ipv4.c 30.2 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
/*
 *  net/dccp/ipv4.c
 *
 *  An implementation of the DCCP protocol
 *  Arnaldo Carvalho de Melo <acme@conectiva.com.br>
 *
 *	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/config.h>
#include <linux/dccp.h>
#include <linux/icmp.h>
#include <linux/module.h>
#include <linux/skbuff.h>
#include <linux/random.h>

#include <net/icmp.h>
#include <net/inet_hashtables.h>
22
#include <net/inet_sock.h>
23
#include <net/sock.h>
24
#include <net/timewait_sock.h>
25 26 27
#include <net/tcp_states.h>
#include <net/xfrm.h>

28
#include "ackvec.h"
29 30
#include "ccid.h"
#include "dccp.h"
31
#include "feat.h"
32 33 34 35

struct inet_hashinfo __cacheline_aligned dccp_hashinfo = {
	.lhash_lock	= RW_LOCK_UNLOCKED,
	.lhash_users	= ATOMIC_INIT(0),
36
	.lhash_wait = __WAIT_QUEUE_HEAD_INITIALIZER(dccp_hashinfo.lhash_wait),
37 38
};

39 40
EXPORT_SYMBOL_GPL(dccp_hashinfo);

41 42
static int dccp_v4_get_port(struct sock *sk, const unsigned short snum)
{
43 44
	return inet_csk_get_port(&dccp_hashinfo, sk, snum,
				 inet_csk_bind_conflict);
45 46 47 48 49 50 51
}

static void dccp_v4_hash(struct sock *sk)
{
	inet_hash(&dccp_hashinfo, sk);
}

52
void dccp_unhash(struct sock *sk)
53 54 55 56
{
	inet_unhash(&dccp_hashinfo, sk);
}

57 58 59
EXPORT_SYMBOL_GPL(dccp_unhash);

int dccp_v4_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len)
60 61 62 63 64 65 66 67 68 69 70
{
	struct inet_sock *inet = inet_sk(sk);
	struct dccp_sock *dp = dccp_sk(sk);
	const struct sockaddr_in *usin = (struct sockaddr_in *)uaddr;
	struct rtable *rt;
	u32 daddr, nexthop;
	int tmp;
	int err;

	dp->dccps_role = DCCP_ROLE_CLIENT;

71 72 73
	if (dccp_service_not_initialized(sk))
		return -EPROTO;

74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108
	if (addr_len < sizeof(struct sockaddr_in))
		return -EINVAL;

	if (usin->sin_family != AF_INET)
		return -EAFNOSUPPORT;

	nexthop = daddr = usin->sin_addr.s_addr;
	if (inet->opt != NULL && inet->opt->srr) {
		if (daddr == 0)
			return -EINVAL;
		nexthop = inet->opt->faddr;
	}

	tmp = ip_route_connect(&rt, nexthop, inet->saddr,
			       RT_CONN_FLAGS(sk), sk->sk_bound_dev_if,
			       IPPROTO_DCCP,
			       inet->sport, usin->sin_port, sk);
	if (tmp < 0)
		return tmp;

	if (rt->rt_flags & (RTCF_MULTICAST | RTCF_BROADCAST)) {
		ip_rt_put(rt);
		return -ENETUNREACH;
	}

	if (inet->opt == NULL || !inet->opt->srr)
		daddr = rt->rt_dst;

	if (inet->saddr == 0)
		inet->saddr = rt->rt_src;
	inet->rcv_saddr = inet->saddr;

	inet->dport = usin->sin_port;
	inet->daddr = daddr;

109
	inet_csk(sk)->icsk_ext_hdr_len = 0;
110
	if (inet->opt != NULL)
111
		inet_csk(sk)->icsk_ext_hdr_len = inet->opt->optlen;
112 113 114 115 116 117 118
	/*
	 * Socket identity is still unknown (sport may be zero).
	 * However we set state to DCCP_REQUESTING and not releasing socket
	 * lock select source port, enter ourselves into the hash tables and
	 * complete initialization after this.
	 */
	dccp_set_state(sk, DCCP_REQUESTING);
119
	err = inet_hash_connect(&dccp_death_row, sk);
120 121 122
	if (err != 0)
		goto failure;

123 124
	err = ip_route_newports(&rt, IPPROTO_DCCP, inet->sport, inet->dport,
	                        sk);
125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146
	if (err != 0)
		goto failure;

	/* OK, now commit destination to socket.  */
	sk_setup_caps(sk, &rt->u.dst);

	dp->dccps_gar =
		dp->dccps_iss = secure_dccp_sequence_number(inet->saddr,
							    inet->daddr,
							    inet->sport,
							    usin->sin_port);
	dccp_update_gss(sk, dp->dccps_iss);

	inet->id = dp->dccps_iss ^ jiffies;

	err = dccp_connect(sk);
	rt = NULL;
	if (err != 0)
		goto failure;
out:
	return err;
failure:
147 148 149
	/*
	 * This unhashes the socket and releases the local port, if necessary.
	 */
150 151 152 153 154 155 156
	dccp_set_state(sk, DCCP_CLOSED);
	ip_rt_put(rt);
	sk->sk_route_caps = 0;
	inet->dport = 0;
	goto out;
}

157 158
EXPORT_SYMBOL_GPL(dccp_v4_connect);

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
/*
 * This routine does path mtu discovery as defined in RFC1191.
 */
static inline void dccp_do_pmtu_discovery(struct sock *sk,
					  const struct iphdr *iph,
					  u32 mtu)
{
	struct dst_entry *dst;
	const struct inet_sock *inet = inet_sk(sk);
	const struct dccp_sock *dp = dccp_sk(sk);

	/* We are not interested in DCCP_LISTEN and request_socks (RESPONSEs
	 * send out by Linux are always < 576bytes so they should go through
	 * unfragmented).
	 */
	if (sk->sk_state == DCCP_LISTEN)
		return;

	/* We don't check in the destentry if pmtu discovery is forbidden
	 * on this route. We just assume that no packet_to_big packets
	 * are send back when pmtu discovery is not active.
     	 * There is a small race when the user changes this flag in the
	 * route, but I think that's acceptable.
	 */
	if ((dst = __sk_dst_check(sk, 0)) == NULL)
		return;

	dst->ops->update_pmtu(dst, mtu);

	/* Something is about to be wrong... Remember soft error
	 * for the case, if this connection will not able to recover.
	 */
	if (mtu < dst_mtu(dst) && ip_dont_fragment(sk, dst))
		sk->sk_err_soft = EMSGSIZE;

	mtu = dst_mtu(dst);

	if (inet->pmtudisc != IP_PMTUDISC_DONT &&
197
	    inet_csk(sk)->icsk_pmtu_cookie > mtu) {
198 199 200 201 202
		dccp_sync_mss(sk, mtu);

		/*
		 * From: draft-ietf-dccp-spec-11.txt
		 *
203 204 205
		 *	DCCP-Sync packets are the best choice for upward
		 *	probing, since DCCP-Sync probes do not risk application
		 *	data loss.
206
		 */
207
		dccp_send_sync(sk, dp->dccps_gsr, DCCP_PKT_SYNC);
208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243
	} /* else let the usual retransmit timer handle it */
}

static void dccp_v4_ctl_send_ack(struct sk_buff *rxskb)
{
	int err;
	struct dccp_hdr *rxdh = dccp_hdr(rxskb), *dh;
	const int dccp_hdr_ack_len = sizeof(struct dccp_hdr) +
				     sizeof(struct dccp_hdr_ext) +
				     sizeof(struct dccp_hdr_ack_bits);
	struct sk_buff *skb;

	if (((struct rtable *)rxskb->dst)->rt_type != RTN_LOCAL)
		return;

	skb = alloc_skb(MAX_DCCP_HEADER + 15, GFP_ATOMIC);
	if (skb == NULL)
		return;

	/* Reserve space for headers. */
	skb_reserve(skb, MAX_DCCP_HEADER);

	skb->dst = dst_clone(rxskb->dst);

	skb->h.raw = skb_push(skb, dccp_hdr_ack_len);
	dh = dccp_hdr(skb);
	memset(dh, 0, dccp_hdr_ack_len);

	/* Build DCCP header and checksum it. */
	dh->dccph_type	   = DCCP_PKT_ACK;
	dh->dccph_sport	   = rxdh->dccph_dport;
	dh->dccph_dport	   = rxdh->dccph_sport;
	dh->dccph_doff	   = dccp_hdr_ack_len / 4;
	dh->dccph_x	   = 1;

	dccp_hdr_set_seq(dh, DCCP_SKB_CB(rxskb)->dccpd_ack_seq);
244 245
	dccp_hdr_set_ack(dccp_hdr_ack_bits(skb),
			 DCCP_SKB_CB(rxskb)->dccpd_seq);
246 247 248

	bh_lock_sock(dccp_ctl_socket->sk);
	err = ip_build_and_send_pkt(skb, dccp_ctl_socket->sk,
249 250
				    rxskb->nh.iph->daddr,
				    rxskb->nh.iph->saddr, NULL);
251 252 253 254 255 256 257 258
	bh_unlock_sock(dccp_ctl_socket->sk);

	if (err == NET_XMIT_CN || err == 0) {
		DCCP_INC_STATS_BH(DCCP_MIB_OUTSEGS);
		DCCP_INC_STATS_BH(DCCP_MIB_OUTRSTS);
	}
}

259 260
static void dccp_v4_reqsk_send_ack(struct sk_buff *skb,
				   struct request_sock *req)
261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279
{
	dccp_v4_ctl_send_ack(skb);
}

static int dccp_v4_send_response(struct sock *sk, struct request_sock *req,
				 struct dst_entry *dst)
{
	int err = -1;
	struct sk_buff *skb;

	/* First, grab a route. */
	
	if (dst == NULL && (dst = inet_csk_route_req(sk, req)) == NULL)
		goto out;

	skb = dccp_make_response(sk, dst, req);
	if (skb != NULL) {
		const struct inet_request_sock *ireq = inet_rsk(req);

H
Herbert Xu 已提交
280
		memset(&(IPCB(skb)->opt), 0, sizeof(IPCB(skb)->opt));
281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307
		err = ip_build_and_send_pkt(skb, sk, ireq->loc_addr,
					    ireq->rmt_addr,
					    ireq->opt);
		if (err == NET_XMIT_CN)
			err = 0;
	}

out:
	dst_release(dst);
	return err;
}

/*
 * This routine is called by the ICMP module when it gets some sort of error
 * condition. If err < 0 then the socket should be closed and the error
 * returned to the user. If err > 0 it's just the icmp type << 8 | icmp code.
 * After adjustment header points to the first 8 bytes of the tcp header. We
 * need to find the appropriate port.
 *
 * The locking strategy used here is very "optimistic". When someone else
 * accesses the socket the ICMP is just dropped and for some paths there is no
 * check at all. A more general error queue to queue errors for later handling
 * is probably better.
 */
void dccp_v4_err(struct sk_buff *skb, u32 info)
{
	const struct iphdr *iph = (struct iphdr *)skb->data;
308 309
	const struct dccp_hdr *dh = (struct dccp_hdr *)(skb->data +
							(iph->ihl << 2));
310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448
	struct dccp_sock *dp;
	struct inet_sock *inet;
	const int type = skb->h.icmph->type;
	const int code = skb->h.icmph->code;
	struct sock *sk;
	__u64 seq;
	int err;

	if (skb->len < (iph->ihl << 2) + 8) {
		ICMP_INC_STATS_BH(ICMP_MIB_INERRORS);
		return;
	}

	sk = inet_lookup(&dccp_hashinfo, iph->daddr, dh->dccph_dport,
			 iph->saddr, dh->dccph_sport, inet_iif(skb));
	if (sk == NULL) {
		ICMP_INC_STATS_BH(ICMP_MIB_INERRORS);
		return;
	}

	if (sk->sk_state == DCCP_TIME_WAIT) {
		inet_twsk_put((struct inet_timewait_sock *)sk);
		return;
	}

	bh_lock_sock(sk);
	/* If too many ICMPs get dropped on busy
	 * servers this needs to be solved differently.
	 */
	if (sock_owned_by_user(sk))
		NET_INC_STATS_BH(LINUX_MIB_LOCKDROPPEDICMPS);

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

	dp = dccp_sk(sk);
	seq = dccp_hdr_seq(skb);
	if (sk->sk_state != DCCP_LISTEN &&
	    !between48(seq, dp->dccps_swl, dp->dccps_swh)) {
		NET_INC_STATS(LINUX_MIB_OUTOFWINDOWICMPS);
		goto out;
	}

	switch (type) {
	case ICMP_SOURCE_QUENCH:
		/* Just silently ignore these. */
		goto out;
	case ICMP_PARAMETERPROB:
		err = EPROTO;
		break;
	case ICMP_DEST_UNREACH:
		if (code > NR_ICMP_UNREACH)
			goto out;

		if (code == ICMP_FRAG_NEEDED) { /* PMTU discovery (RFC1191) */
			if (!sock_owned_by_user(sk))
				dccp_do_pmtu_discovery(sk, iph, info);
			goto out;
		}

		err = icmp_err_convert[code].errno;
		break;
	case ICMP_TIME_EXCEEDED:
		err = EHOSTUNREACH;
		break;
	default:
		goto out;
	}

	switch (sk->sk_state) {
		struct request_sock *req , **prev;
	case DCCP_LISTEN:
		if (sock_owned_by_user(sk))
			goto out;
		req = inet_csk_search_req(sk, &prev, dh->dccph_dport,
					  iph->daddr, iph->saddr);
		if (!req)
			goto out;

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

		if (seq != dccp_rsk(req)->dreq_iss) {
			NET_INC_STATS_BH(LINUX_MIB_OUTOFWINDOWICMPS);
			goto out;
		}
		/*
		 * Still in RESPOND, just remove it silently.
		 * There is no good way to pass the error to the newly
		 * created socket, and POSIX does not want network
		 * errors returned from accept().
		 */
		inet_csk_reqsk_queue_drop(sk, req, prev);
		goto out;

	case DCCP_REQUESTING:
	case DCCP_RESPOND:
		if (!sock_owned_by_user(sk)) {
			DCCP_INC_STATS_BH(DCCP_MIB_ATTEMPTFAILS);
			sk->sk_err = err;

			sk->sk_error_report(sk);

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

	/* If we've already connected we will keep trying
	 * until we time out, or the user gives up.
	 *
	 * rfc1122 4.2.3.9 allows to consider as hard errors
	 * only PROTO_UNREACH and PORT_UNREACH (well, FRAG_FAILED too,
	 * but it is obsoleted by pmtu discovery).
	 *
	 * Note, that in modern internet, where routing is unreliable
	 * and in each dark corner broken firewalls sit, sending random
	 * errors ordered by their masters even this two messages finally lose
	 * their original sense (even Linux sends invalid PORT_UNREACHs)
	 *
	 * Now we are in compliance with RFCs.
	 *							--ANK (980905)
	 */

	inet = inet_sk(sk);
	if (!sock_owned_by_user(sk) && inet->recverr) {
		sk->sk_err = err;
		sk->sk_error_report(sk);
	} else /* Only an error on timeout */
		sk->sk_err_soft = err;
out:
	bh_unlock_sock(sk);
	sock_put(sk);
}

449
/* This routine computes an IPv4 DCCP checksum. */
450
void dccp_v4_send_check(struct sock *sk, int len, struct sk_buff *skb)
451 452 453 454 455 456 457
{
	const struct inet_sock *inet = inet_sk(sk);
	struct dccp_hdr *dh = dccp_hdr(skb);

	dh->dccph_checksum = dccp_v4_checksum(skb, inet->saddr, inet->daddr);
}

458 459
EXPORT_SYMBOL_GPL(dccp_v4_send_check);

460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475
int dccp_v4_send_reset(struct sock *sk, enum dccp_reset_codes code)
{
	struct sk_buff *skb;
	/*
	 * FIXME: what if rebuild_header fails?
	 * Should we be doing a rebuild_header here?
	 */
	int err = inet_sk_rebuild_header(sk);

	if (err != 0)
		return err;

	skb = dccp_make_reset(sk, sk->sk_dst_cache, code);
	if (skb != NULL) {
		const struct inet_sock *inet = inet_sk(sk);

H
Herbert Xu 已提交
476
		memset(&(IPCB(skb)->opt), 0, sizeof(IPCB(skb)->opt));
477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502
		err = ip_build_and_send_pkt(skb, sk,
					    inet->saddr, inet->daddr, NULL);
		if (err == NET_XMIT_CN)
			err = 0;
	}

	return err;
}

static inline u64 dccp_v4_init_sequence(const struct sock *sk,
					const struct sk_buff *skb)
{
	return secure_dccp_sequence_number(skb->nh.iph->daddr,
					   skb->nh.iph->saddr,
					   dccp_hdr(skb)->dccph_dport,
					   dccp_hdr(skb)->dccph_sport);
}

int dccp_v4_conn_request(struct sock *sk, struct sk_buff *skb)
{
	struct inet_request_sock *ireq;
	struct dccp_sock dp;
	struct request_sock *req;
	struct dccp_request_sock *dreq;
	const __u32 saddr = skb->nh.iph->saddr;
	const __u32 daddr = skb->nh.iph->daddr;
503
 	const __u32 service = dccp_hdr_request(skb)->dccph_req_service;
504 505
	struct dccp_skb_cb *dcb = DCCP_SKB_CB(skb);
	__u8 reset_code = DCCP_RESET_CODE_TOO_BUSY;
506 507 508

	/* Never answer to DCCP_PKT_REQUESTs send to broadcast or multicast */
	if (((struct rtable *)skb->dst)->rt_flags &
509 510
	    (RTCF_BROADCAST | RTCF_MULTICAST)) {
		reset_code = DCCP_RESET_CODE_NO_CONNECTION;
511
		goto drop;
512
	}
513

514 515 516 517
	if (dccp_bad_service_code(sk, service)) {
		reset_code = DCCP_RESET_CODE_BAD_SERVICE_CODE;
		goto drop;
 	}
518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538
	/*
	 * TW buckets are converted to open requests without
	 * limitations, they conserve resources and peer is
	 * evidently real one.
	 */
	if (inet_csk_reqsk_queue_is_full(sk))
		goto drop;

	/*
	 * Accept backlog is full. If we have already queued enough
	 * of warm entries in syn queue, drop request. It is better than
	 * clogging syn queue with openreqs with exponentially increasing
	 * timeout.
	 */
	if (sk_acceptq_is_full(sk) && inet_csk_reqsk_queue_young(sk) > 1)
		goto drop;

	req = reqsk_alloc(sk->sk_prot->rsk_prot);
	if (req == NULL)
		goto drop;

539 540
	if (dccp_parse_options(sk, skb))
		goto drop;
541 542 543 544 545 546

	dccp_openreq_init(req, &dp, skb);

	ireq = inet_rsk(req);
	ireq->loc_addr = daddr;
	ireq->rmt_addr = saddr;
547 548
	req->rcv_wnd	= 100; /* Fake, option parsing will get the
				  right value */
549 550 551 552 553 554 555 556 557 558 559
	ireq->opt	= NULL;

	/* 
	 * Step 3: Process LISTEN state
	 *
	 * Set S.ISR, S.GSR, S.SWL, S.SWH from packet or Init Cookie
	 *
	 * In fact we defer setting S.GSR, S.SWL, S.SWH to
	 * dccp_create_openreq_child.
	 */
	dreq = dccp_rsk(req);
560 561 562
	dreq->dreq_isr	   = dcb->dccpd_seq;
	dreq->dreq_iss	   = dccp_v4_init_sequence(sk, skb);
	dreq->dreq_service = service;
563

564
	if (dccp_v4_send_response(sk, req, NULL))
565 566 567 568 569 570
		goto drop_and_free;

	inet_csk_reqsk_queue_hash_add(sk, req, DCCP_TIMEOUT_INIT);
	return 0;

drop_and_free:
571
	reqsk_free(req);
572 573
drop:
	DCCP_INC_STATS_BH(DCCP_MIB_ATTEMPTFAILS);
574
	dcb->dccpd_reset_code = reset_code;
575 576 577
	return -1;
}

578 579
EXPORT_SYMBOL_GPL(dccp_v4_conn_request);

580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633
/*
 * The three way handshake has completed - we got a valid ACK or DATAACK -
 * now create the new socket.
 *
 * This is the equivalent of TCP's tcp_v4_syn_recv_sock
 */
struct sock *dccp_v4_request_recv_sock(struct sock *sk, struct sk_buff *skb,
				       struct request_sock *req,
				       struct dst_entry *dst)
{
	struct inet_request_sock *ireq;
	struct inet_sock *newinet;
	struct dccp_sock *newdp;
	struct sock *newsk;

	if (sk_acceptq_is_full(sk))
		goto exit_overflow;

	if (dst == NULL && (dst = inet_csk_route_req(sk, req)) == NULL)
		goto exit;

	newsk = dccp_create_openreq_child(sk, req, skb);
	if (newsk == NULL)
		goto exit;

	sk_setup_caps(newsk, dst);

	newdp		   = dccp_sk(newsk);
	newinet		   = inet_sk(newsk);
	ireq		   = inet_rsk(req);
	newinet->daddr	   = ireq->rmt_addr;
	newinet->rcv_saddr = ireq->loc_addr;
	newinet->saddr	   = ireq->loc_addr;
	newinet->opt	   = ireq->opt;
	ireq->opt	   = NULL;
	newinet->mc_index  = inet_iif(skb);
	newinet->mc_ttl	   = skb->nh.iph->ttl;
	newinet->id	   = jiffies;

	dccp_sync_mss(newsk, dst_mtu(dst));

	__inet_hash(&dccp_hashinfo, newsk, 0);
	__inet_inherit_port(&dccp_hashinfo, sk, newsk);

	return newsk;

exit_overflow:
	NET_INC_STATS_BH(LINUX_MIB_LISTENOVERFLOWS);
exit:
	NET_INC_STATS_BH(LINUX_MIB_LISTENDROPS);
	dst_release(dst);
	return NULL;
}

634 635
EXPORT_SYMBOL_GPL(dccp_v4_request_recv_sock);

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
static struct sock *dccp_v4_hnd_req(struct sock *sk, struct sk_buff *skb)
{
	const struct dccp_hdr *dh = dccp_hdr(skb);
	const struct iphdr *iph = skb->nh.iph;
	struct sock *nsk;
	struct request_sock **prev;
	/* Find possible connection requests. */
	struct request_sock *req = inet_csk_search_req(sk, &prev,
						       dh->dccph_sport,
						       iph->saddr, iph->daddr);
	if (req != NULL)
		return dccp_check_req(sk, skb, req, prev);

	nsk = __inet_lookup_established(&dccp_hashinfo,
					iph->saddr, dh->dccph_sport,
					iph->daddr, ntohs(dh->dccph_dport),
					inet_iif(skb));
	if (nsk != NULL) {
		if (nsk->sk_state != DCCP_TIME_WAIT) {
			bh_lock_sock(nsk);
			return nsk;
		}
		inet_twsk_put((struct inet_timewait_sock *)nsk);
		return NULL;
	}

	return sk;
}

665 666
int dccp_v4_checksum(const struct sk_buff *skb, const u32 saddr,
		     const u32 daddr)
667
{
668
	const struct dccp_hdr* dh = dccp_hdr(skb);
669 670 671 672 673 674 675
	int checksum_len;
	u32 tmp;

	if (dh->dccph_cscov == 0)
		checksum_len = skb->len;
	else {
		checksum_len = (dh->dccph_cscov + dh->dccph_x) * sizeof(u32);
676 677
		checksum_len = checksum_len < skb->len ? checksum_len :
							 skb->len;
678 679 680
	}

	tmp = csum_partial((unsigned char *)dh, checksum_len, 0);
681 682
	return csum_tcpudp_magic(saddr, daddr, checksum_len,
				 IPPROTO_DCCP, tmp);
683 684
}

685 686
static int dccp_v4_verify_checksum(struct sk_buff *skb,
				   const u32 saddr, const u32 daddr)
687
{
688 689 690
	struct dccp_hdr *dh = dccp_hdr(skb);
	int checksum_len;
	u32 tmp;
691

692 693 694 695
	if (dh->dccph_cscov == 0)
		checksum_len = skb->len;
	else {
		checksum_len = (dh->dccph_cscov + dh->dccph_x) * sizeof(u32);
696 697
		checksum_len = checksum_len < skb->len ? checksum_len :
							 skb->len;
698 699
	}
	tmp = csum_partial((unsigned char *)dh, checksum_len, 0);
700 701
	return csum_tcpudp_magic(saddr, daddr, checksum_len,
				 IPPROTO_DCCP, tmp) == 0 ? 0 : -1;
702 703 704 705 706 707 708 709 710 711 712 713 714 715
}

static struct dst_entry* dccp_v4_route_skb(struct sock *sk,
					   struct sk_buff *skb)
{
	struct rtable *rt;
	struct flowi fl = { .oif = ((struct rtable *)skb->dst)->rt_iif,
			    .nl_u = { .ip4_u =
				      { .daddr = skb->nh.iph->saddr,
					.saddr = skb->nh.iph->daddr,
					.tos = RT_CONN_FLAGS(sk) } },
			    .proto = sk->sk_protocol,
			    .uli_u = { .ports =
				       { .sport = dccp_hdr(skb)->dccph_dport,
716 717 718
					 .dport = dccp_hdr(skb)->dccph_sport }
			   	     }
			  };
719 720 721 722 723 724 725 726 727

	if (ip_route_output_flow(&rt, &fl, sk, 0)) {
		IP_INC_STATS_BH(IPSTATS_MIB_OUTNOROUTES);
		return NULL;
	}

	return &rt->u.dst;
}

728
static void dccp_v4_ctl_send_reset(struct sk_buff *rxskb)
729 730 731 732 733 734 735 736
{
	int err;
	struct dccp_hdr *rxdh = dccp_hdr(rxskb), *dh;
	const int dccp_hdr_reset_len = sizeof(struct dccp_hdr) +
				       sizeof(struct dccp_hdr_ext) +
				       sizeof(struct dccp_hdr_reset);
	struct sk_buff *skb;
	struct dst_entry *dst;
737
	u64 seqno;
738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767

	/* Never send a reset in response to a reset. */
	if (rxdh->dccph_type == DCCP_PKT_RESET)
		return;

	if (((struct rtable *)rxskb->dst)->rt_type != RTN_LOCAL)
		return;

	dst = dccp_v4_route_skb(dccp_ctl_socket->sk, rxskb);
	if (dst == NULL)
		return;

	skb = alloc_skb(MAX_DCCP_HEADER + 15, GFP_ATOMIC);
	if (skb == NULL)
		goto out;

	/* Reserve space for headers. */
	skb_reserve(skb, MAX_DCCP_HEADER);
	skb->dst = dst_clone(dst);

	skb->h.raw = skb_push(skb, dccp_hdr_reset_len);
	dh = dccp_hdr(skb);
	memset(dh, 0, dccp_hdr_reset_len);

	/* Build DCCP header and checksum it. */
	dh->dccph_type	   = DCCP_PKT_RESET;
	dh->dccph_sport	   = rxdh->dccph_dport;
	dh->dccph_dport	   = rxdh->dccph_sport;
	dh->dccph_doff	   = dccp_hdr_reset_len / 4;
	dh->dccph_x	   = 1;
768 769
	dccp_hdr_reset(skb)->dccph_reset_code =
				DCCP_SKB_CB(rxskb)->dccpd_reset_code;
770

771 772 773 774 775 776
	/* See "8.3.1. Abnormal Termination" in draft-ietf-dccp-spec-11 */
	seqno = 0;
	if (DCCP_SKB_CB(rxskb)->dccpd_ack_seq != DCCP_PKT_WITHOUT_ACK_SEQ)
		dccp_set_seqno(&seqno, DCCP_SKB_CB(rxskb)->dccpd_ack_seq + 1);

	dccp_hdr_set_seq(dh, seqno);
777 778
	dccp_hdr_set_ack(dccp_hdr_ack_bits(skb),
			 DCCP_SKB_CB(rxskb)->dccpd_seq);
779

780 781
	dh->dccph_checksum = dccp_v4_checksum(skb, rxskb->nh.iph->saddr,
					      rxskb->nh.iph->daddr);
782 783 784

	bh_lock_sock(dccp_ctl_socket->sk);
	err = ip_build_and_send_pkt(skb, dccp_ctl_socket->sk,
785 786
				    rxskb->nh.iph->daddr,
				    rxskb->nh.iph->saddr, NULL);
787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809
	bh_unlock_sock(dccp_ctl_socket->sk);

	if (err == NET_XMIT_CN || err == 0) {
		DCCP_INC_STATS_BH(DCCP_MIB_OUTSEGS);
		DCCP_INC_STATS_BH(DCCP_MIB_OUTRSTS);
	}
out:
	 dst_release(dst);
}

int dccp_v4_do_rcv(struct sock *sk, struct sk_buff *skb)
{
	struct dccp_hdr *dh = dccp_hdr(skb);

	if (sk->sk_state == DCCP_OPEN) { /* Fast path */
		if (dccp_rcv_established(sk, skb, dh, skb->len))
			goto reset;
		return 0;
	}

	/*
	 *  Step 3: Process LISTEN state
	 *     If S.state == LISTEN,
810 811
	 *	  If P.type == Request or P contains a valid Init Cookie
	 *	  	option,
812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827
	 *	     * Must scan the packet's options to check for an Init
	 *		Cookie.  Only the Init Cookie is processed here,
	 *		however; other options are processed in Step 8.  This
	 *		scan need only be performed if the endpoint uses Init
	 *		Cookies *
	 *	     * Generate a new socket and switch to that socket *
	 *	     Set S := new socket for this port pair
	 *	     S.state = RESPOND
	 *	     Choose S.ISS (initial seqno) or set from Init Cookie
	 *	     Set S.ISR, S.GSR, S.SWL, S.SWH from packet or Init Cookie
	 *	     Continue with S.state == RESPOND
	 *	     * A Response packet will be generated in Step 11 *
	 *	  Otherwise,
	 *	     Generate Reset(No Connection) unless P.type == Reset
	 *	     Drop packet and return
	 *
828 829
	 * NOTE: the check for the packet types is done in
	 *	 dccp_rcv_state_process
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
	 */
	if (sk->sk_state == DCCP_LISTEN) {
		struct sock *nsk = dccp_v4_hnd_req(sk, skb);

		if (nsk == NULL)
			goto discard;

		if (nsk != sk) {
			if (dccp_child_process(sk, nsk, skb))
				goto reset;
			return 0;
		}
	}

	if (dccp_rcv_state_process(sk, skb, dh, skb->len))
		goto reset;
	return 0;

reset:
	dccp_v4_ctl_send_reset(skb);
discard:
	kfree_skb(skb);
	return 0;
}

855 856 857
EXPORT_SYMBOL_GPL(dccp_v4_do_rcv);

int dccp_invalid_packet(struct sk_buff *skb)
858 859 860 861 862 863 864
{
	const struct dccp_hdr *dh;

	if (skb->pkt_type != PACKET_HOST)
		return 1;

	if (!pskb_may_pull(skb, sizeof(struct dccp_hdr))) {
865
		LIMIT_NETDEBUG(KERN_WARNING "DCCP: pskb_may_pull failed\n");
866 867 868 869 870 871 872
		return 1;
	}

	dh = dccp_hdr(skb);

	/* If the packet type is not understood, drop packet and return */
	if (dh->dccph_type >= DCCP_PKT_INVALID) {
873
		LIMIT_NETDEBUG(KERN_WARNING "DCCP: invalid packet type\n");
874 875 876 877 878 879 880 881
		return 1;
	}

	/*
	 * If P.Data Offset is too small for packet type, or too large for
	 * packet, drop packet and return
	 */
	if (dh->dccph_doff < dccp_hdr_len(skb) / sizeof(u32)) {
882 883 884
		LIMIT_NETDEBUG(KERN_WARNING "DCCP: P.Data Offset(%u) "
					    "too small 1\n",
			       dh->dccph_doff);
885 886 887 888
		return 1;
	}

	if (!pskb_may_pull(skb, dh->dccph_doff * sizeof(u32))) {
889 890 891
		LIMIT_NETDEBUG(KERN_WARNING "DCCP: P.Data Offset(%u) "
					    "too small 2\n",
			       dh->dccph_doff);
892 893 894 895 896 897 898 899 900 901 902 903 904
		return 1;
	}

	dh = dccp_hdr(skb);

	/*
	 * If P.type is not Data, Ack, or DataAck and P.X == 0 (the packet
	 * has short sequence numbers), drop packet and return
	 */
	if (dh->dccph_x == 0 &&
	    dh->dccph_type != DCCP_PKT_DATA &&
	    dh->dccph_type != DCCP_PKT_ACK &&
	    dh->dccph_type != DCCP_PKT_DATAACK) {
905 906 907
		LIMIT_NETDEBUG(KERN_WARNING "DCCP: P.type (%s) not Data, Ack "
					    "nor DataAck and P.X == 0\n",
			       dccp_packet_name(dh->dccph_type));
908 909 910 911 912 913
		return 1;
	}

	return 0;
}

914 915
EXPORT_SYMBOL_GPL(dccp_invalid_packet);

916 917 918 919 920 921 922 923 924 925 926
/* this is called when real data arrives */
int dccp_v4_rcv(struct sk_buff *skb)
{
	const struct dccp_hdr *dh;
	struct sock *sk;

	/* Step 1: Check header basics: */

	if (dccp_invalid_packet(skb))
		goto discard_it;

927 928 929 930 931 932 933 934
	/* If the header checksum is incorrect, drop packet and return */
	if (dccp_v4_verify_checksum(skb, skb->nh.iph->saddr,
				    skb->nh.iph->daddr) < 0) {
		LIMIT_NETDEBUG(KERN_WARNING "%s: incorrect header checksum\n",
			       __FUNCTION__);
		goto discard_it;
	}

935 936 937 938 939 940 941 942 943 944 945
	dh = dccp_hdr(skb);

	DCCP_SKB_CB(skb)->dccpd_seq  = dccp_hdr_seq(skb);
	DCCP_SKB_CB(skb)->dccpd_type = dh->dccph_type;

	dccp_pr_debug("%8.8s "
		      "src=%u.%u.%u.%u@%-5d "
		      "dst=%u.%u.%u.%u@%-5d seq=%llu",
		      dccp_packet_name(dh->dccph_type),
		      NIPQUAD(skb->nh.iph->saddr), ntohs(dh->dccph_sport),
		      NIPQUAD(skb->nh.iph->daddr), ntohs(dh->dccph_dport),
946
		      (unsigned long long) DCCP_SKB_CB(skb)->dccpd_seq);
947 948 949 950 951 952

	if (dccp_packet_without_ack(skb)) {
		DCCP_SKB_CB(skb)->dccpd_ack_seq = DCCP_PKT_WITHOUT_ACK_SEQ;
		dccp_pr_debug_cat("\n");
	} else {
		DCCP_SKB_CB(skb)->dccpd_ack_seq = dccp_hdr_ack_seq(skb);
953 954 955
		dccp_pr_debug_cat(", ack=%llu\n",
				  (unsigned long long)
				  DCCP_SKB_CB(skb)->dccpd_ack_seq);
956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984
	}

	/* Step 2:
	 * 	Look up flow ID in table and get corresponding socket */
	sk = __inet_lookup(&dccp_hashinfo,
			   skb->nh.iph->saddr, dh->dccph_sport,
			   skb->nh.iph->daddr, ntohs(dh->dccph_dport),
			   inet_iif(skb));

	/* 
	 * Step 2:
	 * 	If no socket ...
	 *		Generate Reset(No Connection) unless P.type == Reset
	 *		Drop packet and return
	 */
	if (sk == NULL) {
		dccp_pr_debug("failed to look up flow ID in table and "
			      "get corresponding socket\n");
		goto no_dccp_socket;
	}

	/* 
	 * Step 2:
	 * 	... or S.state == TIMEWAIT,
	 *		Generate Reset(No Connection) unless P.type == Reset
	 *		Drop packet and return
	 */
	       
	if (sk->sk_state == DCCP_TIME_WAIT) {
985 986 987
		dccp_pr_debug("sk->sk_state == DCCP_TIME_WAIT: "
			      "do_time_wait\n");
                goto do_time_wait;
988 989
	}

990
	if (!xfrm4_policy_check(sk, XFRM_POLICY_IN, skb))
991
		goto discard_and_relse;
992
	nf_reset(skb);
993

994
	return sk_receive_skb(sk, skb);
995 996 997 998 999 1000 1001 1002 1003 1004

no_dccp_socket:
	if (!xfrm4_policy_check(NULL, XFRM_POLICY_IN, skb))
		goto discard_it;
	/*
	 * Step 2:
	 *		Generate Reset(No Connection) unless P.type == Reset
	 *		Drop packet and return
	 */
	if (dh->dccph_type != DCCP_PKT_RESET) {
1005 1006
		DCCP_SKB_CB(skb)->dccpd_reset_code =
					DCCP_RESET_CODE_NO_CONNECTION;
1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017
		dccp_v4_ctl_send_reset(skb);
	}

discard_it:
	/* Discard frame. */
	kfree_skb(skb);
	return 0;

discard_and_relse:
	sock_put(sk);
	goto discard_it;
1018 1019 1020 1021

do_time_wait:
	inet_twsk_put((struct inet_timewait_sock *)sk);
	goto no_dccp_socket;
1022 1023
}

1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036
struct inet_connection_sock_af_ops dccp_ipv4_af_ops = {
	.queue_xmit	= ip_queue_xmit,
	.send_check	= dccp_v4_send_check,
	.rebuild_header	= inet_sk_rebuild_header,
	.conn_request	= dccp_v4_conn_request,
	.syn_recv_sock	= dccp_v4_request_recv_sock,
	.net_header_len	= sizeof(struct iphdr),
	.setsockopt	= ip_setsockopt,
	.getsockopt	= ip_getsockopt,
	.addr2sockaddr	= inet_csk_addr2sockaddr,
	.sockaddr_len	= sizeof(struct sockaddr_in),
};

1037
int dccp_v4_init_sock(struct sock *sk)
1038 1039
{
	struct dccp_sock *dp = dccp_sk(sk);
1040
	struct inet_connection_sock *icsk = inet_csk(sk);
1041 1042 1043
	static int dccp_ctl_socket_init = 1;

	dccp_options_init(&dp->dccps_options);
1044
	do_gettimeofday(&dp->dccps_epoch);
1045 1046 1047 1048 1049 1050 1051 1052 1053

	/*
	 * FIXME: We're hardcoding the CCID, and doing this at this point makes
	 * the listening (master) sock get CCID control blocks, which is not
	 * necessary, but for now, to not mess with the test userspace apps,
	 * lets leave it here, later the real solution is to do this in a
	 * setsockopt(CCIDs-I-want/accept). -acme
	 */
	if (likely(!dccp_ctl_socket_init)) {
1054 1055
		int rc;

1056 1057 1058 1059 1060
		if (dp->dccps_options.dccpo_send_ack_vector) {
			dp->dccps_hc_rx_ackvec = dccp_ackvec_alloc(GFP_KERNEL);
			if (dp->dccps_hc_rx_ackvec == NULL)
				return -ENOMEM;
		}
1061
		dp->dccps_hc_rx_ccid = ccid_init(dp->dccps_options.dccpo_rx_ccid,
1062
						 sk);
1063
		dp->dccps_hc_tx_ccid = ccid_init(dp->dccps_options.dccpo_tx_ccid,
1064
						 sk);
1065 1066 1067 1068
	    	if (dp->dccps_hc_rx_ccid == NULL ||
		    dp->dccps_hc_tx_ccid == NULL) {
			ccid_exit(dp->dccps_hc_rx_ccid, sk);
			ccid_exit(dp->dccps_hc_tx_ccid, sk);
1069 1070 1071 1072
			if (dp->dccps_options.dccpo_send_ack_vector) {
				dccp_ackvec_free(dp->dccps_hc_rx_ackvec);
				dp->dccps_hc_rx_ackvec = NULL;
			}
1073 1074 1075
			dp->dccps_hc_rx_ccid = dp->dccps_hc_tx_ccid = NULL;
			return -ENOMEM;
		}
1076 1077 1078 1079 1080 1081 1082 1083

		rc = dccp_feat_init(sk);
		if (rc)
			return rc;
	} else {
		/* control socket doesn't need feat nego */
		INIT_LIST_HEAD(&dp->dccps_options.dccpo_pending);
		INIT_LIST_HEAD(&dp->dccps_options.dccpo_conf);
1084
		dccp_ctl_socket_init = 0;
1085
	}
1086 1087

	dccp_init_xmit_timers(sk);
1088
	icsk->icsk_rto = DCCP_TIMEOUT_INIT;
1089
	sk->sk_state = DCCP_CLOSED;
1090
	sk->sk_write_space = dccp_write_space;
1091 1092
	icsk->icsk_af_ops = &dccp_ipv4_af_ops;
	icsk->icsk_sync_mss = dccp_sync_mss;
1093 1094
	dp->dccps_mss_cache = 536;
	dp->dccps_role = DCCP_ROLE_UNDEFINED;
1095
	dp->dccps_service = DCCP_SERVICE_INVALID_VALUE;
1096
	dp->dccps_l_ack_ratio = dp->dccps_r_ack_ratio = 1;
1097 1098 1099 1100

	return 0;
}

1101 1102 1103
EXPORT_SYMBOL_GPL(dccp_v4_init_sock);

int dccp_v4_destroy_sock(struct sock *sk)
1104 1105 1106 1107
{
	struct dccp_sock *dp = dccp_sk(sk);

	/*
I
Ian McDonald 已提交
1108
	 * DCCP doesn't use sk_write_queue, just sk_send_head
1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119
	 * for retransmissions
	 */
	if (sk->sk_send_head != NULL) {
		kfree_skb(sk->sk_send_head);
		sk->sk_send_head = NULL;
	}

	/* Clean up a referenced DCCP bind bucket. */
	if (inet_csk(sk)->icsk_bind_hash != NULL)
		inet_put_port(&dccp_hashinfo, sk);

J
Jesper Juhl 已提交
1120 1121
	kfree(dp->dccps_service_list);
	dp->dccps_service_list = NULL;
1122

1123 1124
	ccid_hc_rx_exit(dp->dccps_hc_rx_ccid, sk);
	ccid_hc_tx_exit(dp->dccps_hc_tx_ccid, sk);
1125 1126 1127 1128
	if (dp->dccps_options.dccpo_send_ack_vector) {
		dccp_ackvec_free(dp->dccps_hc_rx_ackvec);
		dp->dccps_hc_rx_ackvec = NULL;
	}
1129 1130 1131 1132
	ccid_exit(dp->dccps_hc_rx_ccid, sk);
	ccid_exit(dp->dccps_hc_tx_ccid, sk);
	dp->dccps_hc_rx_ccid = dp->dccps_hc_tx_ccid = NULL;

1133 1134 1135
	/* clean up feature negotiation state */
	dccp_feat_clean(sk);

1136 1137 1138
	return 0;
}

1139 1140
EXPORT_SYMBOL_GPL(dccp_v4_destroy_sock);

1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154
static void dccp_v4_reqsk_destructor(struct request_sock *req)
{
	kfree(inet_rsk(req)->opt);
}

static struct request_sock_ops dccp_request_sock_ops = {
	.family		= PF_INET,
	.obj_size	= sizeof(struct dccp_request_sock),
	.rtx_syn_ack	= dccp_v4_send_response,
	.send_ack	= dccp_v4_reqsk_send_ack,
	.destructor	= dccp_v4_reqsk_destructor,
	.send_reset	= dccp_v4_ctl_send_reset,
};

1155 1156 1157 1158
static struct timewait_sock_ops dccp_timewait_sock_ops = {
	.twsk_obj_size	= sizeof(struct inet_timewait_sock),
};

1159
struct proto dccp_prot = {
1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172
	.name			= "DCCP",
	.owner			= THIS_MODULE,
	.close			= dccp_close,
	.connect		= dccp_v4_connect,
	.disconnect		= dccp_disconnect,
	.ioctl			= dccp_ioctl,
	.init			= dccp_v4_init_sock,
	.setsockopt		= dccp_setsockopt,
	.getsockopt		= dccp_getsockopt,
	.sendmsg		= dccp_sendmsg,
	.recvmsg		= dccp_recvmsg,
	.backlog_rcv		= dccp_v4_do_rcv,
	.hash			= dccp_v4_hash,
1173
	.unhash			= dccp_unhash,
1174 1175 1176 1177 1178 1179 1180 1181
	.accept			= inet_csk_accept,
	.get_port		= dccp_v4_get_port,
	.shutdown		= dccp_shutdown,
	.destroy		= dccp_v4_destroy_sock,
	.orphan_count		= &dccp_orphan_count,
	.max_header		= MAX_DCCP_HEADER,
	.obj_size		= sizeof(struct dccp_sock),
	.rsk_prot		= &dccp_request_sock_ops,
1182
	.twsk_prot		= &dccp_timewait_sock_ops,
1183
};
1184 1185

EXPORT_SYMBOL_GPL(dccp_prot);