udp.c 37.6 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);
57 58
	__be32 sk1_rcv_saddr = sk_rcv_saddr(sk);
	__be32 sk2_rcv_saddr = sk_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
static inline int compute_score(struct sock *sk, struct net *net,
				unsigned short hnum,
125 126
				const struct in6_addr *saddr, __be16 sport,
				const struct in6_addr *daddr, __be16 dport,
127 128 129 130
				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
#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
230
		if (unlikely(!atomic_inc_not_zero_hint(&result->sk_refcnt, 2)))
E
Eric Dumazet 已提交
231 232 233 234 235 236 237 238 239 240
			result = NULL;
		else if (unlikely(compute_score2(result, net, saddr, sport,
				  daddr, hnum, dif) < badness)) {
			sock_put(result);
			goto begin;
		}
	}
	return result;
}

241
static struct sock *__udp6_lib_lookup(struct net *net,
242 243
				      const struct in6_addr *saddr, __be16 sport,
				      const 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
	if (result) {
E
Eric Dumazet 已提交
297
		if (unlikely(!atomic_inc_not_zero_hint(&result->sk_refcnt, 2)))
298 299 300 301 302 303 304 305
			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
	const struct ipv6hdr *iph = ipv6_hdr(skb);
315

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
}

323 324 325 326 327 328 329 330
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
331 332 333 334 335
/*
 * 	This should be easy, if there is something there we
 * 	return it, otherwise we block.
 */

336
int udpv6_recvmsg(struct kiocb *iocb, struct sock *sk,
L
Linus Torvalds 已提交
337 338 339 340 341
		  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);
342
	struct sk_buff *skb;
G
Gerrit Renker 已提交
343
	unsigned int ulen;
344
	int peeked;
345 346
	int err;
	int is_udplite = IS_UDPLITE(sk);
347
	int is_udp4;
348
	bool slow;
L
Linus Torvalds 已提交
349

350 351 352
	if (addr_len)
		*addr_len=sizeof(struct sockaddr_in6);

L
Linus Torvalds 已提交
353 354 355
	if (flags & MSG_ERRQUEUE)
		return ipv6_recv_error(sk, msg, len);

356 357 358
	if (np->rxpmtu && np->rxopt.bits.rxpmtu)
		return ipv6_recv_rxpmtu(sk, msg, len);

L
Linus Torvalds 已提交
359
try_again:
360 361
	skb = __skb_recv_datagram(sk, flags | (noblock ? MSG_DONTWAIT : 0),
				  &peeked, &err);
L
Linus Torvalds 已提交
362 363 364
	if (!skb)
		goto out;

365
	ulen = skb->len - sizeof(struct udphdr);
G
Gerrit Renker 已提交
366 367 368
	if (len > ulen)
		len = ulen;
	else if (len < ulen)
369
		msg->msg_flags |= MSG_TRUNC;
L
Linus Torvalds 已提交
370

371 372
	is_udp4 = (skb->protocol == htons(ETH_P_IP));

373
	/*
374 375 376
	 * 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.
377 378
	 */

G
Gerrit Renker 已提交
379
	if (len < ulen || UDP_SKB_CB(skb)->partial_cov) {
380
		if (udp_lib_checksum_complete(skb))
L
Linus Torvalds 已提交
381
			goto csum_copy_err;
382 383
	}

384
	if (skb_csum_unnecessary(skb))
385
		err = skb_copy_datagram_iovec(skb, sizeof(struct udphdr),
G
Gerrit Renker 已提交
386
					      msg->msg_iov,len);
387
	else {
L
Linus Torvalds 已提交
388 389 390 391 392 393 394
		err = skb_copy_and_csum_datagram_iovec(skb, sizeof(struct udphdr), msg->msg_iov);
		if (err == -EINVAL)
			goto csum_copy_err;
	}
	if (err)
		goto out_free;

395 396 397 398 399 400 401 402
	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);
	}
403

404
	sock_recv_ts_and_drops(msg, sk, skb);
L
Linus Torvalds 已提交
405 406 407 408

	/* Copy the address. */
	if (msg->msg_name) {
		struct sockaddr_in6 *sin6;
409

L
Linus Torvalds 已提交
410 411
		sin6 = (struct sockaddr_in6 *) msg->msg_name;
		sin6->sin6_family = AF_INET6;
412
		sin6->sin6_port = udp_hdr(skb)->source;
L
Linus Torvalds 已提交
413 414 415
		sin6->sin6_flowinfo = 0;
		sin6->sin6_scope_id = 0;

416
		if (is_udp4)
B
Brian Haley 已提交
417 418
			ipv6_addr_set_v4mapped(ip_hdr(skb)->saddr,
					       &sin6->sin6_addr);
L
Linus Torvalds 已提交
419
		else {
420 421
			ipv6_addr_copy(&sin6->sin6_addr,
				       &ipv6_hdr(skb)->saddr);
L
Linus Torvalds 已提交
422 423 424 425 426
			if (ipv6_addr_type(&sin6->sin6_addr) & IPV6_ADDR_LINKLOCAL)
				sin6->sin6_scope_id = IP6CB(skb)->iif;
		}

	}
427
	if (is_udp4) {
L
Linus Torvalds 已提交
428 429 430 431 432
		if (inet->cmsg_flags)
			ip_cmsg_recv(msg, skb);
	} else {
		if (np->rxopt.all)
			datagram_recv_ctl(sk, msg, skb);
433
	}
L
Linus Torvalds 已提交
434

G
Gerrit Renker 已提交
435
	err = len;
L
Linus Torvalds 已提交
436
	if (flags & MSG_TRUNC)
437
		err = ulen;
L
Linus Torvalds 已提交
438 439

out_free:
440
	skb_free_datagram_locked(sk, skb);
L
Linus Torvalds 已提交
441 442 443 444
out:
	return err;

csum_copy_err:
445
	slow = lock_sock_fast(sk);
446 447 448 449 450 451 452 453
	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);
	}
454
	unlock_sock_fast(sk, slow);
L
Linus Torvalds 已提交
455

456
	if (noblock)
L
Linus Torvalds 已提交
457
		return -EAGAIN;
458 459 460

	/* starting over for a new packet */
	msg->msg_flags &= ~MSG_TRUNC;
L
Linus Torvalds 已提交
461 462 463
	goto try_again;
}

464
void __udp6_lib_err(struct sk_buff *skb, struct inet6_skb_parm *opt,
465
		    u8 type, u8 code, int offset, __be32 info,
466
		    struct udp_table *udptable)
L
Linus Torvalds 已提交
467 468
{
	struct ipv6_pinfo *np;
469 470 471
	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
472 473 474 475
	struct udphdr *uh = (struct udphdr*)(skb->data+offset);
	struct sock *sk;
	int err;

476
	sk = __udp6_lib_lookup(dev_net(skb->dev), daddr, uh->dest,
477
			       saddr, uh->source, inet6_iif(skb), udptable);
L
Linus Torvalds 已提交
478 479 480 481 482 483 484 485 486 487 488
	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;

489
	if (np->recverr)
L
Linus Torvalds 已提交
490
		ipv6_icmp_error(sk, skb, err, uh->dest, ntohl(info), (u8 *)(uh+1));
491

L
Linus Torvalds 已提交
492 493 494 495 496 497
	sk->sk_err = err;
	sk->sk_error_report(sk);
out:
	sock_put(sk);
}

498
static __inline__ void udpv6_err(struct sk_buff *skb,
499 500
				 struct inet6_skb_parm *opt, u8 type,
				 u8 code, int offset, __be32 info     )
501
{
502
	__udp6_lib_err(skb, opt, type, code, offset, info, &udp_table);
503 504 505
}

int udpv6_queue_rcv_skb(struct sock * sk, struct sk_buff *skb)
L
Linus Torvalds 已提交
506
{
507
	struct udp_sock *up = udp_sk(sk);
508
	int rc;
W
Wang Chen 已提交
509
	int is_udplite = IS_UDPLITE(sk);
510

511
	if (!ipv6_addr_any(&inet6_sk(sk)->daddr))
512
		sock_rps_save_rxhash(sk, skb);
513

514 515
	if (!xfrm6_policy_check(sk, XFRM_POLICY_IN, skb))
		goto drop;
L
Linus Torvalds 已提交
516

517 518 519
	/*
	 * UDP-Lite specific tests, ignored on UDP sockets (see net/ipv4/udp.c).
	 */
W
Wang Chen 已提交
520
	if ((is_udplite & UDPLITE_RECV_CC)  &&  UDP_SKB_CB(skb)->partial_cov) {
521 522 523 524 525 526 527 528 529 530 531 532 533

		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
534 535
	}

536
	if (rcu_access_pointer(sk->sk_filter)) {
537 538 539
		if (udp_lib_checksum_complete(skb))
			goto drop;
	}
540

E
Eric Dumazet 已提交
541
	if ((rc = ip_queue_rcv_skb(sk, skb)) < 0) {
542
		/* Note that an ENOMEM error is charged twice */
E
Eric Dumazet 已提交
543
		if (rc == -ENOMEM)
544 545
			UDP6_INC_STATS_BH(sock_net(sk),
					UDP_MIB_RCVBUFERRORS, is_udplite);
E
Eric Dumazet 已提交
546
		goto drop_no_sk_drops_inc;
L
Linus Torvalds 已提交
547
	}
548

L
Linus Torvalds 已提交
549
	return 0;
550
drop:
E
Eric Dumazet 已提交
551 552
	atomic_inc(&sk->sk_drops);
drop_no_sk_drops_inc:
553
	UDP6_INC_STATS_BH(sock_net(sk), UDP_MIB_INERRORS, is_udplite);
554 555
	kfree_skb(skb);
	return -1;
L
Linus Torvalds 已提交
556 557
}

558
static struct sock *udp_v6_mcast_next(struct net *net, struct sock *sk,
559 560
				      __be16 loc_port, const struct in6_addr *loc_addr,
				      __be16 rmt_port, const struct in6_addr *rmt_addr,
L
Linus Torvalds 已提交
561 562
				      int dif)
{
563
	struct hlist_nulls_node *node;
L
Linus Torvalds 已提交
564 565 566
	struct sock *s = sk;
	unsigned short num = ntohs(loc_port);

567
	sk_nulls_for_each_from(s, node) {
L
Linus Torvalds 已提交
568 569
		struct inet_sock *inet = inet_sk(s);

570
		if (!net_eq(sock_net(s), net))
571 572
			continue;

573 574
		if (udp_sk(s)->udp_port_hash == num &&
		    s->sk_family == PF_INET6) {
L
Linus Torvalds 已提交
575
			struct ipv6_pinfo *np = inet6_sk(s);
E
Eric Dumazet 已提交
576 577
			if (inet->inet_dport) {
				if (inet->inet_dport != rmt_port)
L
Linus Torvalds 已提交
578 579 580 581 582 583 584 585 586 587
					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)) {
588 589
				if (!ipv6_addr_equal(&np->rcv_saddr, loc_addr))
					continue;
L
Linus Torvalds 已提交
590
			}
S
Stephen Hemminger 已提交
591
			if (!inet6_mc_check(s, loc_addr, rmt_addr))
L
Linus Torvalds 已提交
592 593 594 595 596 597 598
				continue;
			return s;
		}
	}
	return NULL;
}

599 600 601 602 603 604 605 606 607 608
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);

609
		sk = stack[i];
610
		if (skb1) {
611
			if (sk_rcvqueues_full(sk, skb1)) {
612 613 614
				kfree_skb(skb1);
				goto drop;
			}
615 616 617
			bh_lock_sock(sk);
			if (!sock_owned_by_user(sk))
				udpv6_queue_rcv_skb(sk, skb1);
Z
Zhu Yi 已提交
618
			else if (sk_add_backlog(sk, skb1)) {
Z
Zhu Yi 已提交
619 620 621 622
				kfree_skb(skb1);
				bh_unlock_sock(sk);
				goto drop;
			}
623
			bh_unlock_sock(sk);
Z
Zhu Yi 已提交
624
			continue;
625
		}
Z
Zhu Yi 已提交
626 627 628 629 630 631
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));
632 633
	}
}
L
Linus Torvalds 已提交
634 635 636 637
/*
 * Note: called only from the BH handler context,
 * so we don't need to lock the hashes.
 */
638
static int __udp6_lib_mcast_deliver(struct net *net, struct sk_buff *skb,
639
		const struct in6_addr *saddr, const struct in6_addr *daddr,
640
		struct udp_table *udptable)
L
Linus Torvalds 已提交
641
{
642
	struct sock *sk, *stack[256 / sizeof(struct sock *)];
643
	const struct udphdr *uh = udp_hdr(skb);
644
	struct udp_hslot *hslot = udp_hashslot(udptable, net, ntohs(uh->dest));
L
Linus Torvalds 已提交
645
	int dif;
646
	unsigned int i, count = 0;
L
Linus Torvalds 已提交
647

648
	spin_lock(&hslot->lock);
649
	sk = sk_nulls_head(&hslot->head);
650
	dif = inet6_iif(skb);
651
	sk = udp_v6_mcast_next(net, sk, uh->dest, daddr, uh->source, saddr, dif);
652 653 654 655 656 657 658 659 660
	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
661
		}
L
Linus Torvalds 已提交
662
	}
663 664 665 666 667 668
	/*
	 * before releasing the lock, we must take reference on sockets
	 */
	for (i = 0; i < count; i++)
		sock_hold(stack[i]);

669
	spin_unlock(&hslot->lock);
670 671 672 673 674 675 676 677 678

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

		for (i = 0; i < count; i++)
			sock_put(stack[i]);
	} else {
		kfree_skb(skb);
	}
679
	return 0;
L
Linus Torvalds 已提交
680 681
}

682 683
static inline int udp6_csum_init(struct sk_buff *skb, struct udphdr *uh,
				 int proto)
684
{
685 686 687 688 689
	int err;

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

690
	if (proto == IPPROTO_UDPLITE) {
691 692 693 694 695
		err = udplite_checksum_init(skb, uh);
		if (err)
			return err;
	}

696 697 698 699 700 701 702 703
	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 &&
704
	    !csum_ipv6_magic(&ipv6_hdr(skb)->saddr, &ipv6_hdr(skb)->daddr,
705
			     skb->len, proto, skb->csum))
706 707
		skb->ip_summed = CHECKSUM_UNNECESSARY;

708
	if (!skb_csum_unnecessary(skb))
709 710
		skb->csum = ~csum_unfold(csum_ipv6_magic(&ipv6_hdr(skb)->saddr,
							 &ipv6_hdr(skb)->daddr,
711
							 skb->len, proto, 0));
712

713
	return 0;
714 715
}

716
int __udp6_lib_rcv(struct sk_buff *skb, struct udp_table *udptable,
717
		   int proto)
L
Linus Torvalds 已提交
718
{
719
	struct net *net = dev_net(skb->dev);
L
Linus Torvalds 已提交
720
	struct sock *sk;
721
	struct udphdr *uh;
722
	const struct in6_addr *saddr, *daddr;
L
Linus Torvalds 已提交
723 724 725
	u32 ulen = 0;

	if (!pskb_may_pull(skb, sizeof(struct udphdr)))
726
		goto discard;
L
Linus Torvalds 已提交
727

728 729
	saddr = &ipv6_hdr(skb)->saddr;
	daddr = &ipv6_hdr(skb)->daddr;
730
	uh = udp_hdr(skb);
L
Linus Torvalds 已提交
731 732

	ulen = ntohs(uh->len);
733 734
	if (ulen > skb->len)
		goto short_packet;
L
Linus Torvalds 已提交
735

736 737
	if (proto == IPPROTO_UDP) {
		/* UDP validates ulen. */
L
Linus Torvalds 已提交
738

739 740 741
		/* Check for jumbo payload */
		if (ulen == 0)
			ulen = skb->len;
L
Linus Torvalds 已提交
742

743 744
		if (ulen < sizeof(*uh))
			goto short_packet;
L
Linus Torvalds 已提交
745

746 747 748
		if (ulen < skb->len) {
			if (pskb_trim_rcsum(skb, ulen))
				goto short_packet;
749 750
			saddr = &ipv6_hdr(skb)->saddr;
			daddr = &ipv6_hdr(skb)->daddr;
751
			uh = udp_hdr(skb);
752 753
		}
	}
L
Linus Torvalds 已提交
754

755 756 757
	if (udp6_csum_init(skb, uh, proto))
		goto discard;

758 759
	/*
	 *	Multicast receive code
L
Linus Torvalds 已提交
760
	 */
761
	if (ipv6_addr_is_multicast(daddr))
762 763
		return __udp6_lib_mcast_deliver(net, skb,
				saddr, daddr, udptable);
L
Linus Torvalds 已提交
764 765 766

	/* Unicast */

767
	/*
L
Linus Torvalds 已提交
768 769 770
	 * check socket cache ... must talk to Alan about his plans
	 * for sock caches... i'll skip this for now.
	 */
771
	sk = __udp6_lib_lookup_skb(skb, uh->source, uh->dest, udptable);
L
Linus Torvalds 已提交
772 773 774 775 776

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

777
		if (udp_lib_checksum_complete(skb))
L
Linus Torvalds 已提交
778
			goto discard;
779 780
		UDP6_INC_STATS_BH(net, UDP_MIB_NOPORTS,
				proto == IPPROTO_UDPLITE);
L
Linus Torvalds 已提交
781

782
		icmpv6_send(skb, ICMPV6_DEST_UNREACH, ICMPV6_PORT_UNREACH, 0);
L
Linus Torvalds 已提交
783 784

		kfree_skb(skb);
S
Stephen Hemminger 已提交
785
		return 0;
L
Linus Torvalds 已提交
786
	}
787

L
Linus Torvalds 已提交
788
	/* deliver */
789

790 791 792 793
	if (sk_rcvqueues_full(sk, skb)) {
		sock_put(sk);
		goto discard;
	}
794
	bh_lock_sock(sk);
H
Hideo Aoki 已提交
795 796
	if (!sock_owned_by_user(sk))
		udpv6_queue_rcv_skb(sk, skb);
Z
Zhu Yi 已提交
797
	else if (sk_add_backlog(sk, skb)) {
Z
Zhu Yi 已提交
798 799 800 801 802
		atomic_inc(&sk->sk_drops);
		bh_unlock_sock(sk);
		sock_put(sk);
		goto discard;
	}
H
Hideo Aoki 已提交
803
	bh_unlock_sock(sk);
L
Linus Torvalds 已提交
804
	sock_put(sk);
S
Stephen Hemminger 已提交
805
	return 0;
L
Linus Torvalds 已提交
806

807
short_packet:
808
	LIMIT_NETDEBUG(KERN_DEBUG "UDP%sv6: short packet: From [%pI6c]:%u %d/%d to [%pI6c]:%u\n",
809
		       proto == IPPROTO_UDPLITE ? "-Lite" : "",
810 811 812 813 814 815
		       saddr,
		       ntohs(uh->source),
		       ulen,
		       skb->len,
		       daddr,
		       ntohs(uh->dest));
L
Linus Torvalds 已提交
816 817

discard:
818
	UDP6_INC_STATS_BH(net, UDP_MIB_INERRORS, proto == IPPROTO_UDPLITE);
L
Linus Torvalds 已提交
819
	kfree_skb(skb);
S
Stephen Hemminger 已提交
820
	return 0;
L
Linus Torvalds 已提交
821
}
822

823
static __inline__ int udpv6_rcv(struct sk_buff *skb)
824
{
825
	return __udp6_lib_rcv(skb, &udp_table, IPPROTO_UDP);
826 827
}

L
Linus Torvalds 已提交
828 829 830 831 832 833 834
/*
 * 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);

835 836 837
	if (up->pending == AF_INET)
		udp_flush_pending_frames(sk);
	else if (up->pending) {
L
Linus Torvalds 已提交
838 839 840
		up->len = 0;
		up->pending = 0;
		ip6_flush_pending_frames(sk);
841
	}
L
Linus Torvalds 已提交
842 843
}

844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884
/**
 * 	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
885 886 887 888
/*
 *	Sending
 */

889
static int udp_v6_push_pending_frames(struct sock *sk)
L
Linus Torvalds 已提交
890 891 892
{
	struct sk_buff *skb;
	struct udphdr *uh;
893
	struct udp_sock  *up = udp_sk(sk);
L
Linus Torvalds 已提交
894
	struct inet_sock *inet = inet_sk(sk);
895
	struct flowi6 *fl6 = &inet->cork.fl.u.ip6;
L
Linus Torvalds 已提交
896
	int err = 0;
W
Wang Chen 已提交
897
	int is_udplite = IS_UDPLITE(sk);
898
	__wsum csum = 0;
L
Linus Torvalds 已提交
899 900 901 902 903 904 905 906

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

	/*
	 * Create a UDP header
	 */
907
	uh = udp_hdr(skb);
908 909
	uh->source = fl6->fl6_sport;
	uh->dest = fl6->fl6_dport;
L
Linus Torvalds 已提交
910 911 912
	uh->len = htons(up->len);
	uh->check = 0;

W
Wang Chen 已提交
913
	if (is_udplite)
914
		csum = udplite_csum_outgoing(sk, skb);
915
	else if (skb->ip_summed == CHECKSUM_PARTIAL) { /* UDP hardware csum */
916
		udp6_hwcsum_outgoing(sk, skb, &fl6->saddr, &fl6->daddr,
917 918 919
				     up->len);
		goto send;
	} else
920
		csum = udp_csum_outgoing(sk, skb);
L
Linus Torvalds 已提交
921

922
	/* add protocol-dependent pseudo-header */
923 924
	uh->check = csum_ipv6_magic(&fl6->saddr, &fl6->daddr,
				    up->len, fl6->flowi6_proto, csum);
L
Linus Torvalds 已提交
925
	if (uh->check == 0)
926
		uh->check = CSUM_MANGLED_0;
L
Linus Torvalds 已提交
927

928
send:
L
Linus Torvalds 已提交
929
	err = ip6_push_pending_frames(sk);
E
Eric Dumazet 已提交
930 931 932 933 934 935 936 937 938
	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
939 940 941 942 943 944
out:
	up->len = 0;
	up->pending = 0;
	return err;
}

945
int udpv6_sendmsg(struct kiocb *iocb, struct sock *sk,
L
Linus Torvalds 已提交
946 947 948 949 950 951 952
		  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;
953
	struct in6_addr *daddr, *final_p, final;
L
Linus Torvalds 已提交
954 955
	struct ipv6_txoptions *opt = NULL;
	struct ip6_flowlabel *flowlabel = NULL;
956
	struct flowi6 fl6;
L
Linus Torvalds 已提交
957 958 959 960
	struct dst_entry *dst;
	int addr_len = msg->msg_namelen;
	int ulen = len;
	int hlimit = -1;
961
	int tclass = -1;
962
	int dontfrag = -1;
L
Linus Torvalds 已提交
963 964
	int corkreq = up->corkflag || msg->msg_flags&MSG_MORE;
	int err;
965
	int connected = 0;
W
Wang Chen 已提交
966
	int is_udplite = IS_UDPLITE(sk);
967
	int (*getfrag)(void *, char *, int, int, int, struct sk_buff *);
L
Linus Torvalds 已提交
968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993

	/* 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;
994
	} else
L
Linus Torvalds 已提交
995 996 997
		daddr = NULL;

	if (daddr) {
B
Brian Haley 已提交
998
		if (ipv6_addr_v4mapped(daddr)) {
L
Linus Torvalds 已提交
999 1000
			struct sockaddr_in sin;
			sin.sin_family = AF_INET;
E
Eric Dumazet 已提交
1001
			sin.sin_port = sin6 ? sin6->sin6_port : inet->inet_dport;
L
Linus Torvalds 已提交
1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015
			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,
1016
	   better check is made in ip6_append_data().
L
Linus Torvalds 已提交
1017 1018 1019
	   */
	if (len > INT_MAX - sizeof(struct udphdr))
		return -EMSGSIZE;
1020

L
Linus Torvalds 已提交
1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038
	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);

1039
	memset(&fl6, 0, sizeof(fl6));
L
Linus Torvalds 已提交
1040 1041 1042 1043 1044

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

1045
		fl6.fl6_dport = sin6->sin6_port;
L
Linus Torvalds 已提交
1046 1047 1048
		daddr = &sin6->sin6_addr;

		if (np->sndflow) {
1049 1050 1051
			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
1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068
				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)
1069
			fl6.flowi6_oif = sin6->sin6_scope_id;
L
Linus Torvalds 已提交
1070 1071 1072 1073
	} else {
		if (sk->sk_state != TCP_ESTABLISHED)
			return -EDESTADDRREQ;

1074
		fl6.fl6_dport = inet->inet_dport;
L
Linus Torvalds 已提交
1075
		daddr = &np->daddr;
1076
		fl6.flowlabel = np->flow_label;
1077
		connected = 1;
L
Linus Torvalds 已提交
1078 1079
	}

1080 1081
	if (!fl6.flowi6_oif)
		fl6.flowi6_oif = sk->sk_bound_dev_if;
L
Linus Torvalds 已提交
1082

1083 1084
	if (!fl6.flowi6_oif)
		fl6.flowi6_oif = np->sticky_pktinfo.ipi6_ifindex;
1085

1086
	fl6.flowi6_mark = sk->sk_mark;
B
Brian Haley 已提交
1087

L
Linus Torvalds 已提交
1088 1089 1090 1091 1092
	if (msg->msg_controllen) {
		opt = &opt_space;
		memset(opt, 0, sizeof(struct ipv6_txoptions));
		opt->tot_len = sizeof(*opt);

1093 1094
		err = datagram_send_ctl(sock_net(sk), sk, msg, &fl6, opt,
					&hlimit, &tclass, &dontfrag);
L
Linus Torvalds 已提交
1095 1096 1097 1098
		if (err < 0) {
			fl6_sock_release(flowlabel);
			return err;
		}
1099 1100
		if ((fl6.flowlabel&IPV6_FLOWLABEL_MASK) && !flowlabel) {
			flowlabel = fl6_sock_lookup(sk, fl6.flowlabel);
L
Linus Torvalds 已提交
1101 1102 1103 1104 1105
			if (flowlabel == NULL)
				return -EINVAL;
		}
		if (!(opt->opt_nflen|opt->opt_flen))
			opt = NULL;
1106
		connected = 0;
L
Linus Torvalds 已提交
1107 1108 1109
	}
	if (opt == NULL)
		opt = np->opt;
1110 1111 1112
	if (flowlabel)
		opt = fl6_merge_options(&opt_space, flowlabel, opt);
	opt = ipv6_fixup_options(&opt_space, opt);
L
Linus Torvalds 已提交
1113

1114
	fl6.flowi6_proto = sk->sk_protocol;
1115
	if (!ipv6_addr_any(daddr))
1116
		ipv6_addr_copy(&fl6.daddr, daddr);
1117
	else
1118 1119 1120
		fl6.daddr.s6_addr[15] = 0x1; /* :: means loopback (BSD'ism) */
	if (ipv6_addr_any(&fl6.saddr) && !ipv6_addr_any(&np->saddr))
		ipv6_addr_copy(&fl6.saddr, &np->saddr);
1121
	fl6.fl6_sport = inet->inet_sport;
1122

1123
	final_p = fl6_update_dst(&fl6, opt, &final);
1124
	if (final_p)
1125
		connected = 0;
L
Linus Torvalds 已提交
1126

1127 1128
	if (!fl6.flowi6_oif && ipv6_addr_is_multicast(&fl6.daddr)) {
		fl6.flowi6_oif = np->mcast_oif;
1129 1130
		connected = 0;
	}
L
Linus Torvalds 已提交
1131

1132
	security_sk_classify_flow(sk, flowi6_to_flowi(&fl6));
V
Venkat Yekkirala 已提交
1133

1134
	dst = ip6_sk_dst_lookup_flow(sk, &fl6, final_p, true);
1135 1136 1137
	if (IS_ERR(dst)) {
		err = PTR_ERR(dst);
		dst = NULL;
L
Linus Torvalds 已提交
1138
		goto out;
1139
	}
L
Linus Torvalds 已提交
1140 1141

	if (hlimit < 0) {
1142
		if (ipv6_addr_is_multicast(&fl6.daddr))
L
Linus Torvalds 已提交
1143 1144 1145 1146
			hlimit = np->mcast_hops;
		else
			hlimit = np->hop_limit;
		if (hlimit < 0)
1147
			hlimit = ip6_dst_hoplimit(dst);
L
Linus Torvalds 已提交
1148 1149
	}

G
Gerrit Renker 已提交
1150
	if (tclass < 0)
1151 1152
		tclass = np->tclass;

1153 1154 1155
	if (dontfrag < 0)
		dontfrag = np->dontfrag;

L
Linus Torvalds 已提交
1156 1157 1158 1159 1160 1161 1162 1163 1164 1165
	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);

1166
		LIMIT_NETDEBUG(KERN_DEBUG "udp cork app bug 2\n");
L
Linus Torvalds 已提交
1167 1168 1169 1170 1171 1172 1173 1174
		err = -EINVAL;
		goto out;
	}

	up->pending = AF_INET6;

do_append_data:
	up->len += ulen;
1175 1176
	getfrag  =  is_udplite ?  udplite_getfrag : ip_generic_getfrag;
	err = ip6_append_data(sk, getfrag, msg->msg_iov, ulen,
1177
		sizeof(struct udphdr), hlimit, tclass, opt, &fl6,
1178
		(struct rt6_info*)dst,
1179
		corkreq ? msg->msg_flags|MSG_MORE : msg->msg_flags, dontfrag);
L
Linus Torvalds 已提交
1180 1181 1182
	if (err)
		udp_v6_flush_pending_frames(sk);
	else if (!corkreq)
1183
		err = udp_v6_push_pending_frames(sk);
H
Herbert Xu 已提交
1184 1185
	else if (unlikely(skb_queue_empty(&sk->sk_write_queue)))
		up->pending = 0;
L
Linus Torvalds 已提交
1186

1187 1188 1189
	if (dst) {
		if (connected) {
			ip6_dst_store(sk, dst,
1190
				      ipv6_addr_equal(&fl6.daddr, &np->daddr) ?
1191 1192
				      &np->daddr : NULL,
#ifdef CONFIG_IPV6_SUBTREES
1193
				      ipv6_addr_equal(&fl6.saddr, &np->saddr) ?
1194 1195 1196
				      &np->saddr :
#endif
				      NULL);
1197 1198 1199
		} else {
			dst_release(dst);
		}
1200
		dst = NULL;
1201 1202
	}

L
Linus Torvalds 已提交
1203 1204 1205 1206
	if (err > 0)
		err = np->recverr ? net_xmit_errno(err) : 0;
	release_sock(sk);
out:
1207
	dst_release(dst);
L
Linus Torvalds 已提交
1208
	fl6_sock_release(flowlabel);
1209
	if (!err)
L
Linus Torvalds 已提交
1210
		return len;
1211 1212 1213 1214 1215 1216 1217 1218
	/*
	 * 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)) {
1219 1220
		UDP6_INC_STATS_USER(sock_net(sk),
				UDP_MIB_SNDBUFERRORS, is_udplite);
1221
	}
L
Linus Torvalds 已提交
1222 1223 1224 1225 1226 1227 1228 1229 1230 1231
	return err;

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

1232
void udpv6_destroy_sock(struct sock *sk)
L
Linus Torvalds 已提交
1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243
{
	lock_sock(sk);
	udp_v6_flush_pending_frames(sk);
	release_sock(sk);

	inet6_destroy_sock(sk);
}

/*
 *	Socket option code for UDP
 */
1244
int udpv6_setsockopt(struct sock *sk, int level, int optname,
1245
		     char __user *optval, unsigned int optlen)
1246
{
1247
	if (level == SOL_UDP  ||  level == SOL_UDPLITE)
1248 1249
		return udp_lib_setsockopt(sk, level, optname, optval, optlen,
					  udp_v6_push_pending_frames);
1250
	return ipv6_setsockopt(sk, level, optname, optval, optlen);
1251 1252 1253
}

#ifdef CONFIG_COMPAT
1254
int compat_udpv6_setsockopt(struct sock *sk, int level, int optname,
1255
			    char __user *optval, unsigned int optlen)
1256
{
1257
	if (level == SOL_UDP  ||  level == SOL_UDPLITE)
1258 1259
		return udp_lib_setsockopt(sk, level, optname, optval, optlen,
					  udp_v6_push_pending_frames);
1260
	return compat_ipv6_setsockopt(sk, level, optname, optval, optlen);
1261 1262 1263
}
#endif

1264 1265
int udpv6_getsockopt(struct sock *sk, int level, int optname,
		     char __user *optval, int __user *optlen)
1266
{
1267
	if (level == SOL_UDP  ||  level == SOL_UDPLITE)
1268
		return udp_lib_getsockopt(sk, level, optname, optval, optlen);
1269
	return ipv6_getsockopt(sk, level, optname, optval, optlen);
1270 1271 1272
}

#ifdef CONFIG_COMPAT
1273 1274
int compat_udpv6_getsockopt(struct sock *sk, int level, int optname,
			    char __user *optval, int __user *optlen)
1275
{
1276
	if (level == SOL_UDP  ||  level == SOL_UDPLITE)
1277
		return udp_lib_getsockopt(sk, level, optname, optval, optlen);
1278
	return compat_ipv6_getsockopt(sk, level, optname, optval, optlen);
1279 1280 1281
}
#endif

1282 1283
static int udp6_ufo_send_check(struct sk_buff *skb)
{
1284
	const struct ipv6hdr *ipv6h;
1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300
	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;
}

1301
static struct sk_buff *udp6_ufo_fragment(struct sk_buff *skb, u32 features)
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
{
	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.
	 */
1334
	offset = skb_checksum_start_offset(skb);
1335 1336 1337 1338 1339 1340
	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. */
1341
	if ((skb_mac_header(skb) < skb->head + frag_hdr_sz) &&
1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361
	    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;
1362
	ipv6_select_ident(fptr, (struct rt6_info *)skb_dst(skb));
1363 1364 1365 1366 1367 1368 1369 1370 1371 1372

	/* 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;
}

1373
static const struct inet6_protocol udpv6_protocol = {
L
Linus Torvalds 已提交
1374 1375
	.handler	=	udpv6_rcv,
	.err_handler	=	udpv6_err,
1376 1377
	.gso_send_check =	udp6_ufo_send_check,
	.gso_segment	=	udp6_ufo_fragment,
L
Linus Torvalds 已提交
1378 1379 1380 1381 1382 1383 1384 1385 1386 1387
	.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);
1388
	const struct in6_addr *dest, *src;
L
Linus Torvalds 已提交
1389 1390 1391 1392
	__u16 destp, srcp;

	dest  = &np->daddr;
	src   = &np->rcv_saddr;
E
Eric Dumazet 已提交
1393 1394
	destp = ntohs(inet->inet_dport);
	srcp  = ntohs(inet->inet_sport);
L
Linus Torvalds 已提交
1395
	seq_printf(seq,
1396
		   "%5d: %08X%08X%08X%08X:%04X %08X%08X%08X%08X:%04X "
D
Dan Rosenberg 已提交
1397
		   "%02X %08X:%08X %02X:%08lX %08X %5d %8d %lu %d %pK %d\n",
L
Linus Torvalds 已提交
1398 1399 1400 1401 1402
		   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,
1403
		   sp->sk_state,
1404 1405
		   sk_wmem_alloc_get(sp),
		   sk_rmem_alloc_get(sp),
L
Linus Torvalds 已提交
1406 1407 1408
		   0, 0L, 0,
		   sock_i_uid(sp), 0,
		   sock_i_ino(sp),
E
Eric Dumazet 已提交
1409 1410
		   atomic_read(&sp->sk_refcnt), sp,
		   atomic_read(&sp->sk_drops));
L
Linus Torvalds 已提交
1411 1412
}

1413
int udp6_seq_show(struct seq_file *seq, void *v)
L
Linus Torvalds 已提交
1414 1415 1416 1417 1418 1419 1420
{
	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
1421
			   "   uid  timeout inode ref pointer drops\n");
L
Linus Torvalds 已提交
1422 1423 1424 1425 1426
	else
		udp6_sock_seq_show(seq, v, ((struct udp_iter_state *)seq->private)->bucket);
	return 0;
}

1427 1428 1429 1430 1431 1432 1433 1434
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
1435 1436 1437
static struct udp_seq_afinfo udp6_seq_afinfo = {
	.name		= "udp6",
	.family		= AF_INET6,
1438
	.udp_table	= &udp_table,
1439
	.seq_fops	= &udp6_afinfo_seq_fops,
1440 1441 1442
	.seq_ops	= {
		.show		= udp6_seq_show,
	},
L
Linus Torvalds 已提交
1443 1444
};

1445
int __net_init udp6_proc_init(struct net *net)
L
Linus Torvalds 已提交
1446
{
1447
	return udp_proc_register(net, &udp6_seq_afinfo);
L
Linus Torvalds 已提交
1448 1449
}

1450 1451
void udp6_proc_exit(struct net *net) {
	udp_proc_unregister(net, &udp6_seq_afinfo);
L
Linus Torvalds 已提交
1452 1453 1454 1455 1456 1457
}
#endif /* CONFIG_PROC_FS */

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

struct proto udpv6_prot = {
1458 1459
	.name		   = "UDPv6",
	.owner		   = THIS_MODULE,
1460
	.close		   = udp_lib_close,
1461 1462 1463 1464 1465 1466 1467 1468 1469
	.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,
1470 1471
	.hash		   = udp_lib_hash,
	.unhash		   = udp_lib_unhash,
E
Eric Dumazet 已提交
1472
	.rehash		   = udp_v6_rehash,
1473
	.get_port	   = udp_v6_get_port,
H
Hideo Aoki 已提交
1474 1475 1476 1477
	.memory_allocated  = &udp_memory_allocated,
	.sysctl_mem	   = sysctl_udp_mem,
	.sysctl_wmem	   = &sysctl_udp_wmem_min,
	.sysctl_rmem	   = &sysctl_udp_rmem_min,
1478
	.obj_size	   = sizeof(struct udp6_sock),
1479
	.slab_flags	   = SLAB_DESTROY_BY_RCU,
1480
	.h.udp_table	   = &udp_table,
1481
#ifdef CONFIG_COMPAT
1482 1483
	.compat_setsockopt = compat_udpv6_setsockopt,
	.compat_getsockopt = compat_udpv6_getsockopt,
1484
#endif
1485
	.clear_sk	   = sk_prot_clear_portaddr_nulls,
L
Linus Torvalds 已提交
1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497
};

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,
};


1498
int __init udpv6_init(void)
L
Linus Torvalds 已提交
1499
{
1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516
	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;
}

1517
void udpv6_exit(void)
1518 1519 1520
{
	inet6_unregister_protosw(&udpv6_protosw);
	inet6_del_protocol(&udpv6_protocol, IPPROTO_UDP);
L
Linus Torvalds 已提交
1521
}