tcp_ipv6.c 55.5 KB
Newer Older
L
Linus Torvalds 已提交
1 2
/*
 *	TCP over IPv6
3
 *	Linux INET6 implementation
L
Linus Torvalds 已提交
4 5
 *
 *	Authors:
6
 *	Pedro Roque		<roque@di.fc.ul.pt>
L
Linus Torvalds 已提交
7
 *
8
 *	Based on:
L
Linus Torvalds 已提交
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
 *	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.
 */

H
Herbert Xu 已提交
26
#include <linux/bottom_half.h>
L
Linus Torvalds 已提交
27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47
#include <linux/module.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>
48
#include <net/inet6_hashtables.h>
49
#include <net/inet6_connection_sock.h>
L
Linus Torvalds 已提交
50 51 52 53 54 55 56 57 58 59
#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/snmp.h>
#include <net/dsfield.h>
60
#include <net/timewait_sock.h>
61
#include <net/netdma.h>
62
#include <net/inet_common.h>
L
Linus Torvalds 已提交
63 64 65 66 67 68

#include <asm/uaccess.h>

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

69 70 71 72
#include <linux/crypto.h>
#include <linux/scatterlist.h>

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

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

78 79
static const struct inet_connection_sock_af_ops ipv6_mapped;
static const struct inet_connection_sock_af_ops ipv6_specific;
80
#ifdef CONFIG_TCP_MD5SIG
81 82
static const struct tcp_sock_af_ops tcp_sock_ipv6_specific;
static const struct tcp_sock_af_ops tcp_sock_ipv6_mapped_specific;
83 84 85 86 87 88
#else
static struct tcp_md5sig_key *tcp_v6_md5_do_lookup(struct sock *sk,
						   struct in6_addr *addr)
{
	return NULL;
}
89
#endif
L
Linus Torvalds 已提交
90 91 92 93

static void tcp_v6_hash(struct sock *sk)
{
	if (sk->sk_state != TCP_CLOSE) {
94
		if (inet_csk(sk)->icsk_af_ops == &ipv6_mapped) {
L
Linus Torvalds 已提交
95 96 97 98
			tcp_prot.hash(sk);
			return;
		}
		local_bh_disable();
99
		__inet6_hash(sk, NULL);
L
Linus Torvalds 已提交
100 101 102 103
		local_bh_enable();
	}
}

H
Herbert Xu 已提交
104
static __inline__ __sum16 tcp_v6_check(int len,
105 106
				   struct in6_addr *saddr,
				   struct in6_addr *daddr,
107
				   __wsum base)
L
Linus Torvalds 已提交
108 109 110 111
{
	return csum_ipv6_magic(saddr, daddr, len, IPPROTO_TCP, base);
}

112
static __u32 tcp_v6_init_sequence(struct sk_buff *skb)
L
Linus Torvalds 已提交
113
{
114 115
	return secure_tcpv6_sequence_number(ipv6_hdr(skb)->daddr.s6_addr32,
					    ipv6_hdr(skb)->saddr.s6_addr32,
116 117
					    tcp_hdr(skb)->dest,
					    tcp_hdr(skb)->source);
L
Linus Torvalds 已提交
118 119
}

120
static int tcp_v6_connect(struct sock *sk, struct sockaddr *uaddr,
L
Linus Torvalds 已提交
121 122 123
			  int addr_len)
{
	struct sockaddr_in6 *usin = (struct sockaddr_in6 *) uaddr;
124
	struct inet_sock *inet = inet_sk(sk);
125
	struct inet_connection_sock *icsk = inet_csk(sk);
L
Linus Torvalds 已提交
126 127 128 129 130 131 132 133
	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;

134
	if (addr_len < SIN6_LEN_RFC2133)
L
Linus Torvalds 已提交
135 136
		return -EINVAL;

137
	if (usin->sin6_family != AF_INET6)
L
Linus Torvalds 已提交
138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155
		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);
		}
	}

	/*
156 157 158 159 160
	 *	connect() to INADDR_ANY means loopback (BSD'ism).
	 */

	if(ipv6_addr_any(&usin->sin6_addr))
		usin->sin6_addr.s6_addr[15] = 0x1;
L
Linus Torvalds 已提交
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

	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) {
200
		u32 exthdrlen = icsk->icsk_ext_hdr_len;
L
Linus Torvalds 已提交
201 202 203 204 205 206 207 208 209 210 211
		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];

212
		icsk->icsk_af_ops = &ipv6_mapped;
L
Linus Torvalds 已提交
213
		sk->sk_backlog_rcv = tcp_v4_do_rcv;
214 215 216
#ifdef CONFIG_TCP_MD5SIG
		tp->af_specific = &tcp_sock_ipv6_mapped_specific;
#endif
L
Linus Torvalds 已提交
217 218 219 220

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

		if (err) {
221 222
			icsk->icsk_ext_hdr_len = exthdrlen;
			icsk->icsk_af_ops = &ipv6_specific;
L
Linus Torvalds 已提交
223
			sk->sk_backlog_rcv = tcp_v6_do_rcv;
224 225 226
#ifdef CONFIG_TCP_MD5SIG
			tp->af_specific = &tcp_sock_ipv6_specific;
#endif
L
Linus Torvalds 已提交
227 228
			goto failure;
		} else {
E
Eric Dumazet 已提交
229 230 231
			ipv6_addr_set_v4mapped(inet->inet_saddr, &np->saddr);
			ipv6_addr_set_v4mapped(inet->inet_rcv_saddr,
					       &np->rcv_saddr);
L
Linus Torvalds 已提交
232 233 234 235 236 237 238 239 240 241 242 243 244
		}

		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;
B
Brian Haley 已提交
245
	fl.mark = sk->sk_mark;
L
Linus Torvalds 已提交
246
	fl.fl_ip_dport = usin->sin6_port;
E
Eric Dumazet 已提交
247
	fl.fl_ip_sport = inet->inet_sport;
L
Linus Torvalds 已提交
248 249 250 251 252 253 254 255

	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;
	}

V
Venkat Yekkirala 已提交
256 257
	security_sk_classify_flow(sk, &fl);

L
Linus Torvalds 已提交
258 259 260 261 262 263
	err = ip6_dst_lookup(sk, &dst, &fl);
	if (err)
		goto failure;
	if (final_p)
		ipv6_addr_copy(&fl.fl6_dst, final_p);

A
Alexey Dobriyan 已提交
264 265
	err = __xfrm_lookup(sock_net(sk), &dst, &fl, sk, XFRM_LOOKUP_WAIT);
	if (err < 0) {
266 267 268 269 270
		if (err == -EREMOTE)
			err = ip6_dst_blackhole(sk, &dst, &fl);
		if (err < 0)
			goto failure;
	}
L
Linus Torvalds 已提交
271 272 273 274 275 276 277 278

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

	/* set the source address */
	ipv6_addr_copy(&np->saddr, saddr);
E
Eric Dumazet 已提交
279
	inet->inet_rcv_saddr = LOOPBACK4_IPV6;
L
Linus Torvalds 已提交
280

H
Herbert Xu 已提交
281
	sk->sk_gso_type = SKB_GSO_TCPV6;
282
	__ip6_dst_store(sk, dst, NULL, NULL);
L
Linus Torvalds 已提交
283

284
	icsk->icsk_ext_hdr_len = 0;
L
Linus Torvalds 已提交
285
	if (np->opt)
286 287
		icsk->icsk_ext_hdr_len = (np->opt->opt_flen +
					  np->opt->opt_nflen);
L
Linus Torvalds 已提交
288 289 290

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

E
Eric Dumazet 已提交
291
	inet->inet_dport = usin->sin6_port;
L
Linus Torvalds 已提交
292 293

	tcp_set_state(sk, TCP_SYN_SENT);
294
	err = inet6_hash_connect(&tcp_death_row, sk);
L
Linus Torvalds 已提交
295 296 297 298 299 300
	if (err)
		goto late_failure;

	if (!tp->write_seq)
		tp->write_seq = secure_tcpv6_sequence_number(np->saddr.s6_addr32,
							     np->daddr.s6_addr32,
E
Eric Dumazet 已提交
301 302
							     inet->inet_sport,
							     inet->inet_dport);
L
Linus Torvalds 已提交
303 304 305 306 307 308 309 310 311 312 313

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

	return 0;

late_failure:
	tcp_set_state(sk, TCP_CLOSE);
	__sk_dst_reset(sk);
failure:
E
Eric Dumazet 已提交
314
	inet->inet_dport = 0;
L
Linus Torvalds 已提交
315 316 317 318 319
	sk->sk_route_caps = 0;
	return err;
}

static void tcp_v6_err(struct sk_buff *skb, struct inet6_skb_parm *opt,
320
		u8 type, u8 code, int offset, __be32 info)
L
Linus Torvalds 已提交
321 322
{
	struct ipv6hdr *hdr = (struct ipv6hdr*)skb->data;
323
	const struct tcphdr *th = (struct tcphdr *)(skb->data+offset);
L
Linus Torvalds 已提交
324 325 326
	struct ipv6_pinfo *np;
	struct sock *sk;
	int err;
327
	struct tcp_sock *tp;
L
Linus Torvalds 已提交
328
	__u32 seq;
329
	struct net *net = dev_net(skb->dev);
L
Linus Torvalds 已提交
330

331
	sk = inet6_lookup(net, &tcp_hashinfo, &hdr->daddr,
332
			th->dest, &hdr->saddr, th->source, skb->dev->ifindex);
L
Linus Torvalds 已提交
333 334

	if (sk == NULL) {
335 336
		ICMP6_INC_STATS_BH(net, __in6_dev_get(skb->dev),
				   ICMP6_MIB_INERRORS);
L
Linus Torvalds 已提交
337 338 339 340
		return;
	}

	if (sk->sk_state == TCP_TIME_WAIT) {
341
		inet_twsk_put(inet_twsk(sk));
L
Linus Torvalds 已提交
342 343 344 345 346
		return;
	}

	bh_lock_sock(sk);
	if (sock_owned_by_user(sk))
347
		NET_INC_STATS_BH(net, LINUX_MIB_LOCKDROPPEDICMPS);
L
Linus Torvalds 已提交
348 349 350 351 352

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

	tp = tcp_sk(sk);
353
	seq = ntohl(th->seq);
L
Linus Torvalds 已提交
354 355
	if (sk->sk_state != TCP_LISTEN &&
	    !between(seq, tp->snd_una, tp->snd_nxt)) {
356
		NET_INC_STATS_BH(net, LINUX_MIB_OUTOFWINDOWICMPS);
L
Linus Torvalds 已提交
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
		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;
B
Brian Haley 已提交
386
			fl.mark = sk->sk_mark;
E
Eric Dumazet 已提交
387 388
			fl.fl_ip_dport = inet->inet_dport;
			fl.fl_ip_sport = inet->inet_sport;
V
Venkat Yekkirala 已提交
389
			security_skb_classify_flow(skb, &fl);
L
Linus Torvalds 已提交
390 391 392 393 394 395

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

A
Alexey Dobriyan 已提交
396
			if ((err = xfrm_lookup(net, &dst, &fl, sk, 0)) < 0) {
L
Linus Torvalds 已提交
397 398 399 400 401 402 403
				sk->sk_err_soft = -err;
				goto out;
			}

		} else
			dst_hold(dst);

404
		if (inet_csk(sk)->icsk_pmtu_cookie > dst_mtu(dst)) {
L
Linus Torvalds 已提交
405 406 407 408 409 410 411 412 413
			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);

414
	/* Might be for an request_sock */
L
Linus Torvalds 已提交
415
	switch (sk->sk_state) {
416
		struct request_sock *req, **prev;
L
Linus Torvalds 已提交
417 418 419 420
	case TCP_LISTEN:
		if (sock_owned_by_user(sk))
			goto out;

421 422
		req = inet6_csk_search_req(sk, &prev, th->dest, &hdr->daddr,
					   &hdr->saddr, inet6_iif(skb));
L
Linus Torvalds 已提交
423 424 425 426 427 428
		if (!req)
			goto out;

		/* ICMPs are not backlogged, hence we cannot get
		 * an established socket here.
		 */
429
		WARN_ON(req->sk != NULL);
L
Linus Torvalds 已提交
430

431
		if (seq != tcp_rsk(req)->snt_isn) {
432
			NET_INC_STATS_BH(net, LINUX_MIB_OUTOFWINDOWICMPS);
L
Linus Torvalds 已提交
433 434 435
			goto out;
		}

436
		inet_csk_reqsk_queue_drop(sk, req, prev);
L
Linus Torvalds 已提交
437 438 439 440
		goto out;

	case TCP_SYN_SENT:
	case TCP_SYN_RECV:  /* Cannot happen.
441
			       It can, it SYNs are crossed. --ANK */
L
Linus Torvalds 已提交
442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463
		if (!sock_owned_by_user(sk)) {
			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);
}


464 465
static int tcp_v6_send_synack(struct sock *sk, struct request_sock *req,
			      struct request_values *rvp)
L
Linus Torvalds 已提交
466
{
467
	struct inet6_request_sock *treq = inet6_rsk(req);
L
Linus Torvalds 已提交
468 469 470 471 472
	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;
473
	struct dst_entry *dst;
L
Linus Torvalds 已提交
474 475 476 477
	int err = -1;

	memset(&fl, 0, sizeof(fl));
	fl.proto = IPPROTO_TCP;
478 479
	ipv6_addr_copy(&fl.fl6_dst, &treq->rmt_addr);
	ipv6_addr_copy(&fl.fl6_src, &treq->loc_addr);
L
Linus Torvalds 已提交
480
	fl.fl6_flowlabel = 0;
481
	fl.oif = treq->iif;
B
Brian Haley 已提交
482
	fl.mark = sk->sk_mark;
483
	fl.fl_ip_dport = inet_rsk(req)->rmt_port;
484
	fl.fl_ip_sport = inet_rsk(req)->loc_port;
485
	security_req_classify_flow(req, &fl);
L
Linus Torvalds 已提交
486

487 488 489 490 491 492
	opt = np->opt;
	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;
L
Linus Torvalds 已提交
493 494
	}

495 496 497 498 499
	err = ip6_dst_lookup(sk, &dst, &fl);
	if (err)
		goto done;
	if (final_p)
		ipv6_addr_copy(&fl.fl6_dst, final_p);
A
Alexey Dobriyan 已提交
500
	if ((err = xfrm_lookup(sock_net(sk), &dst, &fl, sk, 0)) < 0)
501 502
		goto done;

503
	skb = tcp_make_synack(sk, dst, req, rvp);
L
Linus Torvalds 已提交
504
	if (skb) {
505
		struct tcphdr *th = tcp_hdr(skb);
L
Linus Torvalds 已提交
506

H
Herbert Xu 已提交
507
		th->check = tcp_v6_check(skb->len,
508
					 &treq->loc_addr, &treq->rmt_addr,
509
					 csum_partial(th, skb->len, skb->csum));
L
Linus Torvalds 已提交
510

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

done:
517
	if (opt && opt != np->opt)
L
Linus Torvalds 已提交
518
		sock_kfree_s(sk, opt, opt->tot_len);
519
	dst_release(dst);
L
Linus Torvalds 已提交
520 521 522
	return err;
}

523 524 525 526 527 528 529 530 531 532 533 534 535 536
static inline void syn_flood_warning(struct sk_buff *skb)
{
#ifdef CONFIG_SYN_COOKIES
	if (sysctl_tcp_syncookies)
		printk(KERN_INFO
		       "TCPv6: Possible SYN flooding on port %d. "
		       "Sending cookies.\n", ntohs(tcp_hdr(skb)->dest));
	else
#endif
		printk(KERN_INFO
		       "TCPv6: Possible SYN flooding on port %d. "
		       "Dropping request.\n", ntohs(tcp_hdr(skb)->dest));
}

537
static void tcp_v6_reqsk_destructor(struct request_sock *req)
L
Linus Torvalds 已提交
538
{
539
	kfree_skb(inet6_rsk(req)->pktopts);
L
Linus Torvalds 已提交
540 541
}

542 543 544 545 546 547 548 549 550 551 552 553 554
#ifdef CONFIG_TCP_MD5SIG
static struct tcp_md5sig_key *tcp_v6_md5_do_lookup(struct sock *sk,
						   struct in6_addr *addr)
{
	struct tcp_sock *tp = tcp_sk(sk);
	int i;

	BUG_ON(tp == NULL);

	if (!tp->md5sig_info || !tp->md5sig_info->entries6)
		return NULL;

	for (i = 0; i < tp->md5sig_info->entries6; i++) {
555
		if (ipv6_addr_equal(&tp->md5sig_info->keys6[i].addr, addr))
556
			return &tp->md5sig_info->keys6[i].base;
557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576
	}
	return NULL;
}

static struct tcp_md5sig_key *tcp_v6_md5_lookup(struct sock *sk,
						struct sock *addr_sk)
{
	return tcp_v6_md5_do_lookup(sk, &inet6_sk(addr_sk)->daddr);
}

static struct tcp_md5sig_key *tcp_v6_reqsk_md5_lookup(struct sock *sk,
						      struct request_sock *req)
{
	return tcp_v6_md5_do_lookup(sk, &inet6_rsk(req)->rmt_addr);
}

static int tcp_v6_md5_do_add(struct sock *sk, struct in6_addr *peer,
			     char *newkey, u8 newkeylen)
{
	/* Add key to the list */
577
	struct tcp_md5sig_key *key;
578 579 580
	struct tcp_sock *tp = tcp_sk(sk);
	struct tcp6_md5sig_key *keys;

581
	key = tcp_v6_md5_do_lookup(sk, peer);
582 583
	if (key) {
		/* modify existing entry - just update that one */
584 585 586
		kfree(key->key);
		key->key = newkey;
		key->keylen = newkeylen;
587 588 589 590 591 592 593 594
	} else {
		/* reallocate new list if current one is full. */
		if (!tp->md5sig_info) {
			tp->md5sig_info = kzalloc(sizeof(*tp->md5sig_info), GFP_ATOMIC);
			if (!tp->md5sig_info) {
				kfree(newkey);
				return -ENOMEM;
			}
595
			sk->sk_route_caps &= ~NETIF_F_GSO_MASK;
596
		}
597
		if (tcp_alloc_md5sig_pool(sk) == NULL) {
598 599 600
			kfree(newkey);
			return -ENOMEM;
		}
601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622
		if (tp->md5sig_info->alloced6 == tp->md5sig_info->entries6) {
			keys = kmalloc((sizeof (tp->md5sig_info->keys6[0]) *
				       (tp->md5sig_info->entries6 + 1)), GFP_ATOMIC);

			if (!keys) {
				tcp_free_md5sig_pool();
				kfree(newkey);
				return -ENOMEM;
			}

			if (tp->md5sig_info->entries6)
				memmove(keys, tp->md5sig_info->keys6,
					(sizeof (tp->md5sig_info->keys6[0]) *
					 tp->md5sig_info->entries6));

			kfree(tp->md5sig_info->keys6);
			tp->md5sig_info->keys6 = keys;
			tp->md5sig_info->alloced6++;
		}

		ipv6_addr_copy(&tp->md5sig_info->keys6[tp->md5sig_info->entries6].addr,
			       peer);
623 624
		tp->md5sig_info->keys6[tp->md5sig_info->entries6].base.key = newkey;
		tp->md5sig_info->keys6[tp->md5sig_info->entries6].base.keylen = newkeylen;
625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643

		tp->md5sig_info->entries6++;
	}
	return 0;
}

static int tcp_v6_md5_add_func(struct sock *sk, struct sock *addr_sk,
			       u8 *newkey, __u8 newkeylen)
{
	return tcp_v6_md5_do_add(sk, &inet6_sk(addr_sk)->daddr,
				 newkey, newkeylen);
}

static int tcp_v6_md5_do_del(struct sock *sk, struct in6_addr *peer)
{
	struct tcp_sock *tp = tcp_sk(sk);
	int i;

	for (i = 0; i < tp->md5sig_info->entries6; i++) {
644
		if (ipv6_addr_equal(&tp->md5sig_info->keys6[i].addr, peer)) {
645
			/* Free the key */
646
			kfree(tp->md5sig_info->keys6[i].base.key);
647 648 649 650 651
			tp->md5sig_info->entries6--;

			if (tp->md5sig_info->entries6 == 0) {
				kfree(tp->md5sig_info->keys6);
				tp->md5sig_info->keys6 = NULL;
652
				tp->md5sig_info->alloced6 = 0;
653 654 655 656 657 658 659 660
			} else {
				/* shrink the database */
				if (tp->md5sig_info->entries6 != i)
					memmove(&tp->md5sig_info->keys6[i],
						&tp->md5sig_info->keys6[i+1],
						(tp->md5sig_info->entries6 - i)
						* sizeof (tp->md5sig_info->keys6[0]));
			}
661 662
			tcp_free_md5sig_pool();
			return 0;
663 664 665 666 667 668 669 670 671 672 673 674
		}
	}
	return -ENOENT;
}

static void tcp_v6_clear_md5_list (struct sock *sk)
{
	struct tcp_sock *tp = tcp_sk(sk);
	int i;

	if (tp->md5sig_info->entries6) {
		for (i = 0; i < tp->md5sig_info->entries6; i++)
675
			kfree(tp->md5sig_info->keys6[i].base.key);
676 677 678 679 680 681 682 683 684 685
		tp->md5sig_info->entries6 = 0;
		tcp_free_md5sig_pool();
	}

	kfree(tp->md5sig_info->keys6);
	tp->md5sig_info->keys6 = NULL;
	tp->md5sig_info->alloced6 = 0;

	if (tp->md5sig_info->entries4) {
		for (i = 0; i < tp->md5sig_info->entries4; i++)
686
			kfree(tp->md5sig_info->keys4[i].base.key);
687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714
		tp->md5sig_info->entries4 = 0;
		tcp_free_md5sig_pool();
	}

	kfree(tp->md5sig_info->keys4);
	tp->md5sig_info->keys4 = NULL;
	tp->md5sig_info->alloced4 = 0;
}

static int tcp_v6_parse_md5_keys (struct sock *sk, char __user *optval,
				  int optlen)
{
	struct tcp_md5sig cmd;
	struct sockaddr_in6 *sin6 = (struct sockaddr_in6 *)&cmd.tcpm_addr;
	u8 *newkey;

	if (optlen < sizeof(cmd))
		return -EINVAL;

	if (copy_from_user(&cmd, optval, sizeof(cmd)))
		return -EFAULT;

	if (sin6->sin6_family != AF_INET6)
		return -EINVAL;

	if (!cmd.tcpm_keylen) {
		if (!tcp_sk(sk)->md5sig_info)
			return -ENOENT;
B
Brian Haley 已提交
715
		if (ipv6_addr_v4mapped(&sin6->sin6_addr))
716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731
			return tcp_v4_md5_do_del(sk, sin6->sin6_addr.s6_addr32[3]);
		return tcp_v6_md5_do_del(sk, &sin6->sin6_addr);
	}

	if (cmd.tcpm_keylen > TCP_MD5SIG_MAXKEYLEN)
		return -EINVAL;

	if (!tcp_sk(sk)->md5sig_info) {
		struct tcp_sock *tp = tcp_sk(sk);
		struct tcp_md5sig_info *p;

		p = kzalloc(sizeof(struct tcp_md5sig_info), GFP_KERNEL);
		if (!p)
			return -ENOMEM;

		tp->md5sig_info = p;
732
		sk->sk_route_caps &= ~NETIF_F_GSO_MASK;
733 734
	}

A
Arnaldo Carvalho de Melo 已提交
735
	newkey = kmemdup(cmd.tcpm_key, cmd.tcpm_keylen, GFP_KERNEL);
736 737
	if (!newkey)
		return -ENOMEM;
B
Brian Haley 已提交
738
	if (ipv6_addr_v4mapped(&sin6->sin6_addr)) {
739 740 741 742 743 744
		return tcp_v4_md5_do_add(sk, sin6->sin6_addr.s6_addr32[3],
					 newkey, cmd.tcpm_keylen);
	}
	return tcp_v6_md5_do_add(sk, &sin6->sin6_addr, newkey, cmd.tcpm_keylen);
}

745 746 747
static int tcp_v6_md5_hash_pseudoheader(struct tcp_md5sig_pool *hp,
					struct in6_addr *daddr,
					struct in6_addr *saddr, int nbytes)
748 749
{
	struct tcp6_pseudohdr *bp;
750
	struct scatterlist sg;
751

752 753 754 755
	bp = &hp->md5_blk.ip6;
	/* 1. TCP pseudo-header (RFC2460) */
	ipv6_addr_copy(&bp->saddr, saddr);
	ipv6_addr_copy(&bp->daddr, daddr);
756
	bp->protocol = cpu_to_be32(IPPROTO_TCP);
A
Adam Langley 已提交
757
	bp->len = cpu_to_be32(nbytes);
758

759 760 761
	sg_init_one(&sg, bp, sizeof(*bp));
	return crypto_hash_update(&hp->md5_desc, &sg, sizeof(*bp));
}
762

763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783
static int tcp_v6_md5_hash_hdr(char *md5_hash, struct tcp_md5sig_key *key,
			       struct in6_addr *daddr, struct in6_addr *saddr,
			       struct tcphdr *th)
{
	struct tcp_md5sig_pool *hp;
	struct hash_desc *desc;

	hp = tcp_get_md5sig_pool();
	if (!hp)
		goto clear_hash_noput;
	desc = &hp->md5_desc;

	if (crypto_hash_init(desc))
		goto clear_hash;
	if (tcp_v6_md5_hash_pseudoheader(hp, daddr, saddr, th->doff << 2))
		goto clear_hash;
	if (tcp_md5_hash_header(hp, th))
		goto clear_hash;
	if (tcp_md5_hash_key(hp, key))
		goto clear_hash;
	if (crypto_hash_final(desc, md5_hash))
784 785 786 787
		goto clear_hash;

	tcp_put_md5sig_pool();
	return 0;
788

789 790 791 792
clear_hash:
	tcp_put_md5sig_pool();
clear_hash_noput:
	memset(md5_hash, 0, 16);
793
	return 1;
794 795
}

796 797 798
static int tcp_v6_md5_hash_skb(char *md5_hash, struct tcp_md5sig_key *key,
			       struct sock *sk, struct request_sock *req,
			       struct sk_buff *skb)
799 800
{
	struct in6_addr *saddr, *daddr;
801 802 803
	struct tcp_md5sig_pool *hp;
	struct hash_desc *desc;
	struct tcphdr *th = tcp_hdr(skb);
804 805 806 807

	if (sk) {
		saddr = &inet6_sk(sk)->saddr;
		daddr = &inet6_sk(sk)->daddr;
808
	} else if (req) {
809 810
		saddr = &inet6_rsk(req)->loc_addr;
		daddr = &inet6_rsk(req)->rmt_addr;
811 812 813 814
	} else {
		struct ipv6hdr *ip6h = ipv6_hdr(skb);
		saddr = &ip6h->saddr;
		daddr = &ip6h->daddr;
815
	}
816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843

	hp = tcp_get_md5sig_pool();
	if (!hp)
		goto clear_hash_noput;
	desc = &hp->md5_desc;

	if (crypto_hash_init(desc))
		goto clear_hash;

	if (tcp_v6_md5_hash_pseudoheader(hp, daddr, saddr, skb->len))
		goto clear_hash;
	if (tcp_md5_hash_header(hp, th))
		goto clear_hash;
	if (tcp_md5_hash_skb_data(hp, skb, th->doff << 2))
		goto clear_hash;
	if (tcp_md5_hash_key(hp, key))
		goto clear_hash;
	if (crypto_hash_final(desc, md5_hash))
		goto clear_hash;

	tcp_put_md5sig_pool();
	return 0;

clear_hash:
	tcp_put_md5sig_pool();
clear_hash_noput:
	memset(md5_hash, 0, 16);
	return 1;
844 845 846 847 848 849
}

static int tcp_v6_inbound_md5_hash (struct sock *sk, struct sk_buff *skb)
{
	__u8 *hash_location = NULL;
	struct tcp_md5sig_key *hash_expected;
850
	struct ipv6hdr *ip6h = ipv6_hdr(skb);
851
	struct tcphdr *th = tcp_hdr(skb);
852 853 854 855
	int genhash;
	u8 newhash[16];

	hash_expected = tcp_v6_md5_do_lookup(sk, &ip6h->saddr);
856
	hash_location = tcp_parse_md5sig_option(th);
857

858 859 860 861 862 863
	/* We've parsed the options - do we have a hash? */
	if (!hash_expected && !hash_location)
		return 0;

	if (hash_expected && !hash_location) {
		NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_TCPMD5NOTFOUND);
864 865 866
		return 1;
	}

867 868
	if (!hash_expected && hash_location) {
		NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_TCPMD5UNEXPECTED);
869 870 871 872
		return 1;
	}

	/* check the signature */
873 874 875 876
	genhash = tcp_v6_md5_hash_skb(newhash,
				      hash_expected,
				      NULL, NULL, skb);

877 878
	if (genhash || memcmp(hash_location, newhash, 16) != 0) {
		if (net_ratelimit()) {
879
			printk(KERN_INFO "MD5 Hash %s for [%pI6c]:%u->[%pI6c]:%u\n",
880
			       genhash ? "failed" : "mismatch",
881 882
			       &ip6h->saddr, ntohs(th->source),
			       &ip6h->daddr, ntohs(th->dest));
883 884 885 886 887 888 889
		}
		return 1;
	}
	return 0;
}
#endif

890
struct request_sock_ops tcp6_request_sock_ops __read_mostly = {
L
Linus Torvalds 已提交
891
	.family		=	AF_INET6,
892
	.obj_size	=	sizeof(struct tcp6_request_sock),
L
Linus Torvalds 已提交
893
	.rtx_syn_ack	=	tcp_v6_send_synack,
894 895
	.send_ack	=	tcp_v6_reqsk_send_ack,
	.destructor	=	tcp_v6_reqsk_destructor,
L
Linus Torvalds 已提交
896 897 898
	.send_reset	=	tcp_v6_send_reset
};

899
#ifdef CONFIG_TCP_MD5SIG
900
static const struct tcp_request_sock_ops tcp_request_sock_ipv6_ops = {
901
	.md5_lookup	=	tcp_v6_reqsk_md5_lookup,
902
	.calc_md5_hash	=	tcp_v6_md5_hash_skb,
903
};
904
#endif
905

906 907 908
static struct timewait_sock_ops tcp6_timewait_sock_ops = {
	.twsk_obj_size	= sizeof(struct tcp6_timewait_sock),
	.twsk_unique	= tcp_twsk_unique,
909
	.twsk_destructor= tcp_twsk_destructor,
910 911
};

912
static void tcp_v6_send_check(struct sock *sk, int len, struct sk_buff *skb)
L
Linus Torvalds 已提交
913 914
{
	struct ipv6_pinfo *np = inet6_sk(sk);
915
	struct tcphdr *th = tcp_hdr(skb);
L
Linus Torvalds 已提交
916

917
	if (skb->ip_summed == CHECKSUM_PARTIAL) {
L
Linus Torvalds 已提交
918
		th->check = ~csum_ipv6_magic(&np->saddr, &np->daddr, len, IPPROTO_TCP,  0);
919
		skb->csum_start = skb_transport_header(skb) - skb->head;
A
Al Viro 已提交
920
		skb->csum_offset = offsetof(struct tcphdr, check);
L
Linus Torvalds 已提交
921
	} else {
922
		th->check = csum_ipv6_magic(&np->saddr, &np->daddr, len, IPPROTO_TCP,
923
					    csum_partial(th, th->doff<<2,
L
Linus Torvalds 已提交
924 925 926 927
							 skb->csum));
	}
}

928 929 930 931 932 933 934 935
static int tcp_v6_gso_send_check(struct sk_buff *skb)
{
	struct ipv6hdr *ipv6h;
	struct tcphdr *th;

	if (!pskb_may_pull(skb, sizeof(*th)))
		return -EINVAL;

936
	ipv6h = ipv6_hdr(skb);
937
	th = tcp_hdr(skb);
938 939 940 941

	th->check = 0;
	th->check = ~csum_ipv6_magic(&ipv6h->saddr, &ipv6h->daddr, skb->len,
				     IPPROTO_TCP, 0);
942
	skb->csum_start = skb_transport_header(skb) - skb->head;
A
Al Viro 已提交
943
	skb->csum_offset = offsetof(struct tcphdr, check);
944
	skb->ip_summed = CHECKSUM_PARTIAL;
945 946
	return 0;
}
L
Linus Torvalds 已提交
947

H
Herbert Xu 已提交
948 949
static struct sk_buff **tcp6_gro_receive(struct sk_buff **head,
					 struct sk_buff *skb)
H
Herbert Xu 已提交
950
{
H
Herbert Xu 已提交
951
	struct ipv6hdr *iph = skb_gro_network_header(skb);
H
Herbert Xu 已提交
952 953 954

	switch (skb->ip_summed) {
	case CHECKSUM_COMPLETE:
955
		if (!tcp_v6_check(skb_gro_len(skb), &iph->saddr, &iph->daddr,
H
Herbert Xu 已提交
956 957 958 959 960 961 962 963 964 965 966 967 968 969
				  skb->csum)) {
			skb->ip_summed = CHECKSUM_UNNECESSARY;
			break;
		}

		/* fall through */
	case CHECKSUM_NONE:
		NAPI_GRO_CB(skb)->flush = 1;
		return NULL;
	}

	return tcp_gro_receive(head, skb);
}

H
Herbert Xu 已提交
970
static int tcp6_gro_complete(struct sk_buff *skb)
H
Herbert Xu 已提交
971 972 973 974 975 976 977 978 979 980 981
{
	struct ipv6hdr *iph = ipv6_hdr(skb);
	struct tcphdr *th = tcp_hdr(skb);

	th->check = ~tcp_v6_check(skb->len - skb_transport_offset(skb),
				  &iph->saddr, &iph->daddr, 0);
	skb_shinfo(skb)->gso_type = SKB_GSO_TCPV6;

	return tcp_gro_complete(skb);
}

982 983
static void tcp_v6_send_response(struct sk_buff *skb, u32 seq, u32 ack, u32 win,
				 u32 ts, struct tcp_md5sig_key *key, int rst)
L
Linus Torvalds 已提交
984
{
985
	struct tcphdr *th = tcp_hdr(skb), *t1;
L
Linus Torvalds 已提交
986 987
	struct sk_buff *buff;
	struct flowi fl;
E
Eric Dumazet 已提交
988
	struct net *net = dev_net(skb_dst(skb)->dev);
989
	struct sock *ctl_sk = net->ipv6.tcp_sk;
990
	unsigned int tot_len = sizeof(struct tcphdr);
E
Eric Dumazet 已提交
991
	struct dst_entry *dst;
992
	__be32 *topt;
L
Linus Torvalds 已提交
993

994 995
	if (ts)
		tot_len += TCPOLEN_TSTAMP_ALIGNED;
996 997 998 999 1000 1001
#ifdef CONFIG_TCP_MD5SIG
	if (key)
		tot_len += TCPOLEN_MD5SIG_ALIGNED;
#endif

	buff = alloc_skb(MAX_HEADER + sizeof(struct ipv6hdr) + tot_len,
L
Linus Torvalds 已提交
1002
			 GFP_ATOMIC);
1003 1004
	if (buff == NULL)
		return;
L
Linus Torvalds 已提交
1005

1006
	skb_reserve(buff, MAX_HEADER + sizeof(struct ipv6hdr) + tot_len);
L
Linus Torvalds 已提交
1007

1008
	t1 = (struct tcphdr *) skb_push(buff, tot_len);
1009
	skb_reset_transport_header(skb);
L
Linus Torvalds 已提交
1010 1011 1012 1013 1014

	/* Swap the send and the receive. */
	memset(t1, 0, sizeof(*t1));
	t1->dest = th->source;
	t1->source = th->dest;
1015
	t1->doff = tot_len / 4;
1016 1017 1018 1019 1020
	t1->seq = htonl(seq);
	t1->ack_seq = htonl(ack);
	t1->ack = !rst || !th->ack;
	t1->rst = rst;
	t1->window = htons(win);
L
Linus Torvalds 已提交
1021

1022 1023
	topt = (__be32 *)(t1 + 1);

1024 1025 1026 1027 1028 1029 1030
	if (ts) {
		*topt++ = htonl((TCPOPT_NOP << 24) | (TCPOPT_NOP << 16) |
				(TCPOPT_TIMESTAMP << 8) | TCPOLEN_TIMESTAMP);
		*topt++ = htonl(tcp_time_stamp);
		*topt++ = htonl(ts);
	}

1031 1032
#ifdef CONFIG_TCP_MD5SIG
	if (key) {
1033 1034 1035
		*topt++ = htonl((TCPOPT_NOP << 24) | (TCPOPT_NOP << 16) |
				(TCPOPT_MD5SIG << 8) | TCPOLEN_MD5SIG);
		tcp_v6_md5_hash_hdr((__u8 *)topt, key,
1036 1037
				    &ipv6_hdr(skb)->saddr,
				    &ipv6_hdr(skb)->daddr, t1);
1038 1039 1040
	}
#endif

1041
	buff->csum = csum_partial(t1, tot_len, 0);
L
Linus Torvalds 已提交
1042 1043

	memset(&fl, 0, sizeof(fl));
1044 1045
	ipv6_addr_copy(&fl.fl6_dst, &ipv6_hdr(skb)->saddr);
	ipv6_addr_copy(&fl.fl6_src, &ipv6_hdr(skb)->daddr);
L
Linus Torvalds 已提交
1046 1047

	t1->check = csum_ipv6_magic(&fl.fl6_src, &fl.fl6_dst,
1048
				    tot_len, IPPROTO_TCP,
L
Linus Torvalds 已提交
1049 1050 1051
				    buff->csum);

	fl.proto = IPPROTO_TCP;
1052
	fl.oif = inet6_iif(skb);
L
Linus Torvalds 已提交
1053 1054
	fl.fl_ip_dport = t1->dest;
	fl.fl_ip_sport = t1->source;
V
Venkat Yekkirala 已提交
1055
	security_skb_classify_flow(skb, &fl);
L
Linus Torvalds 已提交
1056

1057 1058 1059 1060
	/* Pass a socket to ip6_dst_lookup either it is for RST
	 * Underlying function will use this to retrieve the network
	 * namespace
	 */
E
Eric Dumazet 已提交
1061 1062 1063
	if (!ip6_dst_lookup(ctl_sk, &dst, &fl)) {
		if (xfrm_lookup(net, &dst, &fl, NULL, 0) >= 0) {
			skb_dst_set(buff, dst);
1064
			ip6_xmit(ctl_sk, buff, &fl, NULL, 0);
1065
			TCP_INC_STATS_BH(net, TCP_MIB_OUTSEGS);
1066 1067
			if (rst)
				TCP_INC_STATS_BH(net, TCP_MIB_OUTRSTS);
L
Linus Torvalds 已提交
1068
			return;
A
Arnaldo Carvalho de Melo 已提交
1069
		}
L
Linus Torvalds 已提交
1070 1071 1072 1073 1074
	}

	kfree_skb(buff);
}

1075
static void tcp_v6_send_reset(struct sock *sk, struct sk_buff *skb)
L
Linus Torvalds 已提交
1076
{
1077 1078
	struct tcphdr *th = tcp_hdr(skb);
	u32 seq = 0, ack_seq = 0;
1079
	struct tcp_md5sig_key *key = NULL;
L
Linus Torvalds 已提交
1080

1081
	if (th->rst)
L
Linus Torvalds 已提交
1082 1083
		return;

1084 1085
	if (!ipv6_unicast_destination(skb))
		return;
L
Linus Torvalds 已提交
1086

1087
#ifdef CONFIG_TCP_MD5SIG
1088 1089
	if (sk)
		key = tcp_v6_md5_do_lookup(sk, &ipv6_hdr(skb)->daddr);
1090 1091
#endif

1092 1093 1094 1095 1096
	if (th->ack)
		seq = ntohl(th->ack_seq);
	else
		ack_seq = ntohl(th->seq) + th->syn + th->fin + skb->len -
			  (th->doff << 2);
L
Linus Torvalds 已提交
1097

1098 1099
	tcp_v6_send_response(skb, seq, ack_seq, 0, 0, key, 1);
}
L
Linus Torvalds 已提交
1100

1101 1102 1103 1104
static void tcp_v6_send_ack(struct sk_buff *skb, u32 seq, u32 ack, u32 win, u32 ts,
			    struct tcp_md5sig_key *key)
{
	tcp_v6_send_response(skb, seq, ack, win, ts, key, 0);
L
Linus Torvalds 已提交
1105 1106 1107 1108
}

static void tcp_v6_timewait_ack(struct sock *sk, struct sk_buff *skb)
{
1109
	struct inet_timewait_sock *tw = inet_twsk(sk);
1110
	struct tcp_timewait_sock *tcptw = tcp_twsk(sk);
L
Linus Torvalds 已提交
1111

1112
	tcp_v6_send_ack(skb, tcptw->tw_snd_nxt, tcptw->tw_rcv_nxt,
1113
			tcptw->tw_rcv_wnd >> tw->tw_rcv_wscale,
1114
			tcptw->tw_ts_recent, tcp_twsk_md5_key(tcptw));
L
Linus Torvalds 已提交
1115

1116
	inet_twsk_put(tw);
L
Linus Torvalds 已提交
1117 1118
}

1119 1120
static void tcp_v6_reqsk_send_ack(struct sock *sk, struct sk_buff *skb,
				  struct request_sock *req)
L
Linus Torvalds 已提交
1121
{
1122
	tcp_v6_send_ack(skb, tcp_rsk(req)->snt_isn + 1, tcp_rsk(req)->rcv_isn + 1, req->rcv_wnd, req->ts_recent,
1123
			tcp_v6_md5_do_lookup(sk, &ipv6_hdr(skb)->daddr));
L
Linus Torvalds 已提交
1124 1125 1126 1127 1128
}


static struct sock *tcp_v6_hnd_req(struct sock *sk,struct sk_buff *skb)
{
1129
	struct request_sock *req, **prev;
1130
	const struct tcphdr *th = tcp_hdr(skb);
L
Linus Torvalds 已提交
1131 1132 1133
	struct sock *nsk;

	/* Find possible connection requests. */
1134
	req = inet6_csk_search_req(sk, &prev, th->source,
1135 1136
				   &ipv6_hdr(skb)->saddr,
				   &ipv6_hdr(skb)->daddr, inet6_iif(skb));
L
Linus Torvalds 已提交
1137 1138 1139
	if (req)
		return tcp_check_req(sk, skb, req, prev);

1140
	nsk = __inet6_lookup_established(sock_net(sk), &tcp_hashinfo,
1141 1142
			&ipv6_hdr(skb)->saddr, th->source,
			&ipv6_hdr(skb)->daddr, ntohs(th->dest), inet6_iif(skb));
L
Linus Torvalds 已提交
1143 1144 1145 1146 1147 1148

	if (nsk) {
		if (nsk->sk_state != TCP_TIME_WAIT) {
			bh_lock_sock(nsk);
			return nsk;
		}
1149
		inet_twsk_put(inet_twsk(nsk));
L
Linus Torvalds 已提交
1150 1151 1152
		return NULL;
	}

1153
#ifdef CONFIG_SYN_COOKIES
L
Linus Torvalds 已提交
1154
	if (!th->rst && !th->syn && th->ack)
1155
		sk = cookie_v6_check(sk, skb);
L
Linus Torvalds 已提交
1156 1157 1158 1159 1160 1161 1162 1163 1164
#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)
{
1165
	struct tcp_extend_values tmp_ext;
1166
	struct tcp_options_received tmp_opt;
1167
	u8 *hash_location;
1168
	struct request_sock *req;
1169
	struct inet6_request_sock *treq;
L
Linus Torvalds 已提交
1170 1171
	struct ipv6_pinfo *np = inet6_sk(sk);
	struct tcp_sock *tp = tcp_sk(sk);
1172
	__u32 isn = TCP_SKB_CB(skb)->when;
1173 1174 1175 1176 1177
#ifdef CONFIG_SYN_COOKIES
	int want_cookie = 0;
#else
#define want_cookie 0
#endif
L
Linus Torvalds 已提交
1178 1179 1180 1181 1182

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

	if (!ipv6_unicast_destination(skb))
1183
		goto drop;
L
Linus Torvalds 已提交
1184

1185
	if (inet_csk_reqsk_queue_is_full(sk) && !isn) {
L
Linus Torvalds 已提交
1186
		if (net_ratelimit())
1187 1188 1189 1190 1191 1192
			syn_flood_warning(skb);
#ifdef CONFIG_SYN_COOKIES
		if (sysctl_tcp_syncookies)
			want_cookie = 1;
		else
#endif
1193
		goto drop;
L
Linus Torvalds 已提交
1194 1195
	}

1196
	if (sk_acceptq_is_full(sk) && inet_csk_reqsk_queue_young(sk) > 1)
L
Linus Torvalds 已提交
1197 1198
		goto drop;

1199
	req = inet6_reqsk_alloc(&tcp6_request_sock_ops);
L
Linus Torvalds 已提交
1200 1201 1202
	if (req == NULL)
		goto drop;

1203 1204 1205 1206
#ifdef CONFIG_TCP_MD5SIG
	tcp_rsk(req)->af_specific = &tcp_request_sock_ipv6_ops;
#endif

L
Linus Torvalds 已提交
1207 1208 1209
	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;
1210
	tcp_parse_options(skb, &tmp_opt, &hash_location, 0);
1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241

	if (tmp_opt.cookie_plus > 0 &&
	    tmp_opt.saw_tstamp &&
	    !tp->rx_opt.cookie_out_never &&
	    (sysctl_tcp_cookie_size > 0 ||
	     (tp->cookie_values != NULL &&
	      tp->cookie_values->cookie_desired > 0))) {
		u8 *c;
		u32 *d;
		u32 *mess = &tmp_ext.cookie_bakery[COOKIE_DIGEST_WORDS];
		int l = tmp_opt.cookie_plus - TCPOLEN_COOKIE_BASE;

		if (tcp_cookie_generator(&tmp_ext.cookie_bakery[0]) != 0)
			goto drop_and_free;

		/* Secret recipe starts with IP addresses */
		d = &ipv6_hdr(skb)->daddr.s6_addr32[0];
		*mess++ ^= *d++;
		*mess++ ^= *d++;
		*mess++ ^= *d++;
		*mess++ ^= *d++;
		d = &ipv6_hdr(skb)->saddr.s6_addr32[0];
		*mess++ ^= *d++;
		*mess++ ^= *d++;
		*mess++ ^= *d++;
		*mess++ ^= *d++;

		/* plus variable length Initiator Cookie */
		c = (u8 *)mess;
		while (l-- > 0)
			*c++ ^= *hash_location++;
L
Linus Torvalds 已提交
1242

1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255
#ifdef CONFIG_SYN_COOKIES
		want_cookie = 0;	/* not our kind of cookie */
#endif
		tmp_ext.cookie_out_never = 0; /* false */
		tmp_ext.cookie_plus = tmp_opt.cookie_plus;
	} else if (!tp->rx_opt.cookie_in_always) {
		/* redundant indications, but ensure initialization. */
		tmp_ext.cookie_out_never = 1; /* true */
		tmp_ext.cookie_plus = 0;
	} else {
		goto drop_and_free;
	}
	tmp_ext.cookie_in_always = tp->rx_opt.cookie_in_always;
L
Linus Torvalds 已提交
1256

1257
	if (want_cookie && !tmp_opt.saw_tstamp)
1258 1259
		tcp_clear_options(&tmp_opt);

L
Linus Torvalds 已提交
1260 1261 1262
	tmp_opt.tstamp_ok = tmp_opt.saw_tstamp;
	tcp_openreq_init(req, &tmp_opt, skb);

1263
	treq = inet6_rsk(req);
1264 1265
	ipv6_addr_copy(&treq->rmt_addr, &ipv6_hdr(skb)->saddr);
	ipv6_addr_copy(&treq->loc_addr, &ipv6_hdr(skb)->daddr);
1266 1267 1268 1269 1270
	if (!want_cookie)
		TCP_ECN_create_request(req, tcp_hdr(skb));

	if (want_cookie) {
		isn = cookie_v6_init_sequence(sk, skb, &req->mss);
1271
		req->cookie_ts = tmp_opt.tstamp_ok;
1272 1273 1274 1275 1276 1277 1278 1279
	} else if (!isn) {
		if (ipv6_opt_accepted(sk, skb) ||
		    np->rxopt.bits.rxinfo || np->rxopt.bits.rxoinfo ||
		    np->rxopt.bits.rxhlim || np->rxopt.bits.rxohlim) {
			atomic_inc(&skb->users);
			treq->pktopts = skb;
		}
		treq->iif = sk->sk_bound_dev_if;
L
Linus Torvalds 已提交
1280

1281 1282 1283 1284
		/* So that link locals have meaning */
		if (!sk->sk_bound_dev_if &&
		    ipv6_addr_type(&treq->rmt_addr) & IPV6_ADDR_LINKLOCAL)
			treq->iif = inet6_iif(skb);
L
Linus Torvalds 已提交
1285

1286
		isn = tcp_v6_init_sequence(skb);
1287
	}
1288
	tcp_rsk(req)->snt_isn = isn;
L
Linus Torvalds 已提交
1289

1290 1291
	security_inet_conn_request(sk, skb, req);

1292 1293 1294
	if (tcp_v6_send_synack(sk, req,
			       (struct request_values *)&tmp_ext) ||
	    want_cookie)
1295
		goto drop_and_free;
L
Linus Torvalds 已提交
1296

1297 1298
	inet6_csk_reqsk_queue_hash_add(sk, req, TCP_TIMEOUT_INIT);
	return 0;
L
Linus Torvalds 已提交
1299

1300 1301
drop_and_free:
	reqsk_free(req);
L
Linus Torvalds 已提交
1302 1303 1304 1305 1306
drop:
	return 0; /* don't send reset */
}

static struct sock * tcp_v6_syn_recv_sock(struct sock *sk, struct sk_buff *skb,
1307
					  struct request_sock *req,
L
Linus Torvalds 已提交
1308 1309
					  struct dst_entry *dst)
{
1310
	struct inet6_request_sock *treq;
L
Linus Torvalds 已提交
1311 1312 1313 1314 1315 1316
	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;
1317 1318 1319
#ifdef CONFIG_TCP_MD5SIG
	struct tcp_md5sig_key *key;
#endif
L
Linus Torvalds 已提交
1320 1321 1322 1323 1324 1325 1326 1327

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

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

1328
		if (newsk == NULL)
L
Linus Torvalds 已提交
1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339
			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));

E
Eric Dumazet 已提交
1340
		ipv6_addr_set_v4mapped(newinet->inet_daddr, &newnp->daddr);
L
Linus Torvalds 已提交
1341

E
Eric Dumazet 已提交
1342
		ipv6_addr_set_v4mapped(newinet->inet_saddr, &newnp->saddr);
L
Linus Torvalds 已提交
1343 1344 1345

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

1346
		inet_csk(newsk)->icsk_af_ops = &ipv6_mapped;
L
Linus Torvalds 已提交
1347
		newsk->sk_backlog_rcv = tcp_v4_do_rcv;
1348 1349 1350 1351
#ifdef CONFIG_TCP_MD5SIG
		newtp->af_specific = &tcp_sock_ipv6_mapped_specific;
#endif

L
Linus Torvalds 已提交
1352 1353
		newnp->pktoptions  = NULL;
		newnp->opt	   = NULL;
1354
		newnp->mcast_oif   = inet6_iif(skb);
1355
		newnp->mcast_hops  = ipv6_hdr(skb)->hop_limit;
L
Linus Torvalds 已提交
1356

1357 1358 1359 1360
		/*
		 * 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
1361 1362 1363
		 */

		/* It is tricky place. Until this moment IPv4 tcp
1364
		   worked with IPv6 icsk.icsk_af_ops.
L
Linus Torvalds 已提交
1365 1366
		   Sync it now.
		 */
1367
		tcp_sync_mss(newsk, inet_csk(newsk)->icsk_pmtu_cookie);
L
Linus Torvalds 已提交
1368 1369 1370 1371

		return newsk;
	}

1372
	treq = inet6_rsk(req);
L
Linus Torvalds 已提交
1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383
	opt = np->opt;

	if (sk_acceptq_is_full(sk))
		goto out_overflow;

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

		memset(&fl, 0, sizeof(fl));
		fl.proto = IPPROTO_TCP;
1384
		ipv6_addr_copy(&fl.fl6_dst, &treq->rmt_addr);
L
Linus Torvalds 已提交
1385 1386 1387 1388 1389 1390
		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;
		}
1391
		ipv6_addr_copy(&fl.fl6_src, &treq->loc_addr);
L
Linus Torvalds 已提交
1392
		fl.oif = sk->sk_bound_dev_if;
B
Brian Haley 已提交
1393
		fl.mark = sk->sk_mark;
1394
		fl.fl_ip_dport = inet_rsk(req)->rmt_port;
1395
		fl.fl_ip_sport = inet_rsk(req)->loc_port;
1396
		security_req_classify_flow(req, &fl);
L
Linus Torvalds 已提交
1397 1398 1399 1400 1401 1402 1403

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

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

A
Alexey Dobriyan 已提交
1404
		if ((xfrm_lookup(sock_net(sk), &dst, &fl, sk, 0)) < 0)
L
Linus Torvalds 已提交
1405
			goto out;
1406
	}
L
Linus Torvalds 已提交
1407 1408 1409 1410 1411

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

1412 1413 1414 1415 1416
	/*
	 * 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
1417

1418
	newsk->sk_gso_type = SKB_GSO_TCPV6;
1419
	__ip6_dst_store(newsk, dst, NULL, NULL);
L
Linus Torvalds 已提交
1420 1421 1422 1423 1424 1425 1426 1427 1428 1429

	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));

1430 1431 1432 1433
	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
1434

1435
	/* Now IPv6 options...
L
Linus Torvalds 已提交
1436 1437 1438 1439

	   First: no IPv4 options.
	 */
	newinet->opt = NULL;
1440
	newnp->ipv6_fl_list = NULL;
L
Linus Torvalds 已提交
1441 1442 1443 1444 1445 1446

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

	/* Clone pktoptions received with SYN */
	newnp->pktoptions = NULL;
1447 1448 1449 1450
	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
1451 1452 1453 1454
		if (newnp->pktoptions)
			skb_set_owner_r(newnp->pktoptions, newsk);
	}
	newnp->opt	  = NULL;
1455
	newnp->mcast_oif  = inet6_iif(skb);
1456
	newnp->mcast_hops = ipv6_hdr(skb)->hop_limit;
L
Linus Torvalds 已提交
1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469

	/* 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);
	}

1470
	inet_csk(newsk)->icsk_ext_hdr_len = 0;
L
Linus Torvalds 已提交
1471
	if (newnp->opt)
1472 1473
		inet_csk(newsk)->icsk_ext_hdr_len = (newnp->opt->opt_nflen +
						     newnp->opt->opt_flen);
L
Linus Torvalds 已提交
1474

J
John Heffner 已提交
1475
	tcp_mtup_init(newsk);
L
Linus Torvalds 已提交
1476 1477 1478 1479
	tcp_sync_mss(newsk, dst_mtu(dst));
	newtp->advmss = dst_metric(dst, RTAX_ADVMSS);
	tcp_initialize_rcv_mss(newsk);

E
Eric Dumazet 已提交
1480 1481
	newinet->inet_daddr = newinet->inet_saddr = LOOPBACK4_IPV6;
	newinet->inet_rcv_saddr = LOOPBACK4_IPV6;
L
Linus Torvalds 已提交
1482

1483 1484 1485 1486 1487 1488 1489 1490
#ifdef CONFIG_TCP_MD5SIG
	/* Copy over the MD5 key from the original socket */
	if ((key = tcp_v6_md5_do_lookup(sk, &newnp->daddr)) != 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.
		 */
A
Arnaldo Carvalho de Melo 已提交
1491 1492
		char *newkey = kmemdup(key->key, key->keylen, GFP_ATOMIC);
		if (newkey != NULL)
1493
			tcp_v6_md5_do_add(newsk, &newnp->daddr,
1494 1495 1496 1497
					  newkey, key->keylen);
	}
#endif

1498
	__inet6_hash(newsk, NULL);
1499
	__inet_inherit_port(sk, newsk);
L
Linus Torvalds 已提交
1500 1501 1502 1503

	return newsk;

out_overflow:
1504
	NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_LISTENOVERFLOWS);
L
Linus Torvalds 已提交
1505
out:
1506
	NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_LISTENDROPS);
L
Linus Torvalds 已提交
1507 1508 1509 1510 1511 1512
	if (opt && opt != np->opt)
		sock_kfree_s(sk, opt, opt->tot_len);
	dst_release(dst);
	return NULL;
}

1513
static __sum16 tcp_v6_checksum_init(struct sk_buff *skb)
L
Linus Torvalds 已提交
1514
{
1515
	if (skb->ip_summed == CHECKSUM_COMPLETE) {
H
Herbert Xu 已提交
1516
		if (!tcp_v6_check(skb->len, &ipv6_hdr(skb)->saddr,
1517
				  &ipv6_hdr(skb)->daddr, skb->csum)) {
1518
			skb->ip_summed = CHECKSUM_UNNECESSARY;
L
Linus Torvalds 已提交
1519
			return 0;
1520
		}
L
Linus Torvalds 已提交
1521
	}
1522

H
Herbert Xu 已提交
1523
	skb->csum = ~csum_unfold(tcp_v6_check(skb->len,
1524 1525
					      &ipv6_hdr(skb)->saddr,
					      &ipv6_hdr(skb)->daddr, 0));
1526

L
Linus Torvalds 已提交
1527
	if (skb->len <= 76) {
1528
		return __skb_checksum_complete(skb);
L
Linus Torvalds 已提交
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
	}
	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);

1558 1559 1560 1561 1562
#ifdef CONFIG_TCP_MD5SIG
	if (tcp_v6_inbound_md5_hash (sk, skb))
		goto discard;
#endif

1563
	if (sk_filter(sk, skb))
L
Linus Torvalds 已提交
1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581
		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.
1582
					       --ANK (980728)
L
Linus Torvalds 已提交
1583 1584 1585 1586 1587 1588
	 */
	if (np->rxopt.all)
		opt_skb = skb_clone(skb, GFP_ATOMIC);

	if (sk->sk_state == TCP_ESTABLISHED) { /* Fast path */
		TCP_CHECK_TIMER(sk);
1589
		if (tcp_rcv_established(sk, skb, tcp_hdr(skb), skb->len))
L
Linus Torvalds 已提交
1590 1591 1592 1593 1594 1595 1596
			goto reset;
		TCP_CHECK_TIMER(sk);
		if (opt_skb)
			goto ipv6_pktoptions;
		return 0;
	}

1597
	if (skb->len < tcp_hdrlen(skb) || tcp_checksum_complete(skb))
L
Linus Torvalds 已提交
1598 1599
		goto csum_err;

1600
	if (sk->sk_state == TCP_LISTEN) {
L
Linus Torvalds 已提交
1601 1602 1603 1604 1605 1606 1607 1608 1609
		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..
		 */
1610
		if(nsk != sk) {
L
Linus Torvalds 已提交
1611 1612 1613 1614 1615 1616 1617 1618 1619
			if (tcp_child_process(sk, nsk, skb))
				goto reset;
			if (opt_skb)
				__kfree_skb(opt_skb);
			return 0;
		}
	}

	TCP_CHECK_TIMER(sk);
1620
	if (tcp_rcv_state_process(sk, skb, tcp_hdr(skb), skb->len))
L
Linus Torvalds 已提交
1621 1622 1623 1624 1625 1626 1627
		goto reset;
	TCP_CHECK_TIMER(sk);
	if (opt_skb)
		goto ipv6_pktoptions;
	return 0;

reset:
1628
	tcp_v6_send_reset(sk, skb);
L
Linus Torvalds 已提交
1629 1630 1631 1632 1633 1634
discard:
	if (opt_skb)
		__kfree_skb(opt_skb);
	kfree_skb(skb);
	return 0;
csum_err:
1635
	TCP_INC_STATS_BH(sock_net(sk), TCP_MIB_INERRS);
L
Linus Torvalds 已提交
1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649
	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))) {
1650
		if (np->rxopt.bits.rxinfo || np->rxopt.bits.rxoinfo)
1651
			np->mcast_oif = inet6_iif(opt_skb);
1652
		if (np->rxopt.bits.rxhlim || np->rxopt.bits.rxohlim)
1653
			np->mcast_hops = ipv6_hdr(opt_skb)->hop_limit;
L
Linus Torvalds 已提交
1654 1655 1656 1657 1658 1659 1660 1661 1662
		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);
		}
	}

1663
	kfree_skb(opt_skb);
L
Linus Torvalds 已提交
1664 1665 1666
	return 0;
}

1667
static int tcp_v6_rcv(struct sk_buff *skb)
L
Linus Torvalds 已提交
1668
{
1669
	struct tcphdr *th;
L
Linus Torvalds 已提交
1670 1671
	struct sock *sk;
	int ret;
1672
	struct net *net = dev_net(skb->dev);
L
Linus Torvalds 已提交
1673 1674 1675 1676 1677 1678 1679

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

	/*
	 *	Count it even if it's bad.
	 */
1680
	TCP_INC_STATS_BH(net, TCP_MIB_INSEGS);
L
Linus Torvalds 已提交
1681 1682 1683 1684

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

1685
	th = tcp_hdr(skb);
L
Linus Torvalds 已提交
1686 1687 1688 1689 1690 1691

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

1692
	if (!skb_csum_unnecessary(skb) && tcp_v6_checksum_init(skb))
L
Linus Torvalds 已提交
1693 1694
		goto bad_packet;

1695
	th = tcp_hdr(skb);
L
Linus Torvalds 已提交
1696 1697 1698 1699 1700
	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;
1701
	TCP_SKB_CB(skb)->flags = ipv6_get_dsfield(ipv6_hdr(skb));
L
Linus Torvalds 已提交
1702 1703
	TCP_SKB_CB(skb)->sacked = 0;

1704
	sk = __inet6_lookup_skb(&tcp_hashinfo, skb, th->source, th->dest);
L
Linus Torvalds 已提交
1705 1706 1707 1708 1709 1710 1711 1712 1713 1714
	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;

1715
	if (sk_filter(sk, skb))
L
Linus Torvalds 已提交
1716 1717 1718 1719
		goto discard_and_relse;

	skb->dev = NULL;

1720
	bh_lock_sock_nested(sk);
L
Linus Torvalds 已提交
1721 1722
	ret = 0;
	if (!sock_owned_by_user(sk)) {
C
Chris Leech 已提交
1723
#ifdef CONFIG_NET_DMA
1724
		struct tcp_sock *tp = tcp_sk(sk);
1725
		if (!tp->ucopy.dma_chan && tp->ucopy.pinned_list)
1726
			tp->ucopy.dma_chan = dma_find_channel(DMA_MEMCPY);
1727 1728 1729
		if (tp->ucopy.dma_chan)
			ret = tcp_v6_do_rcv(sk, skb);
		else
C
Chris Leech 已提交
1730 1731 1732 1733 1734
#endif
		{
			if (!tcp_prequeue(sk, skb))
				ret = tcp_v6_do_rcv(sk, skb);
		}
L
Linus Torvalds 已提交
1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747
	} 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:
1748
		TCP_INC_STATS_BH(net, TCP_MIB_INERRS);
L
Linus Torvalds 已提交
1749
	} else {
1750
		tcp_v6_send_reset(NULL, skb);
L
Linus Torvalds 已提交
1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767
	}

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)) {
1768
		inet_twsk_put(inet_twsk(sk));
L
Linus Torvalds 已提交
1769 1770 1771 1772
		goto discard_it;
	}

	if (skb->len < (th->doff<<2) || tcp_checksum_complete(skb)) {
1773
		TCP_INC_STATS_BH(net, TCP_MIB_INERRS);
1774
		inet_twsk_put(inet_twsk(sk));
L
Linus Torvalds 已提交
1775 1776 1777
		goto discard_it;
	}

1778
	switch (tcp_timewait_state_process(inet_twsk(sk), skb, th)) {
L
Linus Torvalds 已提交
1779 1780 1781 1782
	case TCP_TW_SYN:
	{
		struct sock *sk2;

1783
		sk2 = inet6_lookup_listener(dev_net(skb->dev), &tcp_hashinfo,
1784
					    &ipv6_hdr(skb)->daddr,
1785
					    ntohs(th->dest), inet6_iif(skb));
L
Linus Torvalds 已提交
1786
		if (sk2 != NULL) {
1787 1788 1789
			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
1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810
			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;
}

1811
static const struct inet_connection_sock_af_ops ipv6_specific = {
1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822
	.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),
1823
	.bind_conflict	   = inet6_csk_bind_conflict,
1824
#ifdef CONFIG_COMPAT
1825 1826
	.compat_setsockopt = compat_ipv6_setsockopt,
	.compat_getsockopt = compat_ipv6_getsockopt,
1827
#endif
L
Linus Torvalds 已提交
1828 1829
};

1830
#ifdef CONFIG_TCP_MD5SIG
1831
static const struct tcp_sock_af_ops tcp_sock_ipv6_specific = {
1832
	.md5_lookup	=	tcp_v6_md5_lookup,
1833
	.calc_md5_hash	=	tcp_v6_md5_hash_skb,
1834 1835 1836
	.md5_add	=	tcp_v6_md5_add_func,
	.md5_parse	=	tcp_v6_parse_md5_keys,
};
1837
#endif
1838

L
Linus Torvalds 已提交
1839 1840 1841 1842
/*
 *	TCP over IPv4 via INET6 API
 */

1843
static const struct inet_connection_sock_af_ops ipv6_mapped = {
1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854
	.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),
1855
	.bind_conflict	   = inet6_csk_bind_conflict,
1856
#ifdef CONFIG_COMPAT
1857 1858
	.compat_setsockopt = compat_ipv6_setsockopt,
	.compat_getsockopt = compat_ipv6_getsockopt,
1859
#endif
L
Linus Torvalds 已提交
1860 1861
};

1862
#ifdef CONFIG_TCP_MD5SIG
1863
static const struct tcp_sock_af_ops tcp_sock_ipv6_mapped_specific = {
1864
	.md5_lookup	=	tcp_v4_md5_lookup,
1865
	.calc_md5_hash	=	tcp_v4_md5_hash_skb,
1866 1867 1868
	.md5_add	=	tcp_v6_md5_add_func,
	.md5_parse	=	tcp_v6_parse_md5_keys,
};
1869
#endif
1870

L
Linus Torvalds 已提交
1871 1872 1873 1874 1875
/* 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)
{
1876
	struct inet_connection_sock *icsk = inet_csk(sk);
L
Linus Torvalds 已提交
1877 1878 1879 1880 1881 1882
	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);

1883
	icsk->icsk_rto = TCP_TIMEOUT_INIT;
L
Linus Torvalds 已提交
1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895
	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.
	 */
I
Ilpo Järvinen 已提交
1896
	tp->snd_ssthresh = TCP_INFINITE_SSTHRESH;
L
Linus Torvalds 已提交
1897
	tp->snd_cwnd_clamp = ~0;
1898
	tp->mss_cache = TCP_MSS_DEFAULT;
L
Linus Torvalds 已提交
1899 1900 1901 1902 1903

	tp->reordering = sysctl_tcp_reordering;

	sk->sk_state = TCP_CLOSE;

1904
	icsk->icsk_af_ops = &ipv6_specific;
1905
	icsk->icsk_ca_ops = &tcp_init_congestion_ops;
1906
	icsk->icsk_sync_mss = tcp_sync_mss;
L
Linus Torvalds 已提交
1907 1908 1909
	sk->sk_write_space = sk_stream_write_space;
	sock_set_flag(sk, SOCK_USE_WRITE_QUEUE);

1910 1911 1912 1913
#ifdef CONFIG_TCP_MD5SIG
	tp->af_specific = &tcp_sock_ipv6_specific;
#endif

1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926
	/* TCP Cookie Transactions */
	if (sysctl_tcp_cookie_size > 0) {
		/* Default, cookies without s_data_payload. */
		tp->cookie_values =
			kzalloc(sizeof(*tp->cookie_values),
				sk->sk_allocation);
		if (tp->cookie_values != NULL)
			kref_init(&tp->cookie_values->kref);
	}
	/* Presumed zeroed, in order of appearance:
	 *	cookie_in_always, cookie_out_never,
	 *	s_data_constant, s_data_in, s_data_out
	 */
L
Linus Torvalds 已提交
1927 1928 1929
	sk->sk_sndbuf = sysctl_tcp_wmem[1];
	sk->sk_rcvbuf = sysctl_tcp_rmem[1];

H
Herbert Xu 已提交
1930
	local_bh_disable();
1931
	percpu_counter_inc(&tcp_sockets_allocated);
H
Herbert Xu 已提交
1932
	local_bh_enable();
L
Linus Torvalds 已提交
1933 1934 1935 1936

	return 0;
}

1937
static void tcp_v6_destroy_sock(struct sock *sk)
L
Linus Torvalds 已提交
1938
{
1939 1940 1941 1942 1943
#ifdef CONFIG_TCP_MD5SIG
	/* Clean up the MD5 key list */
	if (tcp_sk(sk)->md5sig_info)
		tcp_v6_clear_md5_list(sk);
#endif
L
Linus Torvalds 已提交
1944
	tcp_v4_destroy_sock(sk);
1945
	inet6_destroy_sock(sk);
L
Linus Torvalds 已提交
1946 1947
}

1948
#ifdef CONFIG_PROC_FS
L
Linus Torvalds 已提交
1949
/* Proc filesystem TCPv6 sock list dumping. */
1950
static void get_openreq6(struct seq_file *seq,
1951
			 struct sock *sk, struct request_sock *req, int i, int uid)
L
Linus Torvalds 已提交
1952 1953
{
	int ttd = req->expires - jiffies;
1954 1955
	struct in6_addr *src = &inet6_rsk(req)->loc_addr;
	struct in6_addr *dest = &inet6_rsk(req)->rmt_addr;
L
Linus Torvalds 已提交
1956 1957 1958 1959 1960 1961 1962 1963 1964 1965

	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],
1966
		   ntohs(inet_rsk(req)->loc_port),
L
Linus Torvalds 已提交
1967 1968
		   dest->s6_addr32[0], dest->s6_addr32[1],
		   dest->s6_addr32[2], dest->s6_addr32[3],
1969
		   ntohs(inet_rsk(req)->rmt_port),
L
Linus Torvalds 已提交
1970 1971
		   TCP_SYN_RECV,
		   0,0, /* could print option size, but that is af dependent. */
1972 1973
		   1,   /* timers active (only the expire timer) */
		   jiffies_to_clock_t(ttd),
L
Linus Torvalds 已提交
1974 1975
		   req->retrans,
		   uid,
1976
		   0,  /* non standard timer */
L
Linus Torvalds 已提交
1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988
		   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);
1989
	const struct inet_connection_sock *icsk = inet_csk(sp);
L
Linus Torvalds 已提交
1990 1991 1992 1993
	struct ipv6_pinfo *np = inet6_sk(sp);

	dest  = &np->daddr;
	src   = &np->rcv_saddr;
E
Eric Dumazet 已提交
1994 1995
	destp = ntohs(inet->inet_dport);
	srcp  = ntohs(inet->inet_sport);
1996 1997

	if (icsk->icsk_pending == ICSK_TIME_RETRANS) {
L
Linus Torvalds 已提交
1998
		timer_active	= 1;
1999 2000
		timer_expires	= icsk->icsk_timeout;
	} else if (icsk->icsk_pending == ICSK_TIME_PROBE0) {
L
Linus Torvalds 已提交
2001
		timer_active	= 4;
2002
		timer_expires	= icsk->icsk_timeout;
L
Linus Torvalds 已提交
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
	} 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 "
2013
		   "%02X %08X:%08X %02X:%08lX %08X %5d %8d %lu %d %p %lu %lu %u %u %d\n",
L
Linus Torvalds 已提交
2014 2015 2016 2017 2018
		   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,
2019
		   sp->sk_state,
2020 2021
		   tp->write_seq-tp->snd_una,
		   (sp->sk_state == TCP_LISTEN) ? sp->sk_ack_backlog : (tp->rcv_nxt - tp->copied_seq),
L
Linus Torvalds 已提交
2022 2023
		   timer_active,
		   jiffies_to_clock_t(timer_expires - jiffies),
2024
		   icsk->icsk_retransmits,
L
Linus Torvalds 已提交
2025
		   sock_i_uid(sp),
2026
		   icsk->icsk_probes_out,
L
Linus Torvalds 已提交
2027 2028
		   sock_i_ino(sp),
		   atomic_read(&sp->sk_refcnt), sp,
2029 2030
		   jiffies_to_clock_t(icsk->icsk_rto),
		   jiffies_to_clock_t(icsk->icsk_ack.ato),
2031
		   (icsk->icsk_ack.quick << 1 ) | icsk->icsk_ack.pingpong,
I
Ilpo Järvinen 已提交
2032 2033
		   tp->snd_cwnd,
		   tcp_in_initial_slowstart(tp) ? -1 : tp->snd_ssthresh
L
Linus Torvalds 已提交
2034 2035 2036
		   );
}

2037
static void get_timewait6_sock(struct seq_file *seq,
2038
			       struct inet_timewait_sock *tw, int i)
L
Linus Torvalds 已提交
2039 2040 2041
{
	struct in6_addr *dest, *src;
	__u16 destp, srcp;
2042
	struct inet6_timewait_sock *tw6 = inet6_twsk((struct sock *)tw);
L
Linus Torvalds 已提交
2043 2044 2045 2046 2047
	int ttd = tw->tw_ttd - jiffies;

	if (ttd < 0)
		ttd = 0;

2048 2049
	dest = &tw6->tw_v6_daddr;
	src  = &tw6->tw_v6_rcv_saddr;
L
Linus Torvalds 已提交
2050 2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099
	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);
}

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 tcp_seq_afinfo tcp6_seq_afinfo = {
	.name		= "tcp6",
	.family		= AF_INET6,
2100 2101 2102
	.seq_fops	= {
		.owner		= THIS_MODULE,
	},
2103 2104 2105
	.seq_ops	= {
		.show		= tcp6_seq_show,
	},
L
Linus Torvalds 已提交
2106 2107
};

2108
int tcp6_proc_init(struct net *net)
L
Linus Torvalds 已提交
2109
{
2110
	return tcp_proc_register(net, &tcp6_seq_afinfo);
L
Linus Torvalds 已提交
2111 2112
}

2113
void tcp6_proc_exit(struct net *net)
L
Linus Torvalds 已提交
2114
{
2115
	tcp_proc_unregister(net, &tcp6_seq_afinfo);
L
Linus Torvalds 已提交
2116 2117 2118 2119 2120 2121 2122 2123 2124
}
#endif

struct proto tcpv6_prot = {
	.name			= "TCPv6",
	.owner			= THIS_MODULE,
	.close			= tcp_close,
	.connect		= tcp_v6_connect,
	.disconnect		= tcp_disconnect,
2125
	.accept			= inet_csk_accept,
L
Linus Torvalds 已提交
2126 2127 2128 2129 2130 2131 2132 2133 2134
	.ioctl			= tcp_ioctl,
	.init			= tcp_v6_init_sock,
	.destroy		= tcp_v6_destroy_sock,
	.shutdown		= tcp_shutdown,
	.setsockopt		= tcp_setsockopt,
	.getsockopt		= tcp_getsockopt,
	.recvmsg		= tcp_recvmsg,
	.backlog_rcv		= tcp_v6_do_rcv,
	.hash			= tcp_v6_hash,
2135 2136
	.unhash			= inet_unhash,
	.get_port		= inet_csk_get_port,
L
Linus Torvalds 已提交
2137 2138 2139 2140
	.enter_memory_pressure	= tcp_enter_memory_pressure,
	.sockets_allocated	= &tcp_sockets_allocated,
	.memory_allocated	= &tcp_memory_allocated,
	.memory_pressure	= &tcp_memory_pressure,
2141
	.orphan_count		= &tcp_orphan_count,
L
Linus Torvalds 已提交
2142 2143 2144 2145 2146
	.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),
2147
	.slab_flags		= SLAB_DESTROY_BY_RCU,
2148
	.twsk_prot		= &tcp6_timewait_sock_ops,
2149
	.rsk_prot		= &tcp6_request_sock_ops,
2150
	.h.hashinfo		= &tcp_hashinfo,
2151 2152 2153 2154
#ifdef CONFIG_COMPAT
	.compat_setsockopt	= compat_tcp_setsockopt,
	.compat_getsockopt	= compat_tcp_getsockopt,
#endif
L
Linus Torvalds 已提交
2155 2156
};

2157
static const struct inet6_protocol tcpv6_protocol = {
L
Linus Torvalds 已提交
2158 2159
	.handler	=	tcp_v6_rcv,
	.err_handler	=	tcp_v6_err,
2160
	.gso_send_check	=	tcp_v6_gso_send_check,
H
Herbert Xu 已提交
2161
	.gso_segment	=	tcp_tso_segment,
H
Herbert Xu 已提交
2162 2163
	.gro_receive	=	tcp6_gro_receive,
	.gro_complete	=	tcp6_gro_complete,
L
Linus Torvalds 已提交
2164 2165 2166 2167 2168 2169 2170 2171 2172
	.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,
	.no_check	=	0,
2173 2174
	.flags		=	INET_PROTOSW_PERMANENT |
				INET_PROTOSW_ICSK,
L
Linus Torvalds 已提交
2175 2176
};

2177 2178
static int tcpv6_net_init(struct net *net)
{
2179 2180
	return inet_ctl_sock_create(&net->ipv6.tcp_sk, PF_INET6,
				    SOCK_RAW, IPPROTO_TCP, net);
2181 2182 2183 2184
}

static void tcpv6_net_exit(struct net *net)
{
2185
	inet_ctl_sock_destroy(net->ipv6.tcp_sk);
E
Eric W. Biederman 已提交
2186 2187 2188 2189 2190
}

static void tcpv6_net_exit_batch(struct list_head *net_exit_list)
{
	inet_twsk_purge(&tcp_hashinfo, &tcp_death_row, AF_INET6);
2191 2192 2193
}

static struct pernet_operations tcpv6_net_ops = {
E
Eric W. Biederman 已提交
2194 2195 2196
	.init	    = tcpv6_net_init,
	.exit	    = tcpv6_net_exit,
	.exit_batch = tcpv6_net_exit_batch,
2197 2198
};

2199
int __init tcpv6_init(void)
L
Linus Torvalds 已提交
2200
{
2201 2202 2203 2204 2205 2206
	int ret;

	ret = inet6_add_protocol(&tcpv6_protocol, IPPROTO_TCP);
	if (ret)
		goto out;

L
Linus Torvalds 已提交
2207
	/* register inet6 protocol */
2208 2209 2210 2211
	ret = inet6_register_protosw(&tcpv6_protosw);
	if (ret)
		goto out_tcpv6_protocol;

2212
	ret = register_pernet_subsys(&tcpv6_net_ops);
2213 2214 2215 2216
	if (ret)
		goto out_tcpv6_protosw;
out:
	return ret;
2217

2218 2219 2220 2221 2222 2223 2224
out_tcpv6_protocol:
	inet6_del_protocol(&tcpv6_protocol, IPPROTO_TCP);
out_tcpv6_protosw:
	inet6_unregister_protosw(&tcpv6_protosw);
	goto out;
}

2225
void tcpv6_exit(void)
2226
{
2227
	unregister_pernet_subsys(&tcpv6_net_ops);
2228 2229
	inet6_unregister_protosw(&tcpv6_protosw);
	inet6_del_protocol(&tcpv6_protocol, IPPROTO_TCP);
L
Linus Torvalds 已提交
2230
}