udp.c 36.9 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 "udp_impl.h"
L
Linus Torvalds 已提交
52

53 54 55 56
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);
E
Eric Dumazet 已提交
57
	__be32 sk1_rcv_saddr = inet_sk(sk)->inet_rcv_saddr;
58
	__be32 sk2_rcv_saddr = inet_rcv_saddr(sk2);
59 60 61 62 63 64 65
	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)
66
		return (!sk2_ipv6only &&
E
Eric Dumazet 已提交
67 68
			(!sk1_rcv_saddr || !sk2_rcv_saddr ||
			  sk1_rcv_saddr == sk2_rcv_saddr));
69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84

	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;
}

85 86 87 88 89 90 91 92
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
93
	else if (ipv6_addr_v4mapped(addr6))
94
		hash = jhash_1word((__force u32)addr6->s6_addr32[3], mix);
95
	else
96
		hash = jhash2((__force u32 *)addr6->s6_addr32, 4, mix);
97 98 99 100 101

	return hash ^ port;
}


102
int udp_v6_get_port(struct sock *sk, unsigned short snum)
L
Linus Torvalds 已提交
103
{
E
Eric Dumazet 已提交
104 105 106 107 108
	unsigned int hash2_nulladdr =
		udp6_portaddr_hash(sock_net(sk), &in6addr_any, snum);
	unsigned int hash2_partial = 
		udp6_portaddr_hash(sock_net(sk), &inet6_sk(sk)->rcv_saddr, 0);

109
	/* precompute partial secondary hash */
E
Eric Dumazet 已提交
110 111
	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
112 113
}

E
Eric Dumazet 已提交
114 115 116 117 118 119 120 121 122
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);
}

123 124 125 126 127 128 129 130
static inline int compute_score(struct sock *sk, struct net *net,
				unsigned short hnum,
				struct in6_addr *saddr, __be16 sport,
				struct in6_addr *daddr, __be16 dport,
				int dif)
{
	int score = -1;

131
	if (net_eq(sock_net(sk), net) && udp_sk(sk)->udp_port_hash == hnum &&
132 133 134 135 136
			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
137 138
		if (inet->inet_dport) {
			if (inet->inet_dport != sport)
139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160
				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
161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240
#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:
		if (unlikely(!atomic_inc_not_zero(&result->sk_refcnt)))
			result = NULL;
		else if (unlikely(compute_score2(result, net, saddr, sport,
				  daddr, hnum, dif) < badness)) {
			sock_put(result);
			goto begin;
		}
	}
	return result;
}

241 242
static struct sock *__udp6_lib_lookup(struct net *net,
				      struct in6_addr *saddr, __be16 sport,
243
				      struct in6_addr *daddr, __be16 dport,
244
				      int dif, struct udp_table *udptable)
L
Linus Torvalds 已提交
245
{
246
	struct sock *sk, *result;
247
	struct hlist_nulls_node *node;
L
Linus Torvalds 已提交
248
	unsigned short hnum = ntohs(dport);
E
Eric Dumazet 已提交
249 250
	unsigned int hash2, slot2, slot = udp_hashfn(net, hnum, udptable->mask);
	struct udp_hslot *hslot2, *hslot = &udptable->hash[slot];
251
	int score, badness;
252

253
	rcu_read_lock();
E
Eric Dumazet 已提交
254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270
	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;

271 272
			result = udp6_lib_lookup2(net, saddr, sport,
						  &in6addr_any, hnum, dif,
E
Eric Dumazet 已提交
273 274 275 276 277
						  hslot2, slot2);
		}
		rcu_read_unlock();
		return result;
	}
278 279 280
begin:
	result = NULL;
	badness = -1;
281
	sk_nulls_for_each_rcu(sk, node, &hslot->head) {
282 283 284 285
		score = compute_score(sk, net, hnum, saddr, sport, daddr, dport, dif);
		if (score > badness) {
			result = sk;
			badness = score;
L
Linus Torvalds 已提交
286 287
		}
	}
288 289 290 291 292
	/*
	 * 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
293
	if (get_nulls_value(node) != slot)
294 295
		goto begin;

296 297 298 299 300 301 302 303 304 305
	if (result) {
		if (unlikely(!atomic_inc_not_zero(&result->sk_refcnt)))
			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
306 307 308
	return result;
}

309 310
static struct sock *__udp6_lib_lookup_skb(struct sk_buff *skb,
					  __be16 sport, __be16 dport,
311
					  struct udp_table *udptable)
312
{
313
	struct sock *sk;
314 315
	struct ipv6hdr *iph = ipv6_hdr(skb);

316 317
	if (unlikely(sk = skb_steal_sock(skb)))
		return sk;
E
Eric Dumazet 已提交
318 319 320
	return __udp6_lib_lookup(dev_net(skb_dst(skb)->dev), &iph->saddr, sport,
				 &iph->daddr, dport, inet6_iif(skb),
				 udptable);
321 322
}

L
Linus Torvalds 已提交
323 324 325 326 327
/*
 * 	This should be easy, if there is something there we
 * 	return it, otherwise we block.
 */

328
int udpv6_recvmsg(struct kiocb *iocb, struct sock *sk,
L
Linus Torvalds 已提交
329 330 331 332 333
		  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);
334
	struct sk_buff *skb;
G
Gerrit Renker 已提交
335
	unsigned int ulen;
336
	int peeked;
337 338
	int err;
	int is_udplite = IS_UDPLITE(sk);
339
	int is_udp4;
340
	bool slow;
L
Linus Torvalds 已提交
341

342 343 344
	if (addr_len)
		*addr_len=sizeof(struct sockaddr_in6);

L
Linus Torvalds 已提交
345 346 347
	if (flags & MSG_ERRQUEUE)
		return ipv6_recv_error(sk, msg, len);

348 349 350
	if (np->rxpmtu && np->rxopt.bits.rxpmtu)
		return ipv6_recv_rxpmtu(sk, msg, len);

L
Linus Torvalds 已提交
351
try_again:
352 353
	skb = __skb_recv_datagram(sk, flags | (noblock ? MSG_DONTWAIT : 0),
				  &peeked, &err);
L
Linus Torvalds 已提交
354 355 356
	if (!skb)
		goto out;

357
	ulen = skb->len - sizeof(struct udphdr);
G
Gerrit Renker 已提交
358 359 360
	if (len > ulen)
		len = ulen;
	else if (len < ulen)
361
		msg->msg_flags |= MSG_TRUNC;
L
Linus Torvalds 已提交
362

363 364
	is_udp4 = (skb->protocol == htons(ETH_P_IP));

365
	/*
366 367 368
	 * 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.
369 370
	 */

G
Gerrit Renker 已提交
371
	if (len < ulen || UDP_SKB_CB(skb)->partial_cov) {
372
		if (udp_lib_checksum_complete(skb))
L
Linus Torvalds 已提交
373
			goto csum_copy_err;
374 375
	}

376
	if (skb_csum_unnecessary(skb))
377
		err = skb_copy_datagram_iovec(skb, sizeof(struct udphdr),
G
Gerrit Renker 已提交
378
					      msg->msg_iov,len);
379
	else {
L
Linus Torvalds 已提交
380 381 382 383 384 385 386
		err = skb_copy_and_csum_datagram_iovec(skb, sizeof(struct udphdr), msg->msg_iov);
		if (err == -EINVAL)
			goto csum_copy_err;
	}
	if (err)
		goto out_free;

387 388 389 390 391 392 393 394
	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);
	}
395

396
	sock_recv_ts_and_drops(msg, sk, skb);
L
Linus Torvalds 已提交
397 398 399 400

	/* Copy the address. */
	if (msg->msg_name) {
		struct sockaddr_in6 *sin6;
401

L
Linus Torvalds 已提交
402 403
		sin6 = (struct sockaddr_in6 *) msg->msg_name;
		sin6->sin6_family = AF_INET6;
404
		sin6->sin6_port = udp_hdr(skb)->source;
L
Linus Torvalds 已提交
405 406 407
		sin6->sin6_flowinfo = 0;
		sin6->sin6_scope_id = 0;

408
		if (is_udp4)
B
Brian Haley 已提交
409 410
			ipv6_addr_set_v4mapped(ip_hdr(skb)->saddr,
					       &sin6->sin6_addr);
L
Linus Torvalds 已提交
411
		else {
412 413
			ipv6_addr_copy(&sin6->sin6_addr,
				       &ipv6_hdr(skb)->saddr);
L
Linus Torvalds 已提交
414 415 416 417 418
			if (ipv6_addr_type(&sin6->sin6_addr) & IPV6_ADDR_LINKLOCAL)
				sin6->sin6_scope_id = IP6CB(skb)->iif;
		}

	}
419
	if (is_udp4) {
L
Linus Torvalds 已提交
420 421 422 423 424
		if (inet->cmsg_flags)
			ip_cmsg_recv(msg, skb);
	} else {
		if (np->rxopt.all)
			datagram_recv_ctl(sk, msg, skb);
425
	}
L
Linus Torvalds 已提交
426

G
Gerrit Renker 已提交
427
	err = len;
L
Linus Torvalds 已提交
428
	if (flags & MSG_TRUNC)
429
		err = ulen;
L
Linus Torvalds 已提交
430 431

out_free:
432
	skb_free_datagram_locked(sk, skb);
L
Linus Torvalds 已提交
433 434 435 436
out:
	return err;

csum_copy_err:
437
	slow = lock_sock_fast(sk);
438 439 440 441 442 443 444 445
	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);
	}
446
	unlock_sock_fast(sk, slow);
L
Linus Torvalds 已提交
447

448
	if (flags & MSG_DONTWAIT)
L
Linus Torvalds 已提交
449 450 451 452
		return -EAGAIN;
	goto try_again;
}

453
void __udp6_lib_err(struct sk_buff *skb, struct inet6_skb_parm *opt,
454
		    u8 type, u8 code, int offset, __be32 info,
455
		    struct udp_table *udptable)
L
Linus Torvalds 已提交
456 457 458 459 460 461 462 463 464
{
	struct ipv6_pinfo *np;
	struct ipv6hdr *hdr = (struct ipv6hdr*)skb->data;
	struct in6_addr *saddr = &hdr->saddr;
	struct in6_addr *daddr = &hdr->daddr;
	struct udphdr *uh = (struct udphdr*)(skb->data+offset);
	struct sock *sk;
	int err;

465
	sk = __udp6_lib_lookup(dev_net(skb->dev), daddr, uh->dest,
466
			       saddr, uh->source, inet6_iif(skb), udptable);
L
Linus Torvalds 已提交
467 468 469 470 471 472 473 474 475 476 477
	if (sk == NULL)
		return;

	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;

478
	if (np->recverr)
L
Linus Torvalds 已提交
479
		ipv6_icmp_error(sk, skb, err, uh->dest, ntohl(info), (u8 *)(uh+1));
480

L
Linus Torvalds 已提交
481 482 483 484 485 486
	sk->sk_err = err;
	sk->sk_error_report(sk);
out:
	sock_put(sk);
}

487
static __inline__ void udpv6_err(struct sk_buff *skb,
488 489
				 struct inet6_skb_parm *opt, u8 type,
				 u8 code, int offset, __be32 info     )
490
{
491
	__udp6_lib_err(skb, opt, type, code, offset, info, &udp_table);
492 493 494
}

int udpv6_queue_rcv_skb(struct sock * sk, struct sk_buff *skb)
L
Linus Torvalds 已提交
495
{
496
	struct udp_sock *up = udp_sk(sk);
497
	int rc;
W
Wang Chen 已提交
498
	int is_udplite = IS_UDPLITE(sk);
499

500 501
	if (!xfrm6_policy_check(sk, XFRM_POLICY_IN, skb))
		goto drop;
L
Linus Torvalds 已提交
502

503 504 505
	/*
	 * UDP-Lite specific tests, ignored on UDP sockets (see net/ipv4/udp.c).
	 */
W
Wang Chen 已提交
506
	if ((is_udplite & UDPLITE_RECV_CC)  &&  UDP_SKB_CB(skb)->partial_cov) {
507 508 509 510 511 512 513 514 515 516 517 518 519

		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
520 521
	}

522 523 524 525
	if (sk->sk_filter) {
		if (udp_lib_checksum_complete(skb))
			goto drop;
	}
526

E
Eric Dumazet 已提交
527
	if ((rc = ip_queue_rcv_skb(sk, skb)) < 0) {
528
		/* Note that an ENOMEM error is charged twice */
E
Eric Dumazet 已提交
529
		if (rc == -ENOMEM)
530 531
			UDP6_INC_STATS_BH(sock_net(sk),
					UDP_MIB_RCVBUFERRORS, is_udplite);
E
Eric Dumazet 已提交
532
		goto drop_no_sk_drops_inc;
L
Linus Torvalds 已提交
533
	}
534

L
Linus Torvalds 已提交
535
	return 0;
536
drop:
E
Eric Dumazet 已提交
537 538
	atomic_inc(&sk->sk_drops);
drop_no_sk_drops_inc:
539
	UDP6_INC_STATS_BH(sock_net(sk), UDP_MIB_INERRORS, is_udplite);
540 541
	kfree_skb(skb);
	return -1;
L
Linus Torvalds 已提交
542 543
}

544
static struct sock *udp_v6_mcast_next(struct net *net, struct sock *sk,
A
Al Viro 已提交
545 546
				      __be16 loc_port, struct in6_addr *loc_addr,
				      __be16 rmt_port, struct in6_addr *rmt_addr,
L
Linus Torvalds 已提交
547 548
				      int dif)
{
549
	struct hlist_nulls_node *node;
L
Linus Torvalds 已提交
550 551 552
	struct sock *s = sk;
	unsigned short num = ntohs(loc_port);

553
	sk_nulls_for_each_from(s, node) {
L
Linus Torvalds 已提交
554 555
		struct inet_sock *inet = inet_sk(s);

556
		if (!net_eq(sock_net(s), net))
557 558
			continue;

559 560
		if (udp_sk(s)->udp_port_hash == num &&
		    s->sk_family == PF_INET6) {
L
Linus Torvalds 已提交
561
			struct ipv6_pinfo *np = inet6_sk(s);
E
Eric Dumazet 已提交
562 563
			if (inet->inet_dport) {
				if (inet->inet_dport != rmt_port)
L
Linus Torvalds 已提交
564 565 566 567 568 569 570 571 572 573
					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)) {
574 575
				if (!ipv6_addr_equal(&np->rcv_saddr, loc_addr))
					continue;
L
Linus Torvalds 已提交
576
			}
S
Stephen Hemminger 已提交
577
			if (!inet6_mc_check(s, loc_addr, rmt_addr))
L
Linus Torvalds 已提交
578 579 580 581 582 583 584
				continue;
			return s;
		}
	}
	return NULL;
}

585 586 587 588 589 590 591 592 593 594
static void flush_stack(struct sock **stack, unsigned int count,
			struct sk_buff *skb, unsigned int final)
{
	unsigned int i;
	struct sock *sk;
	struct sk_buff *skb1;

	for (i = 0; i < count; i++) {
		skb1 = (i == final) ? skb : skb_clone(skb, GFP_ATOMIC);

595
		sk = stack[i];
596
		if (skb1) {
597 598 599 600
			if (sk_rcvqueues_full(sk, skb)) {
				kfree_skb(skb1);
				goto drop;
			}
601 602 603
			bh_lock_sock(sk);
			if (!sock_owned_by_user(sk))
				udpv6_queue_rcv_skb(sk, skb1);
Z
Zhu Yi 已提交
604
			else if (sk_add_backlog(sk, skb1)) {
Z
Zhu Yi 已提交
605 606 607 608
				kfree_skb(skb1);
				bh_unlock_sock(sk);
				goto drop;
			}
609
			bh_unlock_sock(sk);
Z
Zhu Yi 已提交
610
			continue;
611
		}
Z
Zhu Yi 已提交
612 613 614 615 616 617
drop:
		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));
618 619
	}
}
L
Linus Torvalds 已提交
620 621 622 623
/*
 * Note: called only from the BH handler context,
 * so we don't need to lock the hashes.
 */
624 625
static int __udp6_lib_mcast_deliver(struct net *net, struct sk_buff *skb,
		struct in6_addr *saddr, struct in6_addr *daddr,
626
		struct udp_table *udptable)
L
Linus Torvalds 已提交
627
{
628
	struct sock *sk, *stack[256 / sizeof(struct sock *)];
629
	const struct udphdr *uh = udp_hdr(skb);
630
	struct udp_hslot *hslot = udp_hashslot(udptable, net, ntohs(uh->dest));
L
Linus Torvalds 已提交
631
	int dif;
632
	unsigned int i, count = 0;
L
Linus Torvalds 已提交
633

634
	spin_lock(&hslot->lock);
635
	sk = sk_nulls_head(&hslot->head);
636
	dif = inet6_iif(skb);
637
	sk = udp_v6_mcast_next(net, sk, uh->dest, daddr, uh->source, saddr, dif);
638 639 640 641 642 643 644 645 646
	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
647
		}
L
Linus Torvalds 已提交
648
	}
649 650 651 652 653 654
	/*
	 * before releasing the lock, we must take reference on sockets
	 */
	for (i = 0; i < count; i++)
		sock_hold(stack[i]);

655
	spin_unlock(&hslot->lock);
656 657 658 659 660 661 662 663 664

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

		for (i = 0; i < count; i++)
			sock_put(stack[i]);
	} else {
		kfree_skb(skb);
	}
665
	return 0;
L
Linus Torvalds 已提交
666 667
}

668 669
static inline int udp6_csum_init(struct sk_buff *skb, struct udphdr *uh,
				 int proto)
670
{
671 672 673 674 675
	int err;

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

676
	if (proto == IPPROTO_UDPLITE) {
677 678 679 680 681
		err = udplite_checksum_init(skb, uh);
		if (err)
			return err;
	}

682 683 684 685 686 687 688 689
	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 &&
690
	    !csum_ipv6_magic(&ipv6_hdr(skb)->saddr, &ipv6_hdr(skb)->daddr,
691
			     skb->len, proto, skb->csum))
692 693
		skb->ip_summed = CHECKSUM_UNNECESSARY;

694
	if (!skb_csum_unnecessary(skb))
695 696
		skb->csum = ~csum_unfold(csum_ipv6_magic(&ipv6_hdr(skb)->saddr,
							 &ipv6_hdr(skb)->daddr,
697
							 skb->len, proto, 0));
698

699
	return 0;
700 701
}

702
int __udp6_lib_rcv(struct sk_buff *skb, struct udp_table *udptable,
703
		   int proto)
L
Linus Torvalds 已提交
704
{
705
	struct net *net = dev_net(skb->dev);
L
Linus Torvalds 已提交
706
	struct sock *sk;
707
	struct udphdr *uh;
L
Linus Torvalds 已提交
708 709 710 711
	struct in6_addr *saddr, *daddr;
	u32 ulen = 0;

	if (!pskb_may_pull(skb, sizeof(struct udphdr)))
712
		goto discard;
L
Linus Torvalds 已提交
713

714 715
	saddr = &ipv6_hdr(skb)->saddr;
	daddr = &ipv6_hdr(skb)->daddr;
716
	uh = udp_hdr(skb);
L
Linus Torvalds 已提交
717 718

	ulen = ntohs(uh->len);
719 720
	if (ulen > skb->len)
		goto short_packet;
L
Linus Torvalds 已提交
721

722 723
	if (proto == IPPROTO_UDP) {
		/* UDP validates ulen. */
L
Linus Torvalds 已提交
724

725 726 727
		/* Check for jumbo payload */
		if (ulen == 0)
			ulen = skb->len;
L
Linus Torvalds 已提交
728

729 730
		if (ulen < sizeof(*uh))
			goto short_packet;
L
Linus Torvalds 已提交
731

732 733 734
		if (ulen < skb->len) {
			if (pskb_trim_rcsum(skb, ulen))
				goto short_packet;
735 736
			saddr = &ipv6_hdr(skb)->saddr;
			daddr = &ipv6_hdr(skb)->daddr;
737
			uh = udp_hdr(skb);
738 739
		}
	}
L
Linus Torvalds 已提交
740

741 742 743
	if (udp6_csum_init(skb, uh, proto))
		goto discard;

744 745
	/*
	 *	Multicast receive code
L
Linus Torvalds 已提交
746
	 */
747
	if (ipv6_addr_is_multicast(daddr))
748 749
		return __udp6_lib_mcast_deliver(net, skb,
				saddr, daddr, udptable);
L
Linus Torvalds 已提交
750 751 752

	/* Unicast */

753
	/*
L
Linus Torvalds 已提交
754 755 756
	 * check socket cache ... must talk to Alan about his plans
	 * for sock caches... i'll skip this for now.
	 */
757
	sk = __udp6_lib_lookup_skb(skb, uh->source, uh->dest, udptable);
L
Linus Torvalds 已提交
758 759 760 761 762

	if (sk == NULL) {
		if (!xfrm6_policy_check(NULL, XFRM_POLICY_IN, skb))
			goto discard;

763
		if (udp_lib_checksum_complete(skb))
L
Linus Torvalds 已提交
764
			goto discard;
765 766
		UDP6_INC_STATS_BH(net, UDP_MIB_NOPORTS,
				proto == IPPROTO_UDPLITE);
L
Linus Torvalds 已提交
767

768
		icmpv6_send(skb, ICMPV6_DEST_UNREACH, ICMPV6_PORT_UNREACH, 0);
L
Linus Torvalds 已提交
769 770

		kfree_skb(skb);
S
Stephen Hemminger 已提交
771
		return 0;
L
Linus Torvalds 已提交
772
	}
773

L
Linus Torvalds 已提交
774
	/* deliver */
775

776 777 778 779
	if (sk_rcvqueues_full(sk, skb)) {
		sock_put(sk);
		goto discard;
	}
780
	bh_lock_sock(sk);
H
Hideo Aoki 已提交
781 782
	if (!sock_owned_by_user(sk))
		udpv6_queue_rcv_skb(sk, skb);
Z
Zhu Yi 已提交
783
	else if (sk_add_backlog(sk, skb)) {
Z
Zhu Yi 已提交
784 785 786 787 788
		atomic_inc(&sk->sk_drops);
		bh_unlock_sock(sk);
		sock_put(sk);
		goto discard;
	}
H
Hideo Aoki 已提交
789
	bh_unlock_sock(sk);
L
Linus Torvalds 已提交
790
	sock_put(sk);
S
Stephen Hemminger 已提交
791
	return 0;
L
Linus Torvalds 已提交
792

793
short_packet:
794
	LIMIT_NETDEBUG(KERN_DEBUG "UDP%sv6: short packet: From [%pI6c]:%u %d/%d to [%pI6c]:%u\n",
795
		       proto == IPPROTO_UDPLITE ? "-Lite" : "",
796 797 798 799 800 801
		       saddr,
		       ntohs(uh->source),
		       ulen,
		       skb->len,
		       daddr,
		       ntohs(uh->dest));
L
Linus Torvalds 已提交
802 803

discard:
804
	UDP6_INC_STATS_BH(net, UDP_MIB_INERRORS, proto == IPPROTO_UDPLITE);
L
Linus Torvalds 已提交
805
	kfree_skb(skb);
S
Stephen Hemminger 已提交
806
	return 0;
L
Linus Torvalds 已提交
807
}
808

809
static __inline__ int udpv6_rcv(struct sk_buff *skb)
810
{
811
	return __udp6_lib_rcv(skb, &udp_table, IPPROTO_UDP);
812 813
}

L
Linus Torvalds 已提交
814 815 816 817 818 819 820
/*
 * 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);

821 822 823
	if (up->pending == AF_INET)
		udp_flush_pending_frames(sk);
	else if (up->pending) {
L
Linus Torvalds 已提交
824 825 826
		up->len = 0;
		up->pending = 0;
		ip6_flush_pending_frames(sk);
827
	}
L
Linus Torvalds 已提交
828 829
}

830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870
/**
 * 	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
871 872 873 874
/*
 *	Sending
 */

875
static int udp_v6_push_pending_frames(struct sock *sk)
L
Linus Torvalds 已提交
876 877 878
{
	struct sk_buff *skb;
	struct udphdr *uh;
879
	struct udp_sock  *up = udp_sk(sk);
L
Linus Torvalds 已提交
880 881 882
	struct inet_sock *inet = inet_sk(sk);
	struct flowi *fl = &inet->cork.fl;
	int err = 0;
W
Wang Chen 已提交
883
	int is_udplite = IS_UDPLITE(sk);
884
	__wsum csum = 0;
L
Linus Torvalds 已提交
885 886 887 888 889 890 891 892

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

	/*
	 * Create a UDP header
	 */
893
	uh = udp_hdr(skb);
L
Linus Torvalds 已提交
894 895 896 897 898
	uh->source = fl->fl_ip_sport;
	uh->dest = fl->fl_ip_dport;
	uh->len = htons(up->len);
	uh->check = 0;

W
Wang Chen 已提交
899
	if (is_udplite)
900
		csum = udplite_csum_outgoing(sk, skb);
901 902 903 904 905
	else if (skb->ip_summed == CHECKSUM_PARTIAL) { /* UDP hardware csum */
		udp6_hwcsum_outgoing(sk, skb, &fl->fl6_src, &fl->fl6_dst,
				     up->len);
		goto send;
	} else
906
		csum = udp_csum_outgoing(sk, skb);
L
Linus Torvalds 已提交
907

908 909 910
	/* add protocol-dependent pseudo-header */
	uh->check = csum_ipv6_magic(&fl->fl6_src, &fl->fl6_dst,
				    up->len, fl->proto, csum   );
L
Linus Torvalds 已提交
911
	if (uh->check == 0)
912
		uh->check = CSUM_MANGLED_0;
L
Linus Torvalds 已提交
913

914
send:
L
Linus Torvalds 已提交
915
	err = ip6_push_pending_frames(sk);
E
Eric Dumazet 已提交
916 917 918 919 920 921 922 923 924
	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
925 926 927 928 929 930
out:
	up->len = 0;
	up->pending = 0;
	return err;
}

931
int udpv6_sendmsg(struct kiocb *iocb, struct sock *sk,
L
Linus Torvalds 已提交
932 933 934 935 936 937 938
		  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;
939
	struct in6_addr *daddr, *final_p, final;
L
Linus Torvalds 已提交
940 941
	struct ipv6_txoptions *opt = NULL;
	struct ip6_flowlabel *flowlabel = NULL;
H
Herbert Xu 已提交
942
	struct flowi fl;
L
Linus Torvalds 已提交
943 944 945 946
	struct dst_entry *dst;
	int addr_len = msg->msg_namelen;
	int ulen = len;
	int hlimit = -1;
947
	int tclass = -1;
948
	int dontfrag = -1;
L
Linus Torvalds 已提交
949 950
	int corkreq = up->corkflag || msg->msg_flags&MSG_MORE;
	int err;
951
	int connected = 0;
W
Wang Chen 已提交
952
	int is_udplite = IS_UDPLITE(sk);
953
	int (*getfrag)(void *, char *, int, int, int, struct sk_buff *);
L
Linus Torvalds 已提交
954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979

	/* 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;
980
	} else
L
Linus Torvalds 已提交
981 982 983
		daddr = NULL;

	if (daddr) {
B
Brian Haley 已提交
984
		if (ipv6_addr_v4mapped(daddr)) {
L
Linus Torvalds 已提交
985 986
			struct sockaddr_in sin;
			sin.sin_family = AF_INET;
E
Eric Dumazet 已提交
987
			sin.sin_port = sin6 ? sin6->sin6_port : inet->inet_dport;
L
Linus Torvalds 已提交
988 989 990 991 992 993 994 995 996 997 998 999 1000 1001
			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,
1002
	   better check is made in ip6_append_data().
L
Linus Torvalds 已提交
1003 1004 1005
	   */
	if (len > INT_MAX - sizeof(struct udphdr))
		return -EMSGSIZE;
1006

L
Linus Torvalds 已提交
1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024
	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);

H
Herbert Xu 已提交
1025
	memset(&fl, 0, sizeof(fl));
L
Linus Torvalds 已提交
1026 1027 1028 1029 1030

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

H
Herbert Xu 已提交
1031
		fl.fl_ip_dport = sin6->sin6_port;
L
Linus Torvalds 已提交
1032 1033 1034
		daddr = &sin6->sin6_addr;

		if (np->sndflow) {
H
Herbert Xu 已提交
1035 1036 1037
			fl.fl6_flowlabel = sin6->sin6_flowinfo&IPV6_FLOWINFO_MASK;
			if (fl.fl6_flowlabel&IPV6_FLOWLABEL_MASK) {
				flowlabel = fl6_sock_lookup(sk, fl.fl6_flowlabel);
L
Linus Torvalds 已提交
1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054
				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)
H
Herbert Xu 已提交
1055
			fl.oif = sin6->sin6_scope_id;
L
Linus Torvalds 已提交
1056 1057 1058 1059
	} else {
		if (sk->sk_state != TCP_ESTABLISHED)
			return -EDESTADDRREQ;

E
Eric Dumazet 已提交
1060
		fl.fl_ip_dport = inet->inet_dport;
L
Linus Torvalds 已提交
1061
		daddr = &np->daddr;
H
Herbert Xu 已提交
1062
		fl.fl6_flowlabel = np->flow_label;
1063
		connected = 1;
L
Linus Torvalds 已提交
1064 1065
	}

H
Herbert Xu 已提交
1066 1067
	if (!fl.oif)
		fl.oif = sk->sk_bound_dev_if;
L
Linus Torvalds 已提交
1068

1069 1070 1071
	if (!fl.oif)
		fl.oif = np->sticky_pktinfo.ipi6_ifindex;

B
Brian Haley 已提交
1072 1073
	fl.mark = sk->sk_mark;

L
Linus Torvalds 已提交
1074 1075 1076 1077 1078
	if (msg->msg_controllen) {
		opt = &opt_space;
		memset(opt, 0, sizeof(struct ipv6_txoptions));
		opt->tot_len = sizeof(*opt);

1079 1080
		err = datagram_send_ctl(sock_net(sk), msg, &fl, opt, &hlimit,
					&tclass, &dontfrag);
L
Linus Torvalds 已提交
1081 1082 1083 1084
		if (err < 0) {
			fl6_sock_release(flowlabel);
			return err;
		}
H
Herbert Xu 已提交
1085 1086
		if ((fl.fl6_flowlabel&IPV6_FLOWLABEL_MASK) && !flowlabel) {
			flowlabel = fl6_sock_lookup(sk, fl.fl6_flowlabel);
L
Linus Torvalds 已提交
1087 1088 1089 1090 1091
			if (flowlabel == NULL)
				return -EINVAL;
		}
		if (!(opt->opt_nflen|opt->opt_flen))
			opt = NULL;
1092
		connected = 0;
L
Linus Torvalds 已提交
1093 1094 1095
	}
	if (opt == NULL)
		opt = np->opt;
1096 1097 1098
	if (flowlabel)
		opt = fl6_merge_options(&opt_space, flowlabel, opt);
	opt = ipv6_fixup_options(&opt_space, opt);
L
Linus Torvalds 已提交
1099

1100
	fl.proto = sk->sk_protocol;
1101 1102 1103 1104
	if (!ipv6_addr_any(daddr))
		ipv6_addr_copy(&fl.fl6_dst, daddr);
	else
		fl.fl6_dst.s6_addr[15] = 0x1; /* :: means loopback (BSD'ism) */
H
Herbert Xu 已提交
1105 1106
	if (ipv6_addr_any(&fl.fl6_src) && !ipv6_addr_any(&np->saddr))
		ipv6_addr_copy(&fl.fl6_src, &np->saddr);
E
Eric Dumazet 已提交
1107
	fl.fl_ip_sport = inet->inet_sport;
1108

1109 1110
	final_p = fl6_update_dst(&fl, opt, &final);
	if (final_p)
1111
		connected = 0;
L
Linus Torvalds 已提交
1112

H
Herbert Xu 已提交
1113 1114
	if (!fl.oif && ipv6_addr_is_multicast(&fl.fl6_dst)) {
		fl.oif = np->mcast_oif;
1115 1116
		connected = 0;
	}
L
Linus Torvalds 已提交
1117

H
Herbert Xu 已提交
1118
	security_sk_classify_flow(sk, &fl);
V
Venkat Yekkirala 已提交
1119

H
Herbert Xu 已提交
1120
	err = ip6_sk_dst_lookup(sk, &dst, &fl);
L
Linus Torvalds 已提交
1121 1122 1123
	if (err)
		goto out;
	if (final_p)
H
Herbert Xu 已提交
1124
		ipv6_addr_copy(&fl.fl6_dst, final_p);
L
Linus Torvalds 已提交
1125

A
Alexey Dobriyan 已提交
1126 1127
	err = __xfrm_lookup(sock_net(sk), &dst, &fl, sk, XFRM_LOOKUP_WAIT);
	if (err < 0) {
1128 1129 1130 1131 1132
		if (err == -EREMOTE)
			err = ip6_dst_blackhole(sk, &dst, &fl);
		if (err < 0)
			goto out;
	}
L
Linus Torvalds 已提交
1133 1134

	if (hlimit < 0) {
H
Herbert Xu 已提交
1135
		if (ipv6_addr_is_multicast(&fl.fl6_dst))
L
Linus Torvalds 已提交
1136 1137 1138 1139
			hlimit = np->mcast_hops;
		else
			hlimit = np->hop_limit;
		if (hlimit < 0)
1140
			hlimit = ip6_dst_hoplimit(dst);
L
Linus Torvalds 已提交
1141 1142
	}

G
Gerrit Renker 已提交
1143
	if (tclass < 0)
1144 1145
		tclass = np->tclass;

1146 1147 1148
	if (dontfrag < 0)
		dontfrag = np->dontfrag;

L
Linus Torvalds 已提交
1149 1150 1151 1152 1153 1154 1155 1156 1157 1158
	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);

1159
		LIMIT_NETDEBUG(KERN_DEBUG "udp cork app bug 2\n");
L
Linus Torvalds 已提交
1160 1161 1162 1163 1164 1165 1166 1167
		err = -EINVAL;
		goto out;
	}

	up->pending = AF_INET6;

do_append_data:
	up->len += ulen;
1168 1169
	getfrag  =  is_udplite ?  udplite_getfrag : ip_generic_getfrag;
	err = ip6_append_data(sk, getfrag, msg->msg_iov, ulen,
H
Herbert Xu 已提交
1170
		sizeof(struct udphdr), hlimit, tclass, opt, &fl,
1171
		(struct rt6_info*)dst,
1172
		corkreq ? msg->msg_flags|MSG_MORE : msg->msg_flags, dontfrag);
L
Linus Torvalds 已提交
1173 1174 1175
	if (err)
		udp_v6_flush_pending_frames(sk);
	else if (!corkreq)
1176
		err = udp_v6_push_pending_frames(sk);
H
Herbert Xu 已提交
1177 1178
	else if (unlikely(skb_queue_empty(&sk->sk_write_queue)))
		up->pending = 0;
L
Linus Torvalds 已提交
1179

1180 1181 1182
	if (dst) {
		if (connected) {
			ip6_dst_store(sk, dst,
H
Herbert Xu 已提交
1183
				      ipv6_addr_equal(&fl.fl6_dst, &np->daddr) ?
1184 1185
				      &np->daddr : NULL,
#ifdef CONFIG_IPV6_SUBTREES
H
Herbert Xu 已提交
1186
				      ipv6_addr_equal(&fl.fl6_src, &np->saddr) ?
1187 1188 1189
				      &np->saddr :
#endif
				      NULL);
1190 1191 1192
		} else {
			dst_release(dst);
		}
1193
		dst = NULL;
1194 1195
	}

L
Linus Torvalds 已提交
1196 1197 1198 1199
	if (err > 0)
		err = np->recverr ? net_xmit_errno(err) : 0;
	release_sock(sk);
out:
1200
	dst_release(dst);
L
Linus Torvalds 已提交
1201
	fl6_sock_release(flowlabel);
1202
	if (!err)
L
Linus Torvalds 已提交
1203
		return len;
1204 1205 1206 1207 1208 1209 1210 1211
	/*
	 * 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)) {
1212 1213
		UDP6_INC_STATS_USER(sock_net(sk),
				UDP_MIB_SNDBUFERRORS, is_udplite);
1214
	}
L
Linus Torvalds 已提交
1215 1216 1217 1218 1219 1220 1221 1222 1223 1224
	return err;

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

1225
void udpv6_destroy_sock(struct sock *sk)
L
Linus Torvalds 已提交
1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236
{
	lock_sock(sk);
	udp_v6_flush_pending_frames(sk);
	release_sock(sk);

	inet6_destroy_sock(sk);
}

/*
 *	Socket option code for UDP
 */
1237
int udpv6_setsockopt(struct sock *sk, int level, int optname,
1238
		     char __user *optval, unsigned int optlen)
1239
{
1240
	if (level == SOL_UDP  ||  level == SOL_UDPLITE)
1241 1242
		return udp_lib_setsockopt(sk, level, optname, optval, optlen,
					  udp_v6_push_pending_frames);
1243
	return ipv6_setsockopt(sk, level, optname, optval, optlen);
1244 1245 1246
}

#ifdef CONFIG_COMPAT
1247
int compat_udpv6_setsockopt(struct sock *sk, int level, int optname,
1248
			    char __user *optval, unsigned int optlen)
1249
{
1250
	if (level == SOL_UDP  ||  level == SOL_UDPLITE)
1251 1252
		return udp_lib_setsockopt(sk, level, optname, optval, optlen,
					  udp_v6_push_pending_frames);
1253
	return compat_ipv6_setsockopt(sk, level, optname, optval, optlen);
1254 1255 1256
}
#endif

1257 1258
int udpv6_getsockopt(struct sock *sk, int level, int optname,
		     char __user *optval, int __user *optlen)
1259
{
1260
	if (level == SOL_UDP  ||  level == SOL_UDPLITE)
1261
		return udp_lib_getsockopt(sk, level, optname, optval, optlen);
1262
	return ipv6_getsockopt(sk, level, optname, optval, optlen);
1263 1264 1265
}

#ifdef CONFIG_COMPAT
1266 1267
int compat_udpv6_getsockopt(struct sock *sk, int level, int optname,
			    char __user *optval, int __user *optlen)
1268
{
1269
	if (level == SOL_UDP  ||  level == SOL_UDPLITE)
1270
		return udp_lib_getsockopt(sk, level, optname, optval, optlen);
1271
	return compat_ipv6_getsockopt(sk, level, optname, optval, optlen);
1272 1273 1274
}
#endif

1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365
static int udp6_ufo_send_check(struct sk_buff *skb)
{
	struct ipv6hdr *ipv6h;
	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;
}

static struct sk_buff *udp6_ufo_fragment(struct sk_buff *skb, int features)
{
	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.
	 */
	offset = skb->csum_start - skb_headroom(skb);
	csum = skb_checksum(skb, offset, skb->len- offset, 0);
	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. */
	if ((skb_headroom(skb) < frag_hdr_sz) &&
	    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;
	ipv6_select_ident(fptr);

	/* 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;
}

1366
static const struct inet6_protocol udpv6_protocol = {
L
Linus Torvalds 已提交
1367 1368
	.handler	=	udpv6_rcv,
	.err_handler	=	udpv6_err,
1369 1370
	.gso_send_check =	udp6_ufo_send_check,
	.gso_segment	=	udp6_ufo_fragment,
L
Linus Torvalds 已提交
1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385
	.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);
	struct in6_addr *dest, *src;
	__u16 destp, srcp;

	dest  = &np->daddr;
	src   = &np->rcv_saddr;
E
Eric Dumazet 已提交
1386 1387
	destp = ntohs(inet->inet_dport);
	srcp  = ntohs(inet->inet_sport);
L
Linus Torvalds 已提交
1388
	seq_printf(seq,
1389
		   "%5d: %08X%08X%08X%08X:%04X %08X%08X%08X%08X:%04X "
E
Eric Dumazet 已提交
1390
		   "%02X %08X:%08X %02X:%08lX %08X %5d %8d %lu %d %p %d\n",
L
Linus Torvalds 已提交
1391 1392 1393 1394 1395
		   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,
1396
		   sp->sk_state,
1397 1398
		   sk_wmem_alloc_get(sp),
		   sk_rmem_alloc_get(sp),
L
Linus Torvalds 已提交
1399 1400 1401
		   0, 0L, 0,
		   sock_i_uid(sp), 0,
		   sock_i_ino(sp),
E
Eric Dumazet 已提交
1402 1403
		   atomic_read(&sp->sk_refcnt), sp,
		   atomic_read(&sp->sk_drops));
L
Linus Torvalds 已提交
1404 1405
}

1406
int udp6_seq_show(struct seq_file *seq, void *v)
L
Linus Torvalds 已提交
1407 1408 1409 1410 1411 1412 1413
{
	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
1414
			   "   uid  timeout inode ref pointer drops\n");
L
Linus Torvalds 已提交
1415 1416 1417 1418 1419 1420 1421 1422
	else
		udp6_sock_seq_show(seq, v, ((struct udp_iter_state *)seq->private)->bucket);
	return 0;
}

static struct udp_seq_afinfo udp6_seq_afinfo = {
	.name		= "udp6",
	.family		= AF_INET6,
1423
	.udp_table	= &udp_table,
1424 1425 1426
	.seq_fops	= {
		.owner	=	THIS_MODULE,
	},
1427 1428 1429
	.seq_ops	= {
		.show		= udp6_seq_show,
	},
L
Linus Torvalds 已提交
1430 1431
};

1432
int __net_init udp6_proc_init(struct net *net)
L
Linus Torvalds 已提交
1433
{
1434
	return udp_proc_register(net, &udp6_seq_afinfo);
L
Linus Torvalds 已提交
1435 1436
}

1437 1438
void udp6_proc_exit(struct net *net) {
	udp_proc_unregister(net, &udp6_seq_afinfo);
L
Linus Torvalds 已提交
1439 1440 1441 1442 1443 1444
}
#endif /* CONFIG_PROC_FS */

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

struct proto udpv6_prot = {
1445 1446
	.name		   = "UDPv6",
	.owner		   = THIS_MODULE,
1447
	.close		   = udp_lib_close,
1448 1449 1450 1451 1452 1453 1454 1455 1456
	.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,
	.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
L
Linus Torvalds 已提交
1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483
};

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,
};


1484
int __init udpv6_init(void)
L
Linus Torvalds 已提交
1485
{
1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502
	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;
}

1503
void udpv6_exit(void)
1504 1505 1506
{
	inet6_unregister_protosw(&udpv6_protosw);
	inet6_del_protocol(&udpv6_protocol, IPPROTO_UDP);
L
Linus Torvalds 已提交
1507
}