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 router.
7 *
8 * Version: @(#)route.h 1.0.4 05/27/93
9 *
10 * Authors: Ross Biro
11 * Fred N. van Kempen, <waltje@uWalt.NL.Mugnet.ORG>
12 * Fixes:
13 * Alan Cox : Reformatted. Added ip_rt_local()
14 * Alan Cox : Support for TCP parameters.
15 * Alexey Kuznetsov: Major changes for new routing code.
16 * Mike McLagan : Routing by source
17 * Robert Olsson : Added rt_cache statistics
18 *
19 * This program is free software; you can redistribute it and/or
20 * modify it under the terms of the GNU General Public License
21 * as published by the Free Software Foundation; either version
22 * 2 of the License, or (at your option) any later version.
23 */
24 #ifndef _ROUTE_H
25 #define _ROUTE_H
26
27 #include <net/dst.h>
28 #include <net/inetpeer.h>
29 #include <net/flow.h>
30 #include <net/inet_sock.h>
31 #include <linux/in_route.h>
32 #include <linux/rtnetlink.h>
33 #include <linux/route.h>
34 #include <linux/ip.h>
35 #include <linux/cache.h>
36 #include <linux/security.h>
37
38 #ifndef __KERNEL__
39 #warning This file is not supposed to be used outside of kernel.
40 #endif
41
42 #define RTO_ONLINK 0x01
43
44 #define RTO_CONN 0
45 /* RTO_CONN is not used (being alias for 0), but preserved not to break
46 * some modules referring to it. */
47
48 #define RT_CONN_FLAGS(sk) (RT_TOS(inet_sk(sk)->tos) | sock_flag(sk, SOCK_LOCALROUTE))
49
50 struct fib_nh;
51 struct inet_peer;
52 struct fib_info;
53 struct rtable {
54 struct dst_entry dst;
55
56 /* Lookup key. */
57 __be32 rt_key_dst;
58 __be32 rt_key_src;
59
60 int rt_genid;
61 unsigned rt_flags;
62 __u16 rt_type;
63 __u8 rt_tos;
64
65 __be32 rt_dst; /* Path destination */
66 __be32 rt_src; /* Path source */
67 int rt_route_iif;
68 int rt_iif;
69 int rt_oif;
70 __u32 rt_mark;
71
72 /* Info on neighbour */
73 __be32 rt_gateway;
74
75 /* Miscellaneous cached information */
76 __be32 rt_spec_dst; /* RFC1122 specific destination */
77 u32 rt_peer_genid;
78 struct inet_peer *peer; /* long-living peer info */
79 struct fib_info *fi; /* for client ref to shared metrics */
80 };
81
rt_is_input_route(struct rtable * rt)82 static inline bool rt_is_input_route(struct rtable *rt)
83 {
84 return rt->rt_route_iif != 0;
85 }
86
rt_is_output_route(struct rtable * rt)87 static inline bool rt_is_output_route(struct rtable *rt)
88 {
89 return rt->rt_route_iif == 0;
90 }
91
92 struct ip_rt_acct {
93 __u32 o_bytes;
94 __u32 o_packets;
95 __u32 i_bytes;
96 __u32 i_packets;
97 };
98
99 struct rt_cache_stat {
100 unsigned int in_hit;
101 unsigned int in_slow_tot;
102 unsigned int in_slow_mc;
103 unsigned int in_no_route;
104 unsigned int in_brd;
105 unsigned int in_martian_dst;
106 unsigned int in_martian_src;
107 unsigned int out_hit;
108 unsigned int out_slow_tot;
109 unsigned int out_slow_mc;
110 unsigned int gc_total;
111 unsigned int gc_ignored;
112 unsigned int gc_goal_miss;
113 unsigned int gc_dst_overflow;
114 unsigned int in_hlist_search;
115 unsigned int out_hlist_search;
116 };
117
118 extern struct ip_rt_acct __percpu *ip_rt_acct;
119
120 struct in_device;
121 extern int ip_rt_init(void);
122 extern void ip_rt_redirect(__be32 old_gw, __be32 dst, __be32 new_gw,
123 __be32 src, struct net_device *dev);
124 extern void rt_cache_flush(struct net *net, int how);
125 extern void rt_cache_flush_batch(struct net *net);
126 extern struct rtable *__ip_route_output_key(struct net *, const struct flowi4 *flp);
127 extern struct rtable *ip_route_output_flow(struct net *, struct flowi4 *flp,
128 struct sock *sk);
129 extern struct dst_entry *ipv4_blackhole_route(struct net *net, struct dst_entry *dst_orig);
130
ip_route_output_key(struct net * net,struct flowi4 * flp)131 static inline struct rtable *ip_route_output_key(struct net *net, struct flowi4 *flp)
132 {
133 return ip_route_output_flow(net, flp, NULL);
134 }
135
ip_route_output(struct net * net,__be32 daddr,__be32 saddr,u8 tos,int oif)136 static inline struct rtable *ip_route_output(struct net *net, __be32 daddr,
137 __be32 saddr, u8 tos, int oif)
138 {
139 struct flowi4 fl4 = {
140 .flowi4_oif = oif,
141 .daddr = daddr,
142 .saddr = saddr,
143 .flowi4_tos = tos,
144 };
145 return ip_route_output_key(net, &fl4);
146 }
147
ip_route_output_ports(struct net * net,struct sock * sk,__be32 daddr,__be32 saddr,__be16 dport,__be16 sport,__u8 proto,__u8 tos,int oif)148 static inline struct rtable *ip_route_output_ports(struct net *net, struct sock *sk,
149 __be32 daddr, __be32 saddr,
150 __be16 dport, __be16 sport,
151 __u8 proto, __u8 tos, int oif)
152 {
153 struct flowi4 fl4 = {
154 .flowi4_oif = oif,
155 .flowi4_flags = sk ? inet_sk_flowi_flags(sk) : 0,
156 .flowi4_mark = sk ? sk->sk_mark : 0,
157 .daddr = daddr,
158 .saddr = saddr,
159 .flowi4_tos = tos,
160 .flowi4_proto = proto,
161 .fl4_dport = dport,
162 .fl4_sport = sport,
163 };
164 if (sk)
165 security_sk_classify_flow(sk, flowi4_to_flowi(&fl4));
166 return ip_route_output_flow(net, &fl4, sk);
167 }
168
ip_route_output_gre(struct net * net,__be32 daddr,__be32 saddr,__be32 gre_key,__u8 tos,int oif)169 static inline struct rtable *ip_route_output_gre(struct net *net,
170 __be32 daddr, __be32 saddr,
171 __be32 gre_key, __u8 tos, int oif)
172 {
173 struct flowi4 fl4 = {
174 .flowi4_oif = oif,
175 .daddr = daddr,
176 .saddr = saddr,
177 .flowi4_tos = tos,
178 .flowi4_proto = IPPROTO_GRE,
179 .fl4_gre_key = gre_key,
180 };
181 return ip_route_output_key(net, &fl4);
182 }
183
184 extern int ip_route_input_common(struct sk_buff *skb, __be32 dst, __be32 src,
185 u8 tos, struct net_device *devin, bool noref);
186
ip_route_input(struct sk_buff * skb,__be32 dst,__be32 src,u8 tos,struct net_device * devin)187 static inline int ip_route_input(struct sk_buff *skb, __be32 dst, __be32 src,
188 u8 tos, struct net_device *devin)
189 {
190 return ip_route_input_common(skb, dst, src, tos, devin, false);
191 }
192
ip_route_input_noref(struct sk_buff * skb,__be32 dst,__be32 src,u8 tos,struct net_device * devin)193 static inline int ip_route_input_noref(struct sk_buff *skb, __be32 dst, __be32 src,
194 u8 tos, struct net_device *devin)
195 {
196 return ip_route_input_common(skb, dst, src, tos, devin, true);
197 }
198
199 extern unsigned short ip_rt_frag_needed(struct net *net, struct iphdr *iph, unsigned short new_mtu, struct net_device *dev);
200 extern void ip_rt_send_redirect(struct sk_buff *skb);
201
202 extern unsigned inet_addr_type(struct net *net, __be32 addr);
203 extern unsigned inet_dev_addr_type(struct net *net, const struct net_device *dev, __be32 addr);
204 extern void ip_rt_multicast_event(struct in_device *);
205 extern int ip_rt_ioctl(struct net *, unsigned int cmd, void __user *arg);
206 extern void ip_rt_get_source(u8 *src, struct rtable *rt);
207 extern int ip_rt_dump(struct sk_buff *skb, struct netlink_callback *cb);
208
209 struct in_ifaddr;
210 extern void fib_add_ifaddr(struct in_ifaddr *);
211 extern void fib_del_ifaddr(struct in_ifaddr *, struct in_ifaddr *);
212
ip_rt_put(struct rtable * rt)213 static inline void ip_rt_put(struct rtable * rt)
214 {
215 if (rt)
216 dst_release(&rt->dst);
217 }
218
219 #define IPTOS_RT_MASK (IPTOS_TOS_MASK & ~3)
220
221 extern const __u8 ip_tos2prio[16];
222
rt_tos2priority(u8 tos)223 static inline char rt_tos2priority(u8 tos)
224 {
225 return ip_tos2prio[IPTOS_TOS(tos)>>1];
226 }
227
ip_route_connect(__be32 dst,__be32 src,u32 tos,int oif,u8 protocol,__be16 sport,__be16 dport,struct sock * sk,bool can_sleep)228 static inline struct rtable *ip_route_connect(__be32 dst, __be32 src, u32 tos,
229 int oif, u8 protocol,
230 __be16 sport, __be16 dport,
231 struct sock *sk, bool can_sleep)
232 {
233 struct flowi4 fl4 = {
234 .flowi4_oif = oif,
235 .flowi4_mark = sk->sk_mark,
236 .daddr = dst,
237 .saddr = src,
238 .flowi4_tos = tos,
239 .flowi4_proto = protocol,
240 .fl4_sport = sport,
241 .fl4_dport = dport,
242 };
243 struct net *net = sock_net(sk);
244 struct rtable *rt;
245
246 if (inet_sk(sk)->transparent)
247 fl4.flowi4_flags |= FLOWI_FLAG_ANYSRC;
248 if (protocol == IPPROTO_TCP)
249 fl4.flowi4_flags |= FLOWI_FLAG_PRECOW_METRICS;
250 if (can_sleep)
251 fl4.flowi4_flags |= FLOWI_FLAG_CAN_SLEEP;
252
253 if (!dst || !src) {
254 rt = __ip_route_output_key(net, &fl4);
255 if (IS_ERR(rt))
256 return rt;
257 fl4.daddr = rt->rt_dst;
258 fl4.saddr = rt->rt_src;
259 ip_rt_put(rt);
260 }
261 security_sk_classify_flow(sk, flowi4_to_flowi(&fl4));
262 return ip_route_output_flow(net, &fl4, sk);
263 }
264
ip_route_newports(struct rtable * rt,u8 protocol,__be16 orig_sport,__be16 orig_dport,__be16 sport,__be16 dport,struct sock * sk)265 static inline struct rtable *ip_route_newports(struct rtable *rt,
266 u8 protocol, __be16 orig_sport,
267 __be16 orig_dport, __be16 sport,
268 __be16 dport, struct sock *sk)
269 {
270 if (sport != orig_sport || dport != orig_dport) {
271 struct flowi4 fl4 = {
272 .flowi4_oif = rt->rt_oif,
273 .flowi4_mark = rt->rt_mark,
274 .daddr = rt->rt_dst,
275 .saddr = rt->rt_src,
276 .flowi4_tos = rt->rt_tos,
277 .flowi4_proto = protocol,
278 .fl4_sport = sport,
279 .fl4_dport = dport
280 };
281 if (inet_sk(sk)->transparent)
282 fl4.flowi4_flags |= FLOWI_FLAG_ANYSRC;
283 if (protocol == IPPROTO_TCP)
284 fl4.flowi4_flags |= FLOWI_FLAG_PRECOW_METRICS;
285 ip_rt_put(rt);
286 security_sk_classify_flow(sk, flowi4_to_flowi(&fl4));
287 return ip_route_output_flow(sock_net(sk), &fl4, sk);
288 }
289 return rt;
290 }
291
292 extern void rt_bind_peer(struct rtable *rt, int create);
293
rt_get_peer(struct rtable * rt)294 static inline struct inet_peer *rt_get_peer(struct rtable *rt)
295 {
296 if (rt->peer)
297 return rt->peer;
298
299 rt_bind_peer(rt, 0);
300 return rt->peer;
301 }
302
inet_iif(const struct sk_buff * skb)303 static inline int inet_iif(const struct sk_buff *skb)
304 {
305 return skb_rtable(skb)->rt_iif;
306 }
307
308 extern int sysctl_ip_default_ttl;
309
ip4_dst_hoplimit(const struct dst_entry * dst)310 static inline int ip4_dst_hoplimit(const struct dst_entry *dst)
311 {
312 int hoplimit = dst_metric_raw(dst, RTAX_HOPLIMIT);
313
314 if (hoplimit == 0)
315 hoplimit = sysctl_ip_default_ttl;
316 return hoplimit;
317 }
318
319 #endif /* _ROUTE_H */
320