udp.c 36.7 KB
Newer Older
L
Linus Torvalds 已提交
1 2
/*
 *	UDP over IPv6
3
 *	Linux INET6 implementation
L
Linus Torvalds 已提交
4 5
 *
 *	Authors:
6
 *	Pedro Roque		<roque@di.fc.ul.pt>
L
Linus Torvalds 已提交
7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34
 *
 *	Based on linux/ipv4/udp.c
 *
 *	Fixes:
 *	Hideaki YOSHIFUJI	:	sin6_scope_id support
 *	YOSHIFUJI Hideaki @USAGI and:	Support IPV6_V6ONLY socket option, which
 *	Alexey Kuznetsov		allow both IPv4 and IPv6 sockets to bind
 *					a single port at the same time.
 *      Kazunori MIYAZAWA @USAGI:       change process style to use ip6_append_data
 *      YOSHIFUJI Hideaki @USAGI:	convert /proc/net/udp6 to seq_file.
 *
 *	This program is free software; you can redistribute it and/or
 *      modify it under the terms of the GNU General Public License
 *      as published by the Free Software Foundation; either version
 *      2 of the License, or (at your option) any later version.
 */

#include <linux/errno.h>
#include <linux/types.h>
#include <linux/socket.h>
#include <linux/sockios.h>
#include <linux/net.h>
#include <linux/in6.h>
#include <linux/netdevice.h>
#include <linux/if_arp.h>
#include <linux/ipv6.h>
#include <linux/icmpv6.h>
#include <linux/init.h>
35
#include <linux/module.h>
36
#include <linux/skbuff.h>
37
#include <linux/slab.h>
L
Linus Torvalds 已提交
38 39 40 41 42 43 44
#include <asm/uaccess.h>

#include <net/ndisc.h>
#include <net/protocol.h>
#include <net/transp_v6.h>
#include <net/ip6_route.h>
#include <net/raw.h>
45
#include <net/tcp_states.h>
L
Linus Torvalds 已提交
46 47 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
}

114 115 116 117 118 119 120 121
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;

122
	if (net_eq(sock_net(sk), net) && udp_sk(sk)->udp_port_hash == hnum &&
123 124 125 126 127
			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
128 129
		if (inet->inet_dport) {
			if (inet->inet_dport != sport)
130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151
				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
152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231
#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;
}

232 233
static struct sock *__udp6_lib_lookup(struct net *net,
				      struct in6_addr *saddr, __be16 sport,
234
				      struct in6_addr *daddr, __be16 dport,
235
				      int dif, struct udp_table *udptable)
L
Linus Torvalds 已提交
236
{
237
	struct sock *sk, *result;
238
	struct hlist_nulls_node *node;
L
Linus Torvalds 已提交
239
	unsigned short hnum = ntohs(dport);
E
Eric Dumazet 已提交
240 241
	unsigned int hash2, slot2, slot = udp_hashfn(net, hnum, udptable->mask);
	struct udp_hslot *hslot2, *hslot = &udptable->hash[slot];
242
	int score, badness;
243

244
	rcu_read_lock();
E
Eric Dumazet 已提交
245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261
	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;

262 263
			result = udp6_lib_lookup2(net, saddr, sport,
						  &in6addr_any, hnum, dif,
E
Eric Dumazet 已提交
264 265 266 267 268
						  hslot2, slot2);
		}
		rcu_read_unlock();
		return result;
	}
269 270 271
begin:
	result = NULL;
	badness = -1;
272
	sk_nulls_for_each_rcu(sk, node, &hslot->head) {
273 274 275 276
		score = compute_score(sk, net, hnum, saddr, sport, daddr, dport, dif);
		if (score > badness) {
			result = sk;
			badness = score;
L
Linus Torvalds 已提交
277 278
		}
	}
279 280 281 282 283
	/*
	 * 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
284
	if (get_nulls_value(node) != slot)
285 286
		goto begin;

287 288 289 290 291 292 293 294 295 296
	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
297 298 299
	return result;
}

300 301
static struct sock *__udp6_lib_lookup_skb(struct sk_buff *skb,
					  __be16 sport, __be16 dport,
302
					  struct udp_table *udptable)
303
{
304
	struct sock *sk;
305 306
	struct ipv6hdr *iph = ipv6_hdr(skb);

307 308
	if (unlikely(sk = skb_steal_sock(skb)))
		return sk;
E
Eric Dumazet 已提交
309 310 311
	return __udp6_lib_lookup(dev_net(skb_dst(skb)->dev), &iph->saddr, sport,
				 &iph->daddr, dport, inet6_iif(skb),
				 udptable);
312 313
}

L
Linus Torvalds 已提交
314 315 316 317 318
/*
 * 	This should be easy, if there is something there we
 * 	return it, otherwise we block.
 */

319
int udpv6_recvmsg(struct kiocb *iocb, struct sock *sk,
L
Linus Torvalds 已提交
320 321 322 323 324
		  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);
325
	struct sk_buff *skb;
G
Gerrit Renker 已提交
326
	unsigned int ulen;
327
	int peeked;
328 329
	int err;
	int is_udplite = IS_UDPLITE(sk);
330
	int is_udp4;
L
Linus Torvalds 已提交
331

332 333 334
	if (addr_len)
		*addr_len=sizeof(struct sockaddr_in6);

L
Linus Torvalds 已提交
335 336 337
	if (flags & MSG_ERRQUEUE)
		return ipv6_recv_error(sk, msg, len);

338 339 340
	if (np->rxpmtu && np->rxopt.bits.rxpmtu)
		return ipv6_recv_rxpmtu(sk, msg, len);

L
Linus Torvalds 已提交
341
try_again:
342 343
	skb = __skb_recv_datagram(sk, flags | (noblock ? MSG_DONTWAIT : 0),
				  &peeked, &err);
L
Linus Torvalds 已提交
344 345 346
	if (!skb)
		goto out;

347
	ulen = skb->len - sizeof(struct udphdr);
G
Gerrit Renker 已提交
348 349 350
	if (len > ulen)
		len = ulen;
	else if (len < ulen)
351
		msg->msg_flags |= MSG_TRUNC;
L
Linus Torvalds 已提交
352

353 354
	is_udp4 = (skb->protocol == htons(ETH_P_IP));

355
	/*
356 357 358
	 * 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.
359 360
	 */

G
Gerrit Renker 已提交
361
	if (len < ulen || UDP_SKB_CB(skb)->partial_cov) {
362
		if (udp_lib_checksum_complete(skb))
L
Linus Torvalds 已提交
363
			goto csum_copy_err;
364 365
	}

366
	if (skb_csum_unnecessary(skb))
367
		err = skb_copy_datagram_iovec(skb, sizeof(struct udphdr),
G
Gerrit Renker 已提交
368
					      msg->msg_iov,len);
369
	else {
L
Linus Torvalds 已提交
370 371 372 373 374 375 376
		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;

377 378 379 380 381 382 383 384
	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);
	}
385

386
	sock_recv_ts_and_drops(msg, sk, skb);
L
Linus Torvalds 已提交
387 388 389 390

	/* Copy the address. */
	if (msg->msg_name) {
		struct sockaddr_in6 *sin6;
391

L
Linus Torvalds 已提交
392 393
		sin6 = (struct sockaddr_in6 *) msg->msg_name;
		sin6->sin6_family = AF_INET6;
394
		sin6->sin6_port = udp_hdr(skb)->source;
L
Linus Torvalds 已提交
395 396 397
		sin6->sin6_flowinfo = 0;
		sin6->sin6_scope_id = 0;

398
		if (is_udp4)
B
Brian Haley 已提交
399 400
			ipv6_addr_set_v4mapped(ip_hdr(skb)->saddr,
					       &sin6->sin6_addr);
L
Linus Torvalds 已提交
401
		else {
402 403
			ipv6_addr_copy(&sin6->sin6_addr,
				       &ipv6_hdr(skb)->saddr);
L
Linus Torvalds 已提交
404 405 406 407 408
			if (ipv6_addr_type(&sin6->sin6_addr) & IPV6_ADDR_LINKLOCAL)
				sin6->sin6_scope_id = IP6CB(skb)->iif;
		}

	}
409
	if (is_udp4) {
L
Linus Torvalds 已提交
410 411 412 413 414
		if (inet->cmsg_flags)
			ip_cmsg_recv(msg, skb);
	} else {
		if (np->rxopt.all)
			datagram_recv_ctl(sk, msg, skb);
415
	}
L
Linus Torvalds 已提交
416

G
Gerrit Renker 已提交
417
	err = len;
L
Linus Torvalds 已提交
418
	if (flags & MSG_TRUNC)
419
		err = ulen;
L
Linus Torvalds 已提交
420 421

out_free:
422
	skb_free_datagram_locked(sk, skb);
L
Linus Torvalds 已提交
423 424 425 426
out:
	return err;

csum_copy_err:
E
Eric Dumazet 已提交
427
	lock_sock_bh(sk);
428 429 430 431 432 433 434 435
	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);
	}
E
Eric Dumazet 已提交
436
	unlock_sock_bh(sk);
L
Linus Torvalds 已提交
437

438
	if (flags & MSG_DONTWAIT)
L
Linus Torvalds 已提交
439 440 441 442
		return -EAGAIN;
	goto try_again;
}

443
void __udp6_lib_err(struct sk_buff *skb, struct inet6_skb_parm *opt,
444
		    u8 type, u8 code, int offset, __be32 info,
445
		    struct udp_table *udptable)
L
Linus Torvalds 已提交
446 447 448 449 450 451 452 453 454
{
	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;

455
	sk = __udp6_lib_lookup(dev_net(skb->dev), daddr, uh->dest,
456
			       saddr, uh->source, inet6_iif(skb), udptable);
L
Linus Torvalds 已提交
457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476
	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;

	if (np->recverr)
		ipv6_icmp_error(sk, skb, err, uh->dest, ntohl(info), (u8 *)(uh+1));

	sk->sk_err = err;
	sk->sk_error_report(sk);
out:
	sock_put(sk);
}

477
static __inline__ void udpv6_err(struct sk_buff *skb,
478 479
				 struct inet6_skb_parm *opt, u8 type,
				 u8 code, int offset, __be32 info     )
480
{
481
	__udp6_lib_err(skb, opt, type, code, offset, info, &udp_table);
482 483 484
}

int udpv6_queue_rcv_skb(struct sock * sk, struct sk_buff *skb)
L
Linus Torvalds 已提交
485
{
486
	struct udp_sock *up = udp_sk(sk);
487
	int rc;
W
Wang Chen 已提交
488
	int is_udplite = IS_UDPLITE(sk);
489

490 491
	if (!xfrm6_policy_check(sk, XFRM_POLICY_IN, skb))
		goto drop;
L
Linus Torvalds 已提交
492

493 494 495
	/*
	 * UDP-Lite specific tests, ignored on UDP sockets (see net/ipv4/udp.c).
	 */
W
Wang Chen 已提交
496
	if ((is_udplite & UDPLITE_RECV_CC)  &&  UDP_SKB_CB(skb)->partial_cov) {
497 498 499 500 501 502 503 504 505 506 507 508 509

		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
510 511
	}

512 513 514 515
	if (sk->sk_filter) {
		if (udp_lib_checksum_complete(skb))
			goto drop;
	}
516

E
Eric Dumazet 已提交
517
	if ((rc = sock_queue_rcv_skb(sk, skb)) < 0) {
518
		/* Note that an ENOMEM error is charged twice */
E
Eric Dumazet 已提交
519
		if (rc == -ENOMEM)
520 521
			UDP6_INC_STATS_BH(sock_net(sk),
					UDP_MIB_RCVBUFERRORS, is_udplite);
E
Eric Dumazet 已提交
522
		goto drop_no_sk_drops_inc;
L
Linus Torvalds 已提交
523
	}
524

L
Linus Torvalds 已提交
525
	return 0;
526
drop:
E
Eric Dumazet 已提交
527 528
	atomic_inc(&sk->sk_drops);
drop_no_sk_drops_inc:
529
	UDP6_INC_STATS_BH(sock_net(sk), UDP_MIB_INERRORS, is_udplite);
530 531
	kfree_skb(skb);
	return -1;
L
Linus Torvalds 已提交
532 533
}

534
static struct sock *udp_v6_mcast_next(struct net *net, struct sock *sk,
A
Al Viro 已提交
535 536
				      __be16 loc_port, struct in6_addr *loc_addr,
				      __be16 rmt_port, struct in6_addr *rmt_addr,
L
Linus Torvalds 已提交
537 538
				      int dif)
{
539
	struct hlist_nulls_node *node;
L
Linus Torvalds 已提交
540 541 542
	struct sock *s = sk;
	unsigned short num = ntohs(loc_port);

543
	sk_nulls_for_each_from(s, node) {
L
Linus Torvalds 已提交
544 545
		struct inet_sock *inet = inet_sk(s);

546
		if (!net_eq(sock_net(s), net))
547 548
			continue;

549 550
		if (udp_sk(s)->udp_port_hash == num &&
		    s->sk_family == PF_INET6) {
L
Linus Torvalds 已提交
551
			struct ipv6_pinfo *np = inet6_sk(s);
E
Eric Dumazet 已提交
552 553
			if (inet->inet_dport) {
				if (inet->inet_dport != rmt_port)
L
Linus Torvalds 已提交
554 555 556 557 558 559 560 561 562 563
					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)) {
564 565
				if (!ipv6_addr_equal(&np->rcv_saddr, loc_addr))
					continue;
L
Linus Torvalds 已提交
566
			}
S
Stephen Hemminger 已提交
567
			if (!inet6_mc_check(s, loc_addr, rmt_addr))
L
Linus Torvalds 已提交
568 569 570 571 572 573 574
				continue;
			return s;
		}
	}
	return NULL;
}

575 576 577 578 579 580 581 582 583 584
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);

585
		sk = stack[i];
586
		if (skb1) {
587 588 589 590
			if (sk_rcvqueues_full(sk, skb)) {
				kfree_skb(skb1);
				goto drop;
			}
591 592 593
			bh_lock_sock(sk);
			if (!sock_owned_by_user(sk))
				udpv6_queue_rcv_skb(sk, skb1);
Z
Zhu Yi 已提交
594
			else if (sk_add_backlog(sk, skb1)) {
Z
Zhu Yi 已提交
595 596 597 598
				kfree_skb(skb1);
				bh_unlock_sock(sk);
				goto drop;
			}
599
			bh_unlock_sock(sk);
Z
Zhu Yi 已提交
600
			continue;
601
		}
Z
Zhu Yi 已提交
602 603 604 605 606 607
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));
608 609
	}
}
L
Linus Torvalds 已提交
610 611 612 613
/*
 * Note: called only from the BH handler context,
 * so we don't need to lock the hashes.
 */
614 615
static int __udp6_lib_mcast_deliver(struct net *net, struct sk_buff *skb,
		struct in6_addr *saddr, struct in6_addr *daddr,
616
		struct udp_table *udptable)
L
Linus Torvalds 已提交
617
{
618
	struct sock *sk, *stack[256 / sizeof(struct sock *)];
619
	const struct udphdr *uh = udp_hdr(skb);
620
	struct udp_hslot *hslot = udp_hashslot(udptable, net, ntohs(uh->dest));
L
Linus Torvalds 已提交
621
	int dif;
622
	unsigned int i, count = 0;
L
Linus Torvalds 已提交
623

624
	spin_lock(&hslot->lock);
625
	sk = sk_nulls_head(&hslot->head);
626
	dif = inet6_iif(skb);
627
	sk = udp_v6_mcast_next(net, sk, uh->dest, daddr, uh->source, saddr, dif);
628 629 630 631 632 633 634 635 636
	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
637
		}
L
Linus Torvalds 已提交
638
	}
639 640 641 642 643 644
	/*
	 * before releasing the lock, we must take reference on sockets
	 */
	for (i = 0; i < count; i++)
		sock_hold(stack[i]);

645
	spin_unlock(&hslot->lock);
646 647 648 649 650 651 652 653 654

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

		for (i = 0; i < count; i++)
			sock_put(stack[i]);
	} else {
		kfree_skb(skb);
	}
655
	return 0;
L
Linus Torvalds 已提交
656 657
}

658 659
static inline int udp6_csum_init(struct sk_buff *skb, struct udphdr *uh,
				 int proto)
660
{
661 662 663 664 665
	int err;

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

666
	if (proto == IPPROTO_UDPLITE) {
667 668 669 670 671
		err = udplite_checksum_init(skb, uh);
		if (err)
			return err;
	}

672 673 674 675 676 677 678 679
	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 &&
680
	    !csum_ipv6_magic(&ipv6_hdr(skb)->saddr, &ipv6_hdr(skb)->daddr,
681
			     skb->len, proto, skb->csum))
682 683
		skb->ip_summed = CHECKSUM_UNNECESSARY;

684
	if (!skb_csum_unnecessary(skb))
685 686
		skb->csum = ~csum_unfold(csum_ipv6_magic(&ipv6_hdr(skb)->saddr,
							 &ipv6_hdr(skb)->daddr,
687
							 skb->len, proto, 0));
688

689
	return 0;
690 691
}

692
int __udp6_lib_rcv(struct sk_buff *skb, struct udp_table *udptable,
693
		   int proto)
L
Linus Torvalds 已提交
694
{
695
	struct net *net = dev_net(skb->dev);
L
Linus Torvalds 已提交
696
	struct sock *sk;
697
	struct udphdr *uh;
L
Linus Torvalds 已提交
698 699 700 701 702 703
	struct in6_addr *saddr, *daddr;
	u32 ulen = 0;

	if (!pskb_may_pull(skb, sizeof(struct udphdr)))
		goto short_packet;

704 705
	saddr = &ipv6_hdr(skb)->saddr;
	daddr = &ipv6_hdr(skb)->daddr;
706
	uh = udp_hdr(skb);
L
Linus Torvalds 已提交
707 708

	ulen = ntohs(uh->len);
709 710
	if (ulen > skb->len)
		goto short_packet;
L
Linus Torvalds 已提交
711

712 713
	if (proto == IPPROTO_UDP) {
		/* UDP validates ulen. */
L
Linus Torvalds 已提交
714

715 716 717
		/* Check for jumbo payload */
		if (ulen == 0)
			ulen = skb->len;
L
Linus Torvalds 已提交
718

719 720
		if (ulen < sizeof(*uh))
			goto short_packet;
L
Linus Torvalds 已提交
721

722 723 724
		if (ulen < skb->len) {
			if (pskb_trim_rcsum(skb, ulen))
				goto short_packet;
725 726
			saddr = &ipv6_hdr(skb)->saddr;
			daddr = &ipv6_hdr(skb)->daddr;
727
			uh = udp_hdr(skb);
728 729
		}
	}
L
Linus Torvalds 已提交
730

731 732 733
	if (udp6_csum_init(skb, uh, proto))
		goto discard;

734 735
	/*
	 *	Multicast receive code
L
Linus Torvalds 已提交
736
	 */
737
	if (ipv6_addr_is_multicast(daddr))
738 739
		return __udp6_lib_mcast_deliver(net, skb,
				saddr, daddr, udptable);
L
Linus Torvalds 已提交
740 741 742

	/* Unicast */

743
	/*
L
Linus Torvalds 已提交
744 745 746
	 * check socket cache ... must talk to Alan about his plans
	 * for sock caches... i'll skip this for now.
	 */
747
	sk = __udp6_lib_lookup_skb(skb, uh->source, uh->dest, udptable);
L
Linus Torvalds 已提交
748 749 750 751 752

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

753
		if (udp_lib_checksum_complete(skb))
L
Linus Torvalds 已提交
754
			goto discard;
755 756
		UDP6_INC_STATS_BH(net, UDP_MIB_NOPORTS,
				proto == IPPROTO_UDPLITE);
L
Linus Torvalds 已提交
757

758
		icmpv6_send(skb, ICMPV6_DEST_UNREACH, ICMPV6_PORT_UNREACH, 0);
L
Linus Torvalds 已提交
759 760

		kfree_skb(skb);
S
Stephen Hemminger 已提交
761
		return 0;
L
Linus Torvalds 已提交
762
	}
763

L
Linus Torvalds 已提交
764
	/* deliver */
765

766 767 768 769
	if (sk_rcvqueues_full(sk, skb)) {
		sock_put(sk);
		goto discard;
	}
770
	bh_lock_sock(sk);
H
Hideo Aoki 已提交
771 772
	if (!sock_owned_by_user(sk))
		udpv6_queue_rcv_skb(sk, skb);
Z
Zhu Yi 已提交
773
	else if (sk_add_backlog(sk, skb)) {
Z
Zhu Yi 已提交
774 775 776 777 778
		atomic_inc(&sk->sk_drops);
		bh_unlock_sock(sk);
		sock_put(sk);
		goto discard;
	}
H
Hideo Aoki 已提交
779
	bh_unlock_sock(sk);
L
Linus Torvalds 已提交
780
	sock_put(sk);
S
Stephen Hemminger 已提交
781
	return 0;
L
Linus Torvalds 已提交
782

783
short_packet:
784
	LIMIT_NETDEBUG(KERN_DEBUG "UDP%sv6: short packet: %d/%u\n",
785
		       proto == IPPROTO_UDPLITE ? "-Lite" : "",
786
		       ulen, skb->len);
L
Linus Torvalds 已提交
787 788

discard:
789
	UDP6_INC_STATS_BH(net, UDP_MIB_INERRORS, proto == IPPROTO_UDPLITE);
L
Linus Torvalds 已提交
790
	kfree_skb(skb);
S
Stephen Hemminger 已提交
791
	return 0;
L
Linus Torvalds 已提交
792
}
793

794
static __inline__ int udpv6_rcv(struct sk_buff *skb)
795
{
796
	return __udp6_lib_rcv(skb, &udp_table, IPPROTO_UDP);
797 798
}

L
Linus Torvalds 已提交
799 800 801 802 803 804 805
/*
 * 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);

806 807 808
	if (up->pending == AF_INET)
		udp_flush_pending_frames(sk);
	else if (up->pending) {
L
Linus Torvalds 已提交
809 810 811
		up->len = 0;
		up->pending = 0;
		ip6_flush_pending_frames(sk);
812
	}
L
Linus Torvalds 已提交
813 814
}

815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855
/**
 * 	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
856 857 858 859
/*
 *	Sending
 */

860
static int udp_v6_push_pending_frames(struct sock *sk)
L
Linus Torvalds 已提交
861 862 863
{
	struct sk_buff *skb;
	struct udphdr *uh;
864
	struct udp_sock  *up = udp_sk(sk);
L
Linus Torvalds 已提交
865 866 867
	struct inet_sock *inet = inet_sk(sk);
	struct flowi *fl = &inet->cork.fl;
	int err = 0;
W
Wang Chen 已提交
868
	int is_udplite = IS_UDPLITE(sk);
869
	__wsum csum = 0;
L
Linus Torvalds 已提交
870 871 872 873 874 875 876 877

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

	/*
	 * Create a UDP header
	 */
878
	uh = udp_hdr(skb);
L
Linus Torvalds 已提交
879 880 881 882 883
	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
884
	if (is_udplite)
885
		csum = udplite_csum_outgoing(sk, skb);
886 887 888 889 890
	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
891
		csum = udp_csum_outgoing(sk, skb);
L
Linus Torvalds 已提交
892

893 894 895
	/* 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
896
	if (uh->check == 0)
897
		uh->check = CSUM_MANGLED_0;
L
Linus Torvalds 已提交
898

899
send:
L
Linus Torvalds 已提交
900
	err = ip6_push_pending_frames(sk);
E
Eric Dumazet 已提交
901 902 903 904 905 906 907 908 909
	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
910 911 912 913 914 915
out:
	up->len = 0;
	up->pending = 0;
	return err;
}

916
int udpv6_sendmsg(struct kiocb *iocb, struct sock *sk,
L
Linus Torvalds 已提交
917 918 919 920 921 922 923 924 925 926
		  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;
	struct in6_addr *daddr, *final_p = NULL, final;
	struct ipv6_txoptions *opt = NULL;
	struct ip6_flowlabel *flowlabel = NULL;
H
Herbert Xu 已提交
927
	struct flowi fl;
L
Linus Torvalds 已提交
928 929 930 931
	struct dst_entry *dst;
	int addr_len = msg->msg_namelen;
	int ulen = len;
	int hlimit = -1;
932
	int tclass = -1;
933
	int dontfrag = -1;
L
Linus Torvalds 已提交
934 935
	int corkreq = up->corkflag || msg->msg_flags&MSG_MORE;
	int err;
936
	int connected = 0;
W
Wang Chen 已提交
937
	int is_udplite = IS_UDPLITE(sk);
938
	int (*getfrag)(void *, char *, int, int, int, struct sk_buff *);
L
Linus Torvalds 已提交
939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964

	/* 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;
965
	} else
L
Linus Torvalds 已提交
966 967 968
		daddr = NULL;

	if (daddr) {
B
Brian Haley 已提交
969
		if (ipv6_addr_v4mapped(daddr)) {
L
Linus Torvalds 已提交
970 971
			struct sockaddr_in sin;
			sin.sin_family = AF_INET;
E
Eric Dumazet 已提交
972
			sin.sin_port = sin6 ? sin6->sin6_port : inet->inet_dport;
L
Linus Torvalds 已提交
973 974 975 976 977 978 979 980 981 982 983 984 985 986
			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,
987
	   better check is made in ip6_append_data().
L
Linus Torvalds 已提交
988 989 990
	   */
	if (len > INT_MAX - sizeof(struct udphdr))
		return -EMSGSIZE;
991

L
Linus Torvalds 已提交
992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009
	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
1010
	memset(&fl, 0, sizeof(fl));
L
Linus Torvalds 已提交
1011 1012 1013 1014 1015

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

H
Herbert Xu 已提交
1016
		fl.fl_ip_dport = sin6->sin6_port;
L
Linus Torvalds 已提交
1017 1018 1019
		daddr = &sin6->sin6_addr;

		if (np->sndflow) {
H
Herbert Xu 已提交
1020 1021 1022
			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
1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039
				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
1040
			fl.oif = sin6->sin6_scope_id;
L
Linus Torvalds 已提交
1041 1042 1043 1044
	} else {
		if (sk->sk_state != TCP_ESTABLISHED)
			return -EDESTADDRREQ;

E
Eric Dumazet 已提交
1045
		fl.fl_ip_dport = inet->inet_dport;
L
Linus Torvalds 已提交
1046
		daddr = &np->daddr;
H
Herbert Xu 已提交
1047
		fl.fl6_flowlabel = np->flow_label;
1048
		connected = 1;
L
Linus Torvalds 已提交
1049 1050
	}

H
Herbert Xu 已提交
1051 1052
	if (!fl.oif)
		fl.oif = sk->sk_bound_dev_if;
L
Linus Torvalds 已提交
1053

1054 1055 1056
	if (!fl.oif)
		fl.oif = np->sticky_pktinfo.ipi6_ifindex;

B
Brian Haley 已提交
1057 1058
	fl.mark = sk->sk_mark;

L
Linus Torvalds 已提交
1059 1060 1061 1062 1063
	if (msg->msg_controllen) {
		opt = &opt_space;
		memset(opt, 0, sizeof(struct ipv6_txoptions));
		opt->tot_len = sizeof(*opt);

1064 1065
		err = datagram_send_ctl(sock_net(sk), msg, &fl, opt, &hlimit,
					&tclass, &dontfrag);
L
Linus Torvalds 已提交
1066 1067 1068 1069
		if (err < 0) {
			fl6_sock_release(flowlabel);
			return err;
		}
H
Herbert Xu 已提交
1070 1071
		if ((fl.fl6_flowlabel&IPV6_FLOWLABEL_MASK) && !flowlabel) {
			flowlabel = fl6_sock_lookup(sk, fl.fl6_flowlabel);
L
Linus Torvalds 已提交
1072 1073 1074 1075 1076
			if (flowlabel == NULL)
				return -EINVAL;
		}
		if (!(opt->opt_nflen|opt->opt_flen))
			opt = NULL;
1077
		connected = 0;
L
Linus Torvalds 已提交
1078 1079 1080
	}
	if (opt == NULL)
		opt = np->opt;
1081 1082 1083
	if (flowlabel)
		opt = fl6_merge_options(&opt_space, flowlabel, opt);
	opt = ipv6_fixup_options(&opt_space, opt);
L
Linus Torvalds 已提交
1084

1085
	fl.proto = sk->sk_protocol;
1086 1087 1088 1089
	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
1090 1091
	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
1092
	fl.fl_ip_sport = inet->inet_sport;
1093

L
Linus Torvalds 已提交
1094 1095 1096
	/* merge ip6_build_xmit from ip6_output */
	if (opt && opt->srcrt) {
		struct rt0_hdr *rt0 = (struct rt0_hdr *) opt->srcrt;
H
Herbert Xu 已提交
1097 1098
		ipv6_addr_copy(&final, &fl.fl6_dst);
		ipv6_addr_copy(&fl.fl6_dst, rt0->addr);
L
Linus Torvalds 已提交
1099
		final_p = &final;
1100
		connected = 0;
L
Linus Torvalds 已提交
1101 1102
	}

H
Herbert Xu 已提交
1103 1104
	if (!fl.oif && ipv6_addr_is_multicast(&fl.fl6_dst)) {
		fl.oif = np->mcast_oif;
1105 1106
		connected = 0;
	}
L
Linus Torvalds 已提交
1107

H
Herbert Xu 已提交
1108
	security_sk_classify_flow(sk, &fl);
V
Venkat Yekkirala 已提交
1109

H
Herbert Xu 已提交
1110
	err = ip6_sk_dst_lookup(sk, &dst, &fl);
L
Linus Torvalds 已提交
1111 1112 1113
	if (err)
		goto out;
	if (final_p)
H
Herbert Xu 已提交
1114
		ipv6_addr_copy(&fl.fl6_dst, final_p);
L
Linus Torvalds 已提交
1115

A
Alexey Dobriyan 已提交
1116 1117
	err = __xfrm_lookup(sock_net(sk), &dst, &fl, sk, XFRM_LOOKUP_WAIT);
	if (err < 0) {
1118 1119 1120 1121 1122
		if (err == -EREMOTE)
			err = ip6_dst_blackhole(sk, &dst, &fl);
		if (err < 0)
			goto out;
	}
L
Linus Torvalds 已提交
1123 1124

	if (hlimit < 0) {
H
Herbert Xu 已提交
1125
		if (ipv6_addr_is_multicast(&fl.fl6_dst))
L
Linus Torvalds 已提交
1126 1127 1128 1129
			hlimit = np->mcast_hops;
		else
			hlimit = np->hop_limit;
		if (hlimit < 0)
1130
			hlimit = ip6_dst_hoplimit(dst);
L
Linus Torvalds 已提交
1131 1132
	}

G
Gerrit Renker 已提交
1133
	if (tclass < 0)
1134 1135
		tclass = np->tclass;

1136 1137 1138
	if (dontfrag < 0)
		dontfrag = np->dontfrag;

L
Linus Torvalds 已提交
1139 1140 1141 1142 1143 1144 1145 1146 1147 1148
	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);

1149
		LIMIT_NETDEBUG(KERN_DEBUG "udp cork app bug 2\n");
L
Linus Torvalds 已提交
1150 1151 1152 1153 1154 1155 1156 1157
		err = -EINVAL;
		goto out;
	}

	up->pending = AF_INET6;

do_append_data:
	up->len += ulen;
1158 1159
	getfrag  =  is_udplite ?  udplite_getfrag : ip_generic_getfrag;
	err = ip6_append_data(sk, getfrag, msg->msg_iov, ulen,
H
Herbert Xu 已提交
1160
		sizeof(struct udphdr), hlimit, tclass, opt, &fl,
1161
		(struct rt6_info*)dst,
1162
		corkreq ? msg->msg_flags|MSG_MORE : msg->msg_flags, dontfrag);
L
Linus Torvalds 已提交
1163 1164 1165
	if (err)
		udp_v6_flush_pending_frames(sk);
	else if (!corkreq)
1166
		err = udp_v6_push_pending_frames(sk);
H
Herbert Xu 已提交
1167 1168
	else if (unlikely(skb_queue_empty(&sk->sk_write_queue)))
		up->pending = 0;
L
Linus Torvalds 已提交
1169

1170 1171 1172
	if (dst) {
		if (connected) {
			ip6_dst_store(sk, dst,
H
Herbert Xu 已提交
1173
				      ipv6_addr_equal(&fl.fl6_dst, &np->daddr) ?
1174 1175
				      &np->daddr : NULL,
#ifdef CONFIG_IPV6_SUBTREES
H
Herbert Xu 已提交
1176
				      ipv6_addr_equal(&fl.fl6_src, &np->saddr) ?
1177 1178 1179
				      &np->saddr :
#endif
				      NULL);
1180 1181 1182
		} else {
			dst_release(dst);
		}
1183
		dst = NULL;
1184 1185
	}

L
Linus Torvalds 已提交
1186 1187 1188 1189
	if (err > 0)
		err = np->recverr ? net_xmit_errno(err) : 0;
	release_sock(sk);
out:
1190
	dst_release(dst);
L
Linus Torvalds 已提交
1191
	fl6_sock_release(flowlabel);
1192
	if (!err)
L
Linus Torvalds 已提交
1193
		return len;
1194 1195 1196 1197 1198 1199 1200 1201
	/*
	 * 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)) {
1202 1203
		UDP6_INC_STATS_USER(sock_net(sk),
				UDP_MIB_SNDBUFERRORS, is_udplite);
1204
	}
L
Linus Torvalds 已提交
1205 1206 1207 1208 1209 1210 1211 1212 1213 1214
	return err;

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

1215
void udpv6_destroy_sock(struct sock *sk)
L
Linus Torvalds 已提交
1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226
{
	lock_sock(sk);
	udp_v6_flush_pending_frames(sk);
	release_sock(sk);

	inet6_destroy_sock(sk);
}

/*
 *	Socket option code for UDP
 */
1227
int udpv6_setsockopt(struct sock *sk, int level, int optname,
1228
		     char __user *optval, unsigned int optlen)
1229
{
1230
	if (level == SOL_UDP  ||  level == SOL_UDPLITE)
1231 1232
		return udp_lib_setsockopt(sk, level, optname, optval, optlen,
					  udp_v6_push_pending_frames);
1233
	return ipv6_setsockopt(sk, level, optname, optval, optlen);
1234 1235 1236
}

#ifdef CONFIG_COMPAT
1237
int compat_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 compat_ipv6_setsockopt(sk, level, optname, optval, optlen);
1244 1245 1246
}
#endif

1247 1248
int udpv6_getsockopt(struct sock *sk, int level, int optname,
		     char __user *optval, int __user *optlen)
1249
{
1250
	if (level == SOL_UDP  ||  level == SOL_UDPLITE)
1251
		return udp_lib_getsockopt(sk, level, optname, optval, optlen);
1252
	return ipv6_getsockopt(sk, level, optname, optval, optlen);
1253 1254 1255
}

#ifdef CONFIG_COMPAT
1256 1257
int compat_udpv6_getsockopt(struct sock *sk, int level, int optname,
			    char __user *optval, int __user *optlen)
1258
{
1259
	if (level == SOL_UDP  ||  level == SOL_UDPLITE)
1260
		return udp_lib_getsockopt(sk, level, optname, optval, optlen);
1261
	return compat_ipv6_getsockopt(sk, level, optname, optval, optlen);
1262 1263 1264
}
#endif

1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 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
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;
}

1356
static const struct inet6_protocol udpv6_protocol = {
L
Linus Torvalds 已提交
1357 1358
	.handler	=	udpv6_rcv,
	.err_handler	=	udpv6_err,
1359 1360
	.gso_send_check =	udp6_ufo_send_check,
	.gso_segment	=	udp6_ufo_fragment,
L
Linus Torvalds 已提交
1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375
	.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
1376 1377
	destp = ntohs(inet->inet_dport);
	srcp  = ntohs(inet->inet_sport);
L
Linus Torvalds 已提交
1378
	seq_printf(seq,
1379
		   "%5d: %08X%08X%08X%08X:%04X %08X%08X%08X%08X:%04X "
E
Eric Dumazet 已提交
1380
		   "%02X %08X:%08X %02X:%08lX %08X %5d %8d %lu %d %p %d\n",
L
Linus Torvalds 已提交
1381 1382 1383 1384 1385
		   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,
1386
		   sp->sk_state,
1387 1388
		   sk_wmem_alloc_get(sp),
		   sk_rmem_alloc_get(sp),
L
Linus Torvalds 已提交
1389 1390 1391
		   0, 0L, 0,
		   sock_i_uid(sp), 0,
		   sock_i_ino(sp),
E
Eric Dumazet 已提交
1392 1393
		   atomic_read(&sp->sk_refcnt), sp,
		   atomic_read(&sp->sk_drops));
L
Linus Torvalds 已提交
1394 1395
}

1396
int udp6_seq_show(struct seq_file *seq, void *v)
L
Linus Torvalds 已提交
1397 1398 1399 1400 1401 1402 1403
{
	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
1404
			   "   uid  timeout inode ref pointer drops\n");
L
Linus Torvalds 已提交
1405 1406 1407 1408 1409 1410 1411 1412
	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,
1413
	.udp_table	= &udp_table,
1414 1415 1416
	.seq_fops	= {
		.owner	=	THIS_MODULE,
	},
1417 1418 1419
	.seq_ops	= {
		.show		= udp6_seq_show,
	},
L
Linus Torvalds 已提交
1420 1421
};

1422
int __net_init udp6_proc_init(struct net *net)
L
Linus Torvalds 已提交
1423
{
1424
	return udp_proc_register(net, &udp6_seq_afinfo);
L
Linus Torvalds 已提交
1425 1426
}

1427 1428
void udp6_proc_exit(struct net *net) {
	udp_proc_unregister(net, &udp6_seq_afinfo);
L
Linus Torvalds 已提交
1429 1430 1431 1432 1433 1434
}
#endif /* CONFIG_PROC_FS */

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

struct proto udpv6_prot = {
1435 1436
	.name		   = "UDPv6",
	.owner		   = THIS_MODULE,
1437
	.close		   = udp_lib_close,
1438 1439 1440 1441 1442 1443 1444 1445 1446
	.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,
1447 1448
	.hash		   = udp_lib_hash,
	.unhash		   = udp_lib_unhash,
1449
	.get_port	   = udp_v6_get_port,
H
Hideo Aoki 已提交
1450 1451 1452 1453
	.memory_allocated  = &udp_memory_allocated,
	.sysctl_mem	   = sysctl_udp_mem,
	.sysctl_wmem	   = &sysctl_udp_wmem_min,
	.sysctl_rmem	   = &sysctl_udp_rmem_min,
1454
	.obj_size	   = sizeof(struct udp6_sock),
1455
	.slab_flags	   = SLAB_DESTROY_BY_RCU,
1456
	.h.udp_table	   = &udp_table,
1457
#ifdef CONFIG_COMPAT
1458 1459
	.compat_setsockopt = compat_udpv6_setsockopt,
	.compat_getsockopt = compat_udpv6_getsockopt,
1460
#endif
L
Linus Torvalds 已提交
1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472
};

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,
};


1473
int __init udpv6_init(void)
L
Linus Torvalds 已提交
1474
{
1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491
	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;
}

1492
void udpv6_exit(void)
1493 1494 1495
{
	inet6_unregister_protosw(&udpv6_protosw);
	inet6_del_protocol(&udpv6_protocol, IPPROTO_UDP);
L
Linus Torvalds 已提交
1496
}