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
		sin6->sin6_flowinfo = 0;

454
		if (is_udp4) {
B
Brian Haley 已提交
455 456
			ipv6_addr_set_v4mapped(ip_hdr(skb)->saddr,
					       &sin6->sin6_addr);
457 458
			sin6->sin6_scope_id = 0;
		} else {
A
Alexey Dobriyan 已提交
459
			sin6->sin6_addr = ipv6_hdr(skb)->saddr;
460 461 462
			sin6->sin6_scope_id =
				ipv6_iface_scope_id(&sin6->sin6_addr,
						    IP6CB(skb)->iif);
L
Linus Torvalds 已提交
463 464 465
		}

	}
466
	if (is_udp4) {
L
Linus Torvalds 已提交
467 468 469 470
		if (inet->cmsg_flags)
			ip_cmsg_recv(msg, skb);
	} else {
		if (np->rxopt.all)
471
			ip6_datagram_recv_ctl(sk, msg, skb);
472
	}
L
Linus Torvalds 已提交
473

474
	err = copied;
L
Linus Torvalds 已提交
475
	if (flags & MSG_TRUNC)
476
		err = ulen;
L
Linus Torvalds 已提交
477 478

out_free:
479
	skb_free_datagram_locked(sk, skb);
L
Linus Torvalds 已提交
480 481 482 483
out:
	return err;

csum_copy_err:
484
	slow = lock_sock_fast(sk);
485
	if (!skb_kill_datagram(sk, skb, flags)) {
486 487 488
		if (is_udp4) {
			UDP_INC_STATS_USER(sock_net(sk),
					UDP_MIB_CSUMERRORS, is_udplite);
489 490
			UDP_INC_STATS_USER(sock_net(sk),
					UDP_MIB_INERRORS, is_udplite);
491 492 493
		} else {
			UDP6_INC_STATS_USER(sock_net(sk),
					UDP_MIB_CSUMERRORS, is_udplite);
494 495
			UDP6_INC_STATS_USER(sock_net(sk),
					UDP_MIB_INERRORS, is_udplite);
496
		}
497
	}
498
	unlock_sock_fast(sk, slow);
L
Linus Torvalds 已提交
499

500
	if (noblock)
L
Linus Torvalds 已提交
501
		return -EAGAIN;
502 503 504

	/* starting over for a new packet */
	msg->msg_flags &= ~MSG_TRUNC;
L
Linus Torvalds 已提交
505 506 507
	goto try_again;
}

508
void __udp6_lib_err(struct sk_buff *skb, struct inet6_skb_parm *opt,
509
		    u8 type, u8 code, int offset, __be32 info,
510
		    struct udp_table *udptable)
L
Linus Torvalds 已提交
511 512
{
	struct ipv6_pinfo *np;
513 514 515
	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
516 517 518 519
	struct udphdr *uh = (struct udphdr*)(skb->data+offset);
	struct sock *sk;
	int err;

520
	sk = __udp6_lib_lookup(dev_net(skb->dev), daddr, uh->dest,
521
			       saddr, uh->source, inet6_iif(skb), udptable);
L
Linus Torvalds 已提交
522 523 524
	if (sk == NULL)
		return;

525 526
	if (type == ICMPV6_PKT_TOOBIG)
		ip6_sk_update_pmtu(skb, sk, info);
527 528
	if (type == NDISC_REDIRECT)
		ip6_sk_redirect(skb, sk);
529

L
Linus Torvalds 已提交
530 531 532 533 534 535 536 537
	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;

538
	if (np->recverr)
L
Linus Torvalds 已提交
539
		ipv6_icmp_error(sk, skb, err, uh->dest, ntohl(info), (u8 *)(uh+1));
540

L
Linus Torvalds 已提交
541 542 543 544 545 546
	sk->sk_err = err;
	sk->sk_error_report(sk);
out:
	sock_put(sk);
}

547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568
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;
}

569
static __inline__ void udpv6_err(struct sk_buff *skb,
570 571
				 struct inet6_skb_parm *opt, u8 type,
				 u8 code, int offset, __be32 info     )
572
{
573
	__udp6_lib_err(skb, opt, type, code, offset, info, &udp_table);
574 575
}

576 577 578 579 580 581 582 583
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);

584
int udpv6_queue_rcv_skb(struct sock *sk, struct sk_buff *skb)
L
Linus Torvalds 已提交
585
{
586
	struct udp_sock *up = udp_sk(sk);
587
	int rc;
W
Wang Chen 已提交
588
	int is_udplite = IS_UDPLITE(sk);
589

590 591
	if (!xfrm6_policy_check(sk, XFRM_POLICY_IN, skb))
		goto drop;
L
Linus Torvalds 已提交
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
	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 */
	}

624 625 626
	/*
	 * UDP-Lite specific tests, ignored on UDP sockets (see net/ipv4/udp.c).
	 */
W
Wang Chen 已提交
627
	if ((is_udplite & UDPLITE_RECV_CC)  &&  UDP_SKB_CB(skb)->partial_cov) {
628 629 630 631 632 633 634 635 636 637 638 639 640

		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
641 642
	}

643
	if (rcu_access_pointer(sk->sk_filter)) {
644
		if (udp_lib_checksum_complete(skb))
645
			goto csum_error;
646
	}
647

648 649 650
	if (sk_rcvqueues_full(sk, skb, sk->sk_rcvbuf))
		goto drop;

651
	skb_dst_drop(skb);
652

653 654 655 656 657 658 659 660 661
	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);
662 663

	return rc;
664 665
csum_error:
	UDP6_INC_STATS_BH(sock_net(sk), UDP_MIB_CSUMERRORS, is_udplite);
666
drop:
667
	UDP6_INC_STATS_BH(sock_net(sk), UDP_MIB_INERRORS, is_udplite);
668
	atomic_inc(&sk->sk_drops);
669 670
	kfree_skb(skb);
	return -1;
L
Linus Torvalds 已提交
671 672
}

673
static struct sock *udp_v6_mcast_next(struct net *net, struct sock *sk,
674 675
				      __be16 loc_port, const struct in6_addr *loc_addr,
				      __be16 rmt_port, const struct in6_addr *rmt_addr,
L
Linus Torvalds 已提交
676 677
				      int dif)
{
678
	struct hlist_nulls_node *node;
L
Linus Torvalds 已提交
679 680 681
	struct sock *s = sk;
	unsigned short num = ntohs(loc_port);

682
	sk_nulls_for_each_from(s, node) {
L
Linus Torvalds 已提交
683 684
		struct inet_sock *inet = inet_sk(s);

685
		if (!net_eq(sock_net(s), net))
686 687
			continue;

688 689
		if (udp_sk(s)->udp_port_hash == num &&
		    s->sk_family == PF_INET6) {
L
Linus Torvalds 已提交
690
			struct ipv6_pinfo *np = inet6_sk(s);
E
Eric Dumazet 已提交
691 692
			if (inet->inet_dport) {
				if (inet->inet_dport != rmt_port)
L
Linus Torvalds 已提交
693 694 695 696 697 698 699 700 701 702
					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)) {
703 704
				if (!ipv6_addr_equal(&np->rcv_saddr, loc_addr))
					continue;
L
Linus Torvalds 已提交
705
			}
S
Stephen Hemminger 已提交
706
			if (!inet6_mc_check(s, loc_addr, rmt_addr))
L
Linus Torvalds 已提交
707 708 709 710 711 712 713
				continue;
			return s;
		}
	}
	return NULL;
}

714 715 716
static void flush_stack(struct sock **stack, unsigned int count,
			struct sk_buff *skb, unsigned int final)
{
717
	struct sk_buff *skb1 = NULL;
718
	struct sock *sk;
719
	unsigned int i;
720 721

	for (i = 0; i < count; i++) {
722
		sk = stack[i];
723 724 725 726 727 728 729 730
		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));
731
		}
732 733 734

		if (skb1 && udpv6_queue_rcv_skb(sk, skb1) <= 0)
			skb1 = NULL;
735
	}
736 737
	if (unlikely(skb1))
		kfree_skb(skb1);
738
}
L
Linus Torvalds 已提交
739 740 741 742
/*
 * Note: called only from the BH handler context,
 * so we don't need to lock the hashes.
 */
743
static int __udp6_lib_mcast_deliver(struct net *net, struct sk_buff *skb,
744
		const struct in6_addr *saddr, const struct in6_addr *daddr,
745
		struct udp_table *udptable)
L
Linus Torvalds 已提交
746
{
747
	struct sock *sk, *stack[256 / sizeof(struct sock *)];
748
	const struct udphdr *uh = udp_hdr(skb);
749
	struct udp_hslot *hslot = udp_hashslot(udptable, net, ntohs(uh->dest));
L
Linus Torvalds 已提交
750
	int dif;
751
	unsigned int i, count = 0;
L
Linus Torvalds 已提交
752

753
	spin_lock(&hslot->lock);
754
	sk = sk_nulls_head(&hslot->head);
755
	dif = inet6_iif(skb);
756
	sk = udp_v6_mcast_next(net, sk, uh->dest, daddr, uh->source, saddr, dif);
757 758 759 760 761 762 763 764 765
	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
766
		}
L
Linus Torvalds 已提交
767
	}
768 769 770 771 772 773
	/*
	 * before releasing the lock, we must take reference on sockets
	 */
	for (i = 0; i < count; i++)
		sock_hold(stack[i]);

774
	spin_unlock(&hslot->lock);
775 776 777 778 779 780 781 782 783

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

		for (i = 0; i < count; i++)
			sock_put(stack[i]);
	} else {
		kfree_skb(skb);
	}
784
	return 0;
L
Linus Torvalds 已提交
785 786
}

787
int __udp6_lib_rcv(struct sk_buff *skb, struct udp_table *udptable,
788
		   int proto)
L
Linus Torvalds 已提交
789
{
790
	struct net *net = dev_net(skb->dev);
L
Linus Torvalds 已提交
791
	struct sock *sk;
792
	struct udphdr *uh;
793
	const struct in6_addr *saddr, *daddr;
L
Linus Torvalds 已提交
794 795 796
	u32 ulen = 0;

	if (!pskb_may_pull(skb, sizeof(struct udphdr)))
797
		goto discard;
L
Linus Torvalds 已提交
798

799 800
	saddr = &ipv6_hdr(skb)->saddr;
	daddr = &ipv6_hdr(skb)->daddr;
801
	uh = udp_hdr(skb);
L
Linus Torvalds 已提交
802 803

	ulen = ntohs(uh->len);
804 805
	if (ulen > skb->len)
		goto short_packet;
L
Linus Torvalds 已提交
806

807 808
	if (proto == IPPROTO_UDP) {
		/* UDP validates ulen. */
L
Linus Torvalds 已提交
809

810 811 812
		/* Check for jumbo payload */
		if (ulen == 0)
			ulen = skb->len;
L
Linus Torvalds 已提交
813

814 815
		if (ulen < sizeof(*uh))
			goto short_packet;
L
Linus Torvalds 已提交
816

817 818 819
		if (ulen < skb->len) {
			if (pskb_trim_rcsum(skb, ulen))
				goto short_packet;
820 821
			saddr = &ipv6_hdr(skb)->saddr;
			daddr = &ipv6_hdr(skb)->daddr;
822
			uh = udp_hdr(skb);
823 824
		}
	}
L
Linus Torvalds 已提交
825

826
	if (udp6_csum_init(skb, uh, proto))
827
		goto csum_error;
828

829 830
	/*
	 *	Multicast receive code
L
Linus Torvalds 已提交
831
	 */
832
	if (ipv6_addr_is_multicast(daddr))
833 834
		return __udp6_lib_mcast_deliver(net, skb,
				saddr, daddr, udptable);
L
Linus Torvalds 已提交
835 836 837

	/* Unicast */

838
	/*
L
Linus Torvalds 已提交
839 840 841
	 * check socket cache ... must talk to Alan about his plans
	 * for sock caches... i'll skip this for now.
	 */
842
	sk = __udp6_lib_lookup_skb(skb, uh->source, uh->dest, udptable);
843 844 845
	if (sk != NULL) {
		int ret = udpv6_queue_rcv_skb(sk, skb);
		sock_put(sk);
L
Linus Torvalds 已提交
846

847 848 849 850 851
		/* 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
852

S
Stephen Hemminger 已提交
853
		return 0;
L
Linus Torvalds 已提交
854
	}
855

856
	if (!xfrm6_policy_check(NULL, XFRM_POLICY_IN, skb))
857
		goto discard;
858 859

	if (udp_lib_checksum_complete(skb))
860
		goto csum_error;
861 862 863 864 865

	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
866
	return 0;
L
Linus Torvalds 已提交
867

868
short_packet:
869
	LIMIT_NETDEBUG(KERN_DEBUG "UDP%sv6: short packet: From [%pI6c]:%u %d/%d to [%pI6c]:%u\n",
870
		       proto == IPPROTO_UDPLITE ? "-Lite" : "",
871 872 873 874 875 876
		       saddr,
		       ntohs(uh->source),
		       ulen,
		       skb->len,
		       daddr,
		       ntohs(uh->dest));
877 878 879
	goto discard;
csum_error:
	UDP6_INC_STATS_BH(net, UDP_MIB_CSUMERRORS, proto == IPPROTO_UDPLITE);
L
Linus Torvalds 已提交
880
discard:
881
	UDP6_INC_STATS_BH(net, UDP_MIB_INERRORS, proto == IPPROTO_UDPLITE);
L
Linus Torvalds 已提交
882
	kfree_skb(skb);
S
Stephen Hemminger 已提交
883
	return 0;
L
Linus Torvalds 已提交
884
}
885

886
static __inline__ int udpv6_rcv(struct sk_buff *skb)
887
{
888
	return __udp6_lib_rcv(skb, &udp_table, IPPROTO_UDP);
889 890
}

L
Linus Torvalds 已提交
891 892 893 894 895 896 897
/*
 * 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);

898 899 900
	if (up->pending == AF_INET)
		udp_flush_pending_frames(sk);
	else if (up->pending) {
L
Linus Torvalds 已提交
901 902 903
		up->len = 0;
		up->pending = 0;
		ip6_flush_pending_frames(sk);
904
	}
L
Linus Torvalds 已提交
905 906
}

907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947
/**
 * 	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
948 949 950 951
/*
 *	Sending
 */

952
static int udp_v6_push_pending_frames(struct sock *sk)
L
Linus Torvalds 已提交
953 954 955
{
	struct sk_buff *skb;
	struct udphdr *uh;
956
	struct udp_sock  *up = udp_sk(sk);
L
Linus Torvalds 已提交
957
	struct inet_sock *inet = inet_sk(sk);
958
	struct flowi6 *fl6 = &inet->cork.fl.u.ip6;
L
Linus Torvalds 已提交
959
	int err = 0;
W
Wang Chen 已提交
960
	int is_udplite = IS_UDPLITE(sk);
961
	__wsum csum = 0;
L
Linus Torvalds 已提交
962 963 964 965 966 967 968 969

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

	/*
	 * Create a UDP header
	 */
970
	uh = udp_hdr(skb);
971 972
	uh->source = fl6->fl6_sport;
	uh->dest = fl6->fl6_dport;
L
Linus Torvalds 已提交
973 974 975
	uh->len = htons(up->len);
	uh->check = 0;

W
Wang Chen 已提交
976
	if (is_udplite)
977
		csum = udplite_csum_outgoing(sk, skb);
978
	else if (skb->ip_summed == CHECKSUM_PARTIAL) { /* UDP hardware csum */
979
		udp6_hwcsum_outgoing(sk, skb, &fl6->saddr, &fl6->daddr,
980 981 982
				     up->len);
		goto send;
	} else
983
		csum = udp_csum_outgoing(sk, skb);
L
Linus Torvalds 已提交
984

985
	/* add protocol-dependent pseudo-header */
986 987
	uh->check = csum_ipv6_magic(&fl6->saddr, &fl6->daddr,
				    up->len, fl6->flowi6_proto, csum);
L
Linus Torvalds 已提交
988
	if (uh->check == 0)
989
		uh->check = CSUM_MANGLED_0;
L
Linus Torvalds 已提交
990

991
send:
L
Linus Torvalds 已提交
992
	err = ip6_push_pending_frames(sk);
E
Eric Dumazet 已提交
993 994 995 996 997 998 999 1000 1001
	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
1002 1003 1004 1005 1006 1007
out:
	up->len = 0;
	up->pending = 0;
	return err;
}

1008
int udpv6_sendmsg(struct kiocb *iocb, struct sock *sk,
L
Linus Torvalds 已提交
1009 1010 1011 1012 1013 1014 1015
		  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;
1016
	struct in6_addr *daddr, *final_p, final;
L
Linus Torvalds 已提交
1017 1018
	struct ipv6_txoptions *opt = NULL;
	struct ip6_flowlabel *flowlabel = NULL;
1019
	struct flowi6 fl6;
L
Linus Torvalds 已提交
1020 1021 1022 1023
	struct dst_entry *dst;
	int addr_len = msg->msg_namelen;
	int ulen = len;
	int hlimit = -1;
1024
	int tclass = -1;
1025
	int dontfrag = -1;
L
Linus Torvalds 已提交
1026 1027
	int corkreq = up->corkflag || msg->msg_flags&MSG_MORE;
	int err;
1028
	int connected = 0;
W
Wang Chen 已提交
1029
	int is_udplite = IS_UDPLITE(sk);
1030
	int (*getfrag)(void *, char *, int, int, int, struct sk_buff *);
L
Linus Torvalds 已提交
1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056

	/* 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;
1057
	} else
L
Linus Torvalds 已提交
1058 1059 1060
		daddr = NULL;

	if (daddr) {
B
Brian Haley 已提交
1061
		if (ipv6_addr_v4mapped(daddr)) {
L
Linus Torvalds 已提交
1062 1063
			struct sockaddr_in sin;
			sin.sin_family = AF_INET;
E
Eric Dumazet 已提交
1064
			sin.sin_port = sin6 ? sin6->sin6_port : inet->inet_dport;
L
Linus Torvalds 已提交
1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078
			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,
1079
	   better check is made in ip6_append_data().
L
Linus Torvalds 已提交
1080 1081 1082
	   */
	if (len > INT_MAX - sizeof(struct udphdr))
		return -EMSGSIZE;
1083

L
Linus Torvalds 已提交
1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101
	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);

1102
	memset(&fl6, 0, sizeof(fl6));
L
Linus Torvalds 已提交
1103 1104 1105 1106 1107

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

1108
		fl6.fl6_dport = sin6->sin6_port;
L
Linus Torvalds 已提交
1109 1110 1111
		daddr = &sin6->sin6_addr;

		if (np->sndflow) {
1112 1113 1114
			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
1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130
				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 &&
1131
		    __ipv6_addr_needs_scope_id(__ipv6_addr_type(daddr)))
1132
			fl6.flowi6_oif = sin6->sin6_scope_id;
L
Linus Torvalds 已提交
1133 1134 1135 1136
	} else {
		if (sk->sk_state != TCP_ESTABLISHED)
			return -EDESTADDRREQ;

1137
		fl6.fl6_dport = inet->inet_dport;
L
Linus Torvalds 已提交
1138
		daddr = &np->daddr;
1139
		fl6.flowlabel = np->flow_label;
1140
		connected = 1;
L
Linus Torvalds 已提交
1141 1142
	}

1143 1144
	if (!fl6.flowi6_oif)
		fl6.flowi6_oif = sk->sk_bound_dev_if;
L
Linus Torvalds 已提交
1145

1146 1147
	if (!fl6.flowi6_oif)
		fl6.flowi6_oif = np->sticky_pktinfo.ipi6_ifindex;
1148

1149
	fl6.flowi6_mark = sk->sk_mark;
B
Brian Haley 已提交
1150

L
Linus Torvalds 已提交
1151 1152 1153 1154 1155
	if (msg->msg_controllen) {
		opt = &opt_space;
		memset(opt, 0, sizeof(struct ipv6_txoptions));
		opt->tot_len = sizeof(*opt);

1156 1157
		err = ip6_datagram_send_ctl(sock_net(sk), sk, msg, &fl6, opt,
					    &hlimit, &tclass, &dontfrag);
L
Linus Torvalds 已提交
1158 1159 1160 1161
		if (err < 0) {
			fl6_sock_release(flowlabel);
			return err;
		}
1162 1163
		if ((fl6.flowlabel&IPV6_FLOWLABEL_MASK) && !flowlabel) {
			flowlabel = fl6_sock_lookup(sk, fl6.flowlabel);
L
Linus Torvalds 已提交
1164 1165 1166 1167 1168
			if (flowlabel == NULL)
				return -EINVAL;
		}
		if (!(opt->opt_nflen|opt->opt_flen))
			opt = NULL;
1169
		connected = 0;
L
Linus Torvalds 已提交
1170 1171 1172
	}
	if (opt == NULL)
		opt = np->opt;
1173 1174 1175
	if (flowlabel)
		opt = fl6_merge_options(&opt_space, flowlabel, opt);
	opt = ipv6_fixup_options(&opt_space, opt);
L
Linus Torvalds 已提交
1176

1177
	fl6.flowi6_proto = sk->sk_protocol;
1178
	if (!ipv6_addr_any(daddr))
A
Alexey Dobriyan 已提交
1179
		fl6.daddr = *daddr;
1180
	else
1181 1182
		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
1183
		fl6.saddr = np->saddr;
1184
	fl6.fl6_sport = inet->inet_sport;
1185

1186
	final_p = fl6_update_dst(&fl6, opt, &final);
1187
	if (final_p)
1188
		connected = 0;
L
Linus Torvalds 已提交
1189

1190 1191
	if (!fl6.flowi6_oif && ipv6_addr_is_multicast(&fl6.daddr)) {
		fl6.flowi6_oif = np->mcast_oif;
1192
		connected = 0;
1193 1194
	} else if (!fl6.flowi6_oif)
		fl6.flowi6_oif = np->ucast_oif;
L
Linus Torvalds 已提交
1195

1196
	security_sk_classify_flow(sk, flowi6_to_flowi(&fl6));
V
Venkat Yekkirala 已提交
1197

1198
	dst = ip6_sk_dst_lookup_flow(sk, &fl6, final_p, true);
1199 1200 1201
	if (IS_ERR(dst)) {
		err = PTR_ERR(dst);
		dst = NULL;
L
Linus Torvalds 已提交
1202
		goto out;
1203
	}
L
Linus Torvalds 已提交
1204 1205

	if (hlimit < 0) {
1206
		if (ipv6_addr_is_multicast(&fl6.daddr))
L
Linus Torvalds 已提交
1207 1208 1209 1210
			hlimit = np->mcast_hops;
		else
			hlimit = np->hop_limit;
		if (hlimit < 0)
1211
			hlimit = ip6_dst_hoplimit(dst);
L
Linus Torvalds 已提交
1212 1213
	}

G
Gerrit Renker 已提交
1214
	if (tclass < 0)
1215 1216
		tclass = np->tclass;

1217 1218 1219
	if (dontfrag < 0)
		dontfrag = np->dontfrag;

L
Linus Torvalds 已提交
1220 1221 1222 1223 1224 1225 1226 1227 1228 1229
	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);

1230
		LIMIT_NETDEBUG(KERN_DEBUG "udp cork app bug 2\n");
L
Linus Torvalds 已提交
1231 1232 1233 1234 1235 1236 1237 1238
		err = -EINVAL;
		goto out;
	}

	up->pending = AF_INET6;

do_append_data:
	up->len += ulen;
1239 1240
	getfrag  =  is_udplite ?  udplite_getfrag : ip_generic_getfrag;
	err = ip6_append_data(sk, getfrag, msg->msg_iov, ulen,
1241
		sizeof(struct udphdr), hlimit, tclass, opt, &fl6,
1242
		(struct rt6_info*)dst,
1243
		corkreq ? msg->msg_flags|MSG_MORE : msg->msg_flags, dontfrag);
L
Linus Torvalds 已提交
1244 1245 1246
	if (err)
		udp_v6_flush_pending_frames(sk);
	else if (!corkreq)
1247
		err = udp_v6_push_pending_frames(sk);
H
Herbert Xu 已提交
1248 1249
	else if (unlikely(skb_queue_empty(&sk->sk_write_queue)))
		up->pending = 0;
L
Linus Torvalds 已提交
1250

1251 1252 1253
	if (dst) {
		if (connected) {
			ip6_dst_store(sk, dst,
1254
				      ipv6_addr_equal(&fl6.daddr, &np->daddr) ?
1255 1256
				      &np->daddr : NULL,
#ifdef CONFIG_IPV6_SUBTREES
1257
				      ipv6_addr_equal(&fl6.saddr, &np->saddr) ?
1258 1259 1260
				      &np->saddr :
#endif
				      NULL);
1261 1262 1263
		} else {
			dst_release(dst);
		}
1264
		dst = NULL;
1265 1266
	}

L
Linus Torvalds 已提交
1267 1268 1269 1270
	if (err > 0)
		err = np->recverr ? net_xmit_errno(err) : 0;
	release_sock(sk);
out:
1271
	dst_release(dst);
L
Linus Torvalds 已提交
1272
	fl6_sock_release(flowlabel);
1273
	if (!err)
L
Linus Torvalds 已提交
1274
		return len;
1275 1276 1277 1278 1279 1280 1281 1282
	/*
	 * 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)) {
1283 1284
		UDP6_INC_STATS_USER(sock_net(sk),
				UDP_MIB_SNDBUFERRORS, is_udplite);
1285
	}
L
Linus Torvalds 已提交
1286 1287 1288 1289 1290 1291 1292 1293 1294 1295
	return err;

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

1296
void udpv6_destroy_sock(struct sock *sk)
L
Linus Torvalds 已提交
1297
{
T
Tom Parkin 已提交
1298
	struct udp_sock *up = udp_sk(sk);
L
Linus Torvalds 已提交
1299 1300 1301 1302
	lock_sock(sk);
	udp_v6_flush_pending_frames(sk);
	release_sock(sk);

T
Tom Parkin 已提交
1303 1304 1305 1306 1307 1308 1309
	if (static_key_false(&udpv6_encap_needed) && up->encap_type) {
		void (*encap_destroy)(struct sock *sk);
		encap_destroy = ACCESS_ONCE(up->encap_destroy);
		if (encap_destroy)
			encap_destroy(sk);
	}

L
Linus Torvalds 已提交
1310 1311 1312 1313 1314 1315
	inet6_destroy_sock(sk);
}

/*
 *	Socket option code for UDP
 */
1316
int udpv6_setsockopt(struct sock *sk, int level, int optname,
1317
		     char __user *optval, unsigned int optlen)
1318
{
1319
	if (level == SOL_UDP  ||  level == SOL_UDPLITE)
1320 1321
		return udp_lib_setsockopt(sk, level, optname, optval, optlen,
					  udp_v6_push_pending_frames);
1322
	return ipv6_setsockopt(sk, level, optname, optval, optlen);
1323 1324 1325
}

#ifdef CONFIG_COMPAT
1326
int compat_udpv6_setsockopt(struct sock *sk, int level, int optname,
1327
			    char __user *optval, unsigned int optlen)
1328
{
1329
	if (level == SOL_UDP  ||  level == SOL_UDPLITE)
1330 1331
		return udp_lib_setsockopt(sk, level, optname, optval, optlen,
					  udp_v6_push_pending_frames);
1332
	return compat_ipv6_setsockopt(sk, level, optname, optval, optlen);
1333 1334 1335
}
#endif

1336 1337
int udpv6_getsockopt(struct sock *sk, int level, int optname,
		     char __user *optval, int __user *optlen)
1338
{
1339
	if (level == SOL_UDP  ||  level == SOL_UDPLITE)
1340
		return udp_lib_getsockopt(sk, level, optname, optval, optlen);
1341
	return ipv6_getsockopt(sk, level, optname, optval, optlen);
1342 1343 1344
}

#ifdef CONFIG_COMPAT
1345 1346
int compat_udpv6_getsockopt(struct sock *sk, int level, int optname,
			    char __user *optval, int __user *optlen)
1347
{
1348
	if (level == SOL_UDP  ||  level == SOL_UDPLITE)
1349
		return udp_lib_getsockopt(sk, level, optname, optval, optlen);
1350
	return compat_ipv6_getsockopt(sk, level, optname, optval, optlen);
1351 1352 1353
}
#endif

1354
static const struct inet6_protocol udpv6_protocol = {
L
Linus Torvalds 已提交
1355 1356 1357 1358 1359 1360 1361
	.handler	=	udpv6_rcv,
	.err_handler	=	udpv6_err,
	.flags		=	INET6_PROTO_NOPOLICY|INET6_PROTO_FINAL,
};

/* ------------------------------------------------------------------------ */
#ifdef CONFIG_PROC_FS
1362
int udp6_seq_show(struct seq_file *seq, void *v)
L
Linus Torvalds 已提交
1363
{
1364 1365 1366 1367 1368 1369 1370 1371 1372
	if (v == SEQ_START_TOKEN) {
		seq_puts(seq, IPV6_SEQ_DGRAM_HEADER);
	} else {
		int bucket = ((struct udp_iter_state *)seq->private)->bucket;
		struct inet_sock *inet = inet_sk(v);
		__u16 srcp = ntohs(inet->inet_sport);
		__u16 destp = ntohs(inet->inet_dport);
		ip6_dgram_sock_seq_show(seq, v, srcp, destp, bucket);
	}
L
Linus Torvalds 已提交
1373 1374 1375
	return 0;
}

1376 1377 1378 1379 1380 1381 1382 1383
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
1384 1385 1386
static struct udp_seq_afinfo udp6_seq_afinfo = {
	.name		= "udp6",
	.family		= AF_INET6,
1387
	.udp_table	= &udp_table,
1388
	.seq_fops	= &udp6_afinfo_seq_fops,
1389 1390 1391
	.seq_ops	= {
		.show		= udp6_seq_show,
	},
L
Linus Torvalds 已提交
1392 1393
};

1394
int __net_init udp6_proc_init(struct net *net)
L
Linus Torvalds 已提交
1395
{
1396
	return udp_proc_register(net, &udp6_seq_afinfo);
L
Linus Torvalds 已提交
1397 1398
}

1399 1400
void udp6_proc_exit(struct net *net) {
	udp_proc_unregister(net, &udp6_seq_afinfo);
L
Linus Torvalds 已提交
1401 1402 1403
}
#endif /* CONFIG_PROC_FS */

E
Eric Dumazet 已提交
1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414
void udp_v6_clear_sk(struct sock *sk, int size)
{
	struct inet_sock *inet = inet_sk(sk);

	/* we do not want to clear pinet6 field, because of RCU lookups */
	sk_prot_clear_portaddr_nulls(sk, offsetof(struct inet_sock, pinet6));

	size -= offsetof(struct inet_sock, pinet6) + sizeof(inet->pinet6);
	memset(&inet->pinet6 + 1, 0, size);
}

L
Linus Torvalds 已提交
1415 1416 1417
/* ------------------------------------------------------------------------ */

struct proto udpv6_prot = {
1418 1419
	.name		   = "UDPv6",
	.owner		   = THIS_MODULE,
1420
	.close		   = udp_lib_close,
1421 1422 1423 1424 1425 1426 1427 1428
	.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,
1429
	.backlog_rcv	   = __udpv6_queue_rcv_skb,
1430 1431
	.hash		   = udp_lib_hash,
	.unhash		   = udp_lib_unhash,
E
Eric Dumazet 已提交
1432
	.rehash		   = udp_v6_rehash,
1433
	.get_port	   = udp_v6_get_port,
H
Hideo Aoki 已提交
1434 1435 1436 1437
	.memory_allocated  = &udp_memory_allocated,
	.sysctl_mem	   = sysctl_udp_mem,
	.sysctl_wmem	   = &sysctl_udp_wmem_min,
	.sysctl_rmem	   = &sysctl_udp_rmem_min,
1438
	.obj_size	   = sizeof(struct udp6_sock),
1439
	.slab_flags	   = SLAB_DESTROY_BY_RCU,
1440
	.h.udp_table	   = &udp_table,
1441
#ifdef CONFIG_COMPAT
1442 1443
	.compat_setsockopt = compat_udpv6_setsockopt,
	.compat_getsockopt = compat_udpv6_getsockopt,
1444
#endif
E
Eric Dumazet 已提交
1445
	.clear_sk	   = udp_v6_clear_sk,
L
Linus Torvalds 已提交
1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457
};

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,
};


1458
int __init udpv6_init(void)
L
Linus Torvalds 已提交
1459
{
1460 1461
	int ret;

1462 1463
	ret = inet6_add_protocol(&udpv6_protocol, IPPROTO_UDP);
	if (ret)
1464
		goto out;
1465

1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476
	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;
}

1477
void udpv6_exit(void)
1478 1479 1480
{
	inet6_unregister_protosw(&udpv6_protosw);
	inet6_del_protocol(&udpv6_protocol, IPPROTO_UDP);
L
Linus Torvalds 已提交
1481
}