udp.c 37.3 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>
49
#include <net/busy_poll.h>
L
Linus Torvalds 已提交
50 51 52

#include <linux/proc_fs.h>
#include <linux/seq_file.h>
53
#include <trace/events/skb.h>
54
#include "udp_impl.h"
L
Linus Torvalds 已提交
55

56 57 58 59 60 61
static unsigned int udp6_ehashfn(struct net *net,
				  const struct in6_addr *laddr,
				  const u16 lport,
				  const struct in6_addr *faddr,
				  const __be16 fport)
{
62 63 64 65 66 67 68 69 70 71 72 73 74
	static u32 udp6_ehash_secret __read_mostly;
	static u32 udp_ipv6_hash_secret __read_mostly;

	u32 lhash, fhash;

	net_get_random_once(&udp6_ehash_secret,
			    sizeof(udp6_ehash_secret));
	net_get_random_once(&udp_ipv6_hash_secret,
			    sizeof(udp_ipv6_hash_secret));

	lhash = (__force u32)laddr->s6_addr32[3];
	fhash = __ipv6_addr_jhash(faddr, udp_ipv6_hash_secret);

75
	return __inet6_ehashfn(lhash, lport, fhash, fport,
76
			       udp_ipv6_hash_secret + net_hash_mix(net));
77 78
}

79 80 81 82 83
int ipv6_rcv_saddr_equal(const struct sock *sk, const struct sock *sk2)
{
	const struct in6_addr *sk2_rcv_saddr6 = inet6_rcv_saddr(sk2);
	int sk_ipv6only = ipv6_only_sock(sk);
	int sk2_ipv6only = inet_v6_ipv6only(sk2);
84
	int addr_type = ipv6_addr_type(&sk->sk_v6_rcv_saddr);
85 86 87 88
	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)
89
		return (!sk2_ipv6only &&
E
Eric Dumazet 已提交
90 91
			(!sk->sk_rcv_saddr || !sk2->sk_rcv_saddr ||
			  sk->sk_rcv_saddr == sk2->sk_rcv_saddr));
92 93 94 95 96 97 98 99 100 101

	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 &&
102
	    ipv6_addr_equal(&sk->sk_v6_rcv_saddr, sk2_rcv_saddr6))
103 104 105 106 107
		return 1;

	return 0;
}

108 109 110 111 112 113 114 115
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
116
	else if (ipv6_addr_v4mapped(addr6))
117
		hash = jhash_1word((__force u32)addr6->s6_addr32[3], mix);
118
	else
119
		hash = jhash2((__force u32 *)addr6->s6_addr32, 4, mix);
120 121 122 123 124

	return hash ^ port;
}


125
int udp_v6_get_port(struct sock *sk, unsigned short snum)
L
Linus Torvalds 已提交
126
{
E
Eric Dumazet 已提交
127 128
	unsigned int hash2_nulladdr =
		udp6_portaddr_hash(sock_net(sk), &in6addr_any, snum);
129
	unsigned int hash2_partial =
130
		udp6_portaddr_hash(sock_net(sk), &sk->sk_v6_rcv_saddr, 0);
E
Eric Dumazet 已提交
131

132
	/* precompute partial secondary hash */
E
Eric Dumazet 已提交
133 134
	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
135 136
}

E
Eric Dumazet 已提交
137 138 139
static void udp_v6_rehash(struct sock *sk)
{
	u16 new_hash = udp6_portaddr_hash(sock_net(sk),
140
					  &sk->sk_v6_rcv_saddr,
E
Eric Dumazet 已提交
141 142 143 144 145
					  inet_sk(sk)->inet_num);

	udp_lib_rehash(sk, new_hash);
}

146 147
static inline int compute_score(struct sock *sk, struct net *net,
				unsigned short hnum,
148 149
				const struct in6_addr *saddr, __be16 sport,
				const struct in6_addr *daddr, __be16 dport,
150 151 152 153
				int dif)
{
	int score = -1;

154
	if (net_eq(sock_net(sk), net) && udp_sk(sk)->udp_port_hash == hnum &&
155 156 157 158
			sk->sk_family == PF_INET6) {
		struct inet_sock *inet = inet_sk(sk);

		score = 0;
E
Eric Dumazet 已提交
159 160
		if (inet->inet_dport) {
			if (inet->inet_dport != sport)
161 162 163
				return -1;
			score++;
		}
164 165
		if (!ipv6_addr_any(&sk->sk_v6_rcv_saddr)) {
			if (!ipv6_addr_equal(&sk->sk_v6_rcv_saddr, daddr))
166 167 168
				return -1;
			score++;
		}
169 170
		if (!ipv6_addr_any(&sk->sk_v6_daddr)) {
			if (!ipv6_addr_equal(&sk->sk_v6_daddr, saddr))
171 172 173 174 175 176 177 178 179 180 181 182
				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
183 184 185 186 187 188 189 190 191 192 193 194
#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 inet_sock *inet = inet_sk(sk);

195
		if (!ipv6_addr_equal(&sk->sk_v6_rcv_saddr, daddr))
E
Eric Dumazet 已提交
196 197 198 199 200 201 202
			return -1;
		score = 0;
		if (inet->inet_dport) {
			if (inet->inet_dport != sport)
				return -1;
			score++;
		}
203 204
		if (!ipv6_addr_any(&sk->sk_v6_daddr)) {
			if (!ipv6_addr_equal(&sk->sk_v6_daddr, saddr))
E
Eric Dumazet 已提交
205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225
				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;
226 227
	int score, badness, matches = 0, reuseport = 0;
	u32 hash = 0;
E
Eric Dumazet 已提交
228 229 230 231 232 233 234 235 236 237

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;
238 239
			reuseport = sk->sk_reuseport;
			if (reuseport) {
240 241
				hash = udp6_ehashfn(net, daddr, hnum,
						    saddr, sport);
242 243
				matches = 1;
			} else if (score == SCORE2_MAX)
E
Eric Dumazet 已提交
244
				goto exact_match;
245 246 247 248 249
		} 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
250 251 252 253 254 255 256 257 258 259 260 261
		}
	}
	/*
	 * 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
262
		if (unlikely(!atomic_inc_not_zero_hint(&result->sk_refcnt, 2)))
E
Eric Dumazet 已提交
263 264 265 266 267 268 269 270 271 272
			result = NULL;
		else if (unlikely(compute_score2(result, net, saddr, sport,
				  daddr, hnum, dif) < badness)) {
			sock_put(result);
			goto begin;
		}
	}
	return result;
}

273
struct sock *__udp6_lib_lookup(struct net *net,
274 275
				      const struct in6_addr *saddr, __be16 sport,
				      const struct in6_addr *daddr, __be16 dport,
276
				      int dif, struct udp_table *udptable)
L
Linus Torvalds 已提交
277
{
278
	struct sock *sk, *result;
279
	struct hlist_nulls_node *node;
L
Linus Torvalds 已提交
280
	unsigned short hnum = ntohs(dport);
E
Eric Dumazet 已提交
281 282
	unsigned int hash2, slot2, slot = udp_hashfn(net, hnum, udptable->mask);
	struct udp_hslot *hslot2, *hslot = &udptable->hash[slot];
283 284
	int score, badness, matches = 0, reuseport = 0;
	u32 hash = 0;
285

286
	rcu_read_lock();
E
Eric Dumazet 已提交
287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303
	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;

304 305
			result = udp6_lib_lookup2(net, saddr, sport,
						  &in6addr_any, hnum, dif,
E
Eric Dumazet 已提交
306 307 308 309 310
						  hslot2, slot2);
		}
		rcu_read_unlock();
		return result;
	}
311 312 313
begin:
	result = NULL;
	badness = -1;
314
	sk_nulls_for_each_rcu(sk, node, &hslot->head) {
315 316 317 318
		score = compute_score(sk, net, hnum, saddr, sport, daddr, dport, dif);
		if (score > badness) {
			result = sk;
			badness = score;
319 320
			reuseport = sk->sk_reuseport;
			if (reuseport) {
321 322
				hash = udp6_ehashfn(net, daddr, hnum,
						    saddr, sport);
323 324 325 326 327 328 329
				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
330 331
		}
	}
332 333 334 335 336
	/*
	 * 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
337
	if (get_nulls_value(node) != slot)
338 339
		goto begin;

340
	if (result) {
E
Eric Dumazet 已提交
341
		if (unlikely(!atomic_inc_not_zero_hint(&result->sk_refcnt, 2)))
342 343 344 345 346 347 348 349
			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
350 351
	return result;
}
352
EXPORT_SYMBOL_GPL(__udp6_lib_lookup);
L
Linus Torvalds 已提交
353

354 355
static struct sock *__udp6_lib_lookup_skb(struct sk_buff *skb,
					  __be16 sport, __be16 dport,
356
					  struct udp_table *udptable)
357
{
358
	struct sock *sk;
359
	const struct ipv6hdr *iph = ipv6_hdr(skb);
360

361 362
	if (unlikely(sk = skb_steal_sock(skb)))
		return sk;
E
Eric Dumazet 已提交
363 364 365
	return __udp6_lib_lookup(dev_net(skb_dst(skb)->dev), &iph->saddr, sport,
				 &iph->daddr, dport, inet6_iif(skb),
				 udptable);
366 367
}

368 369 370 371 372 373 374 375
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
376 377 378 379 380
/*
 * 	This should be easy, if there is something there we
 * 	return it, otherwise we block.
 */

381
int udpv6_recvmsg(struct kiocb *iocb, struct sock *sk,
L
Linus Torvalds 已提交
382 383 384 385 386
		  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);
387
	struct sk_buff *skb;
388
	unsigned int ulen, copied;
389
	int peeked, off = 0;
390 391
	int err;
	int is_udplite = IS_UDPLITE(sk);
392
	int is_udp4;
393
	bool slow;
L
Linus Torvalds 已提交
394 395

	if (flags & MSG_ERRQUEUE)
396
		return ipv6_recv_error(sk, msg, len, addr_len);
L
Linus Torvalds 已提交
397

398
	if (np->rxpmtu && np->rxopt.bits.rxpmtu)
399
		return ipv6_recv_rxpmtu(sk, msg, len, addr_len);
400

L
Linus Torvalds 已提交
401
try_again:
402
	skb = __skb_recv_datagram(sk, flags | (noblock ? MSG_DONTWAIT : 0),
403
				  &peeked, &off, &err);
L
Linus Torvalds 已提交
404 405 406
	if (!skb)
		goto out;

407
	ulen = skb->len - sizeof(struct udphdr);
408 409 410 411
	copied = len;
	if (copied > ulen)
		copied = ulen;
	else if (copied < ulen)
412
		msg->msg_flags |= MSG_TRUNC;
L
Linus Torvalds 已提交
413

414 415
	is_udp4 = (skb->protocol == htons(ETH_P_IP));

416
	/*
417 418 419
	 * 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.
420 421
	 */

422
	if (copied < ulen || UDP_SKB_CB(skb)->partial_cov) {
423
		if (udp_lib_checksum_complete(skb))
L
Linus Torvalds 已提交
424
			goto csum_copy_err;
425 426
	}

427
	if (skb_csum_unnecessary(skb))
428
		err = skb_copy_datagram_iovec(skb, sizeof(struct udphdr),
429
					      msg->msg_iov, copied);
430
	else {
L
Linus Torvalds 已提交
431 432 433 434
		err = skb_copy_and_csum_datagram_iovec(skb, sizeof(struct udphdr), msg->msg_iov);
		if (err == -EINVAL)
			goto csum_copy_err;
	}
435 436
	if (unlikely(err)) {
		trace_kfree_skb(skb, udpv6_recvmsg);
437 438 439 440 441 442 443 444 445 446 447
		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
448
		goto out_free;
449
	}
450 451 452 453 454 455 456 457
	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);
	}
458

459
	sock_recv_ts_and_drops(msg, sk, skb);
L
Linus Torvalds 已提交
460 461 462 463

	/* Copy the address. */
	if (msg->msg_name) {
		struct sockaddr_in6 *sin6;
464

L
Linus Torvalds 已提交
465 466
		sin6 = (struct sockaddr_in6 *) msg->msg_name;
		sin6->sin6_family = AF_INET6;
467
		sin6->sin6_port = udp_hdr(skb)->source;
L
Linus Torvalds 已提交
468 469
		sin6->sin6_flowinfo = 0;

470
		if (is_udp4) {
B
Brian Haley 已提交
471 472
			ipv6_addr_set_v4mapped(ip_hdr(skb)->saddr,
					       &sin6->sin6_addr);
473 474
			sin6->sin6_scope_id = 0;
		} else {
A
Alexey Dobriyan 已提交
475
			sin6->sin6_addr = ipv6_hdr(skb)->saddr;
476 477 478
			sin6->sin6_scope_id =
				ipv6_iface_scope_id(&sin6->sin6_addr,
						    IP6CB(skb)->iif);
L
Linus Torvalds 已提交
479
		}
480
		*addr_len = sizeof(*sin6);
L
Linus Torvalds 已提交
481
	}
482
	if (is_udp4) {
L
Linus Torvalds 已提交
483 484 485 486
		if (inet->cmsg_flags)
			ip_cmsg_recv(msg, skb);
	} else {
		if (np->rxopt.all)
487
			ip6_datagram_recv_ctl(sk, msg, skb);
488
	}
L
Linus Torvalds 已提交
489

490
	err = copied;
L
Linus Torvalds 已提交
491
	if (flags & MSG_TRUNC)
492
		err = ulen;
L
Linus Torvalds 已提交
493 494

out_free:
495
	skb_free_datagram_locked(sk, skb);
L
Linus Torvalds 已提交
496 497 498 499
out:
	return err;

csum_copy_err:
500
	slow = lock_sock_fast(sk);
501
	if (!skb_kill_datagram(sk, skb, flags)) {
502 503 504
		if (is_udp4) {
			UDP_INC_STATS_USER(sock_net(sk),
					UDP_MIB_CSUMERRORS, is_udplite);
505 506
			UDP_INC_STATS_USER(sock_net(sk),
					UDP_MIB_INERRORS, is_udplite);
507 508 509
		} else {
			UDP6_INC_STATS_USER(sock_net(sk),
					UDP_MIB_CSUMERRORS, is_udplite);
510 511
			UDP6_INC_STATS_USER(sock_net(sk),
					UDP_MIB_INERRORS, is_udplite);
512
		}
513
	}
514
	unlock_sock_fast(sk, slow);
L
Linus Torvalds 已提交
515

516
	if (noblock)
L
Linus Torvalds 已提交
517
		return -EAGAIN;
518 519 520

	/* starting over for a new packet */
	msg->msg_flags &= ~MSG_TRUNC;
L
Linus Torvalds 已提交
521 522 523
	goto try_again;
}

524
void __udp6_lib_err(struct sk_buff *skb, struct inet6_skb_parm *opt,
525
		    u8 type, u8 code, int offset, __be32 info,
526
		    struct udp_table *udptable)
L
Linus Torvalds 已提交
527 528
{
	struct ipv6_pinfo *np;
529 530 531
	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
532 533 534 535
	struct udphdr *uh = (struct udphdr*)(skb->data+offset);
	struct sock *sk;
	int err;

536
	sk = __udp6_lib_lookup(dev_net(skb->dev), daddr, uh->dest,
537
			       saddr, uh->source, inet6_iif(skb), udptable);
L
Linus Torvalds 已提交
538 539 540
	if (sk == NULL)
		return;

541 542 543
	if (type == ICMPV6_PKT_TOOBIG) {
		if (!ip6_sk_accept_pmtu(sk))
			goto out;
544
		ip6_sk_update_pmtu(skb, sk, info);
545
	}
546
	if (type == NDISC_REDIRECT) {
547
		ip6_sk_redirect(skb, sk);
548 549
		goto out;
	}
550

L
Linus Torvalds 已提交
551 552 553 554 555 556 557 558
	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;

559
	if (np->recverr)
L
Linus Torvalds 已提交
560
		ipv6_icmp_error(sk, skb, err, uh->dest, ntohl(info), (u8 *)(uh+1));
561

L
Linus Torvalds 已提交
562 563 564 565 566 567
	sk->sk_err = err;
	sk->sk_error_report(sk);
out:
	sock_put(sk);
}

568 569 570 571
static int __udpv6_queue_rcv_skb(struct sock *sk, struct sk_buff *skb)
{
	int rc;

572
	if (!ipv6_addr_any(&sk->sk_v6_daddr)) {
573
		sock_rps_save_rxhash(sk, skb);
574 575
		sk_mark_napi_id(sk, skb);
	}
576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591

	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;
}

592
static __inline__ void udpv6_err(struct sk_buff *skb,
593 594
				 struct inet6_skb_parm *opt, u8 type,
				 u8 code, int offset, __be32 info     )
595
{
596
	__udp6_lib_err(skb, opt, type, code, offset, info, &udp_table);
597 598
}

599 600 601 602 603 604 605 606
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);

607
int udpv6_queue_rcv_skb(struct sock *sk, struct sk_buff *skb)
L
Linus Torvalds 已提交
608
{
609
	struct udp_sock *up = udp_sk(sk);
610
	int rc;
W
Wang Chen 已提交
611
	int is_udplite = IS_UDPLITE(sk);
612

613 614
	if (!xfrm6_policy_check(sk, XFRM_POLICY_IN, skb))
		goto drop;
L
Linus Torvalds 已提交
615

616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646
	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 */
	}

647 648 649
	/*
	 * UDP-Lite specific tests, ignored on UDP sockets (see net/ipv4/udp.c).
	 */
W
Wang Chen 已提交
650
	if ((is_udplite & UDPLITE_RECV_CC)  &&  UDP_SKB_CB(skb)->partial_cov) {
651 652 653 654 655 656 657 658 659 660 661 662 663

		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
664 665
	}

666
	if (rcu_access_pointer(sk->sk_filter)) {
667
		if (udp_lib_checksum_complete(skb))
668
			goto csum_error;
669
	}
670

671 672 673
	if (sk_rcvqueues_full(sk, skb, sk->sk_rcvbuf))
		goto drop;

674
	skb_dst_drop(skb);
675

676 677 678 679 680 681 682 683 684
	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);
685 686

	return rc;
687 688
csum_error:
	UDP6_INC_STATS_BH(sock_net(sk), UDP_MIB_CSUMERRORS, is_udplite);
689
drop:
690
	UDP6_INC_STATS_BH(sock_net(sk), UDP_MIB_INERRORS, is_udplite);
691
	atomic_inc(&sk->sk_drops);
692 693
	kfree_skb(skb);
	return -1;
L
Linus Torvalds 已提交
694 695
}

696
static struct sock *udp_v6_mcast_next(struct net *net, struct sock *sk,
697 698
				      __be16 loc_port, const struct in6_addr *loc_addr,
				      __be16 rmt_port, const struct in6_addr *rmt_addr,
L
Linus Torvalds 已提交
699 700
				      int dif)
{
701
	struct hlist_nulls_node *node;
L
Linus Torvalds 已提交
702 703 704
	struct sock *s = sk;
	unsigned short num = ntohs(loc_port);

705
	sk_nulls_for_each_from(s, node) {
L
Linus Torvalds 已提交
706 707
		struct inet_sock *inet = inet_sk(s);

708
		if (!net_eq(sock_net(s), net))
709 710
			continue;

711 712
		if (udp_sk(s)->udp_port_hash == num &&
		    s->sk_family == PF_INET6) {
E
Eric Dumazet 已提交
713 714
			if (inet->inet_dport) {
				if (inet->inet_dport != rmt_port)
L
Linus Torvalds 已提交
715 716
					continue;
			}
717 718
			if (!ipv6_addr_any(&sk->sk_v6_daddr) &&
			    !ipv6_addr_equal(&sk->sk_v6_daddr, rmt_addr))
L
Linus Torvalds 已提交
719 720 721 722 723
				continue;

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

724 725
			if (!ipv6_addr_any(&sk->sk_v6_rcv_saddr)) {
				if (!ipv6_addr_equal(&sk->sk_v6_rcv_saddr, loc_addr))
726
					continue;
L
Linus Torvalds 已提交
727
			}
S
Stephen Hemminger 已提交
728
			if (!inet6_mc_check(s, loc_addr, rmt_addr))
L
Linus Torvalds 已提交
729 730 731 732 733 734 735
				continue;
			return s;
		}
	}
	return NULL;
}

736 737 738
static void flush_stack(struct sock **stack, unsigned int count,
			struct sk_buff *skb, unsigned int final)
{
739
	struct sk_buff *skb1 = NULL;
740
	struct sock *sk;
741
	unsigned int i;
742 743

	for (i = 0; i < count; i++) {
744
		sk = stack[i];
745 746 747 748 749 750 751 752
		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));
753
		}
754 755 756

		if (skb1 && udpv6_queue_rcv_skb(sk, skb1) <= 0)
			skb1 = NULL;
757
	}
758 759
	if (unlikely(skb1))
		kfree_skb(skb1);
760
}
L
Linus Torvalds 已提交
761 762 763 764
/*
 * Note: called only from the BH handler context,
 * so we don't need to lock the hashes.
 */
765
static int __udp6_lib_mcast_deliver(struct net *net, struct sk_buff *skb,
766
		const struct in6_addr *saddr, const struct in6_addr *daddr,
767
		struct udp_table *udptable)
L
Linus Torvalds 已提交
768
{
769
	struct sock *sk, *stack[256 / sizeof(struct sock *)];
770
	const struct udphdr *uh = udp_hdr(skb);
771
	struct udp_hslot *hslot = udp_hashslot(udptable, net, ntohs(uh->dest));
L
Linus Torvalds 已提交
772
	int dif;
773
	unsigned int i, count = 0;
L
Linus Torvalds 已提交
774

775
	spin_lock(&hslot->lock);
776
	sk = sk_nulls_head(&hslot->head);
777
	dif = inet6_iif(skb);
778
	sk = udp_v6_mcast_next(net, sk, uh->dest, daddr, uh->source, saddr, dif);
779 780 781 782 783 784 785 786 787
	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
788
		}
L
Linus Torvalds 已提交
789
	}
790 791 792 793 794 795
	/*
	 * before releasing the lock, we must take reference on sockets
	 */
	for (i = 0; i < count; i++)
		sock_hold(stack[i]);

796
	spin_unlock(&hslot->lock);
797 798 799 800 801 802 803 804 805

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

		for (i = 0; i < count; i++)
			sock_put(stack[i]);
	} else {
		kfree_skb(skb);
	}
806
	return 0;
L
Linus Torvalds 已提交
807 808
}

809
int __udp6_lib_rcv(struct sk_buff *skb, struct udp_table *udptable,
810
		   int proto)
L
Linus Torvalds 已提交
811
{
812
	struct net *net = dev_net(skb->dev);
L
Linus Torvalds 已提交
813
	struct sock *sk;
814
	struct udphdr *uh;
815
	const struct in6_addr *saddr, *daddr;
L
Linus Torvalds 已提交
816 817 818
	u32 ulen = 0;

	if (!pskb_may_pull(skb, sizeof(struct udphdr)))
819
		goto discard;
L
Linus Torvalds 已提交
820

821 822
	saddr = &ipv6_hdr(skb)->saddr;
	daddr = &ipv6_hdr(skb)->daddr;
823
	uh = udp_hdr(skb);
L
Linus Torvalds 已提交
824 825

	ulen = ntohs(uh->len);
826 827
	if (ulen > skb->len)
		goto short_packet;
L
Linus Torvalds 已提交
828

829 830
	if (proto == IPPROTO_UDP) {
		/* UDP validates ulen. */
L
Linus Torvalds 已提交
831

832 833 834
		/* Check for jumbo payload */
		if (ulen == 0)
			ulen = skb->len;
L
Linus Torvalds 已提交
835

836 837
		if (ulen < sizeof(*uh))
			goto short_packet;
L
Linus Torvalds 已提交
838

839 840 841
		if (ulen < skb->len) {
			if (pskb_trim_rcsum(skb, ulen))
				goto short_packet;
842 843
			saddr = &ipv6_hdr(skb)->saddr;
			daddr = &ipv6_hdr(skb)->daddr;
844
			uh = udp_hdr(skb);
845 846
		}
	}
L
Linus Torvalds 已提交
847

848
	if (udp6_csum_init(skb, uh, proto))
849
		goto csum_error;
850

851 852
	/*
	 *	Multicast receive code
L
Linus Torvalds 已提交
853
	 */
854
	if (ipv6_addr_is_multicast(daddr))
855 856
		return __udp6_lib_mcast_deliver(net, skb,
				saddr, daddr, udptable);
L
Linus Torvalds 已提交
857 858 859

	/* Unicast */

860
	/*
L
Linus Torvalds 已提交
861 862 863
	 * check socket cache ... must talk to Alan about his plans
	 * for sock caches... i'll skip this for now.
	 */
864
	sk = __udp6_lib_lookup_skb(skb, uh->source, uh->dest, udptable);
865
	if (sk != NULL) {
866 867 868
		int ret;

		ret = udpv6_queue_rcv_skb(sk, skb);
869
		sock_put(sk);
L
Linus Torvalds 已提交
870

871 872 873 874 875
		/* 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
876

S
Stephen Hemminger 已提交
877
		return 0;
L
Linus Torvalds 已提交
878
	}
879

880
	if (!xfrm6_policy_check(NULL, XFRM_POLICY_IN, skb))
881
		goto discard;
882 883

	if (udp_lib_checksum_complete(skb))
884
		goto csum_error;
885 886 887 888 889

	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
890
	return 0;
L
Linus Torvalds 已提交
891

892
short_packet:
893
	LIMIT_NETDEBUG(KERN_DEBUG "UDP%sv6: short packet: From [%pI6c]:%u %d/%d to [%pI6c]:%u\n",
894
		       proto == IPPROTO_UDPLITE ? "-Lite" : "",
895 896 897 898 899 900
		       saddr,
		       ntohs(uh->source),
		       ulen,
		       skb->len,
		       daddr,
		       ntohs(uh->dest));
901 902 903
	goto discard;
csum_error:
	UDP6_INC_STATS_BH(net, UDP_MIB_CSUMERRORS, proto == IPPROTO_UDPLITE);
L
Linus Torvalds 已提交
904
discard:
905
	UDP6_INC_STATS_BH(net, UDP_MIB_INERRORS, proto == IPPROTO_UDPLITE);
L
Linus Torvalds 已提交
906
	kfree_skb(skb);
S
Stephen Hemminger 已提交
907
	return 0;
L
Linus Torvalds 已提交
908
}
909

910
static __inline__ int udpv6_rcv(struct sk_buff *skb)
911
{
912
	return __udp6_lib_rcv(skb, &udp_table, IPPROTO_UDP);
913 914
}

L
Linus Torvalds 已提交
915 916 917 918 919 920 921
/*
 * 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);

922 923 924
	if (up->pending == AF_INET)
		udp_flush_pending_frames(sk);
	else if (up->pending) {
L
Linus Torvalds 已提交
925 926 927
		up->len = 0;
		up->pending = 0;
		ip6_flush_pending_frames(sk);
928
	}
L
Linus Torvalds 已提交
929 930
}

931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971
/**
 * 	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
972 973 974 975
/*
 *	Sending
 */

976
static int udp_v6_push_pending_frames(struct sock *sk)
L
Linus Torvalds 已提交
977 978 979
{
	struct sk_buff *skb;
	struct udphdr *uh;
980
	struct udp_sock  *up = udp_sk(sk);
L
Linus Torvalds 已提交
981
	struct inet_sock *inet = inet_sk(sk);
982
	struct flowi6 *fl6;
L
Linus Torvalds 已提交
983
	int err = 0;
W
Wang Chen 已提交
984
	int is_udplite = IS_UDPLITE(sk);
985
	__wsum csum = 0;
L
Linus Torvalds 已提交
986

987 988 989 990 991
	if (up->pending == AF_INET)
		return udp_push_pending_frames(sk);

	fl6 = &inet->cork.fl.u.ip6;

L
Linus Torvalds 已提交
992 993 994 995 996 997 998
	/* Grab the skbuff where UDP header space exists. */
	if ((skb = skb_peek(&sk->sk_write_queue)) == NULL)
		goto out;

	/*
	 * Create a UDP header
	 */
999
	uh = udp_hdr(skb);
1000 1001
	uh->source = fl6->fl6_sport;
	uh->dest = fl6->fl6_dport;
L
Linus Torvalds 已提交
1002 1003 1004
	uh->len = htons(up->len);
	uh->check = 0;

W
Wang Chen 已提交
1005
	if (is_udplite)
1006
		csum = udplite_csum_outgoing(sk, skb);
1007
	else if (skb->ip_summed == CHECKSUM_PARTIAL) { /* UDP hardware csum */
1008
		udp6_hwcsum_outgoing(sk, skb, &fl6->saddr, &fl6->daddr,
1009 1010 1011
				     up->len);
		goto send;
	} else
1012
		csum = udp_csum_outgoing(sk, skb);
L
Linus Torvalds 已提交
1013

1014
	/* add protocol-dependent pseudo-header */
1015 1016
	uh->check = csum_ipv6_magic(&fl6->saddr, &fl6->daddr,
				    up->len, fl6->flowi6_proto, csum);
L
Linus Torvalds 已提交
1017
	if (uh->check == 0)
1018
		uh->check = CSUM_MANGLED_0;
L
Linus Torvalds 已提交
1019

1020
send:
L
Linus Torvalds 已提交
1021
	err = ip6_push_pending_frames(sk);
E
Eric Dumazet 已提交
1022 1023 1024 1025 1026 1027 1028 1029 1030
	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
1031 1032 1033 1034 1035 1036
out:
	up->len = 0;
	up->pending = 0;
	return err;
}

1037
int udpv6_sendmsg(struct kiocb *iocb, struct sock *sk,
L
Linus Torvalds 已提交
1038 1039 1040 1041 1042 1043 1044
		  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;
1045
	struct in6_addr *daddr, *final_p, final;
L
Linus Torvalds 已提交
1046 1047
	struct ipv6_txoptions *opt = NULL;
	struct ip6_flowlabel *flowlabel = NULL;
1048
	struct flowi6 fl6;
L
Linus Torvalds 已提交
1049 1050 1051 1052
	struct dst_entry *dst;
	int addr_len = msg->msg_namelen;
	int ulen = len;
	int hlimit = -1;
1053
	int tclass = -1;
1054
	int dontfrag = -1;
L
Linus Torvalds 已提交
1055 1056
	int corkreq = up->corkflag || msg->msg_flags&MSG_MORE;
	int err;
1057
	int connected = 0;
W
Wang Chen 已提交
1058
	int is_udplite = IS_UDPLITE(sk);
1059
	int (*getfrag)(void *, char *, int, int, int, struct sk_buff *);
L
Linus Torvalds 已提交
1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084

	/* 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;
1085
		daddr = &sk->sk_v6_daddr;
1086
	} else
L
Linus Torvalds 已提交
1087 1088 1089
		daddr = NULL;

	if (daddr) {
B
Brian Haley 已提交
1090
		if (ipv6_addr_v4mapped(daddr)) {
L
Linus Torvalds 已提交
1091 1092
			struct sockaddr_in sin;
			sin.sin_family = AF_INET;
E
Eric Dumazet 已提交
1093
			sin.sin_port = sin6 ? sin6->sin6_port : inet->inet_dport;
L
Linus Torvalds 已提交
1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107
			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,
1108
	   better check is made in ip6_append_data().
L
Linus Torvalds 已提交
1109 1110 1111
	   */
	if (len > INT_MAX - sizeof(struct udphdr))
		return -EMSGSIZE;
1112

L
Linus Torvalds 已提交
1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130
	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);

1131
	memset(&fl6, 0, sizeof(fl6));
L
Linus Torvalds 已提交
1132 1133 1134 1135 1136

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

1137
		fl6.fl6_dport = sin6->sin6_port;
L
Linus Torvalds 已提交
1138 1139 1140
		daddr = &sin6->sin6_addr;

		if (np->sndflow) {
1141 1142 1143
			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
1144 1145 1146 1147 1148 1149 1150 1151 1152 1153
				if (flowlabel == NULL)
					return -EINVAL;
			}
		}

		/*
		 * Otherwise it will be difficult to maintain
		 * sk->sk_dst_cache.
		 */
		if (sk->sk_state == TCP_ESTABLISHED &&
1154 1155
		    ipv6_addr_equal(daddr, &sk->sk_v6_daddr))
			daddr = &sk->sk_v6_daddr;
L
Linus Torvalds 已提交
1156 1157 1158

		if (addr_len >= sizeof(struct sockaddr_in6) &&
		    sin6->sin6_scope_id &&
1159
		    __ipv6_addr_needs_scope_id(__ipv6_addr_type(daddr)))
1160
			fl6.flowi6_oif = sin6->sin6_scope_id;
L
Linus Torvalds 已提交
1161 1162 1163 1164
	} else {
		if (sk->sk_state != TCP_ESTABLISHED)
			return -EDESTADDRREQ;

1165
		fl6.fl6_dport = inet->inet_dport;
1166
		daddr = &sk->sk_v6_daddr;
1167
		fl6.flowlabel = np->flow_label;
1168
		connected = 1;
L
Linus Torvalds 已提交
1169 1170
	}

1171 1172
	if (!fl6.flowi6_oif)
		fl6.flowi6_oif = sk->sk_bound_dev_if;
L
Linus Torvalds 已提交
1173

1174 1175
	if (!fl6.flowi6_oif)
		fl6.flowi6_oif = np->sticky_pktinfo.ipi6_ifindex;
1176

1177
	fl6.flowi6_mark = sk->sk_mark;
B
Brian Haley 已提交
1178

L
Linus Torvalds 已提交
1179 1180 1181 1182 1183
	if (msg->msg_controllen) {
		opt = &opt_space;
		memset(opt, 0, sizeof(struct ipv6_txoptions));
		opt->tot_len = sizeof(*opt);

1184 1185
		err = ip6_datagram_send_ctl(sock_net(sk), sk, msg, &fl6, opt,
					    &hlimit, &tclass, &dontfrag);
L
Linus Torvalds 已提交
1186 1187 1188 1189
		if (err < 0) {
			fl6_sock_release(flowlabel);
			return err;
		}
1190 1191
		if ((fl6.flowlabel&IPV6_FLOWLABEL_MASK) && !flowlabel) {
			flowlabel = fl6_sock_lookup(sk, fl6.flowlabel);
L
Linus Torvalds 已提交
1192 1193 1194 1195 1196
			if (flowlabel == NULL)
				return -EINVAL;
		}
		if (!(opt->opt_nflen|opt->opt_flen))
			opt = NULL;
1197
		connected = 0;
L
Linus Torvalds 已提交
1198 1199 1200
	}
	if (opt == NULL)
		opt = np->opt;
1201 1202 1203
	if (flowlabel)
		opt = fl6_merge_options(&opt_space, flowlabel, opt);
	opt = ipv6_fixup_options(&opt_space, opt);
L
Linus Torvalds 已提交
1204

1205
	fl6.flowi6_proto = sk->sk_protocol;
1206
	if (!ipv6_addr_any(daddr))
A
Alexey Dobriyan 已提交
1207
		fl6.daddr = *daddr;
1208
	else
1209 1210
		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
1211
		fl6.saddr = np->saddr;
1212
	fl6.fl6_sport = inet->inet_sport;
1213

1214
	final_p = fl6_update_dst(&fl6, opt, &final);
1215
	if (final_p)
1216
		connected = 0;
L
Linus Torvalds 已提交
1217

1218 1219
	if (!fl6.flowi6_oif && ipv6_addr_is_multicast(&fl6.daddr)) {
		fl6.flowi6_oif = np->mcast_oif;
1220
		connected = 0;
1221 1222
	} else if (!fl6.flowi6_oif)
		fl6.flowi6_oif = np->ucast_oif;
L
Linus Torvalds 已提交
1223

1224
	security_sk_classify_flow(sk, flowi6_to_flowi(&fl6));
V
Venkat Yekkirala 已提交
1225

1226
	dst = ip6_sk_dst_lookup_flow(sk, &fl6, final_p);
1227 1228 1229
	if (IS_ERR(dst)) {
		err = PTR_ERR(dst);
		dst = NULL;
L
Linus Torvalds 已提交
1230
		goto out;
1231
	}
L
Linus Torvalds 已提交
1232 1233

	if (hlimit < 0) {
1234
		if (ipv6_addr_is_multicast(&fl6.daddr))
L
Linus Torvalds 已提交
1235 1236 1237 1238
			hlimit = np->mcast_hops;
		else
			hlimit = np->hop_limit;
		if (hlimit < 0)
1239
			hlimit = ip6_dst_hoplimit(dst);
L
Linus Torvalds 已提交
1240 1241
	}

G
Gerrit Renker 已提交
1242
	if (tclass < 0)
1243 1244
		tclass = np->tclass;

L
Linus Torvalds 已提交
1245 1246 1247 1248 1249 1250 1251 1252 1253 1254
	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);

1255
		LIMIT_NETDEBUG(KERN_DEBUG "udp cork app bug 2\n");
L
Linus Torvalds 已提交
1256 1257 1258 1259 1260 1261 1262
		err = -EINVAL;
		goto out;
	}

	up->pending = AF_INET6;

do_append_data:
1263 1264
	if (dontfrag < 0)
		dontfrag = np->dontfrag;
L
Linus Torvalds 已提交
1265
	up->len += ulen;
1266 1267
	getfrag  =  is_udplite ?  udplite_getfrag : ip_generic_getfrag;
	err = ip6_append_data(sk, getfrag, msg->msg_iov, ulen,
1268
		sizeof(struct udphdr), hlimit, tclass, opt, &fl6,
1269
		(struct rt6_info*)dst,
1270
		corkreq ? msg->msg_flags|MSG_MORE : msg->msg_flags, dontfrag);
L
Linus Torvalds 已提交
1271 1272 1273
	if (err)
		udp_v6_flush_pending_frames(sk);
	else if (!corkreq)
1274
		err = udp_v6_push_pending_frames(sk);
H
Herbert Xu 已提交
1275 1276
	else if (unlikely(skb_queue_empty(&sk->sk_write_queue)))
		up->pending = 0;
L
Linus Torvalds 已提交
1277

1278 1279 1280
	if (dst) {
		if (connected) {
			ip6_dst_store(sk, dst,
1281 1282
				      ipv6_addr_equal(&fl6.daddr, &sk->sk_v6_daddr) ?
				      &sk->sk_v6_daddr : NULL,
1283
#ifdef CONFIG_IPV6_SUBTREES
1284
				      ipv6_addr_equal(&fl6.saddr, &np->saddr) ?
1285 1286 1287
				      &np->saddr :
#endif
				      NULL);
1288 1289 1290
		} else {
			dst_release(dst);
		}
1291
		dst = NULL;
1292 1293
	}

L
Linus Torvalds 已提交
1294 1295 1296 1297
	if (err > 0)
		err = np->recverr ? net_xmit_errno(err) : 0;
	release_sock(sk);
out:
1298
	dst_release(dst);
L
Linus Torvalds 已提交
1299
	fl6_sock_release(flowlabel);
1300
	if (!err)
L
Linus Torvalds 已提交
1301
		return len;
1302 1303 1304 1305 1306 1307 1308 1309
	/*
	 * 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)) {
1310 1311
		UDP6_INC_STATS_USER(sock_net(sk),
				UDP_MIB_SNDBUFERRORS, is_udplite);
1312
	}
L
Linus Torvalds 已提交
1313 1314 1315 1316 1317 1318 1319 1320 1321 1322
	return err;

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

1323
void udpv6_destroy_sock(struct sock *sk)
L
Linus Torvalds 已提交
1324
{
T
Tom Parkin 已提交
1325
	struct udp_sock *up = udp_sk(sk);
L
Linus Torvalds 已提交
1326 1327 1328 1329
	lock_sock(sk);
	udp_v6_flush_pending_frames(sk);
	release_sock(sk);

T
Tom Parkin 已提交
1330 1331 1332 1333 1334 1335 1336
	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
1337 1338 1339 1340 1341 1342
	inet6_destroy_sock(sk);
}

/*
 *	Socket option code for UDP
 */
1343
int udpv6_setsockopt(struct sock *sk, int level, int optname,
1344
		     char __user *optval, unsigned int optlen)
1345
{
1346
	if (level == SOL_UDP  ||  level == SOL_UDPLITE)
1347 1348
		return udp_lib_setsockopt(sk, level, optname, optval, optlen,
					  udp_v6_push_pending_frames);
1349
	return ipv6_setsockopt(sk, level, optname, optval, optlen);
1350 1351 1352
}

#ifdef CONFIG_COMPAT
1353
int compat_udpv6_setsockopt(struct sock *sk, int level, int optname,
1354
			    char __user *optval, unsigned int optlen)
1355
{
1356
	if (level == SOL_UDP  ||  level == SOL_UDPLITE)
1357 1358
		return udp_lib_setsockopt(sk, level, optname, optval, optlen,
					  udp_v6_push_pending_frames);
1359
	return compat_ipv6_setsockopt(sk, level, optname, optval, optlen);
1360 1361 1362
}
#endif

1363 1364
int udpv6_getsockopt(struct sock *sk, int level, int optname,
		     char __user *optval, int __user *optlen)
1365
{
1366
	if (level == SOL_UDP  ||  level == SOL_UDPLITE)
1367
		return udp_lib_getsockopt(sk, level, optname, optval, optlen);
1368
	return ipv6_getsockopt(sk, level, optname, optval, optlen);
1369 1370 1371
}

#ifdef CONFIG_COMPAT
1372 1373
int compat_udpv6_getsockopt(struct sock *sk, int level, int optname,
			    char __user *optval, int __user *optlen)
1374
{
1375
	if (level == SOL_UDP  ||  level == SOL_UDPLITE)
1376
		return udp_lib_getsockopt(sk, level, optname, optval, optlen);
1377
	return compat_ipv6_getsockopt(sk, level, optname, optval, optlen);
1378 1379 1380
}
#endif

1381
static const struct inet6_protocol udpv6_protocol = {
L
Linus Torvalds 已提交
1382 1383 1384 1385 1386 1387 1388
	.handler	=	udpv6_rcv,
	.err_handler	=	udpv6_err,
	.flags		=	INET6_PROTO_NOPOLICY|INET6_PROTO_FINAL,
};

/* ------------------------------------------------------------------------ */
#ifdef CONFIG_PROC_FS
1389
int udp6_seq_show(struct seq_file *seq, void *v)
L
Linus Torvalds 已提交
1390
{
1391 1392 1393 1394 1395 1396 1397 1398 1399
	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
1400 1401 1402
	return 0;
}

1403 1404 1405 1406 1407 1408 1409 1410
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
1411 1412 1413
static struct udp_seq_afinfo udp6_seq_afinfo = {
	.name		= "udp6",
	.family		= AF_INET6,
1414
	.udp_table	= &udp_table,
1415
	.seq_fops	= &udp6_afinfo_seq_fops,
1416 1417 1418
	.seq_ops	= {
		.show		= udp6_seq_show,
	},
L
Linus Torvalds 已提交
1419 1420
};

1421
int __net_init udp6_proc_init(struct net *net)
L
Linus Torvalds 已提交
1422
{
1423
	return udp_proc_register(net, &udp6_seq_afinfo);
L
Linus Torvalds 已提交
1424 1425
}

1426 1427
void udp6_proc_exit(struct net *net) {
	udp_proc_unregister(net, &udp6_seq_afinfo);
L
Linus Torvalds 已提交
1428 1429 1430
}
#endif /* CONFIG_PROC_FS */

E
Eric Dumazet 已提交
1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441
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
1442 1443 1444
/* ------------------------------------------------------------------------ */

struct proto udpv6_prot = {
1445 1446
	.name		   = "UDPv6",
	.owner		   = THIS_MODULE,
1447
	.close		   = udp_lib_close,
1448 1449 1450 1451 1452 1453 1454 1455
	.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,
1456
	.backlog_rcv	   = __udpv6_queue_rcv_skb,
1457 1458
	.hash		   = udp_lib_hash,
	.unhash		   = udp_lib_unhash,
E
Eric Dumazet 已提交
1459
	.rehash		   = udp_v6_rehash,
1460
	.get_port	   = udp_v6_get_port,
H
Hideo Aoki 已提交
1461 1462 1463 1464
	.memory_allocated  = &udp_memory_allocated,
	.sysctl_mem	   = sysctl_udp_mem,
	.sysctl_wmem	   = &sysctl_udp_wmem_min,
	.sysctl_rmem	   = &sysctl_udp_rmem_min,
1465
	.obj_size	   = sizeof(struct udp6_sock),
1466
	.slab_flags	   = SLAB_DESTROY_BY_RCU,
1467
	.h.udp_table	   = &udp_table,
1468
#ifdef CONFIG_COMPAT
1469 1470
	.compat_setsockopt = compat_udpv6_setsockopt,
	.compat_getsockopt = compat_udpv6_getsockopt,
1471
#endif
E
Eric Dumazet 已提交
1472
	.clear_sk	   = udp_v6_clear_sk,
L
Linus Torvalds 已提交
1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484
};

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,
};


1485
int __init udpv6_init(void)
L
Linus Torvalds 已提交
1486
{
1487 1488
	int ret;

1489 1490
	ret = inet6_add_protocol(&udpv6_protocol, IPPROTO_UDP);
	if (ret)
1491
		goto out;
1492

1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503
	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;
}

1504
void udpv6_exit(void)
1505 1506 1507
{
	inet6_unregister_protosw(&udpv6_protosw);
	inet6_del_protocol(&udpv6_protocol, IPPROTO_UDP);
L
Linus Torvalds 已提交
1508
}