udp.c 39.1 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 48 49 50
#include <net/ip6_checksum.h>
#include <net/xfrm.h>

#include <linux/proc_fs.h>
#include <linux/seq_file.h>
51
#include <trace/events/skb.h>
52
#include "udp_impl.h"
L
Linus Torvalds 已提交
53

54 55 56 57
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);
58 59
	__be32 sk1_rcv_saddr = sk_rcv_saddr(sk);
	__be32 sk2_rcv_saddr = sk_rcv_saddr(sk2);
60 61 62 63 64 65 66
	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)
67
		return (!sk2_ipv6only &&
E
Eric Dumazet 已提交
68 69
			(!sk1_rcv_saddr || !sk2_rcv_saddr ||
			  sk1_rcv_saddr == sk2_rcv_saddr));
70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85

	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;
}

86 87 88 89 90 91 92 93
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
94
	else if (ipv6_addr_v4mapped(addr6))
95
		hash = jhash_1word((__force u32)addr6->s6_addr32[3], mix);
96
	else
97
		hash = jhash2((__force u32 *)addr6->s6_addr32, 4, mix);
98 99 100 101 102

	return hash ^ port;
}


103
int udp_v6_get_port(struct sock *sk, unsigned short snum)
L
Linus Torvalds 已提交
104
{
E
Eric Dumazet 已提交
105 106
	unsigned int hash2_nulladdr =
		udp6_portaddr_hash(sock_net(sk), &in6addr_any, snum);
107
	unsigned int hash2_partial =
E
Eric Dumazet 已提交
108 109
		udp6_portaddr_hash(sock_net(sk), &inet6_sk(sk)->rcv_saddr, 0);

110
	/* precompute partial secondary hash */
E
Eric Dumazet 已提交
111 112
	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
113 114
}

E
Eric Dumazet 已提交
115 116 117 118 119 120 121 122 123
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);
}

124 125
static inline int compute_score(struct sock *sk, struct net *net,
				unsigned short hnum,
126 127
				const struct in6_addr *saddr, __be16 sport,
				const struct in6_addr *daddr, __be16 dport,
128 129 130 131
				int dif)
{
	int score = -1;

132
	if (net_eq(sock_net(sk), net) && udp_sk(sk)->udp_port_hash == hnum &&
133 134 135 136 137
			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
138 139
		if (inet->inet_dport) {
			if (inet->inet_dport != sport)
140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161
				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
162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230
#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;
	int score, badness;

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;
			if (score == SCORE2_MAX)
				goto exact_match;
		}
	}
	/*
	 * 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
231
		if (unlikely(!atomic_inc_not_zero_hint(&result->sk_refcnt, 2)))
E
Eric Dumazet 已提交
232 233 234 235 236 237 238 239 240 241
			result = NULL;
		else if (unlikely(compute_score2(result, net, saddr, sport,
				  daddr, hnum, dif) < badness)) {
			sock_put(result);
			goto begin;
		}
	}
	return result;
}

242
struct sock *__udp6_lib_lookup(struct net *net,
243 244
				      const struct in6_addr *saddr, __be16 sport,
				      const struct in6_addr *daddr, __be16 dport,
245
				      int dif, struct udp_table *udptable)
L
Linus Torvalds 已提交
246
{
247
	struct sock *sk, *result;
248
	struct hlist_nulls_node *node;
L
Linus Torvalds 已提交
249
	unsigned short hnum = ntohs(dport);
E
Eric Dumazet 已提交
250 251
	unsigned int hash2, slot2, slot = udp_hashfn(net, hnum, udptable->mask);
	struct udp_hslot *hslot2, *hslot = &udptable->hash[slot];
252
	int score, badness;
253

254
	rcu_read_lock();
E
Eric Dumazet 已提交
255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271
	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;

272 273
			result = udp6_lib_lookup2(net, saddr, sport,
						  &in6addr_any, hnum, dif,
E
Eric Dumazet 已提交
274 275 276 277 278
						  hslot2, slot2);
		}
		rcu_read_unlock();
		return result;
	}
279 280 281
begin:
	result = NULL;
	badness = -1;
282
	sk_nulls_for_each_rcu(sk, node, &hslot->head) {
283 284 285 286
		score = compute_score(sk, net, hnum, saddr, sport, daddr, dport, dif);
		if (score > badness) {
			result = sk;
			badness = score;
L
Linus Torvalds 已提交
287 288
		}
	}
289 290 291 292 293
	/*
	 * 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
294
	if (get_nulls_value(node) != slot)
295 296
		goto begin;

297
	if (result) {
E
Eric Dumazet 已提交
298
		if (unlikely(!atomic_inc_not_zero_hint(&result->sk_refcnt, 2)))
299 300 301 302 303 304 305 306
			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
307 308
	return result;
}
309
EXPORT_SYMBOL_GPL(__udp6_lib_lookup);
L
Linus Torvalds 已提交
310

311 312
static struct sock *__udp6_lib_lookup_skb(struct sk_buff *skb,
					  __be16 sport, __be16 dport,
313
					  struct udp_table *udptable)
314
{
315
	struct sock *sk;
316
	const struct ipv6hdr *iph = ipv6_hdr(skb);
317

318 319
	if (unlikely(sk = skb_steal_sock(skb)))
		return sk;
E
Eric Dumazet 已提交
320 321 322
	return __udp6_lib_lookup(dev_net(skb_dst(skb)->dev), &iph->saddr, sport,
				 &iph->daddr, dport, inet6_iif(skb),
				 udptable);
323 324
}

325 326 327 328 329 330 331 332
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
333 334 335 336 337
/*
 * 	This should be easy, if there is something there we
 * 	return it, otherwise we block.
 */

338
int udpv6_recvmsg(struct kiocb *iocb, struct sock *sk,
L
Linus Torvalds 已提交
339 340 341 342 343
		  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);
344
	struct sk_buff *skb;
345
	unsigned int ulen, copied;
346
	int peeked, off = 0;
347 348
	int err;
	int is_udplite = IS_UDPLITE(sk);
349
	int is_udp4;
350
	bool slow;
L
Linus Torvalds 已提交
351

352
	if (addr_len)
353
		*addr_len = sizeof(struct sockaddr_in6);
354

L
Linus Torvalds 已提交
355 356 357
	if (flags & MSG_ERRQUEUE)
		return ipv6_recv_error(sk, msg, len);

358 359 360
	if (np->rxpmtu && np->rxopt.bits.rxpmtu)
		return ipv6_recv_rxpmtu(sk, msg, len);

L
Linus Torvalds 已提交
361
try_again:
362
	skb = __skb_recv_datagram(sk, flags | (noblock ? MSG_DONTWAIT : 0),
363
				  &peeked, &off, &err);
L
Linus Torvalds 已提交
364 365 366
	if (!skb)
		goto out;

367
	ulen = skb->len - sizeof(struct udphdr);
368 369 370 371
	copied = len;
	if (copied > ulen)
		copied = ulen;
	else if (copied < ulen)
372
		msg->msg_flags |= MSG_TRUNC;
L
Linus Torvalds 已提交
373

374 375
	is_udp4 = (skb->protocol == htons(ETH_P_IP));

376
	/*
377 378 379
	 * 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.
380 381
	 */

382
	if (copied < ulen || UDP_SKB_CB(skb)->partial_cov) {
383
		if (udp_lib_checksum_complete(skb))
L
Linus Torvalds 已提交
384
			goto csum_copy_err;
385 386
	}

387
	if (skb_csum_unnecessary(skb))
388
		err = skb_copy_datagram_iovec(skb, sizeof(struct udphdr),
389
					      msg->msg_iov, copied);
390
	else {
L
Linus Torvalds 已提交
391 392 393 394
		err = skb_copy_and_csum_datagram_iovec(skb, sizeof(struct udphdr), msg->msg_iov);
		if (err == -EINVAL)
			goto csum_copy_err;
	}
395 396
	if (unlikely(err)) {
		trace_kfree_skb(skb, udpv6_recvmsg);
L
Linus Torvalds 已提交
397
		goto out_free;
398
	}
399 400 401 402 403 404 405 406
	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);
	}
407

408
	sock_recv_ts_and_drops(msg, sk, skb);
L
Linus Torvalds 已提交
409 410 411 412

	/* Copy the address. */
	if (msg->msg_name) {
		struct sockaddr_in6 *sin6;
413

L
Linus Torvalds 已提交
414 415
		sin6 = (struct sockaddr_in6 *) msg->msg_name;
		sin6->sin6_family = AF_INET6;
416
		sin6->sin6_port = udp_hdr(skb)->source;
L
Linus Torvalds 已提交
417 418 419
		sin6->sin6_flowinfo = 0;
		sin6->sin6_scope_id = 0;

420
		if (is_udp4)
B
Brian Haley 已提交
421 422
			ipv6_addr_set_v4mapped(ip_hdr(skb)->saddr,
					       &sin6->sin6_addr);
L
Linus Torvalds 已提交
423
		else {
A
Alexey Dobriyan 已提交
424
			sin6->sin6_addr = ipv6_hdr(skb)->saddr;
L
Linus Torvalds 已提交
425 426 427 428 429
			if (ipv6_addr_type(&sin6->sin6_addr) & IPV6_ADDR_LINKLOCAL)
				sin6->sin6_scope_id = IP6CB(skb)->iif;
		}

	}
430
	if (is_udp4) {
L
Linus Torvalds 已提交
431 432 433 434 435
		if (inet->cmsg_flags)
			ip_cmsg_recv(msg, skb);
	} else {
		if (np->rxopt.all)
			datagram_recv_ctl(sk, msg, skb);
436
	}
L
Linus Torvalds 已提交
437

438
	err = copied;
L
Linus Torvalds 已提交
439
	if (flags & MSG_TRUNC)
440
		err = ulen;
L
Linus Torvalds 已提交
441 442

out_free:
443
	skb_free_datagram_locked(sk, skb);
L
Linus Torvalds 已提交
444 445 446 447
out:
	return err;

csum_copy_err:
448
	slow = lock_sock_fast(sk);
449 450 451 452 453 454 455 456
	if (!skb_kill_datagram(sk, skb, flags)) {
		if (is_udp4)
			UDP_INC_STATS_USER(sock_net(sk),
					UDP_MIB_INERRORS, is_udplite);
		else
			UDP6_INC_STATS_USER(sock_net(sk),
					UDP_MIB_INERRORS, is_udplite);
	}
457
	unlock_sock_fast(sk, slow);
L
Linus Torvalds 已提交
458

459
	if (noblock)
L
Linus Torvalds 已提交
460
		return -EAGAIN;
461 462 463

	/* starting over for a new packet */
	msg->msg_flags &= ~MSG_TRUNC;
L
Linus Torvalds 已提交
464 465 466
	goto try_again;
}

467
void __udp6_lib_err(struct sk_buff *skb, struct inet6_skb_parm *opt,
468
		    u8 type, u8 code, int offset, __be32 info,
469
		    struct udp_table *udptable)
L
Linus Torvalds 已提交
470 471
{
	struct ipv6_pinfo *np;
472 473 474
	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
475 476 477 478
	struct udphdr *uh = (struct udphdr*)(skb->data+offset);
	struct sock *sk;
	int err;

479
	sk = __udp6_lib_lookup(dev_net(skb->dev), daddr, uh->dest,
480
			       saddr, uh->source, inet6_iif(skb), udptable);
L
Linus Torvalds 已提交
481 482 483
	if (sk == NULL)
		return;

484 485 486
	if (type == ICMPV6_PKT_TOOBIG)
		ip6_sk_update_pmtu(skb, sk, info);

L
Linus Torvalds 已提交
487 488 489 490 491 492 493 494
	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;

495
	if (np->recverr)
L
Linus Torvalds 已提交
496
		ipv6_icmp_error(sk, skb, err, uh->dest, ntohl(info), (u8 *)(uh+1));
497

L
Linus Torvalds 已提交
498 499 500 501 502 503
	sk->sk_err = err;
	sk->sk_error_report(sk);
out:
	sock_put(sk);
}

504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525
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;
}

526
static __inline__ void udpv6_err(struct sk_buff *skb,
527 528
				 struct inet6_skb_parm *opt, u8 type,
				 u8 code, int offset, __be32 info     )
529
{
530
	__udp6_lib_err(skb, opt, type, code, offset, info, &udp_table);
531 532
}

533 534 535 536 537 538 539 540
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);

541
int udpv6_queue_rcv_skb(struct sock *sk, struct sk_buff *skb)
L
Linus Torvalds 已提交
542
{
543
	struct udp_sock *up = udp_sk(sk);
544
	int rc;
W
Wang Chen 已提交
545
	int is_udplite = IS_UDPLITE(sk);
546

547 548
	if (!xfrm6_policy_check(sk, XFRM_POLICY_IN, skb))
		goto drop;
L
Linus Torvalds 已提交
549

550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580
	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 */
	}

581 582 583
	/*
	 * UDP-Lite specific tests, ignored on UDP sockets (see net/ipv4/udp.c).
	 */
W
Wang Chen 已提交
584
	if ((is_udplite & UDPLITE_RECV_CC)  &&  UDP_SKB_CB(skb)->partial_cov) {
585 586 587 588 589 590 591 592 593 594 595 596 597

		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
598 599
	}

600
	if (rcu_access_pointer(sk->sk_filter)) {
601 602 603
		if (udp_lib_checksum_complete(skb))
			goto drop;
	}
604

605 606 607
	if (sk_rcvqueues_full(sk, skb, sk->sk_rcvbuf))
		goto drop;

608
	skb_dst_drop(skb);
609

610 611 612 613 614 615 616 617 618
	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);
619 620

	return rc;
621
drop:
622
	UDP6_INC_STATS_BH(sock_net(sk), UDP_MIB_INERRORS, is_udplite);
623
	atomic_inc(&sk->sk_drops);
624 625
	kfree_skb(skb);
	return -1;
L
Linus Torvalds 已提交
626 627
}

628
static struct sock *udp_v6_mcast_next(struct net *net, struct sock *sk,
629 630
				      __be16 loc_port, const struct in6_addr *loc_addr,
				      __be16 rmt_port, const struct in6_addr *rmt_addr,
L
Linus Torvalds 已提交
631 632
				      int dif)
{
633
	struct hlist_nulls_node *node;
L
Linus Torvalds 已提交
634 635 636
	struct sock *s = sk;
	unsigned short num = ntohs(loc_port);

637
	sk_nulls_for_each_from(s, node) {
L
Linus Torvalds 已提交
638 639
		struct inet_sock *inet = inet_sk(s);

640
		if (!net_eq(sock_net(s), net))
641 642
			continue;

643 644
		if (udp_sk(s)->udp_port_hash == num &&
		    s->sk_family == PF_INET6) {
L
Linus Torvalds 已提交
645
			struct ipv6_pinfo *np = inet6_sk(s);
E
Eric Dumazet 已提交
646 647
			if (inet->inet_dport) {
				if (inet->inet_dport != rmt_port)
L
Linus Torvalds 已提交
648 649 650 651 652 653 654 655 656 657
					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)) {
658 659
				if (!ipv6_addr_equal(&np->rcv_saddr, loc_addr))
					continue;
L
Linus Torvalds 已提交
660
			}
S
Stephen Hemminger 已提交
661
			if (!inet6_mc_check(s, loc_addr, rmt_addr))
L
Linus Torvalds 已提交
662 663 664 665 666 667 668
				continue;
			return s;
		}
	}
	return NULL;
}

669 670 671
static void flush_stack(struct sock **stack, unsigned int count,
			struct sk_buff *skb, unsigned int final)
{
672
	struct sk_buff *skb1 = NULL;
673
	struct sock *sk;
674
	unsigned int i;
675 676

	for (i = 0; i < count; i++) {
677
		sk = stack[i];
678 679 680 681 682 683 684 685
		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));
686
		}
687 688 689

		if (skb1 && udpv6_queue_rcv_skb(sk, skb1) <= 0)
			skb1 = NULL;
690
	}
691 692
	if (unlikely(skb1))
		kfree_skb(skb1);
693
}
L
Linus Torvalds 已提交
694 695 696 697
/*
 * Note: called only from the BH handler context,
 * so we don't need to lock the hashes.
 */
698
static int __udp6_lib_mcast_deliver(struct net *net, struct sk_buff *skb,
699
		const struct in6_addr *saddr, const struct in6_addr *daddr,
700
		struct udp_table *udptable)
L
Linus Torvalds 已提交
701
{
702
	struct sock *sk, *stack[256 / sizeof(struct sock *)];
703
	const struct udphdr *uh = udp_hdr(skb);
704
	struct udp_hslot *hslot = udp_hashslot(udptable, net, ntohs(uh->dest));
L
Linus Torvalds 已提交
705
	int dif;
706
	unsigned int i, count = 0;
L
Linus Torvalds 已提交
707

708
	spin_lock(&hslot->lock);
709
	sk = sk_nulls_head(&hslot->head);
710
	dif = inet6_iif(skb);
711
	sk = udp_v6_mcast_next(net, sk, uh->dest, daddr, uh->source, saddr, dif);
712 713 714 715 716 717 718 719 720
	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
721
		}
L
Linus Torvalds 已提交
722
	}
723 724 725 726 727 728
	/*
	 * before releasing the lock, we must take reference on sockets
	 */
	for (i = 0; i < count; i++)
		sock_hold(stack[i]);

729
	spin_unlock(&hslot->lock);
730 731 732 733 734 735 736 737 738

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

		for (i = 0; i < count; i++)
			sock_put(stack[i]);
	} else {
		kfree_skb(skb);
	}
739
	return 0;
L
Linus Torvalds 已提交
740 741
}

742 743
static inline int udp6_csum_init(struct sk_buff *skb, struct udphdr *uh,
				 int proto)
744
{
745 746 747 748 749
	int err;

	UDP_SKB_CB(skb)->partial_cov = 0;
	UDP_SKB_CB(skb)->cscov = skb->len;

750
	if (proto == IPPROTO_UDPLITE) {
751 752 753 754 755
		err = udplite_checksum_init(skb, uh);
		if (err)
			return err;
	}

756 757 758 759 760 761 762 763
	if (uh->check == 0) {
		/* RFC 2460 section 8.1 says that we SHOULD log
		   this error. Well, it is reasonable.
		 */
		LIMIT_NETDEBUG(KERN_INFO "IPv6: udp checksum is 0\n");
		return 1;
	}
	if (skb->ip_summed == CHECKSUM_COMPLETE &&
764
	    !csum_ipv6_magic(&ipv6_hdr(skb)->saddr, &ipv6_hdr(skb)->daddr,
765
			     skb->len, proto, skb->csum))
766 767
		skb->ip_summed = CHECKSUM_UNNECESSARY;

768
	if (!skb_csum_unnecessary(skb))
769 770
		skb->csum = ~csum_unfold(csum_ipv6_magic(&ipv6_hdr(skb)->saddr,
							 &ipv6_hdr(skb)->daddr,
771
							 skb->len, proto, 0));
772

773
	return 0;
774 775
}

776
int __udp6_lib_rcv(struct sk_buff *skb, struct udp_table *udptable,
777
		   int proto)
L
Linus Torvalds 已提交
778
{
779
	struct net *net = dev_net(skb->dev);
L
Linus Torvalds 已提交
780
	struct sock *sk;
781
	struct udphdr *uh;
782
	const struct in6_addr *saddr, *daddr;
L
Linus Torvalds 已提交
783 784 785
	u32 ulen = 0;

	if (!pskb_may_pull(skb, sizeof(struct udphdr)))
786
		goto discard;
L
Linus Torvalds 已提交
787

788 789
	saddr = &ipv6_hdr(skb)->saddr;
	daddr = &ipv6_hdr(skb)->daddr;
790
	uh = udp_hdr(skb);
L
Linus Torvalds 已提交
791 792

	ulen = ntohs(uh->len);
793 794
	if (ulen > skb->len)
		goto short_packet;
L
Linus Torvalds 已提交
795

796 797
	if (proto == IPPROTO_UDP) {
		/* UDP validates ulen. */
L
Linus Torvalds 已提交
798

799 800 801
		/* Check for jumbo payload */
		if (ulen == 0)
			ulen = skb->len;
L
Linus Torvalds 已提交
802

803 804
		if (ulen < sizeof(*uh))
			goto short_packet;
L
Linus Torvalds 已提交
805

806 807 808
		if (ulen < skb->len) {
			if (pskb_trim_rcsum(skb, ulen))
				goto short_packet;
809 810
			saddr = &ipv6_hdr(skb)->saddr;
			daddr = &ipv6_hdr(skb)->daddr;
811
			uh = udp_hdr(skb);
812 813
		}
	}
L
Linus Torvalds 已提交
814

815 816 817
	if (udp6_csum_init(skb, uh, proto))
		goto discard;

818 819
	/*
	 *	Multicast receive code
L
Linus Torvalds 已提交
820
	 */
821
	if (ipv6_addr_is_multicast(daddr))
822 823
		return __udp6_lib_mcast_deliver(net, skb,
				saddr, daddr, udptable);
L
Linus Torvalds 已提交
824 825 826

	/* Unicast */

827
	/*
L
Linus Torvalds 已提交
828 829 830
	 * check socket cache ... must talk to Alan about his plans
	 * for sock caches... i'll skip this for now.
	 */
831
	sk = __udp6_lib_lookup_skb(skb, uh->source, uh->dest, udptable);
832 833 834
	if (sk != NULL) {
		int ret = udpv6_queue_rcv_skb(sk, skb);
		sock_put(sk);
L
Linus Torvalds 已提交
835

836 837 838 839 840
		/* 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
841

S
Stephen Hemminger 已提交
842
		return 0;
L
Linus Torvalds 已提交
843
	}
844

845
	if (!xfrm6_policy_check(NULL, XFRM_POLICY_IN, skb))
846
		goto discard;
847 848

	if (udp_lib_checksum_complete(skb))
Z
Zhu Yi 已提交
849
		goto discard;
850 851 852 853 854

	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
855
	return 0;
L
Linus Torvalds 已提交
856

857
short_packet:
858
	LIMIT_NETDEBUG(KERN_DEBUG "UDP%sv6: short packet: From [%pI6c]:%u %d/%d to [%pI6c]:%u\n",
859
		       proto == IPPROTO_UDPLITE ? "-Lite" : "",
860 861 862 863 864 865
		       saddr,
		       ntohs(uh->source),
		       ulen,
		       skb->len,
		       daddr,
		       ntohs(uh->dest));
L
Linus Torvalds 已提交
866 867

discard:
868
	UDP6_INC_STATS_BH(net, UDP_MIB_INERRORS, proto == IPPROTO_UDPLITE);
L
Linus Torvalds 已提交
869
	kfree_skb(skb);
S
Stephen Hemminger 已提交
870
	return 0;
L
Linus Torvalds 已提交
871
}
872

873
static __inline__ int udpv6_rcv(struct sk_buff *skb)
874
{
875
	return __udp6_lib_rcv(skb, &udp_table, IPPROTO_UDP);
876 877
}

L
Linus Torvalds 已提交
878 879 880 881 882 883 884
/*
 * 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);

885 886 887
	if (up->pending == AF_INET)
		udp_flush_pending_frames(sk);
	else if (up->pending) {
L
Linus Torvalds 已提交
888 889 890
		up->len = 0;
		up->pending = 0;
		ip6_flush_pending_frames(sk);
891
	}
L
Linus Torvalds 已提交
892 893
}

894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934
/**
 * 	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
935 936 937 938
/*
 *	Sending
 */

939
static int udp_v6_push_pending_frames(struct sock *sk)
L
Linus Torvalds 已提交
940 941 942
{
	struct sk_buff *skb;
	struct udphdr *uh;
943
	struct udp_sock  *up = udp_sk(sk);
L
Linus Torvalds 已提交
944
	struct inet_sock *inet = inet_sk(sk);
945
	struct flowi6 *fl6 = &inet->cork.fl.u.ip6;
L
Linus Torvalds 已提交
946
	int err = 0;
W
Wang Chen 已提交
947
	int is_udplite = IS_UDPLITE(sk);
948
	__wsum csum = 0;
L
Linus Torvalds 已提交
949 950 951 952 953 954 955 956

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

	/*
	 * Create a UDP header
	 */
957
	uh = udp_hdr(skb);
958 959
	uh->source = fl6->fl6_sport;
	uh->dest = fl6->fl6_dport;
L
Linus Torvalds 已提交
960 961 962
	uh->len = htons(up->len);
	uh->check = 0;

W
Wang Chen 已提交
963
	if (is_udplite)
964
		csum = udplite_csum_outgoing(sk, skb);
965
	else if (skb->ip_summed == CHECKSUM_PARTIAL) { /* UDP hardware csum */
966
		udp6_hwcsum_outgoing(sk, skb, &fl6->saddr, &fl6->daddr,
967 968 969
				     up->len);
		goto send;
	} else
970
		csum = udp_csum_outgoing(sk, skb);
L
Linus Torvalds 已提交
971

972
	/* add protocol-dependent pseudo-header */
973 974
	uh->check = csum_ipv6_magic(&fl6->saddr, &fl6->daddr,
				    up->len, fl6->flowi6_proto, csum);
L
Linus Torvalds 已提交
975
	if (uh->check == 0)
976
		uh->check = CSUM_MANGLED_0;
L
Linus Torvalds 已提交
977

978
send:
L
Linus Torvalds 已提交
979
	err = ip6_push_pending_frames(sk);
E
Eric Dumazet 已提交
980 981 982 983 984 985 986 987 988
	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
989 990 991 992 993 994
out:
	up->len = 0;
	up->pending = 0;
	return err;
}

995
int udpv6_sendmsg(struct kiocb *iocb, struct sock *sk,
L
Linus Torvalds 已提交
996 997 998 999 1000 1001 1002
		  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;
1003
	struct in6_addr *daddr, *final_p, final;
L
Linus Torvalds 已提交
1004 1005
	struct ipv6_txoptions *opt = NULL;
	struct ip6_flowlabel *flowlabel = NULL;
1006
	struct flowi6 fl6;
L
Linus Torvalds 已提交
1007 1008 1009 1010
	struct dst_entry *dst;
	int addr_len = msg->msg_namelen;
	int ulen = len;
	int hlimit = -1;
1011
	int tclass = -1;
1012
	int dontfrag = -1;
L
Linus Torvalds 已提交
1013 1014
	int corkreq = up->corkflag || msg->msg_flags&MSG_MORE;
	int err;
1015
	int connected = 0;
W
Wang Chen 已提交
1016
	int is_udplite = IS_UDPLITE(sk);
1017
	int (*getfrag)(void *, char *, int, int, int, struct sk_buff *);
L
Linus Torvalds 已提交
1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043

	/* 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;
1044
	} else
L
Linus Torvalds 已提交
1045 1046 1047
		daddr = NULL;

	if (daddr) {
B
Brian Haley 已提交
1048
		if (ipv6_addr_v4mapped(daddr)) {
L
Linus Torvalds 已提交
1049 1050
			struct sockaddr_in sin;
			sin.sin_family = AF_INET;
E
Eric Dumazet 已提交
1051
			sin.sin_port = sin6 ? sin6->sin6_port : inet->inet_dport;
L
Linus Torvalds 已提交
1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065
			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,
1066
	   better check is made in ip6_append_data().
L
Linus Torvalds 已提交
1067 1068 1069
	   */
	if (len > INT_MAX - sizeof(struct udphdr))
		return -EMSGSIZE;
1070

L
Linus Torvalds 已提交
1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088
	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);

1089
	memset(&fl6, 0, sizeof(fl6));
L
Linus Torvalds 已提交
1090 1091 1092 1093 1094

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

1095
		fl6.fl6_dport = sin6->sin6_port;
L
Linus Torvalds 已提交
1096 1097 1098
		daddr = &sin6->sin6_addr;

		if (np->sndflow) {
1099 1100 1101
			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
1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118
				if (flowlabel == NULL)
					return -EINVAL;
				daddr = &flowlabel->dst;
			}
		}

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

		if (addr_len >= sizeof(struct sockaddr_in6) &&
		    sin6->sin6_scope_id &&
		    ipv6_addr_type(daddr)&IPV6_ADDR_LINKLOCAL)
1119
			fl6.flowi6_oif = sin6->sin6_scope_id;
L
Linus Torvalds 已提交
1120 1121 1122 1123
	} else {
		if (sk->sk_state != TCP_ESTABLISHED)
			return -EDESTADDRREQ;

1124
		fl6.fl6_dport = inet->inet_dport;
L
Linus Torvalds 已提交
1125
		daddr = &np->daddr;
1126
		fl6.flowlabel = np->flow_label;
1127
		connected = 1;
L
Linus Torvalds 已提交
1128 1129
	}

1130 1131
	if (!fl6.flowi6_oif)
		fl6.flowi6_oif = sk->sk_bound_dev_if;
L
Linus Torvalds 已提交
1132

1133 1134
	if (!fl6.flowi6_oif)
		fl6.flowi6_oif = np->sticky_pktinfo.ipi6_ifindex;
1135

1136
	fl6.flowi6_mark = sk->sk_mark;
B
Brian Haley 已提交
1137

L
Linus Torvalds 已提交
1138 1139 1140 1141 1142
	if (msg->msg_controllen) {
		opt = &opt_space;
		memset(opt, 0, sizeof(struct ipv6_txoptions));
		opt->tot_len = sizeof(*opt);

1143 1144
		err = datagram_send_ctl(sock_net(sk), sk, msg, &fl6, opt,
					&hlimit, &tclass, &dontfrag);
L
Linus Torvalds 已提交
1145 1146 1147 1148
		if (err < 0) {
			fl6_sock_release(flowlabel);
			return err;
		}
1149 1150
		if ((fl6.flowlabel&IPV6_FLOWLABEL_MASK) && !flowlabel) {
			flowlabel = fl6_sock_lookup(sk, fl6.flowlabel);
L
Linus Torvalds 已提交
1151 1152 1153 1154 1155
			if (flowlabel == NULL)
				return -EINVAL;
		}
		if (!(opt->opt_nflen|opt->opt_flen))
			opt = NULL;
1156
		connected = 0;
L
Linus Torvalds 已提交
1157 1158 1159
	}
	if (opt == NULL)
		opt = np->opt;
1160 1161 1162
	if (flowlabel)
		opt = fl6_merge_options(&opt_space, flowlabel, opt);
	opt = ipv6_fixup_options(&opt_space, opt);
L
Linus Torvalds 已提交
1163

1164
	fl6.flowi6_proto = sk->sk_protocol;
1165
	if (!ipv6_addr_any(daddr))
A
Alexey Dobriyan 已提交
1166
		fl6.daddr = *daddr;
1167
	else
1168 1169
		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
1170
		fl6.saddr = np->saddr;
1171
	fl6.fl6_sport = inet->inet_sport;
1172

1173
	final_p = fl6_update_dst(&fl6, opt, &final);
1174
	if (final_p)
1175
		connected = 0;
L
Linus Torvalds 已提交
1176

1177 1178
	if (!fl6.flowi6_oif && ipv6_addr_is_multicast(&fl6.daddr)) {
		fl6.flowi6_oif = np->mcast_oif;
1179
		connected = 0;
1180 1181
	} else if (!fl6.flowi6_oif)
		fl6.flowi6_oif = np->ucast_oif;
L
Linus Torvalds 已提交
1182

1183
	security_sk_classify_flow(sk, flowi6_to_flowi(&fl6));
V
Venkat Yekkirala 已提交
1184

1185
	dst = ip6_sk_dst_lookup_flow(sk, &fl6, final_p, true);
1186 1187 1188
	if (IS_ERR(dst)) {
		err = PTR_ERR(dst);
		dst = NULL;
L
Linus Torvalds 已提交
1189
		goto out;
1190
	}
L
Linus Torvalds 已提交
1191 1192

	if (hlimit < 0) {
1193
		if (ipv6_addr_is_multicast(&fl6.daddr))
L
Linus Torvalds 已提交
1194 1195 1196 1197
			hlimit = np->mcast_hops;
		else
			hlimit = np->hop_limit;
		if (hlimit < 0)
1198
			hlimit = ip6_dst_hoplimit(dst);
L
Linus Torvalds 已提交
1199 1200
	}

G
Gerrit Renker 已提交
1201
	if (tclass < 0)
1202 1203
		tclass = np->tclass;

1204 1205 1206
	if (dontfrag < 0)
		dontfrag = np->dontfrag;

L
Linus Torvalds 已提交
1207 1208 1209 1210 1211 1212 1213 1214 1215 1216
	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);

1217
		LIMIT_NETDEBUG(KERN_DEBUG "udp cork app bug 2\n");
L
Linus Torvalds 已提交
1218 1219 1220 1221 1222 1223 1224 1225
		err = -EINVAL;
		goto out;
	}

	up->pending = AF_INET6;

do_append_data:
	up->len += ulen;
1226 1227
	getfrag  =  is_udplite ?  udplite_getfrag : ip_generic_getfrag;
	err = ip6_append_data(sk, getfrag, msg->msg_iov, ulen,
1228
		sizeof(struct udphdr), hlimit, tclass, opt, &fl6,
1229
		(struct rt6_info*)dst,
1230
		corkreq ? msg->msg_flags|MSG_MORE : msg->msg_flags, dontfrag);
L
Linus Torvalds 已提交
1231 1232 1233
	if (err)
		udp_v6_flush_pending_frames(sk);
	else if (!corkreq)
1234
		err = udp_v6_push_pending_frames(sk);
H
Herbert Xu 已提交
1235 1236
	else if (unlikely(skb_queue_empty(&sk->sk_write_queue)))
		up->pending = 0;
L
Linus Torvalds 已提交
1237

1238 1239 1240
	if (dst) {
		if (connected) {
			ip6_dst_store(sk, dst,
1241
				      ipv6_addr_equal(&fl6.daddr, &np->daddr) ?
1242 1243
				      &np->daddr : NULL,
#ifdef CONFIG_IPV6_SUBTREES
1244
				      ipv6_addr_equal(&fl6.saddr, &np->saddr) ?
1245 1246 1247
				      &np->saddr :
#endif
				      NULL);
1248 1249 1250
		} else {
			dst_release(dst);
		}
1251
		dst = NULL;
1252 1253
	}

L
Linus Torvalds 已提交
1254 1255 1256 1257
	if (err > 0)
		err = np->recverr ? net_xmit_errno(err) : 0;
	release_sock(sk);
out:
1258
	dst_release(dst);
L
Linus Torvalds 已提交
1259
	fl6_sock_release(flowlabel);
1260
	if (!err)
L
Linus Torvalds 已提交
1261
		return len;
1262 1263 1264 1265 1266 1267 1268 1269
	/*
	 * 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)) {
1270 1271
		UDP6_INC_STATS_USER(sock_net(sk),
				UDP_MIB_SNDBUFERRORS, is_udplite);
1272
	}
L
Linus Torvalds 已提交
1273 1274 1275 1276 1277 1278 1279 1280 1281 1282
	return err;

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

1283
void udpv6_destroy_sock(struct sock *sk)
L
Linus Torvalds 已提交
1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294
{
	lock_sock(sk);
	udp_v6_flush_pending_frames(sk);
	release_sock(sk);

	inet6_destroy_sock(sk);
}

/*
 *	Socket option code for UDP
 */
1295
int udpv6_setsockopt(struct sock *sk, int level, int optname,
1296
		     char __user *optval, unsigned int optlen)
1297
{
1298
	if (level == SOL_UDP  ||  level == SOL_UDPLITE)
1299 1300
		return udp_lib_setsockopt(sk, level, optname, optval, optlen,
					  udp_v6_push_pending_frames);
1301
	return ipv6_setsockopt(sk, level, optname, optval, optlen);
1302 1303 1304
}

#ifdef CONFIG_COMPAT
1305
int compat_udpv6_setsockopt(struct sock *sk, int level, int optname,
1306
			    char __user *optval, unsigned int optlen)
1307
{
1308
	if (level == SOL_UDP  ||  level == SOL_UDPLITE)
1309 1310
		return udp_lib_setsockopt(sk, level, optname, optval, optlen,
					  udp_v6_push_pending_frames);
1311
	return compat_ipv6_setsockopt(sk, level, optname, optval, optlen);
1312 1313 1314
}
#endif

1315 1316
int udpv6_getsockopt(struct sock *sk, int level, int optname,
		     char __user *optval, int __user *optlen)
1317
{
1318
	if (level == SOL_UDP  ||  level == SOL_UDPLITE)
1319
		return udp_lib_getsockopt(sk, level, optname, optval, optlen);
1320
	return ipv6_getsockopt(sk, level, optname, optval, optlen);
1321 1322 1323
}

#ifdef CONFIG_COMPAT
1324 1325
int compat_udpv6_getsockopt(struct sock *sk, int level, int optname,
			    char __user *optval, int __user *optlen)
1326
{
1327
	if (level == SOL_UDP  ||  level == SOL_UDPLITE)
1328
		return udp_lib_getsockopt(sk, level, optname, optval, optlen);
1329
	return compat_ipv6_getsockopt(sk, level, optname, optval, optlen);
1330 1331 1332
}
#endif

1333 1334
static int udp6_ufo_send_check(struct sk_buff *skb)
{
1335
	const struct ipv6hdr *ipv6h;
1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351
	struct udphdr *uh;

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

	ipv6h = ipv6_hdr(skb);
	uh = udp_hdr(skb);

	uh->check = ~csum_ipv6_magic(&ipv6h->saddr, &ipv6h->daddr, skb->len,
				     IPPROTO_UDP, 0);
	skb->csum_start = skb_transport_header(skb) - skb->head;
	skb->csum_offset = offsetof(struct udphdr, check);
	skb->ip_summed = CHECKSUM_PARTIAL;
	return 0;
}

1352 1353
static struct sk_buff *udp6_ufo_fragment(struct sk_buff *skb,
	netdev_features_t features)
1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385
{
	struct sk_buff *segs = ERR_PTR(-EINVAL);
	unsigned int mss;
	unsigned int unfrag_ip6hlen, unfrag_len;
	struct frag_hdr *fptr;
	u8 *mac_start, *prevhdr;
	u8 nexthdr;
	u8 frag_hdr_sz = sizeof(struct frag_hdr);
	int offset;
	__wsum csum;

	mss = skb_shinfo(skb)->gso_size;
	if (unlikely(skb->len <= mss))
		goto out;

	if (skb_gso_ok(skb, features | NETIF_F_GSO_ROBUST)) {
		/* Packet is from an untrusted source, reset gso_segs. */
		int type = skb_shinfo(skb)->gso_type;

		if (unlikely(type & ~(SKB_GSO_UDP | SKB_GSO_DODGY) ||
			     !(type & (SKB_GSO_UDP))))
			goto out;

		skb_shinfo(skb)->gso_segs = DIV_ROUND_UP(skb->len, mss);

		segs = NULL;
		goto out;
	}

	/* Do software UFO. Complete and fill in the UDP checksum as HW cannot
	 * do checksum of UDP packets sent as multiple IP fragments.
	 */
1386
	offset = skb_checksum_start_offset(skb);
1387
	csum = skb_checksum(skb, offset, skb->len - offset, 0);
1388 1389 1390 1391 1392
	offset += skb->csum_offset;
	*(__sum16 *)(skb->data + offset) = csum_fold(csum);
	skb->ip_summed = CHECKSUM_NONE;

	/* Check if there is enough headroom to insert fragment header. */
1393
	if ((skb_mac_header(skb) < skb->head + frag_hdr_sz) &&
1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413
	    pskb_expand_head(skb, frag_hdr_sz, 0, GFP_ATOMIC))
		goto out;

	/* Find the unfragmentable header and shift it left by frag_hdr_sz
	 * bytes to insert fragment header.
	 */
	unfrag_ip6hlen = ip6_find_1stfragopt(skb, &prevhdr);
	nexthdr = *prevhdr;
	*prevhdr = NEXTHDR_FRAGMENT;
	unfrag_len = skb_network_header(skb) - skb_mac_header(skb) +
		     unfrag_ip6hlen;
	mac_start = skb_mac_header(skb);
	memmove(mac_start-frag_hdr_sz, mac_start, unfrag_len);

	skb->mac_header -= frag_hdr_sz;
	skb->network_header -= frag_hdr_sz;

	fptr = (struct frag_hdr *)(skb_network_header(skb) + unfrag_ip6hlen);
	fptr->nexthdr = nexthdr;
	fptr->reserved = 0;
1414
	ipv6_select_ident(fptr, (struct rt6_info *)skb_dst(skb));
1415 1416 1417 1418 1419 1420 1421 1422 1423 1424

	/* Fragment the skb. ipv6 header and the remaining fields of the
	 * fragment header are updated in ipv6_gso_segment()
	 */
	segs = skb_segment(skb, features);

out:
	return segs;
}

1425
static const struct inet6_protocol udpv6_protocol = {
L
Linus Torvalds 已提交
1426 1427
	.handler	=	udpv6_rcv,
	.err_handler	=	udpv6_err,
1428 1429
	.gso_send_check =	udp6_ufo_send_check,
	.gso_segment	=	udp6_ufo_fragment,
L
Linus Torvalds 已提交
1430 1431 1432 1433 1434 1435 1436 1437 1438 1439
	.flags		=	INET6_PROTO_NOPOLICY|INET6_PROTO_FINAL,
};

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

static void udp6_sock_seq_show(struct seq_file *seq, struct sock *sp, int bucket)
{
	struct inet_sock *inet = inet_sk(sp);
	struct ipv6_pinfo *np = inet6_sk(sp);
1440
	const struct in6_addr *dest, *src;
L
Linus Torvalds 已提交
1441 1442 1443 1444
	__u16 destp, srcp;

	dest  = &np->daddr;
	src   = &np->rcv_saddr;
E
Eric Dumazet 已提交
1445 1446
	destp = ntohs(inet->inet_dport);
	srcp  = ntohs(inet->inet_sport);
L
Linus Torvalds 已提交
1447
	seq_printf(seq,
1448
		   "%5d: %08X%08X%08X%08X:%04X %08X%08X%08X%08X:%04X "
D
Dan Rosenberg 已提交
1449
		   "%02X %08X:%08X %02X:%08lX %08X %5d %8d %lu %d %pK %d\n",
L
Linus Torvalds 已提交
1450 1451 1452 1453 1454
		   bucket,
		   src->s6_addr32[0], src->s6_addr32[1],
		   src->s6_addr32[2], src->s6_addr32[3], srcp,
		   dest->s6_addr32[0], dest->s6_addr32[1],
		   dest->s6_addr32[2], dest->s6_addr32[3], destp,
1455
		   sp->sk_state,
1456 1457
		   sk_wmem_alloc_get(sp),
		   sk_rmem_alloc_get(sp),
L
Linus Torvalds 已提交
1458 1459 1460
		   0, 0L, 0,
		   sock_i_uid(sp), 0,
		   sock_i_ino(sp),
E
Eric Dumazet 已提交
1461 1462
		   atomic_read(&sp->sk_refcnt), sp,
		   atomic_read(&sp->sk_drops));
L
Linus Torvalds 已提交
1463 1464
}

1465
int udp6_seq_show(struct seq_file *seq, void *v)
L
Linus Torvalds 已提交
1466 1467 1468 1469 1470 1471 1472
{
	if (v == SEQ_START_TOKEN)
		seq_printf(seq,
			   "  sl  "
			   "local_address                         "
			   "remote_address                        "
			   "st tx_queue rx_queue tr tm->when retrnsmt"
E
Eric Dumazet 已提交
1473
			   "   uid  timeout inode ref pointer drops\n");
L
Linus Torvalds 已提交
1474 1475 1476 1477 1478
	else
		udp6_sock_seq_show(seq, v, ((struct udp_iter_state *)seq->private)->bucket);
	return 0;
}

1479 1480 1481 1482 1483 1484 1485 1486
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
1487 1488 1489
static struct udp_seq_afinfo udp6_seq_afinfo = {
	.name		= "udp6",
	.family		= AF_INET6,
1490
	.udp_table	= &udp_table,
1491
	.seq_fops	= &udp6_afinfo_seq_fops,
1492 1493 1494
	.seq_ops	= {
		.show		= udp6_seq_show,
	},
L
Linus Torvalds 已提交
1495 1496
};

1497
int __net_init udp6_proc_init(struct net *net)
L
Linus Torvalds 已提交
1498
{
1499
	return udp_proc_register(net, &udp6_seq_afinfo);
L
Linus Torvalds 已提交
1500 1501
}

1502 1503
void udp6_proc_exit(struct net *net) {
	udp_proc_unregister(net, &udp6_seq_afinfo);
L
Linus Torvalds 已提交
1504 1505 1506 1507 1508 1509
}
#endif /* CONFIG_PROC_FS */

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

struct proto udpv6_prot = {
1510 1511
	.name		   = "UDPv6",
	.owner		   = THIS_MODULE,
1512
	.close		   = udp_lib_close,
1513 1514 1515 1516 1517 1518 1519 1520
	.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,
1521
	.backlog_rcv	   = __udpv6_queue_rcv_skb,
1522 1523
	.hash		   = udp_lib_hash,
	.unhash		   = udp_lib_unhash,
E
Eric Dumazet 已提交
1524
	.rehash		   = udp_v6_rehash,
1525
	.get_port	   = udp_v6_get_port,
H
Hideo Aoki 已提交
1526 1527 1528 1529
	.memory_allocated  = &udp_memory_allocated,
	.sysctl_mem	   = sysctl_udp_mem,
	.sysctl_wmem	   = &sysctl_udp_wmem_min,
	.sysctl_rmem	   = &sysctl_udp_rmem_min,
1530
	.obj_size	   = sizeof(struct udp6_sock),
1531
	.slab_flags	   = SLAB_DESTROY_BY_RCU,
1532
	.h.udp_table	   = &udp_table,
1533
#ifdef CONFIG_COMPAT
1534 1535
	.compat_setsockopt = compat_udpv6_setsockopt,
	.compat_getsockopt = compat_udpv6_getsockopt,
1536
#endif
1537
	.clear_sk	   = sk_prot_clear_portaddr_nulls,
L
Linus Torvalds 已提交
1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549
};

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,
};


1550
int __init udpv6_init(void)
L
Linus Torvalds 已提交
1551
{
1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568
	int ret;

	ret = inet6_add_protocol(&udpv6_protocol, IPPROTO_UDP);
	if (ret)
		goto out;

	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;
}

1569
void udpv6_exit(void)
1570 1571 1572
{
	inet6_unregister_protosw(&udpv6_protosw);
	inet6_del_protocol(&udpv6_protocol, IPPROTO_UDP);
L
Linus Torvalds 已提交
1573
}