af_inet.c 48.1 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7
/*
 * 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.
 *
 *		PF_INET protocol family socket handler.
 *
8
 * Authors:	Ross Biro
L
Linus Torvalds 已提交
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 41 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
 *		Fred N. van Kempen, <waltje@uWalt.NL.Mugnet.ORG>
 *		Florian La Roche, <flla@stud.uni-sb.de>
 *		Alan Cox, <A.Cox@swansea.ac.uk>
 *
 * Changes (see also sock.c)
 *
 *		piggy,
 *		Karl Knutson	:	Socket protocol table
 *		A.N.Kuznetsov	:	Socket death error in accept().
 *		John Richardson :	Fix non blocking error in connect()
 *					so sockets that fail to connect
 *					don't return -EINPROGRESS.
 *		Alan Cox	:	Asynchronous I/O support
 *		Alan Cox	:	Keep correct socket pointer on sock
 *					structures
 *					when accept() ed
 *		Alan Cox	:	Semantics of SO_LINGER aren't state
 *					moved to close when you look carefully.
 *					With this fixed and the accept bug fixed
 *					some RPC stuff seems happier.
 *		Niibe Yutaka	:	4.4BSD style write async I/O
 *		Alan Cox,
 *		Tony Gale 	:	Fixed reuse semantics.
 *		Alan Cox	:	bind() shouldn't abort existing but dead
 *					sockets. Stops FTP netin:.. I hope.
 *		Alan Cox	:	bind() works correctly for RAW sockets.
 *					Note that FreeBSD at least was broken
 *					in this respect so be careful with
 *					compatibility tests...
 *		Alan Cox	:	routing cache support
 *		Alan Cox	:	memzero the socket structure for
 *					compactness.
 *		Matt Day	:	nonblock connect error handler
 *		Alan Cox	:	Allow large numbers of pending sockets
 *					(eg for big web sites), but only if
 *					specifically application requested.
 *		Alan Cox	:	New buffering throughout IP. Used
 *					dumbly.
 *		Alan Cox	:	New buffering now used smartly.
 *		Alan Cox	:	BSD rather than common sense
 *					interpretation of listen.
 *		Germano Caronni	:	Assorted small races.
 *		Alan Cox	:	sendmsg/recvmsg basic support.
 *		Alan Cox	:	Only sendmsg/recvmsg now supported.
 *		Alan Cox	:	Locked down bind (see security list).
 *		Alan Cox	:	Loosened bind a little.
 *		Mike McLagan	:	ADD/DEL DLCI Ioctls
 *	Willy Konynenberg	:	Transparent proxying support.
 *		David S. Miller	:	New socket lookup architecture.
 *					Some other random speedups.
 *		Cyrus Durgin	:	Cleaned up file for kmod hacks.
 *		Andi Kleen	:	Fix inet_stream_connect TCP race.
 *
 *		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.
 */

68 69
#define pr_fmt(fmt) "IPv4: " fmt

H
Herbert Xu 已提交
70
#include <linux/err.h>
L
Linus Torvalds 已提交
71 72 73 74 75
#include <linux/errno.h>
#include <linux/types.h>
#include <linux/socket.h>
#include <linux/in.h>
#include <linux/kernel.h>
76
#include <linux/kmod.h>
L
Linus Torvalds 已提交
77 78 79 80 81
#include <linux/sched.h>
#include <linux/timer.h>
#include <linux/string.h>
#include <linux/sockios.h>
#include <linux/net.h>
82
#include <linux/capability.h>
L
Linus Torvalds 已提交
83 84 85 86 87 88 89
#include <linux/fcntl.h>
#include <linux/mm.h>
#include <linux/interrupt.h>
#include <linux/stat.h>
#include <linux/init.h>
#include <linux/poll.h>
#include <linux/netfilter_ipv4.h>
90
#include <linux/random.h>
91
#include <linux/slab.h>
L
Linus Torvalds 已提交
92

93
#include <linux/uaccess.h>
L
Linus Torvalds 已提交
94 95 96

#include <linux/inet.h>
#include <linux/igmp.h>
97
#include <linux/inetdevice.h>
L
Linus Torvalds 已提交
98
#include <linux/netdevice.h>
H
Herbert Xu 已提交
99
#include <net/checksum.h>
L
Linus Torvalds 已提交
100 101 102 103 104
#include <net/ip.h>
#include <net/protocol.h>
#include <net/arp.h>
#include <net/route.h>
#include <net/ip_fib.h>
105
#include <net/inet_connection_sock.h>
L
Linus Torvalds 已提交
106 107
#include <net/tcp.h>
#include <net/udp.h>
108
#include <net/udplite.h>
109
#include <net/ping.h>
L
Linus Torvalds 已提交
110 111 112 113 114
#include <linux/skbuff.h>
#include <net/sock.h>
#include <net/raw.h>
#include <net/icmp.h>
#include <net/inet_common.h>
115
#include <net/ip_tunnels.h>
L
Linus Torvalds 已提交
116
#include <net/xfrm.h>
117
#include <net/net_namespace.h>
118
#include <net/secure_seq.h>
L
Linus Torvalds 已提交
119 120 121
#ifdef CONFIG_IP_MROUTE
#include <linux/mroute.h>
#endif
122
#include <net/l3mdev.h>
L
Linus Torvalds 已提交
123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139


/* The inetsw table contains everything that inet_create needs to
 * build a new socket.
 */
static struct list_head inetsw[SOCK_MAX];
static DEFINE_SPINLOCK(inetsw_lock);

/* New destruction routine */

void inet_sock_destruct(struct sock *sk)
{
	struct inet_sock *inet = inet_sk(sk);

	__skb_queue_purge(&sk->sk_receive_queue);
	__skb_queue_purge(&sk->sk_error_queue);

H
Hideo Aoki 已提交
140 141
	sk_mem_reclaim(sk);

L
Linus Torvalds 已提交
142
	if (sk->sk_type == SOCK_STREAM && sk->sk_state != TCP_CLOSE) {
E
Eric Dumazet 已提交
143
		pr_err("Attempt to release TCP socket in state %d %p\n",
L
Linus Torvalds 已提交
144 145 146 147
		       sk->sk_state, sk);
		return;
	}
	if (!sock_flag(sk, SOCK_DEAD)) {
E
Eric Dumazet 已提交
148
		pr_err("Attempt to release alive inet socket %p\n", sk);
L
Linus Torvalds 已提交
149 150 151
		return;
	}

152
	WARN_ON(atomic_read(&sk->sk_rmem_alloc));
153
	WARN_ON(refcount_read(&sk->sk_wmem_alloc));
154 155
	WARN_ON(sk->sk_wmem_queued);
	WARN_ON(sk->sk_forward_alloc);
L
Linus Torvalds 已提交
156

157
	kfree(rcu_dereference_protected(inet->inet_opt, 1));
E
Eric Dumazet 已提交
158
	dst_release(rcu_dereference_check(sk->sk_dst_cache, 1));
D
David S. Miller 已提交
159
	dst_release(sk->sk_rx_dst);
160
	sk_refcnt_debug_dec(sk);
L
Linus Torvalds 已提交
161
}
E
Eric Dumazet 已提交
162
EXPORT_SYMBOL(inet_sock_destruct);
L
Linus Torvalds 已提交
163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179

/*
 *	The routines beyond this point handle the behaviour of an AF_INET
 *	socket object. Mostly it punts to the subprotocols of IP to do
 *	the work.
 */

/*
 *	Automatically bind an unbound socket.
 */

static int inet_autobind(struct sock *sk)
{
	struct inet_sock *inet;
	/* We may need to bind the socket. */
	lock_sock(sk);
	inet = inet_sk(sk);
E
Eric Dumazet 已提交
180
	if (!inet->inet_num) {
L
Linus Torvalds 已提交
181 182 183 184
		if (sk->sk_prot->get_port(sk, 0)) {
			release_sock(sk);
			return -EAGAIN;
		}
E
Eric Dumazet 已提交
185
		inet->inet_sport = htons(inet->inet_num);
L
Linus Torvalds 已提交
186 187 188 189 190 191 192 193 194 195 196 197
	}
	release_sock(sk);
	return 0;
}

/*
 *	Move a socket into listening state.
 */
int inet_listen(struct socket *sock, int backlog)
{
	struct sock *sk = sock->sk;
	unsigned char old_state;
198
	int err, tcp_fastopen;
L
Linus Torvalds 已提交
199 200 201 202 203 204 205 206 207 208 209 210 211 212 213

	lock_sock(sk);

	err = -EINVAL;
	if (sock->state != SS_UNCONNECTED || sock->type != SOCK_STREAM)
		goto out;

	old_state = sk->sk_state;
	if (!((1 << old_state) & (TCPF_CLOSE | TCPF_LISTEN)))
		goto out;

	/* Really, if the socket is already in listen state
	 * we can only allow the backlog to be adjusted.
	 */
	if (old_state != TCP_LISTEN) {
214
		/* Enable TFO w/o requiring TCP_FASTOPEN socket option.
215
		 * Note that only TCP sockets (SOCK_STREAM) will reach here.
216 217 218
		 * Also fastopen backlog may already been set via the option
		 * because the socket was in TCP_LISTEN state previously but
		 * was shutdown() rather than close().
219
		 */
220 221 222
		tcp_fastopen = sock_net(sk)->ipv4.sysctl_tcp_fastopen;
		if ((tcp_fastopen & TFO_SERVER_WO_SOCKOPT1) &&
		    (tcp_fastopen & TFO_SERVER_ENABLE) &&
223
		    !inet_csk(sk)->icsk_accept_queue.fastopenq.max_qlen) {
224
			fastopen_queue_tune(sk, backlog);
225
			tcp_fastopen_init_key_once(sock_net(sk));
226
		}
227

228
		err = inet_csk_listen_start(sk, backlog);
L
Linus Torvalds 已提交
229 230 231 232 233 234 235 236 237 238
		if (err)
			goto out;
	}
	sk->sk_max_ack_backlog = backlog;
	err = 0;

out:
	release_sock(sk);
	return err;
}
E
Eric Dumazet 已提交
239
EXPORT_SYMBOL(inet_listen);
L
Linus Torvalds 已提交
240 241 242 243 244

/*
 *	Create an inet socket.
 */

245 246
static int inet_create(struct net *net, struct socket *sock, int protocol,
		       int kern)
L
Linus Torvalds 已提交
247 248 249 250 251 252
{
	struct sock *sk;
	struct inet_protosw *answer;
	struct inet_sock *inet;
	struct proto *answer_prot;
	unsigned char answer_flags;
253
	int try_loading_module = 0;
254
	int err;
L
Linus Torvalds 已提交
255

256 257 258
	if (protocol < 0 || protocol >= IPPROTO_MAX)
		return -EINVAL;

L
Linus Torvalds 已提交
259 260 261
	sock->state = SS_UNCONNECTED;

	/* Look for the requested type/protocol pair. */
262
lookup_protocol:
263
	err = -ESOCKTNOSUPPORT;
L
Linus Torvalds 已提交
264
	rcu_read_lock();
265
	list_for_each_entry_rcu(answer, &inetsw[sock->type], list) {
L
Linus Torvalds 已提交
266

267
		err = 0;
L
Linus Torvalds 已提交
268 269 270 271 272 273 274 275 276 277 278 279 280
		/* Check the non-wild match. */
		if (protocol == answer->protocol) {
			if (protocol != IPPROTO_IP)
				break;
		} else {
			/* Check for the two wild cases. */
			if (IPPROTO_IP == protocol) {
				protocol = answer->protocol;
				break;
			}
			if (IPPROTO_IP == answer->protocol)
				break;
		}
281
		err = -EPROTONOSUPPORT;
L
Linus Torvalds 已提交
282 283
	}

284
	if (unlikely(err)) {
285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305
		if (try_loading_module < 2) {
			rcu_read_unlock();
			/*
			 * Be more specific, e.g. net-pf-2-proto-132-type-1
			 * (net-pf-PF_INET-proto-IPPROTO_SCTP-type-SOCK_STREAM)
			 */
			if (++try_loading_module == 1)
				request_module("net-pf-%d-proto-%d-type-%d",
					       PF_INET, protocol, sock->type);
			/*
			 * Fall back to generic, e.g. net-pf-2-proto-132
			 * (net-pf-PF_INET-proto-IPPROTO_SCTP)
			 */
			else
				request_module("net-pf-%d-proto-%d",
					       PF_INET, protocol);
			goto lookup_protocol;
		} else
			goto out_rcu_unlock;
	}

L
Linus Torvalds 已提交
306
	err = -EPERM;
307 308
	if (sock->type == SOCK_RAW && !kern &&
	    !ns_capable(net->user_ns, CAP_NET_RAW))
L
Linus Torvalds 已提交
309 310 311 312 313 314 315
		goto out_rcu_unlock;

	sock->ops = answer->ops;
	answer_prot = answer->prot;
	answer_flags = answer->flags;
	rcu_read_unlock();

316
	WARN_ON(!answer_prot->slab);
L
Linus Torvalds 已提交
317 318

	err = -ENOBUFS;
319
	sk = sk_alloc(net, PF_INET, GFP_KERNEL, answer_prot, kern);
320
	if (!sk)
L
Linus Torvalds 已提交
321 322 323 324
		goto out;

	err = 0;
	if (INET_PROTOSW_REUSE & answer_flags)
325
		sk->sk_reuse = SK_CAN_REUSE;
L
Linus Torvalds 已提交
326 327

	inet = inet_sk(sk);
328
	inet->is_icsk = (INET_PROTOSW_ICSK & answer_flags) != 0;
L
Linus Torvalds 已提交
329

330 331
	inet->nodefrag = 0;

L
Linus Torvalds 已提交
332
	if (SOCK_RAW == sock->type) {
E
Eric Dumazet 已提交
333
		inet->inet_num = protocol;
L
Linus Torvalds 已提交
334 335 336 337
		if (IPPROTO_RAW == protocol)
			inet->hdrincl = 1;
	}

338
	if (net->ipv4.sysctl_ip_no_pmtu_disc)
L
Linus Torvalds 已提交
339 340 341 342
		inet->pmtudisc = IP_PMTUDISC_DONT;
	else
		inet->pmtudisc = IP_PMTUDISC_WANT;

E
Eric Dumazet 已提交
343
	inet->inet_id = 0;
L
Linus Torvalds 已提交
344 345 346 347 348 349 350 351 352 353

	sock_init_data(sock, sk);

	sk->sk_destruct	   = inet_sock_destruct;
	sk->sk_protocol	   = protocol;
	sk->sk_backlog_rcv = sk->sk_prot->backlog_rcv;

	inet->uc_ttl	= -1;
	inet->mc_loop	= 1;
	inet->mc_ttl	= 1;
354
	inet->mc_all	= 1;
L
Linus Torvalds 已提交
355 356
	inet->mc_index	= 0;
	inet->mc_list	= NULL;
357
	inet->rcv_tos	= 0;
L
Linus Torvalds 已提交
358

359
	sk_refcnt_debug_inc(sk);
L
Linus Torvalds 已提交
360

E
Eric Dumazet 已提交
361
	if (inet->inet_num) {
L
Linus Torvalds 已提交
362 363 364 365 366
		/* It assumes that any protocol which allows
		 * the user to assign a number at socket
		 * creation time automatically
		 * shares.
		 */
E
Eric Dumazet 已提交
367
		inet->inet_sport = htons(inet->inet_num);
L
Linus Torvalds 已提交
368
		/* Add to protocol hash chains. */
369 370 371 372 373
		err = sk->sk_prot->hash(sk);
		if (err) {
			sk_common_release(sk);
			goto out;
		}
L
Linus Torvalds 已提交
374 375 376 377
	}

	if (sk->sk_prot->init) {
		err = sk->sk_prot->init(sk);
378 379 380 381 382 383 384 385 386
		if (err) {
			sk_common_release(sk);
			goto out;
		}
	}

	if (!kern) {
		err = BPF_CGROUP_RUN_PROG_INET_SOCK(sk);
		if (err) {
L
Linus Torvalds 已提交
387
			sk_common_release(sk);
388 389
			goto out;
		}
L
Linus Torvalds 已提交
390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429
	}
out:
	return err;
out_rcu_unlock:
	rcu_read_unlock();
	goto out;
}


/*
 *	The peer socket should always be NULL (or else). When we call this
 *	function we are destroying the object and from then on nobody
 *	should refer to it.
 */
int inet_release(struct socket *sock)
{
	struct sock *sk = sock->sk;

	if (sk) {
		long timeout;

		/* Applications forget to leave groups before exiting */
		ip_mc_drop_socket(sk);

		/* If linger is set, we don't return until the close
		 * is complete.  Otherwise we return immediately. The
		 * actually closing is done the same either way.
		 *
		 * If the close is due to the process exiting, we never
		 * linger..
		 */
		timeout = 0;
		if (sock_flag(sk, SOCK_LINGER) &&
		    !(current->flags & PF_EXITING))
			timeout = sk->sk_lingertime;
		sock->sk = NULL;
		sk->sk_prot->close(sk, timeout);
	}
	return 0;
}
E
Eric Dumazet 已提交
430
EXPORT_SYMBOL(inet_release);
L
Linus Torvalds 已提交
431 432 433 434 435 436

int inet_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len)
{
	struct sockaddr_in *addr = (struct sockaddr_in *)uaddr;
	struct sock *sk = sock->sk;
	struct inet_sock *inet = inet_sk(sk);
437
	struct net *net = sock_net(sk);
L
Linus Torvalds 已提交
438 439
	unsigned short snum;
	int chk_addr_ret;
D
David Ahern 已提交
440
	u32 tb_id = RT_TABLE_LOCAL;
L
Linus Torvalds 已提交
441 442 443 444 445 446 447 448 449 450 451
	int err;

	/* If the socket has its own bind function then use it. (RAW) */
	if (sk->sk_prot->bind) {
		err = sk->sk_prot->bind(sk, uaddr, addr_len);
		goto out;
	}
	err = -EINVAL;
	if (addr_len < sizeof(struct sockaddr_in))
		goto out;

452
	if (addr->sin_family != AF_INET) {
453 454 455
		/* Compatibility games : accept AF_UNSPEC (mapped to AF_INET)
		 * only if s_addr is INADDR_ANY.
		 */
456
		err = -EAFNOSUPPORT;
457 458 459
		if (addr->sin_family != AF_UNSPEC ||
		    addr->sin_addr.s_addr != htonl(INADDR_ANY))
			goto out;
460
	}
461

462
	tb_id = l3mdev_fib_table_by_index(net, sk->sk_bound_dev_if) ? : tb_id;
D
David Ahern 已提交
463
	chk_addr_ret = inet_addr_type_table(net, addr->sin_addr.s_addr, tb_id);
L
Linus Torvalds 已提交
464 465 466 467 468 469 470 471 472

	/* Not specified by any standard per-se, however it breaks too
	 * many applications when removed.  It is unfortunate since
	 * allowing applications to make a non-local bind solves
	 * several problems with systems using dynamic addressing.
	 * (ie. your servers still start up even if your ISDN link
	 *  is temporarily down)
	 */
	err = -EADDRNOTAVAIL;
473
	if (!net->ipv4.sysctl_ip_nonlocal_bind &&
474
	    !(inet->freebind || inet->transparent) &&
A
Al Viro 已提交
475
	    addr->sin_addr.s_addr != htonl(INADDR_ANY) &&
L
Linus Torvalds 已提交
476 477 478 479 480 481 482
	    chk_addr_ret != RTN_LOCAL &&
	    chk_addr_ret != RTN_MULTICAST &&
	    chk_addr_ret != RTN_BROADCAST)
		goto out;

	snum = ntohs(addr->sin_port);
	err = -EACCES;
483
	if (snum && snum < inet_prot_sock(net) &&
484
	    !ns_capable(net->user_ns, CAP_NET_BIND_SERVICE))
L
Linus Torvalds 已提交
485 486 487 488 489 490 491 492 493 494 495 496 497
		goto out;

	/*      We keep a pair of addresses. rcv_saddr is the one
	 *      used by hash lookups, and saddr is used for transmit.
	 *
	 *      In the BSD API these are the same except where it
	 *      would be illegal to use them (multicast/broadcast) in
	 *      which case the sending device address is used.
	 */
	lock_sock(sk);

	/* Check these errors (active socket, double bind). */
	err = -EINVAL;
E
Eric Dumazet 已提交
498
	if (sk->sk_state != TCP_CLOSE || inet->inet_num)
L
Linus Torvalds 已提交
499 500
		goto out_release_sock;

E
Eric Dumazet 已提交
501
	inet->inet_rcv_saddr = inet->inet_saddr = addr->sin_addr.s_addr;
L
Linus Torvalds 已提交
502
	if (chk_addr_ret == RTN_MULTICAST || chk_addr_ret == RTN_BROADCAST)
E
Eric Dumazet 已提交
503
		inet->inet_saddr = 0;  /* Use device */
L
Linus Torvalds 已提交
504 505

	/* Make sure we are allowed to bind here. */
506 507
	if ((snum || !inet->bind_address_no_port) &&
	    sk->sk_prot->get_port(sk, snum)) {
E
Eric Dumazet 已提交
508
		inet->inet_saddr = inet->inet_rcv_saddr = 0;
L
Linus Torvalds 已提交
509 510 511 512
		err = -EADDRINUSE;
		goto out_release_sock;
	}

E
Eric Dumazet 已提交
513
	if (inet->inet_rcv_saddr)
L
Linus Torvalds 已提交
514 515 516
		sk->sk_userlocks |= SOCK_BINDADDR_LOCK;
	if (snum)
		sk->sk_userlocks |= SOCK_BINDPORT_LOCK;
E
Eric Dumazet 已提交
517 518 519
	inet->inet_sport = htons(inet->inet_num);
	inet->inet_daddr = 0;
	inet->inet_dport = 0;
L
Linus Torvalds 已提交
520 521 522 523 524 525 526
	sk_dst_reset(sk);
	err = 0;
out_release_sock:
	release_sock(sk);
out:
	return err;
}
E
Eric Dumazet 已提交
527
EXPORT_SYMBOL(inet_bind);
L
Linus Torvalds 已提交
528

D
Daniel Baluta 已提交
529
int inet_dgram_connect(struct socket *sock, struct sockaddr *uaddr,
L
Linus Torvalds 已提交
530 531 532 533
		       int addr_len, int flags)
{
	struct sock *sk = sock->sk;

534 535
	if (addr_len < sizeof(uaddr->sa_family))
		return -EINVAL;
L
Linus Torvalds 已提交
536 537 538
	if (uaddr->sa_family == AF_UNSPEC)
		return sk->sk_prot->disconnect(sk, flags);

E
Eric Dumazet 已提交
539
	if (!inet_sk(sk)->inet_num && inet_autobind(sk))
L
Linus Torvalds 已提交
540
		return -EAGAIN;
J
Joe Perches 已提交
541
	return sk->sk_prot->connect(sk, uaddr, addr_len);
L
Linus Torvalds 已提交
542
}
E
Eric Dumazet 已提交
543
EXPORT_SYMBOL(inet_dgram_connect);
L
Linus Torvalds 已提交
544

545
static long inet_wait_for_connect(struct sock *sk, long timeo, int writebias)
L
Linus Torvalds 已提交
546
{
547
	DEFINE_WAIT_FUNC(wait, woken_wake_function);
L
Linus Torvalds 已提交
548

549
	add_wait_queue(sk_sleep(sk), &wait);
550
	sk->sk_write_pending += writebias;
L
Linus Torvalds 已提交
551 552 553 554 555 556 557 558

	/* Basic assumption: if someone sets sk->sk_err, he _must_
	 * change state of the socket from TCP_SYN_*.
	 * Connect() does not allow to get error notifications
	 * without closing the socket.
	 */
	while ((1 << sk->sk_state) & (TCPF_SYN_SENT | TCPF_SYN_RECV)) {
		release_sock(sk);
559
		timeo = wait_woken(&wait, TASK_INTERRUPTIBLE, timeo);
L
Linus Torvalds 已提交
560 561 562 563
		lock_sock(sk);
		if (signal_pending(current) || !timeo)
			break;
	}
564
	remove_wait_queue(sk_sleep(sk), &wait);
565
	sk->sk_write_pending -= writebias;
L
Linus Torvalds 已提交
566 567 568 569 570 571 572
	return timeo;
}

/*
 *	Connect to a remote host. There is regrettably still a little
 *	TCP 'magic' in here.
 */
573
int __inet_stream_connect(struct socket *sock, struct sockaddr *uaddr,
574
			  int addr_len, int flags, int is_sendmsg)
L
Linus Torvalds 已提交
575 576 577 578 579
{
	struct sock *sk = sock->sk;
	int err;
	long timeo;

W
Wei Wang 已提交
580 581 582 583 584 585 586 587 588 589 590 591
	/*
	 * uaddr can be NULL and addr_len can be 0 if:
	 * sk is a TCP fastopen active socket and
	 * TCP_FASTOPEN_CONNECT sockopt is set and
	 * we already have a valid cookie for this socket.
	 * In this case, user can call write() after connect().
	 * write() will invoke tcp_sendmsg_fastopen() which calls
	 * __inet_stream_connect().
	 */
	if (uaddr) {
		if (addr_len < sizeof(uaddr->sa_family))
			return -EINVAL;
592

W
Wei Wang 已提交
593 594 595 596 597
		if (uaddr->sa_family == AF_UNSPEC) {
			err = sk->sk_prot->disconnect(sk, flags);
			sock->state = err ? SS_DISCONNECTING : SS_UNCONNECTED;
			goto out;
		}
L
Linus Torvalds 已提交
598 599 600 601 602 603 604 605 606 607
	}

	switch (sock->state) {
	default:
		err = -EINVAL;
		goto out;
	case SS_CONNECTED:
		err = -EISCONN;
		goto out;
	case SS_CONNECTING:
W
Wei Wang 已提交
608
		if (inet_sk(sk)->defer_connect)
609
			err = is_sendmsg ? -EINPROGRESS : -EISCONN;
W
Wei Wang 已提交
610 611
		else
			err = -EALREADY;
L
Linus Torvalds 已提交
612 613 614 615 616 617 618 619 620 621 622
		/* Fall out of switch with err, set for this state */
		break;
	case SS_UNCONNECTED:
		err = -EISCONN;
		if (sk->sk_state != TCP_CLOSE)
			goto out;

		err = sk->sk_prot->connect(sk, uaddr, addr_len);
		if (err < 0)
			goto out;

623
		sock->state = SS_CONNECTING;
L
Linus Torvalds 已提交
624

W
Wei Wang 已提交
625 626 627
		if (!err && inet_sk(sk)->defer_connect)
			goto out;

L
Linus Torvalds 已提交
628 629 630 631 632 633 634 635 636 637 638
		/* Just entered SS_CONNECTING state; the only
		 * difference is that return value in non-blocking
		 * case is EINPROGRESS, rather than EALREADY.
		 */
		err = -EINPROGRESS;
		break;
	}

	timeo = sock_sndtimeo(sk, flags & O_NONBLOCK);

	if ((1 << sk->sk_state) & (TCPF_SYN_SENT | TCPF_SYN_RECV)) {
639 640 641 642
		int writebias = (sk->sk_protocol == IPPROTO_TCP) &&
				tcp_sk(sk)->fastopen_req &&
				tcp_sk(sk)->fastopen_req->data ? 1 : 0;

L
Linus Torvalds 已提交
643
		/* Error code is set above */
644
		if (!timeo || !inet_wait_for_connect(sk, timeo, writebias))
L
Linus Torvalds 已提交
645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674
			goto out;

		err = sock_intr_errno(timeo);
		if (signal_pending(current))
			goto out;
	}

	/* Connection was closed by RST, timeout, ICMP error
	 * or another process disconnected us.
	 */
	if (sk->sk_state == TCP_CLOSE)
		goto sock_error;

	/* sk->sk_err may be not zero now, if RECVERR was ordered by user
	 * and error was received after socket entered established state.
	 * Hence, it is handled normally after connect() return successfully.
	 */

	sock->state = SS_CONNECTED;
	err = 0;
out:
	return err;

sock_error:
	err = sock_error(sk) ? : -ECONNABORTED;
	sock->state = SS_UNCONNECTED;
	if (sk->sk_prot->disconnect(sk, flags))
		sock->state = SS_DISCONNECTING;
	goto out;
}
675 676 677 678 679 680 681 682
EXPORT_SYMBOL(__inet_stream_connect);

int inet_stream_connect(struct socket *sock, struct sockaddr *uaddr,
			int addr_len, int flags)
{
	int err;

	lock_sock(sock->sk);
683
	err = __inet_stream_connect(sock, uaddr, addr_len, flags, 0);
684 685 686
	release_sock(sock->sk);
	return err;
}
E
Eric Dumazet 已提交
687
EXPORT_SYMBOL(inet_stream_connect);
L
Linus Torvalds 已提交
688 689 690 691 692

/*
 *	Accept a pending connection. The TCP layer now gives BSD semantics.
 */

693 694
int inet_accept(struct socket *sock, struct socket *newsock, int flags,
		bool kern)
L
Linus Torvalds 已提交
695 696 697
{
	struct sock *sk1 = sock->sk;
	int err = -EINVAL;
698
	struct sock *sk2 = sk1->sk_prot->accept(sk1, flags, &err, kern);
L
Linus Torvalds 已提交
699 700 701 702 703 704

	if (!sk2)
		goto do_err;

	lock_sock(sk2);

705
	sock_rps_record_flow(sk2);
706
	WARN_ON(!((1 << sk2->sk_state) &
707 708
		  (TCPF_ESTABLISHED | TCPF_SYN_RECV |
		  TCPF_CLOSE_WAIT | TCPF_CLOSE)));
L
Linus Torvalds 已提交
709 710 711 712 713 714 715 716 717

	sock_graft(sk2, newsock);

	newsock->state = SS_CONNECTED;
	err = 0;
	release_sock(sk2);
do_err:
	return err;
}
E
Eric Dumazet 已提交
718
EXPORT_SYMBOL(inet_accept);
L
Linus Torvalds 已提交
719 720 721 722 723 724 725 726 727 728


/*
 *	This does both peername and sockname.
 */
int inet_getname(struct socket *sock, struct sockaddr *uaddr,
			int *uaddr_len, int peer)
{
	struct sock *sk		= sock->sk;
	struct inet_sock *inet	= inet_sk(sk);
729
	DECLARE_SOCKADDR(struct sockaddr_in *, sin, uaddr);
L
Linus Torvalds 已提交
730 731 732

	sin->sin_family = AF_INET;
	if (peer) {
E
Eric Dumazet 已提交
733
		if (!inet->inet_dport ||
L
Linus Torvalds 已提交
734 735 736
		    (((1 << sk->sk_state) & (TCPF_CLOSE | TCPF_SYN_SENT)) &&
		     peer == 1))
			return -ENOTCONN;
E
Eric Dumazet 已提交
737 738
		sin->sin_port = inet->inet_dport;
		sin->sin_addr.s_addr = inet->inet_daddr;
L
Linus Torvalds 已提交
739
	} else {
E
Eric Dumazet 已提交
740
		__be32 addr = inet->inet_rcv_saddr;
L
Linus Torvalds 已提交
741
		if (!addr)
E
Eric Dumazet 已提交
742 743
			addr = inet->inet_saddr;
		sin->sin_port = inet->inet_sport;
L
Linus Torvalds 已提交
744 745 746 747 748 749
		sin->sin_addr.s_addr = addr;
	}
	memset(sin->sin_zero, 0, sizeof(sin->sin_zero));
	*uaddr_len = sizeof(*sin);
	return 0;
}
E
Eric Dumazet 已提交
750
EXPORT_SYMBOL(inet_getname);
L
Linus Torvalds 已提交
751

752
int inet_sendmsg(struct socket *sock, struct msghdr *msg, size_t size)
L
Linus Torvalds 已提交
753 754 755
{
	struct sock *sk = sock->sk;

756
	sock_rps_record_flow(sk);
T
Tom Herbert 已提交
757

L
Linus Torvalds 已提交
758
	/* We may need to bind the socket. */
759 760
	if (!inet_sk(sk)->inet_num && !sk->sk_prot->no_autobind &&
	    inet_autobind(sk))
L
Linus Torvalds 已提交
761 762
		return -EAGAIN;

763
	return sk->sk_prot->sendmsg(sk, msg, size);
L
Linus Torvalds 已提交
764
}
E
Eric Dumazet 已提交
765
EXPORT_SYMBOL(inet_sendmsg);
L
Linus Torvalds 已提交
766

767 768
ssize_t inet_sendpage(struct socket *sock, struct page *page, int offset,
		      size_t size, int flags)
L
Linus Torvalds 已提交
769 770 771
{
	struct sock *sk = sock->sk;

772
	sock_rps_record_flow(sk);
T
Tom Herbert 已提交
773

L
Linus Torvalds 已提交
774
	/* We may need to bind the socket. */
775 776
	if (!inet_sk(sk)->inet_num && !sk->sk_prot->no_autobind &&
	    inet_autobind(sk))
L
Linus Torvalds 已提交
777 778 779 780 781 782
		return -EAGAIN;

	if (sk->sk_prot->sendpage)
		return sk->sk_prot->sendpage(sk, page, offset, size, flags);
	return sock_no_sendpage(sock, page, offset, size, flags);
}
783
EXPORT_SYMBOL(inet_sendpage);
L
Linus Torvalds 已提交
784

785 786
int inet_recvmsg(struct socket *sock, struct msghdr *msg, size_t size,
		 int flags)
T
Tom Herbert 已提交
787 788 789 790 791
{
	struct sock *sk = sock->sk;
	int addr_len = 0;
	int err;

792
	sock_rps_record_flow(sk);
T
Tom Herbert 已提交
793

794
	err = sk->sk_prot->recvmsg(sk, msg, size, flags & MSG_DONTWAIT,
T
Tom Herbert 已提交
795 796 797 798 799 800
				   flags & ~MSG_DONTWAIT, &addr_len);
	if (err >= 0)
		msg->msg_namelen = addr_len;
	return err;
}
EXPORT_SYMBOL(inet_recvmsg);
L
Linus Torvalds 已提交
801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829

int inet_shutdown(struct socket *sock, int how)
{
	struct sock *sk = sock->sk;
	int err = 0;

	/* This should really check to make sure
	 * the socket is a TCP socket. (WHY AC...)
	 */
	how++; /* maps 0->1 has the advantage of making bit 1 rcvs and
		       1->2 bit 2 snds.
		       2->3 */
	if ((how & ~SHUTDOWN_MASK) || !how)	/* MAXINT->0 */
		return -EINVAL;

	lock_sock(sk);
	if (sock->state == SS_CONNECTING) {
		if ((1 << sk->sk_state) &
		    (TCPF_SYN_SENT | TCPF_SYN_RECV | TCPF_CLOSE))
			sock->state = SS_DISCONNECTING;
		else
			sock->state = SS_CONNECTED;
	}

	switch (sk->sk_state) {
	case TCP_CLOSE:
		err = -ENOTCONN;
		/* Hack to wake up other listeners, who can poll for
		   POLLHUP, even on eg. unconnected UDP sockets -- RR */
830
		/* fall through */
L
Linus Torvalds 已提交
831 832 833 834 835 836 837 838 839 840 841 842 843
	default:
		sk->sk_shutdown |= how;
		if (sk->sk_prot->shutdown)
			sk->sk_prot->shutdown(sk, how);
		break;

	/* Remaining two branches are temporary solution for missing
	 * close() in multithreaded environment. It is _not_ a good idea,
	 * but we have no choice until close() is repaired at VFS level.
	 */
	case TCP_LISTEN:
		if (!(how & RCV_SHUTDOWN))
			break;
844
		/* fall through */
L
Linus Torvalds 已提交
845 846 847 848 849 850 851 852 853 854 855
	case TCP_SYN_SENT:
		err = sk->sk_prot->disconnect(sk, O_NONBLOCK);
		sock->state = err ? SS_DISCONNECTING : SS_UNCONNECTED;
		break;
	}

	/* Wake up anyone sleeping in poll. */
	sk->sk_state_change(sk);
	release_sock(sk);
	return err;
}
E
Eric Dumazet 已提交
856
EXPORT_SYMBOL(inet_shutdown);
L
Linus Torvalds 已提交
857 858 859 860 861 862 863 864 865 866 867 868 869 870 871

/*
 *	ioctl() calls you can issue on an INET socket. Most of these are
 *	device configuration and stuff and very rarely used. Some ioctls
 *	pass on to the socket itself.
 *
 *	NOTE: I like the idea of a module for the config stuff. ie ifconfig
 *	loads the devconfigure module does its configuring and unloads it.
 *	There's a good 20K of config code hanging around the kernel.
 */

int inet_ioctl(struct socket *sock, unsigned int cmd, unsigned long arg)
{
	struct sock *sk = sock->sk;
	int err = 0;
872
	struct net *net = sock_net(sk);
L
Linus Torvalds 已提交
873 874

	switch (cmd) {
E
Eric Dumazet 已提交
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
	case SIOCGSTAMP:
		err = sock_get_timestamp(sk, (struct timeval __user *)arg);
		break;
	case SIOCGSTAMPNS:
		err = sock_get_timestampns(sk, (struct timespec __user *)arg);
		break;
	case SIOCADDRT:
	case SIOCDELRT:
	case SIOCRTMSG:
		err = ip_rt_ioctl(net, cmd, (void __user *)arg);
		break;
	case SIOCDARP:
	case SIOCGARP:
	case SIOCSARP:
		err = arp_ioctl(net, cmd, (void __user *)arg);
		break;
	case SIOCGIFADDR:
	case SIOCSIFADDR:
	case SIOCGIFBRDADDR:
	case SIOCSIFBRDADDR:
	case SIOCGIFNETMASK:
	case SIOCSIFNETMASK:
	case SIOCGIFDSTADDR:
	case SIOCSIFDSTADDR:
	case SIOCSIFPFLAGS:
	case SIOCGIFPFLAGS:
	case SIOCSIFFLAGS:
		err = devinet_ioctl(net, cmd, (void __user *)arg);
		break;
	default:
		if (sk->sk_prot->ioctl)
			err = sk->sk_prot->ioctl(sk, cmd, arg);
		else
			err = -ENOIOCTLCMD;
		break;
L
Linus Torvalds 已提交
910 911 912
	}
	return err;
}
E
Eric Dumazet 已提交
913
EXPORT_SYMBOL(inet_ioctl);
L
Linus Torvalds 已提交
914

915
#ifdef CONFIG_COMPAT
916
static int inet_compat_ioctl(struct socket *sock, unsigned int cmd, unsigned long arg)
917 918 919 920 921 922 923 924 925 926 927
{
	struct sock *sk = sock->sk;
	int err = -ENOIOCTLCMD;

	if (sk->sk_prot->compat_ioctl)
		err = sk->sk_prot->compat_ioctl(sk, cmd, arg);

	return err;
}
#endif

928
const struct proto_ops inet_stream_ops = {
929 930 931 932 933 934 935 936 937 938 939 940 941 942
	.family		   = PF_INET,
	.owner		   = THIS_MODULE,
	.release	   = inet_release,
	.bind		   = inet_bind,
	.connect	   = inet_stream_connect,
	.socketpair	   = sock_no_socketpair,
	.accept		   = inet_accept,
	.getname	   = inet_getname,
	.poll		   = tcp_poll,
	.ioctl		   = inet_ioctl,
	.listen		   = inet_listen,
	.shutdown	   = inet_shutdown,
	.setsockopt	   = sock_common_setsockopt,
	.getsockopt	   = sock_common_getsockopt,
943
	.sendmsg	   = inet_sendmsg,
T
Tom Herbert 已提交
944
	.recvmsg	   = inet_recvmsg,
945
	.mmap		   = sock_no_mmap,
946
	.sendpage	   = inet_sendpage,
J
Jens Axboe 已提交
947
	.splice_read	   = tcp_splice_read,
948
	.read_sock	   = tcp_read_sock,
949 950
	.sendmsg_locked    = tcp_sendmsg_locked,
	.sendpage_locked   = tcp_sendpage_locked,
951
	.peek_len	   = tcp_peek_len,
952
#ifdef CONFIG_COMPAT
953 954
	.compat_setsockopt = compat_sock_common_setsockopt,
	.compat_getsockopt = compat_sock_common_getsockopt,
955
	.compat_ioctl	   = inet_compat_ioctl,
956
#endif
L
Linus Torvalds 已提交
957
};
E
Eric Dumazet 已提交
958
EXPORT_SYMBOL(inet_stream_ops);
L
Linus Torvalds 已提交
959

960
const struct proto_ops inet_dgram_ops = {
961 962 963 964 965 966 967 968 969 970 971 972 973 974 975
	.family		   = PF_INET,
	.owner		   = THIS_MODULE,
	.release	   = inet_release,
	.bind		   = inet_bind,
	.connect	   = inet_dgram_connect,
	.socketpair	   = sock_no_socketpair,
	.accept		   = sock_no_accept,
	.getname	   = inet_getname,
	.poll		   = udp_poll,
	.ioctl		   = inet_ioctl,
	.listen		   = sock_no_listen,
	.shutdown	   = inet_shutdown,
	.setsockopt	   = sock_common_setsockopt,
	.getsockopt	   = sock_common_getsockopt,
	.sendmsg	   = inet_sendmsg,
T
Tom Herbert 已提交
976
	.recvmsg	   = inet_recvmsg,
977 978
	.mmap		   = sock_no_mmap,
	.sendpage	   = inet_sendpage,
979
	.set_peek_off	   = sk_set_peek_off,
980
#ifdef CONFIG_COMPAT
981 982
	.compat_setsockopt = compat_sock_common_setsockopt,
	.compat_getsockopt = compat_sock_common_getsockopt,
983
	.compat_ioctl	   = inet_compat_ioctl,
984
#endif
L
Linus Torvalds 已提交
985
};
E
Eric Dumazet 已提交
986
EXPORT_SYMBOL(inet_dgram_ops);
L
Linus Torvalds 已提交
987 988 989 990 991

/*
 * For SOCK_RAW sockets; should be the same as inet_dgram_ops but without
 * udp_poll
 */
992
static const struct proto_ops inet_sockraw_ops = {
993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007
	.family		   = PF_INET,
	.owner		   = THIS_MODULE,
	.release	   = inet_release,
	.bind		   = inet_bind,
	.connect	   = inet_dgram_connect,
	.socketpair	   = sock_no_socketpair,
	.accept		   = sock_no_accept,
	.getname	   = inet_getname,
	.poll		   = datagram_poll,
	.ioctl		   = inet_ioctl,
	.listen		   = sock_no_listen,
	.shutdown	   = inet_shutdown,
	.setsockopt	   = sock_common_setsockopt,
	.getsockopt	   = sock_common_getsockopt,
	.sendmsg	   = inet_sendmsg,
T
Tom Herbert 已提交
1008
	.recvmsg	   = inet_recvmsg,
1009 1010
	.mmap		   = sock_no_mmap,
	.sendpage	   = inet_sendpage,
1011
#ifdef CONFIG_COMPAT
1012 1013
	.compat_setsockopt = compat_sock_common_setsockopt,
	.compat_getsockopt = compat_sock_common_getsockopt,
1014
	.compat_ioctl	   = inet_compat_ioctl,
1015
#endif
L
Linus Torvalds 已提交
1016 1017
};

1018
static const struct net_proto_family inet_family_ops = {
L
Linus Torvalds 已提交
1019 1020 1021 1022 1023 1024 1025 1026 1027 1028
	.family = PF_INET,
	.create = inet_create,
	.owner	= THIS_MODULE,
};

/* Upon startup we insert all the elements in inetsw_array[] into
 * the linked list inetsw.
 */
static struct inet_protosw inetsw_array[] =
{
1029 1030 1031 1032 1033 1034
	{
		.type =       SOCK_STREAM,
		.protocol =   IPPROTO_TCP,
		.prot =       &tcp_prot,
		.ops =        &inet_stream_ops,
		.flags =      INET_PROTOSW_PERMANENT |
1035
			      INET_PROTOSW_ICSK,
1036 1037 1038 1039 1040 1041 1042 1043
	},

	{
		.type =       SOCK_DGRAM,
		.protocol =   IPPROTO_UDP,
		.prot =       &udp_prot,
		.ops =        &inet_dgram_ops,
		.flags =      INET_PROTOSW_PERMANENT,
L
Linus Torvalds 已提交
1044
       },
1045

1046 1047 1048 1049
       {
		.type =       SOCK_DGRAM,
		.protocol =   IPPROTO_ICMP,
		.prot =       &ping_prot,
1050
		.ops =        &inet_sockraw_ops,
1051 1052
		.flags =      INET_PROTOSW_REUSE,
       },
L
Linus Torvalds 已提交
1053 1054

       {
1055 1056 1057 1058 1059
	       .type =       SOCK_RAW,
	       .protocol =   IPPROTO_IP,	/* wild card */
	       .prot =       &raw_prot,
	       .ops =        &inet_sockraw_ops,
	       .flags =      INET_PROTOSW_REUSE,
L
Linus Torvalds 已提交
1060 1061 1062
       }
};

1063
#define INETSW_ARRAY_LEN ARRAY_SIZE(inetsw_array)
L
Linus Torvalds 已提交
1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081

void inet_register_protosw(struct inet_protosw *p)
{
	struct list_head *lh;
	struct inet_protosw *answer;
	int protocol = p->protocol;
	struct list_head *last_perm;

	spin_lock_bh(&inetsw_lock);

	if (p->type >= SOCK_MAX)
		goto out_illegal;

	/* If we are trying to override a permanent protocol, bail. */
	last_perm = &inetsw[p->type];
	list_for_each(lh, &inetsw[p->type]) {
		answer = list_entry(lh, struct inet_protosw, list);
		/* Check only the non-wild match. */
1082 1083 1084 1085 1086
		if ((INET_PROTOSW_PERMANENT & answer->flags) == 0)
			break;
		if (protocol == answer->protocol)
			goto out_permanent;
		last_perm = lh;
L
Linus Torvalds 已提交
1087 1088 1089 1090 1091
	}

	/* Add the new entry after the last permanent entry if any, so that
	 * the new entry does not override a permanent entry when matched with
	 * a wild-card protocol. But it is allowed to override any existing
1092
	 * non-permanent entry.  This means that when we remove this entry, the
L
Linus Torvalds 已提交
1093 1094 1095 1096 1097 1098 1099 1100 1101
	 * system automatically returns to the old behavior.
	 */
	list_add_rcu(&p->list, last_perm);
out:
	spin_unlock_bh(&inetsw_lock);

	return;

out_permanent:
J
Joe Perches 已提交
1102
	pr_err("Attempt to override permanent protocol %d\n", protocol);
L
Linus Torvalds 已提交
1103 1104 1105
	goto out;

out_illegal:
J
Joe Perches 已提交
1106
	pr_err("Ignoring attempt to register invalid socket type %d\n",
L
Linus Torvalds 已提交
1107 1108 1109
	       p->type);
	goto out;
}
E
Eric Dumazet 已提交
1110
EXPORT_SYMBOL(inet_register_protosw);
L
Linus Torvalds 已提交
1111 1112 1113 1114

void inet_unregister_protosw(struct inet_protosw *p)
{
	if (INET_PROTOSW_PERMANENT & p->flags) {
J
Joe Perches 已提交
1115
		pr_err("Attempt to unregister permanent protocol %d\n",
L
Linus Torvalds 已提交
1116 1117 1118 1119 1120 1121 1122 1123 1124
		       p->protocol);
	} else {
		spin_lock_bh(&inetsw_lock);
		list_del_rcu(&p->list);
		spin_unlock_bh(&inetsw_lock);

		synchronize_net();
	}
}
E
Eric Dumazet 已提交
1125
EXPORT_SYMBOL(inet_unregister_protosw);
L
Linus Torvalds 已提交
1126

1127 1128 1129
static int inet_sk_reselect_saddr(struct sock *sk)
{
	struct inet_sock *inet = inet_sk(sk);
E
Eric Dumazet 已提交
1130 1131
	__be32 old_saddr = inet->inet_saddr;
	__be32 daddr = inet->inet_daddr;
1132
	struct flowi4 *fl4;
1133 1134
	struct rtable *rt;
	__be32 new_saddr;
1135
	struct ip_options_rcu *inet_opt;
1136

1137
	inet_opt = rcu_dereference_protected(inet->inet_opt,
1138
					     lockdep_sock_is_held(sk));
1139 1140
	if (inet_opt && inet_opt->opt.srr)
		daddr = inet_opt->opt.faddr;
1141 1142

	/* Query new route. */
1143 1144
	fl4 = &inet->cork.fl.u.ip4;
	rt = ip_route_connect(fl4, daddr, 0, RT_CONN_FLAGS(sk),
1145
			      sk->sk_bound_dev_if, sk->sk_protocol,
1146
			      inet->inet_sport, inet->inet_dport, sk);
1147 1148
	if (IS_ERR(rt))
		return PTR_ERR(rt);
1149

1150
	sk_setup_caps(sk, &rt->dst);
1151

1152
	new_saddr = fl4->saddr;
1153 1154 1155 1156

	if (new_saddr == old_saddr)
		return 0;

1157
	if (sock_net(sk)->ipv4.sysctl_ip_dynaddr > 1) {
J
Joe Perches 已提交
1158 1159
		pr_info("%s(): shifting inet->saddr from %pI4 to %pI4\n",
			__func__, &old_saddr, &new_saddr);
1160 1161
	}

E
Eric Dumazet 已提交
1162
	inet->inet_saddr = inet->inet_rcv_saddr = new_saddr;
1163 1164 1165 1166 1167 1168 1169 1170 1171

	/*
	 * XXX The only one ugly spot where we need to
	 * XXX really change the sockets identity after
	 * XXX it has entered the hashes. -DaveM
	 *
	 * Besides that, it does not check for connection
	 * uniqueness. Wait for troubles.
	 */
1172
	return __sk_prot_rehash(sk);
1173 1174 1175 1176 1177 1178
}

int inet_sk_rebuild_header(struct sock *sk)
{
	struct inet_sock *inet = inet_sk(sk);
	struct rtable *rt = (struct rtable *)__sk_dst_check(sk, 0);
A
Al Viro 已提交
1179
	__be32 daddr;
1180
	struct ip_options_rcu *inet_opt;
1181
	struct flowi4 *fl4;
1182 1183 1184 1185 1186 1187 1188
	int err;

	/* Route is OK, nothing to do. */
	if (rt)
		return 0;

	/* Reroute. */
1189 1190
	rcu_read_lock();
	inet_opt = rcu_dereference(inet->inet_opt);
E
Eric Dumazet 已提交
1191
	daddr = inet->inet_daddr;
1192 1193 1194
	if (inet_opt && inet_opt->opt.srr)
		daddr = inet_opt->opt.faddr;
	rcu_read_unlock();
1195 1196
	fl4 = &inet->cork.fl.u.ip4;
	rt = ip_route_output_ports(sock_net(sk), fl4, sk, daddr, inet->inet_saddr,
1197 1198 1199
				   inet->inet_dport, inet->inet_sport,
				   sk->sk_protocol, RT_CONN_FLAGS(sk),
				   sk->sk_bound_dev_if);
1200 1201
	if (!IS_ERR(rt)) {
		err = 0;
1202
		sk_setup_caps(sk, &rt->dst);
1203 1204 1205
	} else {
		err = PTR_ERR(rt);

1206 1207 1208 1209 1210 1211
		/* Routing failed... */
		sk->sk_route_caps = 0;
		/*
		 * Other protocols have to map its equivalent state to TCP_SYN_SENT.
		 * DCCP maps its DCCP_REQUESTING state to TCP_SYN_SENT. -acme
		 */
1212
		if (!sock_net(sk)->ipv4.sysctl_ip_dynaddr ||
1213 1214 1215 1216 1217 1218 1219 1220 1221 1222
		    sk->sk_state != TCP_SYN_SENT ||
		    (sk->sk_userlocks & SOCK_BINDADDR_LOCK) ||
		    (err = inet_sk_reselect_saddr(sk)) != 0)
			sk->sk_err_soft = -err;
	}

	return err;
}
EXPORT_SYMBOL(inet_sk_rebuild_header);

T
Tom Herbert 已提交
1223 1224
struct sk_buff *inet_gso_segment(struct sk_buff *skb,
				 netdev_features_t features)
H
Herbert Xu 已提交
1225
{
1226
	bool fixedid = false, gso_partial, encap;
H
Herbert Xu 已提交
1227
	struct sk_buff *segs = ERR_PTR(-EINVAL);
1228
	const struct net_offload *ops;
1229
	struct iphdr *iph;
1230
	int proto, tot_len;
1231
	int nhoff;
H
Herbert Xu 已提交
1232 1233 1234
	int ihl;
	int id;

1235 1236
	skb_reset_network_header(skb);
	nhoff = skb_network_header(skb) - skb_mac_header(skb);
1237
	if (unlikely(!pskb_may_pull(skb, sizeof(*iph))))
H
Herbert Xu 已提交
1238 1239
		goto out;

1240
	iph = ip_hdr(skb);
H
Herbert Xu 已提交
1241 1242 1243 1244
	ihl = iph->ihl * 4;
	if (ihl < sizeof(*iph))
		goto out;

1245 1246 1247 1248
	id = ntohs(iph->id);
	proto = iph->protocol;

	/* Warning: after this point, iph might be no longer valid */
1249
	if (unlikely(!pskb_may_pull(skb, ihl)))
H
Herbert Xu 已提交
1250
		goto out;
1251
	__skb_pull(skb, ihl);
H
Herbert Xu 已提交
1252

E
Eric Dumazet 已提交
1253 1254
	encap = SKB_GSO_CB(skb)->encap_level > 0;
	if (encap)
1255
		features &= skb->dev->hw_enc_features;
1256
	SKB_GSO_CB(skb)->encap_level += ihl;
1257

1258
	skb_reset_transport_header(skb);
1259

H
Herbert Xu 已提交
1260 1261
	segs = ERR_PTR(-EPROTONOSUPPORT);

1262 1263 1264 1265
	if (!skb->encapsulation || encap) {
		fixedid = !!(skb_shinfo(skb)->gso_type & SKB_GSO_TCP_FIXEDID);

		/* fixed ID is invalid if DF bit is not set */
1266
		if (fixedid && !(ip_hdr(skb)->frag_off & htons(IP_DF)))
1267 1268
			goto out;
	}
E
Eric Dumazet 已提交
1269

1270
	ops = rcu_dereference(inet_offloads[proto]);
1271 1272
	if (likely(ops && ops->callbacks.gso_segment))
		segs = ops->callbacks.gso_segment(skb, features);
H
Herbert Xu 已提交
1273

1274
	if (IS_ERR_OR_NULL(segs))
H
Herbert Xu 已提交
1275 1276
		goto out;

1277 1278
	gso_partial = !!(skb_shinfo(segs)->gso_type & SKB_GSO_PARTIAL);

H
Herbert Xu 已提交
1279 1280
	skb = segs;
	do {
1281
		iph = (struct iphdr *)(skb_mac_header(skb) + nhoff);
1282
		if (skb_is_gso(skb)) {
1283 1284 1285 1286
			if (!fixedid) {
				iph->id = htons(id);
				id += skb_shinfo(skb)->gso_segs;
			}
1287 1288 1289 1290 1291 1292 1293

			if (gso_partial)
				tot_len = skb_shinfo(skb)->gso_size +
					  SKB_GSO_CB(skb)->data_offset +
					  skb->head - (unsigned char *)iph;
			else
				tot_len = skb->len - nhoff;
1294 1295 1296 1297
		} else {
			if (!fixedid)
				iph->id = htons(id++);
			tot_len = skb->len - nhoff;
P
Pravin B Shelar 已提交
1298
		}
1299
		iph->tot_len = htons(tot_len);
1300
		ip_send_check(iph);
E
Eric Dumazet 已提交
1301
		if (encap)
1302 1303
			skb_reset_inner_headers(skb);
		skb->network_header = (u8 *)iph - skb->head;
H
Herbert Xu 已提交
1304 1305 1306 1307 1308
	} while ((skb = skb->next));

out:
	return segs;
}
T
Tom Herbert 已提交
1309
EXPORT_SYMBOL(inet_gso_segment);
H
Herbert Xu 已提交
1310

T
Tom Herbert 已提交
1311
struct sk_buff **inet_gro_receive(struct sk_buff **head, struct sk_buff *skb)
H
Herbert Xu 已提交
1312
{
1313
	const struct net_offload *ops;
H
Herbert Xu 已提交
1314 1315
	struct sk_buff **pp = NULL;
	struct sk_buff *p;
1316
	const struct iphdr *iph;
1317 1318
	unsigned int hlen;
	unsigned int off;
1319
	unsigned int id;
H
Herbert Xu 已提交
1320 1321 1322
	int flush = 1;
	int proto;

1323 1324 1325 1326 1327 1328 1329 1330
	off = skb_gro_offset(skb);
	hlen = off + sizeof(*iph);
	iph = skb_gro_header_fast(skb, off);
	if (skb_gro_header_hard(skb, hlen)) {
		iph = skb_gro_header_slow(skb, hlen, off);
		if (unlikely(!iph))
			goto out;
	}
H
Herbert Xu 已提交
1331

1332
	proto = iph->protocol;
H
Herbert Xu 已提交
1333 1334

	rcu_read_lock();
1335
	ops = rcu_dereference(inet_offloads[proto]);
1336
	if (!ops || !ops->callbacks.gro_receive)
H
Herbert Xu 已提交
1337 1338
		goto out_unlock;

H
Herbert Xu 已提交
1339
	if (*(u8 *)iph != 0x45)
H
Herbert Xu 已提交
1340 1341
		goto out_unlock;

1342 1343 1344
	if (ip_is_fragment(iph))
		goto out_unlock;

1345
	if (unlikely(ip_fast_csum((u8 *)iph, 5)))
H
Herbert Xu 已提交
1346 1347
		goto out_unlock;

1348
	id = ntohl(*(__be32 *)&iph->id);
1349
	flush = (u16)((ntohl(*(__be32 *)iph) ^ skb_gro_len(skb)) | (id & ~IP_DF));
1350
	id >>= 16;
H
Herbert Xu 已提交
1351 1352 1353

	for (p = *head; p; p = p->next) {
		struct iphdr *iph2;
1354
		u16 flush_id;
H
Herbert Xu 已提交
1355 1356 1357 1358

		if (!NAPI_GRO_CB(p)->same_flow)
			continue;

1359 1360 1361 1362 1363 1364
		iph2 = (struct iphdr *)(p->data + off);
		/* The above works because, with the exception of the top
		 * (inner most) layer, we only aggregate pkts with the same
		 * hdr length so all the hdrs we'll need to verify will start
		 * at the same offset.
		 */
H
Herbert Xu 已提交
1365
		if ((iph->protocol ^ iph2->protocol) |
1366 1367
		    ((__force u32)iph->saddr ^ (__force u32)iph2->saddr) |
		    ((__force u32)iph->daddr ^ (__force u32)iph2->daddr)) {
H
Herbert Xu 已提交
1368 1369 1370 1371 1372 1373
			NAPI_GRO_CB(p)->same_flow = 0;
			continue;
		}

		/* All fields must match except length and checksum. */
		NAPI_GRO_CB(p)->flush |=
H
Herbert Xu 已提交
1374
			(iph->ttl ^ iph2->ttl) |
1375
			(iph->tos ^ iph2->tos) |
1376
			((iph->frag_off ^ iph2->frag_off) & htons(IP_DF));
H
Herbert Xu 已提交
1377 1378

		NAPI_GRO_CB(p)->flush |= flush;
1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404

		/* We need to store of the IP ID check to be included later
		 * when we can verify that this packet does in fact belong
		 * to a given flow.
		 */
		flush_id = (u16)(id - ntohs(iph2->id));

		/* This bit of code makes it much easier for us to identify
		 * the cases where we are doing atomic vs non-atomic IP ID
		 * checks.  Specifically an atomic check can return IP ID
		 * values 0 - 0xFFFF, while a non-atomic check can only
		 * return 0 or 0xFFFF.
		 */
		if (!NAPI_GRO_CB(p)->is_atomic ||
		    !(iph->frag_off & htons(IP_DF))) {
			flush_id ^= NAPI_GRO_CB(p)->count;
			flush_id = flush_id ? 0xFFFF : 0;
		}

		/* If the previous IP ID value was based on an atomic
		 * datagram we can overwrite the value and ignore it.
		 */
		if (NAPI_GRO_CB(skb)->is_atomic)
			NAPI_GRO_CB(p)->flush_id = flush_id;
		else
			NAPI_GRO_CB(p)->flush_id |= flush_id;
H
Herbert Xu 已提交
1405 1406
	}

1407
	NAPI_GRO_CB(skb)->is_atomic = !!(iph->frag_off & htons(IP_DF));
H
Herbert Xu 已提交
1408
	NAPI_GRO_CB(skb)->flush |= flush;
1409 1410 1411 1412 1413
	skb_set_network_header(skb, off);
	/* The above will be needed by the transport layer if there is one
	 * immediately following this IP hdr.
	 */

1414 1415 1416
	/* Note : No need to call skb_gro_postpull_rcsum() here,
	 * as we already checked checksum over ipv4 header was 0
	 */
1417 1418
	skb_gro_pull(skb, sizeof(*iph));
	skb_set_transport_header(skb, skb_gro_offset(skb));
H
Herbert Xu 已提交
1419

S
Sabrina Dubroca 已提交
1420
	pp = call_gro_receive(ops->callbacks.gro_receive, head, skb);
H
Herbert Xu 已提交
1421 1422 1423 1424 1425

out_unlock:
	rcu_read_unlock();

out:
1426
	skb_gro_flush_final(skb, pp, flush);
H
Herbert Xu 已提交
1427 1428 1429

	return pp;
}
T
Tom Herbert 已提交
1430
EXPORT_SYMBOL(inet_gro_receive);
H
Herbert Xu 已提交
1431

1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444
static struct sk_buff **ipip_gro_receive(struct sk_buff **head,
					 struct sk_buff *skb)
{
	if (NAPI_GRO_CB(skb)->encap_mark) {
		NAPI_GRO_CB(skb)->flush = 1;
		return NULL;
	}

	NAPI_GRO_CB(skb)->encap_mark = 1;

	return inet_gro_receive(head, skb);
}

1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466
#define SECONDS_PER_DAY	86400

/* inet_current_timestamp - Return IP network timestamp
 *
 * Return milliseconds since midnight in network byte order.
 */
__be32 inet_current_timestamp(void)
{
	u32 secs;
	u32 msecs;
	struct timespec64 ts;

	ktime_get_real_ts64(&ts);

	/* Get secs since midnight. */
	(void)div_u64_rem(ts.tv_sec, SECONDS_PER_DAY, &secs);
	/* Convert to msecs. */
	msecs = secs * MSEC_PER_SEC;
	/* Convert nsec to msec. */
	msecs += (u32)ts.tv_nsec / NSEC_PER_MSEC;

	/* Convert to network byte order. */
1467
	return htonl(msecs);
1468 1469 1470
}
EXPORT_SYMBOL(inet_current_timestamp);

1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481
int inet_recv_error(struct sock *sk, struct msghdr *msg, int len, int *addr_len)
{
	if (sk->sk_family == AF_INET)
		return ip_recv_error(sk, msg, len, addr_len);
#if IS_ENABLED(CONFIG_IPV6)
	if (sk->sk_family == AF_INET6)
		return pingv6_ops.ipv6_recv_error(sk, msg, len, addr_len);
#endif
	return -EINVAL;
}

T
Tom Herbert 已提交
1482
int inet_gro_complete(struct sk_buff *skb, int nhoff)
H
Herbert Xu 已提交
1483
{
1484 1485
	__be16 newlen = htons(skb->len - nhoff);
	struct iphdr *iph = (struct iphdr *)(skb->data + nhoff);
1486
	const struct net_offload *ops;
1487
	int proto = iph->protocol;
H
Herbert Xu 已提交
1488 1489
	int err = -ENOSYS;

1490 1491
	if (skb->encapsulation) {
		skb_set_inner_protocol(skb, cpu_to_be16(ETH_P_IP));
1492
		skb_set_inner_network_header(skb, nhoff);
1493
	}
1494

H
Herbert Xu 已提交
1495 1496 1497 1498
	csum_replace2(&iph->check, iph->tot_len, newlen);
	iph->tot_len = newlen;

	rcu_read_lock();
1499
	ops = rcu_dereference(inet_offloads[proto]);
1500
	if (WARN_ON(!ops || !ops->callbacks.gro_complete))
H
Herbert Xu 已提交
1501 1502
		goto out_unlock;

1503 1504 1505 1506 1507
	/* Only need to add sizeof(*iph) to get to the next hdr below
	 * because any hdr with option will have been flushed in
	 * inet_gro_receive().
	 */
	err = ops->callbacks.gro_complete(skb, nhoff + sizeof(*iph));
H
Herbert Xu 已提交
1508 1509 1510 1511 1512 1513

out_unlock:
	rcu_read_unlock();

	return err;
}
T
Tom Herbert 已提交
1514
EXPORT_SYMBOL(inet_gro_complete);
H
Herbert Xu 已提交
1515

1516 1517 1518
static int ipip_gro_complete(struct sk_buff *skb, int nhoff)
{
	skb->encapsulation = 1;
1519
	skb_shinfo(skb)->gso_type |= SKB_GSO_IPXIP4;
1520 1521 1522
	return inet_gro_complete(skb, nhoff);
}

1523
int inet_ctl_sock_create(struct sock **sk, unsigned short family,
1524 1525
			 unsigned short type, unsigned char protocol,
			 struct net *net)
1526
{
1527
	struct socket *sock;
1528
	int rc = sock_create_kern(net, family, type, protocol, &sock);
1529 1530

	if (rc == 0) {
1531 1532
		*sk = sock->sk;
		(*sk)->sk_allocation = GFP_ATOMIC;
1533 1534 1535 1536
		/*
		 * Unhash it so that IP input processing does not even see it,
		 * we do not wish this socket to see incoming packets.
		 */
1537
		(*sk)->sk_prot->unhash(*sk);
1538 1539 1540 1541 1542
	}
	return rc;
}
EXPORT_SYMBOL_GPL(inet_ctl_sock_create);

1543 1544 1545 1546 1547 1548
u64 snmp_get_cpu_field(void __percpu *mib, int cpu, int offt)
{
	return  *(((unsigned long *)per_cpu_ptr(mib, cpu)) + offt);
}
EXPORT_SYMBOL_GPL(snmp_get_cpu_field);

W
WANG Cong 已提交
1549
unsigned long snmp_fold_field(void __percpu *mib, int offt)
1550 1551
{
	unsigned long res = 0;
W
WANG Cong 已提交
1552
	int i;
1553

W
WANG Cong 已提交
1554
	for_each_possible_cpu(i)
1555
		res += snmp_get_cpu_field(mib, i, offt);
1556 1557 1558 1559
	return res;
}
EXPORT_SYMBOL_GPL(snmp_fold_field);

1560 1561
#if BITS_PER_LONG==32

D
David S. Miller 已提交
1562
u64 snmp_get_cpu_field64(void __percpu *mib, int cpu, int offt,
1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580
			 size_t syncp_offset)
{
	void *bhptr;
	struct u64_stats_sync *syncp;
	u64 v;
	unsigned int start;

	bhptr = per_cpu_ptr(mib, cpu);
	syncp = (struct u64_stats_sync *)(bhptr + syncp_offset);
	do {
		start = u64_stats_fetch_begin_irq(syncp);
		v = *(((u64 *)bhptr) + offt);
	} while (u64_stats_fetch_retry_irq(syncp, start));

	return v;
}
EXPORT_SYMBOL_GPL(snmp_get_cpu_field64);

W
WANG Cong 已提交
1581
u64 snmp_fold_field64(void __percpu *mib, int offt, size_t syncp_offset)
1582 1583 1584 1585 1586
{
	u64 res = 0;
	int cpu;

	for_each_possible_cpu(cpu) {
M
Madalin Bucur 已提交
1587
		res += snmp_get_cpu_field64(mib, cpu, offt, syncp_offset);
1588 1589 1590 1591 1592 1593
	}
	return res;
}
EXPORT_SYMBOL_GPL(snmp_fold_field64);
#endif

L
Linus Torvalds 已提交
1594
#ifdef CONFIG_IP_MULTICAST
1595
static const struct net_protocol igmp_protocol = {
L
Linus Torvalds 已提交
1596
	.handler =	igmp_rcv,
1597
	.netns_ok =	1,
L
Linus Torvalds 已提交
1598 1599 1600
};
#endif

1601 1602 1603
/* thinking of making this const? Don't.
 * early_demux can change based on sysctl.
 */
1604
static struct net_protocol tcp_protocol = {
D
David S. Miller 已提交
1605
	.early_demux	=	tcp_v4_early_demux,
1606
	.early_demux_handler =  tcp_v4_early_demux,
D
David S. Miller 已提交
1607 1608 1609 1610
	.handler	=	tcp_v4_rcv,
	.err_handler	=	tcp_v4_err,
	.no_policy	=	1,
	.netns_ok	=	1,
1611
	.icmp_strict_tag_validation = 1,
L
Linus Torvalds 已提交
1612 1613
};

1614 1615 1616
/* thinking of making this const? Don't.
 * early_demux can change based on sysctl.
 */
1617
static struct net_protocol udp_protocol = {
S
Shawn Bohrer 已提交
1618
	.early_demux =	udp_v4_early_demux,
1619
	.early_demux_handler =	udp_v4_early_demux,
L
Linus Torvalds 已提交
1620 1621 1622
	.handler =	udp_rcv,
	.err_handler =	udp_err,
	.no_policy =	1,
1623
	.netns_ok =	1,
L
Linus Torvalds 已提交
1624 1625
};

1626
static const struct net_protocol icmp_protocol = {
L
Linus Torvalds 已提交
1627
	.handler =	icmp_rcv,
1628
	.err_handler =	icmp_err,
1629
	.no_policy =	1,
1630
	.netns_ok =	1,
L
Linus Torvalds 已提交
1631 1632
};

1633 1634
static __net_init int ipv4_mib_init_net(struct net *net)
{
1635 1636
	int i;

W
WANG Cong 已提交
1637 1638
	net->mib.tcp_statistics = alloc_percpu(struct tcp_mib);
	if (!net->mib.tcp_statistics)
1639
		goto err_tcp_mib;
W
WANG Cong 已提交
1640 1641
	net->mib.ip_statistics = alloc_percpu(struct ipstats_mib);
	if (!net->mib.ip_statistics)
1642
		goto err_ip_mib;
1643 1644 1645

	for_each_possible_cpu(i) {
		struct ipstats_mib *af_inet_stats;
W
WANG Cong 已提交
1646
		af_inet_stats = per_cpu_ptr(net->mib.ip_statistics, i);
1647 1648 1649
		u64_stats_init(&af_inet_stats->syncp);
	}

W
WANG Cong 已提交
1650 1651
	net->mib.net_statistics = alloc_percpu(struct linux_mib);
	if (!net->mib.net_statistics)
1652
		goto err_net_mib;
W
WANG Cong 已提交
1653 1654
	net->mib.udp_statistics = alloc_percpu(struct udp_mib);
	if (!net->mib.udp_statistics)
1655
		goto err_udp_mib;
W
WANG Cong 已提交
1656 1657
	net->mib.udplite_statistics = alloc_percpu(struct udp_mib);
	if (!net->mib.udplite_statistics)
1658
		goto err_udplite_mib;
W
WANG Cong 已提交
1659 1660
	net->mib.icmp_statistics = alloc_percpu(struct icmp_mib);
	if (!net->mib.icmp_statistics)
1661
		goto err_icmp_mib;
1662 1663 1664
	net->mib.icmpmsg_statistics = kzalloc(sizeof(struct icmpmsg_mib),
					      GFP_KERNEL);
	if (!net->mib.icmpmsg_statistics)
1665
		goto err_icmpmsg_mib;
1666 1667

	tcp_mib_init(net);
1668
	return 0;
1669

1670
err_icmpmsg_mib:
W
WANG Cong 已提交
1671
	free_percpu(net->mib.icmp_statistics);
1672
err_icmp_mib:
W
WANG Cong 已提交
1673
	free_percpu(net->mib.udplite_statistics);
1674
err_udplite_mib:
W
WANG Cong 已提交
1675
	free_percpu(net->mib.udp_statistics);
1676
err_udp_mib:
W
WANG Cong 已提交
1677
	free_percpu(net->mib.net_statistics);
1678
err_net_mib:
W
WANG Cong 已提交
1679
	free_percpu(net->mib.ip_statistics);
1680
err_ip_mib:
W
WANG Cong 已提交
1681
	free_percpu(net->mib.tcp_statistics);
1682 1683
err_tcp_mib:
	return -ENOMEM;
1684 1685 1686 1687
}

static __net_exit void ipv4_mib_exit_net(struct net *net)
{
1688
	kfree(net->mib.icmpmsg_statistics);
W
WANG Cong 已提交
1689 1690 1691 1692 1693 1694
	free_percpu(net->mib.icmp_statistics);
	free_percpu(net->mib.udplite_statistics);
	free_percpu(net->mib.udp_statistics);
	free_percpu(net->mib.net_statistics);
	free_percpu(net->mib.ip_statistics);
	free_percpu(net->mib.tcp_statistics);
1695 1696 1697 1698 1699 1700 1701
}

static __net_initdata struct pernet_operations ipv4_mib_ops = {
	.init = ipv4_mib_init_net,
	.exit = ipv4_mib_exit_net,
};

L
Linus Torvalds 已提交
1702 1703
static int __init init_ipv4_mibs(void)
{
1704
	return register_pernet_subsys(&ipv4_mib_ops);
L
Linus Torvalds 已提交
1705 1706
}

1707 1708 1709 1710 1711 1712 1713
static __net_init int inet_init_net(struct net *net)
{
	/*
	 * Set defaults for local port range
	 */
	seqlock_init(&net->ipv4.ip_local_ports.lock);
	net->ipv4.ip_local_ports.range[0] =  32768;
1714
	net->ipv4.ip_local_ports.range[1] =  60999;
1715 1716 1717 1718 1719 1720 1721 1722

	seqlock_init(&net->ipv4.ping_group_range.lock);
	/*
	 * Sane defaults - nobody may create ping sockets.
	 * Boot scripts should set this to distro-specific group.
	 */
	net->ipv4.ping_group_range.range[0] = make_kgid(&init_user_ns, 1);
	net->ipv4.ping_group_range.range[1] = make_kgid(&init_user_ns, 0);
1723 1724 1725 1726 1727 1728 1729

	/* Default values for sysctl-controlled parameters.
	 * We set them here, in case sysctl is not compiled.
	 */
	net->ipv4.sysctl_ip_default_ttl = IPDEFTTL;
	net->ipv4.sysctl_ip_dynaddr = 0;
	net->ipv4.sysctl_ip_early_demux = 1;
1730 1731
	net->ipv4.sysctl_udp_early_demux = 1;
	net->ipv4.sysctl_tcp_early_demux = 1;
1732 1733 1734
#ifdef CONFIG_SYSCTL
	net->ipv4.sysctl_ip_prot_sock = PROT_SOCK;
#endif
1735

1736 1737 1738 1739 1740 1741 1742
	/* Some igmp sysctl, whose values are always used */
	net->ipv4.sysctl_igmp_max_memberships = 20;
	net->ipv4.sysctl_igmp_max_msf = 10;
	/* IGMP reports for link-local multicast groups are enabled by default */
	net->ipv4.sysctl_igmp_llm_reports = 1;
	net->ipv4.sysctl_igmp_qrv = 2;

1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759
	return 0;
}

static __net_exit void inet_exit_net(struct net *net)
{
}

static __net_initdata struct pernet_operations af_inet_ops = {
	.init = inet_init_net,
	.exit = inet_exit_net,
};

static int __init init_inet_pernet_ops(void)
{
	return register_pernet_subsys(&af_inet_ops);
}

L
Linus Torvalds 已提交
1760 1761
static int ipv4_proc_init(void);

1762 1763 1764 1765
/*
 *	IP protocol layer initialiser
 */

1766 1767
static struct packet_offload ip_packet_offload __read_mostly = {
	.type = cpu_to_be16(ETH_P_IP),
1768 1769 1770 1771 1772
	.callbacks = {
		.gso_segment = inet_gso_segment,
		.gro_receive = inet_gro_receive,
		.gro_complete = inet_gro_complete,
	},
1773 1774
};

E
Eric Dumazet 已提交
1775 1776 1777
static const struct net_offload ipip_offload = {
	.callbacks = {
		.gso_segment	= inet_gso_segment,
1778
		.gro_receive	= ipip_gro_receive,
1779
		.gro_complete	= ipip_gro_complete,
E
Eric Dumazet 已提交
1780 1781 1782
	},
};

1783 1784 1785 1786 1787
static int __init ipip_offload_init(void)
{
	return inet_add_offload(&ipip_offload, IPPROTO_IPIP);
}

1788 1789 1790 1791 1792
static int __init ipv4_offload_init(void)
{
	/*
	 * Add offloads
	 */
1793
	if (udpv4_offload_init() < 0)
1794
		pr_crit("%s: Cannot add UDP protocol offload\n", __func__);
1795 1796
	if (tcpv4_offload_init() < 0)
		pr_crit("%s: Cannot add TCP protocol offload\n", __func__);
1797 1798
	if (ipip_offload_init() < 0)
		pr_crit("%s: Cannot add IPIP protocol offload\n", __func__);
1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810

	dev_add_offload(&ip_packet_offload);
	return 0;
}

fs_initcall(ipv4_offload_init);

static struct packet_type ip_packet_type __read_mostly = {
	.type = cpu_to_be16(ETH_P_IP),
	.func = ip_rcv,
};

L
Linus Torvalds 已提交
1811 1812 1813 1814 1815 1816
static int __init inet_init(void)
{
	struct inet_protosw *q;
	struct list_head *r;
	int rc = -EINVAL;

1817
	sock_skb_cb_check_size(sizeof(struct inet_skb_parm));
L
Linus Torvalds 已提交
1818 1819 1820

	rc = proto_register(&tcp_prot, 1);
	if (rc)
1821
		goto out;
L
Linus Torvalds 已提交
1822 1823 1824 1825 1826 1827 1828 1829 1830

	rc = proto_register(&udp_prot, 1);
	if (rc)
		goto out_unregister_tcp_proto;

	rc = proto_register(&raw_prot, 1);
	if (rc)
		goto out_unregister_udp_proto;

1831 1832 1833 1834
	rc = proto_register(&ping_prot, 1);
	if (rc)
		goto out_unregister_raw_proto;

L
Linus Torvalds 已提交
1835
	/*
1836
	 *	Tell SOCKET that we are alive...
L
Linus Torvalds 已提交
1837 1838
	 */

1839
	(void)sock_register(&inet_family_ops);
L
Linus Torvalds 已提交
1840

1841 1842 1843 1844
#ifdef CONFIG_SYSCTL
	ip_static_sysctl_init();
#endif

L
Linus Torvalds 已提交
1845 1846 1847 1848 1849
	/*
	 *	Add all the base protocols.
	 */

	if (inet_add_protocol(&icmp_protocol, IPPROTO_ICMP) < 0)
J
Joe Perches 已提交
1850
		pr_crit("%s: Cannot add ICMP protocol\n", __func__);
L
Linus Torvalds 已提交
1851
	if (inet_add_protocol(&udp_protocol, IPPROTO_UDP) < 0)
J
Joe Perches 已提交
1852
		pr_crit("%s: Cannot add UDP protocol\n", __func__);
L
Linus Torvalds 已提交
1853
	if (inet_add_protocol(&tcp_protocol, IPPROTO_TCP) < 0)
J
Joe Perches 已提交
1854
		pr_crit("%s: Cannot add TCP protocol\n", __func__);
L
Linus Torvalds 已提交
1855 1856
#ifdef CONFIG_IP_MULTICAST
	if (inet_add_protocol(&igmp_protocol, IPPROTO_IGMP) < 0)
J
Joe Perches 已提交
1857
		pr_crit("%s: Cannot add IGMP protocol\n", __func__);
L
Linus Torvalds 已提交
1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872
#endif

	/* Register the socket-side information for inet_create. */
	for (r = &inetsw[0]; r < &inetsw[SOCK_MAX]; ++r)
		INIT_LIST_HEAD(r);

	for (q = inetsw_array; q < &inetsw_array[INETSW_ARRAY_LEN]; ++q)
		inet_register_protosw(q);

	/*
	 *	Set the ARP module up
	 */

	arp_init();

1873 1874 1875
	/*
	 *	Set the IP module up
	 */
L
Linus Torvalds 已提交
1876 1877 1878 1879 1880 1881

	ip_init();

	/* Setup TCP slab cache for open requests. */
	tcp_init();

H
Hideo Aoki 已提交
1882 1883 1884
	/* Setup UDP memory threshold */
	udp_init();

1885 1886
	/* Add UDP-Lite (RFC 3828) */
	udplite4_register();
L
Linus Torvalds 已提交
1887

1888 1889
	ping_init();

L
Linus Torvalds 已提交
1890 1891 1892 1893
	/*
	 *	Set the ICMP layer up
	 */

1894 1895
	if (icmp_init() < 0)
		panic("Failed to create the ICMP control socket.\n");
L
Linus Torvalds 已提交
1896 1897 1898 1899 1900

	/*
	 *	Initialise the multicast router
	 */
#if defined(CONFIG_IP_MROUTE)
W
Wang Chen 已提交
1901
	if (ip_mr_init())
J
Joe Perches 已提交
1902
		pr_crit("%s: Cannot init ipv4 mroute\n", __func__);
L
Linus Torvalds 已提交
1903
#endif
1904 1905 1906

	if (init_inet_pernet_ops())
		pr_crit("%s: Cannot init ipv4 inet pernet ops\n", __func__);
L
Linus Torvalds 已提交
1907 1908
	/*
	 *	Initialise per-cpu ipv4 mibs
1909
	 */
L
Linus Torvalds 已提交
1910

S
Stephen Hemminger 已提交
1911
	if (init_ipv4_mibs())
J
Joe Perches 已提交
1912
		pr_crit("%s: Cannot init ipv4 mibs\n", __func__);
1913

L
Linus Torvalds 已提交
1914 1915 1916 1917
	ipv4_proc_init();

	ipfrag_init();

1918 1919
	dev_add_pack(&ip_packet_type);

1920 1921
	ip_tunnel_core_init();

L
Linus Torvalds 已提交
1922 1923 1924
	rc = 0;
out:
	return rc;
1925 1926
out_unregister_raw_proto:
	proto_unregister(&raw_prot);
L
Linus Torvalds 已提交
1927 1928
out_unregister_udp_proto:
	proto_unregister(&udp_prot);
1929 1930
out_unregister_tcp_proto:
	proto_unregister(&tcp_prot);
L
Linus Torvalds 已提交
1931 1932 1933
	goto out;
}

1934
fs_initcall(inet_init);
L
Linus Torvalds 已提交
1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948

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

#ifdef CONFIG_PROC_FS
static int __init ipv4_proc_init(void)
{
	int rc = 0;

	if (raw_proc_init())
		goto out_raw;
	if (tcp4_proc_init())
		goto out_tcp;
	if (udp4_proc_init())
		goto out_udp;
1949 1950
	if (ping_proc_init())
		goto out_ping;
L
Linus Torvalds 已提交
1951 1952 1953 1954 1955
	if (ip_misc_proc_init())
		goto out_misc;
out:
	return rc;
out_misc:
1956 1957
	ping_proc_exit();
out_ping:
L
Linus Torvalds 已提交
1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973
	udp4_proc_exit();
out_udp:
	tcp4_proc_exit();
out_tcp:
	raw_proc_exit();
out_raw:
	rc = -ENOMEM;
	goto out;
}

#else /* CONFIG_PROC_FS */
static int __init ipv4_proc_init(void)
{
	return 0;
}
#endif /* CONFIG_PROC_FS */