ip6_output.c 36.4 KB
Newer Older
L
Linus Torvalds 已提交
1 2
/*
 *	IPv6 output functions
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 35 36 37 38 39 40
 *
 *	$Id: ip6_output.c,v 1.34 2002/02/01 22:01:04 davem Exp $
 *
 *	Based on linux/net/ipv4/ip_output.c
 *
 *	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.
 *
 *	Changes:
 *	A.N.Kuznetsov	:	airthmetics in fragmentation.
 *				extension headers are implemented.
 *				route changes now work.
 *				ip6_forward does not confuse sniffers.
 *				etc.
 *
 *      H. von Brand    :       Added missing #include <linux/string.h>
 *	Imran Patel	: 	frag id should be in NBO
 *      Kazunori MIYAZAWA @USAGI
 *			:       add ip6_append_data and related functions
 *				for datagram xmit
 */

#include <linux/errno.h>
#include <linux/types.h>
#include <linux/string.h>
#include <linux/socket.h>
#include <linux/net.h>
#include <linux/netdevice.h>
#include <linux/if_arp.h>
#include <linux/in6.h>
#include <linux/tcp.h>
#include <linux/route.h>
41
#include <linux/module.h>
L
Linus Torvalds 已提交
42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72

#include <linux/netfilter.h>
#include <linux/netfilter_ipv6.h>

#include <net/sock.h>
#include <net/snmp.h>

#include <net/ipv6.h>
#include <net/ndisc.h>
#include <net/protocol.h>
#include <net/ip6_route.h>
#include <net/addrconf.h>
#include <net/rawv6.h>
#include <net/icmp.h>
#include <net/xfrm.h>
#include <net/checksum.h>

static int ip6_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *));

static __inline__ void ipv6_select_ident(struct sk_buff *skb, struct frag_hdr *fhdr)
{
	static u32 ipv6_fragmentation_id = 1;
	static DEFINE_SPINLOCK(ip6_id_lock);

	spin_lock_bh(&ip6_id_lock);
	fhdr->identification = htonl(ipv6_fragmentation_id);
	if (++ipv6_fragmentation_id == 0)
		ipv6_fragmentation_id = 1;
	spin_unlock_bh(&ip6_id_lock);
}

P
Patrick McHardy 已提交
73
static int ip6_output_finish(struct sk_buff *skb)
L
Linus Torvalds 已提交
74 75 76
{
	struct dst_entry *dst = skb->dst;

77 78 79
	if (dst->hh)
		return neigh_hh_output(dst->hh, skb);
	else if (dst->neighbour)
L
Linus Torvalds 已提交
80 81
		return dst->neighbour->output(skb);

82
	IP6_INC_STATS_BH(ip6_dst_idev(dst), IPSTATS_MIB_OUTNOROUTES);
L
Linus Torvalds 已提交
83 84 85 86 87 88 89 90
	kfree_skb(skb);
	return -EINVAL;

}

/* dev_loopback_xmit for use with netfilter. */
static int ip6_dev_loopback_xmit(struct sk_buff *newskb)
{
91
	skb_reset_mac_header(newskb);
92
	__skb_pull(newskb, skb_network_offset(newskb));
L
Linus Torvalds 已提交
93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109
	newskb->pkt_type = PACKET_LOOPBACK;
	newskb->ip_summed = CHECKSUM_UNNECESSARY;
	BUG_TRAP(newskb->dst);

	netif_rx(newskb);
	return 0;
}


static int ip6_output2(struct sk_buff *skb)
{
	struct dst_entry *dst = skb->dst;
	struct net_device *dev = dst->dev;

	skb->protocol = htons(ETH_P_IPV6);
	skb->dev = dev;

110
	if (ipv6_addr_is_multicast(&ipv6_hdr(skb)->daddr)) {
L
Linus Torvalds 已提交
111
		struct ipv6_pinfo* np = skb->sk ? inet6_sk(skb->sk) : NULL;
112
		struct inet6_dev *idev = ip6_dst_idev(skb->dst);
L
Linus Torvalds 已提交
113 114

		if (!(dev->flags & IFF_LOOPBACK) && (!np || np->mc_loop) &&
115 116
		    ipv6_chk_mcast_addr(dev, &ipv6_hdr(skb)->daddr,
					&ipv6_hdr(skb)->saddr)) {
L
Linus Torvalds 已提交
117 118 119 120 121 122 123 124 125 126
			struct sk_buff *newskb = skb_clone(skb, GFP_ATOMIC);

			/* Do not check for IFF_ALLMULTI; multicast routing
			   is not supported in any case.
			 */
			if (newskb)
				NF_HOOK(PF_INET6, NF_IP6_POST_ROUTING, newskb, NULL,
					newskb->dev,
					ip6_dev_loopback_xmit);

127
			if (ipv6_hdr(skb)->hop_limit == 0) {
128
				IP6_INC_STATS(idev, IPSTATS_MIB_OUTDISCARDS);
L
Linus Torvalds 已提交
129 130 131 132 133
				kfree_skb(skb);
				return 0;
			}
		}

134
		IP6_INC_STATS(idev, IPSTATS_MIB_OUTMCASTPKTS);
L
Linus Torvalds 已提交
135 136 137 138 139
	}

	return NF_HOOK(PF_INET6, NF_IP6_POST_ROUTING, skb,NULL, skb->dev,ip6_output_finish);
}

J
John Heffner 已提交
140 141 142 143 144 145 146 147
static inline int ip6_skb_dst_mtu(struct sk_buff *skb)
{
	struct ipv6_pinfo *np = skb->sk ? inet6_sk(skb->sk) : NULL;

	return (np && np->pmtudisc == IPV6_PMTUDISC_PROBE) ?
	       skb->dst->dev->mtu : dst_mtu(skb->dst);
}

L
Linus Torvalds 已提交
148 149
int ip6_output(struct sk_buff *skb)
{
J
John Heffner 已提交
150
	if ((skb->len > ip6_skb_dst_mtu(skb) && !skb_is_gso(skb)) ||
151
				dst_allfrag(skb->dst))
L
Linus Torvalds 已提交
152 153 154 155 156 157 158 159 160 161 162 163
		return ip6_fragment(skb, ip6_output2);
	else
		return ip6_output2(skb);
}

/*
 *	xmit an sk_buff (used by TCP)
 */

int ip6_xmit(struct sock *sk, struct sk_buff *skb, struct flowi *fl,
	     struct ipv6_txoptions *opt, int ipfragok)
{
164
	struct ipv6_pinfo *np = inet6_sk(sk);
L
Linus Torvalds 已提交
165 166 167 168 169
	struct in6_addr *first_hop = &fl->fl6_dst;
	struct dst_entry *dst = skb->dst;
	struct ipv6hdr *hdr;
	u8  proto = fl->proto;
	int seg_len = skb->len;
170
	int hlimit, tclass;
L
Linus Torvalds 已提交
171 172 173
	u32 mtu;

	if (opt) {
174
		unsigned int head_room;
L
Linus Torvalds 已提交
175 176 177 178 179 180 181 182 183 184

		/* First: exthdrs may take lots of space (~8K for now)
		   MAX_HEADER is not enough.
		 */
		head_room = opt->opt_nflen + opt->opt_flen;
		seg_len += head_room;
		head_room += sizeof(struct ipv6hdr) + LL_RESERVED_SPACE(dst->dev);

		if (skb_headroom(skb) < head_room) {
			struct sk_buff *skb2 = skb_realloc_headroom(skb, head_room);
185 186 187 188
			if (skb2 == NULL) {
				IP6_INC_STATS(ip6_dst_idev(skb->dst),
					      IPSTATS_MIB_OUTDISCARDS);
				kfree_skb(skb);
L
Linus Torvalds 已提交
189 190
				return -ENOBUFS;
			}
191 192
			kfree_skb(skb);
			skb = skb2;
L
Linus Torvalds 已提交
193 194 195 196 197 198 199 200 201
			if (sk)
				skb_set_owner_w(skb, sk);
		}
		if (opt->opt_flen)
			ipv6_push_frag_opts(skb, opt, &proto);
		if (opt->opt_nflen)
			ipv6_push_nfrag_opts(skb, opt, &proto, &first_hop);
	}

202 203
	skb_push(skb, sizeof(struct ipv6hdr));
	skb_reset_network_header(skb);
204
	hdr = ipv6_hdr(skb);
L
Linus Torvalds 已提交
205 206 207 208 209 210 211 212 213 214 215 216 217

	/*
	 *	Fill in the IPv6 header
	 */

	hlimit = -1;
	if (np)
		hlimit = np->hop_limit;
	if (hlimit < 0)
		hlimit = dst_metric(dst, RTAX_HOPLIMIT);
	if (hlimit < 0)
		hlimit = ipv6_get_hoplimit(dst->dev);

218 219 220 221 222 223
	tclass = -1;
	if (np)
		tclass = np->tclass;
	if (tclass < 0)
		tclass = 0;

A
Al Viro 已提交
224
	*(__be32 *)hdr = htonl(0x60000000 | (tclass << 20)) | fl->fl6_flowlabel;
225

L
Linus Torvalds 已提交
226 227 228 229 230 231 232
	hdr->payload_len = htons(seg_len);
	hdr->nexthdr = proto;
	hdr->hop_limit = hlimit;

	ipv6_addr_copy(&hdr->saddr, &fl->fl6_src);
	ipv6_addr_copy(&hdr->daddr, first_hop);

233 234
	skb->priority = sk->sk_priority;

L
Linus Torvalds 已提交
235
	mtu = dst_mtu(dst);
H
Herbert Xu 已提交
236
	if ((skb->len <= mtu) || ipfragok || skb_is_gso(skb)) {
237 238
		IP6_INC_STATS(ip6_dst_idev(skb->dst),
			      IPSTATS_MIB_OUTREQUESTS);
239 240
		return NF_HOOK(PF_INET6, NF_IP6_LOCAL_OUT, skb, NULL, dst->dev,
				dst_output);
L
Linus Torvalds 已提交
241 242 243 244 245 246
	}

	if (net_ratelimit())
		printk(KERN_DEBUG "IPv6: sending pkt_too_big to self\n");
	skb->dev = dst->dev;
	icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu, skb->dev);
247
	IP6_INC_STATS(ip6_dst_idev(skb->dst), IPSTATS_MIB_FRAGFAILS);
L
Linus Torvalds 已提交
248 249 250 251
	kfree_skb(skb);
	return -EMSGSIZE;
}

252 253
EXPORT_SYMBOL(ip6_xmit);

L
Linus Torvalds 已提交
254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273
/*
 *	To avoid extra problems ND packets are send through this
 *	routine. It's code duplication but I really want to avoid
 *	extra checks since ipv6_build_header is used by TCP (which
 *	is for us performance critical)
 */

int ip6_nd_hdr(struct sock *sk, struct sk_buff *skb, struct net_device *dev,
	       struct in6_addr *saddr, struct in6_addr *daddr,
	       int proto, int len)
{
	struct ipv6_pinfo *np = inet6_sk(sk);
	struct ipv6hdr *hdr;
	int totlen;

	skb->protocol = htons(ETH_P_IPV6);
	skb->dev = dev;

	totlen = len + sizeof(struct ipv6hdr);

274 275
	skb_reset_network_header(skb);
	skb_put(skb, sizeof(struct ipv6hdr));
276
	hdr = ipv6_hdr(skb);
L
Linus Torvalds 已提交
277

A
Al Viro 已提交
278
	*(__be32*)hdr = htonl(0x60000000);
L
Linus Torvalds 已提交
279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297

	hdr->payload_len = htons(len);
	hdr->nexthdr = proto;
	hdr->hop_limit = np->hop_limit;

	ipv6_addr_copy(&hdr->saddr, saddr);
	ipv6_addr_copy(&hdr->daddr, daddr);

	return 0;
}

static int ip6_call_ra_chain(struct sk_buff *skb, int sel)
{
	struct ip6_ra_chain *ra;
	struct sock *last = NULL;

	read_lock(&ip6_ra_lock);
	for (ra = ip6_ra_chain; ra; ra = ra->next) {
		struct sock *sk = ra->sk;
298 299 300
		if (sk && ra->sel == sel &&
		    (!sk->sk_bound_dev_if ||
		     sk->sk_bound_dev_if == skb->dev->ifindex)) {
L
Linus Torvalds 已提交
301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318
			if (last) {
				struct sk_buff *skb2 = skb_clone(skb, GFP_ATOMIC);
				if (skb2)
					rawv6_rcv(last, skb2);
			}
			last = sk;
		}
	}

	if (last) {
		rawv6_rcv(last, skb);
		read_unlock(&ip6_ra_lock);
		return 1;
	}
	read_unlock(&ip6_ra_lock);
	return 0;
}

319 320
static int ip6_forward_proxy_check(struct sk_buff *skb)
{
321
	struct ipv6hdr *hdr = ipv6_hdr(skb);
322 323 324 325 326 327 328 329 330 331 332 333 334
	u8 nexthdr = hdr->nexthdr;
	int offset;

	if (ipv6_ext_hdr(nexthdr)) {
		offset = ipv6_skip_exthdr(skb, sizeof(*hdr), &nexthdr);
		if (offset < 0)
			return 0;
	} else
		offset = sizeof(struct ipv6hdr);

	if (nexthdr == IPPROTO_ICMPV6) {
		struct icmp6hdr *icmp6;

335 336
		if (!pskb_may_pull(skb, (skb_network_header(skb) +
					 offset + 1 - skb->data)))
337 338
			return 0;

339
		icmp6 = (struct icmp6hdr *)(skb_network_header(skb) + offset);
340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356

		switch (icmp6->icmp6_type) {
		case NDISC_ROUTER_SOLICITATION:
		case NDISC_ROUTER_ADVERTISEMENT:
		case NDISC_NEIGHBOUR_SOLICITATION:
		case NDISC_NEIGHBOUR_ADVERTISEMENT:
		case NDISC_REDIRECT:
			/* For reaction involving unicast neighbor discovery
			 * message destined to the proxied address, pass it to
			 * input function.
			 */
			return 1;
		default:
			break;
		}
	}

357 358 359 360 361 362 363 364 365 366
	/*
	 * The proxying router can't forward traffic sent to a link-local
	 * address, so signal the sender and discard the packet. This
	 * behavior is clarified by the MIPv6 specification.
	 */
	if (ipv6_addr_type(&hdr->daddr) & IPV6_ADDR_LINKLOCAL) {
		dst_link_failure(skb);
		return -1;
	}

367 368 369
	return 0;
}

L
Linus Torvalds 已提交
370 371 372 373 374 375 376 377
static inline int ip6_forward_finish(struct sk_buff *skb)
{
	return dst_output(skb);
}

int ip6_forward(struct sk_buff *skb)
{
	struct dst_entry *dst = skb->dst;
378
	struct ipv6hdr *hdr = ipv6_hdr(skb);
L
Linus Torvalds 已提交
379
	struct inet6_skb_parm *opt = IP6CB(skb);
380

L
Linus Torvalds 已提交
381 382 383 384
	if (ipv6_devconf.forwarding == 0)
		goto error;

	if (!xfrm6_policy_check(NULL, XFRM_POLICY_FWD, skb)) {
385
		IP6_INC_STATS(ip6_dst_idev(dst), IPSTATS_MIB_INDISCARDS);
L
Linus Torvalds 已提交
386 387 388
		goto drop;
	}

389
	skb_forward_csum(skb);
L
Linus Torvalds 已提交
390 391 392 393 394 395 396 397 398 399 400 401 402 403 404

	/*
	 *	We DO NOT make any processing on
	 *	RA packets, pushing them to user level AS IS
	 *	without ane WARRANTY that application will be able
	 *	to interpret them. The reason is that we
	 *	cannot make anything clever here.
	 *
	 *	We are not end-node, so that if packet contains
	 *	AH/ESP, we cannot make anything.
	 *	Defragmentation also would be mistake, RA packets
	 *	cannot be fragmented, because there is no warranty
	 *	that different fragments will go along one path. --ANK
	 */
	if (opt->ra) {
405
		u8 *ptr = skb_network_header(skb) + opt->ra;
L
Linus Torvalds 已提交
406 407 408 409 410 411 412 413 414 415 416 417
		if (ip6_call_ra_chain(skb, (ptr[2]<<8) + ptr[3]))
			return 0;
	}

	/*
	 *	check and decrement ttl
	 */
	if (hdr->hop_limit <= 1) {
		/* Force OUTPUT device used as source address */
		skb->dev = dst->dev;
		icmpv6_send(skb, ICMPV6_TIME_EXCEED, ICMPV6_EXC_HOPLIMIT,
			    0, skb->dev);
418
		IP6_INC_STATS_BH(ip6_dst_idev(dst), IPSTATS_MIB_INHDRERRORS);
L
Linus Torvalds 已提交
419 420 421 422 423

		kfree_skb(skb);
		return -ETIMEDOUT;
	}

424 425 426
	/* XXX: idev->cnf.proxy_ndp? */
	if (ipv6_devconf.proxy_ndp &&
	    pneigh_lookup(&nd_tbl, &hdr->daddr, skb->dev, 0)) {
427 428
		int proxied = ip6_forward_proxy_check(skb);
		if (proxied > 0)
429
			return ip6_input(skb);
430
		else if (proxied < 0) {
431
			IP6_INC_STATS(ip6_dst_idev(dst), IPSTATS_MIB_INDISCARDS);
432 433
			goto drop;
		}
434 435
	}

L
Linus Torvalds 已提交
436
	if (!xfrm6_route_forward(skb)) {
437
		IP6_INC_STATS(ip6_dst_idev(dst), IPSTATS_MIB_INDISCARDS);
L
Linus Torvalds 已提交
438 439 440 441 442 443
		goto drop;
	}
	dst = skb->dst;

	/* IPv6 specs say nothing about it, but it is clear that we cannot
	   send redirects to source routed frames.
444
	   We don't send redirects to frames decapsulated from IPsec.
L
Linus Torvalds 已提交
445
	 */
446 447
	if (skb->dev == dst->dev && dst->neighbour && opt->srcrt == 0 &&
	    !skb->sp) {
L
Linus Torvalds 已提交
448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467
		struct in6_addr *target = NULL;
		struct rt6_info *rt;
		struct neighbour *n = dst->neighbour;

		/*
		 *	incoming and outgoing devices are the same
		 *	send a redirect.
		 */

		rt = (struct rt6_info *) dst;
		if ((rt->rt6i_flags & RTF_GATEWAY))
			target = (struct in6_addr*)&n->primary_key;
		else
			target = &hdr->daddr;

		/* Limit redirects both by destination (here)
		   and by source (inside ndisc_send_redirect)
		 */
		if (xrlim_allow(dst, 1*HZ))
			ndisc_send_redirect(skb, n, target);
468 469 470
	} else {
		int addrtype = ipv6_addr_type(&hdr->saddr);

L
Linus Torvalds 已提交
471
		/* This check is security critical. */
472 473 474 475 476 477 478
		if (addrtype & (IPV6_ADDR_MULTICAST|IPV6_ADDR_LOOPBACK))
			goto error;
		if (addrtype & IPV6_ADDR_LINKLOCAL) {
			icmpv6_send(skb, ICMPV6_DEST_UNREACH,
				ICMPV6_NOT_NEIGHBOUR, 0, skb->dev);
			goto error;
		}
L
Linus Torvalds 已提交
479 480 481 482 483 484
	}

	if (skb->len > dst_mtu(dst)) {
		/* Again, force OUTPUT device used as source address */
		skb->dev = dst->dev;
		icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, dst_mtu(dst), skb->dev);
485 486
		IP6_INC_STATS_BH(ip6_dst_idev(dst), IPSTATS_MIB_INTOOBIGERRORS);
		IP6_INC_STATS_BH(ip6_dst_idev(dst), IPSTATS_MIB_FRAGFAILS);
L
Linus Torvalds 已提交
487 488 489 490 491
		kfree_skb(skb);
		return -EMSGSIZE;
	}

	if (skb_cow(skb, dst->dev->hard_header_len)) {
492
		IP6_INC_STATS(ip6_dst_idev(dst), IPSTATS_MIB_OUTDISCARDS);
L
Linus Torvalds 已提交
493 494 495
		goto drop;
	}

496
	hdr = ipv6_hdr(skb);
L
Linus Torvalds 已提交
497 498

	/* Mangling hops number delayed to point after skb COW */
499

L
Linus Torvalds 已提交
500 501
	hdr->hop_limit--;

502
	IP6_INC_STATS_BH(ip6_dst_idev(dst), IPSTATS_MIB_OUTFORWDATAGRAMS);
L
Linus Torvalds 已提交
503 504 505
	return NF_HOOK(PF_INET6,NF_IP6_FORWARD, skb, skb->dev, dst->dev, ip6_forward_finish);

error:
506
	IP6_INC_STATS_BH(ip6_dst_idev(dst), IPSTATS_MIB_INADDRERRORS);
L
Linus Torvalds 已提交
507 508 509 510 511 512 513 514 515 516 517 518 519
drop:
	kfree_skb(skb);
	return -EINVAL;
}

static void ip6_copy_metadata(struct sk_buff *to, struct sk_buff *from)
{
	to->pkt_type = from->pkt_type;
	to->priority = from->priority;
	to->protocol = from->protocol;
	dst_release(to->dst);
	to->dst = dst_clone(from->dst);
	to->dev = from->dev;
T
Thomas Graf 已提交
520
	to->mark = from->mark;
L
Linus Torvalds 已提交
521 522 523 524

#ifdef CONFIG_NET_SCHED
	to->tc_index = from->tc_index;
#endif
525
	nf_copy(to, from);
526 527 528 529
#if defined(CONFIG_NETFILTER_XT_TARGET_TRACE) || \
    defined(CONFIG_NETFILTER_XT_TARGET_TRACE_MODULE)
	to->nf_trace = from->nf_trace;
#endif
530
	skb_copy_secmark(to, from);
L
Linus Torvalds 已提交
531 532 533 534 535
}

int ip6_find_1stfragopt(struct sk_buff *skb, u8 **nexthdr)
{
	u16 offset = sizeof(struct ipv6hdr);
536 537
	struct ipv6_opt_hdr *exthdr =
				(struct ipv6_opt_hdr *)(ipv6_hdr(skb) + 1);
538
	unsigned int packet_len = skb->tail - skb->network_header;
L
Linus Torvalds 已提交
539
	int found_rhdr = 0;
540
	*nexthdr = &ipv6_hdr(skb)->nexthdr;
L
Linus Torvalds 已提交
541 542 543 544 545 546

	while (offset + 1 <= packet_len) {

		switch (**nexthdr) {

		case NEXTHDR_HOP:
547
			break;
L
Linus Torvalds 已提交
548
		case NEXTHDR_ROUTING:
549 550
			found_rhdr = 1;
			break;
L
Linus Torvalds 已提交
551
		case NEXTHDR_DEST:
552
#if defined(CONFIG_IPV6_MIP6) || defined(CONFIG_IPV6_MIP6_MODULE)
553 554 555 556 557
			if (ipv6_find_tlv(skb, offset, IPV6_TLV_HAO) >= 0)
				break;
#endif
			if (found_rhdr)
				return offset;
L
Linus Torvalds 已提交
558 559 560 561
			break;
		default :
			return offset;
		}
562 563 564

		offset += ipv6_optlen(exthdr);
		*nexthdr = &exthdr->nexthdr;
565 566
		exthdr = (struct ipv6_opt_hdr *)(skb_network_header(skb) +
						 offset);
L
Linus Torvalds 已提交
567 568 569 570
	}

	return offset;
}
571
EXPORT_SYMBOL_GPL(ip6_find_1stfragopt);
L
Linus Torvalds 已提交
572 573 574 575 576 577

static int ip6_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *))
{
	struct net_device *dev;
	struct sk_buff *frag;
	struct rt6_info *rt = (struct rt6_info*)skb->dst;
578
	struct ipv6_pinfo *np = skb->sk ? inet6_sk(skb->sk) : NULL;
L
Linus Torvalds 已提交
579 580 581
	struct ipv6hdr *tmp_hdr;
	struct frag_hdr *fh;
	unsigned int mtu, hlen, left, len;
A
Al Viro 已提交
582
	__be32 frag_id = 0;
L
Linus Torvalds 已提交
583 584 585 586 587 588 589
	int ptr, offset = 0, err=0;
	u8 *prevhdr, nexthdr = 0;

	dev = rt->u.dst.dev;
	hlen = ip6_find_1stfragopt(skb, &prevhdr);
	nexthdr = *prevhdr;

J
John Heffner 已提交
590
	mtu = ip6_skb_dst_mtu(skb);
591 592 593 594 595 596 597 598 599 600 601 602 603

	/* We must not fragment if the socket is set to force MTU discovery
	 * or if the skb it not generated by a local socket.  (This last
	 * check should be redundant, but it's free.)
	 */
	if (!np || np->pmtudisc >= IPV6_PMTUDISC_DO) {
		skb->dev = skb->dst->dev;
		icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu, skb->dev);
		IP6_INC_STATS(ip6_dst_idev(skb->dst), IPSTATS_MIB_FRAGFAILS);
		kfree_skb(skb);
		return -EMSGSIZE;
	}

604 605 606 607 608
	if (np && np->frag_size < mtu) {
		if (np->frag_size)
			mtu = np->frag_size;
	}
	mtu -= hlen + sizeof(struct frag_hdr);
L
Linus Torvalds 已提交
609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627

	if (skb_shinfo(skb)->frag_list) {
		int first_len = skb_pagelen(skb);

		if (first_len - hlen > mtu ||
		    ((first_len - hlen) & 7) ||
		    skb_cloned(skb))
			goto slow_path;

		for (frag = skb_shinfo(skb)->frag_list; frag; frag = frag->next) {
			/* Correct geometry. */
			if (frag->len > mtu ||
			    ((frag->len & 7) && frag->next) ||
			    skb_headroom(frag) < hlen)
			    goto slow_path;

			/* Partially cloned skb? */
			if (skb_shared(frag))
				goto slow_path;
628 629 630 631 632 633 634 635

			BUG_ON(frag->sk);
			if (skb->sk) {
				sock_hold(skb->sk);
				frag->sk = skb->sk;
				frag->destructor = sock_wfree;
				skb->truesize -= frag->truesize;
			}
L
Linus Torvalds 已提交
636 637 638 639 640 641 642 643
		}

		err = 0;
		offset = 0;
		frag = skb_shinfo(skb)->frag_list;
		skb_shinfo(skb)->frag_list = NULL;
		/* BUILD HEADER */

Y
YOSHIFUJI Hideaki 已提交
644
		*prevhdr = NEXTHDR_FRAGMENT;
645
		tmp_hdr = kmemdup(skb_network_header(skb), hlen, GFP_ATOMIC);
L
Linus Torvalds 已提交
646
		if (!tmp_hdr) {
647
			IP6_INC_STATS(ip6_dst_idev(skb->dst), IPSTATS_MIB_FRAGFAILS);
L
Linus Torvalds 已提交
648 649 650 651 652
			return -ENOMEM;
		}

		__skb_pull(skb, hlen);
		fh = (struct frag_hdr*)__skb_push(skb, sizeof(struct frag_hdr));
653 654
		__skb_push(skb, hlen);
		skb_reset_network_header(skb);
655
		memcpy(skb_network_header(skb), tmp_hdr, hlen);
L
Linus Torvalds 已提交
656 657 658 659 660 661 662 663 664 665

		ipv6_select_ident(skb, fh);
		fh->nexthdr = nexthdr;
		fh->reserved = 0;
		fh->frag_off = htons(IP6_MF);
		frag_id = fh->identification;

		first_len = skb_pagelen(skb);
		skb->data_len = first_len - skb_headlen(skb);
		skb->len = first_len;
666 667
		ipv6_hdr(skb)->payload_len = htons(first_len -
						   sizeof(struct ipv6hdr));
668 669

		dst_hold(&rt->u.dst);
L
Linus Torvalds 已提交
670 671 672 673 674 675

		for (;;) {
			/* Prepare header of the next frame,
			 * before previous one went down. */
			if (frag) {
				frag->ip_summed = CHECKSUM_NONE;
676
				skb_reset_transport_header(frag);
L
Linus Torvalds 已提交
677
				fh = (struct frag_hdr*)__skb_push(frag, sizeof(struct frag_hdr));
678 679
				__skb_push(frag, hlen);
				skb_reset_network_header(frag);
680 681
				memcpy(skb_network_header(frag), tmp_hdr,
				       hlen);
L
Linus Torvalds 已提交
682 683 684 685 686 687 688
				offset += skb->len - hlen - sizeof(struct frag_hdr);
				fh->nexthdr = nexthdr;
				fh->reserved = 0;
				fh->frag_off = htons(offset);
				if (frag->next != NULL)
					fh->frag_off |= htons(IP6_MF);
				fh->identification = frag_id;
689 690 691
				ipv6_hdr(frag)->payload_len =
						htons(frag->len -
						      sizeof(struct ipv6hdr));
L
Linus Torvalds 已提交
692 693
				ip6_copy_metadata(frag, skb);
			}
694

L
Linus Torvalds 已提交
695
			err = output(skb);
696
			if(!err)
697
				IP6_INC_STATS(ip6_dst_idev(&rt->u.dst), IPSTATS_MIB_FRAGCREATES);
698

L
Linus Torvalds 已提交
699 700 701 702 703 704 705 706
			if (err || !frag)
				break;

			skb = frag;
			frag = skb->next;
			skb->next = NULL;
		}

J
Jesper Juhl 已提交
707
		kfree(tmp_hdr);
L
Linus Torvalds 已提交
708 709

		if (err == 0) {
710 711
			IP6_INC_STATS(ip6_dst_idev(&rt->u.dst), IPSTATS_MIB_FRAGOKS);
			dst_release(&rt->u.dst);
L
Linus Torvalds 已提交
712 713 714 715 716 717 718 719 720
			return 0;
		}

		while (frag) {
			skb = frag->next;
			kfree_skb(frag);
			frag = skb;
		}

721 722
		IP6_INC_STATS(ip6_dst_idev(&rt->u.dst), IPSTATS_MIB_FRAGFAILS);
		dst_release(&rt->u.dst);
L
Linus Torvalds 已提交
723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753
		return err;
	}

slow_path:
	left = skb->len - hlen;		/* Space per frame */
	ptr = hlen;			/* Where to start from */

	/*
	 *	Fragment the datagram.
	 */

	*prevhdr = NEXTHDR_FRAGMENT;

	/*
	 *	Keep copying data until we run out.
	 */
	while(left > 0)	{
		len = left;
		/* IF: it doesn't fit, use 'mtu' - the data space left */
		if (len > mtu)
			len = mtu;
		/* IF: we are not sending upto and including the packet end
		   then align the next start on an eight byte boundary */
		if (len < left)	{
			len &= ~7;
		}
		/*
		 *	Allocate buffer.
		 */

		if ((frag = alloc_skb(len+hlen+sizeof(struct frag_hdr)+LL_RESERVED_SPACE(rt->u.dst.dev), GFP_ATOMIC)) == NULL) {
754
			NETDEBUG(KERN_INFO "IPv6: frag: no memory for new fragment!\n");
755 756
			IP6_INC_STATS(ip6_dst_idev(skb->dst),
				      IPSTATS_MIB_FRAGFAILS);
L
Linus Torvalds 已提交
757 758 759 760 761 762 763 764 765 766 767
			err = -ENOMEM;
			goto fail;
		}

		/*
		 *	Set up data on packet
		 */

		ip6_copy_metadata(frag, skb);
		skb_reserve(frag, LL_RESERVED_SPACE(rt->u.dst.dev));
		skb_put(frag, len + hlen + sizeof(struct frag_hdr));
768
		skb_reset_network_header(frag);
769
		fh = (struct frag_hdr *)(skb_network_header(frag) + hlen);
770 771
		frag->transport_header = (frag->network_header + hlen +
					  sizeof(struct frag_hdr));
L
Linus Torvalds 已提交
772 773 774 775 776 777 778 779 780 781 782

		/*
		 *	Charge the memory for the fragment to any owner
		 *	it might possess
		 */
		if (skb->sk)
			skb_set_owner_w(frag, skb->sk);

		/*
		 *	Copy the packet header into the new buffer.
		 */
783
		skb_copy_from_linear_data(skb, skb_network_header(frag), hlen);
L
Linus Torvalds 已提交
784 785 786 787 788 789

		/*
		 *	Build fragment header.
		 */
		fh->nexthdr = nexthdr;
		fh->reserved = 0;
790
		if (!frag_id) {
L
Linus Torvalds 已提交
791 792 793 794 795 796 797 798
			ipv6_select_ident(skb, fh);
			frag_id = fh->identification;
		} else
			fh->identification = frag_id;

		/*
		 *	Copy a block of the IP datagram.
		 */
799
		if (skb_copy_bits(skb, ptr, skb_transport_header(frag), len))
L
Linus Torvalds 已提交
800 801 802 803 804 805
			BUG();
		left -= len;

		fh->frag_off = htons(offset);
		if (left > 0)
			fh->frag_off |= htons(IP6_MF);
806 807
		ipv6_hdr(frag)->payload_len = htons(frag->len -
						    sizeof(struct ipv6hdr));
L
Linus Torvalds 已提交
808 809 810 811 812 813 814 815 816 817

		ptr += len;
		offset += len;

		/*
		 *	Put this fragment into the sending queue.
		 */
		err = output(frag);
		if (err)
			goto fail;
818

819
		IP6_INC_STATS(ip6_dst_idev(skb->dst), IPSTATS_MIB_FRAGCREATES);
L
Linus Torvalds 已提交
820
	}
821 822
	IP6_INC_STATS(ip6_dst_idev(skb->dst),
		      IPSTATS_MIB_FRAGOKS);
L
Linus Torvalds 已提交
823 824 825 826
	kfree_skb(skb);
	return err;

fail:
827 828
	IP6_INC_STATS(ip6_dst_idev(skb->dst),
		      IPSTATS_MIB_FRAGFAILS);
829
	kfree_skb(skb);
L
Linus Torvalds 已提交
830 831 832
	return err;
}

833 834 835 836 837 838 839 840
static inline int ip6_rt_check(struct rt6key *rt_key,
			       struct in6_addr *fl_addr,
			       struct in6_addr *addr_cache)
{
	return ((rt_key->plen != 128 || !ipv6_addr_equal(fl_addr, &rt_key->addr)) &&
		(addr_cache == NULL || !ipv6_addr_equal(fl_addr, addr_cache)));
}

841 842 843
static struct dst_entry *ip6_sk_dst_check(struct sock *sk,
					  struct dst_entry *dst,
					  struct flowi *fl)
L
Linus Torvalds 已提交
844
{
845 846
	struct ipv6_pinfo *np = inet6_sk(sk);
	struct rt6_info *rt = (struct rt6_info *)dst;
L
Linus Torvalds 已提交
847

848 849 850 851 852 853 854 855
	if (!dst)
		goto out;

	/* Yes, checking route validity in not connected
	 * case is not very simple. Take into account,
	 * that we do not support routing by source, TOS,
	 * and MSG_DONTROUTE 		--ANK (980726)
	 *
856 857
	 * 1. ip6_rt_check(): If route was host route,
	 *    check that cached destination is current.
858 859 860 861 862 863 864 865 866 867
	 *    If it is network route, we still may
	 *    check its validity using saved pointer
	 *    to the last used address: daddr_cache.
	 *    We do not want to save whole address now,
	 *    (because main consumer of this service
	 *    is tcp, which has not this problem),
	 *    so that the last trick works only on connected
	 *    sockets.
	 * 2. oif also should be the same.
	 */
868
	if (ip6_rt_check(&rt->rt6i_dst, &fl->fl6_dst, np->daddr_cache) ||
869 870 871
#ifdef CONFIG_IPV6_SUBTREES
	    ip6_rt_check(&rt->rt6i_src, &fl->fl6_src, np->saddr_cache) ||
#endif
872
	    (fl->oif && fl->oif != dst->dev->ifindex)) {
873 874
		dst_release(dst);
		dst = NULL;
L
Linus Torvalds 已提交
875 876
	}

877 878 879 880 881 882 883 884 885
out:
	return dst;
}

static int ip6_dst_lookup_tail(struct sock *sk,
			       struct dst_entry **dst, struct flowi *fl)
{
	int err;

L
Linus Torvalds 已提交
886 887 888 889 890 891 892 893
	if (*dst == NULL)
		*dst = ip6_route_output(sk, fl);

	if ((err = (*dst)->error))
		goto out_err_release;

	if (ipv6_addr_any(&fl->fl6_src)) {
		err = ipv6_get_saddr(*dst, &fl->fl6_dst, &fl->fl6_src);
894
		if (err)
L
Linus Torvalds 已提交
895 896 897
			goto out_err_release;
	}

898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932
#ifdef CONFIG_IPV6_OPTIMISTIC_DAD
		/*
		 * Here if the dst entry we've looked up
		 * has a neighbour entry that is in the INCOMPLETE
		 * state and the src address from the flow is
		 * marked as OPTIMISTIC, we release the found
		 * dst entry and replace it instead with the
		 * dst entry of the nexthop router
		 */
		if (!((*dst)->neighbour->nud_state & NUD_VALID)) {
			struct inet6_ifaddr *ifp;
			struct flowi fl_gw;
			int redirect;

			ifp = ipv6_get_ifaddr(&fl->fl6_src, (*dst)->dev, 1);

			redirect = (ifp && ifp->flags & IFA_F_OPTIMISTIC);
			if (ifp)
				in6_ifa_put(ifp);

			if (redirect) {
				/*
				 * We need to get the dst entry for the
				 * default router instead
				 */
				dst_release(*dst);
				memcpy(&fl_gw, fl, sizeof(struct flowi));
				memset(&fl_gw.fl6_dst, 0, sizeof(struct in6_addr));
				*dst = ip6_route_output(sk, &fl_gw);
				if ((err = (*dst)->error))
					goto out_err_release;
			}
		}
#endif

L
Linus Torvalds 已提交
933 934 935
	return 0;

out_err_release:
936 937
	if (err == -ENETUNREACH)
		IP6_INC_STATS_BH(NULL, IPSTATS_MIB_OUTNOROUTES);
L
Linus Torvalds 已提交
938 939 940 941
	dst_release(*dst);
	*dst = NULL;
	return err;
}
A
Adrian Bunk 已提交
942

943 944 945 946 947 948 949 950 951 952 953 954 955 956 957
/**
 *	ip6_dst_lookup - perform route lookup on flow
 *	@sk: socket which provides route info
 *	@dst: pointer to dst_entry * for result
 *	@fl: flow to lookup
 *
 *	This function performs a route lookup on the given flow.
 *
 *	It returns zero on success, or a standard errno code on error.
 */
int ip6_dst_lookup(struct sock *sk, struct dst_entry **dst, struct flowi *fl)
{
	*dst = NULL;
	return ip6_dst_lookup_tail(sk, dst, fl);
}
958 959
EXPORT_SYMBOL_GPL(ip6_dst_lookup);

960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984
/**
 *	ip6_sk_dst_lookup - perform socket cached route lookup on flow
 *	@sk: socket which provides the dst cache and route info
 *	@dst: pointer to dst_entry * for result
 *	@fl: flow to lookup
 *
 *	This function performs a route lookup on the given flow with the
 *	possibility of using the cached route in the socket if it is valid.
 *	It will take the socket dst lock when operating on the dst cache.
 *	As a result, this function can only be used in process context.
 *
 *	It returns zero on success, or a standard errno code on error.
 */
int ip6_sk_dst_lookup(struct sock *sk, struct dst_entry **dst, struct flowi *fl)
{
	*dst = NULL;
	if (sk) {
		*dst = sk_dst_check(sk, inet6_sk(sk)->dst_cookie);
		*dst = ip6_sk_dst_check(sk, *dst, fl);
	}

	return ip6_dst_lookup_tail(sk, dst, fl);
}
EXPORT_SYMBOL_GPL(ip6_sk_dst_lookup);

A
Adrian Bunk 已提交
985
static inline int ip6_ufo_append_data(struct sock *sk,
986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012
			int getfrag(void *from, char *to, int offset, int len,
			int odd, struct sk_buff *skb),
			void *from, int length, int hh_len, int fragheaderlen,
			int transhdrlen, int mtu,unsigned int flags)

{
	struct sk_buff *skb;
	int err;

	/* There is support for UDP large send offload by network
	 * device, so create one single skb packet containing complete
	 * udp datagram
	 */
	if ((skb = skb_peek_tail(&sk->sk_write_queue)) == NULL) {
		skb = sock_alloc_send_skb(sk,
			hh_len + fragheaderlen + transhdrlen + 20,
			(flags & MSG_DONTWAIT), &err);
		if (skb == NULL)
			return -ENOMEM;

		/* reserve space for Hardware header */
		skb_reserve(skb, hh_len);

		/* create space for UDP/IP header */
		skb_put(skb,fragheaderlen + transhdrlen);

		/* initialize network header pointer */
1013
		skb_reset_network_header(skb);
1014 1015

		/* initialize protocol header pointer */
1016
		skb->transport_header = skb->network_header + fragheaderlen;
1017

1018
		skb->ip_summed = CHECKSUM_PARTIAL;
1019 1020 1021 1022 1023 1024 1025 1026 1027 1028
		skb->csum = 0;
		sk->sk_sndmsg_off = 0;
	}

	err = skb_append_datato_frags(sk,skb, getfrag, from,
				      (length - transhdrlen));
	if (!err) {
		struct frag_hdr fhdr;

		/* specify the length of each IP datagram fragment*/
1029
		skb_shinfo(skb)->gso_size = mtu - fragheaderlen -
1030
					    sizeof(struct frag_hdr);
H
Herbert Xu 已提交
1031
		skb_shinfo(skb)->gso_type = SKB_GSO_UDP;
1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044
		ipv6_select_ident(skb, &fhdr);
		skb_shinfo(skb)->ip6_frag_id = fhdr.identification;
		__skb_queue_tail(&sk->sk_write_queue, skb);

		return 0;
	}
	/* There is not enough support do UPD LSO,
	 * so follow normal path
	 */
	kfree_skb(skb);

	return err;
}
L
Linus Torvalds 已提交
1045

1046 1047 1048 1049 1050
int ip6_append_data(struct sock *sk, int getfrag(void *from, char *to,
	int offset, int len, int odd, struct sk_buff *skb),
	void *from, int length, int transhdrlen,
	int hlimit, int tclass, struct ipv6_txoptions *opt, struct flowi *fl,
	struct rt6_info *rt, unsigned int flags)
L
Linus Torvalds 已提交
1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087
{
	struct inet_sock *inet = inet_sk(sk);
	struct ipv6_pinfo *np = inet6_sk(sk);
	struct sk_buff *skb;
	unsigned int maxfraglen, fragheaderlen;
	int exthdrlen;
	int hh_len;
	int mtu;
	int copy;
	int err;
	int offset = 0;
	int csummode = CHECKSUM_NONE;

	if (flags&MSG_PROBE)
		return 0;
	if (skb_queue_empty(&sk->sk_write_queue)) {
		/*
		 * setup for corking
		 */
		if (opt) {
			if (np->cork.opt == NULL) {
				np->cork.opt = kmalloc(opt->tot_len,
						       sk->sk_allocation);
				if (unlikely(np->cork.opt == NULL))
					return -ENOBUFS;
			} else if (np->cork.opt->tot_len < opt->tot_len) {
				printk(KERN_DEBUG "ip6_append_data: invalid option length\n");
				return -EINVAL;
			}
			memcpy(np->cork.opt, opt, opt->tot_len);
			inet->cork.flags |= IPCORK_OPT;
			/* need source address above miyazawa*/
		}
		dst_hold(&rt->u.dst);
		np->cork.rt = rt;
		inet->cork.fl = *fl;
		np->cork.hop_limit = hlimit;
1088
		np->cork.tclass = tclass;
J
John Heffner 已提交
1089 1090
		mtu = np->pmtudisc == IPV6_PMTUDISC_PROBE ?
		      rt->u.dst.dev->mtu : dst_mtu(rt->u.dst.path);
1091
		if (np->frag_size < mtu) {
1092 1093 1094 1095
			if (np->frag_size)
				mtu = np->frag_size;
		}
		inet->cork.fragsize = mtu;
L
Linus Torvalds 已提交
1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115
		if (dst_allfrag(rt->u.dst.path))
			inet->cork.flags |= IPCORK_ALLFRAG;
		inet->cork.length = 0;
		sk->sk_sndmsg_page = NULL;
		sk->sk_sndmsg_off = 0;
		exthdrlen = rt->u.dst.header_len + (opt ? opt->opt_flen : 0);
		length += exthdrlen;
		transhdrlen += exthdrlen;
	} else {
		rt = np->cork.rt;
		fl = &inet->cork.fl;
		if (inet->cork.flags & IPCORK_OPT)
			opt = np->cork.opt;
		transhdrlen = 0;
		exthdrlen = 0;
		mtu = inet->cork.fragsize;
	}

	hh_len = LL_RESERVED_SPACE(rt->u.dst.dev);

1116
	fragheaderlen = sizeof(struct ipv6hdr) + rt->u.dst.nfheader_len + (opt ? opt->opt_nflen : 0);
L
Linus Torvalds 已提交
1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132
	maxfraglen = ((mtu - fragheaderlen) & ~7) + fragheaderlen - sizeof(struct frag_hdr);

	if (mtu <= sizeof(struct ipv6hdr) + IPV6_MAXPLEN) {
		if (inet->cork.length + length > sizeof(struct ipv6hdr) + IPV6_MAXPLEN - fragheaderlen) {
			ipv6_local_error(sk, EMSGSIZE, fl, mtu-exthdrlen);
			return -EMSGSIZE;
		}
	}

	/*
	 * Let's try using as much space as possible.
	 * Use MTU if total length of the message fits into the MTU.
	 * Otherwise, we need to reserve fragment header and
	 * fragment alignment (= 8-15 octects, in total).
	 *
	 * Note that we may need to "move" the data from the tail of
1133
	 * of the buffer to the new fragment when we split
L
Linus Torvalds 已提交
1134 1135
	 * the message.
	 *
1136
	 * FIXME: It may be fragmented into multiple chunks
L
Linus Torvalds 已提交
1137 1138
	 *        at once if non-fragmentable extension headers
	 *        are too large.
1139
	 * --yoshfuji
L
Linus Torvalds 已提交
1140 1141 1142
	 */

	inet->cork.length += length;
1143 1144 1145
	if (((length > mtu) && (sk->sk_protocol == IPPROTO_UDP)) &&
	    (rt->u.dst.dev->features & NETIF_F_UFO)) {

1146 1147 1148 1149
		err = ip6_ufo_append_data(sk, getfrag, from, length, hh_len,
					  fragheaderlen, transhdrlen, mtu,
					  flags);
		if (err)
1150 1151 1152
			goto error;
		return 0;
	}
L
Linus Torvalds 已提交
1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203

	if ((skb = skb_peek_tail(&sk->sk_write_queue)) == NULL)
		goto alloc_new_skb;

	while (length > 0) {
		/* Check if the remaining data fits into current packet. */
		copy = (inet->cork.length <= mtu && !(inet->cork.flags & IPCORK_ALLFRAG) ? mtu : maxfraglen) - skb->len;
		if (copy < length)
			copy = maxfraglen - skb->len;

		if (copy <= 0) {
			char *data;
			unsigned int datalen;
			unsigned int fraglen;
			unsigned int fraggap;
			unsigned int alloclen;
			struct sk_buff *skb_prev;
alloc_new_skb:
			skb_prev = skb;

			/* There's no room in the current skb */
			if (skb_prev)
				fraggap = skb_prev->len - maxfraglen;
			else
				fraggap = 0;

			/*
			 * If remaining data exceeds the mtu,
			 * we know we need more fragment(s).
			 */
			datalen = length + fraggap;
			if (datalen > (inet->cork.length <= mtu && !(inet->cork.flags & IPCORK_ALLFRAG) ? mtu : maxfraglen) - fragheaderlen)
				datalen = maxfraglen - fragheaderlen;

			fraglen = datalen + fragheaderlen;
			if ((flags & MSG_MORE) &&
			    !(rt->u.dst.dev->features&NETIF_F_SG))
				alloclen = mtu;
			else
				alloclen = datalen + fragheaderlen;

			/*
			 * The last fragment gets additional space at tail.
			 * Note: we overallocate on fragments with MSG_MODE
			 * because we have no idea if we're the last one.
			 */
			if (datalen == length + fraggap)
				alloclen += rt->u.dst.trailer_len;

			/*
			 * We just reserve space for fragment header.
1204
			 * Note: this may be overallocation if the message
L
Linus Torvalds 已提交
1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236
			 * (without MSG_MORE) fits into the MTU.
			 */
			alloclen += sizeof(struct frag_hdr);

			if (transhdrlen) {
				skb = sock_alloc_send_skb(sk,
						alloclen + hh_len,
						(flags & MSG_DONTWAIT), &err);
			} else {
				skb = NULL;
				if (atomic_read(&sk->sk_wmem_alloc) <=
				    2 * sk->sk_sndbuf)
					skb = sock_wmalloc(sk,
							   alloclen + hh_len, 1,
							   sk->sk_allocation);
				if (unlikely(skb == NULL))
					err = -ENOBUFS;
			}
			if (skb == NULL)
				goto error;
			/*
			 *	Fill in the control structures
			 */
			skb->ip_summed = csummode;
			skb->csum = 0;
			/* reserve for fragmentation */
			skb_reserve(skb, hh_len+sizeof(struct frag_hdr));

			/*
			 *	Find where to start putting bytes
			 */
			data = skb_put(skb, fraglen);
1237
			skb_set_network_header(skb, exthdrlen);
L
Linus Torvalds 已提交
1238
			data += fragheaderlen;
1239 1240
			skb->transport_header = (skb->network_header +
						 fragheaderlen);
L
Linus Torvalds 已提交
1241 1242 1243 1244 1245 1246 1247
			if (fraggap) {
				skb->csum = skb_copy_and_csum_bits(
					skb_prev, maxfraglen,
					data + transhdrlen, fraggap, 0);
				skb_prev->csum = csum_sub(skb_prev->csum,
							  skb->csum);
				data += fraggap;
1248
				pskb_trim_unique(skb_prev, maxfraglen);
L
Linus Torvalds 已提交
1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 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
			}
			copy = datalen - transhdrlen - fraggap;
			if (copy < 0) {
				err = -EINVAL;
				kfree_skb(skb);
				goto error;
			} else if (copy > 0 && getfrag(from, data + transhdrlen, offset, copy, fraggap, skb) < 0) {
				err = -EFAULT;
				kfree_skb(skb);
				goto error;
			}

			offset += copy;
			length -= datalen - fraggap;
			transhdrlen = 0;
			exthdrlen = 0;
			csummode = CHECKSUM_NONE;

			/*
			 * Put the packet on the pending queue
			 */
			__skb_queue_tail(&sk->sk_write_queue, skb);
			continue;
		}

		if (copy > length)
			copy = length;

		if (!(rt->u.dst.dev->features&NETIF_F_SG)) {
			unsigned int off;

			off = skb->len;
			if (getfrag(from, skb_put(skb, copy),
						offset, copy, off, skb) < 0) {
				__skb_trim(skb, off);
				err = -EFAULT;
				goto error;
			}
		} else {
			int i = skb_shinfo(skb)->nr_frags;
			skb_frag_t *frag = &skb_shinfo(skb)->frags[i-1];
			struct page *page = sk->sk_sndmsg_page;
			int off = sk->sk_sndmsg_off;
			unsigned int left;

			if (page && (left = PAGE_SIZE - off) > 0) {
				if (copy >= left)
					copy = left;
				if (page != frag->page) {
					if (i == MAX_SKB_FRAGS) {
						err = -EMSGSIZE;
						goto error;
					}
					get_page(page);
					skb_fill_page_desc(skb, i, page, sk->sk_sndmsg_off, 0);
					frag = &skb_shinfo(skb)->frags[i];
				}
			} else if(i < MAX_SKB_FRAGS) {
				if (copy > PAGE_SIZE)
					copy = PAGE_SIZE;
				page = alloc_pages(sk->sk_allocation, 0);
				if (page == NULL) {
					err = -ENOMEM;
					goto error;
				}
				sk->sk_sndmsg_page = page;
				sk->sk_sndmsg_off = 0;

				skb_fill_page_desc(skb, i, page, 0, 0);
				frag = &skb_shinfo(skb)->frags[i];
			} else {
				err = -EMSGSIZE;
				goto error;
			}
			if (getfrag(from, page_address(frag->page)+frag->page_offset+frag->size, offset, copy, skb->len, skb) < 0) {
				err = -EFAULT;
				goto error;
			}
			sk->sk_sndmsg_off += copy;
			frag->size += copy;
			skb->len += copy;
			skb->data_len += copy;
1331 1332
			skb->truesize += copy;
			atomic_add(copy, &sk->sk_wmem_alloc);
L
Linus Torvalds 已提交
1333 1334 1335 1336 1337 1338 1339
		}
		offset += copy;
		length -= copy;
	}
	return 0;
error:
	inet->cork.length -= length;
1340
	IP6_INC_STATS(rt->rt6i_idev, IPSTATS_MIB_OUTDISCARDS);
L
Linus Torvalds 已提交
1341 1342 1343
	return err;
}

1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356
static void ip6_cork_release(struct inet_sock *inet, struct ipv6_pinfo *np)
{
	inet->cork.flags &= ~IPCORK_OPT;
	kfree(np->cork.opt);
	np->cork.opt = NULL;
	if (np->cork.rt) {
		dst_release(&np->cork.rt->u.dst);
		np->cork.rt = NULL;
		inet->cork.flags &= ~IPCORK_ALLFRAG;
	}
	memset(&inet->cork.fl, 0, sizeof(inet->cork.fl));
}

L
Linus Torvalds 已提交
1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375
int ip6_push_pending_frames(struct sock *sk)
{
	struct sk_buff *skb, *tmp_skb;
	struct sk_buff **tail_skb;
	struct in6_addr final_dst_buf, *final_dst = &final_dst_buf;
	struct inet_sock *inet = inet_sk(sk);
	struct ipv6_pinfo *np = inet6_sk(sk);
	struct ipv6hdr *hdr;
	struct ipv6_txoptions *opt = np->cork.opt;
	struct rt6_info *rt = np->cork.rt;
	struct flowi *fl = &inet->cork.fl;
	unsigned char proto = fl->proto;
	int err = 0;

	if ((skb = __skb_dequeue(&sk->sk_write_queue)) == NULL)
		goto out;
	tail_skb = &(skb_shinfo(skb)->frag_list);

	/* move skb->data to ip header from ext header */
1376
	if (skb->data < skb_network_header(skb))
1377
		__skb_pull(skb, skb_network_offset(skb));
L
Linus Torvalds 已提交
1378
	while ((tmp_skb = __skb_dequeue(&sk->sk_write_queue)) != NULL) {
1379
		__skb_pull(tmp_skb, skb_network_header_len(skb));
L
Linus Torvalds 已提交
1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390
		*tail_skb = tmp_skb;
		tail_skb = &(tmp_skb->next);
		skb->len += tmp_skb->len;
		skb->data_len += tmp_skb->len;
		skb->truesize += tmp_skb->truesize;
		__sock_put(tmp_skb->sk);
		tmp_skb->destructor = NULL;
		tmp_skb->sk = NULL;
	}

	ipv6_addr_copy(final_dst, &fl->fl6_dst);
1391
	__skb_pull(skb, skb_network_header_len(skb));
L
Linus Torvalds 已提交
1392 1393 1394 1395 1396
	if (opt && opt->opt_flen)
		ipv6_push_frag_opts(skb, opt, &proto);
	if (opt && opt->opt_nflen)
		ipv6_push_nfrag_opts(skb, opt, &proto, &final_dst);

1397 1398
	skb_push(skb, sizeof(struct ipv6hdr));
	skb_reset_network_header(skb);
1399
	hdr = ipv6_hdr(skb);
1400

A
Al Viro 已提交
1401
	*(__be32*)hdr = fl->fl6_flowlabel |
1402
		     htonl(0x60000000 | ((int)np->cork.tclass << 20));
L
Linus Torvalds 已提交
1403 1404 1405 1406 1407 1408 1409 1410 1411 1412

	if (skb->len <= sizeof(struct ipv6hdr) + IPV6_MAXPLEN)
		hdr->payload_len = htons(skb->len - sizeof(struct ipv6hdr));
	else
		hdr->payload_len = 0;
	hdr->hop_limit = np->cork.hop_limit;
	hdr->nexthdr = proto;
	ipv6_addr_copy(&hdr->saddr, &fl->fl6_src);
	ipv6_addr_copy(&hdr->daddr, final_dst);

1413 1414
	skb->priority = sk->sk_priority;

L
Linus Torvalds 已提交
1415
	skb->dst = dst_clone(&rt->u.dst);
1416
	IP6_INC_STATS(rt->rt6i_idev, IPSTATS_MIB_OUTREQUESTS);
1417 1418 1419 1420 1421 1422 1423
	if (proto == IPPROTO_ICMPV6) {
		struct inet6_dev *idev = ip6_dst_idev(skb->dst);

		ICMP6MSGOUT_INC_STATS_BH(idev, icmp6_hdr(skb)->icmp6_type);
		ICMP6_INC_STATS_BH(idev, ICMP6_MIB_OUTMSGS);
	}

L
Linus Torvalds 已提交
1424 1425 1426
	err = NF_HOOK(PF_INET6, NF_IP6_LOCAL_OUT, skb, NULL, skb->dst->dev, dst_output);
	if (err) {
		if (err > 0)
1427
			err = np->recverr ? net_xmit_errno(err) : 0;
L
Linus Torvalds 已提交
1428 1429 1430 1431 1432
		if (err)
			goto error;
	}

out:
1433
	ip6_cork_release(inet, np);
L
Linus Torvalds 已提交
1434 1435 1436 1437 1438 1439 1440 1441 1442 1443
	return err;
error:
	goto out;
}

void ip6_flush_pending_frames(struct sock *sk)
{
	struct sk_buff *skb;

	while ((skb = __skb_dequeue_tail(&sk->sk_write_queue)) != NULL) {
1444 1445 1446
		if (skb->dst)
			IP6_INC_STATS(ip6_dst_idev(skb->dst),
				      IPSTATS_MIB_OUTDISCARDS);
L
Linus Torvalds 已提交
1447 1448 1449
		kfree_skb(skb);
	}

1450
	ip6_cork_release(inet_sk(sk), inet6_sk(sk));
L
Linus Torvalds 已提交
1451
}