include/net/ip.h

   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
  25 #include <linux/types.h>
  26 #include <linux/ip.h>
  27 #include <linux/in.h>
  28 #include <linux/skbuff.h>
  29
  30 #include <net/inet_sock.h>
  31 #include <net/snmp.h>
  32 #include <net/flow.h>
  33
  34 struct sock;
  35
  36 struct inet_skb_parm {
  37         struct ip_options       opt;            /* Compiled IP options          */
  38         unsigned char           flags;
  39
  40 #define IPSKB_FORWARDED         1
  41 #define IPSKB_XFRM_TUNNEL_SIZE  2
  42 #define IPSKB_XFRM_TRANSFORMED  4
  43 #define IPSKB_FRAG_COMPLETE     8
  44 #define IPSKB_REROUTED          16
  45 };
  46
  47 static inline unsigned int ip_hdrlen(const struct sk_buff *skb)
  48 {
  49         return ip_hdr(skb)->ihl * 4;
  50 }
  51
  52 struct ipcm_cookie {
  53         __be32                  addr;
  54         int                     oif;
  55         struct ip_options_rcu   *opt;
  56         __u8                    tx_flags;
  57 };
  58
  59 #define IPCB(skb) ((struct inet_skb_parm*)((skb)->cb))
  60
  61 struct ip_ra_chain {
  62         struct ip_ra_chain __rcu *next;
  63         struct sock             *sk;
  64         union {
  65                 void                    (*destructor)(struct sock *);
  66                 struct sock             *saved_sk;
  67         };
  68         struct rcu_head         rcu;
  69 };
  70
  71 extern struct ip_ra_chain __rcu *ip_ra_chain;
  72
  73 /* IP flags. */
  74 #define IP_CE           0x8000          /* Flag: "Congestion"           */
  75 #define IP_DF           0x4000          /* Flag: "Don't Fragment"       */
  76 #define IP_MF           0x2000          /* Flag: "More Fragments"       */
  77 #define IP_OFFSET       0x1FFF          /* "Fragment Offset" part       */
  78
  79 #define IP_FRAG_TIME    (30 * HZ)               /* fragment lifetime    */
  80
  81 struct msghdr;
  82 struct net_device;
  83 struct packet_type;
  84 struct rtable;
  85 struct sockaddr;
  86
  87 extern int              igmp_mc_proc_init(void);
  88
  89 /*
  90  *      Functions provided by ip.c
  91  */
  92
  93 extern int              ip_build_and_send_pkt(struct sk_buff *skb, struct sock *sk,
  94                                               __be32 saddr, __be32 daddr,
  95                                               struct ip_options_rcu *opt);
  96 extern int              ip_rcv(struct sk_buff *skb, struct net_device *dev,
  97                                struct packet_type *pt, struct net_device *orig_dev);
  98 extern int              ip_local_deliver(struct sk_buff *skb);
  99 extern int              ip_mr_input(struct sk_buff *skb);
 100 extern int              ip_output(struct sk_buff *skb);
 101 extern int              ip_mc_output(struct sk_buff *skb);
 102 extern int              ip_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *));
 103 extern int              ip_do_nat(struct sk_buff *skb);
 104 extern void             ip_send_check(struct iphdr *ip);
 105 extern int              __ip_local_out(struct sk_buff *skb);
 106 extern int              ip_local_out(struct sk_buff *skb);
 107 extern int              ip_queue_xmit(struct sk_buff *skb, struct flowi *fl);
 108 extern void             ip_init(void);
 109 extern int              ip_append_data(struct sock *sk, struct flowi4 *fl4,
 110                                        int getfrag(void *from, char *to, int offset, int len,
 111                                                    int odd, struct sk_buff *skb),
 112                                 void *from, int len, int protolen,
 113                                 struct ipcm_cookie *ipc,
 114                                 struct rtable **rt,
 115                                 unsigned int flags);
 116 extern int              ip_generic_getfrag(void *from, char *to, int offset, int len, int odd, struct sk_buff *skb);
 117 extern ssize_t          ip_append_page(struct sock *sk, struct flowi4 *fl4, struct page *page,
 118                                 int offset, size_t size, int flags);
 119 extern struct sk_buff  *__ip_make_skb(struct sock *sk,
 120                                       struct flowi4 *fl4,
 121                                       struct sk_buff_head *queue,
 122                                       struct inet_cork *cork);
 123 extern int              ip_send_skb(struct sk_buff *skb);
 124 extern int              ip_push_pending_frames(struct sock *sk, struct flowi4 *fl4);
 125 extern void             ip_flush_pending_frames(struct sock *sk);
 126 extern struct sk_buff  *ip_make_skb(struct sock *sk,
 127                                     struct flowi4 *fl4,
 128                                     int getfrag(void *from, char *to, int offset, int len,
 129                                                 int odd, struct sk_buff *skb),
 130                                     void *from, int length, int transhdrlen,
 131                                     struct ipcm_cookie *ipc,
 132                                     struct rtable **rtp,
 133                                     unsigned int flags);
 134
 135 static inline struct sk_buff *ip_finish_skb(struct sock *sk, struct flowi4 *fl4)
 136 {
 137         return __ip_make_skb(sk, fl4, &sk->sk_write_queue, &inet_sk(sk)->cork.base);
 138 }
 139
 140 /* datagram.c */
 141 extern int              ip4_datagram_connect(struct sock *sk,
 142                                              struct sockaddr *uaddr, int addr_len);
 143
 144 /*
 145  *      Map a multicast IP onto multicast MAC for type Token Ring.
 146  *      This conforms to RFC1469 Option 2 Multicasting i.e.
 147  *      using a functional address to transmit / receive
 148  *      multicast packets.
 149  */
 150
 151 static inline void ip_tr_mc_map(__be32 addr, char *buf)
 152 {
 153         buf[0]=0xC0;
 154         buf[1]=0x00;
 155         buf[2]=0x00;
 156         buf[3]=0x04;
 157         buf[4]=0x00;
 158         buf[5]=0x00;
 159 }
 160
 161 struct ip_reply_arg {
 162         struct kvec iov[1];
 163         int         flags;
 164         __wsum      csum;
 165         int         csumoffset; /* u16 offset of csum in iov[0].iov_base */
 166                                 /* -1 if not needed */
 167         int         bound_dev_if;
 168         u8          tos;
 169 };
 170
 171 #define IP_REPLY_ARG_NOSRCCHECK 1
 172
 173 static inline __u8 ip_reply_arg_flowi_flags(const struct ip_reply_arg *arg)
 174 {
 175         return (arg->flags & IP_REPLY_ARG_NOSRCCHECK) ? FLOWI_FLAG_ANYSRC : 0;
 176 }
 177
 178 void ip_send_reply(struct sock *sk, struct sk_buff *skb, __be32 daddr,
 179                    const struct ip_reply_arg *arg, unsigned int len);
 180
 181 struct ipv4_config {
 182         int     log_martians;
 183         int     no_pmtu_disc;
 184 };
 185
 186 extern struct ipv4_config ipv4_config;
 187 #define IP_INC_STATS(net, field)        SNMP_INC_STATS64((net)->mib.ip_statistics, field)
 188 #define IP_INC_STATS_BH(net, field)     SNMP_INC_STATS64_BH((net)->mib.ip_statistics, field)
 189 #define IP_ADD_STATS(net, field, val)   SNMP_ADD_STATS64((net)->mib.ip_statistics, field, val)
 190 #define IP_ADD_STATS_BH(net, field, val) SNMP_ADD_STATS64_BH((net)->mib.ip_statistics, field, val)
 191 #define IP_UPD_PO_STATS(net, field, val) SNMP_UPD_PO_STATS64((net)->mib.ip_statistics, field, val)
 192 #define IP_UPD_PO_STATS_BH(net, field, val) SNMP_UPD_PO_STATS64_BH((net)->mib.ip_statistics, field, val)
 193 #define NET_INC_STATS(net, field)       SNMP_INC_STATS((net)->mib.net_statistics, field)
 194 #define NET_INC_STATS_BH(net, field)    SNMP_INC_STATS_BH((net)->mib.net_statistics, field)
 195 #define NET_INC_STATS_USER(net, field)  SNMP_INC_STATS_USER((net)->mib.net_statistics, field)
 196 #define NET_ADD_STATS_BH(net, field, adnd) SNMP_ADD_STATS_BH((net)->mib.net_statistics, field, adnd)
 197 #define NET_ADD_STATS_USER(net, field, adnd) SNMP_ADD_STATS_USER((net)->mib.net_statistics, field, adnd)
 198
 199 extern unsigned long snmp_fold_field(void __percpu *mib[], int offt);
 200 #if BITS_PER_LONG==32
 201 extern u64 snmp_fold_field64(void __percpu *mib[], int offt, size_t sync_off);
 202 #else
 203 static inline u64 snmp_fold_field64(void __percpu *mib[], int offt, size_t syncp_off)
 204 {
 205         return snmp_fold_field(mib, offt);
 206 }
 207 #endif
 208 extern int snmp_mib_init(void __percpu *ptr[2], size_t mibsize, size_t align);
 209 extern void snmp_mib_free(void __percpu *ptr[2]);
 210
 211 extern struct local_ports {
 212         seqlock_t       lock;
 213         int             range[2];
 214 } sysctl_local_ports;
 215 extern void inet_get_local_port_range(int *low, int *high);
 216
 217 extern unsigned long *sysctl_local_reserved_ports;
 218 static inline int inet_is_reserved_local_port(int port)
 219 {
 220         return test_bit(port, sysctl_local_reserved_ports);
 221 }
 222
 223 extern int sysctl_ip_nonlocal_bind;
 224
 225 extern struct ctl_path net_core_path[];
 226 extern struct ctl_path net_ipv4_ctl_path[];
 227
 228 /* From inetpeer.c */
 229 extern int inet_peer_threshold;
 230 extern int inet_peer_minttl;
 231 extern int inet_peer_maxttl;
 232
 233 /* From ip_output.c */
 234 extern int sysctl_ip_dynaddr;
 235
 236 extern void ipfrag_init(void);
 237
 238 extern void ip_static_sysctl_init(void);
 239
 240 static inline bool ip_is_fragment(const struct iphdr *iph)
 241 {
 242         return (iph->frag_off & htons(IP_MF | IP_OFFSET)) != 0;
 243 }
 244
 245 #ifdef CONFIG_INET
 246 #include <net/dst.h>
 247
 248 /* The function in 2.2 was invalid, producing wrong result for
 249  * check=0xFEFF. It was noticed by Arthur Skawina _year_ ago. --ANK(000625) */
 250 static inline
 251 int ip_decrease_ttl(struct iphdr *iph)
 252 {
 253         u32 check = (__force u32)iph->check;
 254         check += (__force u32)htons(0x0100);
 255         iph->check = (__force __sum16)(check + (check>=0xFFFF));
 256         return --iph->ttl;
 257 }
 258
 259 static inline
 260 int ip_dont_fragment(struct sock *sk, struct dst_entry *dst)
 261 {
 262         return  inet_sk(sk)->pmtudisc == IP_PMTUDISC_DO ||
 263                 (inet_sk(sk)->pmtudisc == IP_PMTUDISC_WANT &&
 264                  !(dst_metric_locked(dst, RTAX_MTU)));
 265 }
 266
 267 extern void __ip_select_ident(struct iphdr *iph, struct dst_entry *dst, int more);
 268
 269 static inline void ip_select_ident(struct iphdr *iph, struct dst_entry *dst, struct sock *sk)
 270 {
 271         if (iph->frag_off & htons(IP_DF)) {
 272                 /* This is only to work around buggy Windows95/2000
 273                  * VJ compression implementations.  If the ID field
 274                  * does not change, they drop every other packet in
 275                  * a TCP stream using header compression.
 276                  */
 277                 iph->id = (sk && inet_sk(sk)->inet_daddr) ?
 278                                         htons(inet_sk(sk)->inet_id++) : 0;
 279         } else
 280                 __ip_select_ident(iph, dst, 0);
 281 }
 282
 283 static inline void ip_select_ident_more(struct iphdr *iph, struct dst_entry *dst, struct sock *sk, int more)
 284 {
 285         if (iph->frag_off & htons(IP_DF)) {
 286                 if (sk && inet_sk(sk)->inet_daddr) {
 287                         iph->id = htons(inet_sk(sk)->inet_id);
 288                         inet_sk(sk)->inet_id += 1 + more;
 289                 } else
 290                         iph->id = 0;
 291         } else
 292                 __ip_select_ident(iph, dst, more);
 293 }
 294
 295 /*
 296  *      Map a multicast IP onto multicast MAC for type ethernet.
 297  */
 298
 299 static inline void ip_eth_mc_map(__be32 naddr, char *buf)
 300 {
 301         __u32 addr=ntohl(naddr);
 302         buf[0]=0x01;
 303         buf[1]=0x00;
 304         buf[2]=0x5e;
 305         buf[5]=addr&0xFF;
 306         addr>>=8;
 307         buf[4]=addr&0xFF;
 308         addr>>=8;
 309         buf[3]=addr&0x7F;
 310 }
 311
 312 /*
 313  *      Map a multicast IP onto multicast MAC for type IP-over-InfiniBand.
 314  *      Leave P_Key as 0 to be filled in by driver.
 315  */
 316
 317 static inline void ip_ib_mc_map(__be32 naddr, const unsigned char *broadcast, char *buf)
 318 {
 319         __u32 addr;
 320         unsigned char scope = broadcast[5] & 0xF;
 321
 322         buf[0]  = 0;            /* Reserved */
 323         buf[1]  = 0xff;         /* Multicast QPN */
 324         buf[2]  = 0xff;
 325         buf[3]  = 0xff;
 326         addr    = ntohl(naddr);
 327         buf[4]  = 0xff;
 328         buf[5]  = 0x10 | scope; /* scope from broadcast address */
 329         buf[6]  = 0x40;         /* IPv4 signature */
 330         buf[7]  = 0x1b;
 331         buf[8]  = broadcast[8];         /* P_Key */
 332         buf[9]  = broadcast[9];
 333         buf[10] = 0;
 334         buf[11] = 0;
 335         buf[12] = 0;
 336         buf[13] = 0;
 337         buf[14] = 0;
 338         buf[15] = 0;
 339         buf[19] = addr & 0xff;
 340         addr  >>= 8;
 341         buf[18] = addr & 0xff;
 342         addr  >>= 8;
 343         buf[17] = addr & 0xff;
 344         addr  >>= 8;
 345         buf[16] = addr & 0x0f;
 346 }
 347
 348 static inline void ip_ipgre_mc_map(__be32 naddr, const unsigned char *broadcast, char *buf)
 349 {
 350         if ((broadcast[0] | broadcast[1] | broadcast[2] | broadcast[3]) != 0)
 351                 memcpy(buf, broadcast, 4);
 352         else
 353                 memcpy(buf, &naddr, sizeof(naddr));
 354 }
 355
 356 #if IS_ENABLED(CONFIG_IPV6)
 357 #include <linux/ipv6.h>
 358 #endif
 359
 360 static __inline__ void inet_reset_saddr(struct sock *sk)
 361 {
 362         inet_sk(sk)->inet_rcv_saddr = inet_sk(sk)->inet_saddr = 0;
 363 #if IS_ENABLED(CONFIG_IPV6)
 364         if (sk->sk_family == PF_INET6) {
 365                 struct ipv6_pinfo *np = inet6_sk(sk);
 366
 367                 memset(&np->saddr, 0, sizeof(np->saddr));
 368                 memset(&np->rcv_saddr, 0, sizeof(np->rcv_saddr));
 369         }
 370 #endif
 371 }
 372
 373 #endif
 374
 375 static inline int sk_mc_loop(struct sock *sk)
 376 {
 377         if (!sk)
 378                 return 1;
 379         switch (sk->sk_family) {
 380         case AF_INET:
 381                 return inet_sk(sk)->mc_loop;
 382 #if IS_ENABLED(CONFIG_IPV6)
 383         case AF_INET6:
 384                 return inet6_sk(sk)->mc_loop;
 385 #endif
 386         }
 387         WARN_ON(1);
 388         return 1;
 389 }
 390
 391 extern int      ip_call_ra_chain(struct sk_buff *skb);
 392
 393 /*
 394  *      Functions provided by ip_fragment.c
 395  */
 396
 397 enum ip_defrag_users {
 398         IP_DEFRAG_LOCAL_DELIVER,
 399         IP_DEFRAG_CALL_RA_CHAIN,
 400         IP_DEFRAG_CONNTRACK_IN,
 401         __IP_DEFRAG_CONNTRACK_IN_END    = IP_DEFRAG_CONNTRACK_IN + USHRT_MAX,
 402         IP_DEFRAG_CONNTRACK_OUT,
 403         __IP_DEFRAG_CONNTRACK_OUT_END   = IP_DEFRAG_CONNTRACK_OUT + USHRT_MAX,
 404         IP_DEFRAG_CONNTRACK_BRIDGE_IN,
 405         __IP_DEFRAG_CONNTRACK_BRIDGE_IN = IP_DEFRAG_CONNTRACK_BRIDGE_IN + USHRT_MAX,
 406         IP_DEFRAG_VS_IN,
 407         IP_DEFRAG_VS_OUT,
 408         IP_DEFRAG_VS_FWD,
 409         IP_DEFRAG_AF_PACKET,
 410         IP_DEFRAG_MACVLAN,
 411 };
 412
 413 int ip_defrag(struct sk_buff *skb, u32 user);
 414 #ifdef CONFIG_INET
 415 struct sk_buff *ip_check_defrag(struct sk_buff *skb, u32 user);
 416 #else
 417 static inline struct sk_buff *ip_check_defrag(struct sk_buff *skb, u32 user)
 418 {
 419         return skb;
 420 }
 421 #endif
 422 int ip_frag_mem(struct net *net);
 423 int ip_frag_nqueues(struct net *net);
 424
 425 /*
 426  *      Functions provided by ip_forward.c
 427  */
 428
 429 extern int ip_forward(struct sk_buff *skb);
 430
 431 /*
 432  *      Functions provided by ip_options.c
 433  */
 434
 435 extern void ip_options_build(struct sk_buff *skb, struct ip_options *opt,
 436                              __be32 daddr, struct rtable *rt, int is_frag);
 437 extern int ip_options_echo(struct ip_options *dopt, struct sk_buff *skb);
 438 extern void ip_options_fragment(struct sk_buff *skb);
 439 extern int ip_options_compile(struct net *net,
 440                               struct ip_options *opt, struct sk_buff *skb);
 441 extern int ip_options_get(struct net *net, struct ip_options_rcu **optp,
 442                           unsigned char *data, int optlen);
 443 extern int ip_options_get_from_user(struct net *net, struct ip_options_rcu **optp,
 444                                     unsigned char __user *data, int optlen);
 445 extern void ip_options_undo(struct ip_options * opt);
 446 extern void ip_forward_options(struct sk_buff *skb);
 447 extern int ip_options_rcv_srr(struct sk_buff *skb);
 448
 449 /*
 450  *      Functions provided by ip_sockglue.c
 451  */
 452
 453 extern void     ipv4_pktinfo_prepare(struct sk_buff *skb);
 454 extern void     ip_cmsg_recv(struct msghdr *msg, struct sk_buff *skb);
 455 extern int      ip_cmsg_send(struct net *net,
 456                              struct msghdr *msg, struct ipcm_cookie *ipc);
 457 extern int      ip_setsockopt(struct sock *sk, int level, int optname, char __user *optval, unsigned int optlen);
 458 extern int      ip_getsockopt(struct sock *sk, int level, int optname, char __user *optval, int __user *optlen);
 459 extern int      compat_ip_setsockopt(struct sock *sk, int level,
 460                         int optname, char __user *optval, unsigned int optlen);
 461 extern int      compat_ip_getsockopt(struct sock *sk, int level,
 462                         int optname, char __user *optval, int __user *optlen);
 463 extern int      ip_ra_control(struct sock *sk, unsigned char on, void (*destructor)(struct sock *));
 464
 465 extern int      ip_recv_error(struct sock *sk, struct msghdr *msg, int len);
 466 extern void     ip_icmp_error(struct sock *sk, struct sk_buff *skb, int err,
 467                               __be16 port, u32 info, u8 *payload);
 468 extern void     ip_local_error(struct sock *sk, int err, __be32 daddr, __be16 dport,
 469                                u32 info);
 470
 471 #ifdef CONFIG_PROC_FS
 472 extern int ip_misc_proc_init(void);
 473 #endif
 474
 475 #endif  /* _IP_H */