ip_output.c 34.8 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8 9
/*
 * INET		An implementation of the TCP/IP protocol suite for the LINUX
 *		operating system.  INET is implemented using the  BSD Socket
 *		interface as the means of communication with the user level.
 *
 *		The Internet Protocol (IP) output module.
 *
 * Version:	$Id: ip_output.c,v 1.100 2002/02/01 22:01:03 davem Exp $
 *
10
 * Authors:	Ross Biro
L
Linus Torvalds 已提交
11 12 13 14 15 16 17 18 19 20 21 22 23 24
 *		Fred N. van Kempen, <waltje@uWalt.NL.Mugnet.ORG>
 *		Donald Becker, <becker@super.org>
 *		Alan Cox, <Alan.Cox@linux.org>
 *		Richard Underwood
 *		Stefan Becker, <stefanb@yello.ping.de>
 *		Jorge Cwik, <jorge@laser.satlink.net>
 *		Arnt Gulbrandsen, <agulbra@nvg.unit.no>
 *		Hirokazu Takahashi, <taka@valinux.co.jp>
 *
 *	See ip_input.c for original log
 *
 *	Fixes:
 *		Alan Cox	:	Missing nonblock feature in ip_build_xmit.
 *		Mike Kilburn	:	htons() missing in ip_build_xmit.
25
 *		Bradford Johnson:	Fix faulty handling of some frames when
L
Linus Torvalds 已提交
26 27 28 29 30 31 32 33 34 35
 *					no route is found.
 *		Alexander Demenshin:	Missing sk/skb free in ip_queue_xmit
 *					(in case if packet not accepted by
 *					output firewall rules)
 *		Mike McLagan	:	Routing by source
 *		Alexey Kuznetsov:	use new route cache
 *		Andi Kleen:		Fix broken PMTU recovery and remove
 *					some redundant tests.
 *	Vitaly E. Lavrov	:	Transparent proxy revived after year coma.
 *		Andi Kleen	: 	Replace ip_reply with ip_send_reply.
36 37 38
 *		Andi Kleen	:	Split fast and slow ip_build_xmit path
 *					for decreased register pressure on x86
 *					and more readibility.
L
Linus Torvalds 已提交
39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54
 *		Marc Boucher	:	When call_out_firewall returns FW_QUEUE,
 *					silently drop skb instead of failing with -EPERM.
 *		Detlev Wengorz	:	Copy protocol for fragments.
 *		Hirokazu Takahashi:	HW checksumming for outgoing UDP
 *					datagrams.
 *		Hirokazu Takahashi:	sendfile() on UDP works now.
 */

#include <asm/uaccess.h>
#include <asm/system.h>
#include <linux/module.h>
#include <linux/types.h>
#include <linux/kernel.h>
#include <linux/mm.h>
#include <linux/string.h>
#include <linux/errno.h>
A
Al Viro 已提交
55
#include <linux/highmem.h>
L
Linus Torvalds 已提交
56 57 58 59 60 61 62 63 64 65 66 67 68 69 70

#include <linux/socket.h>
#include <linux/sockios.h>
#include <linux/in.h>
#include <linux/inet.h>
#include <linux/netdevice.h>
#include <linux/etherdevice.h>
#include <linux/proc_fs.h>
#include <linux/stat.h>
#include <linux/init.h>

#include <net/snmp.h>
#include <net/ip.h>
#include <net/protocol.h>
#include <net/route.h>
71
#include <net/xfrm.h>
L
Linus Torvalds 已提交
72 73 74 75 76 77 78 79 80 81 82
#include <linux/skbuff.h>
#include <net/sock.h>
#include <net/arp.h>
#include <net/icmp.h>
#include <net/checksum.h>
#include <net/inetpeer.h>
#include <linux/igmp.h>
#include <linux/netfilter_ipv4.h>
#include <linux/netfilter_bridge.h>
#include <linux/mroute.h>
#include <linux/netlink.h>
83
#include <linux/tcp.h>
L
Linus Torvalds 已提交
84

85
int sysctl_ip_default_ttl __read_mostly = IPDEFTTL;
L
Linus Torvalds 已提交
86 87 88 89 90 91 92 93

/* Generate a checksum for an outgoing IP datagram. */
__inline__ void ip_send_check(struct iphdr *iph)
{
	iph->check = 0;
	iph->check = ip_fast_csum((unsigned char *)iph, iph->ihl);
}

H
Herbert Xu 已提交
94 95 96 97 98 99
int __ip_local_out(struct sk_buff *skb)
{
	struct iphdr *iph = ip_hdr(skb);

	iph->tot_len = htons(skb->len);
	ip_send_check(iph);
100
	return nf_hook(PF_INET, NF_INET_LOCAL_OUT, skb, NULL, skb->dst->dev,
H
Herbert Xu 已提交
101 102 103 104 105 106 107 108 109 110 111 112 113 114 115
		       dst_output);
}

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

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

	return err;
}
EXPORT_SYMBOL_GPL(ip_local_out);

L
Linus Torvalds 已提交
116 117 118
/* dev_loopback_xmit for use with netfilter. */
static int ip_dev_loopback_xmit(struct sk_buff *newskb)
{
119
	skb_reset_mac_header(newskb);
120
	__skb_pull(newskb, skb_network_offset(newskb));
L
Linus Torvalds 已提交
121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136
	newskb->pkt_type = PACKET_LOOPBACK;
	newskb->ip_summed = CHECKSUM_UNNECESSARY;
	BUG_TRAP(newskb->dst);
	netif_rx(newskb);
	return 0;
}

static inline int ip_select_ttl(struct inet_sock *inet, struct dst_entry *dst)
{
	int ttl = inet->uc_ttl;

	if (ttl < 0)
		ttl = dst_metric(dst, RTAX_HOPLIMIT);
	return ttl;
}

137
/*
L
Linus Torvalds 已提交
138 139 140 141
 *		Add an ip header to a skbuff and send it out.
 *
 */
int ip_build_and_send_pkt(struct sk_buff *skb, struct sock *sk,
142
			  __be32 saddr, __be32 daddr, struct ip_options *opt)
L
Linus Torvalds 已提交
143 144 145 146 147 148
{
	struct inet_sock *inet = inet_sk(sk);
	struct rtable *rt = (struct rtable *)skb->dst;
	struct iphdr *iph;

	/* Build the IP header. */
149 150
	skb_push(skb, sizeof(struct iphdr) + (opt ? opt->optlen : 0));
	skb_reset_network_header(skb);
151
	iph = ip_hdr(skb);
L
Linus Torvalds 已提交
152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172
	iph->version  = 4;
	iph->ihl      = 5;
	iph->tos      = inet->tos;
	if (ip_dont_fragment(sk, &rt->u.dst))
		iph->frag_off = htons(IP_DF);
	else
		iph->frag_off = 0;
	iph->ttl      = ip_select_ttl(inet, &rt->u.dst);
	iph->daddr    = rt->rt_dst;
	iph->saddr    = rt->rt_src;
	iph->protocol = sk->sk_protocol;
	ip_select_ident(iph, &rt->u.dst, sk);

	if (opt && opt->optlen) {
		iph->ihl += opt->optlen>>2;
		ip_options_build(skb, opt, daddr, rt, 0);
	}

	skb->priority = sk->sk_priority;

	/* Send it out. */
H
Herbert Xu 已提交
173
	return ip_local_out(skb);
L
Linus Torvalds 已提交
174 175
}

176 177
EXPORT_SYMBOL_GPL(ip_build_and_send_pkt);

L
Linus Torvalds 已提交
178 179 180
static inline int ip_finish_output2(struct sk_buff *skb)
{
	struct dst_entry *dst = skb->dst;
181
	struct rtable *rt = (struct rtable *)dst;
L
Linus Torvalds 已提交
182
	struct net_device *dev = dst->dev;
183
	unsigned int hh_len = LL_RESERVED_SPACE(dev);
L
Linus Torvalds 已提交
184

185 186 187 188 189
	if (rt->rt_type == RTN_MULTICAST)
		IP_INC_STATS(IPSTATS_MIB_OUTMCASTPKTS);
	else if (rt->rt_type == RTN_BROADCAST)
		IP_INC_STATS(IPSTATS_MIB_OUTBCASTPKTS);

L
Linus Torvalds 已提交
190
	/* Be paranoid, rather than too clever. */
191
	if (unlikely(skb_headroom(skb) < hh_len && dev->header_ops)) {
L
Linus Torvalds 已提交
192 193 194 195 196 197 198 199 200 201 202 203 204
		struct sk_buff *skb2;

		skb2 = skb_realloc_headroom(skb, LL_RESERVED_SPACE(dev));
		if (skb2 == NULL) {
			kfree_skb(skb);
			return -ENOMEM;
		}
		if (skb->sk)
			skb_set_owner_w(skb2, skb->sk);
		kfree_skb(skb);
		skb = skb2;
	}

205 206 207
	if (dst->hh)
		return neigh_hh_output(dst->hh, skb);
	else if (dst->neighbour)
L
Linus Torvalds 已提交
208 209 210 211 212 213 214 215
		return dst->neighbour->output(skb);

	if (net_ratelimit())
		printk(KERN_DEBUG "ip_finish_output2: No header cache and no neighbour!\n");
	kfree_skb(skb);
	return -EINVAL;
}

J
John Heffner 已提交
216 217 218 219 220 221 222 223
static inline int ip_skb_dst_mtu(struct sk_buff *skb)
{
	struct inet_sock *inet = skb->sk ? inet_sk(skb->sk) : NULL;

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

P
Patrick McHardy 已提交
224
static int ip_finish_output(struct sk_buff *skb)
L
Linus Torvalds 已提交
225
{
226 227
#if defined(CONFIG_NETFILTER) && defined(CONFIG_XFRM)
	/* Policy lookup after SNAT yielded a new policy */
228 229 230 231
	if (skb->dst->xfrm != NULL) {
		IPCB(skb)->flags |= IPSKB_REROUTED;
		return dst_output(skb);
	}
232
#endif
J
John Heffner 已提交
233
	if (skb->len > ip_skb_dst_mtu(skb) && !skb_is_gso(skb))
234 235 236
		return ip_fragment(skb, ip_finish_output2);
	else
		return ip_finish_output2(skb);
L
Linus Torvalds 已提交
237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272
}

int ip_mc_output(struct sk_buff *skb)
{
	struct sock *sk = skb->sk;
	struct rtable *rt = (struct rtable*)skb->dst;
	struct net_device *dev = rt->u.dst.dev;

	/*
	 *	If the indicated interface is up and running, send the packet.
	 */
	IP_INC_STATS(IPSTATS_MIB_OUTREQUESTS);

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

	/*
	 *	Multicasts are looped back for other local users
	 */

	if (rt->rt_flags&RTCF_MULTICAST) {
		if ((!sk || inet_sk(sk)->mc_loop)
#ifdef CONFIG_IP_MROUTE
		/* Small optimization: do not loopback not local frames,
		   which returned after forwarding; they will be  dropped
		   by ip_mr_input in any case.
		   Note, that local frames are looped back to be delivered
		   to local recipients.

		   This check is duplicated in ip_mr_input at the moment.
		 */
		    && ((rt->rt_flags&RTCF_LOCAL) || !(IPCB(skb)->flags&IPSKB_FORWARDED))
#endif
		) {
			struct sk_buff *newskb = skb_clone(skb, GFP_ATOMIC);
			if (newskb)
273 274
				NF_HOOK(PF_INET, NF_INET_POST_ROUTING, newskb,
					NULL, newskb->dev,
L
Linus Torvalds 已提交
275 276 277 278 279
					ip_dev_loopback_xmit);
		}

		/* Multicasts with ttl 0 must not go beyond the host */

280
		if (ip_hdr(skb)->ttl == 0) {
L
Linus Torvalds 已提交
281 282 283 284 285 286 287 288
			kfree_skb(skb);
			return 0;
		}
	}

	if (rt->rt_flags&RTCF_BROADCAST) {
		struct sk_buff *newskb = skb_clone(skb, GFP_ATOMIC);
		if (newskb)
289
			NF_HOOK(PF_INET, NF_INET_POST_ROUTING, newskb, NULL,
L
Linus Torvalds 已提交
290 291 292
				newskb->dev, ip_dev_loopback_xmit);
	}

293
	return NF_HOOK_COND(PF_INET, NF_INET_POST_ROUTING, skb, NULL, skb->dev,
294 295
			    ip_finish_output,
			    !(IPCB(skb)->flags & IPSKB_REROUTED));
L
Linus Torvalds 已提交
296 297 298 299
}

int ip_output(struct sk_buff *skb)
{
300 301
	struct net_device *dev = skb->dst->dev;

L
Linus Torvalds 已提交
302 303
	IP_INC_STATS(IPSTATS_MIB_OUTREQUESTS);

304 305 306
	skb->dev = dev;
	skb->protocol = htons(ETH_P_IP);

307
	return NF_HOOK_COND(PF_INET, NF_INET_POST_ROUTING, skb, NULL, dev,
308
			    ip_finish_output,
309
			    !(IPCB(skb)->flags & IPSKB_REROUTED));
L
Linus Torvalds 已提交
310 311
}

312
int ip_queue_xmit(struct sk_buff *skb, int ipfragok)
L
Linus Torvalds 已提交
313
{
314
	struct sock *sk = skb->sk;
L
Linus Torvalds 已提交
315 316 317 318 319 320 321 322 323 324 325 326 327 328 329
	struct inet_sock *inet = inet_sk(sk);
	struct ip_options *opt = inet->opt;
	struct rtable *rt;
	struct iphdr *iph;

	/* Skip all of this if the packet is already routed,
	 * f.e. by something like SCTP.
	 */
	rt = (struct rtable *) skb->dst;
	if (rt != NULL)
		goto packet_routed;

	/* Make sure we can route this packet. */
	rt = (struct rtable *)__sk_dst_check(sk, 0);
	if (rt == NULL) {
A
Al Viro 已提交
330
		__be32 daddr;
L
Linus Torvalds 已提交
331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351

		/* Use correct destination address if we have options. */
		daddr = inet->daddr;
		if(opt && opt->srr)
			daddr = opt->faddr;

		{
			struct flowi fl = { .oif = sk->sk_bound_dev_if,
					    .nl_u = { .ip4_u =
						      { .daddr = daddr,
							.saddr = inet->saddr,
							.tos = RT_CONN_FLAGS(sk) } },
					    .proto = sk->sk_protocol,
					    .uli_u = { .ports =
						       { .sport = inet->sport,
							 .dport = inet->dport } } };

			/* If this fails, retransmit mechanism of transport layer will
			 * keep trying until route appears or the connection times
			 * itself out.
			 */
V
Venkat Yekkirala 已提交
352
			security_sk_classify_flow(sk, &fl);
L
Linus Torvalds 已提交
353 354 355
			if (ip_route_output_flow(&rt, &fl, sk, 0))
				goto no_route;
		}
356
		sk_setup_caps(sk, &rt->u.dst);
L
Linus Torvalds 已提交
357 358 359 360 361 362 363 364
	}
	skb->dst = dst_clone(&rt->u.dst);

packet_routed:
	if (opt && opt->is_strictroute && rt->rt_dst != rt->rt_gateway)
		goto no_route;

	/* OK, we know where to send it, allocate and build IP header. */
365 366
	skb_push(skb, sizeof(struct iphdr) + (opt ? opt->optlen : 0));
	skb_reset_network_header(skb);
367
	iph = ip_hdr(skb);
368
	*((__be16 *)iph) = htons((4 << 12) | (5 << 8) | (inet->tos & 0xff));
L
Linus Torvalds 已提交
369 370 371 372 373 374 375 376 377 378 379 380 381 382 383
	if (ip_dont_fragment(sk, &rt->u.dst) && !ipfragok)
		iph->frag_off = htons(IP_DF);
	else
		iph->frag_off = 0;
	iph->ttl      = ip_select_ttl(inet, &rt->u.dst);
	iph->protocol = sk->sk_protocol;
	iph->saddr    = rt->rt_src;
	iph->daddr    = rt->rt_dst;
	/* Transport layer set skb->h.foo itself. */

	if (opt && opt->optlen) {
		iph->ihl += opt->optlen >> 2;
		ip_options_build(skb, opt, inet->daddr, rt, 0);
	}

384
	ip_select_ident_more(iph, &rt->u.dst, sk,
385
			     (skb_shinfo(skb)->gso_segs ?: 1) - 1);
L
Linus Torvalds 已提交
386 387 388

	skb->priority = sk->sk_priority;

H
Herbert Xu 已提交
389
	return ip_local_out(skb);
L
Linus Torvalds 已提交
390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405

no_route:
	IP_INC_STATS(IPSTATS_MIB_OUTNOROUTES);
	kfree_skb(skb);
	return -EHOSTUNREACH;
}


static void ip_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
406
	to->mark = from->mark;
L
Linus Torvalds 已提交
407 408 409 410 411 412 413

	/* Copy the flags to each fragment. */
	IPCB(to)->flags = IPCB(from)->flags;

#ifdef CONFIG_NET_SCHED
	to->tc_index = from->tc_index;
#endif
414
	nf_copy(to, from);
415 416 417 418
#if defined(CONFIG_NETFILTER_XT_TARGET_TRACE) || \
    defined(CONFIG_NETFILTER_XT_TARGET_TRACE_MODULE)
	to->nf_trace = from->nf_trace;
#endif
419 420
#if defined(CONFIG_IP_VS) || defined(CONFIG_IP_VS_MODULE)
	to->ipvs_property = from->ipvs_property;
L
Linus Torvalds 已提交
421
#endif
422
	skb_copy_secmark(to, from);
L
Linus Torvalds 已提交
423 424 425 426 427 428 429 430 431
}

/*
 *	This IP datagram is too large to be sent in one piece.  Break it up into
 *	smaller pieces (each of size equal to IP header plus
 *	a block of the data of the original IP data part) that will yet fit in a
 *	single device frame, and queue such a frame for sending.
 */

432
int ip_fragment(struct sk_buff *skb, int (*output)(struct sk_buff*))
L
Linus Torvalds 已提交
433 434 435 436 437 438
{
	struct iphdr *iph;
	int raw = 0;
	int ptr;
	struct net_device *dev;
	struct sk_buff *skb2;
439
	unsigned int mtu, hlen, left, len, ll_rs, pad;
L
Linus Torvalds 已提交
440
	int offset;
441
	__be16 not_last_frag;
L
Linus Torvalds 已提交
442 443 444 445 446 447 448 449 450
	struct rtable *rt = (struct rtable*)skb->dst;
	int err = 0;

	dev = rt->u.dst.dev;

	/*
	 *	Point into the IP datagram header.
	 */

451
	iph = ip_hdr(skb);
L
Linus Torvalds 已提交
452 453

	if (unlikely((iph->frag_off & htons(IP_DF)) && !skb->local_df)) {
454
		IP_INC_STATS(IPSTATS_MIB_FRAGFAILS);
L
Linus Torvalds 已提交
455
		icmp_send(skb, ICMP_DEST_UNREACH, ICMP_FRAG_NEEDED,
J
John Heffner 已提交
456
			  htonl(ip_skb_dst_mtu(skb)));
L
Linus Torvalds 已提交
457 458 459 460 461 462 463 464 465 466
		kfree_skb(skb);
		return -EMSGSIZE;
	}

	/*
	 *	Setup starting values.
	 */

	hlen = iph->ihl * 4;
	mtu = dst_mtu(&rt->u.dst) - hlen;	/* Size of data space */
H
Herbert Xu 已提交
467
	IPCB(skb)->flags |= IPSKB_FRAG_COMPLETE;
L
Linus Torvalds 已提交
468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495

	/* When frag_list is given, use it. First, check its validity:
	 * some transformers could create wrong frag_list or break existing
	 * one, it is not prohibited. In this case fall back to copying.
	 *
	 * LATER: this step can be merged to real generation of fragments,
	 * we can switch to copy when see the first bad fragment.
	 */
	if (skb_shinfo(skb)->frag_list) {
		struct sk_buff *frag;
		int first_len = skb_pagelen(skb);

		if (first_len - hlen > mtu ||
		    ((first_len - hlen) & 7) ||
		    (iph->frag_off & htons(IP_MF|IP_OFFSET)) ||
		    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;
496 497 498 499 500 501 502 503

			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
504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522
		}

		/* Everything is OK. Generate! */

		err = 0;
		offset = 0;
		frag = skb_shinfo(skb)->frag_list;
		skb_shinfo(skb)->frag_list = NULL;
		skb->data_len = first_len - skb_headlen(skb);
		skb->len = first_len;
		iph->tot_len = htons(first_len);
		iph->frag_off = htons(IP_MF);
		ip_send_check(iph);

		for (;;) {
			/* Prepare header of the next frame,
			 * before previous one went down. */
			if (frag) {
				frag->ip_summed = CHECKSUM_NONE;
523
				skb_reset_transport_header(frag);
524 525
				__skb_push(frag, hlen);
				skb_reset_network_header(frag);
526
				memcpy(skb_network_header(frag), iph, hlen);
527
				iph = ip_hdr(frag);
L
Linus Torvalds 已提交
528 529 530 531 532 533 534 535 536 537 538 539 540 541
				iph->tot_len = htons(frag->len);
				ip_copy_metadata(frag, skb);
				if (offset == 0)
					ip_options_fragment(frag);
				offset += skb->len - hlen;
				iph->frag_off = htons(offset>>3);
				if (frag->next != NULL)
					iph->frag_off |= htons(IP_MF);
				/* Ready, complete checksum */
				ip_send_check(iph);
			}

			err = output(skb);

542 543
			if (!err)
				IP_INC_STATS(IPSTATS_MIB_FRAGCREATES);
L
Linus Torvalds 已提交
544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570
			if (err || !frag)
				break;

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

		if (err == 0) {
			IP_INC_STATS(IPSTATS_MIB_FRAGOKS);
			return 0;
		}

		while (frag) {
			skb = frag->next;
			kfree_skb(frag);
			frag = skb;
		}
		IP_INC_STATS(IPSTATS_MIB_FRAGFAILS);
		return err;
	}

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

	/* for bridged IP traffic encapsulated inside f.e. a vlan header,
571 572 573 574 575 576
	 * we need to make room for the encapsulating header
	 */
	pad = nf_bridge_pad(skb);
	ll_rs = LL_RESERVED_SPACE_EXTRA(rt->u.dst.dev, pad);
	mtu -= pad;

L
Linus Torvalds 已提交
577 578 579 580 581 582 583 584 585 586 587
	/*
	 *	Fragment the datagram.
	 */

	offset = (ntohs(iph->frag_off) & IP_OFFSET) << 3;
	not_last_frag = iph->frag_off & htons(IP_MF);

	/*
	 *	Keep copying data until we run out.
	 */

S
Stephen Hemminger 已提交
588
	while (left > 0) {
L
Linus Torvalds 已提交
589 590 591 592 593 594 595 596 597 598 599 600 601 602
		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 ((skb2 = alloc_skb(len+hlen+ll_rs, GFP_ATOMIC)) == NULL) {
603
			NETDEBUG(KERN_INFO "IP: frag: no memory for new fragment!\n");
L
Linus Torvalds 已提交
604 605 606 607 608 609 610 611 612 613 614
			err = -ENOMEM;
			goto fail;
		}

		/*
		 *	Set up data on packet
		 */

		ip_copy_metadata(skb2, skb);
		skb_reserve(skb2, ll_rs);
		skb_put(skb2, len + hlen);
615
		skb_reset_network_header(skb2);
616
		skb2->transport_header = skb2->network_header + hlen;
L
Linus Torvalds 已提交
617 618 619 620 621 622 623 624 625 626 627 628 629

		/*
		 *	Charge the memory for the fragment to any owner
		 *	it might possess
		 */

		if (skb->sk)
			skb_set_owner_w(skb2, skb->sk);

		/*
		 *	Copy the packet header into the new buffer.
		 */

630
		skb_copy_from_linear_data(skb, skb_network_header(skb2), hlen);
L
Linus Torvalds 已提交
631 632 633 634

		/*
		 *	Copy a block of the IP datagram.
		 */
635
		if (skb_copy_bits(skb, ptr, skb_transport_header(skb2), len))
L
Linus Torvalds 已提交
636 637 638 639 640 641
			BUG();
		left -= len;

		/*
		 *	Fill in the new header fields.
		 */
642
		iph = ip_hdr(skb2);
L
Linus Torvalds 已提交
643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672
		iph->frag_off = htons((offset >> 3));

		/* ANK: dirty, but effective trick. Upgrade options only if
		 * the segment to be fragmented was THE FIRST (otherwise,
		 * options are already fixed) and make it ONCE
		 * on the initial skb, so that all the following fragments
		 * will inherit fixed options.
		 */
		if (offset == 0)
			ip_options_fragment(skb);

		/*
		 *	Added AC : If we are fragmenting a fragment that's not the
		 *		   last fragment then keep MF on each bit
		 */
		if (left > 0 || not_last_frag)
			iph->frag_off |= htons(IP_MF);
		ptr += len;
		offset += len;

		/*
		 *	Put this fragment into the sending queue.
		 */
		iph->tot_len = htons(len + hlen);

		ip_send_check(iph);

		err = output(skb2);
		if (err)
			goto fail;
673 674

		IP_INC_STATS(IPSTATS_MIB_FRAGCREATES);
L
Linus Torvalds 已提交
675 676 677 678 679 680
	}
	kfree_skb(skb);
	IP_INC_STATS(IPSTATS_MIB_FRAGOKS);
	return err;

fail:
681
	kfree_skb(skb);
L
Linus Torvalds 已提交
682 683 684 685
	IP_INC_STATS(IPSTATS_MIB_FRAGFAILS);
	return err;
}

686 687
EXPORT_SYMBOL(ip_fragment);

L
Linus Torvalds 已提交
688 689 690 691 692
int
ip_generic_getfrag(void *from, char *to, int offset, int len, int odd, struct sk_buff *skb)
{
	struct iovec *iov = from;

693
	if (skb->ip_summed == CHECKSUM_PARTIAL) {
L
Linus Torvalds 已提交
694 695 696
		if (memcpy_fromiovecend(to, iov, offset, len) < 0)
			return -EFAULT;
	} else {
697
		__wsum csum = 0;
L
Linus Torvalds 已提交
698 699 700 701 702 703 704
		if (csum_partial_copy_fromiovecend(to, iov, offset, len, &csum) < 0)
			return -EFAULT;
		skb->csum = csum_block_add(skb->csum, csum, odd);
	}
	return 0;
}

705
static inline __wsum
L
Linus Torvalds 已提交
706 707 708
csum_page(struct page *page, int offset, int copy)
{
	char *kaddr;
709
	__wsum csum;
L
Linus Torvalds 已提交
710 711 712 713 714 715
	kaddr = kmap(page);
	csum = csum_partial(kaddr + offset, copy, 0);
	kunmap(page);
	return csum;
}

A
Adrian Bunk 已提交
716
static inline int ip_ufo_append_data(struct sock *sk,
717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743
			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 fragmentation 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 err;

		/* 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 */
744
		skb_reset_network_header(skb);
745 746

		/* initialize protocol header pointer */
747
		skb->transport_header = skb->network_header + fragheaderlen;
748

749
		skb->ip_summed = CHECKSUM_PARTIAL;
750 751 752 753 754 755 756 757
		skb->csum = 0;
		sk->sk_sndmsg_off = 0;
	}

	err = skb_append_datato_frags(sk,skb, getfrag, from,
			       (length - transhdrlen));
	if (!err) {
		/* specify the length of each IP datagram fragment*/
758
		skb_shinfo(skb)->gso_size = mtu - fragheaderlen;
H
Herbert Xu 已提交
759
		skb_shinfo(skb)->gso_type = SKB_GSO_UDP;
760 761 762 763 764 765 766 767 768 769 770
		__skb_queue_tail(&sk->sk_write_queue, skb);

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

L
Linus Torvalds 已提交
771 772 773 774 775
/*
 *	ip_append_data() and ip_append_page() can make one large IP datagram
 *	from many pieces of data. Each pieces will be holded on the socket
 *	until ip_push_pending_frames() is called. Each piece can be a page
 *	or non-page data.
776
 *
L
Linus Torvalds 已提交
777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820
 *	Not only UDP, other transport protocols - e.g. raw sockets - can use
 *	this interface potentially.
 *
 *	LATER: length must be adjusted by pad at tail, when it is required.
 */
int ip_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,
		   struct ipcm_cookie *ipc, struct rtable *rt,
		   unsigned int flags)
{
	struct inet_sock *inet = inet_sk(sk);
	struct sk_buff *skb;

	struct ip_options *opt = NULL;
	int hh_len;
	int exthdrlen;
	int mtu;
	int copy;
	int err;
	int offset = 0;
	unsigned int maxfraglen, fragheaderlen;
	int csummode = CHECKSUM_NONE;

	if (flags&MSG_PROBE)
		return 0;

	if (skb_queue_empty(&sk->sk_write_queue)) {
		/*
		 * setup for corking.
		 */
		opt = ipc->opt;
		if (opt) {
			if (inet->cork.opt == NULL) {
				inet->cork.opt = kmalloc(sizeof(struct ip_options) + 40, sk->sk_allocation);
				if (unlikely(inet->cork.opt == NULL))
					return -ENOBUFS;
			}
			memcpy(inet->cork.opt, opt, sizeof(struct ip_options)+opt->optlen);
			inet->cork.flags |= IPCORK_OPT;
			inet->cork.addr = ipc->addr;
		}
		dst_hold(&rt->u.dst);
J
John Heffner 已提交
821 822 823
		inet->cork.fragsize = mtu = inet->pmtudisc == IP_PMTUDISC_PROBE ?
					    rt->u.dst.dev->mtu :
					    dst_mtu(rt->u.dst.path);
L
Linus Torvalds 已提交
824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856
		inet->cork.rt = rt;
		inet->cork.length = 0;
		sk->sk_sndmsg_page = NULL;
		sk->sk_sndmsg_off = 0;
		if ((exthdrlen = rt->u.dst.header_len) != 0) {
			length += exthdrlen;
			transhdrlen += exthdrlen;
		}
	} else {
		rt = inet->cork.rt;
		if (inet->cork.flags & IPCORK_OPT)
			opt = inet->cork.opt;

		transhdrlen = 0;
		exthdrlen = 0;
		mtu = inet->cork.fragsize;
	}
	hh_len = LL_RESERVED_SPACE(rt->u.dst.dev);

	fragheaderlen = sizeof(struct iphdr) + (opt ? opt->optlen : 0);
	maxfraglen = ((mtu - fragheaderlen) & ~7) + fragheaderlen;

	if (inet->cork.length + length > 0xFFFF - fragheaderlen) {
		ip_local_error(sk, EMSGSIZE, rt->rt_dst, inet->dport, mtu-exthdrlen);
		return -EMSGSIZE;
	}

	/*
	 * transhdrlen > 0 means that this is the first fragment and we wish
	 * it won't be fragmented in the future.
	 */
	if (transhdrlen &&
	    length + fragheaderlen <= mtu &&
857
	    rt->u.dst.dev->features & NETIF_F_V4_CSUM &&
L
Linus Torvalds 已提交
858
	    !exthdrlen)
859
		csummode = CHECKSUM_PARTIAL;
L
Linus Torvalds 已提交
860 861

	inet->cork.length += length;
862 863 864
	if (((length > mtu) && (sk->sk_protocol == IPPROTO_UDP)) &&
			(rt->u.dst.dev->features & NETIF_F_UFO)) {

865 866 867 868
		err = ip_ufo_append_data(sk, getfrag, from, length, hh_len,
					 fragheaderlen, transhdrlen, mtu,
					 flags);
		if (err)
869 870 871
			goto error;
		return 0;
	}
L
Linus Torvalds 已提交
872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910

	/* So, what's going on in the loop below?
	 *
	 * We use calculated fragment length to generate chained skb,
	 * each of segments is IP fragment ready for sending to network after
	 * adding appropriate IP header.
	 */

	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 = mtu - 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;
			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 > mtu - fragheaderlen)
				datalen = maxfraglen - fragheaderlen;
			fraglen = datalen + fragheaderlen;

911
			if ((flags & MSG_MORE) &&
L
Linus Torvalds 已提交
912 913 914 915 916 917 918 919 920 921
			    !(rt->u.dst.dev->features&NETIF_F_SG))
				alloclen = mtu;
			else
				alloclen = datalen + fragheaderlen;

			/* The last fragment gets additional space at tail.
			 * Note, with MSG_MORE we overallocate on fragments,
			 * because we have no idea what fragment will be
			 * the last.
			 */
922
			if (datalen == length + fraggap)
L
Linus Torvalds 已提交
923 924 925
				alloclen += rt->u.dst.trailer_len;

			if (transhdrlen) {
926
				skb = sock_alloc_send_skb(sk,
L
Linus Torvalds 已提交
927 928 929 930 931 932
						alloclen + hh_len + 15,
						(flags & MSG_DONTWAIT), &err);
			} else {
				skb = NULL;
				if (atomic_read(&sk->sk_wmem_alloc) <=
				    2 * sk->sk_sndbuf)
933
					skb = sock_wmalloc(sk,
L
Linus Torvalds 已提交
934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952
							   alloclen + hh_len + 15, 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;
			skb_reserve(skb, hh_len);

			/*
			 *	Find where to start putting bytes.
			 */
			data = skb_put(skb, fraglen);
953
			skb_set_network_header(skb, exthdrlen);
954 955
			skb->transport_header = (skb->network_header +
						 fragheaderlen);
L
Linus Torvalds 已提交
956 957 958 959 960 961 962 963 964
			data += fragheaderlen;

			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;
965
				pskb_trim_unique(skb_prev, maxfraglen);
L
Linus Torvalds 已提交
966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994
			}

			copy = datalen - transhdrlen - fraggap;
			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;
995
			if (getfrag(from, skb_put(skb, copy),
L
Linus Torvalds 已提交
996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016
					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);
1017
					skb_fill_page_desc(skb, i, page, sk->sk_sndmsg_off, 0);
L
Linus Torvalds 已提交
1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044
					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;
1045 1046
			skb->truesize += copy;
			atomic_add(copy, &sk->sk_wmem_alloc);
L
Linus Torvalds 已提交
1047 1048 1049 1050 1051 1052 1053 1054 1055 1056
		}
		offset += copy;
		length -= copy;
	}

	return 0;

error:
	inet->cork.length -= length;
	IP_INC_STATS(IPSTATS_MIB_OUTDISCARDS);
1057
	return err;
L
Linus Torvalds 已提交
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 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103
}

ssize_t	ip_append_page(struct sock *sk, struct page *page,
		       int offset, size_t size, int flags)
{
	struct inet_sock *inet = inet_sk(sk);
	struct sk_buff *skb;
	struct rtable *rt;
	struct ip_options *opt = NULL;
	int hh_len;
	int mtu;
	int len;
	int err;
	unsigned int maxfraglen, fragheaderlen, fraggap;

	if (inet->hdrincl)
		return -EPERM;

	if (flags&MSG_PROBE)
		return 0;

	if (skb_queue_empty(&sk->sk_write_queue))
		return -EINVAL;

	rt = inet->cork.rt;
	if (inet->cork.flags & IPCORK_OPT)
		opt = inet->cork.opt;

	if (!(rt->u.dst.dev->features&NETIF_F_SG))
		return -EOPNOTSUPP;

	hh_len = LL_RESERVED_SPACE(rt->u.dst.dev);
	mtu = inet->cork.fragsize;

	fragheaderlen = sizeof(struct iphdr) + (opt ? opt->optlen : 0);
	maxfraglen = ((mtu - fragheaderlen) & ~7) + fragheaderlen;

	if (inet->cork.length + size > 0xFFFF - fragheaderlen) {
		ip_local_error(sk, EMSGSIZE, rt->rt_dst, inet->dport, mtu);
		return -EMSGSIZE;
	}

	if ((skb = skb_peek_tail(&sk->sk_write_queue)) == NULL)
		return -EINVAL;

	inet->cork.length += size;
1104
	if ((sk->sk_protocol == IPPROTO_UDP) &&
1105 1106
	    (rt->u.dst.dev->features & NETIF_F_UFO)) {
		skb_shinfo(skb)->gso_size = mtu - fragheaderlen;
H
Herbert Xu 已提交
1107
		skb_shinfo(skb)->gso_type = SKB_GSO_UDP;
1108
	}
1109

L
Linus Torvalds 已提交
1110 1111 1112 1113

	while (size > 0) {
		int i;

H
Herbert Xu 已提交
1114
		if (skb_is_gso(skb))
1115 1116 1117 1118 1119 1120 1121 1122
			len = size;
		else {

			/* Check if the remaining data fits into current packet. */
			len = mtu - skb->len;
			if (len < size)
				len = maxfraglen - skb->len;
		}
L
Linus Torvalds 已提交
1123 1124 1125 1126 1127
		if (len <= 0) {
			struct sk_buff *skb_prev;
			int alloclen;

			skb_prev = skb;
1128
			fraggap = skb_prev->len - maxfraglen;
L
Linus Torvalds 已提交
1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146

			alloclen = fragheaderlen + hh_len + fraggap + 15;
			skb = sock_wmalloc(sk, alloclen, 1, sk->sk_allocation);
			if (unlikely(!skb)) {
				err = -ENOBUFS;
				goto error;
			}

			/*
			 *	Fill in the control structures
			 */
			skb->ip_summed = CHECKSUM_NONE;
			skb->csum = 0;
			skb_reserve(skb, hh_len);

			/*
			 *	Find where to start putting bytes.
			 */
1147
			skb_put(skb, fragheaderlen + fraggap);
1148
			skb_reset_network_header(skb);
1149 1150
			skb->transport_header = (skb->network_header +
						 fragheaderlen);
L
Linus Torvalds 已提交
1151
			if (fraggap) {
1152 1153
				skb->csum = skb_copy_and_csum_bits(skb_prev,
								   maxfraglen,
1154
						    skb_transport_header(skb),
1155
								   fraggap, 0);
L
Linus Torvalds 已提交
1156 1157
				skb_prev->csum = csum_sub(skb_prev->csum,
							  skb->csum);
1158
				pskb_trim_unique(skb_prev, maxfraglen);
L
Linus Torvalds 已提交
1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181
			}

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

		i = skb_shinfo(skb)->nr_frags;
		if (len > size)
			len = size;
		if (skb_can_coalesce(skb, i, page, offset)) {
			skb_shinfo(skb)->frags[i-1].size += len;
		} else if (i < MAX_SKB_FRAGS) {
			get_page(page);
			skb_fill_page_desc(skb, i, page, offset, len);
		} else {
			err = -EMSGSIZE;
			goto error;
		}

		if (skb->ip_summed == CHECKSUM_NONE) {
1182
			__wsum csum;
L
Linus Torvalds 已提交
1183 1184 1185 1186 1187 1188
			csum = csum_page(page, offset, len);
			skb->csum = csum_block_add(skb->csum, csum, skb->len);
		}

		skb->len += len;
		skb->data_len += len;
1189 1190
		skb->truesize += len;
		atomic_add(len, &sk->sk_wmem_alloc);
L
Linus Torvalds 已提交
1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201
		offset += len;
		size -= len;
	}
	return 0;

error:
	inet->cork.length -= size;
	IP_INC_STATS(IPSTATS_MIB_OUTDISCARDS);
	return err;
}

1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212
static void ip_cork_release(struct inet_sock *inet)
{
	inet->cork.flags &= ~IPCORK_OPT;
	kfree(inet->cork.opt);
	inet->cork.opt = NULL;
	if (inet->cork.rt) {
		ip_rt_put(inet->cork.rt);
		inet->cork.rt = NULL;
	}
}

L
Linus Torvalds 已提交
1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224
/*
 *	Combined all pending IP fragments on the socket as one IP datagram
 *	and push them out.
 */
int ip_push_pending_frames(struct sock *sk)
{
	struct sk_buff *skb, *tmp_skb;
	struct sk_buff **tail_skb;
	struct inet_sock *inet = inet_sk(sk);
	struct ip_options *opt = NULL;
	struct rtable *rt = inet->cork.rt;
	struct iphdr *iph;
1225
	__be16 df = 0;
L
Linus Torvalds 已提交
1226 1227 1228 1229 1230 1231 1232 1233
	__u8 ttl;
	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 */
1234
	if (skb->data < skb_network_header(skb))
1235
		__skb_pull(skb, skb_network_offset(skb));
L
Linus Torvalds 已提交
1236
	while ((tmp_skb = __skb_dequeue(&sk->sk_write_queue)) != NULL) {
1237
		__skb_pull(tmp_skb, skb_network_header_len(skb));
L
Linus Torvalds 已提交
1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251
		*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;
	}

	/* Unless user demanded real pmtu discovery (IP_PMTUDISC_DO), we allow
	 * to fragment the frame generated here. No matter, what transforms
	 * how transforms change size of the packet, it will come out.
	 */
J
John Heffner 已提交
1252
	if (inet->pmtudisc < IP_PMTUDISC_DO)
L
Linus Torvalds 已提交
1253 1254 1255 1256 1257
		skb->local_df = 1;

	/* DF bit is set when we want to see DF on outgoing frames.
	 * If local_df is set too, we still allow to fragment this frame
	 * locally. */
J
John Heffner 已提交
1258
	if (inet->pmtudisc >= IP_PMTUDISC_DO ||
L
Linus Torvalds 已提交
1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279
	    (skb->len <= dst_mtu(&rt->u.dst) &&
	     ip_dont_fragment(sk, &rt->u.dst)))
		df = htons(IP_DF);

	if (inet->cork.flags & IPCORK_OPT)
		opt = inet->cork.opt;

	if (rt->rt_type == RTN_MULTICAST)
		ttl = inet->mc_ttl;
	else
		ttl = ip_select_ttl(inet, &rt->u.dst);

	iph = (struct iphdr *)skb->data;
	iph->version = 4;
	iph->ihl = 5;
	if (opt) {
		iph->ihl += opt->optlen>>2;
		ip_options_build(skb, opt, inet->cork.addr, rt, 0);
	}
	iph->tos = inet->tos;
	iph->frag_off = df;
1280
	ip_select_ident(iph, &rt->u.dst, sk);
L
Linus Torvalds 已提交
1281 1282 1283 1284 1285 1286 1287 1288
	iph->ttl = ttl;
	iph->protocol = sk->sk_protocol;
	iph->saddr = rt->rt_src;
	iph->daddr = rt->rt_dst;

	skb->priority = sk->sk_priority;
	skb->dst = dst_clone(&rt->u.dst);

1289 1290 1291 1292
	if (iph->protocol == IPPROTO_ICMP)
		icmp_out_count(((struct icmphdr *)
			skb_transport_header(skb))->type);

L
Linus Torvalds 已提交
1293
	/* Netfilter gets whole the not fragmented skb. */
H
Herbert Xu 已提交
1294
	err = ip_local_out(skb);
L
Linus Torvalds 已提交
1295 1296 1297 1298 1299 1300 1301 1302
	if (err) {
		if (err > 0)
			err = inet->recverr ? net_xmit_errno(err) : 0;
		if (err)
			goto error;
	}

out:
1303
	ip_cork_release(inet);
L
Linus Torvalds 已提交
1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320
	return err;

error:
	IP_INC_STATS(IPSTATS_MIB_OUTDISCARDS);
	goto out;
}

/*
 *	Throw away all pending data on the socket.
 */
void ip_flush_pending_frames(struct sock *sk)
{
	struct sk_buff *skb;

	while ((skb = __skb_dequeue_tail(&sk->sk_write_queue)) != NULL)
		kfree_skb(skb);

1321
	ip_cork_release(inet_sk(sk));
L
Linus Torvalds 已提交
1322 1323 1324 1325 1326 1327
}


/*
 *	Fetch data from kernel space and fill in checksum if needed.
 */
1328
static int ip_reply_glue_bits(void *dptr, char *to, int offset,
L
Linus Torvalds 已提交
1329 1330
			      int len, int odd, struct sk_buff *skb)
{
1331
	__wsum csum;
L
Linus Torvalds 已提交
1332 1333 1334

	csum = csum_partial_copy_nocheck(dptr+offset, to, len, 0);
	skb->csum = csum_block_add(skb->csum, csum, odd);
1335
	return 0;
L
Linus Torvalds 已提交
1336 1337
}

1338
/*
L
Linus Torvalds 已提交
1339 1340 1341
 *	Generic function to send a packet as reply to another packet.
 *	Used to send TCP resets so far. ICMP should use this function too.
 *
1342
 *	Should run single threaded per socket because it uses the sock
L
Linus Torvalds 已提交
1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353
 *     	structure to pass arguments.
 */
void ip_send_reply(struct sock *sk, struct sk_buff *skb, struct ip_reply_arg *arg,
		   unsigned int len)
{
	struct inet_sock *inet = inet_sk(sk);
	struct {
		struct ip_options	opt;
		char			data[40];
	} replyopts;
	struct ipcm_cookie ipc;
A
Al Viro 已提交
1354
	__be32 daddr;
L
Linus Torvalds 已提交
1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370
	struct rtable *rt = (struct rtable*)skb->dst;

	if (ip_options_echo(&replyopts.opt, skb))
		return;

	daddr = ipc.addr = rt->rt_src;
	ipc.opt = NULL;

	if (replyopts.opt.optlen) {
		ipc.opt = &replyopts.opt;

		if (ipc.opt->srr)
			daddr = replyopts.opt.faddr;
	}

	{
1371 1372
		struct flowi fl = { .oif = arg->bound_dev_if,
				    .nl_u = { .ip4_u =
L
Linus Torvalds 已提交
1373 1374
					      { .daddr = daddr,
						.saddr = rt->rt_spec_dst,
1375
						.tos = RT_TOS(ip_hdr(skb)->tos) } },
L
Linus Torvalds 已提交
1376 1377
				    /* Not quite clean, but right. */
				    .uli_u = { .ports =
1378 1379
					       { .sport = tcp_hdr(skb)->dest,
						 .dport = tcp_hdr(skb)->source } },
L
Linus Torvalds 已提交
1380
				    .proto = sk->sk_protocol };
V
Venkat Yekkirala 已提交
1381
		security_skb_classify_flow(skb, &fl);
L
Linus Torvalds 已提交
1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392
		if (ip_route_output_key(&rt, &fl))
			return;
	}

	/* And let IP do all the hard work.

	   This chunk is not reenterable, hence spinlock.
	   Note that it uses the fact, that this function is called
	   with locally disabled BH and that sk cannot be already spinlocked.
	 */
	bh_lock_sock(sk);
1393
	inet->tos = ip_hdr(skb)->tos;
L
Linus Torvalds 已提交
1394
	sk->sk_priority = skb->priority;
1395
	sk->sk_protocol = ip_hdr(skb)->protocol;
1396
	sk->sk_bound_dev_if = arg->bound_dev_if;
L
Linus Torvalds 已提交
1397 1398 1399 1400
	ip_append_data(sk, ip_reply_glue_bits, arg->iov->iov_base, len, 0,
		       &ipc, rt, MSG_DONTWAIT);
	if ((skb = skb_peek(&sk->sk_write_queue)) != NULL) {
		if (arg->csumoffset >= 0)
1401 1402 1403
			*((__sum16 *)skb_transport_header(skb) +
			  arg->csumoffset) = csum_fold(csum_add(skb->csum,
								arg->csum));
L
Linus Torvalds 已提交
1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425
		skb->ip_summed = CHECKSUM_NONE;
		ip_push_pending_frames(sk);
	}

	bh_unlock_sock(sk);

	ip_rt_put(rt);
}

void __init ip_init(void)
{
	ip_rt_init();
	inet_initpeers();

#if defined(CONFIG_IP_MULTICAST) && defined(CONFIG_PROC_FS)
	igmp_mc_proc_init();
#endif
}

EXPORT_SYMBOL(ip_generic_getfrag);
EXPORT_SYMBOL(ip_queue_xmit);
EXPORT_SYMBOL(ip_send_check);