tcp_minisocks.c 22.6 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.
 *
 *		Implementation of the Transmission Control Protocol(TCP).
 *
 * Version:	$Id: tcp_minisocks.c,v 1.15 2002/02/01 22:01:04 davem Exp $
 *
10
 * Authors:	Ross Biro
L
Linus Torvalds 已提交
11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36
 *		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>

#ifdef CONFIG_SYSCTL
#define SYNC_INIT 0 /* let the user enable it */
#else
#define SYNC_INIT 1
#endif

37 38
int sysctl_tcp_syncookies __read_mostly = SYNC_INIT;
int sysctl_tcp_abort_on_overflow __read_mostly;
L
Linus Torvalds 已提交
39

40 41 42
struct inet_timewait_death_row tcp_death_row = {
	.sysctl_max_tw_buckets = NR_FILE * 2,
	.period		= TCP_TIMEWAIT_LEN / INET_TWDR_TWKILL_SLOTS,
43
	.death_lock	= __SPIN_LOCK_UNLOCKED(tcp_death_row.death_lock),
44 45 46 47
	.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,
48
					     inet_twdr_twkill_work),
49 50 51 52 53 54 55 56 57
/* 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
58 59 60 61 62 63 64 65 66
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);
}

67
/*
L
Linus Torvalds 已提交
68 69 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
 * * 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
96 97
tcp_timewait_state_process(struct inet_timewait_sock *tw, struct sk_buff *skb,
			   const struct tcphdr *th)
L
Linus Torvalds 已提交
98
{
99
	struct tcp_timewait_sock *tcptw = tcp_twsk((struct sock *)tw);
L
Linus Torvalds 已提交
100 101 102 103
	struct tcp_options_received tmp_opt;
	int paws_reject = 0;

	tmp_opt.saw_tstamp = 0;
104
	if (th->doff > (sizeof(*th) >> 2) && tcptw->tw_ts_recent_stamp) {
L
Linus Torvalds 已提交
105 106 107
		tcp_parse_options(skb, &tmp_opt, 0);

		if (tmp_opt.saw_tstamp) {
108 109
			tmp_opt.ts_recent	= tcptw->tw_ts_recent;
			tmp_opt.ts_recent_stamp	= tcptw->tw_ts_recent_stamp;
L
Linus Torvalds 已提交
110 111 112 113 114 115 116 117 118 119
			paws_reject = tcp_paws_check(&tmp_opt, th->rst);
		}
	}

	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,
120 121
				   tcptw->tw_rcv_nxt,
				   tcptw->tw_rcv_nxt + tcptw->tw_rcv_wnd))
L
Linus Torvalds 已提交
122 123 124 125 126
			return TCP_TW_ACK;

		if (th->rst)
			goto kill;

127
		if (th->syn && !before(TCP_SKB_CB(skb)->seq, tcptw->tw_rcv_nxt))
L
Linus Torvalds 已提交
128 129 130
			goto kill_with_rst;

		/* Dup ACK? */
131
		if (!after(TCP_SKB_CB(skb)->end_seq, tcptw->tw_rcv_nxt) ||
L
Linus Torvalds 已提交
132
		    TCP_SKB_CB(skb)->end_seq == TCP_SKB_CB(skb)->seq) {
133
			inet_twsk_put(tw);
L
Linus Torvalds 已提交
134 135 136 137 138 139 140
			return TCP_TW_SUCCESS;
		}

		/* New data or FIN. If new data arrive after half-duplex close,
		 * reset.
		 */
		if (!th->fin ||
141
		    TCP_SKB_CB(skb)->end_seq != tcptw->tw_rcv_nxt + 1) {
L
Linus Torvalds 已提交
142
kill_with_rst:
143
			inet_twsk_deschedule(tw, &tcp_death_row);
144
			inet_twsk_put(tw);
L
Linus Torvalds 已提交
145 146 147 148
			return TCP_TW_RST;
		}

		/* FIN arrived, enter true time-wait state. */
149 150
		tw->tw_substate	  = TCP_TIME_WAIT;
		tcptw->tw_rcv_nxt = TCP_SKB_CB(skb)->end_seq;
L
Linus Torvalds 已提交
151
		if (tmp_opt.saw_tstamp) {
152
			tcptw->tw_ts_recent_stamp = get_seconds();
153
			tcptw->tw_ts_recent	  = tmp_opt.rcv_tsval;
L
Linus Torvalds 已提交
154 155 156 157 158
		}

		/* 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
159
		 * do not understand recycling in any case, it not
L
Linus Torvalds 已提交
160 161
		 * a big problem in practice. --ANK */
		if (tw->tw_family == AF_INET &&
162
		    tcp_death_row.sysctl_tw_recycle && tcptw->tw_ts_recent_stamp &&
L
Linus Torvalds 已提交
163
		    tcp_v4_tw_remember_stamp(tw))
164 165
			inet_twsk_schedule(tw, &tcp_death_row, tw->tw_timeout,
					   TCP_TIMEWAIT_LEN);
L
Linus Torvalds 已提交
166
		else
167 168
			inet_twsk_schedule(tw, &tcp_death_row, TCP_TIMEWAIT_LEN,
					   TCP_TIMEWAIT_LEN);
L
Linus Torvalds 已提交
169 170 171 172 173 174 175 176 177 178
		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:
179
	 *
L
Linus Torvalds 已提交
180 181 182 183 184
	 *	(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
	 *
185
	 *	(2)  returns to TIME-WAIT state if the SYN turns out
L
Linus Torvalds 已提交
186 187 188 189
	 *	to be an old duplicate".
	 */

	if (!paws_reject &&
190
	    (TCP_SKB_CB(skb)->seq == tcptw->tw_rcv_nxt &&
L
Linus Torvalds 已提交
191 192 193 194
	     (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
195
			/* This is TIME_WAIT assassination, in two flavors.
L
Linus Torvalds 已提交
196 197 198 199 200
			 * Oh well... nobody has a sufficient solution to this
			 * protocol bug yet.
			 */
			if (sysctl_tcp_rfc1337 == 0) {
kill:
201
				inet_twsk_deschedule(tw, &tcp_death_row);
202
				inet_twsk_put(tw);
L
Linus Torvalds 已提交
203 204 205
				return TCP_TW_SUCCESS;
			}
		}
206 207
		inet_twsk_schedule(tw, &tcp_death_row, TCP_TIMEWAIT_LEN,
				   TCP_TIMEWAIT_LEN);
L
Linus Torvalds 已提交
208 209

		if (tmp_opt.saw_tstamp) {
210
			tcptw->tw_ts_recent	  = tmp_opt.rcv_tsval;
211
			tcptw->tw_ts_recent_stamp = get_seconds();
L
Linus Torvalds 已提交
212 213
		}

214
		inet_twsk_put(tw);
L
Linus Torvalds 已提交
215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235
		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 &&
236 237 238 239
	    (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
240 241 242 243 244 245 246 247 248
		if (isn == 0)
			isn++;
		TCP_SKB_CB(skb)->when = isn;
		return TCP_TW_SYN;
	}

	if (paws_reject)
		NET_INC_STATS_BH(LINUX_MIB_PAWSESTABREJECTED);

S
Stephen Hemminger 已提交
249
	if (!th->rst) {
L
Linus Torvalds 已提交
250 251 252 253 254 255 256
		/* 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)
257 258
			inet_twsk_schedule(tw, &tcp_death_row, TCP_TIMEWAIT_LEN,
					   TCP_TIMEWAIT_LEN);
L
Linus Torvalds 已提交
259 260 261 262 263 264

		/* Send ACK. Note, we do not put the bucket,
		 * it will be released by caller.
		 */
		return TCP_TW_ACK;
	}
265
	inet_twsk_put(tw);
L
Linus Torvalds 已提交
266 267 268
	return TCP_TW_SUCCESS;
}

269
/*
L
Linus Torvalds 已提交
270
 * Move a socket to time-wait or dead fin-wait-2 state.
271
 */
L
Linus Torvalds 已提交
272 273
void tcp_time_wait(struct sock *sk, int state, int timeo)
{
274
	struct inet_timewait_sock *tw = NULL;
275
	const struct inet_connection_sock *icsk = inet_csk(sk);
276
	const struct tcp_sock *tp = tcp_sk(sk);
L
Linus Torvalds 已提交
277 278
	int recycle_ok = 0;

279
	if (tcp_death_row.sysctl_tw_recycle && tp->rx_opt.ts_recent_stamp)
280
		recycle_ok = icsk->icsk_af_ops->remember_stamp(sk);
L
Linus Torvalds 已提交
281

282
	if (tcp_death_row.tw_count < tcp_death_row.sysctl_max_tw_buckets)
283
		tw = inet_twsk_alloc(sk, state);
L
Linus Torvalds 已提交
284

285 286
	if (tw != NULL) {
		struct tcp_timewait_sock *tcptw = tcp_twsk((struct sock *)tw);
287
		const int rto = (icsk->icsk_rto << 2) - (icsk->icsk_rto >> 1);
288

L
Linus Torvalds 已提交
289
		tw->tw_rcv_wscale	= tp->rx_opt.rcv_wscale;
290 291 292 293 294
		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
295 296 297 298

#if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE)
		if (tw->tw_family == PF_INET6) {
			struct ipv6_pinfo *np = inet6_sk(sk);
299
			struct inet6_timewait_sock *tw6;
L
Linus Torvalds 已提交
300

301 302 303 304
			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);
305
			tw->tw_ipv6only = np->ipv6only;
306
		}
L
Linus Torvalds 已提交
307
#endif
308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326

#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;
				if (tcp_alloc_md5sig_pool() == NULL)
					BUG();
			}
S
Stephen Hemminger 已提交
327
		} while (0);
328 329
#endif

L
Linus Torvalds 已提交
330
		/* Linkage updates. */
331
		__inet_twsk_hashdance(tw, sk, &tcp_hashinfo);
L
Linus Torvalds 已提交
332 333 334 335 336 337 338 339 340 341 342 343 344

		/* 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;
		}

345 346
		inet_twsk_schedule(tw, &tcp_death_row, timeo,
				   TCP_TIMEWAIT_LEN);
347
		inet_twsk_put(tw);
L
Linus Torvalds 已提交
348 349 350 351 352
	} else {
		/* Sorry, if we're out of memory, just CLOSE this
		 * socket up.  We've got bigger problems than
		 * non-graceful socket closings.
		 */
353
		LIMIT_NETDEBUG(KERN_INFO "TCP: time wait bucket table overflow\n");
L
Linus Torvalds 已提交
354 355 356 357 358 359
	}

	tcp_update_metrics(sk);
	tcp_done(sk);
}

360 361 362
void tcp_twsk_destructor(struct sock *sk)
{
#ifdef CONFIG_TCP_MD5SIG
363
	struct tcp_timewait_sock *twsk = tcp_twsk(sk);
364 365 366 367 368 369 370
	if (twsk->tw_md5_keylen)
		tcp_put_md5sig_pool();
#endif
}

EXPORT_SYMBOL_GPL(tcp_twsk_destructor);

L
Linus Torvalds 已提交
371 372 373 374 375 376
/* 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.
 */
377
struct sock *tcp_create_openreq_child(struct sock *sk, struct request_sock *req, struct sk_buff *skb)
L
Linus Torvalds 已提交
378
{
379
	struct sock *newsk = inet_csk_clone(sk, req, GFP_ATOMIC);
L
Linus Torvalds 已提交
380

381
	if (newsk != NULL) {
382
		const struct inet_request_sock *ireq = inet_rsk(req);
383
		struct tcp_request_sock *treq = tcp_rsk(req);
384
		struct inet_connection_sock *newicsk = inet_csk(newsk);
L
Linus Torvalds 已提交
385 386 387 388 389
		struct tcp_sock *newtp;

		/* Now setup tcp_sock */
		newtp = tcp_sk(newsk);
		newtp->pred_flags = 0;
390 391
		newtp->rcv_wup = newtp->copied_seq = newtp->rcv_nxt = treq->rcv_isn + 1;
		newtp->snd_sml = newtp->snd_una = newtp->snd_nxt = treq->snt_isn + 1;
L
Linus Torvalds 已提交
392 393 394

		tcp_prequeue_init(newtp);

395
		tcp_init_wl(newtp, treq->snt_isn, treq->rcv_isn);
L
Linus Torvalds 已提交
396 397 398

		newtp->srtt = 0;
		newtp->mdev = TCP_TIMEOUT_INIT;
399
		newicsk->icsk_rto = TCP_TIMEOUT_INIT;
L
Linus Torvalds 已提交
400 401 402 403 404 405 406 407 408 409 410 411 412 413 414

		newtp->packets_out = 0;
		newtp->left_out = 0;
		newtp->retrans_out = 0;
		newtp->sacked_out = 0;
		newtp->fackets_out = 0;
		newtp->snd_ssthresh = 0x7fffffff;

		/* 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;
415
		newtp->bytes_acked = 0;
L
Linus Torvalds 已提交
416 417 418 419

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

420
		newicsk->icsk_ca_ops = &tcp_init_congestion_ops;
421

422
		tcp_set_ca_state(newsk, TCP_CA_Open);
L
Linus Torvalds 已提交
423 424
		tcp_init_xmit_timers(newsk);
		skb_queue_head_init(&newtp->out_of_order_queue);
425
		newtp->write_seq = treq->snt_isn + 1;
L
Linus Torvalds 已提交
426 427 428 429 430 431 432 433 434 435 436
		newtp->pushed_seq = newtp->write_seq;

		newtp->rx_opt.saw_tstamp = 0;

		newtp->rx_opt.dsack = 0;
		newtp->rx_opt.eff_sacks = 0;

		newtp->rx_opt.num_sacks = 0;
		newtp->urg_data = 0;

		if (sock_flag(newsk, SOCK_KEEPOPEN))
437 438
			inet_csk_reset_keepalive_timer(newsk,
						       keepalive_time_when(newtp));
L
Linus Torvalds 已提交
439

440
		newtp->rx_opt.tstamp_ok = ireq->tstamp_ok;
S
Stephen Hemminger 已提交
441
		if ((newtp->rx_opt.sack_ok = ireq->sack_ok) != 0) {
L
Linus Torvalds 已提交
442 443 444 445 446 447
			if (sysctl_tcp_fack)
				newtp->rx_opt.sack_ok |= 2;
		}
		newtp->window_clamp = req->window_clamp;
		newtp->rcv_ssthresh = req->rcv_wnd;
		newtp->rcv_wnd = req->rcv_wnd;
448
		newtp->rx_opt.wscale_ok = ireq->wscale_ok;
L
Linus Torvalds 已提交
449
		if (newtp->rx_opt.wscale_ok) {
450 451
			newtp->rx_opt.snd_wscale = ireq->snd_wscale;
			newtp->rx_opt.rcv_wscale = ireq->rcv_wscale;
L
Linus Torvalds 已提交
452 453 454 455 456 457 458 459 460
		} else {
			newtp->rx_opt.snd_wscale = newtp->rx_opt.rcv_wscale = 0;
			newtp->window_clamp = min(newtp->window_clamp, 65535U);
		}
		newtp->snd_wnd = ntohs(skb->h.th->window) << newtp->rx_opt.snd_wscale;
		newtp->max_window = newtp->snd_wnd;

		if (newtp->rx_opt.tstamp_ok) {
			newtp->rx_opt.ts_recent = req->ts_recent;
461
			newtp->rx_opt.ts_recent_stamp = get_seconds();
L
Linus Torvalds 已提交
462 463 464 465 466
			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);
		}
467 468 469 470 471
#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
L
Linus Torvalds 已提交
472
		if (skb->len >= TCP_MIN_RCVMSS+newtp->tcp_header_len)
473
			newicsk->icsk_ack.last_seg_size = skb->len - newtp->tcp_header_len;
L
Linus Torvalds 已提交
474 475 476 477 478 479 480 481
		newtp->rx_opt.mss_clamp = req->mss;
		TCP_ECN_openreq_child(newtp, req);

		TCP_INC_STATS_BH(TCP_MIB_PASSIVEOPENS);
	}
	return newsk;
}

482
/*
L
Linus Torvalds 已提交
483
 *	Process an incoming packet for SYN_RECV sockets represented
484
 *	as a request_sock.
L
Linus Torvalds 已提交
485 486 487
 */

struct sock *tcp_check_req(struct sock *sk,struct sk_buff *skb,
488 489
			   struct request_sock *req,
			   struct request_sock **prev)
L
Linus Torvalds 已提交
490 491
{
	struct tcphdr *th = skb->h.th;
492
	__be32 flg = tcp_flag_word(th) & (TCP_FLAG_RST|TCP_FLAG_SYN|TCP_FLAG_ACK);
L
Linus Torvalds 已提交
493 494 495 496 497 498 499 500 501 502 503 504 505 506
	int paws_reject = 0;
	struct tcp_options_received tmp_opt;
	struct sock *child;

	tmp_opt.saw_tstamp = 0;
	if (th->doff > (sizeof(struct tcphdr)>>2)) {
		tcp_parse_options(skb, &tmp_opt, 0);

		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.
			 */
507
			tmp_opt.ts_recent_stamp = get_seconds() - ((TCP_TIMEOUT_INIT/HZ)<<req->retrans);
L
Linus Torvalds 已提交
508 509 510 511 512
			paws_reject = tcp_paws_check(&tmp_opt, th->rst);
		}
	}

	/* Check for pure retransmitted SYN. */
513
	if (TCP_SKB_CB(skb)->seq == tcp_rsk(req)->rcv_isn &&
L
Linus Torvalds 已提交
514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532
	    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.
		 */
533
		req->rsk_ops->rtx_syn_ack(sk, req, NULL);
L
Linus Torvalds 已提交
534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565
		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.

566 567 568 569
	   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
570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585

	   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
586
	 *                  sent (the segment carries an unacceptable ACK) ...
L
Linus Torvalds 已提交
587 588 589 590 591
	 *                  a reset is sent."
	 *
	 * Invalid ACK: reset will be sent by listening socket
	 */
	if ((flg & TCP_FLAG_ACK) &&
592
	    (TCP_SKB_CB(skb)->ack_seq != tcp_rsk(req)->snt_isn + 1))
L
Linus Torvalds 已提交
593 594 595 596 597 598 599 600 601 602
		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,
603
					  tcp_rsk(req)->rcv_isn + 1, tcp_rsk(req)->rcv_isn + 1 + req->rcv_wnd)) {
L
Linus Torvalds 已提交
604 605
		/* Out of window: send ACK and drop. */
		if (!(flg & TCP_FLAG_RST))
606
			req->rsk_ops->send_ack(skb, req);
L
Linus Torvalds 已提交
607 608 609 610 611 612 613
		if (paws_reject)
			NET_INC_STATS_BH(LINUX_MIB_PAWSESTABREJECTED);
		return NULL;
	}

	/* In sequence, PAWS is OK. */

614
	if (tmp_opt.saw_tstamp && !after(TCP_SKB_CB(skb)->seq, tcp_rsk(req)->rcv_isn + 1))
L
Linus Torvalds 已提交
615 616
			req->ts_recent = tmp_opt.rcv_tsval;

617
		if (TCP_SKB_CB(skb)->seq == tcp_rsk(req)->rcv_isn) {
L
Linus Torvalds 已提交
618
			/* Truncate SYN, it is out of window starting
619
			   at tcp_rsk(req)->rcv_isn + 1. */
L
Linus Torvalds 已提交
620 621 622 623 624 625
			flg &= ~TCP_FLAG_SYN;
		}

		/* RFC793: "second check the RST bit" and
		 *	   "fourth, check the SYN bit"
		 */
626 627
		if (flg & (TCP_FLAG_RST|TCP_FLAG_SYN)) {
			TCP_INC_STATS_BH(TCP_MIB_ATTEMPTFAILS);
L
Linus Torvalds 已提交
628
			goto embryonic_reset;
629
		}
L
Linus Torvalds 已提交
630 631 632 633 634 635 636 637

		/* 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;

		/* If TCP_DEFER_ACCEPT is set, drop bare ACK. */
638 639
		if (inet_csk(sk)->icsk_accept_queue.rskq_defer_accept &&
		    TCP_SKB_CB(skb)->end_seq == tcp_rsk(req)->rcv_isn + 1) {
640
			inet_rsk(req)->acked = 1;
L
Linus Torvalds 已提交
641 642 643 644 645 646 647 648 649
			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.
		 */
650 651
		child = inet_csk(sk)->icsk_af_ops->syn_recv_sock(sk, skb,
								 req, NULL);
L
Linus Torvalds 已提交
652 653
		if (child == NULL)
			goto listen_overflow;
654 655 656 657 658 659 660 661 662 663 664 665
#ifdef CONFIG_TCP_MD5SIG
		else {
			/* Copy over the MD5 key from the original socket */
			struct tcp_md5sig_key *key;
			struct tcp_sock *tp = tcp_sk(sk);
			key = tp->af_specific->md5_lookup(sk, child);
			if (key != NULL) {
				/*
				 * We're using one, so create a matching key on the
				 * newsk structure. If we fail to get memory then we
				 * end up not copying the key across. Shucks.
				 */
666 667
				char *newkey = kmemdup(key->key, key->keylen,
						       GFP_ATOMIC);
668 669 670 671 672 673 674 675 676 677
				if (newkey) {
					if (!tcp_alloc_md5sig_pool())
						BUG();
					tp->af_specific->md5_add(child, child,
								 newkey,
								 key->keylen);
				}
			}
		}
#endif
L
Linus Torvalds 已提交
678

679 680
		inet_csk_reqsk_queue_unlink(sk, req, prev);
		inet_csk_reqsk_queue_removed(sk, req);
L
Linus Torvalds 已提交
681

682
		inet_csk_reqsk_queue_add(sk, req, child);
L
Linus Torvalds 已提交
683 684 685 686
		return child;

	listen_overflow:
		if (!sysctl_tcp_abort_on_overflow) {
687
			inet_rsk(req)->acked = 1;
L
Linus Torvalds 已提交
688 689 690 691 692 693
			return NULL;
		}

	embryonic_reset:
		NET_INC_STATS_BH(LINUX_MIB_EMBRYONICRSTS);
		if (!(flg & TCP_FLAG_RST))
694
			req->rsk_ops->send_reset(sk, skb);
L
Linus Torvalds 已提交
695

696
		inet_csk_reqsk_queue_drop(sk, req, prev);
L
Linus Torvalds 已提交
697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734
		return NULL;
}

/*
 * 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)) {
		ret = tcp_rcv_state_process(child, skb, skb->h.th, skb->len);

		/* 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);