_wl(struct tcp_sock *tp, u32 seq)
L
Linus Torvalds 已提交
1006 1007 1008 1009 1010 1011 1012
{
	tp->snd_wl1 = seq;
}

/*
 * Calculate(/check) TCP checksum
 */
1013 1014
static inline __sum16 tcp_v4_check(int len, __be32 saddr,
				   __be32 daddr, __wsum base)
L
Linus Torvalds 已提交
1015 1016 1017 1018
{
	return csum_tcpudp_magic(saddr,daddr,len,IPPROTO_TCP,base);
}

1019
static inline __sum16 __tcp_checksum_complete(struct sk_buff *skb)
L
Linus Torvalds 已提交
1020
{
1021
	return __skb_checksum_complete(skb);
L
Linus Torvalds 已提交
1022 1023
}

E
Eric Dumazet 已提交
1024
static inline bool tcp_checksum_complete(struct sk_buff *skb)
L
Linus Torvalds 已提交
1025
{
1026
	return !skb_csum_unnecessary(skb) &&
L
Linus Torvalds 已提交
1027 1028 1029 1030 1031
		__tcp_checksum_complete(skb);
}

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

S
Stephen Hemminger 已提交
1032
static inline void tcp_prequeue_init(struct tcp_sock *tp)
L
Linus Torvalds 已提交
1033 1034 1035 1036 1037
{
	tp->ucopy.task = NULL;
	tp->ucopy.len = 0;
	tp->ucopy.memory = 0;
	skb_queue_head_init(&tp->ucopy.prequeue);
1038 1039 1040 1041 1042 1043
#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
1044 1045
}

1046
bool tcp_prequeue(struct sock *sk, struct sk_buff *skb);
L
Linus Torvalds 已提交
1047 1048 1049 1050 1051 1052 1053 1054 1055 1056

#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
1057
void tcp_set_state(struct sock *sk, int state);
L
Linus Torvalds 已提交
1058

1059
void tcp_done(struct sock *sk);
L
Linus Torvalds 已提交
1060

S
Stephen Hemminger 已提交
1061
static inline void tcp_sack_reset(struct tcp_options_received *rx_opt)
L
Linus Torvalds 已提交
1062 1063 1064 1065 1066
{
	rx_opt->dsack = 0;
	rx_opt->num_sacks = 0;
}

1067
u32 tcp_default_init_rwnd(u32 mss);
1068

L
Linus Torvalds 已提交
1069
/* Determine a window scaling and initial window to offer. */
1070 1071 1072
void tcp_select_initial_window(int __space, __u32 mss, __u32 *rcv_wnd,
			       __u32 *window_clamp, int wscale_ok,
			       __u8 *rcv_wscale, __u32 init_rcv_wnd);
L
Linus Torvalds 已提交
1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092

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
1093 1094 1095
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
1096
{
1097 1098
	struct inet_request_sock *ireq = inet_rsk(req);

L
Linus Torvalds 已提交
1099
	req->rcv_wnd = 0;		/* So that tcp_send_synack() knows! */
1100
	req->cookie_ts = 0;
1101
	tcp_rsk(req)->rcv_isn = TCP_SKB_CB(skb)->seq;
1102
	tcp_rsk(req)->rcv_nxt = TCP_SKB_CB(skb)->seq + 1;
1103
	tcp_rsk(req)->snt_synack = 0;
L
Linus Torvalds 已提交
1104 1105
	req->mss = rx_opt->mss_clamp;
	req->ts_recent = rx_opt->saw_tstamp ? rx_opt->rcv_tsval : 0;
1106 1107 1108 1109 1110 1111
	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;
1112 1113
	ireq->ir_rmt_port = tcp_hdr(skb)->source;
	ireq->ir_loc_port = tcp_hdr(skb)->dest;
L
Linus Torvalds 已提交
1114 1115
}

1116
void tcp_enter_memory_pressure(struct sock *sk);
L
Linus Torvalds 已提交
1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127

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
1128 1129 1130 1131 1132
static inline int keepalive_probes(const struct tcp_sock *tp)
{
	return tp->keepalive_probes ? : sysctl_tcp_keepalive_probes;
}

1133 1134 1135 1136 1137 1138 1139 1140
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);
}

1141
static inline int tcp_fin_time(const struct sock *sk)
L
Linus Torvalds 已提交
1142
{
1143 1144
	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
1145

1146 1147
	if (fin_timeout < (rto << 2) - (rto >> 1))
		fin_timeout = (rto << 2) - (rto >> 1);
L
Linus Torvalds 已提交
1148 1149 1150 1151

	return fin_timeout;
}

E
Eric Dumazet 已提交
1152 1153
static inline bool tcp_paws_check(const struct tcp_options_received *rx_opt,
				  int paws_win)
L
Linus Torvalds 已提交
1154
{
I
Ilpo Järvinen 已提交
1155
	if ((s32)(rx_opt->ts_recent - rx_opt->rcv_tsval) <= paws_win)
E
Eric Dumazet 已提交
1156
		return true;
I
Ilpo Järvinen 已提交
1157
	if (unlikely(get_seconds() >= rx_opt->ts_recent_stamp + TCP_PAWS_24DAYS))
E
Eric Dumazet 已提交
1158
		return true;
E
Eric Dumazet 已提交
1159 1160 1161 1162 1163 1164
	/*
	 * 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
1165 1166
		return true;
	return false;
I
Ilpo Järvinen 已提交
1167 1168
}

E
Eric Dumazet 已提交
1169 1170
static inline bool tcp_paws_reject(const struct tcp_options_received *rx_opt,
				   int rst)
I
Ilpo Järvinen 已提交
1171 1172
{
	if (tcp_paws_check(rx_opt, 0))
E
Eric Dumazet 已提交
1173
		return false;
L
Linus Torvalds 已提交
1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186

	/* 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.
	 */
1187
	if (rst && get_seconds() >= rx_opt->ts_recent_stamp + TCP_PAWS_MSL)
E
Eric Dumazet 已提交
1188 1189
		return false;
	return true;
L
Linus Torvalds 已提交
1190 1191
}

P
Pavel Emelyanov 已提交
1192
static inline void tcp_mib_init(struct net *net)
L
Linus Torvalds 已提交
1193 1194
{
	/* See RFC 2012 */
1195 1196 1197 1198
	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
1199 1200
}

1201
/* from STCP */
1202
static inline void tcp_clear_retrans_hints_partial(struct tcp_sock *tp)
D
David S. Miller 已提交
1203
{
1204
	tp->lost_skb_hint = NULL;
1205 1206 1207 1208 1209
}

static inline void tcp_clear_all_retrans_hints(struct tcp_sock *tp)
{
	tcp_clear_retrans_hints_partial(tp);
1210
	tp->retransmit_skb_hint = NULL;
1211 1212
}

1213 1214 1215
/* MD5 Signature */
struct crypto_hash;

E
Eric Dumazet 已提交
1216 1217 1218 1219 1220 1221 1222
union tcp_md5_addr {
	struct in_addr  a4;
#if IS_ENABLED(CONFIG_IPV6)
	struct in6_addr	a6;
#endif
};

1223 1224
/* - key database */
struct tcp_md5sig_key {
E
Eric Dumazet 已提交
1225
	struct hlist_node	node;
1226
	u8			keylen;
E
Eric Dumazet 已提交
1227 1228 1229 1230
	u8			family; /* AF_INET or AF_INET6 */
	union tcp_md5_addr	addr;
	u8			key[TCP_MD5SIG_MAXKEYLEN];
	struct rcu_head		rcu;
1231 1232 1233 1234
};

/* - sock block */
struct tcp_md5sig_info {
E
Eric Dumazet 已提交
1235
	struct hlist_head	head;
1236
	struct rcu_head		rcu;
1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256
};

/* - 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
1257
#if IS_ENABLED(CONFIG_IPV6)
1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268
	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 */
1269 1270 1271 1272 1273 1274 1275 1276
int tcp_v4_md5_hash_skb(char *md5_hash, struct tcp_md5sig_key *key,
			const struct sock *sk, const struct request_sock *req,
			const struct sk_buff *skb);
int tcp_md5_do_add(struct sock *sk, const union tcp_md5_addr *addr,
		   int family, const u8 *newkey, u8 newkeylen, gfp_t gfp);
int tcp_md5_do_del(struct sock *sk, const union tcp_md5_addr *addr,
		   int family);
struct tcp_md5sig_key *tcp_v4_md5_lookup(struct sock *sk,
E
Eric Dumazet 已提交
1277
					 struct sock *addr_sk);
1278

1279
#ifdef CONFIG_TCP_MD5SIG
1280 1281 1282
struct tcp_md5sig_key *tcp_md5_do_lookup(struct sock *sk,
					 const union tcp_md5_addr *addr,
					 int family);
E
Eric Dumazet 已提交
1283
#define tcp_twsk_md5_key(twsk)	((twsk)->tw_md5_key)
1284
#else
E
Eric Dumazet 已提交
1285 1286 1287 1288 1289 1290
static inline struct tcp_md5sig_key *tcp_md5_do_lookup(struct sock *sk,
					 const union tcp_md5_addr *addr,
					 int family)
{
	return NULL;
}
1291 1292 1293
#define tcp_twsk_md5_key(twsk)	NULL
#endif

1294
bool tcp_alloc_md5sig_pool(void);
1295

1296
struct tcp_md5sig_pool *tcp_get_md5sig_pool(void);
1297 1298 1299 1300
static inline void tcp_put_md5sig_pool(void)
{
	local_bh_enable();
}
E
Eric Dumazet 已提交
1301

1302 1303 1304 1305 1306
int tcp_md5_hash_header(struct tcp_md5sig_pool *, const struct tcphdr *);
int tcp_md5_hash_skb_data(struct tcp_md5sig_pool *, const struct sk_buff *,
			  unsigned int header_len);
int tcp_md5_hash_key(struct tcp_md5sig_pool *hp,
		     const struct tcp_md5sig_key *key);
1307

1308
/* From tcp_fastopen.c */
1309 1310 1311 1312 1313
void tcp_fastopen_cache_get(struct sock *sk, u16 *mss,
			    struct tcp_fastopen_cookie *cookie, int *syn_loss,
			    unsigned long *last_syn_loss);
void tcp_fastopen_cache_set(struct sock *sk, u16 mss,
			    struct tcp_fastopen_cookie *cookie, bool syn_lost);
1314 1315 1316 1317 1318 1319 1320 1321
struct tcp_fastopen_request {
	/* Fast Open cookie. Size 0 means a cookie request */
	struct tcp_fastopen_cookie	cookie;
	struct msghdr			*data;  /* data in MSG_FASTOPEN */
	u16				copied;	/* queued in tcp_connect() */
};
void tcp_free_fastopen_req(struct tcp_sock *tp);

1322 1323
extern struct tcp_fastopen_context __rcu *tcp_fastopen_ctx;
int tcp_fastopen_reset_cipher(void *key, unsigned int len);
1324 1325
void tcp_fastopen_cookie_gen(__be32 src, __be32 dst,
			     struct tcp_fastopen_cookie *foc);
1326 1327 1328 1329 1330

#define TCP_FASTOPEN_KEY_LENGTH 16

/* Fastopen key context */
struct tcp_fastopen_context {
1331 1332 1333
	struct crypto_cipher	*tfm;
	__u8			key[TCP_FASTOPEN_KEY_LENGTH];
	struct rcu_head		rcu;
1334 1335
};

1336 1337 1338 1339 1340 1341
/* 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)
1342 1343
		sk_wmem_free_skb(sk, skb);
	sk_mem_reclaim(sk);
1344
	tcp_clear_all_retrans_hints(tcp_sk(sk));
1345 1346
}

1347
static inline struct sk_buff *tcp_write_queue_head(const struct sock *sk)
1348
{
1349
	return skb_peek(&sk->sk_write_queue);
1350 1351
}

1352
static inline struct sk_buff *tcp_write_queue_tail(const struct sock *sk)
1353
{
1354
	return skb_peek_tail(&sk->sk_write_queue);
1355 1356
}

1357 1358
static inline struct sk_buff *tcp_write_queue_next(const struct sock *sk,
						   const struct sk_buff *skb)
1359
{
1360
	return skb_queue_next(&sk->sk_write_queue, skb);
1361 1362
}

1363 1364
static inline struct sk_buff *tcp_write_queue_prev(const struct sock *sk,
						   const struct sk_buff *skb)
1365 1366 1367 1368
{
	return skb_queue_prev(&sk->sk_write_queue, skb);
}

1369
#define tcp_for_write_queue(skb, sk)					\
1370
	skb_queue_walk(&(sk)->sk_write_queue, skb)
1371 1372

#define tcp_for_write_queue_from(skb, sk)				\
1373
	skb_queue_walk_from(&(sk)->sk_write_queue, skb)
1374

1375
#define tcp_for_write_queue_from_safe(skb, tmp, sk)			\
1376
	skb_queue_walk_from_safe(&(sk)->sk_write_queue, skb, tmp)
1377

1378
static inline struct sk_buff *tcp_send_head(const struct sock *sk)
1379 1380 1381 1382
{
	return sk->sk_send_head;
}

1383 1384 1385 1386 1387 1388
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);
}

1389
static inline void tcp_advance_send_head(struct sock *sk, const struct sk_buff *skb)
1390
{
1391
	if (tcp_skb_is_last(sk, skb))
1392
		sk->sk_send_head = NULL;
1393 1394
	else
		sk->sk_send_head = tcp_write_queue_next(sk, skb);
1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417
}

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. */
1418
	if (sk->sk_send_head == NULL) {
1419
		sk->sk_send_head = skb;
1420 1421 1422 1423

		if (tcp_sk(sk)->highest_sack == NULL)
			tcp_sk(sk)->highest_sack = skb;
	}
1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435
}

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)
{
1436
	__skb_queue_after(&sk->sk_write_queue, skb, buff);
1437 1438
}

1439
/* Insert new before skb on the write queue of sk.  */
1440 1441 1442 1443
static inline void tcp_insert_write_queue_before(struct sk_buff *new,
						  struct sk_buff *skb,
						  struct sock *sk)
{
1444
	__skb_queue_before(&sk->sk_write_queue, skb, new);
1445 1446 1447

	if (sk->sk_send_head == skb)
		sk->sk_send_head = new;
1448 1449 1450 1451 1452 1453 1454
}

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
1455
static inline bool tcp_write_queue_empty(struct sock *sk)
1456 1457 1458 1459
{
	return skb_queue_empty(&sk->sk_write_queue);
}

1460 1461 1462 1463 1464 1465 1466 1467 1468
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);
	}
}

1469 1470 1471
/* 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.
1472 1473 1474 1475 1476
 */
static inline u32 tcp_highest_sack_seq(struct tcp_sock *tp)
{
	if (!tp->sacked_out)
		return tp->snd_una;
1477 1478 1479 1480

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

1481 1482 1483
	return TCP_SKB_CB(tp->highest_sack)->seq;
}

1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508
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
1509 1510 1511
/* 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
1512
static inline bool tcp_stream_is_thin(struct tcp_sock *tp)
A
Andreas Petlund 已提交
1513 1514 1515 1516
{
	return tp->packets_out < 4 && !tcp_in_initial_slowstart(tp);
}

L
Linus Torvalds 已提交
1517 1518 1519 1520 1521 1522 1523
/* /proc */
enum tcp_seq_states {
	TCP_SEQ_STATE_LISTENING,
	TCP_SEQ_STATE_OPENREQ,
	TCP_SEQ_STATE_ESTABLISHED,
};

1524 1525
int tcp_seq_open(struct inode *inode, struct file *file);

L
Linus Torvalds 已提交
1526
struct tcp_seq_afinfo {
1527 1528 1529 1530
	char				*name;
	sa_family_t			family;
	const struct file_operations	*seq_fops;
	struct seq_operations		seq_ops;
L
Linus Torvalds 已提交
1531 1532 1533
};

struct tcp_iter_state {
1534
	struct seq_net_private	p;
L
Linus Torvalds 已提交
1535 1536 1537
	sa_family_t		family;
	enum tcp_seq_states	state;
	struct sock		*syn_wait_sk;
1538 1539
	int			bucket, offset, sbucket, num;
	kuid_t			uid;
1540
	loff_t			last_pos;
L
Linus Torvalds 已提交
1541 1542
};

1543 1544
int tcp_proc_register(struct net *net, struct tcp_seq_afinfo *afinfo);
void tcp_proc_unregister(struct net *net, struct tcp_seq_afinfo *afinfo);
L
Linus Torvalds 已提交
1545

1546
extern struct request_sock_ops tcp_request_sock_ops;
1547
extern struct request_sock_ops tcp6_request_sock_ops;
1548

1549
void tcp_v4_destroy_sock(struct sock *sk);
1550

1551 1552 1553 1554
struct sk_buff *tcp_tso_segment(struct sk_buff *skb,
				netdev_features_t features);
struct sk_buff **tcp_gro_receive(struct sk_buff **head, struct sk_buff *skb);
int tcp_gro_complete(struct sk_buff *skb);
1555

1556
void __tcp_v4_send_check(struct sk_buff *skb, __be32 saddr, __be32 daddr);
H
Herbert Xu 已提交
1557

1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570
static inline u32 tcp_notsent_lowat(const struct tcp_sock *tp)
{
	return tp->notsent_lowat ?: sysctl_tcp_notsent_lowat;
}

static inline bool tcp_stream_memory_free(const struct sock *sk)
{
	const struct tcp_sock *tp = tcp_sk(sk);
	u32 notsent_bytes = tp->write_seq - tp->snd_nxt;

	return notsent_bytes < tcp_notsent_lowat(tp);
}

1571
#ifdef CONFIG_PROC_FS
1572 1573
int tcp4_proc_init(void);
void tcp4_proc_exit(void);
1574 1575
#endif

1576 1577 1578 1579 1580 1581 1582
/* 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
1583 1584 1585
						  const struct sock *sk,
						  const struct request_sock *req,
						  const struct sk_buff *skb);
1586 1587 1588 1589 1590 1591 1592 1593 1594 1595
	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);
1596 1597
	int			(*calc_md5_hash) (char *location,
						  struct tcp_md5sig_key *md5,
E
Eric Dumazet 已提交
1598 1599 1600
						  const struct sock *sk,
						  const struct request_sock *req,
						  const struct sk_buff *skb);
1601 1602 1603
#endif
};

1604
int tcpv4_offload_init(void);
1605

1606 1607
void tcp_v4_init(void);
void tcp_init(void);
1608

L
Linus Torvalds 已提交
1609
#endif	/* _TCP_H */