tcp_minisocks.c 23.0 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7
/*
 * 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.
 *
 *		Implementation of the Transmission Control Protocol(TCP).
 *
8
 * Authors:	Ross Biro
L
Linus Torvalds 已提交
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
 *		Fred N. van Kempen, <waltje@uWalt.NL.Mugnet.ORG>
 *		Mark Evans, <evansmp@uhura.aston.ac.uk>
 *		Corey Minyard <wf-rch!minyard@relay.EU.net>
 *		Florian La Roche, <flla@stud.uni-sb.de>
 *		Charles Hedrick, <hedrick@klinzhai.rutgers.edu>
 *		Linus Torvalds, <torvalds@cs.helsinki.fi>
 *		Alan Cox, <gw4pts@gw4pts.ampr.org>
 *		Matthew Dillon, <dillon@apollo.west.oic.com>
 *		Arnt Gulbrandsen, <agulbra@nvg.unit.no>
 *		Jorge Cwik, <jorge@laser.satlink.net>
 */

#include <linux/mm.h>
#include <linux/module.h>
#include <linux/sysctl.h>
#include <linux/workqueue.h>
#include <net/tcp.h>
#include <net/inet_common.h>
#include <net/xfrm.h>

29
int sysctl_tcp_syncookies __read_mostly = 1;
30 31
EXPORT_SYMBOL(sysctl_tcp_syncookies);

32
int sysctl_tcp_abort_on_overflow __read_mostly;
L
Linus Torvalds 已提交
33

34 35 36
struct inet_timewait_death_row tcp_death_row = {
	.sysctl_max_tw_buckets = NR_FILE * 2,
	.period		= TCP_TIMEWAIT_LEN / INET_TWDR_TWKILL_SLOTS,
37
	.death_lock	= __SPIN_LOCK_UNLOCKED(tcp_death_row.death_lock),
38 39 40 41
	.hashinfo	= &tcp_hashinfo,
	.tw_timer	= TIMER_INITIALIZER(inet_twdr_hangman, 0,
					    (unsigned long)&tcp_death_row),
	.twkill_work	= __WORK_INITIALIZER(tcp_death_row.twkill_work,
42
					     inet_twdr_twkill_work),
43 44 45 46 47 48 49 50 51
/* Short-time timewait calendar */

	.twcal_hand	= -1,
	.twcal_timer	= TIMER_INITIALIZER(inet_twdr_twcal_tick, 0,
					    (unsigned long)&tcp_death_row),
};

EXPORT_SYMBOL_GPL(tcp_death_row);

L
Linus Torvalds 已提交
52 53 54 55 56 57 58 59 60
static __inline__ int tcp_in_window(u32 seq, u32 end_seq, u32 s_win, u32 e_win)
{
	if (seq == s_win)
		return 1;
	if (after(end_seq, s_win) && before(seq, e_win))
		return 1;
	return (seq == e_win && seq == end_seq);
}

61
/*
L
Linus Torvalds 已提交
62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89
 * * Main purpose of TIME-WAIT state is to close connection gracefully,
 *   when one of ends sits in LAST-ACK or CLOSING retransmitting FIN
 *   (and, probably, tail of data) and one or more our ACKs are lost.
 * * What is TIME-WAIT timeout? It is associated with maximal packet
 *   lifetime in the internet, which results in wrong conclusion, that
 *   it is set to catch "old duplicate segments" wandering out of their path.
 *   It is not quite correct. This timeout is calculated so that it exceeds
 *   maximal retransmission timeout enough to allow to lose one (or more)
 *   segments sent by peer and our ACKs. This time may be calculated from RTO.
 * * When TIME-WAIT socket receives RST, it means that another end
 *   finally closed and we are allowed to kill TIME-WAIT too.
 * * Second purpose of TIME-WAIT is catching old duplicate segments.
 *   Well, certainly it is pure paranoia, but if we load TIME-WAIT
 *   with this semantics, we MUST NOT kill TIME-WAIT state with RSTs.
 * * If we invented some more clever way to catch duplicates
 *   (f.e. based on PAWS), we could truncate TIME-WAIT to several RTOs.
 *
 * The algorithm below is based on FORMAL INTERPRETATION of RFCs.
 * When you compare it to RFCs, please, read section SEGMENT ARRIVES
 * from the very beginning.
 *
 * NOTE. With recycling (and later with fin-wait-2) TW bucket
 * is _not_ stateless. It means, that strictly speaking we must
 * spinlock it. I do not want! Well, probability of misbehaviour
 * is ridiculously low and, seems, we could use some mb() tricks
 * to avoid misread sequence numbers, states etc.  --ANK
 */
enum tcp_tw_status
90 91
tcp_timewait_state_process(struct inet_timewait_sock *tw, struct sk_buff *skb,
			   const struct tcphdr *th)
L
Linus Torvalds 已提交
92 93
{
	struct tcp_options_received tmp_opt;
94 95
	u8 *hash_location;
	struct tcp_timewait_sock *tcptw = tcp_twsk((struct sock *)tw);
L
Linus Torvalds 已提交
96 97
	int paws_reject = 0;

98
	if (th->doff > (sizeof(*th) >> 2) && tcptw->tw_ts_recent_stamp) {
99
		tmp_opt.tstamp_ok = 1;
100
		tcp_parse_options(skb, &tmp_opt, &hash_location, 1, NULL);
L
Linus Torvalds 已提交
101 102

		if (tmp_opt.saw_tstamp) {
103 104
			tmp_opt.ts_recent	= tcptw->tw_ts_recent;
			tmp_opt.ts_recent_stamp	= tcptw->tw_ts_recent_stamp;
I
Ilpo Järvinen 已提交
105
			paws_reject = tcp_paws_reject(&tmp_opt, th->rst);
L
Linus Torvalds 已提交
106 107 108 109 110 111 112 113 114
		}
	}

	if (tw->tw_substate == TCP_FIN_WAIT2) {
		/* Just repeat all the checks of tcp_rcv_state_process() */

		/* Out of window, send ACK */
		if (paws_reject ||
		    !tcp_in_window(TCP_SKB_CB(skb)->seq, TCP_SKB_CB(skb)->end_seq,
115 116
				   tcptw->tw_rcv_nxt,
				   tcptw->tw_rcv_nxt + tcptw->tw_rcv_wnd))
L
Linus Torvalds 已提交
117 118 119 120 121
			return TCP_TW_ACK;

		if (th->rst)
			goto kill;

122
		if (th->syn && !before(TCP_SKB_CB(skb)->seq, tcptw->tw_rcv_nxt))
L
Linus Torvalds 已提交
123 124 125
			goto kill_with_rst;

		/* Dup ACK? */
126 127
		if (!th->ack ||
		    !after(TCP_SKB_CB(skb)->end_seq, tcptw->tw_rcv_nxt) ||
L
Linus Torvalds 已提交
128
		    TCP_SKB_CB(skb)->end_seq == TCP_SKB_CB(skb)->seq) {
129
			inet_twsk_put(tw);
L
Linus Torvalds 已提交
130 131 132 133 134 135 136
			return TCP_TW_SUCCESS;
		}

		/* New data or FIN. If new data arrive after half-duplex close,
		 * reset.
		 */
		if (!th->fin ||
137
		    TCP_SKB_CB(skb)->end_seq != tcptw->tw_rcv_nxt + 1) {
L
Linus Torvalds 已提交
138
kill_with_rst:
139
			inet_twsk_deschedule(tw, &tcp_death_row);
140
			inet_twsk_put(tw);
L
Linus Torvalds 已提交
141 142 143 144
			return TCP_TW_RST;
		}

		/* FIN arrived, enter true time-wait state. */
145 146
		tw->tw_substate	  = TCP_TIME_WAIT;
		tcptw->tw_rcv_nxt = TCP_SKB_CB(skb)->end_seq;
L
Linus Torvalds 已提交
147
		if (tmp_opt.saw_tstamp) {
148
			tcptw->tw_ts_recent_stamp = get_seconds();
149
			tcptw->tw_ts_recent	  = tmp_opt.rcv_tsval;
L
Linus Torvalds 已提交
150 151 152 153 154
		}

		/* I am shamed, but failed to make it more elegant.
		 * Yes, it is direct reference to IP, which is impossible
		 * to generalize to IPv6. Taking into account that IPv6
S
Stephen Hemminger 已提交
155
		 * do not understand recycling in any case, it not
L
Linus Torvalds 已提交
156 157
		 * a big problem in practice. --ANK */
		if (tw->tw_family == AF_INET &&
158
		    tcp_death_row.sysctl_tw_recycle && tcptw->tw_ts_recent_stamp &&
L
Linus Torvalds 已提交
159
		    tcp_v4_tw_remember_stamp(tw))
160 161
			inet_twsk_schedule(tw, &tcp_death_row, tw->tw_timeout,
					   TCP_TIMEWAIT_LEN);
L
Linus Torvalds 已提交
162
		else
163 164
			inet_twsk_schedule(tw, &tcp_death_row, TCP_TIMEWAIT_LEN,
					   TCP_TIMEWAIT_LEN);
L
Linus Torvalds 已提交
165 166 167 168 169 170 171 172 173 174
		return TCP_TW_ACK;
	}

	/*
	 *	Now real TIME-WAIT state.
	 *
	 *	RFC 1122:
	 *	"When a connection is [...] on TIME-WAIT state [...]
	 *	[a TCP] MAY accept a new SYN from the remote TCP to
	 *	reopen the connection directly, if it:
175
	 *
L
Linus Torvalds 已提交
176 177 178 179 180
	 *	(1)  assigns its initial sequence number for the new
	 *	connection to be larger than the largest sequence
	 *	number it used on the previous connection incarnation,
	 *	and
	 *
181
	 *	(2)  returns to TIME-WAIT state if the SYN turns out
L
Linus Torvalds 已提交
182 183 184 185
	 *	to be an old duplicate".
	 */

	if (!paws_reject &&
186
	    (TCP_SKB_CB(skb)->seq == tcptw->tw_rcv_nxt &&
L
Linus Torvalds 已提交
187 188 189 190
	     (TCP_SKB_CB(skb)->seq == TCP_SKB_CB(skb)->end_seq || th->rst))) {
		/* In window segment, it may be only reset or bare ack. */

		if (th->rst) {
S
Stephen Hemminger 已提交
191
			/* This is TIME_WAIT assassination, in two flavors.
L
Linus Torvalds 已提交
192 193 194 195 196
			 * Oh well... nobody has a sufficient solution to this
			 * protocol bug yet.
			 */
			if (sysctl_tcp_rfc1337 == 0) {
kill:
197
				inet_twsk_deschedule(tw, &tcp_death_row);
198
				inet_twsk_put(tw);
L
Linus Torvalds 已提交
199 200 201
				return TCP_TW_SUCCESS;
			}
		}
202 203
		inet_twsk_schedule(tw, &tcp_death_row, TCP_TIMEWAIT_LEN,
				   TCP_TIMEWAIT_LEN);
L
Linus Torvalds 已提交
204 205

		if (tmp_opt.saw_tstamp) {
206
			tcptw->tw_ts_recent	  = tmp_opt.rcv_tsval;
207
			tcptw->tw_ts_recent_stamp = get_seconds();
L
Linus Torvalds 已提交
208 209
		}

210
		inet_twsk_put(tw);
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
		return TCP_TW_SUCCESS;
	}

	/* Out of window segment.

	   All the segments are ACKed immediately.

	   The only exception is new SYN. We accept it, if it is
	   not old duplicate and we are not in danger to be killed
	   by delayed old duplicates. RFC check is that it has
	   newer sequence number works at rates <40Mbit/sec.
	   However, if paws works, it is reliable AND even more,
	   we even may relax silly seq space cutoff.

	   RED-PEN: we violate main RFC requirement, if this SYN will appear
	   old duplicate (i.e. we receive RST in reply to SYN-ACK),
	   we must return socket to time-wait state. It is not good,
	   but not fatal yet.
	 */

	if (th->syn && !th->rst && !th->ack && !paws_reject &&
232 233 234 235
	    (after(TCP_SKB_CB(skb)->seq, tcptw->tw_rcv_nxt) ||
	     (tmp_opt.saw_tstamp &&
	      (s32)(tcptw->tw_ts_recent - tmp_opt.rcv_tsval) < 0))) {
		u32 isn = tcptw->tw_snd_nxt + 65535 + 2;
L
Linus Torvalds 已提交
236 237 238 239 240 241 242
		if (isn == 0)
			isn++;
		TCP_SKB_CB(skb)->when = isn;
		return TCP_TW_SYN;
	}

	if (paws_reject)
243
		NET_INC_STATS_BH(twsk_net(tw), LINUX_MIB_PAWSESTABREJECTED);
L
Linus Torvalds 已提交
244

S
Stephen Hemminger 已提交
245
	if (!th->rst) {
L
Linus Torvalds 已提交
246 247 248 249 250 251 252
		/* In this case we must reset the TIMEWAIT timer.
		 *
		 * If it is ACKless SYN it may be both old duplicate
		 * and new good SYN with random sequence number <rcv_nxt.
		 * Do not reschedule in the last case.
		 */
		if (paws_reject || th->ack)
253 254
			inet_twsk_schedule(tw, &tcp_death_row, TCP_TIMEWAIT_LEN,
					   TCP_TIMEWAIT_LEN);
L
Linus Torvalds 已提交
255 256 257 258 259 260

		/* Send ACK. Note, we do not put the bucket,
		 * it will be released by caller.
		 */
		return TCP_TW_ACK;
	}
261
	inet_twsk_put(tw);
L
Linus Torvalds 已提交
262 263 264
	return TCP_TW_SUCCESS;
}

265
/*
L
Linus Torvalds 已提交
266
 * Move a socket to time-wait or dead fin-wait-2 state.
267
 */
L
Linus Torvalds 已提交
268 269
void tcp_time_wait(struct sock *sk, int state, int timeo)
{
270
	struct inet_timewait_sock *tw = NULL;
271
	const struct inet_connection_sock *icsk = inet_csk(sk);
272
	const struct tcp_sock *tp = tcp_sk(sk);
L
Linus Torvalds 已提交
273 274
	int recycle_ok = 0;

275
	if (tcp_death_row.sysctl_tw_recycle && tp->rx_opt.ts_recent_stamp)
276
		recycle_ok = icsk->icsk_af_ops->remember_stamp(sk);
L
Linus Torvalds 已提交
277

278
	if (tcp_death_row.tw_count < tcp_death_row.sysctl_max_tw_buckets)
279
		tw = inet_twsk_alloc(sk, state);
L
Linus Torvalds 已提交
280

281 282
	if (tw != NULL) {
		struct tcp_timewait_sock *tcptw = tcp_twsk((struct sock *)tw);
283
		const int rto = (icsk->icsk_rto << 2) - (icsk->icsk_rto >> 1);
284

L
Linus Torvalds 已提交
285
		tw->tw_rcv_wscale	= tp->rx_opt.rcv_wscale;
286 287 288 289 290
		tcptw->tw_rcv_nxt	= tp->rcv_nxt;
		tcptw->tw_snd_nxt	= tp->snd_nxt;
		tcptw->tw_rcv_wnd	= tcp_receive_window(tp);
		tcptw->tw_ts_recent	= tp->rx_opt.ts_recent;
		tcptw->tw_ts_recent_stamp = tp->rx_opt.ts_recent_stamp;
L
Linus Torvalds 已提交
291 292 293 294

#if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE)
		if (tw->tw_family == PF_INET6) {
			struct ipv6_pinfo *np = inet6_sk(sk);
295
			struct inet6_timewait_sock *tw6;
L
Linus Torvalds 已提交
296

297 298 299 300
			tw->tw_ipv6_offset = inet6_tw_offset(sk->sk_prot);
			tw6 = inet6_twsk((struct sock *)tw);
			ipv6_addr_copy(&tw6->tw_v6_daddr, &np->daddr);
			ipv6_addr_copy(&tw6->tw_v6_rcv_saddr, &np->rcv_saddr);
301
			tw->tw_ipv6only = np->ipv6only;
302
		}
L
Linus Torvalds 已提交
303
#endif
304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319

#ifdef CONFIG_TCP_MD5SIG
		/*
		 * The timewait bucket does not have the key DB from the
		 * sock structure. We just make a quick copy of the
		 * md5 key being used (if indeed we are using one)
		 * so the timewait ack generating code has the key.
		 */
		do {
			struct tcp_md5sig_key *key;
			memset(tcptw->tw_md5_key, 0, sizeof(tcptw->tw_md5_key));
			tcptw->tw_md5_keylen = 0;
			key = tp->af_specific->md5_lookup(sk, sk);
			if (key != NULL) {
				memcpy(&tcptw->tw_md5_key, key->key, key->keylen);
				tcptw->tw_md5_keylen = key->keylen;
320
				if (tcp_alloc_md5sig_pool(sk) == NULL)
321 322
					BUG();
			}
S
Stephen Hemminger 已提交
323
		} while (0);
324 325
#endif

L
Linus Torvalds 已提交
326
		/* Linkage updates. */
327
		__inet_twsk_hashdance(tw, sk, &tcp_hashinfo);
L
Linus Torvalds 已提交
328 329 330 331 332 333 334 335 336 337 338 339 340

		/* Get the TIME_WAIT timeout firing. */
		if (timeo < rto)
			timeo = rto;

		if (recycle_ok) {
			tw->tw_timeout = rto;
		} else {
			tw->tw_timeout = TCP_TIMEWAIT_LEN;
			if (state == TCP_TIME_WAIT)
				timeo = TCP_TIMEWAIT_LEN;
		}

341 342
		inet_twsk_schedule(tw, &tcp_death_row, timeo,
				   TCP_TIMEWAIT_LEN);
343
		inet_twsk_put(tw);
L
Linus Torvalds 已提交
344 345 346 347 348
	} else {
		/* Sorry, if we're out of memory, just CLOSE this
		 * socket up.  We've got bigger problems than
		 * non-graceful socket closings.
		 */
349
		LIMIT_NETDEBUG(KERN_INFO "TCP: time wait bucket table overflow\n");
L
Linus Torvalds 已提交
350 351 352 353 354 355
	}

	tcp_update_metrics(sk);
	tcp_done(sk);
}

356 357 358
void tcp_twsk_destructor(struct sock *sk)
{
#ifdef CONFIG_TCP_MD5SIG
359
	struct tcp_timewait_sock *twsk = tcp_twsk(sk);
360
	if (twsk->tw_md5_keylen)
361
		tcp_free_md5sig_pool();
362 363 364 365 366
#endif
}

EXPORT_SYMBOL_GPL(tcp_twsk_destructor);

367 368 369 370 371 372
static inline void TCP_ECN_openreq_child(struct tcp_sock *tp,
					 struct request_sock *req)
{
	tp->ecn_flags = inet_rsk(req)->ecn_ok ? TCP_ECN_OK : 0;
}

L
Linus Torvalds 已提交
373 374 375 376 377 378
/* This is not only more efficient than what we used to do, it eliminates
 * a lot of code duplication between IPv4/IPv6 SYN recv processing. -DaveM
 *
 * Actually, we could lots of memory writes here. tp of listening
 * socket contains all necessary default parameters.
 */
379
struct sock *tcp_create_openreq_child(struct sock *sk, struct request_sock *req, struct sk_buff *skb)
L
Linus Torvalds 已提交
380
{
381
	struct sock *newsk = inet_csk_clone(sk, req, GFP_ATOMIC);
L
Linus Torvalds 已提交
382

383
	if (newsk != NULL) {
384
		const struct inet_request_sock *ireq = inet_rsk(req);
385
		struct tcp_request_sock *treq = tcp_rsk(req);
386
		struct inet_connection_sock *newicsk = inet_csk(newsk);
387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413
		struct tcp_sock *newtp = tcp_sk(newsk);
		struct tcp_sock *oldtp = tcp_sk(sk);
		struct tcp_cookie_values *oldcvp = oldtp->cookie_values;

		/* TCP Cookie Transactions require space for the cookie pair,
		 * as it differs for each connection.  There is no need to
		 * copy any s_data_payload stored at the original socket.
		 * Failure will prevent resuming the connection.
		 *
		 * Presumed copied, in order of appearance:
		 *	cookie_in_always, cookie_out_never
		 */
		if (oldcvp != NULL) {
			struct tcp_cookie_values *newcvp =
				kzalloc(sizeof(*newtp->cookie_values),
					GFP_ATOMIC);

			if (newcvp != NULL) {
				kref_init(&newcvp->kref);
				newcvp->cookie_desired =
						oldcvp->cookie_desired;
				newtp->cookie_values = newcvp;
			} else {
				/* Not Yet Implemented */
				newtp->cookie_values = NULL;
			}
		}
L
Linus Torvalds 已提交
414 415 416

		/* Now setup tcp_sock */
		newtp->pred_flags = 0;
417 418 419 420 421 422 423

		newtp->rcv_wup = newtp->copied_seq =
		newtp->rcv_nxt = treq->rcv_isn + 1;

		newtp->snd_sml = newtp->snd_una =
		newtp->snd_nxt = newtp->snd_up =
			treq->snt_isn + 1 + tcp_s_data_size(oldtp);
L
Linus Torvalds 已提交
424 425 426

		tcp_prequeue_init(newtp);

427
		tcp_init_wl(newtp, treq->rcv_isn);
L
Linus Torvalds 已提交
428 429 430

		newtp->srtt = 0;
		newtp->mdev = TCP_TIMEOUT_INIT;
431
		newicsk->icsk_rto = TCP_TIMEOUT_INIT;
L
Linus Torvalds 已提交
432 433 434 435 436

		newtp->packets_out = 0;
		newtp->retrans_out = 0;
		newtp->sacked_out = 0;
		newtp->fackets_out = 0;
I
Ilpo Järvinen 已提交
437
		newtp->snd_ssthresh = TCP_INFINITE_SSTHRESH;
L
Linus Torvalds 已提交
438 439 440 441 442 443 444 445

		/* So many TCP implementations out there (incorrectly) count the
		 * initial SYN frame in their delayed-ACK and congestion control
		 * algorithms that we must have the following bandaid to talk
		 * efficiently to them.  -DaveM
		 */
		newtp->snd_cwnd = 2;
		newtp->snd_cwnd_cnt = 0;
446
		newtp->bytes_acked = 0;
L
Linus Torvalds 已提交
447 448 449 450

		newtp->frto_counter = 0;
		newtp->frto_highmark = 0;

451
		newicsk->icsk_ca_ops = &tcp_init_congestion_ops;
452

453
		tcp_set_ca_state(newsk, TCP_CA_Open);
L
Linus Torvalds 已提交
454 455
		tcp_init_xmit_timers(newsk);
		skb_queue_head_init(&newtp->out_of_order_queue);
456 457
		newtp->write_seq = newtp->pushed_seq =
			treq->snt_isn + 1 + tcp_s_data_size(oldtp);
L
Linus Torvalds 已提交
458 459 460 461 462

		newtp->rx_opt.saw_tstamp = 0;

		newtp->rx_opt.dsack = 0;
		newtp->rx_opt.num_sacks = 0;
463

L
Linus Torvalds 已提交
464 465 466
		newtp->urg_data = 0;

		if (sock_flag(newsk, SOCK_KEEPOPEN))
467 468
			inet_csk_reset_keepalive_timer(newsk,
						       keepalive_time_when(newtp));
L
Linus Torvalds 已提交
469

470
		newtp->rx_opt.tstamp_ok = ireq->tstamp_ok;
S
Stephen Hemminger 已提交
471
		if ((newtp->rx_opt.sack_ok = ireq->sack_ok) != 0) {
L
Linus Torvalds 已提交
472
			if (sysctl_tcp_fack)
473
				tcp_enable_fack(newtp);
L
Linus Torvalds 已提交
474 475 476 477
		}
		newtp->window_clamp = req->window_clamp;
		newtp->rcv_ssthresh = req->rcv_wnd;
		newtp->rcv_wnd = req->rcv_wnd;
478
		newtp->rx_opt.wscale_ok = ireq->wscale_ok;
L
Linus Torvalds 已提交
479
		if (newtp->rx_opt.wscale_ok) {
480 481
			newtp->rx_opt.snd_wscale = ireq->snd_wscale;
			newtp->rx_opt.rcv_wscale = ireq->rcv_wscale;
L
Linus Torvalds 已提交
482 483 484 485
		} else {
			newtp->rx_opt.snd_wscale = newtp->rx_opt.rcv_wscale = 0;
			newtp->window_clamp = min(newtp->window_clamp, 65535U);
		}
486 487
		newtp->snd_wnd = (ntohs(tcp_hdr(skb)->window) <<
				  newtp->rx_opt.snd_wscale);
L
Linus Torvalds 已提交
488 489 490 491
		newtp->max_window = newtp->snd_wnd;

		if (newtp->rx_opt.tstamp_ok) {
			newtp->rx_opt.ts_recent = req->ts_recent;
492
			newtp->rx_opt.ts_recent_stamp = get_seconds();
L
Linus Torvalds 已提交
493 494 495 496 497
			newtp->tcp_header_len = sizeof(struct tcphdr) + TCPOLEN_TSTAMP_ALIGNED;
		} else {
			newtp->rx_opt.ts_recent_stamp = 0;
			newtp->tcp_header_len = sizeof(struct tcphdr);
		}
498 499 500 501 502
#ifdef CONFIG_TCP_MD5SIG
		newtp->md5sig_info = NULL;	/*XXX*/
		if (newtp->af_specific->md5_lookup(sk, newsk))
			newtp->tcp_header_len += TCPOLEN_MD5SIG_ALIGNED;
#endif
503
		if (skb->len >= TCP_MSS_DEFAULT + newtp->tcp_header_len)
504
			newicsk->icsk_ack.last_seg_size = skb->len - newtp->tcp_header_len;
L
Linus Torvalds 已提交
505 506 507
		newtp->rx_opt.mss_clamp = req->mss;
		TCP_ECN_openreq_child(newtp, req);

508
		TCP_INC_STATS_BH(sock_net(sk), TCP_MIB_PASSIVEOPENS);
L
Linus Torvalds 已提交
509 510 511 512
	}
	return newsk;
}

513
/*
L
Linus Torvalds 已提交
514
 *	Process an incoming packet for SYN_RECV sockets represented
515
 *	as a request_sock.
L
Linus Torvalds 已提交
516 517
 */

518
struct sock *tcp_check_req(struct sock *sk, struct sk_buff *skb,
519 520
			   struct request_sock *req,
			   struct request_sock **prev)
L
Linus Torvalds 已提交
521
{
522 523 524
	struct tcp_options_received tmp_opt;
	u8 *hash_location;
	struct sock *child;
525
	const struct tcphdr *th = tcp_hdr(skb);
526
	__be32 flg = tcp_flag_word(th) & (TCP_FLAG_RST|TCP_FLAG_SYN|TCP_FLAG_ACK);
L
Linus Torvalds 已提交
527 528
	int paws_reject = 0;

529
	if ((th->doff > (sizeof(*th) >> 2)) && (req->ts_recent)) {
530
		tmp_opt.tstamp_ok = 1;
531
		tcp_parse_options(skb, &tmp_opt, &hash_location, 1, NULL);
L
Linus Torvalds 已提交
532 533 534 535 536 537 538

		if (tmp_opt.saw_tstamp) {
			tmp_opt.ts_recent = req->ts_recent;
			/* We do not store true stamp, but it is not required,
			 * it can be estimated (approximately)
			 * from another data.
			 */
539
			tmp_opt.ts_recent_stamp = get_seconds() - ((TCP_TIMEOUT_INIT/HZ)<<req->retrans);
I
Ilpo Järvinen 已提交
540
			paws_reject = tcp_paws_reject(&tmp_opt, th->rst);
L
Linus Torvalds 已提交
541 542 543 544
		}
	}

	/* Check for pure retransmitted SYN. */
545
	if (TCP_SKB_CB(skb)->seq == tcp_rsk(req)->rcv_isn &&
L
Linus Torvalds 已提交
546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564
	    flg == TCP_FLAG_SYN &&
	    !paws_reject) {
		/*
		 * RFC793 draws (Incorrectly! It was fixed in RFC1122)
		 * this case on figure 6 and figure 8, but formal
		 * protocol description says NOTHING.
		 * To be more exact, it says that we should send ACK,
		 * because this segment (at least, if it has no data)
		 * is out of window.
		 *
		 *  CONCLUSION: RFC793 (even with RFC1122) DOES NOT
		 *  describe SYN-RECV state. All the description
		 *  is wrong, we cannot believe to it and should
		 *  rely only on common sense and implementation
		 *  experience.
		 *
		 * Enforce "SYN-ACK" according to figure 8, figure 6
		 * of RFC793, fixed by RFC1122.
		 */
565
		req->rsk_ops->rtx_syn_ack(sk, req, NULL);
L
Linus Torvalds 已提交
566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597
		return NULL;
	}

	/* Further reproduces section "SEGMENT ARRIVES"
	   for state SYN-RECEIVED of RFC793.
	   It is broken, however, it does not work only
	   when SYNs are crossed.

	   You would think that SYN crossing is impossible here, since
	   we should have a SYN_SENT socket (from connect()) on our end,
	   but this is not true if the crossed SYNs were sent to both
	   ends by a malicious third party.  We must defend against this,
	   and to do that we first verify the ACK (as per RFC793, page
	   36) and reset if it is invalid.  Is this a true full defense?
	   To convince ourselves, let us consider a way in which the ACK
	   test can still pass in this 'malicious crossed SYNs' case.
	   Malicious sender sends identical SYNs (and thus identical sequence
	   numbers) to both A and B:

		A: gets SYN, seq=7
		B: gets SYN, seq=7

	   By our good fortune, both A and B select the same initial
	   send sequence number of seven :-)

		A: sends SYN|ACK, seq=7, ack_seq=8
		B: sends SYN|ACK, seq=7, ack_seq=8

	   So we are now A eating this SYN|ACK, ACK test passes.  So
	   does sequence test, SYN is truncated, and thus we consider
	   it a bare ACK.

598 599 600 601
	   If icsk->icsk_accept_queue.rskq_defer_accept, we silently drop this
	   bare ACK.  Otherwise, we create an established connection.  Both
	   ends (listening sockets) accept the new incoming connection and try
	   to talk to each other. 8-)
L
Linus Torvalds 已提交
602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617

	   Note: This case is both harmless, and rare.  Possibility is about the
	   same as us discovering intelligent life on another plant tomorrow.

	   But generally, we should (RFC lies!) to accept ACK
	   from SYNACK both here and in tcp_rcv_state_process().
	   tcp_rcv_state_process() does not, hence, we do not too.

	   Note that the case is absolutely generic:
	   we cannot optimize anything here without
	   violating protocol. All the checks must be made
	   before attempt to create socket.
	 */

	/* RFC793 page 36: "If the connection is in any non-synchronized state ...
	 *                  and the incoming segment acknowledges something not yet
S
Stephen Hemminger 已提交
618
	 *                  sent (the segment carries an unacceptable ACK) ...
L
Linus Torvalds 已提交
619 620 621 622 623
	 *                  a reset is sent."
	 *
	 * Invalid ACK: reset will be sent by listening socket
	 */
	if ((flg & TCP_FLAG_ACK) &&
624 625
	    (TCP_SKB_CB(skb)->ack_seq !=
	     tcp_rsk(req)->snt_isn + 1 + tcp_s_data_size(tcp_sk(sk))))
L
Linus Torvalds 已提交
626 627 628 629 630 631 632 633 634 635
		return sk;

	/* Also, it would be not so bad idea to check rcv_tsecr, which
	 * is essentially ACK extension and too early or too late values
	 * should cause reset in unsynchronized states.
	 */

	/* RFC793: "first check sequence number". */

	if (paws_reject || !tcp_in_window(TCP_SKB_CB(skb)->seq, TCP_SKB_CB(skb)->end_seq,
636
					  tcp_rsk(req)->rcv_isn + 1, tcp_rsk(req)->rcv_isn + 1 + req->rcv_wnd)) {
L
Linus Torvalds 已提交
637 638
		/* Out of window: send ACK and drop. */
		if (!(flg & TCP_FLAG_RST))
639
			req->rsk_ops->send_ack(sk, skb, req);
L
Linus Torvalds 已提交
640
		if (paws_reject)
641
			NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_PAWSESTABREJECTED);
L
Linus Torvalds 已提交
642 643 644 645 646
		return NULL;
	}

	/* In sequence, PAWS is OK. */

647
	if (tmp_opt.saw_tstamp && !after(TCP_SKB_CB(skb)->seq, tcp_rsk(req)->rcv_isn + 1))
648
		req->ts_recent = tmp_opt.rcv_tsval;
L
Linus Torvalds 已提交
649

650 651 652 653 654
	if (TCP_SKB_CB(skb)->seq == tcp_rsk(req)->rcv_isn) {
		/* Truncate SYN, it is out of window starting
		   at tcp_rsk(req)->rcv_isn + 1. */
		flg &= ~TCP_FLAG_SYN;
	}
L
Linus Torvalds 已提交
655

656 657 658 659 660 661 662
	/* RFC793: "second check the RST bit" and
	 *	   "fourth, check the SYN bit"
	 */
	if (flg & (TCP_FLAG_RST|TCP_FLAG_SYN)) {
		TCP_INC_STATS_BH(sock_net(sk), TCP_MIB_ATTEMPTFAILS);
		goto embryonic_reset;
	}
L
Linus Torvalds 已提交
663

664 665 666 667 668
	/* ACK sequence verified above, just make sure ACK is
	 * set.  If ACK not set, just silently drop the packet.
	 */
	if (!(flg & TCP_FLAG_ACK))
		return NULL;
669

670 671
	/* While TCP_DEFER_ACCEPT is active, drop bare ACK. */
	if (req->retrans < inet_csk(sk)->icsk_accept_queue.rskq_defer_accept &&
672 673 674 675 676 677 678 679 680 681 682 683 684 685
	    TCP_SKB_CB(skb)->end_seq == tcp_rsk(req)->rcv_isn + 1) {
		inet_rsk(req)->acked = 1;
		return NULL;
	}

	/* OK, ACK is valid, create big socket and
	 * feed this segment to it. It will repeat all
	 * the tests. THIS SEGMENT MUST MOVE SOCKET TO
	 * ESTABLISHED STATE. If it will be dropped after
	 * socket is created, wait for troubles.
	 */
	child = inet_csk(sk)->icsk_af_ops->syn_recv_sock(sk, skb, req, NULL);
	if (child == NULL)
		goto listen_overflow;
L
Linus Torvalds 已提交
686

687 688
	inet_csk_reqsk_queue_unlink(sk, req, prev);
	inet_csk_reqsk_queue_removed(sk, req);
L
Linus Torvalds 已提交
689

690 691
	inet_csk_reqsk_queue_add(sk, req, child);
	return child;
L
Linus Torvalds 已提交
692

693 694 695 696 697
listen_overflow:
	if (!sysctl_tcp_abort_on_overflow) {
		inet_rsk(req)->acked = 1;
		return NULL;
	}
L
Linus Torvalds 已提交
698

699 700 701 702
embryonic_reset:
	NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_EMBRYONICRSTS);
	if (!(flg & TCP_FLAG_RST))
		req->rsk_ops->send_reset(sk, skb);
L
Linus Torvalds 已提交
703

704 705
	inet_csk_reqsk_queue_drop(sk, req, prev);
	return NULL;
L
Linus Torvalds 已提交
706 707 708 709 710 711 712 713 714 715 716 717 718 719 720
}

/*
 * Queue segment on the new socket if the new socket is active,
 * otherwise we just shortcircuit this and continue with
 * the new socket.
 */

int tcp_child_process(struct sock *parent, struct sock *child,
		      struct sk_buff *skb)
{
	int ret = 0;
	int state = child->sk_state;

	if (!sock_owned_by_user(child)) {
721 722
		ret = tcp_rcv_state_process(child, skb, tcp_hdr(skb),
					    skb->len);
L
Linus Torvalds 已提交
723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742
		/* Wakeup parent, send SIGIO */
		if (state == TCP_SYN_RECV && child->sk_state != state)
			parent->sk_data_ready(parent, 0);
	} else {
		/* Alas, it is possible again, because we do lookup
		 * in main socket hash table and lock on listening
		 * socket does not protect us more.
		 */
		sk_add_backlog(child, skb);
	}

	bh_unlock_sock(child);
	sock_put(child);
	return ret;
}

EXPORT_SYMBOL(tcp_check_req);
EXPORT_SYMBOL(tcp_child_process);
EXPORT_SYMBOL(tcp_create_openreq_child);
EXPORT_SYMBOL(tcp_timewait_state_process);