tcp_ipv6.c 40.7 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8 9 10 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 37 38 39 40 41 42 43 44 45 46 47 48 49
/*
 *	TCP over IPv6
 *	Linux INET6 implementation 
 *
 *	Authors:
 *	Pedro Roque		<roque@di.fc.ul.pt>	
 *
 *	$Id: tcp_ipv6.c,v 1.144 2002/02/01 22:01:04 davem Exp $
 *
 *	Based on: 
 *	linux/net/ipv4/tcp.c
 *	linux/net/ipv4/tcp_input.c
 *	linux/net/ipv4/tcp_output.c
 *
 *	Fixes:
 *	Hideaki YOSHIFUJI	:	sin6_scope_id support
 *	YOSHIFUJI Hideaki @USAGI and:	Support IPV6_V6ONLY socket option, which
 *	Alexey Kuznetsov		allow both IPv4 and IPv6 sockets to bind
 *					a single port at the same time.
 *	YOSHIFUJI Hideaki @USAGI:	convert /proc/net/tcp6 to seq_file.
 *
 *	This program is free software; you can redistribute it and/or
 *      modify it under the terms of the GNU General Public License
 *      as published by the Free Software Foundation; either version
 *      2 of the License, or (at your option) any later version.
 */

#include <linux/module.h>
#include <linux/config.h>
#include <linux/errno.h>
#include <linux/types.h>
#include <linux/socket.h>
#include <linux/sockios.h>
#include <linux/net.h>
#include <linux/jiffies.h>
#include <linux/in.h>
#include <linux/in6.h>
#include <linux/netdevice.h>
#include <linux/init.h>
#include <linux/jhash.h>
#include <linux/ipsec.h>
#include <linux/times.h>

#include <linux/ipv6.h>
#include <linux/icmpv6.h>
#include <linux/random.h>

#include <net/tcp.h>
#include <net/ndisc.h>
50
#include <net/inet6_hashtables.h>
51
#include <net/inet6_connection_sock.h>
L
Linus Torvalds 已提交
52 53 54 55 56 57 58 59 60 61 62
#include <net/ipv6.h>
#include <net/transp_v6.h>
#include <net/addrconf.h>
#include <net/ip6_route.h>
#include <net/ip6_checksum.h>
#include <net/inet_ecn.h>
#include <net/protocol.h>
#include <net/xfrm.h>
#include <net/addrconf.h>
#include <net/snmp.h>
#include <net/dsfield.h>
63
#include <net/timewait_sock.h>
L
Linus Torvalds 已提交
64 65 66 67 68 69

#include <asm/uaccess.h>

#include <linux/proc_fs.h>
#include <linux/seq_file.h>

70 71 72
/* Socket used for sending RSTs and ACKs */
static struct socket *tcp6_socket;

L
Linus Torvalds 已提交
73
static void	tcp_v6_send_reset(struct sk_buff *skb);
74
static void	tcp_v6_reqsk_send_ack(struct sk_buff *skb, struct request_sock *req);
75
static void	tcp_v6_send_check(struct sock *sk, int len, 
L
Linus Torvalds 已提交
76 77 78 79
				  struct sk_buff *skb);

static int	tcp_v6_do_rcv(struct sock *sk, struct sk_buff *skb);

80 81
static struct inet_connection_sock_af_ops ipv6_mapped;
static struct inet_connection_sock_af_ops ipv6_specific;
L
Linus Torvalds 已提交
82 83 84

static int tcp_v6_get_port(struct sock *sk, unsigned short snum)
{
85 86
	return inet_csk_get_port(&tcp_hashinfo, sk, snum,
				 inet6_csk_bind_conflict);
L
Linus Torvalds 已提交
87 88 89 90 91
}

static void tcp_v6_hash(struct sock *sk)
{
	if (sk->sk_state != TCP_CLOSE) {
92
		if (inet_csk(sk)->icsk_af_ops == &ipv6_mapped) {
L
Linus Torvalds 已提交
93 94 95 96
			tcp_prot.hash(sk);
			return;
		}
		local_bh_disable();
97
		__inet6_hash(&tcp_hashinfo, sk);
L
Linus Torvalds 已提交
98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128
		local_bh_enable();
	}
}

static __inline__ u16 tcp_v6_check(struct tcphdr *th, int len,
				   struct in6_addr *saddr, 
				   struct in6_addr *daddr, 
				   unsigned long base)
{
	return csum_ipv6_magic(saddr, daddr, len, IPPROTO_TCP, base);
}

static __u32 tcp_v6_init_sequence(struct sock *sk, struct sk_buff *skb)
{
	if (skb->protocol == htons(ETH_P_IPV6)) {
		return secure_tcpv6_sequence_number(skb->nh.ipv6h->daddr.s6_addr32,
						    skb->nh.ipv6h->saddr.s6_addr32,
						    skb->h.th->dest,
						    skb->h.th->source);
	} else {
		return secure_tcp_sequence_number(skb->nh.iph->daddr,
						  skb->nh.iph->saddr,
						  skb->h.th->dest,
						  skb->h.th->source);
	}
}

static int tcp_v6_connect(struct sock *sk, struct sockaddr *uaddr, 
			  int addr_len)
{
	struct sockaddr_in6 *usin = (struct sockaddr_in6 *) uaddr;
129 130
 	struct inet_sock *inet = inet_sk(sk);
	struct inet_connection_sock *icsk = inet_csk(sk);
L
Linus Torvalds 已提交
131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204
	struct ipv6_pinfo *np = inet6_sk(sk);
	struct tcp_sock *tp = tcp_sk(sk);
	struct in6_addr *saddr = NULL, *final_p = NULL, final;
	struct flowi fl;
	struct dst_entry *dst;
	int addr_type;
	int err;

	if (addr_len < SIN6_LEN_RFC2133) 
		return -EINVAL;

	if (usin->sin6_family != AF_INET6) 
		return(-EAFNOSUPPORT);

	memset(&fl, 0, sizeof(fl));

	if (np->sndflow) {
		fl.fl6_flowlabel = usin->sin6_flowinfo&IPV6_FLOWINFO_MASK;
		IP6_ECN_flow_init(fl.fl6_flowlabel);
		if (fl.fl6_flowlabel&IPV6_FLOWLABEL_MASK) {
			struct ip6_flowlabel *flowlabel;
			flowlabel = fl6_sock_lookup(sk, fl.fl6_flowlabel);
			if (flowlabel == NULL)
				return -EINVAL;
			ipv6_addr_copy(&usin->sin6_addr, &flowlabel->dst);
			fl6_sock_release(flowlabel);
		}
	}

	/*
  	 *	connect() to INADDR_ANY means loopback (BSD'ism).
  	 */
  	
  	if(ipv6_addr_any(&usin->sin6_addr))
		usin->sin6_addr.s6_addr[15] = 0x1; 

	addr_type = ipv6_addr_type(&usin->sin6_addr);

	if(addr_type & IPV6_ADDR_MULTICAST)
		return -ENETUNREACH;

	if (addr_type&IPV6_ADDR_LINKLOCAL) {
		if (addr_len >= sizeof(struct sockaddr_in6) &&
		    usin->sin6_scope_id) {
			/* If interface is set while binding, indices
			 * must coincide.
			 */
			if (sk->sk_bound_dev_if &&
			    sk->sk_bound_dev_if != usin->sin6_scope_id)
				return -EINVAL;

			sk->sk_bound_dev_if = usin->sin6_scope_id;
		}

		/* Connect to link-local address requires an interface */
		if (!sk->sk_bound_dev_if)
			return -EINVAL;
	}

	if (tp->rx_opt.ts_recent_stamp &&
	    !ipv6_addr_equal(&np->daddr, &usin->sin6_addr)) {
		tp->rx_opt.ts_recent = 0;
		tp->rx_opt.ts_recent_stamp = 0;
		tp->write_seq = 0;
	}

	ipv6_addr_copy(&np->daddr, &usin->sin6_addr);
	np->flow_label = fl.fl6_flowlabel;

	/*
	 *	TCP over IPv4
	 */

	if (addr_type == IPV6_ADDR_MAPPED) {
205
		u32 exthdrlen = icsk->icsk_ext_hdr_len;
L
Linus Torvalds 已提交
206 207 208 209 210 211 212 213 214 215 216
		struct sockaddr_in sin;

		SOCK_DEBUG(sk, "connect: ipv4 mapped\n");

		if (__ipv6_only_sock(sk))
			return -ENETUNREACH;

		sin.sin_family = AF_INET;
		sin.sin_port = usin->sin6_port;
		sin.sin_addr.s_addr = usin->sin6_addr.s6_addr32[3];

217
		icsk->icsk_af_ops = &ipv6_mapped;
L
Linus Torvalds 已提交
218 219 220 221 222
		sk->sk_backlog_rcv = tcp_v4_do_rcv;

		err = tcp_v4_connect(sk, (struct sockaddr *)&sin, sizeof(sin));

		if (err) {
223 224
			icsk->icsk_ext_hdr_len = exthdrlen;
			icsk->icsk_af_ops = &ipv6_specific;
L
Linus Torvalds 已提交
225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260
			sk->sk_backlog_rcv = tcp_v6_do_rcv;
			goto failure;
		} else {
			ipv6_addr_set(&np->saddr, 0, 0, htonl(0x0000FFFF),
				      inet->saddr);
			ipv6_addr_set(&np->rcv_saddr, 0, 0, htonl(0x0000FFFF),
				      inet->rcv_saddr);
		}

		return err;
	}

	if (!ipv6_addr_any(&np->rcv_saddr))
		saddr = &np->rcv_saddr;

	fl.proto = IPPROTO_TCP;
	ipv6_addr_copy(&fl.fl6_dst, &np->daddr);
	ipv6_addr_copy(&fl.fl6_src,
		       (saddr ? saddr : &np->saddr));
	fl.oif = sk->sk_bound_dev_if;
	fl.fl_ip_dport = usin->sin6_port;
	fl.fl_ip_sport = inet->sport;

	if (np->opt && np->opt->srcrt) {
		struct rt0_hdr *rt0 = (struct rt0_hdr *)np->opt->srcrt;
		ipv6_addr_copy(&final, &fl.fl6_dst);
		ipv6_addr_copy(&fl.fl6_dst, rt0->addr);
		final_p = &final;
	}

	err = ip6_dst_lookup(sk, &dst, &fl);
	if (err)
		goto failure;
	if (final_p)
		ipv6_addr_copy(&fl.fl6_dst, final_p);

261
	if ((err = xfrm_lookup(&dst, &fl, sk, 0)) < 0)
L
Linus Torvalds 已提交
262 263 264 265 266 267 268 269 270 271 272 273 274 275 276
		goto failure;

	if (saddr == NULL) {
		saddr = &fl.fl6_src;
		ipv6_addr_copy(&np->rcv_saddr, saddr);
	}

	/* set the source address */
	ipv6_addr_copy(&np->saddr, saddr);
	inet->rcv_saddr = LOOPBACK4_IPV6;

	ip6_dst_store(sk, dst, NULL);
	sk->sk_route_caps = dst->dev->features &
		~(NETIF_F_IP_CSUM | NETIF_F_TSO);

277
	icsk->icsk_ext_hdr_len = 0;
L
Linus Torvalds 已提交
278
	if (np->opt)
279 280
		icsk->icsk_ext_hdr_len = (np->opt->opt_flen +
					  np->opt->opt_nflen);
L
Linus Torvalds 已提交
281 282 283 284 285 286

	tp->rx_opt.mss_clamp = IPV6_MIN_MTU - sizeof(struct tcphdr) - sizeof(struct ipv6hdr);

	inet->dport = usin->sin6_port;

	tcp_set_state(sk, TCP_SYN_SENT);
287
	err = inet6_hash_connect(&tcp_death_row, sk);
L
Linus Torvalds 已提交
288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315
	if (err)
		goto late_failure;

	if (!tp->write_seq)
		tp->write_seq = secure_tcpv6_sequence_number(np->saddr.s6_addr32,
							     np->daddr.s6_addr32,
							     inet->sport,
							     inet->dport);

	err = tcp_connect(sk);
	if (err)
		goto late_failure;

	return 0;

late_failure:
	tcp_set_state(sk, TCP_CLOSE);
	__sk_dst_reset(sk);
failure:
	inet->dport = 0;
	sk->sk_route_caps = 0;
	return err;
}

static void tcp_v6_err(struct sk_buff *skb, struct inet6_skb_parm *opt,
		int type, int code, int offset, __u32 info)
{
	struct ipv6hdr *hdr = (struct ipv6hdr*)skb->data;
316
	const struct tcphdr *th = (struct tcphdr *)(skb->data+offset);
L
Linus Torvalds 已提交
317 318 319 320 321 322
	struct ipv6_pinfo *np;
	struct sock *sk;
	int err;
	struct tcp_sock *tp; 
	__u32 seq;

323 324
	sk = inet6_lookup(&tcp_hashinfo, &hdr->daddr, th->dest, &hdr->saddr,
			  th->source, skb->dev->ifindex);
L
Linus Torvalds 已提交
325 326 327 328 329 330 331

	if (sk == NULL) {
		ICMP6_INC_STATS_BH(__in6_dev_get(skb->dev), ICMP6_MIB_INERRORS);
		return;
	}

	if (sk->sk_state == TCP_TIME_WAIT) {
332
		inet_twsk_put((struct inet_timewait_sock *)sk);
L
Linus Torvalds 已提交
333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392
		return;
	}

	bh_lock_sock(sk);
	if (sock_owned_by_user(sk))
		NET_INC_STATS_BH(LINUX_MIB_LOCKDROPPEDICMPS);

	if (sk->sk_state == TCP_CLOSE)
		goto out;

	tp = tcp_sk(sk);
	seq = ntohl(th->seq); 
	if (sk->sk_state != TCP_LISTEN &&
	    !between(seq, tp->snd_una, tp->snd_nxt)) {
		NET_INC_STATS_BH(LINUX_MIB_OUTOFWINDOWICMPS);
		goto out;
	}

	np = inet6_sk(sk);

	if (type == ICMPV6_PKT_TOOBIG) {
		struct dst_entry *dst = NULL;

		if (sock_owned_by_user(sk))
			goto out;
		if ((1 << sk->sk_state) & (TCPF_LISTEN | TCPF_CLOSE))
			goto out;

		/* icmp should have updated the destination cache entry */
		dst = __sk_dst_check(sk, np->dst_cookie);

		if (dst == NULL) {
			struct inet_sock *inet = inet_sk(sk);
			struct flowi fl;

			/* BUGGG_FUTURE: Again, it is not clear how
			   to handle rthdr case. Ignore this complexity
			   for now.
			 */
			memset(&fl, 0, sizeof(fl));
			fl.proto = IPPROTO_TCP;
			ipv6_addr_copy(&fl.fl6_dst, &np->daddr);
			ipv6_addr_copy(&fl.fl6_src, &np->saddr);
			fl.oif = sk->sk_bound_dev_if;
			fl.fl_ip_dport = inet->dport;
			fl.fl_ip_sport = inet->sport;

			if ((err = ip6_dst_lookup(sk, &dst, &fl))) {
				sk->sk_err_soft = -err;
				goto out;
			}

			if ((err = xfrm_lookup(&dst, &fl, sk, 0)) < 0) {
				sk->sk_err_soft = -err;
				goto out;
			}

		} else
			dst_hold(dst);

393
		if (inet_csk(sk)->icsk_pmtu_cookie > dst_mtu(dst)) {
L
Linus Torvalds 已提交
394 395 396 397 398 399 400 401 402
			tcp_sync_mss(sk, dst_mtu(dst));
			tcp_simple_retransmit(sk);
		} /* else let the usual retransmit timer handle it */
		dst_release(dst);
		goto out;
	}

	icmpv6_err_convert(type, code, &err);

403
	/* Might be for an request_sock */
L
Linus Torvalds 已提交
404
	switch (sk->sk_state) {
405
		struct request_sock *req, **prev;
L
Linus Torvalds 已提交
406 407 408 409
	case TCP_LISTEN:
		if (sock_owned_by_user(sk))
			goto out;

410 411
		req = inet6_csk_search_req(sk, &prev, th->dest, &hdr->daddr,
					   &hdr->saddr, inet6_iif(skb));
L
Linus Torvalds 已提交
412 413 414 415 416 417 418 419
		if (!req)
			goto out;

		/* ICMPs are not backlogged, hence we cannot get
		 * an established socket here.
		 */
		BUG_TRAP(req->sk == NULL);

420
		if (seq != tcp_rsk(req)->snt_isn) {
L
Linus Torvalds 已提交
421 422 423 424
			NET_INC_STATS_BH(LINUX_MIB_OUTOFWINDOWICMPS);
			goto out;
		}

425
		inet_csk_reqsk_queue_drop(sk, req, prev);
L
Linus Torvalds 已提交
426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453
		goto out;

	case TCP_SYN_SENT:
	case TCP_SYN_RECV:  /* Cannot happen.
			       It can, it SYNs are crossed. --ANK */ 
		if (!sock_owned_by_user(sk)) {
			TCP_INC_STATS_BH(TCP_MIB_ATTEMPTFAILS);
			sk->sk_err = err;
			sk->sk_error_report(sk);		/* Wake people up to see the error (see connect in sock.c) */

			tcp_done(sk);
		} else
			sk->sk_err_soft = err;
		goto out;
	}

	if (!sock_owned_by_user(sk) && np->recverr) {
		sk->sk_err = err;
		sk->sk_error_report(sk);
	} else
		sk->sk_err_soft = err;

out:
	bh_unlock_sock(sk);
	sock_put(sk);
}


454
static int tcp_v6_send_synack(struct sock *sk, struct request_sock *req,
L
Linus Torvalds 已提交
455 456
			      struct dst_entry *dst)
{
457
	struct inet6_request_sock *treq = inet6_rsk(req);
L
Linus Torvalds 已提交
458 459 460 461 462 463 464 465 466
	struct ipv6_pinfo *np = inet6_sk(sk);
	struct sk_buff * skb;
	struct ipv6_txoptions *opt = NULL;
	struct in6_addr * final_p = NULL, final;
	struct flowi fl;
	int err = -1;

	memset(&fl, 0, sizeof(fl));
	fl.proto = IPPROTO_TCP;
467 468
	ipv6_addr_copy(&fl.fl6_dst, &treq->rmt_addr);
	ipv6_addr_copy(&fl.fl6_src, &treq->loc_addr);
L
Linus Torvalds 已提交
469
	fl.fl6_flowlabel = 0;
470 471
	fl.oif = treq->iif;
	fl.fl_ip_dport = inet_rsk(req)->rmt_port;
L
Linus Torvalds 已提交
472 473 474 475 476
	fl.fl_ip_sport = inet_sk(sk)->sport;

	if (dst == NULL) {
		opt = np->opt;
		if (opt == NULL &&
477
		    np->rxopt.bits.osrcrt == 2 &&
478 479
		    treq->pktopts) {
			struct sk_buff *pktopts = treq->pktopts;
L
Linus Torvalds 已提交
480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505
			struct inet6_skb_parm *rxopt = IP6CB(pktopts);
			if (rxopt->srcrt)
				opt = ipv6_invert_rthdr(sk, (struct ipv6_rt_hdr*)(pktopts->nh.raw + rxopt->srcrt));
		}

		if (opt && opt->srcrt) {
			struct rt0_hdr *rt0 = (struct rt0_hdr *) opt->srcrt;
			ipv6_addr_copy(&final, &fl.fl6_dst);
			ipv6_addr_copy(&fl.fl6_dst, rt0->addr);
			final_p = &final;
		}

		err = ip6_dst_lookup(sk, &dst, &fl);
		if (err)
			goto done;
		if (final_p)
			ipv6_addr_copy(&fl.fl6_dst, final_p);
		if ((err = xfrm_lookup(&dst, &fl, sk, 0)) < 0)
			goto done;
	}

	skb = tcp_make_synack(sk, dst, req);
	if (skb) {
		struct tcphdr *th = skb->h.th;

		th->check = tcp_v6_check(th, skb->len,
506
					 &treq->loc_addr, &treq->rmt_addr,
L
Linus Torvalds 已提交
507 508
					 csum_partial((char *)th, skb->len, skb->csum));

509
		ipv6_addr_copy(&fl.fl6_dst, &treq->rmt_addr);
L
Linus Torvalds 已提交
510 511 512 513 514 515 516 517
		err = ip6_xmit(sk, skb, &fl, opt, 0);
		if (err == NET_XMIT_CN)
			err = 0;
	}

done:
        if (opt && opt != np->opt)
		sock_kfree_s(sk, opt, opt->tot_len);
518
	dst_release(dst);
L
Linus Torvalds 已提交
519 520 521
	return err;
}

522
static void tcp_v6_reqsk_destructor(struct request_sock *req)
L
Linus Torvalds 已提交
523
{
524 525
	if (inet6_rsk(req)->pktopts)
		kfree_skb(inet6_rsk(req)->pktopts);
L
Linus Torvalds 已提交
526 527
}

528
static struct request_sock_ops tcp6_request_sock_ops = {
L
Linus Torvalds 已提交
529
	.family		=	AF_INET6,
530
	.obj_size	=	sizeof(struct tcp6_request_sock),
L
Linus Torvalds 已提交
531
	.rtx_syn_ack	=	tcp_v6_send_synack,
532 533
	.send_ack	=	tcp_v6_reqsk_send_ack,
	.destructor	=	tcp_v6_reqsk_destructor,
L
Linus Torvalds 已提交
534 535 536
	.send_reset	=	tcp_v6_send_reset
};

537 538 539 540 541
static struct timewait_sock_ops tcp6_timewait_sock_ops = {
	.twsk_obj_size	= sizeof(struct tcp6_timewait_sock),
	.twsk_unique	= tcp_twsk_unique,
};

542
static void tcp_v6_send_check(struct sock *sk, int len, struct sk_buff *skb)
L
Linus Torvalds 已提交
543 544
{
	struct ipv6_pinfo *np = inet6_sk(sk);
545
	struct tcphdr *th = skb->h.th;
L
Linus Torvalds 已提交
546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609

	if (skb->ip_summed == CHECKSUM_HW) {
		th->check = ~csum_ipv6_magic(&np->saddr, &np->daddr, len, IPPROTO_TCP,  0);
		skb->csum = offsetof(struct tcphdr, check);
	} else {
		th->check = csum_ipv6_magic(&np->saddr, &np->daddr, len, IPPROTO_TCP, 
					    csum_partial((char *)th, th->doff<<2, 
							 skb->csum));
	}
}


static void tcp_v6_send_reset(struct sk_buff *skb)
{
	struct tcphdr *th = skb->h.th, *t1; 
	struct sk_buff *buff;
	struct flowi fl;

	if (th->rst)
		return;

	if (!ipv6_unicast_destination(skb))
		return; 

	/*
	 * We need to grab some memory, and put together an RST,
	 * and then put it into the queue to be sent.
	 */

	buff = alloc_skb(MAX_HEADER + sizeof(struct ipv6hdr) + sizeof(struct tcphdr),
			 GFP_ATOMIC);
	if (buff == NULL) 
	  	return;

	skb_reserve(buff, MAX_HEADER + sizeof(struct ipv6hdr) + sizeof(struct tcphdr));

	t1 = (struct tcphdr *) skb_push(buff,sizeof(struct tcphdr));

	/* Swap the send and the receive. */
	memset(t1, 0, sizeof(*t1));
	t1->dest = th->source;
	t1->source = th->dest;
	t1->doff = sizeof(*t1)/4;
	t1->rst = 1;
  
	if(th->ack) {
	  	t1->seq = th->ack_seq;
	} else {
		t1->ack = 1;
		t1->ack_seq = htonl(ntohl(th->seq) + th->syn + th->fin
				    + skb->len - (th->doff<<2));
	}

	buff->csum = csum_partial((char *)t1, sizeof(*t1), 0);

	memset(&fl, 0, sizeof(fl));
	ipv6_addr_copy(&fl.fl6_dst, &skb->nh.ipv6h->saddr);
	ipv6_addr_copy(&fl.fl6_src, &skb->nh.ipv6h->daddr);

	t1->check = csum_ipv6_magic(&fl.fl6_src, &fl.fl6_dst,
				    sizeof(*t1), IPPROTO_TCP,
				    buff->csum);

	fl.proto = IPPROTO_TCP;
610
	fl.oif = inet6_iif(skb);
L
Linus Torvalds 已提交
611 612 613 614 615 616
	fl.fl_ip_dport = t1->dest;
	fl.fl_ip_sport = t1->source;

	/* sk = NULL, but it is safe for now. RST socket required. */
	if (!ip6_dst_lookup(NULL, &buff->dst, &fl)) {

A
Arnaldo Carvalho de Melo 已提交
617
		if (xfrm_lookup(&buff->dst, &fl, NULL, 0) >= 0) {
618
			ip6_xmit(tcp6_socket->sk, buff, &fl, NULL, 0);
A
Arnaldo Carvalho de Melo 已提交
619 620
			TCP_INC_STATS_BH(TCP_MIB_OUTSEGS);
			TCP_INC_STATS_BH(TCP_MIB_OUTRSTS);
L
Linus Torvalds 已提交
621
			return;
A
Arnaldo Carvalho de Melo 已提交
622
		}
L
Linus Torvalds 已提交
623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675
	}

	kfree_skb(buff);
}

static void tcp_v6_send_ack(struct sk_buff *skb, u32 seq, u32 ack, u32 win, u32 ts)
{
	struct tcphdr *th = skb->h.th, *t1;
	struct sk_buff *buff;
	struct flowi fl;
	int tot_len = sizeof(struct tcphdr);

	if (ts)
		tot_len += 3*4;

	buff = alloc_skb(MAX_HEADER + sizeof(struct ipv6hdr) + tot_len,
			 GFP_ATOMIC);
	if (buff == NULL)
		return;

	skb_reserve(buff, MAX_HEADER + sizeof(struct ipv6hdr) + tot_len);

	t1 = (struct tcphdr *) skb_push(buff,tot_len);

	/* Swap the send and the receive. */
	memset(t1, 0, sizeof(*t1));
	t1->dest = th->source;
	t1->source = th->dest;
	t1->doff = tot_len/4;
	t1->seq = htonl(seq);
	t1->ack_seq = htonl(ack);
	t1->ack = 1;
	t1->window = htons(win);
	
	if (ts) {
		u32 *ptr = (u32*)(t1 + 1);
		*ptr++ = htonl((TCPOPT_NOP << 24) | (TCPOPT_NOP << 16) |
			       (TCPOPT_TIMESTAMP << 8) | TCPOLEN_TIMESTAMP);
		*ptr++ = htonl(tcp_time_stamp);
		*ptr = htonl(ts);
	}

	buff->csum = csum_partial((char *)t1, tot_len, 0);

	memset(&fl, 0, sizeof(fl));
	ipv6_addr_copy(&fl.fl6_dst, &skb->nh.ipv6h->saddr);
	ipv6_addr_copy(&fl.fl6_src, &skb->nh.ipv6h->daddr);

	t1->check = csum_ipv6_magic(&fl.fl6_src, &fl.fl6_dst,
				    tot_len, IPPROTO_TCP,
				    buff->csum);

	fl.proto = IPPROTO_TCP;
676
	fl.oif = inet6_iif(skb);
L
Linus Torvalds 已提交
677 678 679 680
	fl.fl_ip_dport = t1->dest;
	fl.fl_ip_sport = t1->source;

	if (!ip6_dst_lookup(NULL, &buff->dst, &fl)) {
A
Arnaldo Carvalho de Melo 已提交
681
		if (xfrm_lookup(&buff->dst, &fl, NULL, 0) >= 0) {
682
			ip6_xmit(tcp6_socket->sk, buff, &fl, NULL, 0);
A
Arnaldo Carvalho de Melo 已提交
683
			TCP_INC_STATS_BH(TCP_MIB_OUTSEGS);
L
Linus Torvalds 已提交
684
			return;
A
Arnaldo Carvalho de Melo 已提交
685
		}
L
Linus Torvalds 已提交
686 687 688 689 690 691 692
	}

	kfree_skb(buff);
}

static void tcp_v6_timewait_ack(struct sock *sk, struct sk_buff *skb)
{
693 694
	struct inet_timewait_sock *tw = inet_twsk(sk);
	const struct tcp_timewait_sock *tcptw = tcp_twsk(sk);
L
Linus Torvalds 已提交
695

696 697 698
	tcp_v6_send_ack(skb, tcptw->tw_snd_nxt, tcptw->tw_rcv_nxt,
			tcptw->tw_rcv_wnd >> tw->tw_rcv_wscale,
			tcptw->tw_ts_recent);
L
Linus Torvalds 已提交
699

700
	inet_twsk_put(tw);
L
Linus Torvalds 已提交
701 702
}

703
static void tcp_v6_reqsk_send_ack(struct sk_buff *skb, struct request_sock *req)
L
Linus Torvalds 已提交
704
{
705
	tcp_v6_send_ack(skb, tcp_rsk(req)->snt_isn + 1, tcp_rsk(req)->rcv_isn + 1, req->rcv_wnd, req->ts_recent);
L
Linus Torvalds 已提交
706 707 708 709 710
}


static struct sock *tcp_v6_hnd_req(struct sock *sk,struct sk_buff *skb)
{
711
	struct request_sock *req, **prev;
712
	const struct tcphdr *th = skb->h.th;
L
Linus Torvalds 已提交
713 714 715
	struct sock *nsk;

	/* Find possible connection requests. */
716 717 718
	req = inet6_csk_search_req(sk, &prev, th->source,
				   &skb->nh.ipv6h->saddr,
				   &skb->nh.ipv6h->daddr, inet6_iif(skb));
L
Linus Torvalds 已提交
719 720 721
	if (req)
		return tcp_check_req(sk, skb, req, prev);

722 723 724
	nsk = __inet6_lookup_established(&tcp_hashinfo, &skb->nh.ipv6h->saddr,
					 th->source, &skb->nh.ipv6h->daddr,
					 ntohs(th->dest), inet6_iif(skb));
L
Linus Torvalds 已提交
725 726 727 728 729 730

	if (nsk) {
		if (nsk->sk_state != TCP_TIME_WAIT) {
			bh_lock_sock(nsk);
			return nsk;
		}
731
		inet_twsk_put((struct inet_timewait_sock *)nsk);
L
Linus Torvalds 已提交
732 733 734 735 736 737 738 739 740 741 742 743 744 745 746
		return NULL;
	}

#if 0 /*def CONFIG_SYN_COOKIES*/
	if (!th->rst && !th->syn && th->ack)
		sk = cookie_v6_check(sk, skb, &(IPCB(skb)->opt));
#endif
	return sk;
}

/* FIXME: this is substantially similar to the ipv4 code.
 * Can some kind of merge be done? -- erics
 */
static int tcp_v6_conn_request(struct sock *sk, struct sk_buff *skb)
{
747
	struct inet6_request_sock *treq;
L
Linus Torvalds 已提交
748 749 750
	struct ipv6_pinfo *np = inet6_sk(sk);
	struct tcp_options_received tmp_opt;
	struct tcp_sock *tp = tcp_sk(sk);
751
	struct request_sock *req = NULL;
L
Linus Torvalds 已提交
752 753 754 755 756 757 758 759 760 761 762
	__u32 isn = TCP_SKB_CB(skb)->when;

	if (skb->protocol == htons(ETH_P_IP))
		return tcp_v4_conn_request(sk, skb);

	if (!ipv6_unicast_destination(skb))
		goto drop; 

	/*
	 *	There are no SYN attacks on IPv6, yet...	
	 */
763
	if (inet_csk_reqsk_queue_is_full(sk) && !isn) {
L
Linus Torvalds 已提交
764 765 766 767 768
		if (net_ratelimit())
			printk(KERN_INFO "TCPv6: dropping request, synflood is possible\n");
		goto drop;		
	}

769
	if (sk_acceptq_is_full(sk) && inet_csk_reqsk_queue_young(sk) > 1)
L
Linus Torvalds 已提交
770 771
		goto drop;

772
	req = inet6_reqsk_alloc(&tcp6_request_sock_ops);
L
Linus Torvalds 已提交
773 774 775 776 777 778 779 780 781 782 783 784
	if (req == NULL)
		goto drop;

	tcp_clear_options(&tmp_opt);
	tmp_opt.mss_clamp = IPV6_MIN_MTU - sizeof(struct tcphdr) - sizeof(struct ipv6hdr);
	tmp_opt.user_mss = tp->rx_opt.user_mss;

	tcp_parse_options(skb, &tmp_opt, 0);

	tmp_opt.tstamp_ok = tmp_opt.saw_tstamp;
	tcp_openreq_init(req, &tmp_opt, skb);

785
	treq = inet6_rsk(req);
786 787
	ipv6_addr_copy(&treq->rmt_addr, &skb->nh.ipv6h->saddr);
	ipv6_addr_copy(&treq->loc_addr, &skb->nh.ipv6h->daddr);
L
Linus Torvalds 已提交
788
	TCP_ECN_create_request(req, skb->h.th);
789
	treq->pktopts = NULL;
L
Linus Torvalds 已提交
790
	if (ipv6_opt_accepted(sk, skb) ||
791 792
	    np->rxopt.bits.rxinfo || np->rxopt.bits.rxoinfo ||
	    np->rxopt.bits.rxhlim || np->rxopt.bits.rxohlim) {
L
Linus Torvalds 已提交
793
		atomic_inc(&skb->users);
794
		treq->pktopts = skb;
L
Linus Torvalds 已提交
795
	}
796
	treq->iif = sk->sk_bound_dev_if;
L
Linus Torvalds 已提交
797 798 799

	/* So that link locals have meaning */
	if (!sk->sk_bound_dev_if &&
800
	    ipv6_addr_type(&treq->rmt_addr) & IPV6_ADDR_LINKLOCAL)
801
		treq->iif = inet6_iif(skb);
L
Linus Torvalds 已提交
802 803 804 805

	if (isn == 0) 
		isn = tcp_v6_init_sequence(sk,skb);

806
	tcp_rsk(req)->snt_isn = isn;
L
Linus Torvalds 已提交
807 808 809 810

	if (tcp_v6_send_synack(sk, req, NULL))
		goto drop;

811
	inet6_csk_reqsk_queue_hash_add(sk, req, TCP_TIMEOUT_INIT);
L
Linus Torvalds 已提交
812 813 814 815
	return 0;

drop:
	if (req)
816
		reqsk_free(req);
L
Linus Torvalds 已提交
817 818 819 820 821 822

	TCP_INC_STATS_BH(TCP_MIB_ATTEMPTFAILS);
	return 0; /* don't send reset */
}

static struct sock * tcp_v6_syn_recv_sock(struct sock *sk, struct sk_buff *skb,
823
					  struct request_sock *req,
L
Linus Torvalds 已提交
824 825
					  struct dst_entry *dst)
{
826
	struct inet6_request_sock *treq = inet6_rsk(req);
L
Linus Torvalds 已提交
827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860
	struct ipv6_pinfo *newnp, *np = inet6_sk(sk);
	struct tcp6_sock *newtcp6sk;
	struct inet_sock *newinet;
	struct tcp_sock *newtp;
	struct sock *newsk;
	struct ipv6_txoptions *opt;

	if (skb->protocol == htons(ETH_P_IP)) {
		/*
		 *	v6 mapped
		 */

		newsk = tcp_v4_syn_recv_sock(sk, skb, req, dst);

		if (newsk == NULL) 
			return NULL;

		newtcp6sk = (struct tcp6_sock *)newsk;
		inet_sk(newsk)->pinet6 = &newtcp6sk->inet6;

		newinet = inet_sk(newsk);
		newnp = inet6_sk(newsk);
		newtp = tcp_sk(newsk);

		memcpy(newnp, np, sizeof(struct ipv6_pinfo));

		ipv6_addr_set(&newnp->daddr, 0, 0, htonl(0x0000FFFF),
			      newinet->daddr);

		ipv6_addr_set(&newnp->saddr, 0, 0, htonl(0x0000FFFF),
			      newinet->saddr);

		ipv6_addr_copy(&newnp->rcv_saddr, &newnp->saddr);

861
		inet_csk(newsk)->icsk_af_ops = &ipv6_mapped;
L
Linus Torvalds 已提交
862 863 864
		newsk->sk_backlog_rcv = tcp_v4_do_rcv;
		newnp->pktoptions  = NULL;
		newnp->opt	   = NULL;
865
		newnp->mcast_oif   = inet6_iif(skb);
L
Linus Torvalds 已提交
866 867
		newnp->mcast_hops  = skb->nh.ipv6h->hop_limit;

868 869 870 871
		/*
		 * No need to charge this sock to the relevant IPv6 refcnt debug socks count
		 * here, tcp_create_openreq_child now does this for us, see the comment in
		 * that function for the gory details. -acme
L
Linus Torvalds 已提交
872 873 874
		 */

		/* It is tricky place. Until this moment IPv4 tcp
875
		   worked with IPv6 icsk.icsk_af_ops.
L
Linus Torvalds 已提交
876 877
		   Sync it now.
		 */
878
		tcp_sync_mss(newsk, inet_csk(newsk)->icsk_pmtu_cookie);
L
Linus Torvalds 已提交
879 880 881 882 883 884 885 886 887

		return newsk;
	}

	opt = np->opt;

	if (sk_acceptq_is_full(sk))
		goto out_overflow;

888
	if (np->rxopt.bits.osrcrt == 2 &&
889 890
	    opt == NULL && treq->pktopts) {
		struct inet6_skb_parm *rxopt = IP6CB(treq->pktopts);
L
Linus Torvalds 已提交
891
		if (rxopt->srcrt)
892
			opt = ipv6_invert_rthdr(sk, (struct ipv6_rt_hdr *)(treq->pktopts->nh.raw + rxopt->srcrt));
L
Linus Torvalds 已提交
893 894 895 896 897 898 899 900
	}

	if (dst == NULL) {
		struct in6_addr *final_p = NULL, final;
		struct flowi fl;

		memset(&fl, 0, sizeof(fl));
		fl.proto = IPPROTO_TCP;
901
		ipv6_addr_copy(&fl.fl6_dst, &treq->rmt_addr);
L
Linus Torvalds 已提交
902 903 904 905 906 907
		if (opt && opt->srcrt) {
			struct rt0_hdr *rt0 = (struct rt0_hdr *) opt->srcrt;
			ipv6_addr_copy(&final, &fl.fl6_dst);
			ipv6_addr_copy(&fl.fl6_dst, rt0->addr);
			final_p = &final;
		}
908
		ipv6_addr_copy(&fl.fl6_src, &treq->loc_addr);
L
Linus Torvalds 已提交
909
		fl.oif = sk->sk_bound_dev_if;
910
		fl.fl_ip_dport = inet_rsk(req)->rmt_port;
L
Linus Torvalds 已提交
911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926
		fl.fl_ip_sport = inet_sk(sk)->sport;

		if (ip6_dst_lookup(sk, &dst, &fl))
			goto out;

		if (final_p)
			ipv6_addr_copy(&fl.fl6_dst, final_p);

		if ((xfrm_lookup(&dst, &fl, sk, 0)) < 0)
			goto out;
	} 

	newsk = tcp_create_openreq_child(sk, req, skb);
	if (newsk == NULL)
		goto out;

927 928 929 930 931
	/*
	 * No need to charge this sock to the relevant IPv6 refcnt debug socks
	 * count here, tcp_create_openreq_child now does this for us, see the
	 * comment in that function for the gory details. -acme
	 */
L
Linus Torvalds 已提交
932 933 934 935 936 937 938 939 940 941 942 943 944 945

	ip6_dst_store(newsk, dst, NULL);
	newsk->sk_route_caps = dst->dev->features &
		~(NETIF_F_IP_CSUM | NETIF_F_TSO);

	newtcp6sk = (struct tcp6_sock *)newsk;
	inet_sk(newsk)->pinet6 = &newtcp6sk->inet6;

	newtp = tcp_sk(newsk);
	newinet = inet_sk(newsk);
	newnp = inet6_sk(newsk);

	memcpy(newnp, np, sizeof(struct ipv6_pinfo));

946 947 948 949
	ipv6_addr_copy(&newnp->daddr, &treq->rmt_addr);
	ipv6_addr_copy(&newnp->saddr, &treq->loc_addr);
	ipv6_addr_copy(&newnp->rcv_saddr, &treq->loc_addr);
	newsk->sk_bound_dev_if = treq->iif;
L
Linus Torvalds 已提交
950 951 952 953 954 955 956 957 958 959 960 961

	/* Now IPv6 options... 

	   First: no IPv4 options.
	 */
	newinet->opt = NULL;

	/* Clone RX bits */
	newnp->rxopt.all = np->rxopt.all;

	/* Clone pktoptions received with SYN */
	newnp->pktoptions = NULL;
962 963 964 965
	if (treq->pktopts != NULL) {
		newnp->pktoptions = skb_clone(treq->pktopts, GFP_ATOMIC);
		kfree_skb(treq->pktopts);
		treq->pktopts = NULL;
L
Linus Torvalds 已提交
966 967 968 969
		if (newnp->pktoptions)
			skb_set_owner_r(newnp->pktoptions, newsk);
	}
	newnp->opt	  = NULL;
970
	newnp->mcast_oif  = inet6_iif(skb);
L
Linus Torvalds 已提交
971 972 973 974 975 976 977 978 979 980 981 982 983 984
	newnp->mcast_hops = skb->nh.ipv6h->hop_limit;

	/* Clone native IPv6 options from listening socket (if any)

	   Yes, keeping reference count would be much more clever,
	   but we make one more one thing there: reattach optmem
	   to newsk.
	 */
	if (opt) {
		newnp->opt = ipv6_dup_options(newsk, opt);
		if (opt != np->opt)
			sock_kfree_s(sk, opt, opt->tot_len);
	}

985
	inet_csk(newsk)->icsk_ext_hdr_len = 0;
L
Linus Torvalds 已提交
986
	if (newnp->opt)
987 988
		inet_csk(newsk)->icsk_ext_hdr_len = (newnp->opt->opt_nflen +
						     newnp->opt->opt_flen);
L
Linus Torvalds 已提交
989

J
John Heffner 已提交
990
	tcp_mtup_init(newsk);
L
Linus Torvalds 已提交
991 992 993 994 995 996
	tcp_sync_mss(newsk, dst_mtu(dst));
	newtp->advmss = dst_metric(dst, RTAX_ADVMSS);
	tcp_initialize_rcv_mss(newsk);

	newinet->daddr = newinet->saddr = newinet->rcv_saddr = LOOPBACK4_IPV6;

997
	__inet6_hash(&tcp_hashinfo, newsk);
998
	inet_inherit_port(&tcp_hashinfo, sk, newsk);
L
Linus Torvalds 已提交
999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015

	return newsk;

out_overflow:
	NET_INC_STATS_BH(LINUX_MIB_LISTENOVERFLOWS);
out:
	NET_INC_STATS_BH(LINUX_MIB_LISTENDROPS);
	if (opt && opt != np->opt)
		sock_kfree_s(sk, opt, opt->tot_len);
	dst_release(dst);
	return NULL;
}

static int tcp_v6_checksum_init(struct sk_buff *skb)
{
	if (skb->ip_summed == CHECKSUM_HW) {
		if (!tcp_v6_check(skb->h.th,skb->len,&skb->nh.ipv6h->saddr,
1016 1017
				  &skb->nh.ipv6h->daddr,skb->csum)) {
			skb->ip_summed = CHECKSUM_UNNECESSARY;
L
Linus Torvalds 已提交
1018
			return 0;
1019
		}
L
Linus Torvalds 已提交
1020
	}
1021 1022 1023 1024

	skb->csum = ~tcp_v6_check(skb->h.th,skb->len,&skb->nh.ipv6h->saddr,
				  &skb->nh.ipv6h->daddr, 0);

L
Linus Torvalds 已提交
1025
	if (skb->len <= 76) {
1026
		return __skb_checksum_complete(skb);
L
Linus Torvalds 已提交
1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142
	}
	return 0;
}

/* The socket must have it's spinlock held when we get
 * here.
 *
 * We have a potential double-lock case here, so even when
 * doing backlog processing we use the BH locking scheme.
 * This is because we cannot sleep with the original spinlock
 * held.
 */
static int tcp_v6_do_rcv(struct sock *sk, struct sk_buff *skb)
{
	struct ipv6_pinfo *np = inet6_sk(sk);
	struct tcp_sock *tp;
	struct sk_buff *opt_skb = NULL;

	/* Imagine: socket is IPv6. IPv4 packet arrives,
	   goes to IPv4 receive handler and backlogged.
	   From backlog it always goes here. Kerboom...
	   Fortunately, tcp_rcv_established and rcv_established
	   handle them correctly, but it is not case with
	   tcp_v6_hnd_req and tcp_v6_send_reset().   --ANK
	 */

	if (skb->protocol == htons(ETH_P_IP))
		return tcp_v4_do_rcv(sk, skb);

	if (sk_filter(sk, skb, 0))
		goto discard;

	/*
	 *	socket locking is here for SMP purposes as backlog rcv
	 *	is currently called with bh processing disabled.
	 */

	/* Do Stevens' IPV6_PKTOPTIONS.

	   Yes, guys, it is the only place in our code, where we
	   may make it not affecting IPv4.
	   The rest of code is protocol independent,
	   and I do not like idea to uglify IPv4.

	   Actually, all the idea behind IPV6_PKTOPTIONS
	   looks not very well thought. For now we latch
	   options, received in the last packet, enqueued
	   by tcp. Feel free to propose better solution.
	                                       --ANK (980728)
	 */
	if (np->rxopt.all)
		opt_skb = skb_clone(skb, GFP_ATOMIC);

	if (sk->sk_state == TCP_ESTABLISHED) { /* Fast path */
		TCP_CHECK_TIMER(sk);
		if (tcp_rcv_established(sk, skb, skb->h.th, skb->len))
			goto reset;
		TCP_CHECK_TIMER(sk);
		if (opt_skb)
			goto ipv6_pktoptions;
		return 0;
	}

	if (skb->len < (skb->h.th->doff<<2) || tcp_checksum_complete(skb))
		goto csum_err;

	if (sk->sk_state == TCP_LISTEN) { 
		struct sock *nsk = tcp_v6_hnd_req(sk, skb);
		if (!nsk)
			goto discard;

		/*
		 * Queue it on the new socket if the new socket is active,
		 * otherwise we just shortcircuit this and continue with
		 * the new socket..
		 */
 		if(nsk != sk) {
			if (tcp_child_process(sk, nsk, skb))
				goto reset;
			if (opt_skb)
				__kfree_skb(opt_skb);
			return 0;
		}
	}

	TCP_CHECK_TIMER(sk);
	if (tcp_rcv_state_process(sk, skb, skb->h.th, skb->len))
		goto reset;
	TCP_CHECK_TIMER(sk);
	if (opt_skb)
		goto ipv6_pktoptions;
	return 0;

reset:
	tcp_v6_send_reset(skb);
discard:
	if (opt_skb)
		__kfree_skb(opt_skb);
	kfree_skb(skb);
	return 0;
csum_err:
	TCP_INC_STATS_BH(TCP_MIB_INERRS);
	goto discard;


ipv6_pktoptions:
	/* Do you ask, what is it?

	   1. skb was enqueued by tcp.
	   2. skb is added to tail of read queue, rather than out of order.
	   3. socket is not in passive state.
	   4. Finally, it really contains options, which user wants to receive.
	 */
	tp = tcp_sk(sk);
	if (TCP_SKB_CB(opt_skb)->end_seq == tp->rcv_nxt &&
	    !((1 << sk->sk_state) & (TCPF_CLOSE | TCPF_LISTEN))) {
1143
		if (np->rxopt.bits.rxinfo || np->rxopt.bits.rxoinfo)
1144
			np->mcast_oif = inet6_iif(opt_skb);
1145
		if (np->rxopt.bits.rxhlim || np->rxopt.bits.rxohlim)
L
Linus Torvalds 已提交
1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160
			np->mcast_hops = opt_skb->nh.ipv6h->hop_limit;
		if (ipv6_opt_accepted(sk, opt_skb)) {
			skb_set_owner_r(opt_skb, sk);
			opt_skb = xchg(&np->pktoptions, opt_skb);
		} else {
			__kfree_skb(opt_skb);
			opt_skb = xchg(&np->pktoptions, NULL);
		}
	}

	if (opt_skb)
		kfree_skb(opt_skb);
	return 0;
}

1161
static int tcp_v6_rcv(struct sk_buff **pskb)
L
Linus Torvalds 已提交
1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186
{
	struct sk_buff *skb = *pskb;
	struct tcphdr *th;	
	struct sock *sk;
	int ret;

	if (skb->pkt_type != PACKET_HOST)
		goto discard_it;

	/*
	 *	Count it even if it's bad.
	 */
	TCP_INC_STATS_BH(TCP_MIB_INSEGS);

	if (!pskb_may_pull(skb, sizeof(struct tcphdr)))
		goto discard_it;

	th = skb->h.th;

	if (th->doff < sizeof(struct tcphdr)/4)
		goto bad_packet;
	if (!pskb_may_pull(skb, th->doff*4))
		goto discard_it;

	if ((skb->ip_summed != CHECKSUM_UNNECESSARY &&
1187
	     tcp_v6_checksum_init(skb)))
L
Linus Torvalds 已提交
1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198
		goto bad_packet;

	th = skb->h.th;
	TCP_SKB_CB(skb)->seq = ntohl(th->seq);
	TCP_SKB_CB(skb)->end_seq = (TCP_SKB_CB(skb)->seq + th->syn + th->fin +
				    skb->len - th->doff*4);
	TCP_SKB_CB(skb)->ack_seq = ntohl(th->ack_seq);
	TCP_SKB_CB(skb)->when = 0;
	TCP_SKB_CB(skb)->flags = ipv6_get_dsfield(skb->nh.ipv6h);
	TCP_SKB_CB(skb)->sacked = 0;

1199 1200 1201
	sk = __inet6_lookup(&tcp_hashinfo, &skb->nh.ipv6h->saddr, th->source,
			    &skb->nh.ipv6h->daddr, ntohs(th->dest),
			    inet6_iif(skb));
L
Linus Torvalds 已提交
1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220

	if (!sk)
		goto no_tcp_socket;

process:
	if (sk->sk_state == TCP_TIME_WAIT)
		goto do_time_wait;

	if (!xfrm6_policy_check(sk, XFRM_POLICY_IN, skb))
		goto discard_and_relse;

	if (sk_filter(sk, skb, 0))
		goto discard_and_relse;

	skb->dev = NULL;

	bh_lock_sock(sk);
	ret = 0;
	if (!sock_owned_by_user(sk)) {
C
Chris Leech 已提交
1221 1222 1223 1224 1225 1226 1227 1228 1229 1230
#ifdef CONFIG_NET_DMA
                struct tcp_sock *tp = tcp_sk(sk);
                if (tp->ucopy.dma_chan)
                        ret = tcp_v6_do_rcv(sk, skb);
                else
#endif
		{
			if (!tcp_prequeue(sk, skb))
				ret = tcp_v6_do_rcv(sk, skb);
		}
L
Linus Torvalds 已提交
1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263
	} else
		sk_add_backlog(sk, skb);
	bh_unlock_sock(sk);

	sock_put(sk);
	return ret ? -1 : 0;

no_tcp_socket:
	if (!xfrm6_policy_check(NULL, XFRM_POLICY_IN, skb))
		goto discard_it;

	if (skb->len < (th->doff<<2) || tcp_checksum_complete(skb)) {
bad_packet:
		TCP_INC_STATS_BH(TCP_MIB_INERRS);
	} else {
		tcp_v6_send_reset(skb);
	}

discard_it:

	/*
	 *	Discard frame
	 */

	kfree_skb(skb);
	return 0;

discard_and_relse:
	sock_put(sk);
	goto discard_it;

do_time_wait:
	if (!xfrm6_policy_check(NULL, XFRM_POLICY_IN, skb)) {
1264
		inet_twsk_put((struct inet_timewait_sock *)sk);
L
Linus Torvalds 已提交
1265 1266 1267 1268 1269
		goto discard_it;
	}

	if (skb->len < (th->doff<<2) || tcp_checksum_complete(skb)) {
		TCP_INC_STATS_BH(TCP_MIB_INERRS);
1270
		inet_twsk_put((struct inet_timewait_sock *)sk);
L
Linus Torvalds 已提交
1271 1272 1273
		goto discard_it;
	}

1274 1275
	switch (tcp_timewait_state_process((struct inet_timewait_sock *)sk,
					   skb, th)) {
L
Linus Torvalds 已提交
1276 1277 1278 1279
	case TCP_TW_SYN:
	{
		struct sock *sk2;

1280 1281 1282
		sk2 = inet6_lookup_listener(&tcp_hashinfo,
					    &skb->nh.ipv6h->daddr,
					    ntohs(th->dest), inet6_iif(skb));
L
Linus Torvalds 已提交
1283
		if (sk2 != NULL) {
1284 1285 1286
			struct inet_timewait_sock *tw = inet_twsk(sk);
			inet_twsk_deschedule(tw, &tcp_death_row);
			inet_twsk_put(tw);
L
Linus Torvalds 已提交
1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307
			sk = sk2;
			goto process;
		}
		/* Fall through to ACK */
	}
	case TCP_TW_ACK:
		tcp_v6_timewait_ack(sk, skb);
		break;
	case TCP_TW_RST:
		goto no_tcp_socket;
	case TCP_TW_SUCCESS:;
	}
	goto discard_it;
}

static int tcp_v6_remember_stamp(struct sock *sk)
{
	/* Alas, not yet... */
	return 0;
}

1308
static struct inet_connection_sock_af_ops ipv6_specific = {
1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319
	.queue_xmit	   = inet6_csk_xmit,
	.send_check	   = tcp_v6_send_check,
	.rebuild_header	   = inet6_sk_rebuild_header,
	.conn_request	   = tcp_v6_conn_request,
	.syn_recv_sock	   = tcp_v6_syn_recv_sock,
	.remember_stamp	   = tcp_v6_remember_stamp,
	.net_header_len	   = sizeof(struct ipv6hdr),
	.setsockopt	   = ipv6_setsockopt,
	.getsockopt	   = ipv6_getsockopt,
	.addr2sockaddr	   = inet6_csk_addr2sockaddr,
	.sockaddr_len	   = sizeof(struct sockaddr_in6),
1320
#ifdef CONFIG_COMPAT
1321 1322
	.compat_setsockopt = compat_ipv6_setsockopt,
	.compat_getsockopt = compat_ipv6_getsockopt,
1323
#endif
L
Linus Torvalds 已提交
1324 1325 1326 1327 1328 1329
};

/*
 *	TCP over IPv4 via INET6 API
 */

1330
static struct inet_connection_sock_af_ops ipv6_mapped = {
1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341
	.queue_xmit	   = ip_queue_xmit,
	.send_check	   = tcp_v4_send_check,
	.rebuild_header	   = inet_sk_rebuild_header,
	.conn_request	   = tcp_v6_conn_request,
	.syn_recv_sock	   = tcp_v6_syn_recv_sock,
	.remember_stamp	   = tcp_v4_remember_stamp,
	.net_header_len	   = sizeof(struct iphdr),
	.setsockopt	   = ipv6_setsockopt,
	.getsockopt	   = ipv6_getsockopt,
	.addr2sockaddr	   = inet6_csk_addr2sockaddr,
	.sockaddr_len	   = sizeof(struct sockaddr_in6),
1342
#ifdef CONFIG_COMPAT
1343 1344
	.compat_setsockopt = compat_ipv6_setsockopt,
	.compat_getsockopt = compat_ipv6_getsockopt,
1345
#endif
L
Linus Torvalds 已提交
1346 1347 1348 1349 1350 1351 1352
};

/* NOTE: A lot of things set to zero explicitly by call to
 *       sk_alloc() so need not be done here.
 */
static int tcp_v6_init_sock(struct sock *sk)
{
1353
	struct inet_connection_sock *icsk = inet_csk(sk);
L
Linus Torvalds 已提交
1354 1355 1356 1357 1358 1359
	struct tcp_sock *tp = tcp_sk(sk);

	skb_queue_head_init(&tp->out_of_order_queue);
	tcp_init_xmit_timers(sk);
	tcp_prequeue_init(tp);

1360
	icsk->icsk_rto = TCP_TIMEOUT_INIT;
L
Linus Torvalds 已提交
1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374
	tp->mdev = TCP_TIMEOUT_INIT;

	/* 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
	 */
	tp->snd_cwnd = 2;

	/* See draft-stevens-tcpca-spec-01 for discussion of the
	 * initialization of these values.
	 */
	tp->snd_ssthresh = 0x7fffffff;
	tp->snd_cwnd_clamp = ~0;
1375
	tp->mss_cache = 536;
L
Linus Torvalds 已提交
1376 1377 1378 1379 1380

	tp->reordering = sysctl_tcp_reordering;

	sk->sk_state = TCP_CLOSE;

1381
	icsk->icsk_af_ops = &ipv6_specific;
1382
	icsk->icsk_ca_ops = &tcp_init_congestion_ops;
1383
	icsk->icsk_sync_mss = tcp_sync_mss;
L
Linus Torvalds 已提交
1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402
	sk->sk_write_space = sk_stream_write_space;
	sock_set_flag(sk, SOCK_USE_WRITE_QUEUE);

	sk->sk_sndbuf = sysctl_tcp_wmem[1];
	sk->sk_rcvbuf = sysctl_tcp_rmem[1];

	atomic_inc(&tcp_sockets_allocated);

	return 0;
}

static int tcp_v6_destroy_sock(struct sock *sk)
{
	tcp_v4_destroy_sock(sk);
	return inet6_destroy_sock(sk);
}

/* Proc filesystem TCPv6 sock list dumping. */
static void get_openreq6(struct seq_file *seq, 
1403
			 struct sock *sk, struct request_sock *req, int i, int uid)
L
Linus Torvalds 已提交
1404 1405
{
	int ttd = req->expires - jiffies;
1406 1407
	struct in6_addr *src = &inet6_rsk(req)->loc_addr;
	struct in6_addr *dest = &inet6_rsk(req)->rmt_addr;
L
Linus Torvalds 已提交
1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420

	if (ttd < 0)
		ttd = 0;

	seq_printf(seq,
		   "%4d: %08X%08X%08X%08X:%04X %08X%08X%08X%08X:%04X "
		   "%02X %08X:%08X %02X:%08lX %08X %5d %8d %d %d %p\n",
		   i,
		   src->s6_addr32[0], src->s6_addr32[1],
		   src->s6_addr32[2], src->s6_addr32[3],
		   ntohs(inet_sk(sk)->sport),
		   dest->s6_addr32[0], dest->s6_addr32[1],
		   dest->s6_addr32[2], dest->s6_addr32[3],
1421
		   ntohs(inet_rsk(req)->rmt_port),
L
Linus Torvalds 已提交
1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440
		   TCP_SYN_RECV,
		   0,0, /* could print option size, but that is af dependent. */
		   1,   /* timers active (only the expire timer) */  
		   jiffies_to_clock_t(ttd), 
		   req->retrans,
		   uid,
		   0,  /* non standard timer */  
		   0, /* open_requests have no inode */
		   0, req);
}

static void get_tcp6_sock(struct seq_file *seq, struct sock *sp, int i)
{
	struct in6_addr *dest, *src;
	__u16 destp, srcp;
	int timer_active;
	unsigned long timer_expires;
	struct inet_sock *inet = inet_sk(sp);
	struct tcp_sock *tp = tcp_sk(sp);
1441
	const struct inet_connection_sock *icsk = inet_csk(sp);
L
Linus Torvalds 已提交
1442 1443 1444 1445 1446 1447
	struct ipv6_pinfo *np = inet6_sk(sp);

	dest  = &np->daddr;
	src   = &np->rcv_saddr;
	destp = ntohs(inet->dport);
	srcp  = ntohs(inet->sport);
1448 1449

	if (icsk->icsk_pending == ICSK_TIME_RETRANS) {
L
Linus Torvalds 已提交
1450
		timer_active	= 1;
1451 1452
		timer_expires	= icsk->icsk_timeout;
	} else if (icsk->icsk_pending == ICSK_TIME_PROBE0) {
L
Linus Torvalds 已提交
1453
		timer_active	= 4;
1454
		timer_expires	= icsk->icsk_timeout;
L
Linus Torvalds 已提交
1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474
	} else if (timer_pending(&sp->sk_timer)) {
		timer_active	= 2;
		timer_expires	= sp->sk_timer.expires;
	} else {
		timer_active	= 0;
		timer_expires = jiffies;
	}

	seq_printf(seq,
		   "%4d: %08X%08X%08X%08X:%04X %08X%08X%08X%08X:%04X "
		   "%02X %08X:%08X %02X:%08lX %08X %5d %8d %lu %d %p %u %u %u %u %d\n",
		   i,
		   src->s6_addr32[0], src->s6_addr32[1],
		   src->s6_addr32[2], src->s6_addr32[3], srcp,
		   dest->s6_addr32[0], dest->s6_addr32[1],
		   dest->s6_addr32[2], dest->s6_addr32[3], destp,
		   sp->sk_state, 
		   tp->write_seq-tp->snd_una, tp->rcv_nxt-tp->copied_seq,
		   timer_active,
		   jiffies_to_clock_t(timer_expires - jiffies),
1475
		   icsk->icsk_retransmits,
L
Linus Torvalds 已提交
1476
		   sock_i_uid(sp),
1477
		   icsk->icsk_probes_out,
L
Linus Torvalds 已提交
1478 1479
		   sock_i_ino(sp),
		   atomic_read(&sp->sk_refcnt), sp,
1480 1481 1482
		   icsk->icsk_rto,
		   icsk->icsk_ack.ato,
		   (icsk->icsk_ack.quick << 1 ) | icsk->icsk_ack.pingpong,
L
Linus Torvalds 已提交
1483 1484 1485 1486 1487
		   tp->snd_cwnd, tp->snd_ssthresh>=0xFFFF?-1:tp->snd_ssthresh
		   );
}

static void get_timewait6_sock(struct seq_file *seq, 
1488
			       struct inet_timewait_sock *tw, int i)
L
Linus Torvalds 已提交
1489 1490 1491
{
	struct in6_addr *dest, *src;
	__u16 destp, srcp;
1492
	struct inet6_timewait_sock *tw6 = inet6_twsk((struct sock *)tw);
L
Linus Torvalds 已提交
1493 1494 1495 1496 1497
	int ttd = tw->tw_ttd - jiffies;

	if (ttd < 0)
		ttd = 0;

1498 1499
	dest = &tw6->tw_v6_daddr;
	src  = &tw6->tw_v6_rcv_saddr;
L
Linus Torvalds 已提交
1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573
	destp = ntohs(tw->tw_dport);
	srcp  = ntohs(tw->tw_sport);

	seq_printf(seq,
		   "%4d: %08X%08X%08X%08X:%04X %08X%08X%08X%08X:%04X "
		   "%02X %08X:%08X %02X:%08lX %08X %5d %8d %d %d %p\n",
		   i,
		   src->s6_addr32[0], src->s6_addr32[1],
		   src->s6_addr32[2], src->s6_addr32[3], srcp,
		   dest->s6_addr32[0], dest->s6_addr32[1],
		   dest->s6_addr32[2], dest->s6_addr32[3], destp,
		   tw->tw_substate, 0, 0,
		   3, jiffies_to_clock_t(ttd), 0, 0, 0, 0,
		   atomic_read(&tw->tw_refcnt), tw);
}

#ifdef CONFIG_PROC_FS
static int tcp6_seq_show(struct seq_file *seq, void *v)
{
	struct tcp_iter_state *st;

	if (v == SEQ_START_TOKEN) {
		seq_puts(seq,
			 "  sl  "
			 "local_address                         "
			 "remote_address                        "
			 "st tx_queue rx_queue tr tm->when retrnsmt"
			 "   uid  timeout inode\n");
		goto out;
	}
	st = seq->private;

	switch (st->state) {
	case TCP_SEQ_STATE_LISTENING:
	case TCP_SEQ_STATE_ESTABLISHED:
		get_tcp6_sock(seq, v, st->num);
		break;
	case TCP_SEQ_STATE_OPENREQ:
		get_openreq6(seq, st->syn_wait_sk, v, st->num, st->uid);
		break;
	case TCP_SEQ_STATE_TIME_WAIT:
		get_timewait6_sock(seq, v, st->num);
		break;
	}
out:
	return 0;
}

static struct file_operations tcp6_seq_fops;
static struct tcp_seq_afinfo tcp6_seq_afinfo = {
	.owner		= THIS_MODULE,
	.name		= "tcp6",
	.family		= AF_INET6,
	.seq_show	= tcp6_seq_show,
	.seq_fops	= &tcp6_seq_fops,
};

int __init tcp6_proc_init(void)
{
	return tcp_proc_register(&tcp6_seq_afinfo);
}

void tcp6_proc_exit(void)
{
	tcp_proc_unregister(&tcp6_seq_afinfo);
}
#endif

struct proto tcpv6_prot = {
	.name			= "TCPv6",
	.owner			= THIS_MODULE,
	.close			= tcp_close,
	.connect		= tcp_v6_connect,
	.disconnect		= tcp_disconnect,
1574
	.accept			= inet_csk_accept,
L
Linus Torvalds 已提交
1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590
	.ioctl			= tcp_ioctl,
	.init			= tcp_v6_init_sock,
	.destroy		= tcp_v6_destroy_sock,
	.shutdown		= tcp_shutdown,
	.setsockopt		= tcp_setsockopt,
	.getsockopt		= tcp_getsockopt,
	.sendmsg		= tcp_sendmsg,
	.recvmsg		= tcp_recvmsg,
	.backlog_rcv		= tcp_v6_do_rcv,
	.hash			= tcp_v6_hash,
	.unhash			= tcp_unhash,
	.get_port		= tcp_v6_get_port,
	.enter_memory_pressure	= tcp_enter_memory_pressure,
	.sockets_allocated	= &tcp_sockets_allocated,
	.memory_allocated	= &tcp_memory_allocated,
	.memory_pressure	= &tcp_memory_pressure,
1591
	.orphan_count		= &tcp_orphan_count,
L
Linus Torvalds 已提交
1592 1593 1594 1595 1596
	.sysctl_mem		= sysctl_tcp_mem,
	.sysctl_wmem		= sysctl_tcp_wmem,
	.sysctl_rmem		= sysctl_tcp_rmem,
	.max_header		= MAX_TCP_HEADER,
	.obj_size		= sizeof(struct tcp6_sock),
1597
	.twsk_prot		= &tcp6_timewait_sock_ops,
1598
	.rsk_prot		= &tcp6_request_sock_ops,
1599 1600 1601 1602
#ifdef CONFIG_COMPAT
	.compat_setsockopt	= compat_tcp_setsockopt,
	.compat_getsockopt	= compat_tcp_getsockopt,
#endif
L
Linus Torvalds 已提交
1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617
};

static struct inet6_protocol tcpv6_protocol = {
	.handler	=	tcp_v6_rcv,
	.err_handler	=	tcp_v6_err,
	.flags		=	INET6_PROTO_NOPOLICY|INET6_PROTO_FINAL,
};

static struct inet_protosw tcpv6_protosw = {
	.type		=	SOCK_STREAM,
	.protocol	=	IPPROTO_TCP,
	.prot		=	&tcpv6_prot,
	.ops		=	&inet6_stream_ops,
	.capability	=	-1,
	.no_check	=	0,
1618 1619
	.flags		=	INET_PROTOSW_PERMANENT |
				INET_PROTOSW_ICSK,
L
Linus Torvalds 已提交
1620 1621 1622 1623 1624 1625 1626 1627
};

void __init tcpv6_init(void)
{
	/* register inet6 protocol */
	if (inet6_add_protocol(&tcpv6_protocol, IPPROTO_TCP) < 0)
		printk(KERN_ERR "tcpv6_init: Could not register protocol\n");
	inet6_register_protosw(&tcpv6_protosw);
1628

1629 1630
	if (inet_csk_ctl_sock_create(&tcp6_socket, PF_INET6, SOCK_RAW,
				     IPPROTO_TCP) < 0)
1631
		panic("Failed to create the TCPv6 control socket.\n");
L
Linus Torvalds 已提交
1632
}