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.
6 * Definitions for the IP module.
8 * Version: @(#)ip.h 1.0.2 05/07/93
15 * Mike McLagan : Routing by source
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.
25 #include <linux/types.h>
28 #include <linux/skbuff.h>
30 #include <net/inet_sock.h>
31 #include <net/route.h>
37 struct inet_skb_parm {
38 struct ip_options opt; /* Compiled IP options */
41 #define IPSKB_FORWARDED 1
42 #define IPSKB_XFRM_TUNNEL_SIZE 2
43 #define IPSKB_XFRM_TRANSFORMED 4
44 #define IPSKB_FRAG_COMPLETE 8
45 #define IPSKB_REROUTED 16
50 static inline unsigned int ip_hdrlen(const struct sk_buff *skb)
52 return ip_hdr(skb)->ihl * 4;
58 struct ip_options_rcu *opt;
65 #define IPCB(skb) ((struct inet_skb_parm*)((skb)->cb))
68 struct ip_ra_chain __rcu *next;
71 void (*destructor)(struct sock *);
72 struct sock *saved_sk;
77 extern struct ip_ra_chain __rcu *ip_ra_chain;
80 #define IP_CE 0x8000 /* Flag: "Congestion" */
81 #define IP_DF 0x4000 /* Flag: "Don't Fragment" */
82 #define IP_MF 0x2000 /* Flag: "More Fragments" */
83 #define IP_OFFSET 0x1FFF /* "Fragment Offset" part */
85 #define IP_FRAG_TIME (30 * HZ) /* fragment lifetime */
93 int igmp_mc_proc_init(void);
96 * Functions provided by ip.c
99 int ip_build_and_send_pkt(struct sk_buff *skb, struct sock *sk,
100 __be32 saddr, __be32 daddr,
101 struct ip_options_rcu *opt);
102 int ip_rcv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt,
103 struct net_device *orig_dev);
104 int ip_local_deliver(struct sk_buff *skb);
105 int ip_mr_input(struct sk_buff *skb);
106 int ip_output(struct sk_buff *skb);
107 int ip_mc_output(struct sk_buff *skb);
108 int ip_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *));
109 int ip_do_nat(struct sk_buff *skb);
110 void ip_send_check(struct iphdr *ip);
111 int __ip_local_out(struct sk_buff *skb);
112 int ip_local_out(struct sk_buff *skb);
113 int ip_queue_xmit(struct sk_buff *skb, struct flowi *fl);
115 int ip_append_data(struct sock *sk, struct flowi4 *fl4,
116 int getfrag(void *from, char *to, int offset, int len,
117 int odd, struct sk_buff *skb),
118 void *from, int len, int protolen,
119 struct ipcm_cookie *ipc,
122 int ip_generic_getfrag(void *from, char *to, int offset, int len, int odd,
123 struct sk_buff *skb);
124 ssize_t ip_append_page(struct sock *sk, struct flowi4 *fl4, struct page *page,
125 int offset, size_t size, int flags);
126 struct sk_buff *__ip_make_skb(struct sock *sk, struct flowi4 *fl4,
127 struct sk_buff_head *queue,
128 struct inet_cork *cork);
129 int ip_send_skb(struct net *net, struct sk_buff *skb);
130 int ip_push_pending_frames(struct sock *sk, struct flowi4 *fl4);
131 void ip_flush_pending_frames(struct sock *sk);
132 struct sk_buff *ip_make_skb(struct sock *sk, struct flowi4 *fl4,
133 int getfrag(void *from, char *to, int offset,
134 int len, int odd, struct sk_buff *skb),
135 void *from, int length, int transhdrlen,
136 struct ipcm_cookie *ipc, struct rtable **rtp,
139 static inline struct sk_buff *ip_finish_skb(struct sock *sk, struct flowi4 *fl4)
141 return __ip_make_skb(sk, fl4, &sk->sk_write_queue, &inet_sk(sk)->cork.base);
144 static inline __u8 get_rttos(struct ipcm_cookie* ipc, struct inet_sock *inet)
146 return (ipc->tos != -1) ? RT_TOS(ipc->tos) : RT_TOS(inet->tos);
149 static inline __u8 get_rtconn_flags(struct ipcm_cookie* ipc, struct sock* sk)
151 return (ipc->tos != -1) ? RT_CONN_FLAGS_TOS(sk, ipc->tos) : RT_CONN_FLAGS(sk);
155 int ip4_datagram_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len);
157 void ip4_datagram_release_cb(struct sock *sk);
159 struct ip_reply_arg {
163 int csumoffset; /* u16 offset of csum in iov[0].iov_base */
164 /* -1 if not needed */
169 #define IP_REPLY_ARG_NOSRCCHECK 1
171 static inline __u8 ip_reply_arg_flowi_flags(const struct ip_reply_arg *arg)
173 return (arg->flags & IP_REPLY_ARG_NOSRCCHECK) ? FLOWI_FLAG_ANYSRC : 0;
176 void ip_send_unicast_reply(struct net *net, struct sk_buff *skb, __be32 daddr,
177 __be32 saddr, const struct ip_reply_arg *arg,
185 extern struct ipv4_config ipv4_config;
186 #define IP_INC_STATS(net, field) SNMP_INC_STATS64((net)->mib.ip_statistics, field)
187 #define IP_INC_STATS_BH(net, field) SNMP_INC_STATS64_BH((net)->mib.ip_statistics, field)
188 #define IP_ADD_STATS(net, field, val) SNMP_ADD_STATS64((net)->mib.ip_statistics, field, val)
189 #define IP_ADD_STATS_BH(net, field, val) SNMP_ADD_STATS64_BH((net)->mib.ip_statistics, field, val)
190 #define IP_UPD_PO_STATS(net, field, val) SNMP_UPD_PO_STATS64((net)->mib.ip_statistics, field, val)
191 #define IP_UPD_PO_STATS_BH(net, field, val) SNMP_UPD_PO_STATS64_BH((net)->mib.ip_statistics, field, val)
192 #define NET_INC_STATS(net, field) SNMP_INC_STATS((net)->mib.net_statistics, field)
193 #define NET_INC_STATS_BH(net, field) SNMP_INC_STATS_BH((net)->mib.net_statistics, field)
194 #define NET_INC_STATS_USER(net, field) SNMP_INC_STATS_USER((net)->mib.net_statistics, field)
195 #define NET_ADD_STATS_BH(net, field, adnd) SNMP_ADD_STATS_BH((net)->mib.net_statistics, field, adnd)
196 #define NET_ADD_STATS_USER(net, field, adnd) SNMP_ADD_STATS_USER((net)->mib.net_statistics, field, adnd)
198 unsigned long snmp_fold_field(void __percpu *mib[], int offt);
199 #if BITS_PER_LONG==32
200 u64 snmp_fold_field64(void __percpu *mib[], int offt, size_t sync_off);
202 static inline u64 snmp_fold_field64(void __percpu *mib[], int offt, size_t syncp_off)
204 return snmp_fold_field(mib, offt);
207 int snmp_mib_init(void __percpu *ptr[2], size_t mibsize, size_t align);
209 static inline void snmp_mib_free(void __percpu *ptr[SNMP_ARRAY_SZ])
214 for (i = 0; i < SNMP_ARRAY_SZ; i++) {
220 void inet_get_local_port_range(struct net *net, int *low, int *high);
222 extern unsigned long *sysctl_local_reserved_ports;
223 static inline int inet_is_reserved_local_port(int port)
225 return test_bit(port, sysctl_local_reserved_ports);
228 extern int sysctl_ip_nonlocal_bind;
230 /* From inetpeer.c */
231 extern int inet_peer_threshold;
232 extern int inet_peer_minttl;
233 extern int inet_peer_maxttl;
235 /* From ip_input.c */
236 extern int sysctl_ip_early_demux;
238 /* From ip_output.c */
239 extern int sysctl_ip_dynaddr;
241 void ipfrag_init(void);
243 void ip_static_sysctl_init(void);
245 static inline bool ip_is_fragment(const struct iphdr *iph)
247 return (iph->frag_off & htons(IP_MF | IP_OFFSET)) != 0;
253 /* The function in 2.2 was invalid, producing wrong result for
254 * check=0xFEFF. It was noticed by Arthur Skawina _year_ ago. --ANK(000625) */
256 int ip_decrease_ttl(struct iphdr *iph)
258 u32 check = (__force u32)iph->check;
259 check += (__force u32)htons(0x0100);
260 iph->check = (__force __sum16)(check + (check>=0xFFFF));
265 int ip_dont_fragment(struct sock *sk, struct dst_entry *dst)
267 return inet_sk(sk)->pmtudisc == IP_PMTUDISC_DO ||
268 (inet_sk(sk)->pmtudisc == IP_PMTUDISC_WANT &&
269 !(dst_metric_locked(dst, RTAX_MTU)));
272 void __ip_select_ident(struct iphdr *iph, struct dst_entry *dst, int more);
274 static inline void ip_select_ident(struct sk_buff *skb, struct dst_entry *dst, struct sock *sk)
276 struct iphdr *iph = ip_hdr(skb);
278 if ((iph->frag_off & htons(IP_DF)) && !skb->local_df) {
279 /* This is only to work around buggy Windows95/2000
280 * VJ compression implementations. If the ID field
281 * does not change, they drop every other packet in
282 * a TCP stream using header compression.
284 iph->id = (sk && inet_sk(sk)->inet_daddr) ?
285 htons(inet_sk(sk)->inet_id++) : 0;
287 __ip_select_ident(iph, dst, 0);
290 static inline void ip_select_ident_more(struct sk_buff *skb, struct dst_entry *dst, struct sock *sk, int more)
292 struct iphdr *iph = ip_hdr(skb);
294 if ((iph->frag_off & htons(IP_DF)) && !skb->local_df) {
295 if (sk && inet_sk(sk)->inet_daddr) {
296 iph->id = htons(inet_sk(sk)->inet_id);
297 inet_sk(sk)->inet_id += 1 + more;
301 __ip_select_ident(iph, dst, more);
305 * Map a multicast IP onto multicast MAC for type ethernet.
308 static inline void ip_eth_mc_map(__be32 naddr, char *buf)
310 __u32 addr=ntohl(naddr);
322 * Map a multicast IP onto multicast MAC for type IP-over-InfiniBand.
323 * Leave P_Key as 0 to be filled in by driver.
326 static inline void ip_ib_mc_map(__be32 naddr, const unsigned char *broadcast, char *buf)
329 unsigned char scope = broadcast[5] & 0xF;
331 buf[0] = 0; /* Reserved */
332 buf[1] = 0xff; /* Multicast QPN */
337 buf[5] = 0x10 | scope; /* scope from broadcast address */
338 buf[6] = 0x40; /* IPv4 signature */
340 buf[8] = broadcast[8]; /* P_Key */
341 buf[9] = broadcast[9];
348 buf[19] = addr & 0xff;
350 buf[18] = addr & 0xff;
352 buf[17] = addr & 0xff;
354 buf[16] = addr & 0x0f;
357 static inline void ip_ipgre_mc_map(__be32 naddr, const unsigned char *broadcast, char *buf)
359 if ((broadcast[0] | broadcast[1] | broadcast[2] | broadcast[3]) != 0)
360 memcpy(buf, broadcast, 4);
362 memcpy(buf, &naddr, sizeof(naddr));
365 #if IS_ENABLED(CONFIG_IPV6)
366 #include <linux/ipv6.h>
369 static __inline__ void inet_reset_saddr(struct sock *sk)
371 inet_sk(sk)->inet_rcv_saddr = inet_sk(sk)->inet_saddr = 0;
372 #if IS_ENABLED(CONFIG_IPV6)
373 if (sk->sk_family == PF_INET6) {
374 struct ipv6_pinfo *np = inet6_sk(sk);
376 memset(&np->saddr, 0, sizeof(np->saddr));
377 memset(&sk->sk_v6_rcv_saddr, 0, sizeof(sk->sk_v6_rcv_saddr));
384 static inline int sk_mc_loop(struct sock *sk)
388 switch (sk->sk_family) {
390 return inet_sk(sk)->mc_loop;
391 #if IS_ENABLED(CONFIG_IPV6)
393 return inet6_sk(sk)->mc_loop;
400 bool ip_call_ra_chain(struct sk_buff *skb);
403 * Functions provided by ip_fragment.c
406 enum ip_defrag_users {
407 IP_DEFRAG_LOCAL_DELIVER,
408 IP_DEFRAG_CALL_RA_CHAIN,
409 IP_DEFRAG_CONNTRACK_IN,
410 __IP_DEFRAG_CONNTRACK_IN_END = IP_DEFRAG_CONNTRACK_IN + USHRT_MAX,
411 IP_DEFRAG_CONNTRACK_OUT,
412 __IP_DEFRAG_CONNTRACK_OUT_END = IP_DEFRAG_CONNTRACK_OUT + USHRT_MAX,
413 IP_DEFRAG_CONNTRACK_BRIDGE_IN,
414 __IP_DEFRAG_CONNTRACK_BRIDGE_IN = IP_DEFRAG_CONNTRACK_BRIDGE_IN + USHRT_MAX,
422 int ip_defrag(struct sk_buff *skb, u32 user);
424 struct sk_buff *ip_check_defrag(struct sk_buff *skb, u32 user);
426 static inline struct sk_buff *ip_check_defrag(struct sk_buff *skb, u32 user)
431 int ip_frag_mem(struct net *net);
432 int ip_frag_nqueues(struct net *net);
435 * Functions provided by ip_forward.c
438 int ip_forward(struct sk_buff *skb);
441 * Functions provided by ip_options.c
444 void ip_options_build(struct sk_buff *skb, struct ip_options *opt,
445 __be32 daddr, struct rtable *rt, int is_frag);
446 int ip_options_echo(struct ip_options *dopt, struct sk_buff *skb);
447 void ip_options_fragment(struct sk_buff *skb);
448 int ip_options_compile(struct net *net, struct ip_options *opt,
449 struct sk_buff *skb);
450 int ip_options_get(struct net *net, struct ip_options_rcu **optp,
451 unsigned char *data, int optlen);
452 int ip_options_get_from_user(struct net *net, struct ip_options_rcu **optp,
453 unsigned char __user *data, int optlen);
454 void ip_options_undo(struct ip_options *opt);
455 void ip_forward_options(struct sk_buff *skb);
456 int ip_options_rcv_srr(struct sk_buff *skb);
459 * Functions provided by ip_sockglue.c
462 void ipv4_pktinfo_prepare(const struct sock *sk, struct sk_buff *skb);
463 void ip_cmsg_recv(struct msghdr *msg, struct sk_buff *skb);
464 int ip_cmsg_send(struct net *net, struct msghdr *msg, struct ipcm_cookie *ipc);
465 int ip_setsockopt(struct sock *sk, int level, int optname, char __user *optval,
466 unsigned int optlen);
467 int ip_getsockopt(struct sock *sk, int level, int optname, char __user *optval,
469 int compat_ip_setsockopt(struct sock *sk, int level, int optname,
470 char __user *optval, unsigned int optlen);
471 int compat_ip_getsockopt(struct sock *sk, int level, int optname,
472 char __user *optval, int __user *optlen);
473 int ip_ra_control(struct sock *sk, unsigned char on,
474 void (*destructor)(struct sock *));
476 int ip_recv_error(struct sock *sk, struct msghdr *msg, int len, int *addr_len);
477 void ip_icmp_error(struct sock *sk, struct sk_buff *skb, int err, __be16 port,
478 u32 info, u8 *payload);
479 void ip_local_error(struct sock *sk, int err, __be32 daddr, __be16 dport,
482 #ifdef CONFIG_PROC_FS
483 int ip_misc_proc_init(void);