route.h 10 KB
Newer Older
Linus Torvalds's avatar
Linus Torvalds committed
1 2 3 4 5 6 7 8 9
/*
 * INET		An implementation of the TCP/IP protocol suite for the LINUX
 *		operating system.  INET  is implemented using the  BSD Socket
 *		interface as the means of communication with the user level.
 *
 *		Definitions for the IP router.
 *
 * Version:	@(#)route.h	1.0.4	05/27/93
 *
10
 * Authors:	Ross Biro
Linus Torvalds's avatar
Linus Torvalds committed
11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
 *		Fred N. van Kempen, <waltje@uWalt.NL.Mugnet.ORG>
 * Fixes:
 *		Alan Cox	:	Reformatted. Added ip_rt_local()
 *		Alan Cox	:	Support for TCP parameters.
 *		Alexey Kuznetsov:	Major changes for new routing code.
 *		Mike McLagan    :	Routing by source
 *		Robert Olsson   :	Added rt_cache statistics
 *
 *		This program is free software; you can redistribute it and/or
 *		modify it under the terms of the GNU General Public License
 *		as published by the Free Software Foundation; either version
 *		2 of the License, or (at your option) any later version.
 */
#ifndef _ROUTE_H
#define _ROUTE_H

#include <net/dst.h>
#include <net/inetpeer.h>
#include <net/flow.h>
30
#include <net/inet_sock.h>
31
#include <net/ip_fib.h>
Linus Torvalds's avatar
Linus Torvalds committed
32 33
#include <linux/in_route.h>
#include <linux/rtnetlink.h>
34
#include <linux/rcupdate.h>
Linus Torvalds's avatar
Linus Torvalds committed
35 36 37
#include <linux/route.h>
#include <linux/ip.h>
#include <linux/cache.h>
38
#include <linux/security.h>
Linus Torvalds's avatar
Linus Torvalds committed
39

40 41 42
/* IPv4 datagram length is stored into 16bit field (tot_len) */
#define IP_MAX_MTU	0xFFFFU

Linus Torvalds's avatar
Linus Torvalds committed
43 44 45
#define RTO_ONLINK	0x01

#define RT_CONN_FLAGS(sk)   (RT_TOS(inet_sk(sk)->tos) | sock_flag(sk, SOCK_LOCALROUTE))
46
#define RT_CONN_FLAGS_TOS(sk,tos)   (RT_TOS(tos) | sock_flag(sk, SOCK_LOCALROUTE))
Linus Torvalds's avatar
Linus Torvalds committed
47 48

struct fib_nh;
49
struct fib_info;
Eric Dumazet's avatar
Eric Dumazet committed
50
struct uncached_list;
Eric Dumazet's avatar
Eric Dumazet committed
51
struct rtable {
52
	struct dst_entry	dst;
Linus Torvalds's avatar
Linus Torvalds committed
53

54
	int			rt_genid;
55
	unsigned int		rt_flags;
Linus Torvalds's avatar
Linus Torvalds committed
56
	__u16			rt_type;
57 58
	__u8			rt_is_input;
	__u8			rt_uses_gateway;
Linus Torvalds's avatar
Linus Torvalds committed
59 60 61 62

	int			rt_iif;

	/* Info on neighbour */
63
	__be32			rt_gateway;
Linus Torvalds's avatar
Linus Torvalds committed
64 65

	/* Miscellaneous cached information */
66 67
	u32			rt_mtu_locked:1,
				rt_pmtu:31;
68 69

	struct list_head	rt_uncached;
Eric Dumazet's avatar
Eric Dumazet committed
70
	struct uncached_list	*rt_uncached_list;
Linus Torvalds's avatar
Linus Torvalds committed
71 72
};

73
static inline bool rt_is_input_route(const struct rtable *rt)
74
{
75
	return rt->rt_is_input != 0;
76 77
}

78
static inline bool rt_is_output_route(const struct rtable *rt)
79
{
80
	return rt->rt_is_input == 0;
81 82
}

83 84 85 86 87 88 89
static inline __be32 rt_nexthop(const struct rtable *rt, __be32 daddr)
{
	if (rt->rt_gateway)
		return rt->rt_gateway;
	return daddr;
}

Eric Dumazet's avatar
Eric Dumazet committed
90
struct ip_rt_acct {
Linus Torvalds's avatar
Linus Torvalds committed
91 92 93 94 95 96
	__u32 	o_bytes;
	__u32 	o_packets;
	__u32 	i_bytes;
	__u32 	i_packets;
};

Eric Dumazet's avatar
Eric Dumazet committed
97
struct rt_cache_stat {
Linus Torvalds's avatar
Linus Torvalds committed
98 99 100 101 102 103 104 105 106 107
        unsigned int in_slow_tot;
        unsigned int in_slow_mc;
        unsigned int in_no_route;
        unsigned int in_brd;
        unsigned int in_martian_dst;
        unsigned int in_martian_src;
        unsigned int out_slow_tot;
        unsigned int out_slow_mc;
};

108
extern struct ip_rt_acct __percpu *ip_rt_acct;
Linus Torvalds's avatar
Linus Torvalds committed
109 110

struct in_device;
111 112 113 114

int ip_rt_init(void);
void rt_cache_flush(struct net *net);
void rt_flush_dev(struct net_device *dev);
115 116 117 118 119
struct rtable *ip_route_output_key_hash(struct net *net, struct flowi4 *flp,
					const struct sk_buff *skb);
struct rtable *ip_route_output_key_hash_rcu(struct net *net, struct flowi4 *flp,
					    struct fib_result *res,
					    const struct sk_buff *skb);
120 121 122 123

static inline struct rtable *__ip_route_output_key(struct net *net,
						   struct flowi4 *flp)
{
124
	return ip_route_output_key_hash(net, flp, NULL);
125 126
}

127
struct rtable *ip_route_output_flow(struct net *, struct flowi4 *flp,
128
				    const struct sock *sk);
129 130
struct dst_entry *ipv4_blackhole_route(struct net *net,
				       struct dst_entry *dst_orig);
131

132
static inline struct rtable *ip_route_output_key(struct net *net, struct flowi4 *flp)
133 134 135 136
{
	return ip_route_output_flow(net, flp, NULL);
}

137 138 139
static inline struct rtable *ip_route_output(struct net *net, __be32 daddr,
					     __be32 saddr, u8 tos, int oif)
{
140 141
	struct flowi4 fl4 = {
		.flowi4_oif = oif,
142
		.flowi4_tos = tos,
143 144
		.daddr = daddr,
		.saddr = saddr,
145
	};
146
	return ip_route_output_key(net, &fl4);
147 148
}

149 150
static inline struct rtable *ip_route_output_ports(struct net *net, struct flowi4 *fl4,
						   struct sock *sk,
151 152 153 154
						   __be32 daddr, __be32 saddr,
						   __be16 dport, __be16 sport,
						   __u8 proto, __u8 tos, int oif)
{
155
	flowi4_init_output(fl4, oif, sk ? sk->sk_mark : 0, tos,
156 157
			   RT_SCOPE_UNIVERSE, proto,
			   sk ? inet_sk_flowi_flags(sk) : 0,
158
			   daddr, saddr, dport, sport, sock_net_uid(net, sk));
159
	if (sk)
160 161
		security_sk_classify_flow(sk, flowi4_to_flowi(fl4));
	return ip_route_output_flow(net, fl4, sk);
162 163
}

164
static inline struct rtable *ip_route_output_gre(struct net *net, struct flowi4 *fl4,
165 166 167
						 __be32 daddr, __be32 saddr,
						 __be32 gre_key, __u8 tos, int oif)
{
168 169 170 171 172 173 174 175
	memset(fl4, 0, sizeof(*fl4));
	fl4->flowi4_oif = oif;
	fl4->daddr = daddr;
	fl4->saddr = saddr;
	fl4->flowi4_tos = tos;
	fl4->flowi4_proto = IPPROTO_GRE;
	fl4->fl4_gre_key = gre_key;
	return ip_route_output_key(net, fl4);
176
}
177 178 179
int ip_mc_validate_source(struct sk_buff *skb, __be32 daddr, __be32 saddr,
			  u8 tos, struct net_device *dev,
			  struct in_device *in_dev, u32 *itag);
180 181
int ip_route_input_noref(struct sk_buff *skb, __be32 dst, __be32 src,
			 u8 tos, struct net_device *devin);
182 183 184
int ip_route_input_rcu(struct sk_buff *skb, __be32 dst, __be32 src,
		       u8 tos, struct net_device *devin,
		       struct fib_result *res);
185 186 187 188 189 190 191 192

static inline int ip_route_input(struct sk_buff *skb, __be32 dst, __be32 src,
				 u8 tos, struct net_device *devin)
{
	int err;

	rcu_read_lock();
	err = ip_route_input_noref(skb, dst, src, tos, devin);
193
	if (!err) {
194
		skb_dst_force(skb);
195 196 197
		if (!skb_dst(skb))
			err = -EINVAL;
	}
198 199 200 201
	rcu_read_unlock();

	return err;
}
202

203
void ipv4_update_pmtu(struct sk_buff *skb, struct net *net, u32 mtu, int oif,
204
		      u8 protocol);
205
void ipv4_sk_update_pmtu(struct sk_buff *skb, struct sock *sk, u32 mtu);
206
void ipv4_redirect(struct sk_buff *skb, struct net *net, int oif, u8 protocol);
207 208 209 210
void ipv4_sk_redirect(struct sk_buff *skb, struct sock *sk);
void ip_rt_send_redirect(struct sk_buff *skb);

unsigned int inet_addr_type(struct net *net, __be32 addr);
David Ahern's avatar
David Ahern committed
211
unsigned int inet_addr_type_table(struct net *net, __be32 addr, u32 tb_id);
212 213
unsigned int inet_dev_addr_type(struct net *net, const struct net_device *dev,
				__be32 addr);
214 215 216
unsigned int inet_addr_type_dev_table(struct net *net,
				      const struct net_device *dev,
				      __be32 addr);
217
void ip_rt_multicast_event(struct in_device *);
218
int ip_rt_ioctl(struct net *, unsigned int cmd, struct rtentry *rt);
219
void ip_rt_get_source(u8 *src, struct sk_buff *skb, struct rtable *rt);
220 221 222
struct rtable *rt_dst_alloc(struct net_device *dev,
			     unsigned int flags, u16 type,
			     bool nopolicy, bool noxfrm, bool will_cache);
Linus Torvalds's avatar
Linus Torvalds committed
223

224
struct in_ifaddr;
225 226
void fib_add_ifaddr(struct in_ifaddr *);
void fib_del_ifaddr(struct in_ifaddr *, struct in_ifaddr *);
227
void fib_modify_prefix_metric(struct in_ifaddr *ifa, u32 new_metric);
228

229 230 231
void rt_add_uncached_list(struct rtable *rt);
void rt_del_uncached_list(struct rtable *rt);

232
static inline void ip_rt_put(struct rtable *rt)
Linus Torvalds's avatar
Linus Torvalds committed
233
{
234 235 236 237 238
	/* dst_release() accepts a NULL parameter.
	 * We rely on dst being first structure in struct rtable
	 */
	BUILD_BUG_ON(offsetof(struct rtable, dst) != 0);
	dst_release(&rt->dst);
Linus Torvalds's avatar
Linus Torvalds committed
239 240 241 242
}

#define IPTOS_RT_MASK	(IPTOS_TOS_MASK & ~3)

243
extern const __u8 ip_tos2prio[16];
Linus Torvalds's avatar
Linus Torvalds committed
244 245 246 247 248 249

static inline char rt_tos2priority(u8 tos)
{
	return ip_tos2prio[IPTOS_TOS(tos)>>1];
}

250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276
/* ip_route_connect() and ip_route_newports() work in tandem whilst
 * binding a socket for a new outgoing connection.
 *
 * In order to use IPSEC properly, we must, in the end, have a
 * route that was looked up using all available keys including source
 * and destination ports.
 *
 * However, if a source port needs to be allocated (the user specified
 * a wildcard source port) we need to obtain addressing information
 * in order to perform that allocation.
 *
 * So ip_route_connect() looks up a route using wildcarded source and
 * destination ports in the key, simply so that we can get a pair of
 * addresses to use for port allocation.
 *
 * Later, once the ports are allocated, ip_route_newports() will make
 * another route lookup if needed to make sure we catch any IPSEC
 * rules keyed on the port information.
 *
 * The callers allocate the flow key on their stack, and must pass in
 * the same flowi4 object to both the ip_route_connect() and the
 * ip_route_newports() calls.
 */

static inline void ip_route_connect_init(struct flowi4 *fl4, __be32 dst, __be32 src,
					 u32 tos, int oif, u8 protocol,
					 __be16 sport, __be16 dport,
277
					 struct sock *sk)
Linus Torvalds's avatar
Linus Torvalds committed
278
{
279
	__u8 flow_flags = 0;
280 281

	if (inet_sk(sk)->transparent)
282 283
		flow_flags |= FLOWI_FLAG_ANYSRC;

284
	flowi4_init_output(fl4, oif, sk->sk_mark, tos, RT_SCOPE_UNIVERSE,
285 286
			   protocol, flow_flags, dst, src, dport, sport,
			   sk->sk_uid);
287 288 289 290 291 292
}

static inline struct rtable *ip_route_connect(struct flowi4 *fl4,
					      __be32 dst, __be32 src, u32 tos,
					      int oif, u8 protocol,
					      __be16 sport, __be16 dport,
293
					      struct sock *sk)
294 295 296 297 298
{
	struct net *net = sock_net(sk);
	struct rtable *rt;

	ip_route_connect_init(fl4, dst, src, tos, oif, protocol,
299
			      sport, dport, sk);
300

Linus Torvalds's avatar
Linus Torvalds committed
301
	if (!dst || !src) {
302
		rt = __ip_route_output_key(net, fl4);
303 304 305
		if (IS_ERR(rt))
			return rt;
		ip_rt_put(rt);
306
		flowi4_update_output(fl4, oif, tos, fl4->daddr, fl4->saddr);
Linus Torvalds's avatar
Linus Torvalds committed
307
	}
308 309
	security_sk_classify_flow(sk, flowi4_to_flowi(fl4));
	return ip_route_output_flow(net, fl4, sk);
Linus Torvalds's avatar
Linus Torvalds committed
310 311
}

312 313 314 315
static inline struct rtable *ip_route_newports(struct flowi4 *fl4, struct rtable *rt,
					       __be16 orig_sport, __be16 orig_dport,
					       __be16 sport, __be16 dport,
					       struct sock *sk)
Linus Torvalds's avatar
Linus Torvalds committed
316
{
317
	if (sport != orig_sport || dport != orig_dport) {
318 319
		fl4->fl4_dport = dport;
		fl4->fl4_sport = sport;
320
		ip_rt_put(rt);
321 322 323
		flowi4_update_output(fl4, sk->sk_bound_dev_if,
				     RT_CONN_FLAGS(sk), fl4->daddr,
				     fl4->saddr);
324 325
		security_sk_classify_flow(sk, flowi4_to_flowi(fl4));
		return ip_route_output_flow(sock_net(sk), fl4, sk);
Linus Torvalds's avatar
Linus Torvalds committed
326
	}
327
	return rt;
Linus Torvalds's avatar
Linus Torvalds committed
328 329
}

330 331
static inline int inet_iif(const struct sk_buff *skb)
{
332 333 334 335
	struct rtable *rt = skb_rtable(skb);

	if (rt && rt->rt_iif)
		return rt->rt_iif;
336 337

	return skb->skb_iif;
338 339
}

340 341 342
static inline int ip4_dst_hoplimit(const struct dst_entry *dst)
{
	int hoplimit = dst_metric_raw(dst, RTAX_HOPLIMIT);
343
	struct net *net = dev_net(dst->dev);
344 345

	if (hoplimit == 0)
346
		hoplimit = net->ipv4.sysctl_ip_default_ttl;
347 348 349
	return hoplimit;
}

Linus Torvalds's avatar
Linus Torvalds committed
350
#endif	/* _ROUTE_H */