ip6_output.c 36.7 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
 *
 *	$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>
H
Herbert Xu 已提交
32
#include <linux/kernel.h>
L
Linus Torvalds 已提交
33 34 35 36 37 38 39 40
#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);
}

H
Herbert Xu 已提交
73 74 75 76 77 78 79 80 81
int __ip6_local_out(struct sk_buff *skb)
{
	int len;

	len = skb->len - sizeof(struct ipv6hdr);
	if (len > IPV6_MAXPLEN)
		len = 0;
	ipv6_hdr(skb)->payload_len = htons(len);

82
	return nf_hook(PF_INET6, NF_INET_LOCAL_OUT, skb, NULL, skb->dst->dev,
H
Herbert Xu 已提交
83 84 85 86 87 88 89 90 91 92 93 94 95 96 97
		       dst_output);
}

int ip6_local_out(struct sk_buff *skb)
{
	int err;

	err = __ip6_local_out(skb);
	if (likely(err == 1))
		err = dst_output(skb);

	return err;
}
EXPORT_SYMBOL_GPL(ip6_local_out);

P
Patrick McHardy 已提交
98
static int ip6_output_finish(struct sk_buff *skb)
L
Linus Torvalds 已提交
99 100 101
{
	struct dst_entry *dst = skb->dst;

102 103 104
	if (dst->hh)
		return neigh_hh_output(dst->hh, skb);
	else if (dst->neighbour)
L
Linus Torvalds 已提交
105 106
		return dst->neighbour->output(skb);

107
	IP6_INC_STATS_BH(ip6_dst_idev(dst), IPSTATS_MIB_OUTNOROUTES);
L
Linus Torvalds 已提交
108 109 110 111 112 113 114 115
	kfree_skb(skb);
	return -EINVAL;

}

/* dev_loopback_xmit for use with netfilter. */
static int ip6_dev_loopback_xmit(struct sk_buff *newskb)
{
116
	skb_reset_mac_header(newskb);
117
	__skb_pull(newskb, skb_network_offset(newskb));
L
Linus Torvalds 已提交
118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134
	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;

135
	if (ipv6_addr_is_multicast(&ipv6_hdr(skb)->daddr)) {
L
Linus Torvalds 已提交
136
		struct ipv6_pinfo* np = skb->sk ? inet6_sk(skb->sk) : NULL;
137
		struct inet6_dev *idev = ip6_dst_idev(skb->dst);
L
Linus Torvalds 已提交
138 139

		if (!(dev->flags & IFF_LOOPBACK) && (!np || np->mc_loop) &&
140 141
		    ipv6_chk_mcast_addr(dev, &ipv6_hdr(skb)->daddr,
					&ipv6_hdr(skb)->saddr)) {
L
Linus Torvalds 已提交
142 143 144 145 146 147
			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)
148 149
				NF_HOOK(PF_INET6, NF_INET_POST_ROUTING, newskb,
					NULL, newskb->dev,
L
Linus Torvalds 已提交
150 151
					ip6_dev_loopback_xmit);

152
			if (ipv6_hdr(skb)->hop_limit == 0) {
153
				IP6_INC_STATS(idev, IPSTATS_MIB_OUTDISCARDS);
L
Linus Torvalds 已提交
154 155 156 157 158
				kfree_skb(skb);
				return 0;
			}
		}

159
		IP6_INC_STATS(idev, IPSTATS_MIB_OUTMCASTPKTS);
L
Linus Torvalds 已提交
160 161
	}

162 163
	return NF_HOOK(PF_INET6, NF_INET_POST_ROUTING, skb, NULL, skb->dev,
		       ip6_output_finish);
L
Linus Torvalds 已提交
164 165
}

J
John Heffner 已提交
166 167 168 169 170 171 172 173
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
174 175
int ip6_output(struct sk_buff *skb)
{
J
John Heffner 已提交
176
	if ((skb->len > ip6_skb_dst_mtu(skb) && !skb_is_gso(skb)) ||
177
				dst_allfrag(skb->dst))
L
Linus Torvalds 已提交
178 179 180 181 182 183 184 185 186 187 188 189
		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)
{
190
	struct ipv6_pinfo *np = inet6_sk(sk);
L
Linus Torvalds 已提交
191 192 193 194 195
	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;
196
	int hlimit, tclass;
L
Linus Torvalds 已提交
197 198 199
	u32 mtu;

	if (opt) {
200
		unsigned int head_room;
L
Linus Torvalds 已提交
201 202 203 204 205 206 207 208 209 210

		/* 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);
211 212 213 214
			if (skb2 == NULL) {
				IP6_INC_STATS(ip6_dst_idev(skb->dst),
					      IPSTATS_MIB_OUTDISCARDS);
				kfree_skb(skb);
L
Linus Torvalds 已提交
215 216
				return -ENOBUFS;
			}
217 218
			kfree_skb(skb);
			skb = skb2;
L
Linus Torvalds 已提交
219 220 221 222 223 224 225 226 227
			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);
	}

228 229
	skb_push(skb, sizeof(struct ipv6hdr));
	skb_reset_network_header(skb);
230
	hdr = ipv6_hdr(skb);
L
Linus Torvalds 已提交
231 232 233 234 235 236 237 238 239 240 241 242 243

	/*
	 *	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);

244 245 246 247 248 249
	tclass = -1;
	if (np)
		tclass = np->tclass;
	if (tclass < 0)
		tclass = 0;

A
Al Viro 已提交
250
	*(__be32 *)hdr = htonl(0x60000000 | (tclass << 20)) | fl->fl6_flowlabel;
251

L
Linus Torvalds 已提交
252 253 254 255 256 257 258
	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);

259 260
	skb->priority = sk->sk_priority;

L
Linus Torvalds 已提交
261
	mtu = dst_mtu(dst);
H
Herbert Xu 已提交
262
	if ((skb->len <= mtu) || ipfragok || skb_is_gso(skb)) {
263 264
		IP6_INC_STATS(ip6_dst_idev(skb->dst),
			      IPSTATS_MIB_OUTREQUESTS);
265
		return NF_HOOK(PF_INET6, NF_INET_LOCAL_OUT, skb, NULL, dst->dev,
266
				dst_output);
L
Linus Torvalds 已提交
267 268 269 270 271 272
	}

	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);
273
	IP6_INC_STATS(ip6_dst_idev(skb->dst), IPSTATS_MIB_FRAGFAILS);
L
Linus Torvalds 已提交
274 275 276 277
	kfree_skb(skb);
	return -EMSGSIZE;
}

278 279
EXPORT_SYMBOL(ip6_xmit);

L
Linus Torvalds 已提交
280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299
/*
 *	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);

300 301
	skb_reset_network_header(skb);
	skb_put(skb, sizeof(struct ipv6hdr));
302
	hdr = ipv6_hdr(skb);
L
Linus Torvalds 已提交
303

A
Al Viro 已提交
304
	*(__be32*)hdr = htonl(0x60000000);
L
Linus Torvalds 已提交
305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323

	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;
324 325 326
		if (sk && ra->sel == sel &&
		    (!sk->sk_bound_dev_if ||
		     sk->sk_bound_dev_if == skb->dev->ifindex)) {
L
Linus Torvalds 已提交
327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344
			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;
}

345 346
static int ip6_forward_proxy_check(struct sk_buff *skb)
{
347
	struct ipv6hdr *hdr = ipv6_hdr(skb);
348 349 350 351 352 353 354 355 356 357 358 359 360
	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;

361 362
		if (!pskb_may_pull(skb, (skb_network_header(skb) +
					 offset + 1 - skb->data)))
363 364
			return 0;

365
		icmp6 = (struct icmp6hdr *)(skb_network_header(skb) + offset);
366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382

		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;
		}
	}

383 384 385 386 387 388 389 390 391 392
	/*
	 * 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;
	}

393 394 395
	return 0;
}

L
Linus Torvalds 已提交
396 397 398 399 400 401 402 403
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;
404
	struct ipv6hdr *hdr = ipv6_hdr(skb);
L
Linus Torvalds 已提交
405
	struct inet6_skb_parm *opt = IP6CB(skb);
406

L
Linus Torvalds 已提交
407 408 409 410
	if (ipv6_devconf.forwarding == 0)
		goto error;

	if (!xfrm6_policy_check(NULL, XFRM_POLICY_FWD, skb)) {
411
		IP6_INC_STATS(ip6_dst_idev(dst), IPSTATS_MIB_INDISCARDS);
L
Linus Torvalds 已提交
412 413 414
		goto drop;
	}

415
	skb_forward_csum(skb);
L
Linus Torvalds 已提交
416 417 418 419 420 421 422 423 424 425 426 427 428 429 430

	/*
	 *	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) {
431
		u8 *ptr = skb_network_header(skb) + opt->ra;
L
Linus Torvalds 已提交
432 433 434 435 436 437 438 439 440 441 442 443
		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);
444
		IP6_INC_STATS_BH(ip6_dst_idev(dst), IPSTATS_MIB_INHDRERRORS);
L
Linus Torvalds 已提交
445 446 447 448 449

		kfree_skb(skb);
		return -ETIMEDOUT;
	}

450 451 452
	/* XXX: idev->cnf.proxy_ndp? */
	if (ipv6_devconf.proxy_ndp &&
	    pneigh_lookup(&nd_tbl, &hdr->daddr, skb->dev, 0)) {
453 454
		int proxied = ip6_forward_proxy_check(skb);
		if (proxied > 0)
455
			return ip6_input(skb);
456
		else if (proxied < 0) {
457
			IP6_INC_STATS(ip6_dst_idev(dst), IPSTATS_MIB_INDISCARDS);
458 459
			goto drop;
		}
460 461
	}

L
Linus Torvalds 已提交
462
	if (!xfrm6_route_forward(skb)) {
463
		IP6_INC_STATS(ip6_dst_idev(dst), IPSTATS_MIB_INDISCARDS);
L
Linus Torvalds 已提交
464 465 466 467 468 469
		goto drop;
	}
	dst = skb->dst;

	/* IPv6 specs say nothing about it, but it is clear that we cannot
	   send redirects to source routed frames.
470
	   We don't send redirects to frames decapsulated from IPsec.
L
Linus Torvalds 已提交
471
	 */
472 473
	if (skb->dev == dst->dev && dst->neighbour && opt->srcrt == 0 &&
	    !skb->sp) {
L
Linus Torvalds 已提交
474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493
		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);
494 495 496
	} else {
		int addrtype = ipv6_addr_type(&hdr->saddr);

L
Linus Torvalds 已提交
497
		/* This check is security critical. */
498 499 500 501 502 503 504
		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
505 506 507 508 509 510
	}

	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);
511 512
		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
513 514 515 516 517
		kfree_skb(skb);
		return -EMSGSIZE;
	}

	if (skb_cow(skb, dst->dev->hard_header_len)) {
518
		IP6_INC_STATS(ip6_dst_idev(dst), IPSTATS_MIB_OUTDISCARDS);
L
Linus Torvalds 已提交
519 520 521
		goto drop;
	}

522
	hdr = ipv6_hdr(skb);
L
Linus Torvalds 已提交
523 524

	/* Mangling hops number delayed to point after skb COW */
525

L
Linus Torvalds 已提交
526 527
	hdr->hop_limit--;

528
	IP6_INC_STATS_BH(ip6_dst_idev(dst), IPSTATS_MIB_OUTFORWDATAGRAMS);
529 530
	return NF_HOOK(PF_INET6, NF_INET_FORWARD, skb, skb->dev, dst->dev,
		       ip6_forward_finish);
L
Linus Torvalds 已提交
531 532

error:
533
	IP6_INC_STATS_BH(ip6_dst_idev(dst), IPSTATS_MIB_INADDRERRORS);
L
Linus Torvalds 已提交
534 535 536 537 538 539 540 541 542 543 544 545 546
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
547
	to->mark = from->mark;
L
Linus Torvalds 已提交
548 549 550 551

#ifdef CONFIG_NET_SCHED
	to->tc_index = from->tc_index;
#endif
552
	nf_copy(to, from);
553 554 555 556
#if defined(CONFIG_NETFILTER_XT_TARGET_TRACE) || \
    defined(CONFIG_NETFILTER_XT_TARGET_TRACE_MODULE)
	to->nf_trace = from->nf_trace;
#endif
557
	skb_copy_secmark(to, from);
L
Linus Torvalds 已提交
558 559 560 561 562
}

int ip6_find_1stfragopt(struct sk_buff *skb, u8 **nexthdr)
{
	u16 offset = sizeof(struct ipv6hdr);
563 564
	struct ipv6_opt_hdr *exthdr =
				(struct ipv6_opt_hdr *)(ipv6_hdr(skb) + 1);
565
	unsigned int packet_len = skb->tail - skb->network_header;
L
Linus Torvalds 已提交
566
	int found_rhdr = 0;
567
	*nexthdr = &ipv6_hdr(skb)->nexthdr;
L
Linus Torvalds 已提交
568 569 570 571 572 573

	while (offset + 1 <= packet_len) {

		switch (**nexthdr) {

		case NEXTHDR_HOP:
574
			break;
L
Linus Torvalds 已提交
575
		case NEXTHDR_ROUTING:
576 577
			found_rhdr = 1;
			break;
L
Linus Torvalds 已提交
578
		case NEXTHDR_DEST:
579
#if defined(CONFIG_IPV6_MIP6) || defined(CONFIG_IPV6_MIP6_MODULE)
580 581 582 583 584
			if (ipv6_find_tlv(skb, offset, IPV6_TLV_HAO) >= 0)
				break;
#endif
			if (found_rhdr)
				return offset;
L
Linus Torvalds 已提交
585 586 587 588
			break;
		default :
			return offset;
		}
589 590 591

		offset += ipv6_optlen(exthdr);
		*nexthdr = &exthdr->nexthdr;
592 593
		exthdr = (struct ipv6_opt_hdr *)(skb_network_header(skb) +
						 offset);
L
Linus Torvalds 已提交
594 595 596 597
	}

	return offset;
}
598
EXPORT_SYMBOL_GPL(ip6_find_1stfragopt);
L
Linus Torvalds 已提交
599 600 601 602 603 604

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;
605
	struct ipv6_pinfo *np = skb->sk ? inet6_sk(skb->sk) : NULL;
L
Linus Torvalds 已提交
606 607 608
	struct ipv6hdr *tmp_hdr;
	struct frag_hdr *fh;
	unsigned int mtu, hlen, left, len;
A
Al Viro 已提交
609
	__be32 frag_id = 0;
L
Linus Torvalds 已提交
610 611 612 613 614 615 616
	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
617
	mtu = ip6_skb_dst_mtu(skb);
618 619 620 621 622 623 624 625 626 627 628 629 630

	/* 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;
	}

631 632 633 634 635
	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
636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654

	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;
655 656 657 658 659 660 661 662

			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
663 664 665 666 667 668 669 670
		}

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

Y
YOSHIFUJI Hideaki 已提交
671
		*prevhdr = NEXTHDR_FRAGMENT;
672
		tmp_hdr = kmemdup(skb_network_header(skb), hlen, GFP_ATOMIC);
L
Linus Torvalds 已提交
673
		if (!tmp_hdr) {
674
			IP6_INC_STATS(ip6_dst_idev(skb->dst), IPSTATS_MIB_FRAGFAILS);
L
Linus Torvalds 已提交
675 676 677 678 679
			return -ENOMEM;
		}

		__skb_pull(skb, hlen);
		fh = (struct frag_hdr*)__skb_push(skb, sizeof(struct frag_hdr));
680 681
		__skb_push(skb, hlen);
		skb_reset_network_header(skb);
682
		memcpy(skb_network_header(skb), tmp_hdr, hlen);
L
Linus Torvalds 已提交
683 684 685 686 687 688 689 690 691 692

		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;
693 694
		ipv6_hdr(skb)->payload_len = htons(first_len -
						   sizeof(struct ipv6hdr));
695 696

		dst_hold(&rt->u.dst);
L
Linus Torvalds 已提交
697 698 699 700 701 702

		for (;;) {
			/* Prepare header of the next frame,
			 * before previous one went down. */
			if (frag) {
				frag->ip_summed = CHECKSUM_NONE;
703
				skb_reset_transport_header(frag);
L
Linus Torvalds 已提交
704
				fh = (struct frag_hdr*)__skb_push(frag, sizeof(struct frag_hdr));
705 706
				__skb_push(frag, hlen);
				skb_reset_network_header(frag);
707 708
				memcpy(skb_network_header(frag), tmp_hdr,
				       hlen);
L
Linus Torvalds 已提交
709 710 711 712 713 714 715
				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;
716 717 718
				ipv6_hdr(frag)->payload_len =
						htons(frag->len -
						      sizeof(struct ipv6hdr));
L
Linus Torvalds 已提交
719 720
				ip6_copy_metadata(frag, skb);
			}
721

L
Linus Torvalds 已提交
722
			err = output(skb);
723
			if(!err)
724
				IP6_INC_STATS(ip6_dst_idev(&rt->u.dst), IPSTATS_MIB_FRAGCREATES);
725

L
Linus Torvalds 已提交
726 727 728 729 730 731 732 733
			if (err || !frag)
				break;

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

J
Jesper Juhl 已提交
734
		kfree(tmp_hdr);
L
Linus Torvalds 已提交
735 736

		if (err == 0) {
737 738
			IP6_INC_STATS(ip6_dst_idev(&rt->u.dst), IPSTATS_MIB_FRAGOKS);
			dst_release(&rt->u.dst);
L
Linus Torvalds 已提交
739 740 741 742 743 744 745 746 747
			return 0;
		}

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

748 749
		IP6_INC_STATS(ip6_dst_idev(&rt->u.dst), IPSTATS_MIB_FRAGFAILS);
		dst_release(&rt->u.dst);
L
Linus Torvalds 已提交
750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780
		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) {
781
			NETDEBUG(KERN_INFO "IPv6: frag: no memory for new fragment!\n");
782 783
			IP6_INC_STATS(ip6_dst_idev(skb->dst),
				      IPSTATS_MIB_FRAGFAILS);
L
Linus Torvalds 已提交
784 785 786 787 788 789 790 791 792 793 794
			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));
795
		skb_reset_network_header(frag);
796
		fh = (struct frag_hdr *)(skb_network_header(frag) + hlen);
797 798
		frag->transport_header = (frag->network_header + hlen +
					  sizeof(struct frag_hdr));
L
Linus Torvalds 已提交
799 800 801 802 803 804 805 806 807 808 809

		/*
		 *	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.
		 */
810
		skb_copy_from_linear_data(skb, skb_network_header(frag), hlen);
L
Linus Torvalds 已提交
811 812 813 814 815 816

		/*
		 *	Build fragment header.
		 */
		fh->nexthdr = nexthdr;
		fh->reserved = 0;
817
		if (!frag_id) {
L
Linus Torvalds 已提交
818 819 820 821 822 823 824 825
			ipv6_select_ident(skb, fh);
			frag_id = fh->identification;
		} else
			fh->identification = frag_id;

		/*
		 *	Copy a block of the IP datagram.
		 */
826
		if (skb_copy_bits(skb, ptr, skb_transport_header(frag), len))
L
Linus Torvalds 已提交
827 828 829 830 831 832
			BUG();
		left -= len;

		fh->frag_off = htons(offset);
		if (left > 0)
			fh->frag_off |= htons(IP6_MF);
833 834
		ipv6_hdr(frag)->payload_len = htons(frag->len -
						    sizeof(struct ipv6hdr));
L
Linus Torvalds 已提交
835 836 837 838 839 840 841 842 843 844

		ptr += len;
		offset += len;

		/*
		 *	Put this fragment into the sending queue.
		 */
		err = output(frag);
		if (err)
			goto fail;
845

846
		IP6_INC_STATS(ip6_dst_idev(skb->dst), IPSTATS_MIB_FRAGCREATES);
L
Linus Torvalds 已提交
847
	}
848 849
	IP6_INC_STATS(ip6_dst_idev(skb->dst),
		      IPSTATS_MIB_FRAGOKS);
L
Linus Torvalds 已提交
850 851 852 853
	kfree_skb(skb);
	return err;

fail:
854 855
	IP6_INC_STATS(ip6_dst_idev(skb->dst),
		      IPSTATS_MIB_FRAGFAILS);
856
	kfree_skb(skb);
L
Linus Torvalds 已提交
857 858 859
	return err;
}

860 861 862 863 864 865 866 867
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)));
}

868 869 870
static struct dst_entry *ip6_sk_dst_check(struct sock *sk,
					  struct dst_entry *dst,
					  struct flowi *fl)
L
Linus Torvalds 已提交
871
{
872 873
	struct ipv6_pinfo *np = inet6_sk(sk);
	struct rt6_info *rt = (struct rt6_info *)dst;
L
Linus Torvalds 已提交
874

875 876 877 878 879 880 881 882
	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)
	 *
883 884
	 * 1. ip6_rt_check(): If route was host route,
	 *    check that cached destination is current.
885 886 887 888 889 890 891 892 893 894
	 *    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.
	 */
895
	if (ip6_rt_check(&rt->rt6i_dst, &fl->fl6_dst, np->daddr_cache) ||
896 897 898
#ifdef CONFIG_IPV6_SUBTREES
	    ip6_rt_check(&rt->rt6i_src, &fl->fl6_src, np->saddr_cache) ||
#endif
899
	    (fl->oif && fl->oif != dst->dev->ifindex)) {
900 901
		dst_release(dst);
		dst = NULL;
L
Linus Torvalds 已提交
902 903
	}

904 905 906 907 908 909 910 911 912
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
913 914 915 916 917 918 919 920
	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);
921
		if (err)
L
Linus Torvalds 已提交
922 923 924
			goto out_err_release;
	}

925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959
#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
960 961 962
	return 0;

out_err_release:
963 964
	if (err == -ENETUNREACH)
		IP6_INC_STATS_BH(NULL, IPSTATS_MIB_OUTNOROUTES);
L
Linus Torvalds 已提交
965 966 967 968
	dst_release(*dst);
	*dst = NULL;
	return err;
}
A
Adrian Bunk 已提交
969

970 971 972 973 974 975 976 977 978 979 980 981 982 983 984
/**
 *	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);
}
985 986
EXPORT_SYMBOL_GPL(ip6_dst_lookup);

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
/**
 *	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
1012
static inline int ip6_ufo_append_data(struct sock *sk,
1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039
			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 */
1040
		skb_reset_network_header(skb);
1041 1042

		/* initialize protocol header pointer */
1043
		skb->transport_header = skb->network_header + fragheaderlen;
1044

1045
		skb->ip_summed = CHECKSUM_PARTIAL;
1046 1047 1048 1049 1050 1051 1052 1053 1054 1055
		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*/
1056
		skb_shinfo(skb)->gso_size = mtu - fragheaderlen -
1057
					    sizeof(struct frag_hdr);
H
Herbert Xu 已提交
1058
		skb_shinfo(skb)->gso_type = SKB_GSO_UDP;
1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071
		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
1072

1073 1074 1075 1076 1077
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
1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114
{
	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;
1115
		np->cork.tclass = tclass;
J
John Heffner 已提交
1116 1117
		mtu = np->pmtudisc == IPV6_PMTUDISC_PROBE ?
		      rt->u.dst.dev->mtu : dst_mtu(rt->u.dst.path);
1118
		if (np->frag_size < mtu) {
1119 1120 1121 1122
			if (np->frag_size)
				mtu = np->frag_size;
		}
		inet->cork.fragsize = mtu;
L
Linus Torvalds 已提交
1123 1124 1125 1126 1127
		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;
1128
		exthdrlen = rt->u.dst.header_len + (opt ? opt->opt_flen : 0) -
1129
			    rt->nfheader_len;
L
Linus Torvalds 已提交
1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143
		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);

1144 1145
	fragheaderlen = sizeof(struct ipv6hdr) + rt->nfheader_len +
			(opt ? opt->opt_nflen : 0);
L
Linus Torvalds 已提交
1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161
	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
1162
	 * of the buffer to the new fragment when we split
L
Linus Torvalds 已提交
1163 1164
	 * the message.
	 *
1165
	 * FIXME: It may be fragmented into multiple chunks
L
Linus Torvalds 已提交
1166 1167
	 *        at once if non-fragmentable extension headers
	 *        are too large.
1168
	 * --yoshfuji
L
Linus Torvalds 已提交
1169 1170 1171
	 */

	inet->cork.length += length;
1172 1173 1174
	if (((length > mtu) && (sk->sk_protocol == IPPROTO_UDP)) &&
	    (rt->u.dst.dev->features & NETIF_F_UFO)) {

1175 1176 1177 1178
		err = ip6_ufo_append_data(sk, getfrag, from, length, hh_len,
					  fragheaderlen, transhdrlen, mtu,
					  flags);
		if (err)
1179 1180 1181
			goto error;
		return 0;
	}
L
Linus Torvalds 已提交
1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 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

	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.
1233
			 * Note: this may be overallocation if the message
L
Linus Torvalds 已提交
1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265
			 * (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);
1266
			skb_set_network_header(skb, exthdrlen);
L
Linus Torvalds 已提交
1267
			data += fragheaderlen;
1268 1269
			skb->transport_header = (skb->network_header +
						 fragheaderlen);
L
Linus Torvalds 已提交
1270 1271 1272 1273 1274 1275 1276
			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;
1277
				pskb_trim_unique(skb_prev, maxfraglen);
L
Linus Torvalds 已提交
1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359
			}
			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;
1360 1361
			skb->truesize += copy;
			atomic_add(copy, &sk->sk_wmem_alloc);
L
Linus Torvalds 已提交
1362 1363 1364 1365 1366 1367 1368
		}
		offset += copy;
		length -= copy;
	}
	return 0;
error:
	inet->cork.length -= length;
1369
	IP6_INC_STATS(rt->rt6i_idev, IPSTATS_MIB_OUTDISCARDS);
L
Linus Torvalds 已提交
1370 1371 1372
	return err;
}

1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385
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
1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404
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 */
1405
	if (skb->data < skb_network_header(skb))
1406
		__skb_pull(skb, skb_network_offset(skb));
L
Linus Torvalds 已提交
1407
	while ((tmp_skb = __skb_dequeue(&sk->sk_write_queue)) != NULL) {
1408
		__skb_pull(tmp_skb, skb_network_header_len(skb));
L
Linus Torvalds 已提交
1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419
		*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);
1420
	__skb_pull(skb, skb_network_header_len(skb));
L
Linus Torvalds 已提交
1421 1422 1423 1424 1425
	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);

1426 1427
	skb_push(skb, sizeof(struct ipv6hdr));
	skb_reset_network_header(skb);
1428
	hdr = ipv6_hdr(skb);
1429

A
Al Viro 已提交
1430
	*(__be32*)hdr = fl->fl6_flowlabel |
1431
		     htonl(0x60000000 | ((int)np->cork.tclass << 20));
L
Linus Torvalds 已提交
1432 1433 1434 1435 1436 1437

	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);

1438 1439
	skb->priority = sk->sk_priority;

L
Linus Torvalds 已提交
1440
	skb->dst = dst_clone(&rt->u.dst);
1441
	IP6_INC_STATS(rt->rt6i_idev, IPSTATS_MIB_OUTREQUESTS);
1442 1443 1444 1445 1446 1447 1448
	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);
	}

H
Herbert Xu 已提交
1449
	err = ip6_local_out(skb);
L
Linus Torvalds 已提交
1450 1451
	if (err) {
		if (err > 0)
1452
			err = np->recverr ? net_xmit_errno(err) : 0;
L
Linus Torvalds 已提交
1453 1454 1455 1456 1457
		if (err)
			goto error;
	}

out:
1458
	ip6_cork_release(inet, np);
L
Linus Torvalds 已提交
1459 1460 1461 1462 1463 1464 1465 1466 1467 1468
	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) {
1469 1470 1471
		if (skb->dst)
			IP6_INC_STATS(ip6_dst_idev(skb->dst),
				      IPSTATS_MIB_OUTDISCARDS);
L
Linus Torvalds 已提交
1472 1473 1474
		kfree_skb(skb);
	}

1475
	ip6_cork_release(inet_sk(sk), inet6_sk(sk));
L
Linus Torvalds 已提交
1476
}