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
11 * Fred N. van Kempen, <waltje@uWalt.NL.Mugnet.ORG>
12 * Alan Cox, <gw4pts@gw4pts.ampr.org>
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>
36 struct inet_skb_parm {
37 struct ip_options opt; /* Compiled IP options */
40 #define IPSKB_FORWARDED BIT(0)
41 #define IPSKB_XFRM_TUNNEL_SIZE BIT(1)
42 #define IPSKB_XFRM_TRANSFORMED BIT(2)
43 #define IPSKB_FRAG_COMPLETE BIT(3)
44 #define IPSKB_REROUTED BIT(4)
45 #define IPSKB_DOREDIRECT BIT(5)
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;
62 #define IPCB(skb) ((struct inet_skb_parm*)((skb)->cb))
65 struct ip_ra_chain __rcu *next;
68 void (*destructor)(struct sock *);
69 struct sock *saved_sk;
74 extern struct ip_ra_chain __rcu *ip_ra_chain;
77 #define IP_CE 0x8000 /* Flag: "Congestion" */
78 #define IP_DF 0x4000 /* Flag: "Don't Fragment" */
79 #define IP_MF 0x2000 /* Flag: "More Fragments" */
80 #define IP_OFFSET 0x1FFF /* "Fragment Offset" part */
82 #define IP_FRAG_TIME (30 * HZ) /* fragment lifetime */
90 extern int igmp_mc_proc_init(void);
93 * Functions provided by ip.c
96 extern int ip_build_and_send_pkt(struct sk_buff *skb, struct sock *sk,
97 __be32 saddr, __be32 daddr,
98 struct ip_options_rcu *opt);
99 extern int ip_rcv(struct sk_buff *skb, struct net_device *dev,
100 struct packet_type *pt, struct net_device *orig_dev);
101 extern int ip_local_deliver(struct sk_buff *skb);
102 extern int ip_mr_input(struct sk_buff *skb);
103 extern int ip_output(struct sk_buff *skb);
104 extern int ip_mc_output(struct sk_buff *skb);
105 extern int ip_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *));
106 extern int ip_do_nat(struct sk_buff *skb);
107 extern void ip_send_check(struct iphdr *ip);
108 extern int __ip_local_out(struct sk_buff *skb);
109 extern int ip_local_out(struct sk_buff *skb);
110 extern int ip_queue_xmit(struct sk_buff *skb, struct flowi *fl);
111 extern void ip_init(void);
112 extern int ip_append_data(struct sock *sk, 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 len, int protolen,
116 struct ipcm_cookie *ipc,
119 extern int ip_generic_getfrag(void *from, char *to, int offset, int len, int odd, struct sk_buff *skb);
120 extern ssize_t ip_append_page(struct sock *sk, struct flowi4 *fl4, struct page *page,
121 int offset, size_t size, int flags);
122 extern struct sk_buff *__ip_make_skb(struct sock *sk,
124 struct sk_buff_head *queue,
125 struct inet_cork *cork);
126 extern int ip_send_skb(struct net *net, struct sk_buff *skb);
127 extern int ip_push_pending_frames(struct sock *sk, struct flowi4 *fl4);
128 extern void ip_flush_pending_frames(struct sock *sk);
129 extern struct sk_buff *ip_make_skb(struct sock *sk,
131 int getfrag(void *from, char *to, int offset, int len,
132 int odd, struct sk_buff *skb),
133 void *from, int length, int transhdrlen,
134 struct ipcm_cookie *ipc,
138 static inline struct sk_buff *ip_finish_skb(struct sock *sk, struct flowi4 *fl4)
140 return __ip_make_skb(sk, fl4, &sk->sk_write_queue, &inet_sk(sk)->cork.base);
144 extern int ip4_datagram_connect(struct sock *sk,
145 struct sockaddr *uaddr, int addr_len);
147 extern void ip4_datagram_release_cb(struct sock *sk);
149 struct ip_reply_arg {
153 int csumoffset; /* u16 offset of csum in iov[0].iov_base */
154 /* -1 if not needed */
159 #define IP_REPLY_ARG_NOSRCCHECK 1
161 static inline __u8 ip_reply_arg_flowi_flags(const struct ip_reply_arg *arg)
163 return (arg->flags & IP_REPLY_ARG_NOSRCCHECK) ? FLOWI_FLAG_ANYSRC : 0;
166 void ip_send_unicast_reply(struct net *net, struct sk_buff *skb, __be32 daddr,
167 __be32 saddr, const struct ip_reply_arg *arg,
175 extern struct ipv4_config ipv4_config;
176 #define IP_INC_STATS(net, field) SNMP_INC_STATS64((net)->mib.ip_statistics, field)
177 #define IP_INC_STATS_BH(net, field) SNMP_INC_STATS64_BH((net)->mib.ip_statistics, field)
178 #define IP_ADD_STATS(net, field, val) SNMP_ADD_STATS64((net)->mib.ip_statistics, field, val)
179 #define IP_ADD_STATS_BH(net, field, val) SNMP_ADD_STATS64_BH((net)->mib.ip_statistics, field, val)
180 #define IP_UPD_PO_STATS(net, field, val) SNMP_UPD_PO_STATS64((net)->mib.ip_statistics, field, val)
181 #define IP_UPD_PO_STATS_BH(net, field, val) SNMP_UPD_PO_STATS64_BH((net)->mib.ip_statistics, field, val)
182 #define NET_INC_STATS(net, field) SNMP_INC_STATS((net)->mib.net_statistics, field)
183 #define NET_INC_STATS_BH(net, field) SNMP_INC_STATS_BH((net)->mib.net_statistics, field)
184 #define NET_INC_STATS_USER(net, field) SNMP_INC_STATS_USER((net)->mib.net_statistics, field)
185 #define NET_ADD_STATS_BH(net, field, adnd) SNMP_ADD_STATS_BH((net)->mib.net_statistics, field, adnd)
186 #define NET_ADD_STATS_USER(net, field, adnd) SNMP_ADD_STATS_USER((net)->mib.net_statistics, field, adnd)
188 extern unsigned long snmp_fold_field(void __percpu *mib[], int offt);
189 #if BITS_PER_LONG==32
190 extern u64 snmp_fold_field64(void __percpu *mib[], int offt, size_t sync_off);
192 static inline u64 snmp_fold_field64(void __percpu *mib[], int offt, size_t syncp_off)
194 return snmp_fold_field(mib, offt);
197 extern int snmp_mib_init(void __percpu *ptr[2], size_t mibsize, size_t align);
198 extern void snmp_mib_free(void __percpu *ptr[2]);
200 extern struct local_ports {
203 } sysctl_local_ports;
204 extern void inet_get_local_port_range(int *low, int *high);
206 extern unsigned long *sysctl_local_reserved_ports;
207 static inline int inet_is_reserved_local_port(int port)
209 return test_bit(port, sysctl_local_reserved_ports);
212 extern int sysctl_ip_nonlocal_bind;
214 /* From inetpeer.c */
215 extern int inet_peer_threshold;
216 extern int inet_peer_minttl;
217 extern int inet_peer_maxttl;
219 /* From ip_input.c */
220 extern int sysctl_ip_early_demux;
222 /* From ip_output.c */
223 extern int sysctl_ip_dynaddr;
225 extern void ipfrag_init(void);
227 extern void ip_static_sysctl_init(void);
229 static inline bool ip_is_fragment(const struct iphdr *iph)
231 return (iph->frag_off & htons(IP_MF | IP_OFFSET)) != 0;
237 /* The function in 2.2 was invalid, producing wrong result for
238 * check=0xFEFF. It was noticed by Arthur Skawina _year_ ago. --ANK(000625) */
240 int ip_decrease_ttl(struct iphdr *iph)
242 u32 check = (__force u32)iph->check;
243 check += (__force u32)htons(0x0100);
244 iph->check = (__force __sum16)(check + (check>=0xFFFF));
249 int ip_dont_fragment(struct sock *sk, struct dst_entry *dst)
251 return inet_sk(sk)->pmtudisc == IP_PMTUDISC_DO ||
252 (inet_sk(sk)->pmtudisc == IP_PMTUDISC_WANT &&
253 !(dst_metric_locked(dst, RTAX_MTU)));
256 u32 ip_idents_reserve(u32 hash, int segs);
257 void __ip_select_ident(struct iphdr *iph, int segs);
259 static inline void ip_select_ident_segs(struct sk_buff *skb, struct sock *sk, int segs)
261 struct iphdr *iph = ip_hdr(skb);
263 if ((iph->frag_off & htons(IP_DF)) && !skb->local_df) {
264 /* This is only to work around buggy Windows95/2000
265 * VJ compression implementations. If the ID field
266 * does not change, they drop every other packet in
267 * a TCP stream using header compression.
269 if (sk && inet_sk(sk)->inet_daddr) {
270 iph->id = htons(inet_sk(sk)->inet_id);
271 inet_sk(sk)->inet_id += segs;
276 __ip_select_ident(iph, segs);
280 static inline void ip_select_ident(struct sk_buff *skb, struct sock *sk)
282 ip_select_ident_segs(skb, sk, 1);
286 * Map a multicast IP onto multicast MAC for type ethernet.
289 static inline void ip_eth_mc_map(__be32 naddr, char *buf)
291 __u32 addr=ntohl(naddr);
303 * Map a multicast IP onto multicast MAC for type IP-over-InfiniBand.
304 * Leave P_Key as 0 to be filled in by driver.
307 static inline void ip_ib_mc_map(__be32 naddr, const unsigned char *broadcast, char *buf)
310 unsigned char scope = broadcast[5] & 0xF;
312 buf[0] = 0; /* Reserved */
313 buf[1] = 0xff; /* Multicast QPN */
318 buf[5] = 0x10 | scope; /* scope from broadcast address */
319 buf[6] = 0x40; /* IPv4 signature */
321 buf[8] = broadcast[8]; /* P_Key */
322 buf[9] = broadcast[9];
329 buf[19] = addr & 0xff;
331 buf[18] = addr & 0xff;
333 buf[17] = addr & 0xff;
335 buf[16] = addr & 0x0f;
338 static inline void ip_ipgre_mc_map(__be32 naddr, const unsigned char *broadcast, char *buf)
340 if ((broadcast[0] | broadcast[1] | broadcast[2] | broadcast[3]) != 0)
341 memcpy(buf, broadcast, 4);
343 memcpy(buf, &naddr, sizeof(naddr));
346 #if IS_ENABLED(CONFIG_IPV6)
347 #include <linux/ipv6.h>
350 static __inline__ void inet_reset_saddr(struct sock *sk)
352 inet_sk(sk)->inet_rcv_saddr = inet_sk(sk)->inet_saddr = 0;
353 #if IS_ENABLED(CONFIG_IPV6)
354 if (sk->sk_family == PF_INET6) {
355 struct ipv6_pinfo *np = inet6_sk(sk);
357 memset(&np->saddr, 0, sizeof(np->saddr));
358 memset(&np->rcv_saddr, 0, sizeof(np->rcv_saddr));
365 static inline int sk_mc_loop(struct sock *sk)
369 switch (sk->sk_family) {
371 return inet_sk(sk)->mc_loop;
372 #if IS_ENABLED(CONFIG_IPV6)
374 return inet6_sk(sk)->mc_loop;
381 extern bool ip_call_ra_chain(struct sk_buff *skb);
384 * Functions provided by ip_fragment.c
387 enum ip_defrag_users {
388 IP_DEFRAG_LOCAL_DELIVER,
389 IP_DEFRAG_CALL_RA_CHAIN,
390 IP_DEFRAG_CONNTRACK_IN,
391 __IP_DEFRAG_CONNTRACK_IN_END = IP_DEFRAG_CONNTRACK_IN + USHRT_MAX,
392 IP_DEFRAG_CONNTRACK_OUT,
393 __IP_DEFRAG_CONNTRACK_OUT_END = IP_DEFRAG_CONNTRACK_OUT + USHRT_MAX,
394 IP_DEFRAG_CONNTRACK_BRIDGE_IN,
395 __IP_DEFRAG_CONNTRACK_BRIDGE_IN = IP_DEFRAG_CONNTRACK_BRIDGE_IN + USHRT_MAX,
403 int ip_defrag(struct sk_buff *skb, u32 user);
405 struct sk_buff *ip_check_defrag(struct sk_buff *skb, u32 user);
407 static inline struct sk_buff *ip_check_defrag(struct sk_buff *skb, u32 user)
412 int ip_frag_mem(struct net *net);
413 int ip_frag_nqueues(struct net *net);
416 * Functions provided by ip_forward.c
419 extern int ip_forward(struct sk_buff *skb);
422 * Functions provided by ip_options.c
425 extern void ip_options_build(struct sk_buff *skb, struct ip_options *opt,
426 __be32 daddr, struct rtable *rt, int is_frag);
427 extern int ip_options_echo(struct ip_options *dopt, struct sk_buff *skb);
428 extern void ip_options_fragment(struct sk_buff *skb);
429 extern int ip_options_compile(struct net *net,
430 struct ip_options *opt, struct sk_buff *skb);
431 extern int ip_options_get(struct net *net, struct ip_options_rcu **optp,
432 unsigned char *data, int optlen);
433 extern int ip_options_get_from_user(struct net *net, struct ip_options_rcu **optp,
434 unsigned char __user *data, int optlen);
435 extern void ip_options_undo(struct ip_options * opt);
436 extern void ip_forward_options(struct sk_buff *skb);
437 extern int ip_options_rcv_srr(struct sk_buff *skb);
440 * Functions provided by ip_sockglue.c
443 extern void ipv4_pktinfo_prepare(struct sk_buff *skb);
444 extern void ip_cmsg_recv(struct msghdr *msg, struct sk_buff *skb);
445 extern int ip_cmsg_send(struct net *net,
446 struct msghdr *msg, struct ipcm_cookie *ipc);
447 extern int ip_setsockopt(struct sock *sk, int level, int optname, char __user *optval, unsigned int optlen);
448 extern int ip_getsockopt(struct sock *sk, int level, int optname, char __user *optval, int __user *optlen);
449 extern int compat_ip_setsockopt(struct sock *sk, int level,
450 int optname, char __user *optval, unsigned int optlen);
451 extern int compat_ip_getsockopt(struct sock *sk, int level,
452 int optname, char __user *optval, int __user *optlen);
453 extern int ip_ra_control(struct sock *sk, unsigned char on, void (*destructor)(struct sock *));
455 extern int ip_recv_error(struct sock *sk, struct msghdr *msg, int len, int *addr_len);
456 extern void ip_icmp_error(struct sock *sk, struct sk_buff *skb, int err,
457 __be16 port, u32 info, u8 *payload);
458 extern void ip_local_error(struct sock *sk, int err, __be32 daddr, __be16 dport,
461 #ifdef CONFIG_PROC_FS
462 extern int ip_misc_proc_init(void);