1// SPDX-License-Identifier: GPL-2.0-or-later 2/* 3 * PF_INET6 socket protocol family 4 * Linux INET6 implementation 5 * 6 * Authors: 7 * Pedro Roque <roque@di.fc.ul.pt> 8 * 9 * Adapted from linux/net/ipv4/af_inet.c 10 * 11 * Fixes: 12 * piggy, Karl Knutson : Socket protocol table 13 * Hideaki YOSHIFUJI : sin6_scope_id support 14 * Arnaldo Melo : check proc_net_create return, cleanups 15 */ 16 17#define pr_fmt(fmt) "IPv6: " fmt 18 19#include <linux/module.h> 20#include <linux/capability.h> 21#include <linux/errno.h> 22#include <linux/types.h> 23#include <linux/socket.h> 24#include <linux/in.h> 25#include <linux/kernel.h> 26#include <linux/timer.h> 27#include <linux/string.h> 28#include <linux/sockios.h> 29#include <linux/net.h> 30#include <linux/fcntl.h> 31#include <linux/mm.h> 32#include <linux/interrupt.h> 33#include <linux/proc_fs.h> 34#include <linux/stat.h> 35#include <linux/init.h> 36#include <linux/slab.h> 37 38#include <linux/inet.h> 39#include <linux/netdevice.h> 40#include <linux/icmpv6.h> 41#include <linux/netfilter_ipv6.h> 42 43#include <net/ip.h> 44#include <net/ipv6.h> 45#include <net/udp.h> 46#include <net/udplite.h> 47#include <net/tcp.h> 48#include <net/ping.h> 49#include <net/protocol.h> 50#include <net/inet_common.h> 51#include <net/route.h> 52#include <net/transp_v6.h> 53#include <net/ip6_route.h> 54#include <net/addrconf.h> 55#include <net/ipv6_stubs.h> 56#include <net/ndisc.h> 57#ifdef CONFIG_IPV6_TUNNEL 58#include <net/ip6_tunnel.h> 59#endif 60#include <net/calipso.h> 61#include <net/seg6.h> 62#include <net/rpl.h> 63#include <net/compat.h> 64#include <net/xfrm.h> 65#include <net/ioam6.h> 66#include <net/rawv6.h> 67 68#include <linux/uaccess.h> 69#include <linux/mroute6.h> 70 71#include "ip6_offload.h" 72 73MODULE_AUTHOR("Cast of dozens"); 74MODULE_DESCRIPTION("IPv6 protocol stack for Linux"); 75MODULE_LICENSE("GPL"); 76 77/* The inetsw6 table contains everything that inet6_create needs to 78 * build a new socket. 79 */ 80static struct list_head inetsw6[SOCK_MAX]; 81static DEFINE_SPINLOCK(inetsw6_lock); 82 83struct ipv6_params ipv6_defaults = { 84 .disable_ipv6 = 0, 85 .autoconf = 1, 86}; 87 88static int disable_ipv6_mod; 89 90module_param_named(disable, disable_ipv6_mod, int, 0444); 91MODULE_PARM_DESC(disable, "Disable IPv6 module such that it is non-functional"); 92 93module_param_named(disable_ipv6, ipv6_defaults.disable_ipv6, int, 0444); 94MODULE_PARM_DESC(disable_ipv6, "Disable IPv6 on all interfaces"); 95 96module_param_named(autoconf, ipv6_defaults.autoconf, int, 0444); 97MODULE_PARM_DESC(autoconf, "Enable IPv6 address autoconfiguration on all interfaces"); 98 99bool ipv6_mod_enabled(void) 100{ 101 return disable_ipv6_mod == 0; 102} 103EXPORT_SYMBOL_GPL(ipv6_mod_enabled); 104 105static struct ipv6_pinfo *inet6_sk_generic(struct sock *sk) 106{ 107 const int offset = sk->sk_prot->ipv6_pinfo_offset; 108 109 return (struct ipv6_pinfo *)(((u8 *)sk) + offset); 110} 111 112void inet6_sock_destruct(struct sock *sk) 113{ 114 inet6_cleanup_sock(sk); 115 inet_sock_destruct(sk); 116} 117EXPORT_SYMBOL_GPL(inet6_sock_destruct); 118 119static int inet6_create(struct net *net, struct socket *sock, int protocol, 120 int kern) 121{ 122 struct inet_sock *inet; 123 struct ipv6_pinfo *np; 124 struct sock *sk; 125 struct inet_protosw *answer; 126 struct proto *answer_prot; 127 unsigned char answer_flags; 128 int try_loading_module = 0; 129 int err; 130 131 if (protocol < 0 || protocol >= IPPROTO_MAX) 132 return -EINVAL; 133 134 /* Look for the requested type/protocol pair. */ 135lookup_protocol: 136 err = -ESOCKTNOSUPPORT; 137 rcu_read_lock(); 138 list_for_each_entry_rcu(answer, &inetsw6[sock->type], list) { 139 140 err = 0; 141 /* Check the non-wild match. */ 142 if (protocol == answer->protocol) { 143 if (protocol != IPPROTO_IP) 144 break; 145 } else { 146 /* Check for the two wild cases. */ 147 if (IPPROTO_IP == protocol) { 148 protocol = answer->protocol; 149 break; 150 } 151 if (IPPROTO_IP == answer->protocol) 152 break; 153 } 154 err = -EPROTONOSUPPORT; 155 } 156 157 if (err) { 158 if (try_loading_module < 2) { 159 rcu_read_unlock(); 160 /* 161 * Be more specific, e.g. net-pf-10-proto-132-type-1 162 * (net-pf-PF_INET6-proto-IPPROTO_SCTP-type-SOCK_STREAM) 163 */ 164 if (++try_loading_module == 1) 165 request_module("net-pf-%d-proto-%d-type-%d", 166 PF_INET6, protocol, sock->type); 167 /* 168 * Fall back to generic, e.g. net-pf-10-proto-132 169 * (net-pf-PF_INET6-proto-IPPROTO_SCTP) 170 */ 171 else 172 request_module("net-pf-%d-proto-%d", 173 PF_INET6, protocol); 174 goto lookup_protocol; 175 } else 176 goto out_rcu_unlock; 177 } 178 179 err = -EPERM; 180 if (sock->type == SOCK_RAW && !kern && 181 !ns_capable(net->user_ns, CAP_NET_RAW)) 182 goto out_rcu_unlock; 183 184 sock->ops = answer->ops; 185 answer_prot = answer->prot; 186 answer_flags = answer->flags; 187 rcu_read_unlock(); 188 189 WARN_ON(!answer_prot->slab); 190 191 err = -ENOBUFS; 192 sk = sk_alloc(net, PF_INET6, GFP_KERNEL, answer_prot, kern); 193 if (!sk) 194 goto out; 195 196 sock_init_data(sock, sk); 197 198 err = 0; 199 if (INET_PROTOSW_REUSE & answer_flags) 200 sk->sk_reuse = SK_CAN_REUSE; 201 202 if (INET_PROTOSW_ICSK & answer_flags) 203 inet_init_csk_locks(sk); 204 205 inet = inet_sk(sk); 206 inet_assign_bit(IS_ICSK, sk, INET_PROTOSW_ICSK & answer_flags); 207 208 if (SOCK_RAW == sock->type) { 209 inet->inet_num = protocol; 210 if (IPPROTO_RAW == protocol) 211 inet_set_bit(HDRINCL, sk); 212 } 213 214 sk->sk_destruct = inet6_sock_destruct; 215 sk->sk_family = PF_INET6; 216 sk->sk_protocol = protocol; 217 218 sk->sk_backlog_rcv = answer->prot->backlog_rcv; 219 220 inet_sk(sk)->pinet6 = np = inet6_sk_generic(sk); 221 np->hop_limit = -1; 222 np->mcast_hops = IPV6_DEFAULT_MCASTHOPS; 223 np->mc_loop = 1; 224 np->mc_all = 1; 225 np->pmtudisc = IPV6_PMTUDISC_WANT; 226 np->repflow = net->ipv6.sysctl.flowlabel_reflect & FLOWLABEL_REFLECT_ESTABLISHED; 227 sk->sk_ipv6only = net->ipv6.sysctl.bindv6only; 228 sk->sk_txrehash = READ_ONCE(net->core.sysctl_txrehash); 229 230 /* Init the ipv4 part of the socket since we can have sockets 231 * using v6 API for ipv4. 232 */ 233 inet->uc_ttl = -1; 234 235 inet_set_bit(MC_LOOP, sk); 236 inet->mc_ttl = 1; 237 inet->mc_index = 0; 238 RCU_INIT_POINTER(inet->mc_list, NULL); 239 inet->rcv_tos = 0; 240 241 if (READ_ONCE(net->ipv4.sysctl_ip_no_pmtu_disc)) 242 inet->pmtudisc = IP_PMTUDISC_DONT; 243 else 244 inet->pmtudisc = IP_PMTUDISC_WANT; 245 246 if (inet->inet_num) { 247 /* It assumes that any protocol which allows 248 * the user to assign a number at socket 249 * creation time automatically shares. 250 */ 251 inet->inet_sport = htons(inet->inet_num); 252 err = sk->sk_prot->hash(sk); 253 if (err) { 254 sk_common_release(sk); 255 goto out; 256 } 257 } 258 if (sk->sk_prot->init) { 259 err = sk->sk_prot->init(sk); 260 if (err) { 261 sk_common_release(sk); 262 goto out; 263 } 264 } 265 266 if (!kern) { 267 err = BPF_CGROUP_RUN_PROG_INET_SOCK(sk); 268 if (err) { 269 sk_common_release(sk); 270 goto out; 271 } 272 } 273out: 274 return err; 275out_rcu_unlock: 276 rcu_read_unlock(); 277 goto out; 278} 279 280static int __inet6_bind(struct sock *sk, struct sockaddr *uaddr, int addr_len, 281 u32 flags) 282{ 283 struct sockaddr_in6 *addr = (struct sockaddr_in6 *)uaddr; 284 struct inet_sock *inet = inet_sk(sk); 285 struct ipv6_pinfo *np = inet6_sk(sk); 286 struct net *net = sock_net(sk); 287 __be32 v4addr = 0; 288 unsigned short snum; 289 bool saved_ipv6only; 290 int addr_type = 0; 291 int err = 0; 292 293 if (addr->sin6_family != AF_INET6) 294 return -EAFNOSUPPORT; 295 296 addr_type = ipv6_addr_type(&addr->sin6_addr); 297 if ((addr_type & IPV6_ADDR_MULTICAST) && sk->sk_type == SOCK_STREAM) 298 return -EINVAL; 299 300 snum = ntohs(addr->sin6_port); 301 if (!(flags & BIND_NO_CAP_NET_BIND_SERVICE) && 302 snum && inet_port_requires_bind_service(net, snum) && 303 !ns_capable(net->user_ns, CAP_NET_BIND_SERVICE)) 304 return -EACCES; 305 306 if (flags & BIND_WITH_LOCK) 307 lock_sock(sk); 308 309 /* Check these errors (active socket, double bind). */ 310 if (sk->sk_state != TCP_CLOSE || inet->inet_num) { 311 err = -EINVAL; 312 goto out; 313 } 314 315 /* Check if the address belongs to the host. */ 316 if (addr_type == IPV6_ADDR_MAPPED) { 317 struct net_device *dev = NULL; 318 int chk_addr_ret; 319 320 /* Binding to v4-mapped address on a v6-only socket 321 * makes no sense 322 */ 323 if (ipv6_only_sock(sk)) { 324 err = -EINVAL; 325 goto out; 326 } 327 328 rcu_read_lock(); 329 if (sk->sk_bound_dev_if) { 330 dev = dev_get_by_index_rcu(net, sk->sk_bound_dev_if); 331 if (!dev) { 332 err = -ENODEV; 333 goto out_unlock; 334 } 335 } 336 337 /* Reproduce AF_INET checks to make the bindings consistent */ 338 v4addr = addr->sin6_addr.s6_addr32[3]; 339 chk_addr_ret = inet_addr_type_dev_table(net, dev, v4addr); 340 rcu_read_unlock(); 341 342 if (!inet_addr_valid_or_nonlocal(net, inet, v4addr, 343 chk_addr_ret)) { 344 err = -EADDRNOTAVAIL; 345 goto out; 346 } 347 } else { 348 if (addr_type != IPV6_ADDR_ANY) { 349 struct net_device *dev = NULL; 350 351 rcu_read_lock(); 352 if (__ipv6_addr_needs_scope_id(addr_type)) { 353 if (addr_len >= sizeof(struct sockaddr_in6) && 354 addr->sin6_scope_id) { 355 /* Override any existing binding, if another one 356 * is supplied by user. 357 */ 358 sk->sk_bound_dev_if = addr->sin6_scope_id; 359 } 360 361 /* Binding to link-local address requires an interface */ 362 if (!sk->sk_bound_dev_if) { 363 err = -EINVAL; 364 goto out_unlock; 365 } 366 } 367 368 if (sk->sk_bound_dev_if) { 369 dev = dev_get_by_index_rcu(net, sk->sk_bound_dev_if); 370 if (!dev) { 371 err = -ENODEV; 372 goto out_unlock; 373 } 374 } 375 376 /* ipv4 addr of the socket is invalid. Only the 377 * unspecified and mapped address have a v4 equivalent. 378 */ 379 v4addr = LOOPBACK4_IPV6; 380 if (!(addr_type & IPV6_ADDR_MULTICAST)) { 381 if (!ipv6_can_nonlocal_bind(net, inet) && 382 !ipv6_chk_addr(net, &addr->sin6_addr, 383 dev, 0)) { 384 err = -EADDRNOTAVAIL; 385 goto out_unlock; 386 } 387 } 388 rcu_read_unlock(); 389 } 390 } 391 392 inet->inet_rcv_saddr = v4addr; 393 inet->inet_saddr = v4addr; 394 395 sk->sk_v6_rcv_saddr = addr->sin6_addr; 396 397 if (!(addr_type & IPV6_ADDR_MULTICAST)) 398 np->saddr = addr->sin6_addr; 399 400 saved_ipv6only = sk->sk_ipv6only; 401 if (addr_type != IPV6_ADDR_ANY && addr_type != IPV6_ADDR_MAPPED) 402 sk->sk_ipv6only = 1; 403 404 /* Make sure we are allowed to bind here. */ 405 if (snum || !(inet_test_bit(BIND_ADDRESS_NO_PORT, sk) || 406 (flags & BIND_FORCE_ADDRESS_NO_PORT))) { 407 err = sk->sk_prot->get_port(sk, snum); 408 if (err) { 409 sk->sk_ipv6only = saved_ipv6only; 410 inet_reset_saddr(sk); 411 goto out; 412 } 413 if (!(flags & BIND_FROM_BPF)) { 414 err = BPF_CGROUP_RUN_PROG_INET6_POST_BIND(sk); 415 if (err) { 416 sk->sk_ipv6only = saved_ipv6only; 417 inet_reset_saddr(sk); 418 if (sk->sk_prot->put_port) 419 sk->sk_prot->put_port(sk); 420 goto out; 421 } 422 } 423 } 424 425 if (addr_type != IPV6_ADDR_ANY) 426 sk->sk_userlocks |= SOCK_BINDADDR_LOCK; 427 if (snum) 428 sk->sk_userlocks |= SOCK_BINDPORT_LOCK; 429 inet->inet_sport = htons(inet->inet_num); 430 inet->inet_dport = 0; 431 inet->inet_daddr = 0; 432out: 433 if (flags & BIND_WITH_LOCK) 434 release_sock(sk); 435 return err; 436out_unlock: 437 rcu_read_unlock(); 438 goto out; 439} 440 441int inet6_bind_sk(struct sock *sk, struct sockaddr *uaddr, int addr_len) 442{ 443 u32 flags = BIND_WITH_LOCK; 444 const struct proto *prot; 445 int err = 0; 446 447 /* IPV6_ADDRFORM can change sk->sk_prot under us. */ 448 prot = READ_ONCE(sk->sk_prot); 449 /* If the socket has its own bind function then use it. */ 450 if (prot->bind) 451 return prot->bind(sk, uaddr, addr_len); 452 453 if (addr_len < SIN6_LEN_RFC2133) 454 return -EINVAL; 455 456 /* BPF prog is run before any checks are done so that if the prog 457 * changes context in a wrong way it will be caught. 458 */ 459 err = BPF_CGROUP_RUN_PROG_INET_BIND_LOCK(sk, uaddr, &addr_len, 460 CGROUP_INET6_BIND, &flags); 461 if (err) 462 return err; 463 464 return __inet6_bind(sk, uaddr, addr_len, flags); 465} 466 467/* bind for INET6 API */ 468int inet6_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len) 469{ 470 return inet6_bind_sk(sock->sk, uaddr, addr_len); 471} 472EXPORT_SYMBOL(inet6_bind); 473 474int inet6_release(struct socket *sock) 475{ 476 struct sock *sk = sock->sk; 477 478 if (!sk) 479 return -EINVAL; 480 481 /* Free mc lists */ 482 ipv6_sock_mc_close(sk); 483 484 /* Free ac lists */ 485 ipv6_sock_ac_close(sk); 486 487 return inet_release(sock); 488} 489EXPORT_SYMBOL(inet6_release); 490 491void inet6_cleanup_sock(struct sock *sk) 492{ 493 struct ipv6_pinfo *np = inet6_sk(sk); 494 struct sk_buff *skb; 495 struct ipv6_txoptions *opt; 496 497 /* Release rx options */ 498 499 skb = xchg(&np->pktoptions, NULL); 500 kfree_skb(skb); 501 502 skb = xchg(&np->rxpmtu, NULL); 503 kfree_skb(skb); 504 505 /* Free flowlabels */ 506 fl6_free_socklist(sk); 507 508 /* Free tx options */ 509 510 opt = xchg((__force struct ipv6_txoptions **)&np->opt, NULL); 511 if (opt) { 512 atomic_sub(opt->tot_len, &sk->sk_omem_alloc); 513 txopt_put(opt); 514 } 515} 516EXPORT_SYMBOL_GPL(inet6_cleanup_sock); 517 518/* 519 * This does both peername and sockname. 520 */ 521int inet6_getname(struct socket *sock, struct sockaddr *uaddr, 522 int peer) 523{ 524 struct sockaddr_in6 *sin = (struct sockaddr_in6 *)uaddr; 525 int sin_addr_len = sizeof(*sin); 526 struct sock *sk = sock->sk; 527 struct inet_sock *inet = inet_sk(sk); 528 struct ipv6_pinfo *np = inet6_sk(sk); 529 530 sin->sin6_family = AF_INET6; 531 sin->sin6_flowinfo = 0; 532 sin->sin6_scope_id = 0; 533 lock_sock(sk); 534 if (peer) { 535 if (!inet->inet_dport || 536 (((1 << sk->sk_state) & (TCPF_CLOSE | TCPF_SYN_SENT)) && 537 peer == 1)) { 538 release_sock(sk); 539 return -ENOTCONN; 540 } 541 sin->sin6_port = inet->inet_dport; 542 sin->sin6_addr = sk->sk_v6_daddr; 543 if (np->sndflow) 544 sin->sin6_flowinfo = np->flow_label; 545 BPF_CGROUP_RUN_SA_PROG(sk, (struct sockaddr *)sin, &sin_addr_len, 546 CGROUP_INET6_GETPEERNAME); 547 } else { 548 if (ipv6_addr_any(&sk->sk_v6_rcv_saddr)) 549 sin->sin6_addr = np->saddr; 550 else 551 sin->sin6_addr = sk->sk_v6_rcv_saddr; 552 sin->sin6_port = inet->inet_sport; 553 BPF_CGROUP_RUN_SA_PROG(sk, (struct sockaddr *)sin, &sin_addr_len, 554 CGROUP_INET6_GETSOCKNAME); 555 } 556 sin->sin6_scope_id = ipv6_iface_scope_id(&sin->sin6_addr, 557 sk->sk_bound_dev_if); 558 release_sock(sk); 559 return sin_addr_len; 560} 561EXPORT_SYMBOL(inet6_getname); 562 563int inet6_ioctl(struct socket *sock, unsigned int cmd, unsigned long arg) 564{ 565 void __user *argp = (void __user *)arg; 566 struct sock *sk = sock->sk; 567 struct net *net = sock_net(sk); 568 const struct proto *prot; 569 570 switch (cmd) { 571 case SIOCADDRT: 572 case SIOCDELRT: { 573 struct in6_rtmsg rtmsg; 574 575 if (copy_from_user(&rtmsg, argp, sizeof(rtmsg))) 576 return -EFAULT; 577 return ipv6_route_ioctl(net, cmd, &rtmsg); 578 } 579 case SIOCSIFADDR: 580 return addrconf_add_ifaddr(net, argp); 581 case SIOCDIFADDR: 582 return addrconf_del_ifaddr(net, argp); 583 case SIOCSIFDSTADDR: 584 return addrconf_set_dstaddr(net, argp); 585 default: 586 /* IPV6_ADDRFORM can change sk->sk_prot under us. */ 587 prot = READ_ONCE(sk->sk_prot); 588 if (!prot->ioctl) 589 return -ENOIOCTLCMD; 590 return sk_ioctl(sk, cmd, (void __user *)arg); 591 } 592 /*NOTREACHED*/ 593 return 0; 594} 595EXPORT_SYMBOL(inet6_ioctl); 596 597#ifdef CONFIG_COMPAT 598struct compat_in6_rtmsg { 599 struct in6_addr rtmsg_dst; 600 struct in6_addr rtmsg_src; 601 struct in6_addr rtmsg_gateway; 602 u32 rtmsg_type; 603 u16 rtmsg_dst_len; 604 u16 rtmsg_src_len; 605 u32 rtmsg_metric; 606 u32 rtmsg_info; 607 u32 rtmsg_flags; 608 s32 rtmsg_ifindex; 609}; 610 611static int inet6_compat_routing_ioctl(struct sock *sk, unsigned int cmd, 612 struct compat_in6_rtmsg __user *ur) 613{ 614 struct in6_rtmsg rt; 615 616 if (copy_from_user(&rt.rtmsg_dst, &ur->rtmsg_dst, 617 3 * sizeof(struct in6_addr)) || 618 get_user(rt.rtmsg_type, &ur->rtmsg_type) || 619 get_user(rt.rtmsg_dst_len, &ur->rtmsg_dst_len) || 620 get_user(rt.rtmsg_src_len, &ur->rtmsg_src_len) || 621 get_user(rt.rtmsg_metric, &ur->rtmsg_metric) || 622 get_user(rt.rtmsg_info, &ur->rtmsg_info) || 623 get_user(rt.rtmsg_flags, &ur->rtmsg_flags) || 624 get_user(rt.rtmsg_ifindex, &ur->rtmsg_ifindex)) 625 return -EFAULT; 626 627 628 return ipv6_route_ioctl(sock_net(sk), cmd, &rt); 629} 630 631int inet6_compat_ioctl(struct socket *sock, unsigned int cmd, unsigned long arg) 632{ 633 void __user *argp = compat_ptr(arg); 634 struct sock *sk = sock->sk; 635 636 switch (cmd) { 637 case SIOCADDRT: 638 case SIOCDELRT: 639 return inet6_compat_routing_ioctl(sk, cmd, argp); 640 default: 641 return -ENOIOCTLCMD; 642 } 643} 644EXPORT_SYMBOL_GPL(inet6_compat_ioctl); 645#endif /* CONFIG_COMPAT */ 646 647INDIRECT_CALLABLE_DECLARE(int udpv6_sendmsg(struct sock *, struct msghdr *, 648 size_t)); 649int inet6_sendmsg(struct socket *sock, struct msghdr *msg, size_t size) 650{ 651 struct sock *sk = sock->sk; 652 const struct proto *prot; 653 654 if (unlikely(inet_send_prepare(sk))) 655 return -EAGAIN; 656 657 /* IPV6_ADDRFORM can change sk->sk_prot under us. */ 658 prot = READ_ONCE(sk->sk_prot); 659 return INDIRECT_CALL_2(prot->sendmsg, tcp_sendmsg, udpv6_sendmsg, 660 sk, msg, size); 661} 662 663INDIRECT_CALLABLE_DECLARE(int udpv6_recvmsg(struct sock *, struct msghdr *, 664 size_t, int, int *)); 665int inet6_recvmsg(struct socket *sock, struct msghdr *msg, size_t size, 666 int flags) 667{ 668 struct sock *sk = sock->sk; 669 const struct proto *prot; 670 int addr_len = 0; 671 int err; 672 673 if (likely(!(flags & MSG_ERRQUEUE))) 674 sock_rps_record_flow(sk); 675 676 /* IPV6_ADDRFORM can change sk->sk_prot under us. */ 677 prot = READ_ONCE(sk->sk_prot); 678 err = INDIRECT_CALL_2(prot->recvmsg, tcp_recvmsg, udpv6_recvmsg, 679 sk, msg, size, flags, &addr_len); 680 if (err >= 0) 681 msg->msg_namelen = addr_len; 682 return err; 683} 684 685const struct proto_ops inet6_stream_ops = { 686 .family = PF_INET6, 687 .owner = THIS_MODULE, 688 .release = inet6_release, 689 .bind = inet6_bind, 690 .connect = inet_stream_connect, /* ok */ 691 .socketpair = sock_no_socketpair, /* a do nothing */ 692 .accept = inet_accept, /* ok */ 693 .getname = inet6_getname, 694 .poll = tcp_poll, /* ok */ 695 .ioctl = inet6_ioctl, /* must change */ 696 .gettstamp = sock_gettstamp, 697 .listen = inet_listen, /* ok */ 698 .shutdown = inet_shutdown, /* ok */ 699 .setsockopt = sock_common_setsockopt, /* ok */ 700 .getsockopt = sock_common_getsockopt, /* ok */ 701 .sendmsg = inet6_sendmsg, /* retpoline's sake */ 702 .recvmsg = inet6_recvmsg, /* retpoline's sake */ 703#ifdef CONFIG_MMU 704 .mmap = tcp_mmap, 705#endif 706 .splice_eof = inet_splice_eof, 707 .sendmsg_locked = tcp_sendmsg_locked, 708 .splice_read = tcp_splice_read, 709 .read_sock = tcp_read_sock, 710 .read_skb = tcp_read_skb, 711 .peek_len = tcp_peek_len, 712#ifdef CONFIG_COMPAT 713 .compat_ioctl = inet6_compat_ioctl, 714#endif 715 .set_rcvlowat = tcp_set_rcvlowat, 716}; 717 718const struct proto_ops inet6_dgram_ops = { 719 .family = PF_INET6, 720 .owner = THIS_MODULE, 721 .release = inet6_release, 722 .bind = inet6_bind, 723 .connect = inet_dgram_connect, /* ok */ 724 .socketpair = sock_no_socketpair, /* a do nothing */ 725 .accept = sock_no_accept, /* a do nothing */ 726 .getname = inet6_getname, 727 .poll = udp_poll, /* ok */ 728 .ioctl = inet6_ioctl, /* must change */ 729 .gettstamp = sock_gettstamp, 730 .listen = sock_no_listen, /* ok */ 731 .shutdown = inet_shutdown, /* ok */ 732 .setsockopt = sock_common_setsockopt, /* ok */ 733 .getsockopt = sock_common_getsockopt, /* ok */ 734 .sendmsg = inet6_sendmsg, /* retpoline's sake */ 735 .recvmsg = inet6_recvmsg, /* retpoline's sake */ 736 .read_skb = udp_read_skb, 737 .mmap = sock_no_mmap, 738 .set_peek_off = sk_set_peek_off, 739#ifdef CONFIG_COMPAT 740 .compat_ioctl = inet6_compat_ioctl, 741#endif 742}; 743 744static const struct net_proto_family inet6_family_ops = { 745 .family = PF_INET6, 746 .create = inet6_create, 747 .owner = THIS_MODULE, 748}; 749 750int inet6_register_protosw(struct inet_protosw *p) 751{ 752 struct list_head *lh; 753 struct inet_protosw *answer; 754 struct list_head *last_perm; 755 int protocol = p->protocol; 756 int ret; 757 758 spin_lock_bh(&inetsw6_lock); 759 760 ret = -EINVAL; 761 if (p->type >= SOCK_MAX) 762 goto out_illegal; 763 764 /* If we are trying to override a permanent protocol, bail. */ 765 answer = NULL; 766 ret = -EPERM; 767 last_perm = &inetsw6[p->type]; 768 list_for_each(lh, &inetsw6[p->type]) { 769 answer = list_entry(lh, struct inet_protosw, list); 770 771 /* Check only the non-wild match. */ 772 if (INET_PROTOSW_PERMANENT & answer->flags) { 773 if (protocol == answer->protocol) 774 break; 775 last_perm = lh; 776 } 777 778 answer = NULL; 779 } 780 if (answer) 781 goto out_permanent; 782 783 /* Add the new entry after the last permanent entry if any, so that 784 * the new entry does not override a permanent entry when matched with 785 * a wild-card protocol. But it is allowed to override any existing 786 * non-permanent entry. This means that when we remove this entry, the 787 * system automatically returns to the old behavior. 788 */ 789 list_add_rcu(&p->list, last_perm); 790 ret = 0; 791out: 792 spin_unlock_bh(&inetsw6_lock); 793 return ret; 794 795out_permanent: 796 pr_err("Attempt to override permanent protocol %d\n", protocol); 797 goto out; 798 799out_illegal: 800 pr_err("Ignoring attempt to register invalid socket type %d\n", 801 p->type); 802 goto out; 803} 804EXPORT_SYMBOL(inet6_register_protosw); 805 806void 807inet6_unregister_protosw(struct inet_protosw *p) 808{ 809 if (INET_PROTOSW_PERMANENT & p->flags) { 810 pr_err("Attempt to unregister permanent protocol %d\n", 811 p->protocol); 812 } else { 813 spin_lock_bh(&inetsw6_lock); 814 list_del_rcu(&p->list); 815 spin_unlock_bh(&inetsw6_lock); 816 817 synchronize_net(); 818 } 819} 820EXPORT_SYMBOL(inet6_unregister_protosw); 821 822int inet6_sk_rebuild_header(struct sock *sk) 823{ 824 struct ipv6_pinfo *np = inet6_sk(sk); 825 struct dst_entry *dst; 826 827 dst = __sk_dst_check(sk, np->dst_cookie); 828 829 if (!dst) { 830 struct inet_sock *inet = inet_sk(sk); 831 struct in6_addr *final_p, final; 832 struct flowi6 fl6; 833 834 memset(&fl6, 0, sizeof(fl6)); 835 fl6.flowi6_proto = sk->sk_protocol; 836 fl6.daddr = sk->sk_v6_daddr; 837 fl6.saddr = np->saddr; 838 fl6.flowlabel = np->flow_label; 839 fl6.flowi6_oif = sk->sk_bound_dev_if; 840 fl6.flowi6_mark = sk->sk_mark; 841 fl6.fl6_dport = inet->inet_dport; 842 fl6.fl6_sport = inet->inet_sport; 843 fl6.flowi6_uid = sk->sk_uid; 844 security_sk_classify_flow(sk, flowi6_to_flowi_common(&fl6)); 845 846 rcu_read_lock(); 847 final_p = fl6_update_dst(&fl6, rcu_dereference(np->opt), 848 &final); 849 rcu_read_unlock(); 850 851 dst = ip6_dst_lookup_flow(sock_net(sk), sk, &fl6, final_p); 852 if (IS_ERR(dst)) { 853 sk->sk_route_caps = 0; 854 WRITE_ONCE(sk->sk_err_soft, -PTR_ERR(dst)); 855 return PTR_ERR(dst); 856 } 857 858 ip6_dst_store(sk, dst, NULL, NULL); 859 } 860 861 return 0; 862} 863EXPORT_SYMBOL_GPL(inet6_sk_rebuild_header); 864 865bool ipv6_opt_accepted(const struct sock *sk, const struct sk_buff *skb, 866 const struct inet6_skb_parm *opt) 867{ 868 const struct ipv6_pinfo *np = inet6_sk(sk); 869 870 if (np->rxopt.all) { 871 if (((opt->flags & IP6SKB_HOPBYHOP) && 872 (np->rxopt.bits.hopopts || np->rxopt.bits.ohopopts)) || 873 (ip6_flowinfo((struct ipv6hdr *) skb_network_header(skb)) && 874 np->rxopt.bits.rxflow) || 875 (opt->srcrt && (np->rxopt.bits.srcrt || 876 np->rxopt.bits.osrcrt)) || 877 ((opt->dst1 || opt->dst0) && 878 (np->rxopt.bits.dstopts || np->rxopt.bits.odstopts))) 879 return true; 880 } 881 return false; 882} 883EXPORT_SYMBOL_GPL(ipv6_opt_accepted); 884 885static struct packet_type ipv6_packet_type __read_mostly = { 886 .type = cpu_to_be16(ETH_P_IPV6), 887 .func = ipv6_rcv, 888 .list_func = ipv6_list_rcv, 889}; 890 891static int __init ipv6_packet_init(void) 892{ 893 dev_add_pack(&ipv6_packet_type); 894 return 0; 895} 896 897static void ipv6_packet_cleanup(void) 898{ 899 dev_remove_pack(&ipv6_packet_type); 900} 901 902static int __net_init ipv6_init_mibs(struct net *net) 903{ 904 int i; 905 906 net->mib.udp_stats_in6 = alloc_percpu(struct udp_mib); 907 if (!net->mib.udp_stats_in6) 908 return -ENOMEM; 909 net->mib.udplite_stats_in6 = alloc_percpu(struct udp_mib); 910 if (!net->mib.udplite_stats_in6) 911 goto err_udplite_mib; 912 net->mib.ipv6_statistics = alloc_percpu(struct ipstats_mib); 913 if (!net->mib.ipv6_statistics) 914 goto err_ip_mib; 915 916 for_each_possible_cpu(i) { 917 struct ipstats_mib *af_inet6_stats; 918 af_inet6_stats = per_cpu_ptr(net->mib.ipv6_statistics, i); 919 u64_stats_init(&af_inet6_stats->syncp); 920 } 921 922 923 net->mib.icmpv6_statistics = alloc_percpu(struct icmpv6_mib); 924 if (!net->mib.icmpv6_statistics) 925 goto err_icmp_mib; 926 net->mib.icmpv6msg_statistics = kzalloc(sizeof(struct icmpv6msg_mib), 927 GFP_KERNEL); 928 if (!net->mib.icmpv6msg_statistics) 929 goto err_icmpmsg_mib; 930 return 0; 931 932err_icmpmsg_mib: 933 free_percpu(net->mib.icmpv6_statistics); 934err_icmp_mib: 935 free_percpu(net->mib.ipv6_statistics); 936err_ip_mib: 937 free_percpu(net->mib.udplite_stats_in6); 938err_udplite_mib: 939 free_percpu(net->mib.udp_stats_in6); 940 return -ENOMEM; 941} 942 943static void ipv6_cleanup_mibs(struct net *net) 944{ 945 free_percpu(net->mib.udp_stats_in6); 946 free_percpu(net->mib.udplite_stats_in6); 947 free_percpu(net->mib.ipv6_statistics); 948 free_percpu(net->mib.icmpv6_statistics); 949 kfree(net->mib.icmpv6msg_statistics); 950} 951 952static int __net_init inet6_net_init(struct net *net) 953{ 954 int err = 0; 955 956 net->ipv6.sysctl.bindv6only = 0; 957 net->ipv6.sysctl.icmpv6_time = 1*HZ; 958 net->ipv6.sysctl.icmpv6_echo_ignore_all = 0; 959 net->ipv6.sysctl.icmpv6_echo_ignore_multicast = 0; 960 net->ipv6.sysctl.icmpv6_echo_ignore_anycast = 0; 961 net->ipv6.sysctl.icmpv6_error_anycast_as_unicast = 0; 962 963 /* By default, rate limit error messages. 964 * Except for pmtu discovery, it would break it. 965 * proc_do_large_bitmap needs pointer to the bitmap. 966 */ 967 bitmap_set(net->ipv6.sysctl.icmpv6_ratemask, 0, ICMPV6_ERRMSG_MAX + 1); 968 bitmap_clear(net->ipv6.sysctl.icmpv6_ratemask, ICMPV6_PKT_TOOBIG, 1); 969 net->ipv6.sysctl.icmpv6_ratemask_ptr = net->ipv6.sysctl.icmpv6_ratemask; 970 971 net->ipv6.sysctl.flowlabel_consistency = 1; 972 net->ipv6.sysctl.auto_flowlabels = IP6_DEFAULT_AUTO_FLOW_LABELS; 973 net->ipv6.sysctl.idgen_retries = 3; 974 net->ipv6.sysctl.idgen_delay = 1 * HZ; 975 net->ipv6.sysctl.flowlabel_state_ranges = 0; 976 net->ipv6.sysctl.max_dst_opts_cnt = IP6_DEFAULT_MAX_DST_OPTS_CNT; 977 net->ipv6.sysctl.max_hbh_opts_cnt = IP6_DEFAULT_MAX_HBH_OPTS_CNT; 978 net->ipv6.sysctl.max_dst_opts_len = IP6_DEFAULT_MAX_DST_OPTS_LEN; 979 net->ipv6.sysctl.max_hbh_opts_len = IP6_DEFAULT_MAX_HBH_OPTS_LEN; 980 net->ipv6.sysctl.fib_notify_on_flag_change = 0; 981 atomic_set(&net->ipv6.fib6_sernum, 1); 982 983 net->ipv6.sysctl.ioam6_id = IOAM6_DEFAULT_ID; 984 net->ipv6.sysctl.ioam6_id_wide = IOAM6_DEFAULT_ID_WIDE; 985 986 err = ipv6_init_mibs(net); 987 if (err) 988 return err; 989#ifdef CONFIG_PROC_FS 990 err = udp6_proc_init(net); 991 if (err) 992 goto out; 993 err = tcp6_proc_init(net); 994 if (err) 995 goto proc_tcp6_fail; 996 err = ac6_proc_init(net); 997 if (err) 998 goto proc_ac6_fail; 999#endif 1000 return err; 1001 1002#ifdef CONFIG_PROC_FS 1003proc_ac6_fail: 1004 tcp6_proc_exit(net); 1005proc_tcp6_fail: 1006 udp6_proc_exit(net); 1007out: 1008 ipv6_cleanup_mibs(net); 1009 return err; 1010#endif 1011} 1012 1013static void __net_exit inet6_net_exit(struct net *net) 1014{ 1015#ifdef CONFIG_PROC_FS 1016 udp6_proc_exit(net); 1017 tcp6_proc_exit(net); 1018 ac6_proc_exit(net); 1019#endif 1020 ipv6_cleanup_mibs(net); 1021} 1022 1023static struct pernet_operations inet6_net_ops = { 1024 .init = inet6_net_init, 1025 .exit = inet6_net_exit, 1026}; 1027 1028static int ipv6_route_input(struct sk_buff *skb) 1029{ 1030 ip6_route_input(skb); 1031 return skb_dst(skb)->error; 1032} 1033 1034static const struct ipv6_stub ipv6_stub_impl = { 1035 .ipv6_sock_mc_join = ipv6_sock_mc_join, 1036 .ipv6_sock_mc_drop = ipv6_sock_mc_drop, 1037 .ipv6_dst_lookup_flow = ip6_dst_lookup_flow, 1038 .ipv6_route_input = ipv6_route_input, 1039 .fib6_get_table = fib6_get_table, 1040 .fib6_table_lookup = fib6_table_lookup, 1041 .fib6_lookup = fib6_lookup, 1042 .fib6_select_path = fib6_select_path, 1043 .ip6_mtu_from_fib6 = ip6_mtu_from_fib6, 1044 .fib6_nh_init = fib6_nh_init, 1045 .fib6_nh_release = fib6_nh_release, 1046 .fib6_nh_release_dsts = fib6_nh_release_dsts, 1047 .fib6_update_sernum = fib6_update_sernum_stub, 1048 .fib6_rt_update = fib6_rt_update, 1049 .ip6_del_rt = ip6_del_rt, 1050 .udpv6_encap_enable = udpv6_encap_enable, 1051 .ndisc_send_na = ndisc_send_na, 1052#if IS_ENABLED(CONFIG_XFRM) 1053 .xfrm6_local_rxpmtu = xfrm6_local_rxpmtu, 1054 .xfrm6_udp_encap_rcv = xfrm6_udp_encap_rcv, 1055 .xfrm6_rcv_encap = xfrm6_rcv_encap, 1056#endif 1057 .nd_tbl = &nd_tbl, 1058 .ipv6_fragment = ip6_fragment, 1059 .ipv6_dev_find = ipv6_dev_find, 1060}; 1061 1062static const struct ipv6_bpf_stub ipv6_bpf_stub_impl = { 1063 .inet6_bind = __inet6_bind, 1064 .udp6_lib_lookup = __udp6_lib_lookup, 1065 .ipv6_setsockopt = do_ipv6_setsockopt, 1066 .ipv6_getsockopt = do_ipv6_getsockopt, 1067 .ipv6_dev_get_saddr = ipv6_dev_get_saddr, 1068}; 1069 1070static int __init inet6_init(void) 1071{ 1072 struct list_head *r; 1073 int err = 0; 1074 1075 sock_skb_cb_check_size(sizeof(struct inet6_skb_parm)); 1076 1077 /* Register the socket-side information for inet6_create. */ 1078 for (r = &inetsw6[0]; r < &inetsw6[SOCK_MAX]; ++r) 1079 INIT_LIST_HEAD(r); 1080 1081 raw_hashinfo_init(&raw_v6_hashinfo); 1082 1083 if (disable_ipv6_mod) { 1084 pr_info("Loaded, but administratively disabled, reboot required to enable\n"); 1085 goto out; 1086 } 1087 1088 err = proto_register(&tcpv6_prot, 1); 1089 if (err) 1090 goto out; 1091 1092 err = proto_register(&udpv6_prot, 1); 1093 if (err) 1094 goto out_unregister_tcp_proto; 1095 1096 err = proto_register(&udplitev6_prot, 1); 1097 if (err) 1098 goto out_unregister_udp_proto; 1099 1100 err = proto_register(&rawv6_prot, 1); 1101 if (err) 1102 goto out_unregister_udplite_proto; 1103 1104 err = proto_register(&pingv6_prot, 1); 1105 if (err) 1106 goto out_unregister_raw_proto; 1107 1108 /* We MUST register RAW sockets before we create the ICMP6, 1109 * IGMP6, or NDISC control sockets. 1110 */ 1111 err = rawv6_init(); 1112 if (err) 1113 goto out_unregister_ping_proto; 1114 1115 /* Register the family here so that the init calls below will 1116 * be able to create sockets. (?? is this dangerous ??) 1117 */ 1118 err = sock_register(&inet6_family_ops); 1119 if (err) 1120 goto out_sock_register_fail; 1121 1122 /* 1123 * ipngwg API draft makes clear that the correct semantics 1124 * for TCP and UDP is to consider one TCP and UDP instance 1125 * in a host available by both INET and INET6 APIs and 1126 * able to communicate via both network protocols. 1127 */ 1128 1129 err = register_pernet_subsys(&inet6_net_ops); 1130 if (err) 1131 goto register_pernet_fail; 1132 err = ip6_mr_init(); 1133 if (err) 1134 goto ipmr_fail; 1135 err = icmpv6_init(); 1136 if (err) 1137 goto icmp_fail; 1138 err = ndisc_init(); 1139 if (err) 1140 goto ndisc_fail; 1141 err = igmp6_init(); 1142 if (err) 1143 goto igmp_fail; 1144 1145 err = ipv6_netfilter_init(); 1146 if (err) 1147 goto netfilter_fail; 1148 /* Create /proc/foo6 entries. */ 1149#ifdef CONFIG_PROC_FS 1150 err = -ENOMEM; 1151 if (raw6_proc_init()) 1152 goto proc_raw6_fail; 1153 if (udplite6_proc_init()) 1154 goto proc_udplite6_fail; 1155 if (ipv6_misc_proc_init()) 1156 goto proc_misc6_fail; 1157 if (if6_proc_init()) 1158 goto proc_if6_fail; 1159#endif 1160 err = ip6_route_init(); 1161 if (err) 1162 goto ip6_route_fail; 1163 err = ndisc_late_init(); 1164 if (err) 1165 goto ndisc_late_fail; 1166 err = ip6_flowlabel_init(); 1167 if (err) 1168 goto ip6_flowlabel_fail; 1169 err = ipv6_anycast_init(); 1170 if (err) 1171 goto ipv6_anycast_fail; 1172 err = addrconf_init(); 1173 if (err) 1174 goto addrconf_fail; 1175 1176 /* Init v6 extension headers. */ 1177 err = ipv6_exthdrs_init(); 1178 if (err) 1179 goto ipv6_exthdrs_fail; 1180 1181 err = ipv6_frag_init(); 1182 if (err) 1183 goto ipv6_frag_fail; 1184 1185 /* Init v6 transport protocols. */ 1186 err = udpv6_init(); 1187 if (err) 1188 goto udpv6_fail; 1189 1190 err = udplitev6_init(); 1191 if (err) 1192 goto udplitev6_fail; 1193 1194 err = udpv6_offload_init(); 1195 if (err) 1196 goto udpv6_offload_fail; 1197 1198 err = tcpv6_init(); 1199 if (err) 1200 goto tcpv6_fail; 1201 1202 err = ipv6_packet_init(); 1203 if (err) 1204 goto ipv6_packet_fail; 1205 1206 err = pingv6_init(); 1207 if (err) 1208 goto pingv6_fail; 1209 1210 err = calipso_init(); 1211 if (err) 1212 goto calipso_fail; 1213 1214 err = seg6_init(); 1215 if (err) 1216 goto seg6_fail; 1217 1218 err = rpl_init(); 1219 if (err) 1220 goto rpl_fail; 1221 1222 err = ioam6_init(); 1223 if (err) 1224 goto ioam6_fail; 1225 1226 err = igmp6_late_init(); 1227 if (err) 1228 goto igmp6_late_err; 1229 1230#ifdef CONFIG_SYSCTL 1231 err = ipv6_sysctl_register(); 1232 if (err) 1233 goto sysctl_fail; 1234#endif 1235 1236 /* ensure that ipv6 stubs are visible only after ipv6 is ready */ 1237 wmb(); 1238 ipv6_stub = &ipv6_stub_impl; 1239 ipv6_bpf_stub = &ipv6_bpf_stub_impl; 1240out: 1241 return err; 1242 1243#ifdef CONFIG_SYSCTL 1244sysctl_fail: 1245 igmp6_late_cleanup(); 1246#endif 1247igmp6_late_err: 1248 ioam6_exit(); 1249ioam6_fail: 1250 rpl_exit(); 1251rpl_fail: 1252 seg6_exit(); 1253seg6_fail: 1254 calipso_exit(); 1255calipso_fail: 1256 pingv6_exit(); 1257pingv6_fail: 1258 ipv6_packet_cleanup(); 1259ipv6_packet_fail: 1260 tcpv6_exit(); 1261tcpv6_fail: 1262 udpv6_offload_exit(); 1263udpv6_offload_fail: 1264 udplitev6_exit(); 1265udplitev6_fail: 1266 udpv6_exit(); 1267udpv6_fail: 1268 ipv6_frag_exit(); 1269ipv6_frag_fail: 1270 ipv6_exthdrs_exit(); 1271ipv6_exthdrs_fail: 1272 addrconf_cleanup(); 1273addrconf_fail: 1274 ipv6_anycast_cleanup(); 1275ipv6_anycast_fail: 1276 ip6_flowlabel_cleanup(); 1277ip6_flowlabel_fail: 1278 ndisc_late_cleanup(); 1279ndisc_late_fail: 1280 ip6_route_cleanup(); 1281ip6_route_fail: 1282#ifdef CONFIG_PROC_FS 1283 if6_proc_exit(); 1284proc_if6_fail: 1285 ipv6_misc_proc_exit(); 1286proc_misc6_fail: 1287 udplite6_proc_exit(); 1288proc_udplite6_fail: 1289 raw6_proc_exit(); 1290proc_raw6_fail: 1291#endif 1292 ipv6_netfilter_fini(); 1293netfilter_fail: 1294 igmp6_cleanup(); 1295igmp_fail: 1296 ndisc_cleanup(); 1297ndisc_fail: 1298 icmpv6_cleanup(); 1299icmp_fail: 1300 ip6_mr_cleanup(); 1301ipmr_fail: 1302 unregister_pernet_subsys(&inet6_net_ops); 1303register_pernet_fail: 1304 sock_unregister(PF_INET6); 1305 rtnl_unregister_all(PF_INET6); 1306out_sock_register_fail: 1307 rawv6_exit(); 1308out_unregister_ping_proto: 1309 proto_unregister(&pingv6_prot); 1310out_unregister_raw_proto: 1311 proto_unregister(&rawv6_prot); 1312out_unregister_udplite_proto: 1313 proto_unregister(&udplitev6_prot); 1314out_unregister_udp_proto: 1315 proto_unregister(&udpv6_prot); 1316out_unregister_tcp_proto: 1317 proto_unregister(&tcpv6_prot); 1318 goto out; 1319} 1320module_init(inet6_init); 1321 1322MODULE_ALIAS_NETPROTO(PF_INET6); 1323