udp.c 36.7 KB
Newer Older
L
Linus Torvalds 已提交
1 2
/*
 *	UDP 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 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34
 *
 *	Based on linux/ipv4/udp.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.
 *      Kazunori MIYAZAWA @USAGI:       change process style to use ip6_append_data
 *      YOSHIFUJI Hideaki @USAGI:	convert /proc/net/udp6 to seq_file.
 *
 *	This program is free software; you can redistribute it and/or
 *      modify it under the terms of the GNU General Public License
 *      as published by the Free Software Foundation; either version
 *      2 of the License, or (at your option) any later version.
 */

#include <linux/errno.h>
#include <linux/types.h>
#include <linux/socket.h>
#include <linux/sockios.h>
#include <linux/net.h>
#include <linux/in6.h>
#include <linux/netdevice.h>
#include <linux/if_arp.h>
#include <linux/ipv6.h>
#include <linux/icmpv6.h>
#include <linux/init.h>
35
#include <linux/module.h>
36
#include <linux/skbuff.h>
37
#include <linux/slab.h>
L
Linus Torvalds 已提交
38 39 40 41 42 43 44
#include <asm/uaccess.h>

#include <net/ndisc.h>
#include <net/protocol.h>
#include <net/transp_v6.h>
#include <net/ip6_route.h>
#include <net/raw.h>
45
#include <net/tcp_states.h>
L
Linus Torvalds 已提交
46 47
#include <net/ip6_checksum.h>
#include <net/xfrm.h>
48
#include <net/inet6_hashtables.h>
L
Linus Torvalds 已提交
49 50 51

#include <linux/proc_fs.h>
#include <linux/seq_file.h>
52
#include <trace/events/skb.h>
53
#include "udp_impl.h"
L
Linus Torvalds 已提交
54

55 56 57 58
int ipv6_rcv_saddr_equal(const struct sock *sk, const struct sock *sk2)
{
	const struct in6_addr *sk_rcv_saddr6 = &inet6_sk(sk)->rcv_saddr;
	const struct in6_addr *sk2_rcv_saddr6 = inet6_rcv_saddr(sk2);
59 60
	__be32 sk1_rcv_saddr = sk_rcv_saddr(sk);
	__be32 sk2_rcv_saddr = sk_rcv_saddr(sk2);
61 62 63 64 65 66 67
	int sk_ipv6only = ipv6_only_sock(sk);
	int sk2_ipv6only = inet_v6_ipv6only(sk2);
	int addr_type = ipv6_addr_type(sk_rcv_saddr6);
	int addr_type2 = sk2_rcv_saddr6 ? ipv6_addr_type(sk2_rcv_saddr6) : IPV6_ADDR_MAPPED;

	/* if both are mapped, treat as IPv4 */
	if (addr_type == IPV6_ADDR_MAPPED && addr_type2 == IPV6_ADDR_MAPPED)
68
		return (!sk2_ipv6only &&
E
Eric Dumazet 已提交
69 70
			(!sk1_rcv_saddr || !sk2_rcv_saddr ||
			  sk1_rcv_saddr == sk2_rcv_saddr));
71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86

	if (addr_type2 == IPV6_ADDR_ANY &&
	    !(sk2_ipv6only && addr_type == IPV6_ADDR_MAPPED))
		return 1;

	if (addr_type == IPV6_ADDR_ANY &&
	    !(sk_ipv6only && addr_type2 == IPV6_ADDR_MAPPED))
		return 1;

	if (sk2_rcv_saddr6 &&
	    ipv6_addr_equal(sk_rcv_saddr6, sk2_rcv_saddr6))
		return 1;

	return 0;
}

87 88 89 90 91 92 93 94
static unsigned int udp6_portaddr_hash(struct net *net,
				       const struct in6_addr *addr6,
				       unsigned int port)
{
	unsigned int hash, mix = net_hash_mix(net);

	if (ipv6_addr_any(addr6))
		hash = jhash_1word(0, mix);
B
Brian Haley 已提交
95
	else if (ipv6_addr_v4mapped(addr6))
96
		hash = jhash_1word((__force u32)addr6->s6_addr32[3], mix);
97
	else
98
		hash = jhash2((__force u32 *)addr6->s6_addr32, 4, mix);
99 100 101 102 103

	return hash ^ port;
}


104
int udp_v6_get_port(struct sock *sk, unsigned short snum)
L
Linus Torvalds 已提交
105
{
E
Eric Dumazet 已提交
106 107
	unsigned int hash2_nulladdr =
		udp6_portaddr_hash(sock_net(sk), &in6addr_any, snum);
108
	unsigned int hash2_partial =
E
Eric Dumazet 已提交
109 110
		udp6_portaddr_hash(sock_net(sk), &inet6_sk(sk)->rcv_saddr, 0);

111
	/* precompute partial secondary hash */
E
Eric Dumazet 已提交
112 113
	udp_sk(sk)->udp_portaddr_hash = hash2_partial;
	return udp_lib_get_port(sk, snum, ipv6_rcv_saddr_equal, hash2_nulladdr);
L
Linus Torvalds 已提交
114 115
}

E
Eric Dumazet 已提交
116 117 118 119 120 121 122 123 124
static void udp_v6_rehash(struct sock *sk)
{
	u16 new_hash = udp6_portaddr_hash(sock_net(sk),
					  &inet6_sk(sk)->rcv_saddr,
					  inet_sk(sk)->inet_num);

	udp_lib_rehash(sk, new_hash);
}

125 126
static inline int compute_score(struct sock *sk, struct net *net,
				unsigned short hnum,
127 128
				const struct in6_addr *saddr, __be16 sport,
				const struct in6_addr *daddr, __be16 dport,
129 130 131 132
				int dif)
{
	int score = -1;

133
	if (net_eq(sock_net(sk), net) && udp_sk(sk)->udp_port_hash == hnum &&
134 135 136 137 138
			sk->sk_family == PF_INET6) {
		struct ipv6_pinfo *np = inet6_sk(sk);
		struct inet_sock *inet = inet_sk(sk);

		score = 0;
E
Eric Dumazet 已提交
139 140
		if (inet->inet_dport) {
			if (inet->inet_dport != sport)
141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162
				return -1;
			score++;
		}
		if (!ipv6_addr_any(&np->rcv_saddr)) {
			if (!ipv6_addr_equal(&np->rcv_saddr, daddr))
				return -1;
			score++;
		}
		if (!ipv6_addr_any(&np->daddr)) {
			if (!ipv6_addr_equal(&np->daddr, saddr))
				return -1;
			score++;
		}
		if (sk->sk_bound_dev_if) {
			if (sk->sk_bound_dev_if != dif)
				return -1;
			score++;
		}
	}
	return score;
}

E
Eric Dumazet 已提交
163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206
#define SCORE2_MAX (1 + 1 + 1)
static inline int compute_score2(struct sock *sk, struct net *net,
				const struct in6_addr *saddr, __be16 sport,
				const struct in6_addr *daddr, unsigned short hnum,
				int dif)
{
	int score = -1;

	if (net_eq(sock_net(sk), net) && udp_sk(sk)->udp_port_hash == hnum &&
			sk->sk_family == PF_INET6) {
		struct ipv6_pinfo *np = inet6_sk(sk);
		struct inet_sock *inet = inet_sk(sk);

		if (!ipv6_addr_equal(&np->rcv_saddr, daddr))
			return -1;
		score = 0;
		if (inet->inet_dport) {
			if (inet->inet_dport != sport)
				return -1;
			score++;
		}
		if (!ipv6_addr_any(&np->daddr)) {
			if (!ipv6_addr_equal(&np->daddr, saddr))
				return -1;
			score++;
		}
		if (sk->sk_bound_dev_if) {
			if (sk->sk_bound_dev_if != dif)
				return -1;
			score++;
		}
	}
	return score;
}


/* called with read_rcu_lock() */
static struct sock *udp6_lib_lookup2(struct net *net,
		const struct in6_addr *saddr, __be16 sport,
		const struct in6_addr *daddr, unsigned int hnum, int dif,
		struct udp_hslot *hslot2, unsigned int slot2)
{
	struct sock *sk, *result;
	struct hlist_nulls_node *node;
207 208
	int score, badness, matches = 0, reuseport = 0;
	u32 hash = 0;
E
Eric Dumazet 已提交
209 210 211 212 213 214 215 216 217 218

begin:
	result = NULL;
	badness = -1;
	udp_portaddr_for_each_entry_rcu(sk, node, &hslot2->head) {
		score = compute_score2(sk, net, saddr, sport,
				      daddr, hnum, dif);
		if (score > badness) {
			result = sk;
			badness = score;
219 220 221 222 223 224
			reuseport = sk->sk_reuseport;
			if (reuseport) {
				hash = inet6_ehashfn(net, daddr, hnum,
						     saddr, sport);
				matches = 1;
			} else if (score == SCORE2_MAX)
E
Eric Dumazet 已提交
225
				goto exact_match;
226 227 228 229 230
		} else if (score == badness && reuseport) {
			matches++;
			if (((u64)hash * matches) >> 32 == 0)
				result = sk;
			hash = next_pseudo_random32(hash);
E
Eric Dumazet 已提交
231 232 233 234 235 236 237 238 239 240 241 242
		}
	}
	/*
	 * if the nulls value we got at the end of this lookup is
	 * not the expected one, we must restart lookup.
	 * We probably met an item that was moved to another chain.
	 */
	if (get_nulls_value(node) != slot2)
		goto begin;

	if (result) {
exact_match:
E
Eric Dumazet 已提交
243
		if (unlikely(!atomic_inc_not_zero_hint(&result->sk_refcnt, 2)))
E
Eric Dumazet 已提交
244 245 246 247 248 249 250 251 252 253
			result = NULL;
		else if (unlikely(compute_score2(result, net, saddr, sport,
				  daddr, hnum, dif) < badness)) {
			sock_put(result);
			goto begin;
		}
	}
	return result;
}

254
struct sock *__udp6_lib_lookup(struct net *net,
255 256
				      const struct in6_addr *saddr, __be16 sport,
				      const struct in6_addr *daddr, __be16 dport,
257
				      int dif, struct udp_table *udptable)
L
Linus Torvalds 已提交
258
{
259
	struct sock *sk, *result;
260
	struct hlist_nulls_node *node;
L
Linus Torvalds 已提交
261
	unsigned short hnum = ntohs(dport);
E
Eric Dumazet 已提交
262 263
	unsigned int hash2, slot2, slot = udp_hashfn(net, hnum, udptable->mask);
	struct udp_hslot *hslot2, *hslot = &udptable->hash[slot];
264 265
	int score, badness, matches = 0, reuseport = 0;
	u32 hash = 0;
266

267
	rcu_read_lock();
E
Eric Dumazet 已提交
268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284
	if (hslot->count > 10) {
		hash2 = udp6_portaddr_hash(net, daddr, hnum);
		slot2 = hash2 & udptable->mask;
		hslot2 = &udptable->hash2[slot2];
		if (hslot->count < hslot2->count)
			goto begin;

		result = udp6_lib_lookup2(net, saddr, sport,
					  daddr, hnum, dif,
					  hslot2, slot2);
		if (!result) {
			hash2 = udp6_portaddr_hash(net, &in6addr_any, hnum);
			slot2 = hash2 & udptable->mask;
			hslot2 = &udptable->hash2[slot2];
			if (hslot->count < hslot2->count)
				goto begin;

285 286
			result = udp6_lib_lookup2(net, saddr, sport,
						  &in6addr_any, hnum, dif,
E
Eric Dumazet 已提交
287 288 289 290 291
						  hslot2, slot2);
		}
		rcu_read_unlock();
		return result;
	}
292 293 294
begin:
	result = NULL;
	badness = -1;
295
	sk_nulls_for_each_rcu(sk, node, &hslot->head) {
296 297 298 299
		score = compute_score(sk, net, hnum, saddr, sport, daddr, dport, dif);
		if (score > badness) {
			result = sk;
			badness = score;
300 301 302 303 304 305 306 307 308 309 310
			reuseport = sk->sk_reuseport;
			if (reuseport) {
				hash = inet6_ehashfn(net, daddr, hnum,
						     saddr, sport);
				matches = 1;
			}
		} else if (score == badness && reuseport) {
			matches++;
			if (((u64)hash * matches) >> 32 == 0)
				result = sk;
			hash = next_pseudo_random32(hash);
L
Linus Torvalds 已提交
311 312
		}
	}
313 314 315 316 317
	/*
	 * if the nulls value we got at the end of this lookup is
	 * not the expected one, we must restart lookup.
	 * We probably met an item that was moved to another chain.
	 */
E
Eric Dumazet 已提交
318
	if (get_nulls_value(node) != slot)
319 320
		goto begin;

321
	if (result) {
E
Eric Dumazet 已提交
322
		if (unlikely(!atomic_inc_not_zero_hint(&result->sk_refcnt, 2)))
323 324 325 326 327 328 329 330
			result = NULL;
		else if (unlikely(compute_score(result, net, hnum, saddr, sport,
					daddr, dport, dif) < badness)) {
			sock_put(result);
			goto begin;
		}
	}
	rcu_read_unlock();
L
Linus Torvalds 已提交
331 332
	return result;
}
333
EXPORT_SYMBOL_GPL(__udp6_lib_lookup);
L
Linus Torvalds 已提交
334

335 336
static struct sock *__udp6_lib_lookup_skb(struct sk_buff *skb,
					  __be16 sport, __be16 dport,
337
					  struct udp_table *udptable)
338
{
339
	struct sock *sk;
340
	const struct ipv6hdr *iph = ipv6_hdr(skb);
341

342 343
	if (unlikely(sk = skb_steal_sock(skb)))
		return sk;
E
Eric Dumazet 已提交
344 345 346
	return __udp6_lib_lookup(dev_net(skb_dst(skb)->dev), &iph->saddr, sport,
				 &iph->daddr, dport, inet6_iif(skb),
				 udptable);
347 348
}

349 350 351 352 353 354 355 356
struct sock *udp6_lib_lookup(struct net *net, const struct in6_addr *saddr, __be16 sport,
			     const struct in6_addr *daddr, __be16 dport, int dif)
{
	return __udp6_lib_lookup(net, saddr, sport, daddr, dport, dif, &udp_table);
}
EXPORT_SYMBOL_GPL(udp6_lib_lookup);


L
Linus Torvalds 已提交
357 358 359 360 361
/*
 * 	This should be easy, if there is something there we
 * 	return it, otherwise we block.
 */

362
int udpv6_recvmsg(struct kiocb *iocb, struct sock *sk,
L
Linus Torvalds 已提交
363 364 365 366 367
		  struct msghdr *msg, size_t len,
		  int noblock, int flags, int *addr_len)
{
	struct ipv6_pinfo *np = inet6_sk(sk);
	struct inet_sock *inet = inet_sk(sk);
368
	struct sk_buff *skb;
369
	unsigned int ulen, copied;
370
	int peeked, off = 0;
371 372
	int err;
	int is_udplite = IS_UDPLITE(sk);
373
	int is_udp4;
374
	bool slow;
L
Linus Torvalds 已提交
375

376
	if (addr_len)
377
		*addr_len = sizeof(struct sockaddr_in6);
378

L
Linus Torvalds 已提交
379 380 381
	if (flags & MSG_ERRQUEUE)
		return ipv6_recv_error(sk, msg, len);

382 383 384
	if (np->rxpmtu && np->rxopt.bits.rxpmtu)
		return ipv6_recv_rxpmtu(sk, msg, len);

L
Linus Torvalds 已提交
385
try_again:
386
	skb = __skb_recv_datagram(sk, flags | (noblock ? MSG_DONTWAIT : 0),
387
				  &peeked, &off, &err);
L
Linus Torvalds 已提交
388 389 390
	if (!skb)
		goto out;

391
	ulen = skb->len - sizeof(struct udphdr);
392 393 394 395
	copied = len;
	if (copied > ulen)
		copied = ulen;
	else if (copied < ulen)
396
		msg->msg_flags |= MSG_TRUNC;
L
Linus Torvalds 已提交
397

398 399
	is_udp4 = (skb->protocol == htons(ETH_P_IP));

400
	/*
401 402 403
	 * If checksum is needed at all, try to do it while copying the
	 * data.  If the data is truncated, or if we only want a partial
	 * coverage checksum (UDP-Lite), do it before the copy.
404 405
	 */

406
	if (copied < ulen || UDP_SKB_CB(skb)->partial_cov) {
407
		if (udp_lib_checksum_complete(skb))
L
Linus Torvalds 已提交
408
			goto csum_copy_err;
409 410
	}

411
	if (skb_csum_unnecessary(skb))
412
		err = skb_copy_datagram_iovec(skb, sizeof(struct udphdr),
413
					      msg->msg_iov, copied);
414
	else {
L
Linus Torvalds 已提交
415 416 417 418
		err = skb_copy_and_csum_datagram_iovec(skb, sizeof(struct udphdr), msg->msg_iov);
		if (err == -EINVAL)
			goto csum_copy_err;
	}
419 420
	if (unlikely(err)) {
		trace_kfree_skb(skb, udpv6_recvmsg);
421 422 423 424 425 426 427 428 429 430 431
		if (!peeked) {
			atomic_inc(&sk->sk_drops);
			if (is_udp4)
				UDP_INC_STATS_USER(sock_net(sk),
						   UDP_MIB_INERRORS,
						   is_udplite);
			else
				UDP6_INC_STATS_USER(sock_net(sk),
						    UDP_MIB_INERRORS,
						    is_udplite);
		}
L
Linus Torvalds 已提交
432
		goto out_free;
433
	}
434 435 436 437 438 439 440 441
	if (!peeked) {
		if (is_udp4)
			UDP_INC_STATS_USER(sock_net(sk),
					UDP_MIB_INDATAGRAMS, is_udplite);
		else
			UDP6_INC_STATS_USER(sock_net(sk),
					UDP_MIB_INDATAGRAMS, is_udplite);
	}
442

443
	sock_recv_ts_and_drops(msg, sk, skb);
L
Linus Torvalds 已提交
444 445 446 447

	/* Copy the address. */
	if (msg->msg_name) {
		struct sockaddr_in6 *sin6;
448

L
Linus Torvalds 已提交
449 450
		sin6 = (struct sockaddr_in6 *) msg->msg_name;
		sin6->sin6_family = AF_INET6;
451
		sin6->sin6_port = udp_hdr(skb)->source;
L
Linus Torvalds 已提交
452 453 454
		sin6->sin6_flowinfo = 0;
		sin6->sin6_scope_id = 0;

455
		if (is_udp4)
B
Brian Haley 已提交
456 457
			ipv6_addr_set_v4mapped(ip_hdr(skb)->saddr,
					       &sin6->sin6_addr);
L
Linus Torvalds 已提交
458
		else {
A
Alexey Dobriyan 已提交
459
			sin6->sin6_addr = ipv6_hdr(skb)->saddr;
L
Linus Torvalds 已提交
460 461 462 463 464
			if (ipv6_addr_type(&sin6->sin6_addr) & IPV6_ADDR_LINKLOCAL)
				sin6->sin6_scope_id = IP6CB(skb)->iif;
		}

	}
465
	if (is_udp4) {
L
Linus Torvalds 已提交
466 467 468 469
		if (inet->cmsg_flags)
			ip_cmsg_recv(msg, skb);
	} else {
		if (np->rxopt.all)
470
			ip6_datagram_recv_ctl(sk, msg, skb);
471
	}
L
Linus Torvalds 已提交
472

473
	err = copied;
L
Linus Torvalds 已提交
474
	if (flags & MSG_TRUNC)
475
		err = ulen;
L
Linus Torvalds 已提交
476 477

out_free:
478
	skb_free_datagram_locked(sk, skb);
L
Linus Torvalds 已提交
479 480 481 482
out:
	return err;

csum_copy_err:
483
	slow = lock_sock_fast(sk);
484 485 486 487 488 489 490 491
	if (!skb_kill_datagram(sk, skb, flags)) {
		if (is_udp4)
			UDP_INC_STATS_USER(sock_net(sk),
					UDP_MIB_INERRORS, is_udplite);
		else
			UDP6_INC_STATS_USER(sock_net(sk),
					UDP_MIB_INERRORS, is_udplite);
	}
492
	unlock_sock_fast(sk, slow);
L
Linus Torvalds 已提交
493

494
	if (noblock)
L
Linus Torvalds 已提交
495
		return -EAGAIN;
496 497 498

	/* starting over for a new packet */
	msg->msg_flags &= ~MSG_TRUNC;
L
Linus Torvalds 已提交
499 500 501
	goto try_again;
}

502
void __udp6_lib_err(struct sk_buff *skb, struct inet6_skb_parm *opt,
503
		    u8 type, u8 code, int offset, __be32 info,
504
		    struct udp_table *udptable)
L
Linus Torvalds 已提交
505 506
{
	struct ipv6_pinfo *np;
507 508 509
	const struct ipv6hdr *hdr = (const struct ipv6hdr *)skb->data;
	const struct in6_addr *saddr = &hdr->saddr;
	const struct in6_addr *daddr = &hdr->daddr;
L
Linus Torvalds 已提交
510 511 512 513
	struct udphdr *uh = (struct udphdr*)(skb->data+offset);
	struct sock *sk;
	int err;

514
	sk = __udp6_lib_lookup(dev_net(skb->dev), daddr, uh->dest,
515
			       saddr, uh->source, inet6_iif(skb), udptable);
L
Linus Torvalds 已提交
516 517 518
	if (sk == NULL)
		return;

519 520
	if (type == ICMPV6_PKT_TOOBIG)
		ip6_sk_update_pmtu(skb, sk, info);
521 522
	if (type == NDISC_REDIRECT)
		ip6_sk_redirect(skb, sk);
523

L
Linus Torvalds 已提交
524 525 526 527 528 529 530 531
	np = inet6_sk(sk);

	if (!icmpv6_err_convert(type, code, &err) && !np->recverr)
		goto out;

	if (sk->sk_state != TCP_ESTABLISHED && !np->recverr)
		goto out;

532
	if (np->recverr)
L
Linus Torvalds 已提交
533
		ipv6_icmp_error(sk, skb, err, uh->dest, ntohl(info), (u8 *)(uh+1));
534

L
Linus Torvalds 已提交
535 536 537 538 539 540
	sk->sk_err = err;
	sk->sk_error_report(sk);
out:
	sock_put(sk);
}

541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562
static int __udpv6_queue_rcv_skb(struct sock *sk, struct sk_buff *skb)
{
	int rc;

	if (!ipv6_addr_any(&inet6_sk(sk)->daddr))
		sock_rps_save_rxhash(sk, skb);

	rc = sock_queue_rcv_skb(sk, skb);
	if (rc < 0) {
		int is_udplite = IS_UDPLITE(sk);

		/* Note that an ENOMEM error is charged twice */
		if (rc == -ENOMEM)
			UDP6_INC_STATS_BH(sock_net(sk),
					UDP_MIB_RCVBUFERRORS, is_udplite);
		UDP6_INC_STATS_BH(sock_net(sk), UDP_MIB_INERRORS, is_udplite);
		kfree_skb(skb);
		return -1;
	}
	return 0;
}

563
static __inline__ void udpv6_err(struct sk_buff *skb,
564 565
				 struct inet6_skb_parm *opt, u8 type,
				 u8 code, int offset, __be32 info     )
566
{
567
	__udp6_lib_err(skb, opt, type, code, offset, info, &udp_table);
568 569
}

570 571 572 573 574 575 576 577
static struct static_key udpv6_encap_needed __read_mostly;
void udpv6_encap_enable(void)
{
	if (!static_key_enabled(&udpv6_encap_needed))
		static_key_slow_inc(&udpv6_encap_needed);
}
EXPORT_SYMBOL(udpv6_encap_enable);

578
int udpv6_queue_rcv_skb(struct sock *sk, struct sk_buff *skb)
L
Linus Torvalds 已提交
579
{
580
	struct udp_sock *up = udp_sk(sk);
581
	int rc;
W
Wang Chen 已提交
582
	int is_udplite = IS_UDPLITE(sk);
583

584 585
	if (!xfrm6_policy_check(sk, XFRM_POLICY_IN, skb))
		goto drop;
L
Linus Torvalds 已提交
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
	if (static_key_false(&udpv6_encap_needed) && up->encap_type) {
		int (*encap_rcv)(struct sock *sk, struct sk_buff *skb);

		/*
		 * This is an encapsulation socket so pass the skb to
		 * the socket's udp_encap_rcv() hook. Otherwise, just
		 * fall through and pass this up the UDP socket.
		 * up->encap_rcv() returns the following value:
		 * =0 if skb was successfully passed to the encap
		 *    handler or was discarded by it.
		 * >0 if skb should be passed on to UDP.
		 * <0 if skb should be resubmitted as proto -N
		 */

		/* if we're overly short, let UDP handle it */
		encap_rcv = ACCESS_ONCE(up->encap_rcv);
		if (skb->len > sizeof(struct udphdr) && encap_rcv != NULL) {
			int ret;

			ret = encap_rcv(sk, skb);
			if (ret <= 0) {
				UDP_INC_STATS_BH(sock_net(sk),
						 UDP_MIB_INDATAGRAMS,
						 is_udplite);
				return -ret;
			}
		}

		/* FALLTHROUGH -- it's a UDP Packet */
	}

618 619 620
	/*
	 * UDP-Lite specific tests, ignored on UDP sockets (see net/ipv4/udp.c).
	 */
W
Wang Chen 已提交
621
	if ((is_udplite & UDPLITE_RECV_CC)  &&  UDP_SKB_CB(skb)->partial_cov) {
622 623 624 625 626 627 628 629 630 631 632 633 634

		if (up->pcrlen == 0) {          /* full coverage was set  */
			LIMIT_NETDEBUG(KERN_WARNING "UDPLITE6: partial coverage"
				" %d while full coverage %d requested\n",
				UDP_SKB_CB(skb)->cscov, skb->len);
			goto drop;
		}
		if (UDP_SKB_CB(skb)->cscov  <  up->pcrlen) {
			LIMIT_NETDEBUG(KERN_WARNING "UDPLITE6: coverage %d "
						    "too small, need min %d\n",
				       UDP_SKB_CB(skb)->cscov, up->pcrlen);
			goto drop;
		}
L
Linus Torvalds 已提交
635 636
	}

637
	if (rcu_access_pointer(sk->sk_filter)) {
638 639 640
		if (udp_lib_checksum_complete(skb))
			goto drop;
	}
641

642 643 644
	if (sk_rcvqueues_full(sk, skb, sk->sk_rcvbuf))
		goto drop;

645
	skb_dst_drop(skb);
646

647 648 649 650 651 652 653 654 655
	bh_lock_sock(sk);
	rc = 0;
	if (!sock_owned_by_user(sk))
		rc = __udpv6_queue_rcv_skb(sk, skb);
	else if (sk_add_backlog(sk, skb, sk->sk_rcvbuf)) {
		bh_unlock_sock(sk);
		goto drop;
	}
	bh_unlock_sock(sk);
656 657

	return rc;
658
drop:
659
	UDP6_INC_STATS_BH(sock_net(sk), UDP_MIB_INERRORS, is_udplite);
660
	atomic_inc(&sk->sk_drops);
661 662
	kfree_skb(skb);
	return -1;
L
Linus Torvalds 已提交
663 664
}

665
static struct sock *udp_v6_mcast_next(struct net *net, struct sock *sk,
666 667
				      __be16 loc_port, const struct in6_addr *loc_addr,
				      __be16 rmt_port, const struct in6_addr *rmt_addr,
L
Linus Torvalds 已提交
668 669
				      int dif)
{
670
	struct hlist_nulls_node *node;
L
Linus Torvalds 已提交
671 672 673
	struct sock *s = sk;
	unsigned short num = ntohs(loc_port);

674
	sk_nulls_for_each_from(s, node) {
L
Linus Torvalds 已提交
675 676
		struct inet_sock *inet = inet_sk(s);

677
		if (!net_eq(sock_net(s), net))
678 679
			continue;

680 681
		if (udp_sk(s)->udp_port_hash == num &&
		    s->sk_family == PF_INET6) {
L
Linus Torvalds 已提交
682
			struct ipv6_pinfo *np = inet6_sk(s);
E
Eric Dumazet 已提交
683 684
			if (inet->inet_dport) {
				if (inet->inet_dport != rmt_port)
L
Linus Torvalds 已提交
685 686 687 688 689 690 691 692 693 694
					continue;
			}
			if (!ipv6_addr_any(&np->daddr) &&
			    !ipv6_addr_equal(&np->daddr, rmt_addr))
				continue;

			if (s->sk_bound_dev_if && s->sk_bound_dev_if != dif)
				continue;

			if (!ipv6_addr_any(&np->rcv_saddr)) {
695 696
				if (!ipv6_addr_equal(&np->rcv_saddr, loc_addr))
					continue;
L
Linus Torvalds 已提交
697
			}
S
Stephen Hemminger 已提交
698
			if (!inet6_mc_check(s, loc_addr, rmt_addr))
L
Linus Torvalds 已提交
699 700 701 702 703 704 705
				continue;
			return s;
		}
	}
	return NULL;
}

706 707 708
static void flush_stack(struct sock **stack, unsigned int count,
			struct sk_buff *skb, unsigned int final)
{
709
	struct sk_buff *skb1 = NULL;
710
	struct sock *sk;
711
	unsigned int i;
712 713

	for (i = 0; i < count; i++) {
714
		sk = stack[i];
715 716 717 718 719 720 721 722
		if (likely(skb1 == NULL))
			skb1 = (i == final) ? skb : skb_clone(skb, GFP_ATOMIC);
		if (!skb1) {
			atomic_inc(&sk->sk_drops);
			UDP6_INC_STATS_BH(sock_net(sk), UDP_MIB_RCVBUFERRORS,
					  IS_UDPLITE(sk));
			UDP6_INC_STATS_BH(sock_net(sk), UDP_MIB_INERRORS,
					  IS_UDPLITE(sk));
723
		}
724 725 726

		if (skb1 && udpv6_queue_rcv_skb(sk, skb1) <= 0)
			skb1 = NULL;
727
	}
728 729
	if (unlikely(skb1))
		kfree_skb(skb1);
730
}
L
Linus Torvalds 已提交
731 732 733 734
/*
 * Note: called only from the BH handler context,
 * so we don't need to lock the hashes.
 */
735
static int __udp6_lib_mcast_deliver(struct net *net, struct sk_buff *skb,
736
		const struct in6_addr *saddr, const struct in6_addr *daddr,
737
		struct udp_table *udptable)
L
Linus Torvalds 已提交
738
{
739
	struct sock *sk, *stack[256 / sizeof(struct sock *)];
740
	const struct udphdr *uh = udp_hdr(skb);
741
	struct udp_hslot *hslot = udp_hashslot(udptable, net, ntohs(uh->dest));
L
Linus Torvalds 已提交
742
	int dif;
743
	unsigned int i, count = 0;
L
Linus Torvalds 已提交
744

745
	spin_lock(&hslot->lock);
746
	sk = sk_nulls_head(&hslot->head);
747
	dif = inet6_iif(skb);
748
	sk = udp_v6_mcast_next(net, sk, uh->dest, daddr, uh->source, saddr, dif);
749 750 751 752 753 754 755 756 757
	while (sk) {
		stack[count++] = sk;
		sk = udp_v6_mcast_next(net, sk_nulls_next(sk), uh->dest, daddr,
				       uh->source, saddr, dif);
		if (unlikely(count == ARRAY_SIZE(stack))) {
			if (!sk)
				break;
			flush_stack(stack, count, skb, ~0);
			count = 0;
H
Hideo Aoki 已提交
758
		}
L
Linus Torvalds 已提交
759
	}
760 761 762 763 764 765
	/*
	 * before releasing the lock, we must take reference on sockets
	 */
	for (i = 0; i < count; i++)
		sock_hold(stack[i]);

766
	spin_unlock(&hslot->lock);
767 768 769 770 771 772 773 774 775

	if (count) {
		flush_stack(stack, count, skb, count - 1);

		for (i = 0; i < count; i++)
			sock_put(stack[i]);
	} else {
		kfree_skb(skb);
	}
776
	return 0;
L
Linus Torvalds 已提交
777 778
}

779
int __udp6_lib_rcv(struct sk_buff *skb, struct udp_table *udptable,
780
		   int proto)
L
Linus Torvalds 已提交
781
{
782
	struct net *net = dev_net(skb->dev);
L
Linus Torvalds 已提交
783
	struct sock *sk;
784
	struct udphdr *uh;
785
	const struct in6_addr *saddr, *daddr;
L
Linus Torvalds 已提交
786 787 788
	u32 ulen = 0;

	if (!pskb_may_pull(skb, sizeof(struct udphdr)))
789
		goto discard;
L
Linus Torvalds 已提交
790

791 792
	saddr = &ipv6_hdr(skb)->saddr;
	daddr = &ipv6_hdr(skb)->daddr;
793
	uh = udp_hdr(skb);
L
Linus Torvalds 已提交
794 795

	ulen = ntohs(uh->len);
796 797
	if (ulen > skb->len)
		goto short_packet;
L
Linus Torvalds 已提交
798

799 800
	if (proto == IPPROTO_UDP) {
		/* UDP validates ulen. */
L
Linus Torvalds 已提交
801

802 803 804
		/* Check for jumbo payload */
		if (ulen == 0)
			ulen = skb->len;
L
Linus Torvalds 已提交
805

806 807
		if (ulen < sizeof(*uh))
			goto short_packet;
L
Linus Torvalds 已提交
808

809 810 811
		if (ulen < skb->len) {
			if (pskb_trim_rcsum(skb, ulen))
				goto short_packet;
812 813
			saddr = &ipv6_hdr(skb)->saddr;
			daddr = &ipv6_hdr(skb)->daddr;
814
			uh = udp_hdr(skb);
815 816
		}
	}
L
Linus Torvalds 已提交
817

818 819 820
	if (udp6_csum_init(skb, uh, proto))
		goto discard;

821 822
	/*
	 *	Multicast receive code
L
Linus Torvalds 已提交
823
	 */
824
	if (ipv6_addr_is_multicast(daddr))
825 826
		return __udp6_lib_mcast_deliver(net, skb,
				saddr, daddr, udptable);
L
Linus Torvalds 已提交
827 828 829

	/* Unicast */

830
	/*
L
Linus Torvalds 已提交
831 832 833
	 * check socket cache ... must talk to Alan about his plans
	 * for sock caches... i'll skip this for now.
	 */
834
	sk = __udp6_lib_lookup_skb(skb, uh->source, uh->dest, udptable);
835 836 837
	if (sk != NULL) {
		int ret = udpv6_queue_rcv_skb(sk, skb);
		sock_put(sk);
L
Linus Torvalds 已提交
838

839 840 841 842 843
		/* a return value > 0 means to resubmit the input, but
		 * it wants the return to be -protocol, or 0
		 */
		if (ret > 0)
			return -ret;
L
Linus Torvalds 已提交
844

S
Stephen Hemminger 已提交
845
		return 0;
L
Linus Torvalds 已提交
846
	}
847

848
	if (!xfrm6_policy_check(NULL, XFRM_POLICY_IN, skb))
849
		goto discard;
850 851

	if (udp_lib_checksum_complete(skb))
Z
Zhu Yi 已提交
852
		goto discard;
853 854 855 856 857

	UDP6_INC_STATS_BH(net, UDP_MIB_NOPORTS, proto == IPPROTO_UDPLITE);
	icmpv6_send(skb, ICMPV6_DEST_UNREACH, ICMPV6_PORT_UNREACH, 0);

	kfree_skb(skb);
S
Stephen Hemminger 已提交
858
	return 0;
L
Linus Torvalds 已提交
859

860
short_packet:
861
	LIMIT_NETDEBUG(KERN_DEBUG "UDP%sv6: short packet: From [%pI6c]:%u %d/%d to [%pI6c]:%u\n",
862
		       proto == IPPROTO_UDPLITE ? "-Lite" : "",
863 864 865 866 867 868
		       saddr,
		       ntohs(uh->source),
		       ulen,
		       skb->len,
		       daddr,
		       ntohs(uh->dest));
L
Linus Torvalds 已提交
869 870

discard:
871
	UDP6_INC_STATS_BH(net, UDP_MIB_INERRORS, proto == IPPROTO_UDPLITE);
L
Linus Torvalds 已提交
872
	kfree_skb(skb);
S
Stephen Hemminger 已提交
873
	return 0;
L
Linus Torvalds 已提交
874
}
875

876
static __inline__ int udpv6_rcv(struct sk_buff *skb)
877
{
878
	return __udp6_lib_rcv(skb, &udp_table, IPPROTO_UDP);
879 880
}

L
Linus Torvalds 已提交
881 882 883 884 885 886 887
/*
 * Throw away all pending data and cancel the corking. Socket is locked.
 */
static void udp_v6_flush_pending_frames(struct sock *sk)
{
	struct udp_sock *up = udp_sk(sk);

888 889 890
	if (up->pending == AF_INET)
		udp_flush_pending_frames(sk);
	else if (up->pending) {
L
Linus Torvalds 已提交
891 892 893
		up->len = 0;
		up->pending = 0;
		ip6_flush_pending_frames(sk);
894
	}
L
Linus Torvalds 已提交
895 896
}

897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937
/**
 * 	udp6_hwcsum_outgoing  -  handle outgoing HW checksumming
 * 	@sk: 	socket we are sending on
 * 	@skb: 	sk_buff containing the filled-in UDP header
 * 	        (checksum field must be zeroed out)
 */
static void udp6_hwcsum_outgoing(struct sock *sk, struct sk_buff *skb,
				 const struct in6_addr *saddr,
				 const struct in6_addr *daddr, int len)
{
	unsigned int offset;
	struct udphdr *uh = udp_hdr(skb);
	__wsum csum = 0;

	if (skb_queue_len(&sk->sk_write_queue) == 1) {
		/* Only one fragment on the socket.  */
		skb->csum_start = skb_transport_header(skb) - skb->head;
		skb->csum_offset = offsetof(struct udphdr, check);
		uh->check = ~csum_ipv6_magic(saddr, daddr, len, IPPROTO_UDP, 0);
	} else {
		/*
		 * HW-checksum won't work as there are two or more
		 * fragments on the socket so that all csums of sk_buffs
		 * should be together
		 */
		offset = skb_transport_offset(skb);
		skb->csum = skb_checksum(skb, offset, skb->len - offset, 0);

		skb->ip_summed = CHECKSUM_NONE;

		skb_queue_walk(&sk->sk_write_queue, skb) {
			csum = csum_add(csum, skb->csum);
		}

		uh->check = csum_ipv6_magic(saddr, daddr, len, IPPROTO_UDP,
					    csum);
		if (uh->check == 0)
			uh->check = CSUM_MANGLED_0;
	}
}

L
Linus Torvalds 已提交
938 939 940 941
/*
 *	Sending
 */

942
static int udp_v6_push_pending_frames(struct sock *sk)
L
Linus Torvalds 已提交
943 944 945
{
	struct sk_buff *skb;
	struct udphdr *uh;
946
	struct udp_sock  *up = udp_sk(sk);
L
Linus Torvalds 已提交
947
	struct inet_sock *inet = inet_sk(sk);
948
	struct flowi6 *fl6 = &inet->cork.fl.u.ip6;
L
Linus Torvalds 已提交
949
	int err = 0;
W
Wang Chen 已提交
950
	int is_udplite = IS_UDPLITE(sk);
951
	__wsum csum = 0;
L
Linus Torvalds 已提交
952 953 954 955 956 957 958 959

	/* Grab the skbuff where UDP header space exists. */
	if ((skb = skb_peek(&sk->sk_write_queue)) == NULL)
		goto out;

	/*
	 * Create a UDP header
	 */
960
	uh = udp_hdr(skb);
961 962
	uh->source = fl6->fl6_sport;
	uh->dest = fl6->fl6_dport;
L
Linus Torvalds 已提交
963 964 965
	uh->len = htons(up->len);
	uh->check = 0;

W
Wang Chen 已提交
966
	if (is_udplite)
967
		csum = udplite_csum_outgoing(sk, skb);
968
	else if (skb->ip_summed == CHECKSUM_PARTIAL) { /* UDP hardware csum */
969
		udp6_hwcsum_outgoing(sk, skb, &fl6->saddr, &fl6->daddr,
970 971 972
				     up->len);
		goto send;
	} else
973
		csum = udp_csum_outgoing(sk, skb);
L
Linus Torvalds 已提交
974

975
	/* add protocol-dependent pseudo-header */
976 977
	uh->check = csum_ipv6_magic(&fl6->saddr, &fl6->daddr,
				    up->len, fl6->flowi6_proto, csum);
L
Linus Torvalds 已提交
978
	if (uh->check == 0)
979
		uh->check = CSUM_MANGLED_0;
L
Linus Torvalds 已提交
980

981
send:
L
Linus Torvalds 已提交
982
	err = ip6_push_pending_frames(sk);
E
Eric Dumazet 已提交
983 984 985 986 987 988 989 990 991
	if (err) {
		if (err == -ENOBUFS && !inet6_sk(sk)->recverr) {
			UDP6_INC_STATS_USER(sock_net(sk),
					    UDP_MIB_SNDBUFERRORS, is_udplite);
			err = 0;
		}
	} else
		UDP6_INC_STATS_USER(sock_net(sk),
				    UDP_MIB_OUTDATAGRAMS, is_udplite);
L
Linus Torvalds 已提交
992 993 994 995 996 997
out:
	up->len = 0;
	up->pending = 0;
	return err;
}

998
int udpv6_sendmsg(struct kiocb *iocb, struct sock *sk,
L
Linus Torvalds 已提交
999 1000 1001 1002 1003 1004 1005
		  struct msghdr *msg, size_t len)
{
	struct ipv6_txoptions opt_space;
	struct udp_sock *up = udp_sk(sk);
	struct inet_sock *inet = inet_sk(sk);
	struct ipv6_pinfo *np = inet6_sk(sk);
	struct sockaddr_in6 *sin6 = (struct sockaddr_in6 *) msg->msg_name;
1006
	struct in6_addr *daddr, *final_p, final;
L
Linus Torvalds 已提交
1007 1008
	struct ipv6_txoptions *opt = NULL;
	struct ip6_flowlabel *flowlabel = NULL;
1009
	struct flowi6 fl6;
L
Linus Torvalds 已提交
1010 1011 1012 1013
	struct dst_entry *dst;
	int addr_len = msg->msg_namelen;
	int ulen = len;
	int hlimit = -1;
1014
	int tclass = -1;
1015
	int dontfrag = -1;
L
Linus Torvalds 已提交
1016 1017
	int corkreq = up->corkflag || msg->msg_flags&MSG_MORE;
	int err;
1018
	int connected = 0;
W
Wang Chen 已提交
1019
	int is_udplite = IS_UDPLITE(sk);
1020
	int (*getfrag)(void *, char *, int, int, int, struct sk_buff *);
L
Linus Torvalds 已提交
1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046

	/* destination address check */
	if (sin6) {
		if (addr_len < offsetof(struct sockaddr, sa_data))
			return -EINVAL;

		switch (sin6->sin6_family) {
		case AF_INET6:
			if (addr_len < SIN6_LEN_RFC2133)
				return -EINVAL;
			daddr = &sin6->sin6_addr;
			break;
		case AF_INET:
			goto do_udp_sendmsg;
		case AF_UNSPEC:
			msg->msg_name = sin6 = NULL;
			msg->msg_namelen = addr_len = 0;
			daddr = NULL;
			break;
		default:
			return -EINVAL;
		}
	} else if (!up->pending) {
		if (sk->sk_state != TCP_ESTABLISHED)
			return -EDESTADDRREQ;
		daddr = &np->daddr;
1047
	} else
L
Linus Torvalds 已提交
1048 1049 1050
		daddr = NULL;

	if (daddr) {
B
Brian Haley 已提交
1051
		if (ipv6_addr_v4mapped(daddr)) {
L
Linus Torvalds 已提交
1052 1053
			struct sockaddr_in sin;
			sin.sin_family = AF_INET;
E
Eric Dumazet 已提交
1054
			sin.sin_port = sin6 ? sin6->sin6_port : inet->inet_dport;
L
Linus Torvalds 已提交
1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068
			sin.sin_addr.s_addr = daddr->s6_addr32[3];
			msg->msg_name = &sin;
			msg->msg_namelen = sizeof(sin);
do_udp_sendmsg:
			if (__ipv6_only_sock(sk))
				return -ENETUNREACH;
			return udp_sendmsg(iocb, sk, msg, len);
		}
	}

	if (up->pending == AF_INET)
		return udp_sendmsg(iocb, sk, msg, len);

	/* Rough check on arithmetic overflow,
1069
	   better check is made in ip6_append_data().
L
Linus Torvalds 已提交
1070 1071 1072
	   */
	if (len > INT_MAX - sizeof(struct udphdr))
		return -EMSGSIZE;
1073

L
Linus Torvalds 已提交
1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091
	if (up->pending) {
		/*
		 * There are pending frames.
		 * The socket lock must be held while it's corked.
		 */
		lock_sock(sk);
		if (likely(up->pending)) {
			if (unlikely(up->pending != AF_INET6)) {
				release_sock(sk);
				return -EAFNOSUPPORT;
			}
			dst = NULL;
			goto do_append_data;
		}
		release_sock(sk);
	}
	ulen += sizeof(struct udphdr);

1092
	memset(&fl6, 0, sizeof(fl6));
L
Linus Torvalds 已提交
1093 1094 1095 1096 1097

	if (sin6) {
		if (sin6->sin6_port == 0)
			return -EINVAL;

1098
		fl6.fl6_dport = sin6->sin6_port;
L
Linus Torvalds 已提交
1099 1100 1101
		daddr = &sin6->sin6_addr;

		if (np->sndflow) {
1102 1103 1104
			fl6.flowlabel = sin6->sin6_flowinfo&IPV6_FLOWINFO_MASK;
			if (fl6.flowlabel&IPV6_FLOWLABEL_MASK) {
				flowlabel = fl6_sock_lookup(sk, fl6.flowlabel);
L
Linus Torvalds 已提交
1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121
				if (flowlabel == NULL)
					return -EINVAL;
				daddr = &flowlabel->dst;
			}
		}

		/*
		 * Otherwise it will be difficult to maintain
		 * sk->sk_dst_cache.
		 */
		if (sk->sk_state == TCP_ESTABLISHED &&
		    ipv6_addr_equal(daddr, &np->daddr))
			daddr = &np->daddr;

		if (addr_len >= sizeof(struct sockaddr_in6) &&
		    sin6->sin6_scope_id &&
		    ipv6_addr_type(daddr)&IPV6_ADDR_LINKLOCAL)
1122
			fl6.flowi6_oif = sin6->sin6_scope_id;
L
Linus Torvalds 已提交
1123 1124 1125 1126
	} else {
		if (sk->sk_state != TCP_ESTABLISHED)
			return -EDESTADDRREQ;

1127
		fl6.fl6_dport = inet->inet_dport;
L
Linus Torvalds 已提交
1128
		daddr = &np->daddr;
1129
		fl6.flowlabel = np->flow_label;
1130
		connected = 1;
L
Linus Torvalds 已提交
1131 1132
	}

1133 1134
	if (!fl6.flowi6_oif)
		fl6.flowi6_oif = sk->sk_bound_dev_if;
L
Linus Torvalds 已提交
1135

1136 1137
	if (!fl6.flowi6_oif)
		fl6.flowi6_oif = np->sticky_pktinfo.ipi6_ifindex;
1138

1139
	fl6.flowi6_mark = sk->sk_mark;
B
Brian Haley 已提交
1140

L
Linus Torvalds 已提交
1141 1142 1143 1144 1145
	if (msg->msg_controllen) {
		opt = &opt_space;
		memset(opt, 0, sizeof(struct ipv6_txoptions));
		opt->tot_len = sizeof(*opt);

1146 1147
		err = ip6_datagram_send_ctl(sock_net(sk), sk, msg, &fl6, opt,
					    &hlimit, &tclass, &dontfrag);
L
Linus Torvalds 已提交
1148 1149 1150 1151
		if (err < 0) {
			fl6_sock_release(flowlabel);
			return err;
		}
1152 1153
		if ((fl6.flowlabel&IPV6_FLOWLABEL_MASK) && !flowlabel) {
			flowlabel = fl6_sock_lookup(sk, fl6.flowlabel);
L
Linus Torvalds 已提交
1154 1155 1156 1157 1158
			if (flowlabel == NULL)
				return -EINVAL;
		}
		if (!(opt->opt_nflen|opt->opt_flen))
			opt = NULL;
1159
		connected = 0;
L
Linus Torvalds 已提交
1160 1161 1162
	}
	if (opt == NULL)
		opt = np->opt;
1163 1164 1165
	if (flowlabel)
		opt = fl6_merge_options(&opt_space, flowlabel, opt);
	opt = ipv6_fixup_options(&opt_space, opt);
L
Linus Torvalds 已提交
1166

1167
	fl6.flowi6_proto = sk->sk_protocol;
1168
	if (!ipv6_addr_any(daddr))
A
Alexey Dobriyan 已提交
1169
		fl6.daddr = *daddr;
1170
	else
1171 1172
		fl6.daddr.s6_addr[15] = 0x1; /* :: means loopback (BSD'ism) */
	if (ipv6_addr_any(&fl6.saddr) && !ipv6_addr_any(&np->saddr))
A
Alexey Dobriyan 已提交
1173
		fl6.saddr = np->saddr;
1174
	fl6.fl6_sport = inet->inet_sport;
1175

1176
	final_p = fl6_update_dst(&fl6, opt, &final);
1177
	if (final_p)
1178
		connected = 0;
L
Linus Torvalds 已提交
1179

1180 1181
	if (!fl6.flowi6_oif && ipv6_addr_is_multicast(&fl6.daddr)) {
		fl6.flowi6_oif = np->mcast_oif;
1182
		connected = 0;
1183 1184
	} else if (!fl6.flowi6_oif)
		fl6.flowi6_oif = np->ucast_oif;
L
Linus Torvalds 已提交
1185

1186
	security_sk_classify_flow(sk, flowi6_to_flowi(&fl6));
V
Venkat Yekkirala 已提交
1187

1188
	dst = ip6_sk_dst_lookup_flow(sk, &fl6, final_p, true);
1189 1190 1191
	if (IS_ERR(dst)) {
		err = PTR_ERR(dst);
		dst = NULL;
L
Linus Torvalds 已提交
1192
		goto out;
1193
	}
L
Linus Torvalds 已提交
1194 1195

	if (hlimit < 0) {
1196
		if (ipv6_addr_is_multicast(&fl6.daddr))
L
Linus Torvalds 已提交
1197 1198 1199 1200
			hlimit = np->mcast_hops;
		else
			hlimit = np->hop_limit;
		if (hlimit < 0)
1201
			hlimit = ip6_dst_hoplimit(dst);
L
Linus Torvalds 已提交
1202 1203
	}

G
Gerrit Renker 已提交
1204
	if (tclass < 0)
1205 1206
		tclass = np->tclass;

1207 1208 1209
	if (dontfrag < 0)
		dontfrag = np->dontfrag;

L
Linus Torvalds 已提交
1210 1211 1212 1213 1214 1215 1216 1217 1218 1219
	if (msg->msg_flags&MSG_CONFIRM)
		goto do_confirm;
back_from_confirm:

	lock_sock(sk);
	if (unlikely(up->pending)) {
		/* The socket is already corked while preparing it. */
		/* ... which is an evident application bug. --ANK */
		release_sock(sk);

1220
		LIMIT_NETDEBUG(KERN_DEBUG "udp cork app bug 2\n");
L
Linus Torvalds 已提交
1221 1222 1223 1224 1225 1226 1227 1228
		err = -EINVAL;
		goto out;
	}

	up->pending = AF_INET6;

do_append_data:
	up->len += ulen;
1229 1230
	getfrag  =  is_udplite ?  udplite_getfrag : ip_generic_getfrag;
	err = ip6_append_data(sk, getfrag, msg->msg_iov, ulen,
1231
		sizeof(struct udphdr), hlimit, tclass, opt, &fl6,
1232
		(struct rt6_info*)dst,
1233
		corkreq ? msg->msg_flags|MSG_MORE : msg->msg_flags, dontfrag);
L
Linus Torvalds 已提交
1234 1235 1236
	if (err)
		udp_v6_flush_pending_frames(sk);
	else if (!corkreq)
1237
		err = udp_v6_push_pending_frames(sk);
H
Herbert Xu 已提交
1238 1239
	else if (unlikely(skb_queue_empty(&sk->sk_write_queue)))
		up->pending = 0;
L
Linus Torvalds 已提交
1240

1241 1242 1243
	if (dst) {
		if (connected) {
			ip6_dst_store(sk, dst,
1244
				      ipv6_addr_equal(&fl6.daddr, &np->daddr) ?
1245 1246
				      &np->daddr : NULL,
#ifdef CONFIG_IPV6_SUBTREES
1247
				      ipv6_addr_equal(&fl6.saddr, &np->saddr) ?
1248 1249 1250
				      &np->saddr :
#endif
				      NULL);
1251 1252 1253
		} else {
			dst_release(dst);
		}
1254
		dst = NULL;
1255 1256
	}

L
Linus Torvalds 已提交
1257 1258 1259 1260
	if (err > 0)
		err = np->recverr ? net_xmit_errno(err) : 0;
	release_sock(sk);
out:
1261
	dst_release(dst);
L
Linus Torvalds 已提交
1262
	fl6_sock_release(flowlabel);
1263
	if (!err)
L
Linus Torvalds 已提交
1264
		return len;
1265 1266 1267 1268 1269 1270 1271 1272
	/*
	 * ENOBUFS = no kernel mem, SOCK_NOSPACE = no sndbuf space.  Reporting
	 * ENOBUFS might not be good (it's not tunable per se), but otherwise
	 * we don't have a good statistic (IpOutDiscards but it can be too many
	 * things).  We could add another new stat but at least for now that
	 * seems like overkill.
	 */
	if (err == -ENOBUFS || test_bit(SOCK_NOSPACE, &sk->sk_socket->flags)) {
1273 1274
		UDP6_INC_STATS_USER(sock_net(sk),
				UDP_MIB_SNDBUFERRORS, is_udplite);
1275
	}
L
Linus Torvalds 已提交
1276 1277 1278 1279 1280 1281 1282 1283 1284 1285
	return err;

do_confirm:
	dst_confirm(dst);
	if (!(msg->msg_flags&MSG_PROBE) || len)
		goto back_from_confirm;
	err = 0;
	goto out;
}

1286
void udpv6_destroy_sock(struct sock *sk)
L
Linus Torvalds 已提交
1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297
{
	lock_sock(sk);
	udp_v6_flush_pending_frames(sk);
	release_sock(sk);

	inet6_destroy_sock(sk);
}

/*
 *	Socket option code for UDP
 */
1298
int udpv6_setsockopt(struct sock *sk, int level, int optname,
1299
		     char __user *optval, unsigned int optlen)
1300
{
1301
	if (level == SOL_UDP  ||  level == SOL_UDPLITE)
1302 1303
		return udp_lib_setsockopt(sk, level, optname, optval, optlen,
					  udp_v6_push_pending_frames);
1304
	return ipv6_setsockopt(sk, level, optname, optval, optlen);
1305 1306 1307
}

#ifdef CONFIG_COMPAT
1308
int compat_udpv6_setsockopt(struct sock *sk, int level, int optname,
1309
			    char __user *optval, unsigned int optlen)
1310
{
1311
	if (level == SOL_UDP  ||  level == SOL_UDPLITE)
1312 1313
		return udp_lib_setsockopt(sk, level, optname, optval, optlen,
					  udp_v6_push_pending_frames);
1314
	return compat_ipv6_setsockopt(sk, level, optname, optval, optlen);
1315 1316 1317
}
#endif

1318 1319
int udpv6_getsockopt(struct sock *sk, int level, int optname,
		     char __user *optval, int __user *optlen)
1320
{
1321
	if (level == SOL_UDP  ||  level == SOL_UDPLITE)
1322
		return udp_lib_getsockopt(sk, level, optname, optval, optlen);
1323
	return ipv6_getsockopt(sk, level, optname, optval, optlen);
1324 1325 1326
}

#ifdef CONFIG_COMPAT
1327 1328
int compat_udpv6_getsockopt(struct sock *sk, int level, int optname,
			    char __user *optval, int __user *optlen)
1329
{
1330
	if (level == SOL_UDP  ||  level == SOL_UDPLITE)
1331
		return udp_lib_getsockopt(sk, level, optname, optval, optlen);
1332
	return compat_ipv6_getsockopt(sk, level, optname, optval, optlen);
1333 1334 1335
}
#endif

1336
static const struct inet6_protocol udpv6_protocol = {
L
Linus Torvalds 已提交
1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348
	.handler	=	udpv6_rcv,
	.err_handler	=	udpv6_err,
	.flags		=	INET6_PROTO_NOPOLICY|INET6_PROTO_FINAL,
};

/* ------------------------------------------------------------------------ */
#ifdef CONFIG_PROC_FS

static void udp6_sock_seq_show(struct seq_file *seq, struct sock *sp, int bucket)
{
	struct inet_sock *inet = inet_sk(sp);
	struct ipv6_pinfo *np = inet6_sk(sp);
1349
	const struct in6_addr *dest, *src;
L
Linus Torvalds 已提交
1350 1351 1352 1353
	__u16 destp, srcp;

	dest  = &np->daddr;
	src   = &np->rcv_saddr;
E
Eric Dumazet 已提交
1354 1355
	destp = ntohs(inet->inet_dport);
	srcp  = ntohs(inet->inet_sport);
L
Linus Torvalds 已提交
1356
	seq_printf(seq,
1357
		   "%5d: %08X%08X%08X%08X:%04X %08X%08X%08X%08X:%04X "
D
Dan Rosenberg 已提交
1358
		   "%02X %08X:%08X %02X:%08lX %08X %5d %8d %lu %d %pK %d\n",
L
Linus Torvalds 已提交
1359 1360 1361 1362 1363
		   bucket,
		   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,
1364
		   sp->sk_state,
1365 1366
		   sk_wmem_alloc_get(sp),
		   sk_rmem_alloc_get(sp),
L
Linus Torvalds 已提交
1367
		   0, 0L, 0,
1368 1369
		   from_kuid_munged(seq_user_ns(seq), sock_i_uid(sp)),
		   0,
L
Linus Torvalds 已提交
1370
		   sock_i_ino(sp),
E
Eric Dumazet 已提交
1371 1372
		   atomic_read(&sp->sk_refcnt), sp,
		   atomic_read(&sp->sk_drops));
L
Linus Torvalds 已提交
1373 1374
}

1375
int udp6_seq_show(struct seq_file *seq, void *v)
L
Linus Torvalds 已提交
1376 1377 1378 1379 1380 1381 1382
{
	if (v == SEQ_START_TOKEN)
		seq_printf(seq,
			   "  sl  "
			   "local_address                         "
			   "remote_address                        "
			   "st tx_queue rx_queue tr tm->when retrnsmt"
E
Eric Dumazet 已提交
1383
			   "   uid  timeout inode ref pointer drops\n");
L
Linus Torvalds 已提交
1384 1385 1386 1387 1388
	else
		udp6_sock_seq_show(seq, v, ((struct udp_iter_state *)seq->private)->bucket);
	return 0;
}

1389 1390 1391 1392 1393 1394 1395 1396
static const struct file_operations udp6_afinfo_seq_fops = {
	.owner    = THIS_MODULE,
	.open     = udp_seq_open,
	.read     = seq_read,
	.llseek   = seq_lseek,
	.release  = seq_release_net
};

L
Linus Torvalds 已提交
1397 1398 1399
static struct udp_seq_afinfo udp6_seq_afinfo = {
	.name		= "udp6",
	.family		= AF_INET6,
1400
	.udp_table	= &udp_table,
1401
	.seq_fops	= &udp6_afinfo_seq_fops,
1402 1403 1404
	.seq_ops	= {
		.show		= udp6_seq_show,
	},
L
Linus Torvalds 已提交
1405 1406
};

1407
int __net_init udp6_proc_init(struct net *net)
L
Linus Torvalds 已提交
1408
{
1409
	return udp_proc_register(net, &udp6_seq_afinfo);
L
Linus Torvalds 已提交
1410 1411
}

1412 1413
void udp6_proc_exit(struct net *net) {
	udp_proc_unregister(net, &udp6_seq_afinfo);
L
Linus Torvalds 已提交
1414 1415 1416 1417 1418 1419
}
#endif /* CONFIG_PROC_FS */

/* ------------------------------------------------------------------------ */

struct proto udpv6_prot = {
1420 1421
	.name		   = "UDPv6",
	.owner		   = THIS_MODULE,
1422
	.close		   = udp_lib_close,
1423 1424 1425 1426 1427 1428 1429 1430
	.connect	   = ip6_datagram_connect,
	.disconnect	   = udp_disconnect,
	.ioctl		   = udp_ioctl,
	.destroy	   = udpv6_destroy_sock,
	.setsockopt	   = udpv6_setsockopt,
	.getsockopt	   = udpv6_getsockopt,
	.sendmsg	   = udpv6_sendmsg,
	.recvmsg	   = udpv6_recvmsg,
1431
	.backlog_rcv	   = __udpv6_queue_rcv_skb,
1432 1433
	.hash		   = udp_lib_hash,
	.unhash		   = udp_lib_unhash,
E
Eric Dumazet 已提交
1434
	.rehash		   = udp_v6_rehash,
1435
	.get_port	   = udp_v6_get_port,
H
Hideo Aoki 已提交
1436 1437 1438 1439
	.memory_allocated  = &udp_memory_allocated,
	.sysctl_mem	   = sysctl_udp_mem,
	.sysctl_wmem	   = &sysctl_udp_wmem_min,
	.sysctl_rmem	   = &sysctl_udp_rmem_min,
1440
	.obj_size	   = sizeof(struct udp6_sock),
1441
	.slab_flags	   = SLAB_DESTROY_BY_RCU,
1442
	.h.udp_table	   = &udp_table,
1443
#ifdef CONFIG_COMPAT
1444 1445
	.compat_setsockopt = compat_udpv6_setsockopt,
	.compat_getsockopt = compat_udpv6_getsockopt,
1446
#endif
1447
	.clear_sk	   = sk_prot_clear_portaddr_nulls,
L
Linus Torvalds 已提交
1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459
};

static struct inet_protosw udpv6_protosw = {
	.type =      SOCK_DGRAM,
	.protocol =  IPPROTO_UDP,
	.prot =      &udpv6_prot,
	.ops =       &inet6_dgram_ops,
	.no_check =  UDP_CSUM_DEFAULT,
	.flags =     INET_PROTOSW_PERMANENT,
};


1460
int __init udpv6_init(void)
L
Linus Torvalds 已提交
1461
{
1462 1463
	int ret;

1464 1465
	ret = inet6_add_protocol(&udpv6_protocol, IPPROTO_UDP);
	if (ret)
1466
		goto out;
1467

1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478
	ret = inet6_register_protosw(&udpv6_protosw);
	if (ret)
		goto out_udpv6_protocol;
out:
	return ret;

out_udpv6_protocol:
	inet6_del_protocol(&udpv6_protocol, IPPROTO_UDP);
	goto out;
}

1479
void udpv6_exit(void)
1480 1481 1482
{
	inet6_unregister_protosw(&udpv6_protosw);
	inet6_del_protocol(&udpv6_protocol, IPPROTO_UDP);
L
Linus Torvalds 已提交
1483
}