ip.h 14 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413414415416417418419420421422423424425426427428429430431432433434435436437438439440441442443444445446447448449450451452453454455456457458459460461462463464465466467468469470471472473
  1. /*
  2. * INET An implementation of the TCP/IP protocol suite for the LINUX
  3. * operating system. INET is implemented using the BSD Socket
  4. * interface as the means of communication with the user level.
  5. *
  6. * Definitions for the IP module.
  7. *
  8. * Version: @(#)ip.h 1.0.2 05/07/93
  9. *
  10. * Authors: Ross Biro
  11. * Fred N. van Kempen, <waltje@uWalt.NL.Mugnet.ORG>
  12. * Alan Cox, <gw4pts@gw4pts.ampr.org>
  13. *
  14. * Changes:
  15. * Mike McLagan : Routing by source
  16. *
  17. * This program is free software; you can redistribute it and/or
  18. * modify it under the terms of the GNU General Public License
  19. * as published by the Free Software Foundation; either version
  20. * 2 of the License, or (at your option) any later version.
  21. */
  22. #ifndef _IP_H
  23. #define _IP_H
  24. #include <linux/types.h>
  25. #include <linux/ip.h>
  26. #include <linux/in.h>
  27. #include <linux/skbuff.h>
  28. #include <net/inet_sock.h>
  29. #include <net/snmp.h>
  30. #include <net/flow.h>
  31. struct sock;
  32. struct inet_skb_parm {
  33. struct ip_options opt; /* Compiled IP options */
  34. unsigned char flags;
  35. #define IPSKB_FORWARDED 1
  36. #define IPSKB_XFRM_TUNNEL_SIZE 2
  37. #define IPSKB_XFRM_TRANSFORMED 4
  38. #define IPSKB_FRAG_COMPLETE 8
  39. #define IPSKB_REROUTED 16
  40. u16 frag_max_size;
  41. };
  42. static inline unsigned int ip_hdrlen(const struct sk_buff *skb)
  43. {
  44. return ip_hdr(skb)->ihl * 4;
  45. }
  46. struct ipcm_cookie {
  47. __be32 addr;
  48. int oif;
  49. struct ip_options_rcu *opt;
  50. __u8 tx_flags;
  51. };
  52. #define IPCB(skb) ((struct inet_skb_parm*)((skb)->cb))
  53. struct ip_ra_chain {
  54. struct ip_ra_chain __rcu *next;
  55. struct sock *sk;
  56. union {
  57. void (*destructor)(struct sock *);
  58. struct sock *saved_sk;
  59. };
  60. struct rcu_head rcu;
  61. };
  62. extern struct ip_ra_chain __rcu *ip_ra_chain;
  63. /* IP flags. */
  64. #define IP_CE 0x8000 /* Flag: "Congestion" */
  65. #define IP_DF 0x4000 /* Flag: "Don't Fragment" */
  66. #define IP_MF 0x2000 /* Flag: "More Fragments" */
  67. #define IP_OFFSET 0x1FFF /* "Fragment Offset" part */
  68. #define IP_FRAG_TIME (30 * HZ) /* fragment lifetime */
  69. struct msghdr;
  70. struct net_device;
  71. struct packet_type;
  72. struct rtable;
  73. struct sockaddr;
  74. extern int igmp_mc_proc_init(void);
  75. /*
  76. * Functions provided by ip.c
  77. */
  78. extern int ip_build_and_send_pkt(struct sk_buff *skb, struct sock *sk,
  79. __be32 saddr, __be32 daddr,
  80. struct ip_options_rcu *opt);
  81. extern int ip_rcv(struct sk_buff *skb, struct net_device *dev,
  82. struct packet_type *pt, struct net_device *orig_dev);
  83. extern int ip_local_deliver(struct sk_buff *skb);
  84. extern int ip_mr_input(struct sk_buff *skb);
  85. extern int ip_output(struct sk_buff *skb);
  86. extern int ip_mc_output(struct sk_buff *skb);
  87. extern int ip_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *));
  88. extern int ip_do_nat(struct sk_buff *skb);
  89. extern void ip_send_check(struct iphdr *ip);
  90. extern int __ip_local_out(struct sk_buff *skb);
  91. extern int ip_local_out(struct sk_buff *skb);
  92. extern int ip_queue_xmit(struct sk_buff *skb, struct flowi *fl);
  93. extern void ip_init(void);
  94. extern int ip_append_data(struct sock *sk, struct flowi4 *fl4,
  95. int getfrag(void *from, char *to, int offset, int len,
  96. int odd, struct sk_buff *skb),
  97. void *from, int len, int protolen,
  98. struct ipcm_cookie *ipc,
  99. struct rtable **rt,
  100. unsigned int flags);
  101. extern int ip_generic_getfrag(void *from, char *to, int offset, int len, int odd, struct sk_buff *skb);
  102. extern ssize_t ip_append_page(struct sock *sk, struct flowi4 *fl4, struct page *page,
  103. int offset, size_t size, int flags);
  104. extern struct sk_buff *__ip_make_skb(struct sock *sk,
  105. struct flowi4 *fl4,
  106. struct sk_buff_head *queue,
  107. struct inet_cork *cork);
  108. extern int ip_send_skb(struct net *net, struct sk_buff *skb);
  109. extern int ip_push_pending_frames(struct sock *sk, struct flowi4 *fl4);
  110. extern void ip_flush_pending_frames(struct sock *sk);
  111. extern struct sk_buff *ip_make_skb(struct sock *sk,
  112. struct flowi4 *fl4,
  113. int getfrag(void *from, char *to, int offset, int len,
  114. int odd, struct sk_buff *skb),
  115. void *from, int length, int transhdrlen,
  116. struct ipcm_cookie *ipc,
  117. struct rtable **rtp,
  118. unsigned int flags);
  119. static inline struct sk_buff *ip_finish_skb(struct sock *sk, struct flowi4 *fl4)
  120. {
  121. return __ip_make_skb(sk, fl4, &sk->sk_write_queue, &inet_sk(sk)->cork.base);
  122. }
  123. /* datagram.c */
  124. extern int ip4_datagram_connect(struct sock *sk,
  125. struct sockaddr *uaddr, int addr_len);
  126. extern void ip4_datagram_release_cb(struct sock *sk);
  127. struct ip_reply_arg {
  128. struct kvec iov[1];
  129. int flags;
  130. __wsum csum;
  131. int csumoffset; /* u16 offset of csum in iov[0].iov_base */
  132. /* -1 if not needed */
  133. int bound_dev_if;
  134. u8 tos;
  135. };
  136. #define IP_REPLY_ARG_NOSRCCHECK 1
  137. static inline __u8 ip_reply_arg_flowi_flags(const struct ip_reply_arg *arg)
  138. {
  139. return (arg->flags & IP_REPLY_ARG_NOSRCCHECK) ? FLOWI_FLAG_ANYSRC : 0;
  140. }
  141. void ip_send_unicast_reply(struct net *net, struct sk_buff *skb, __be32 daddr,
  142. __be32 saddr, const struct ip_reply_arg *arg,
  143. unsigned int len);
  144. struct ipv4_config {
  145. int log_martians;
  146. int no_pmtu_disc;
  147. };
  148. extern struct ipv4_config ipv4_config;
  149. #define IP_INC_STATS(net, field) SNMP_INC_STATS64((net)->mib.ip_statistics, field)
  150. #define IP_INC_STATS_BH(net, field) SNMP_INC_STATS64_BH((net)->mib.ip_statistics, field)
  151. #define IP_ADD_STATS(net, field, val) SNMP_ADD_STATS64((net)->mib.ip_statistics, field, val)
  152. #define IP_ADD_STATS_BH(net, field, val) SNMP_ADD_STATS64_BH((net)->mib.ip_statistics, field, val)
  153. #define IP_UPD_PO_STATS(net, field, val) SNMP_UPD_PO_STATS64((net)->mib.ip_statistics, field, val)
  154. #define IP_UPD_PO_STATS_BH(net, field, val) SNMP_UPD_PO_STATS64_BH((net)->mib.ip_statistics, field, val)
  155. #define NET_INC_STATS(net, field) SNMP_INC_STATS((net)->mib.net_statistics, field)
  156. #define NET_INC_STATS_BH(net, field) SNMP_INC_STATS_BH((net)->mib.net_statistics, field)
  157. #define NET_INC_STATS_USER(net, field) SNMP_INC_STATS_USER((net)->mib.net_statistics, field)
  158. #define NET_ADD_STATS_BH(net, field, adnd) SNMP_ADD_STATS_BH((net)->mib.net_statistics, field, adnd)
  159. #define NET_ADD_STATS_USER(net, field, adnd) SNMP_ADD_STATS_USER((net)->mib.net_statistics, field, adnd)
  160. extern unsigned long snmp_fold_field(void __percpu *mib[], int offt);
  161. #if BITS_PER_LONG==32
  162. extern u64 snmp_fold_field64(void __percpu *mib[], int offt, size_t sync_off);
  163. #else
  164. static inline u64 snmp_fold_field64(void __percpu *mib[], int offt, size_t syncp_off)
  165. {
  166. return snmp_fold_field(mib, offt);
  167. }
  168. #endif
  169. extern int snmp_mib_init(void __percpu *ptr[2], size_t mibsize, size_t align);
  170. static inline void snmp_mib_free(void __percpu *ptr[SNMP_ARRAY_SZ])
  171. {
  172. int i;
  173. BUG_ON(ptr == NULL);
  174. for (i = 0; i < SNMP_ARRAY_SZ; i++) {
  175. free_percpu(ptr[i]);
  176. ptr[i] = NULL;
  177. }
  178. }
  179. extern struct local_ports {
  180. seqlock_t lock;
  181. int range[2];
  182. } sysctl_local_ports;
  183. extern void inet_get_local_port_range(int *low, int *high);
  184. extern unsigned long *sysctl_local_reserved_ports;
  185. static inline int inet_is_reserved_local_port(int port)
  186. {
  187. return test_bit(port, sysctl_local_reserved_ports);
  188. }
  189. extern int sysctl_ip_nonlocal_bind;
  190. /* From inetpeer.c */
  191. extern int inet_peer_threshold;
  192. extern int inet_peer_minttl;
  193. extern int inet_peer_maxttl;
  194. /* From ip_input.c */
  195. extern int sysctl_ip_early_demux;
  196. /* From ip_output.c */
  197. extern int sysctl_ip_dynaddr;
  198. extern void ipfrag_init(void);
  199. extern void ip_static_sysctl_init(void);
  200. static inline bool ip_is_fragment(const struct iphdr *iph)
  201. {
  202. return (iph->frag_off & htons(IP_MF | IP_OFFSET)) != 0;
  203. }
  204. #ifdef CONFIG_INET
  205. #include <net/dst.h>
  206. /* The function in 2.2 was invalid, producing wrong result for
  207. * check=0xFEFF. It was noticed by Arthur Skawina _year_ ago. --ANK(000625) */
  208. static inline
  209. int ip_decrease_ttl(struct iphdr *iph)
  210. {
  211. u32 check = (__force u32)iph->check;
  212. check += (__force u32)htons(0x0100);
  213. iph->check = (__force __sum16)(check + (check>=0xFFFF));
  214. return --iph->ttl;
  215. }
  216. static inline
  217. int ip_dont_fragment(struct sock *sk, struct dst_entry *dst)
  218. {
  219. return inet_sk(sk)->pmtudisc == IP_PMTUDISC_DO ||
  220. (inet_sk(sk)->pmtudisc == IP_PMTUDISC_WANT &&
  221. !(dst_metric_locked(dst, RTAX_MTU)));
  222. }
  223. extern void __ip_select_ident(struct iphdr *iph, struct dst_entry *dst, int more);
  224. static inline void ip_select_ident(struct iphdr *iph, struct dst_entry *dst, struct sock *sk)
  225. {
  226. if (iph->frag_off & htons(IP_DF)) {
  227. /* This is only to work around buggy Windows95/2000
  228. * VJ compression implementations. If the ID field
  229. * does not change, they drop every other packet in
  230. * a TCP stream using header compression.
  231. */
  232. iph->id = (sk && inet_sk(sk)->inet_daddr) ?
  233. htons(inet_sk(sk)->inet_id++) : 0;
  234. } else
  235. __ip_select_ident(iph, dst, 0);
  236. }
  237. static inline void ip_select_ident_more(struct iphdr *iph, struct dst_entry *dst, struct sock *sk, int more)
  238. {
  239. if (iph->frag_off & htons(IP_DF)) {
  240. if (sk && inet_sk(sk)->inet_daddr) {
  241. iph->id = htons(inet_sk(sk)->inet_id);
  242. inet_sk(sk)->inet_id += 1 + more;
  243. } else
  244. iph->id = 0;
  245. } else
  246. __ip_select_ident(iph, dst, more);
  247. }
  248. /*
  249. * Map a multicast IP onto multicast MAC for type ethernet.
  250. */
  251. static inline void ip_eth_mc_map(__be32 naddr, char *buf)
  252. {
  253. __u32 addr=ntohl(naddr);
  254. buf[0]=0x01;
  255. buf[1]=0x00;
  256. buf[2]=0x5e;
  257. buf[5]=addr&0xFF;
  258. addr>>=8;
  259. buf[4]=addr&0xFF;
  260. addr>>=8;
  261. buf[3]=addr&0x7F;
  262. }
  263. /*
  264. * Map a multicast IP onto multicast MAC for type IP-over-InfiniBand.
  265. * Leave P_Key as 0 to be filled in by driver.
  266. */
  267. static inline void ip_ib_mc_map(__be32 naddr, const unsigned char *broadcast, char *buf)
  268. {
  269. __u32 addr;
  270. unsigned char scope = broadcast[5] & 0xF;
  271. buf[0] = 0; /* Reserved */
  272. buf[1] = 0xff; /* Multicast QPN */
  273. buf[2] = 0xff;
  274. buf[3] = 0xff;
  275. addr = ntohl(naddr);
  276. buf[4] = 0xff;
  277. buf[5] = 0x10 | scope; /* scope from broadcast address */
  278. buf[6] = 0x40; /* IPv4 signature */
  279. buf[7] = 0x1b;
  280. buf[8] = broadcast[8]; /* P_Key */
  281. buf[9] = broadcast[9];
  282. buf[10] = 0;
  283. buf[11] = 0;
  284. buf[12] = 0;
  285. buf[13] = 0;
  286. buf[14] = 0;
  287. buf[15] = 0;
  288. buf[19] = addr & 0xff;
  289. addr >>= 8;
  290. buf[18] = addr & 0xff;
  291. addr >>= 8;
  292. buf[17] = addr & 0xff;
  293. addr >>= 8;
  294. buf[16] = addr & 0x0f;
  295. }
  296. static inline void ip_ipgre_mc_map(__be32 naddr, const unsigned char *broadcast, char *buf)
  297. {
  298. if ((broadcast[0] | broadcast[1] | broadcast[2] | broadcast[3]) != 0)
  299. memcpy(buf, broadcast, 4);
  300. else
  301. memcpy(buf, &naddr, sizeof(naddr));
  302. }
  303. #if IS_ENABLED(CONFIG_IPV6)
  304. #include <linux/ipv6.h>
  305. #endif
  306. static __inline__ void inet_reset_saddr(struct sock *sk)
  307. {
  308. inet_sk(sk)->inet_rcv_saddr = inet_sk(sk)->inet_saddr = 0;
  309. #if IS_ENABLED(CONFIG_IPV6)
  310. if (sk->sk_family == PF_INET6) {
  311. struct ipv6_pinfo *np = inet6_sk(sk);
  312. memset(&np->saddr, 0, sizeof(np->saddr));
  313. memset(&np->rcv_saddr, 0, sizeof(np->rcv_saddr));
  314. }
  315. #endif
  316. }
  317. #endif
  318. static inline int sk_mc_loop(struct sock *sk)
  319. {
  320. if (!sk)
  321. return 1;
  322. switch (sk->sk_family) {
  323. case AF_INET:
  324. return inet_sk(sk)->mc_loop;
  325. #if IS_ENABLED(CONFIG_IPV6)
  326. case AF_INET6:
  327. return inet6_sk(sk)->mc_loop;
  328. #endif
  329. }
  330. WARN_ON(1);
  331. return 1;
  332. }
  333. extern bool ip_call_ra_chain(struct sk_buff *skb);
  334. /*
  335. * Functions provided by ip_fragment.c
  336. */
  337. enum ip_defrag_users {
  338. IP_DEFRAG_LOCAL_DELIVER,
  339. IP_DEFRAG_CALL_RA_CHAIN,
  340. IP_DEFRAG_CONNTRACK_IN,
  341. __IP_DEFRAG_CONNTRACK_IN_END = IP_DEFRAG_CONNTRACK_IN + USHRT_MAX,
  342. IP_DEFRAG_CONNTRACK_OUT,
  343. __IP_DEFRAG_CONNTRACK_OUT_END = IP_DEFRAG_CONNTRACK_OUT + USHRT_MAX,
  344. IP_DEFRAG_CONNTRACK_BRIDGE_IN,
  345. __IP_DEFRAG_CONNTRACK_BRIDGE_IN = IP_DEFRAG_CONNTRACK_BRIDGE_IN + USHRT_MAX,
  346. IP_DEFRAG_VS_IN,
  347. IP_DEFRAG_VS_OUT,
  348. IP_DEFRAG_VS_FWD,
  349. IP_DEFRAG_AF_PACKET,
  350. IP_DEFRAG_MACVLAN,
  351. };
  352. int ip_defrag(struct sk_buff *skb, u32 user);
  353. #ifdef CONFIG_INET
  354. struct sk_buff *ip_check_defrag(struct sk_buff *skb, u32 user);
  355. #else
  356. static inline struct sk_buff *ip_check_defrag(struct sk_buff *skb, u32 user)
  357. {
  358. return skb;
  359. }
  360. #endif
  361. int ip_frag_mem(struct net *net);
  362. int ip_frag_nqueues(struct net *net);
  363. /*
  364. * Functions provided by ip_forward.c
  365. */
  366. extern int ip_forward(struct sk_buff *skb);
  367. /*
  368. * Functions provided by ip_options.c
  369. */
  370. extern void ip_options_build(struct sk_buff *skb, struct ip_options *opt,
  371. __be32 daddr, struct rtable *rt, int is_frag);
  372. extern int ip_options_echo(struct ip_options *dopt, struct sk_buff *skb);
  373. extern void ip_options_fragment(struct sk_buff *skb);
  374. extern int ip_options_compile(struct net *net,
  375. struct ip_options *opt, struct sk_buff *skb);
  376. extern int ip_options_get(struct net *net, struct ip_options_rcu **optp,
  377. unsigned char *data, int optlen);
  378. extern int ip_options_get_from_user(struct net *net, struct ip_options_rcu **optp,
  379. unsigned char __user *data, int optlen);
  380. extern void ip_options_undo(struct ip_options * opt);
  381. extern void ip_forward_options(struct sk_buff *skb);
  382. extern int ip_options_rcv_srr(struct sk_buff *skb);
  383. /*
  384. * Functions provided by ip_sockglue.c
  385. */
  386. extern void ipv4_pktinfo_prepare(struct sk_buff *skb);
  387. extern void ip_cmsg_recv(struct msghdr *msg, struct sk_buff *skb);
  388. extern int ip_cmsg_send(struct net *net,
  389. struct msghdr *msg, struct ipcm_cookie *ipc);
  390. extern int ip_setsockopt(struct sock *sk, int level, int optname, char __user *optval, unsigned int optlen);
  391. extern int ip_getsockopt(struct sock *sk, int level, int optname, char __user *optval, int __user *optlen);
  392. extern int compat_ip_setsockopt(struct sock *sk, int level,
  393. int optname, char __user *optval, unsigned int optlen);
  394. extern int compat_ip_getsockopt(struct sock *sk, int level,
  395. int optname, char __user *optval, int __user *optlen);
  396. extern int ip_ra_control(struct sock *sk, unsigned char on, void (*destructor)(struct sock *));
  397. extern int ip_recv_error(struct sock *sk, struct msghdr *msg, int len);
  398. extern void ip_icmp_error(struct sock *sk, struct sk_buff *skb, int err,
  399. __be16 port, u32 info, u8 *payload);
  400. extern void ip_local_error(struct sock *sk, int err, __be32 daddr, __be16 dport,
  401. u32 info);
  402. #ifdef CONFIG_PROC_FS
  403. extern int ip_misc_proc_init(void);
  404. #endif
  405. #endif /* _IP_H */