[deliverable/linux.git] / include / net / route.h

/*
 * 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
 *
 * Authors:	Ross Biro
 *		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>
#include <net/inet_sock.h>
#include <linux/in_route.h>
#include <linux/rtnetlink.h>
#include <linux/route.h>
#include <linux/ip.h>
#include <linux/cache.h>
#include <linux/security.h>

#define RTO_ONLINK	0x01

#define RT_CONN_FLAGS(sk)   (RT_TOS(inet_sk(sk)->tos) | sock_flag(sk, SOCK_LOCALROUTE))

struct fib_nh;
struct fib_info;
struct rtable {
	struct dst_entry	dst;

	int			rt_genid;
	unsigned int		rt_flags;
	__u16			rt_type;

	int			rt_route_iif;
	int			rt_iif;
	int			rt_oif;

	/* Info on neighbour */
	__be32			rt_gateway;

	/* Miscellaneous cached information */
	u32			rt_pmtu;
	struct fib_info		*fi; /* for client ref to shared metrics */
};

static inline bool rt_is_input_route(const struct rtable *rt)
{
	return rt->rt_route_iif != 0;
}

static inline bool rt_is_output_route(const struct rtable *rt)
{
	return rt->rt_route_iif == 0;
}

struct ip_rt_acct {
	__u32 	o_bytes;
	__u32 	o_packets;
	__u32 	i_bytes;
	__u32 	i_packets;
};

struct rt_cache_stat {
        unsigned int in_hit;
        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_hit;
        unsigned int out_slow_tot;
        unsigned int out_slow_mc;
        unsigned int gc_total;
        unsigned int gc_ignored;
        unsigned int gc_goal_miss;
        unsigned int gc_dst_overflow;
        unsigned int in_hlist_search;
        unsigned int out_hlist_search;
};

extern struct ip_rt_acct __percpu *ip_rt_acct;

struct in_device;
extern int		ip_rt_init(void);
extern void		rt_cache_flush(struct net *net, int how);
extern struct rtable *__ip_route_output_key(struct net *, struct flowi4 *flp);
extern struct rtable *ip_route_output_flow(struct net *, struct flowi4 *flp,
					   struct sock *sk);
extern struct dst_entry *ipv4_blackhole_route(struct net *net, struct dst_entry *dst_orig);

static inline struct rtable *ip_route_output_key(struct net *net, struct flowi4 *flp)
{
	return ip_route_output_flow(net, flp, NULL);
}

static inline struct rtable *ip_route_output(struct net *net, __be32 daddr,
					     __be32 saddr, u8 tos, int oif)
{
	struct flowi4 fl4 = {
		.flowi4_oif = oif,
		.flowi4_tos = tos,
		.daddr = daddr,
		.saddr = saddr,
	};
	return ip_route_output_key(net, &fl4);
}

static inline struct rtable *ip_route_output_ports(struct net *net, struct flowi4 *fl4,
						   struct sock *sk,
						   __be32 daddr, __be32 saddr,
						   __be16 dport, __be16 sport,
						   __u8 proto, __u8 tos, int oif)
{
	flowi4_init_output(fl4, oif, sk ? sk->sk_mark : 0, tos,
			   RT_SCOPE_UNIVERSE, proto,
			   sk ? inet_sk_flowi_flags(sk) : 0,
			   daddr, saddr, dport, sport);
	if (sk)
		security_sk_classify_flow(sk, flowi4_to_flowi(fl4));
	return ip_route_output_flow(net, fl4, sk);
}

static inline struct rtable *ip_route_output_gre(struct net *net, struct flowi4 *fl4,
						 __be32 daddr, __be32 saddr,
						 __be32 gre_key, __u8 tos, int oif)
{
	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);
}

extern int ip_route_input(struct sk_buff *skb, __be32 dst, __be32 src,
			  u8 tos, struct net_device *devin);

extern void ipv4_update_pmtu(struct sk_buff *skb, struct net *net, u32 mtu,
			     int oif, u32 mark, u8 protocol, int flow_flags);
extern void ipv4_sk_update_pmtu(struct sk_buff *skb, struct sock *sk, u32 mtu);
extern void ipv4_redirect(struct sk_buff *skb, struct net *net,
			  int oif, u32 mark, u8 protocol, int flow_flags);
extern void ipv4_sk_redirect(struct sk_buff *skb, struct sock *sk);
extern void ip_rt_send_redirect(struct sk_buff *skb);

extern unsigned int		inet_addr_type(struct net *net, __be32 addr);
extern unsigned int		inet_dev_addr_type(struct net *net, const struct net_device *dev, __be32 addr);
extern void		ip_rt_multicast_event(struct in_device *);
extern int		ip_rt_ioctl(struct net *, unsigned int cmd, void __user *arg);
extern void		ip_rt_get_source(u8 *src, struct sk_buff *skb, struct rtable *rt);
extern int		ip_rt_dump(struct sk_buff *skb,  struct netlink_callback *cb);

struct in_ifaddr;
extern void fib_add_ifaddr(struct in_ifaddr *);
extern void fib_del_ifaddr(struct in_ifaddr *, struct in_ifaddr *);

static inline void ip_rt_put(struct rtable * rt)
{
	if (rt)
		dst_release(&rt->dst);
}

#define IPTOS_RT_MASK	(IPTOS_TOS_MASK & ~3)

extern const __u8 ip_tos2prio[16];

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

/* 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,
					 struct sock *sk, bool can_sleep)
{
	__u8 flow_flags = 0;

	if (inet_sk(sk)->transparent)
		flow_flags |= FLOWI_FLAG_ANYSRC;
	if (can_sleep)
		flow_flags |= FLOWI_FLAG_CAN_SLEEP;

	flowi4_init_output(fl4, oif, sk->sk_mark, tos, RT_SCOPE_UNIVERSE,
			   protocol, flow_flags, dst, src, dport, sport);
}

static inline struct rtable *ip_route_connect(struct flowi4 *fl4,
					      __be32 dst, __be32 src, u32 tos,
					      int oif, u8 protocol,
					      __be16 sport, __be16 dport,
					      struct sock *sk, bool can_sleep)
{
	struct net *net = sock_net(sk);
	struct rtable *rt;

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

	if (!dst || !src) {
		rt = __ip_route_output_key(net, fl4);
		if (IS_ERR(rt))
			return rt;
		ip_rt_put(rt);
		flowi4_update_output(fl4, oif, tos, fl4->daddr, fl4->saddr);
	}
	security_sk_classify_flow(sk, flowi4_to_flowi(fl4));
	return ip_route_output_flow(net, fl4, sk);
}

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)
{
	if (sport != orig_sport || dport != orig_dport) {
		fl4->fl4_dport = dport;
		fl4->fl4_sport = sport;
		ip_rt_put(rt);
		flowi4_update_output(fl4, sk->sk_bound_dev_if,
				     RT_CONN_FLAGS(sk), fl4->daddr,
				     fl4->saddr);
		security_sk_classify_flow(sk, flowi4_to_flowi(fl4));
		return ip_route_output_flow(sock_net(sk), fl4, sk);
	}
	return rt;
}

static inline int inet_iif(const struct sk_buff *skb)
{
	return skb_rtable(skb)->rt_iif;
}

extern int sysctl_ip_default_ttl;

static inline int ip4_dst_hoplimit(const struct dst_entry *dst)
{
	int hoplimit = dst_metric_raw(dst, RTAX_HOPLIMIT);

	if (hoplimit == 0)
		hoplimit = sysctl_ip_default_ttl;
	return hoplimit;
}

#endif	/* _ROUTE_H */
Commit	Line	Data
1da177e4 LT	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	*
02c30a84	10	* Authors: Ross Biro
1da177e4 LT	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
1da177e4 LT	27	#include <net/dst.h>
	28	#include <net/inetpeer.h>
	29	#include <net/flow.h>
79876874	30	#include <net/inet_sock.h>
1da177e4 LT	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>
beb8d13b	36	#include <linux/security.h>
1da177e4	37
1da177e4 LT	38	#define RTO_ONLINK 0x01
1da177e4 LT	39
1da177e4 LT	40	#define RT_CONN_FLAGS(sk) (RT_TOS(inet_sk(sk)->tos) \| sock_flag(sk, SOCK_LOCALROUTE))
	41
	42	struct fib_nh;
62fa8a84	43	struct fib_info;
fd2c3ef7	44	struct rtable {
d8d1f30b	45	struct dst_entry dst;
1da177e4	46
29e75252	47	int rt_genid;
95c96174	48	unsigned int rt_flags;
1da177e4	49	__u16 rt_type;
1da177e4	50
1b86a58f	51	int rt_route_iif;
1da177e4	52	int rt_iif;
5e2b61f7	53	int rt_oif;
1da177e4 LT	54
1da177e4 LT	55	/* Info on neighbour */
f2c3fe24	56	__be32 rt_gateway;
1da177e4	57
1da177e4	58	/* Miscellaneous cached information */
5943634f	59	u32 rt_pmtu;
62fa8a84	60	struct fib_info fi; / for client ref to shared metrics */
1da177e4 LT	61	};
1da177e4 LT	62
b8400f37	63	static inline bool rt_is_input_route(const struct rtable *rt)
c7537967	64	{
1b86a58f	65	return rt->rt_route_iif != 0;
c7537967 DM	66	}
c7537967 DM	67
b8400f37	68	static inline bool rt_is_output_route(const struct rtable *rt)
c7537967	69	{
1b86a58f	70	return rt->rt_route_iif == 0;
c7537967 DM	71	}
c7537967 DM	72
fd2c3ef7	73	struct ip_rt_acct {
1da177e4 LT	74	__u32 o_bytes;
	75	__u32 o_packets;
	76	__u32 i_bytes;
	77	__u32 i_packets;
	78	};
	79
fd2c3ef7	80	struct rt_cache_stat {
1da177e4 LT	81	unsigned int in_hit;
	82	unsigned int in_slow_tot;
	83	unsigned int in_slow_mc;
	84	unsigned int in_no_route;
	85	unsigned int in_brd;
	86	unsigned int in_martian_dst;
	87	unsigned int in_martian_src;
	88	unsigned int out_hit;
	89	unsigned int out_slow_tot;
	90	unsigned int out_slow_mc;
	91	unsigned int gc_total;
	92	unsigned int gc_ignored;
	93	unsigned int gc_goal_miss;
	94	unsigned int gc_dst_overflow;
	95	unsigned int in_hlist_search;
	96	unsigned int out_hlist_search;
	97	};
	98
7d720c3e	99	extern struct ip_rt_acct __percpu *ip_rt_acct;
1da177e4 LT	100
	101	struct in_device;
	102	extern int ip_rt_init(void);
76e6ebfb	103	extern void rt_cache_flush(struct net *net, int how);
813b3b5d	104	extern struct rtable __ip_route_output_key(struct net , struct flowi4 *flp);
9d6ec938	105	extern struct rtable ip_route_output_flow(struct net , struct flowi4 *flp,
b23dd4fe	106	struct sock *sk);
2774c131	107	extern struct dst_entry ipv4_blackhole_route(struct net net, struct dst_entry *dst_orig);
407eadd9	108
9d6ec938	109	static inline struct rtable ip_route_output_key(struct net net, struct flowi4 *flp)
5bfa787f DM	110	{
	111	return ip_route_output_flow(net, flp, NULL);
	112	}
	113
78fbfd8a DM	114	static inline struct rtable ip_route_output(struct net net, __be32 daddr,
	115	__be32 saddr, u8 tos, int oif)
	116	{
9d6ec938 DM	117	struct flowi4 fl4 = {
9d6ec938 DM	118	.flowi4_oif = oif,
c5d21c4b	119	.flowi4_tos = tos,
9d6ec938 DM	120	.daddr = daddr,
9d6ec938 DM	121	.saddr = saddr,
78fbfd8a	122	};
9d6ec938	123	return ip_route_output_key(net, &fl4);
78fbfd8a DM	124	}
78fbfd8a DM	125
31e4543d DM	126	static inline struct rtable ip_route_output_ports(struct net net, struct flowi4 *fl4,
31e4543d DM	127	struct sock *sk,
78fbfd8a DM	128	__be32 daddr, __be32 saddr,
	129	__be16 dport, __be16 sport,
	130	__u8 proto, __u8 tos, int oif)
	131	{
31e4543d	132	flowi4_init_output(fl4, oif, sk ? sk->sk_mark : 0, tos,
94b92b88 DM	133	RT_SCOPE_UNIVERSE, proto,
	134	sk ? inet_sk_flowi_flags(sk) : 0,
	135	daddr, saddr, dport, sport);
78fbfd8a	136	if (sk)
31e4543d DM	137	security_sk_classify_flow(sk, flowi4_to_flowi(fl4));
31e4543d DM	138	return ip_route_output_flow(net, fl4, sk);
78fbfd8a DM	139	}
78fbfd8a DM	140
cbb1e85f	141	static inline struct rtable ip_route_output_gre(struct net net, struct flowi4 *fl4,
78fbfd8a DM	142	__be32 daddr, __be32 saddr,
	143	__be32 gre_key, __u8 tos, int oif)
	144	{
cbb1e85f DM	145	memset(fl4, 0, sizeof(*fl4));
	146	fl4->flowi4_oif = oif;
	147	fl4->daddr = daddr;
	148	fl4->saddr = saddr;
	149	fl4->flowi4_tos = tos;
	150	fl4->flowi4_proto = IPPROTO_GRE;
	151	fl4->fl4_gre_key = gre_key;
	152	return ip_route_output_key(net, fl4);
78fbfd8a DM	153	}
78fbfd8a DM	154
38a424e4 DM	155	extern int ip_route_input(struct sk_buff *skb, __be32 dst, __be32 src,
38a424e4 DM	156	u8 tos, struct net_device *devin);
407eadd9	157
36393395 DM	158	extern void ipv4_update_pmtu(struct sk_buff skb, struct net net, u32 mtu,
	159	int oif, u32 mark, u8 protocol, int flow_flags);
	160	extern void ipv4_sk_update_pmtu(struct sk_buff skb, struct sock sk, u32 mtu);
b42597e2 DM	161	extern void ipv4_redirect(struct sk_buff skb, struct net net,
	162	int oif, u32 mark, u8 protocol, int flow_flags);
	163	extern void ipv4_sk_redirect(struct sk_buff skb, struct sock sk);
36393395	164	extern void ip_rt_send_redirect(struct sk_buff *skb);
1da177e4	165
95c96174 ED	166	extern unsigned int inet_addr_type(struct net *net, __be32 addr);
95c96174 ED	167	extern unsigned int inet_dev_addr_type(struct net net, const struct net_device dev, __be32 addr);
1da177e4	168	extern void ip_rt_multicast_event(struct in_device *);
1bad118a	169	extern int ip_rt_ioctl(struct net , unsigned int cmd, void __user arg);
8e36360a	170	extern void ip_rt_get_source(u8 src, struct sk_buff skb, struct rtable *rt);
1da177e4 LT	171	extern int ip_rt_dump(struct sk_buff skb, struct netlink_callback cb);
1da177e4 LT	172
0ff60a45 JHS	173	struct in_ifaddr;
0ff60a45 JHS	174	extern void fib_add_ifaddr(struct in_ifaddr *);
e6abbaa2	175	extern void fib_del_ifaddr(struct in_ifaddr , struct in_ifaddr );
0ff60a45	176
1da177e4 LT	177	static inline void ip_rt_put(struct rtable * rt)
	178	{
	179	if (rt)
d8d1f30b	180	dst_release(&rt->dst);
1da177e4 LT	181	}
	182
	183	#define IPTOS_RT_MASK (IPTOS_TOS_MASK & ~3)
	184
4839c52b	185	extern const __u8 ip_tos2prio[16];
1da177e4 LT	186
	187	static inline char rt_tos2priority(u8 tos)
	188	{
	189	return ip_tos2prio[IPTOS_TOS(tos)>>1];
	190	}
	191
2d7192d6 DM	192	/* ip_route_connect() and ip_route_newports() work in tandem whilst
	193	* binding a socket for a new outgoing connection.
	194	*
	195	* In order to use IPSEC properly, we must, in the end, have a
	196	* route that was looked up using all available keys including source
	197	* and destination ports.
	198	*
	199	* However, if a source port needs to be allocated (the user specified
	200	* a wildcard source port) we need to obtain addressing information
	201	* in order to perform that allocation.
	202	*
	203	* So ip_route_connect() looks up a route using wildcarded source and
	204	* destination ports in the key, simply so that we can get a pair of
	205	* addresses to use for port allocation.
	206	*
	207	* Later, once the ports are allocated, ip_route_newports() will make
	208	* another route lookup if needed to make sure we catch any IPSEC
	209	* rules keyed on the port information.
	210	*
	211	* The callers allocate the flow key on their stack, and must pass in
	212	* the same flowi4 object to both the ip_route_connect() and the
	213	* ip_route_newports() calls.
	214	*/
	215
	216	static inline void ip_route_connect_init(struct flowi4 *fl4, __be32 dst, __be32 src,
	217	u32 tos, int oif, u8 protocol,
	218	__be16 sport, __be16 dport,
	219	struct sock *sk, bool can_sleep)
1da177e4	220	{
2d7192d6	221	__u8 flow_flags = 0;
79876874 KK	222
79876874 KK	223	if (inet_sk(sk)->transparent)
94b92b88	224	flow_flags \|= FLOWI_FLAG_ANYSRC;
5df65e55	225	if (can_sleep)
94b92b88 DM	226	flow_flags \|= FLOWI_FLAG_CAN_SLEEP;
94b92b88 DM	227
2d7192d6	228	flowi4_init_output(fl4, oif, sk->sk_mark, tos, RT_SCOPE_UNIVERSE,
94b92b88	229	protocol, flow_flags, dst, src, dport, sport);
2d7192d6 DM	230	}
	231
	232	static inline struct rtable ip_route_connect(struct flowi4 fl4,
	233	__be32 dst, __be32 src, u32 tos,
	234	int oif, u8 protocol,
	235	__be16 sport, __be16 dport,
	236	struct sock *sk, bool can_sleep)
	237	{
	238	struct net *net = sock_net(sk);
	239	struct rtable *rt;
	240
	241	ip_route_connect_init(fl4, dst, src, tos, oif, protocol,
	242	sport, dport, sk, can_sleep);
79876874	243
1da177e4	244	if (!dst \|\| !src) {
2d7192d6	245	rt = __ip_route_output_key(net, fl4);
b23dd4fe DM	246	if (IS_ERR(rt))
b23dd4fe DM	247	return rt;
b23dd4fe	248	ip_rt_put(rt);
e6b45241	249	flowi4_update_output(fl4, oif, tos, fl4->daddr, fl4->saddr);
1da177e4	250	}
2d7192d6 DM	251	security_sk_classify_flow(sk, flowi4_to_flowi(fl4));
2d7192d6 DM	252	return ip_route_output_flow(net, fl4, sk);
1da177e4 LT	253	}
1da177e4 LT	254
2d7192d6 DM	255	static inline struct rtable ip_route_newports(struct flowi4 fl4, struct rtable *rt,
	256	__be16 orig_sport, __be16 orig_dport,
	257	__be16 sport, __be16 dport,
	258	struct sock *sk)
1da177e4	259	{
dca8b089	260	if (sport != orig_sport \|\| dport != orig_dport) {
2d7192d6 DM	261	fl4->fl4_dport = dport;
2d7192d6 DM	262	fl4->fl4_sport = sport;
b23dd4fe	263	ip_rt_put(rt);
e6b45241 JA	264	flowi4_update_output(fl4, sk->sk_bound_dev_if,
	265	RT_CONN_FLAGS(sk), fl4->daddr,
	266	fl4->saddr);
2d7192d6 DM	267	security_sk_classify_flow(sk, flowi4_to_flowi(fl4));
2d7192d6 DM	268	return ip_route_output_flow(sock_net(sk), fl4, sk);
1da177e4	269	}
b23dd4fe	270	return rt;
1da177e4 LT	271	}
1da177e4 LT	272
1668e010 KK	273	static inline int inet_iif(const struct sk_buff *skb)
1668e010 KK	274	{
511c3f92	275	return skb_rtable(skb)->rt_iif;
1668e010 KK	276	}
1668e010 KK	277
323e126f DM	278	extern int sysctl_ip_default_ttl;
	279
	280	static inline int ip4_dst_hoplimit(const struct dst_entry *dst)
	281	{
	282	int hoplimit = dst_metric_raw(dst, RTAX_HOPLIMIT);
	283
	284	if (hoplimit == 0)
	285	hoplimit = sysctl_ip_default_ttl;
	286	return hoplimit;
	287	}
	288
1da177e4	289	#endif /* _ROUTE_H */