[deliverable/linux.git] / net / ipv4 / fib_rules.c

/*
 * 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.
 *
 *		IPv4 Forwarding Information Base: policy rules.
 *
 * Authors:	Alexey Kuznetsov, <kuznet@ms2.inr.ac.ru>
 *		Thomas Graf <tgraf@suug.ch>
 *
 *		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.
 *
 * Fixes:
 *		Rani Assaf	:	local_rule cannot be deleted
 *		Marc Boucher	:	routing by fwmark
 */

#include <linux/types.h>
#include <linux/kernel.h>
#include <linux/netdevice.h>
#include <linux/netlink.h>
#include <linux/inetdevice.h>
#include <linux/init.h>
#include <linux/list.h>
#include <linux/rcupdate.h>
#include <net/ip.h>
#include <net/route.h>
#include <net/tcp.h>
#include <net/ip_fib.h>
#include <net/fib_rules.h>

struct fib4_rule {
	struct fib_rule		common;
	u8			dst_len;
	u8			src_len;
	u8			tos;
	__be32			src;
	__be32			srcmask;
	__be32			dst;
	__be32			dstmask;
#ifdef CONFIG_IP_ROUTE_CLASSID
	u32			tclassid;
#endif
};

#ifdef CONFIG_IP_ROUTE_CLASSID
u32 fib_rules_tclass(struct fib_result *res)
{
	return res->r ? ((struct fib4_rule *) res->r)->tclassid : 0;
}
#endif

int fib_lookup(struct net *net, struct flowi *flp, struct fib_result *res)
{
	struct fib_lookup_arg arg = {
		.result = res,
		.flags = FIB_LOOKUP_NOREF,
	};
	int err;

	err = fib_rules_lookup(net->ipv4.rules_ops, flp, 0, &arg);
	res->r = arg.rule;

	return err;
}

static int fib4_rule_action(struct fib_rule *rule, struct flowi *flp,
			    int flags, struct fib_lookup_arg *arg)
{
	int err = -EAGAIN;
	struct fib_table *tbl;

	switch (rule->action) {
	case FR_ACT_TO_TBL:
		break;

	case FR_ACT_UNREACHABLE:
		err = -ENETUNREACH;
		goto errout;

	case FR_ACT_PROHIBIT:
		err = -EACCES;
		goto errout;

	case FR_ACT_BLACKHOLE:
	default:
		err = -EINVAL;
		goto errout;
	}

	tbl = fib_get_table(rule->fr_net, rule->table);
	if (!tbl)
		goto errout;

	err = fib_table_lookup(tbl, flp, (struct fib_result *) arg->result, arg->flags);
	if (err > 0)
		err = -EAGAIN;
errout:
	return err;
}


static int fib4_rule_match(struct fib_rule *rule, struct flowi *fl, int flags)
{
	struct fib4_rule *r = (struct fib4_rule *) rule;
	__be32 daddr = fl->fl4_dst;
	__be32 saddr = fl->fl4_src;

	if (((saddr ^ r->src) & r->srcmask) ||
	    ((daddr ^ r->dst) & r->dstmask))
		return 0;

	if (r->tos && (r->tos != fl->fl4_tos))
		return 0;

	return 1;
}

static struct fib_table *fib_empty_table(struct net *net)
{
	u32 id;

	for (id = 1; id <= RT_TABLE_MAX; id++)
		if (fib_get_table(net, id) == NULL)
			return fib_new_table(net, id);
	return NULL;
}

static const struct nla_policy fib4_rule_policy[FRA_MAX+1] = {
	FRA_GENERIC_POLICY,
	[FRA_FLOW]	= { .type = NLA_U32 },
};

static int fib4_rule_configure(struct fib_rule *rule, struct sk_buff *skb,
			       struct fib_rule_hdr *frh,
			       struct nlattr **tb)
{
	struct net *net = sock_net(skb->sk);
	int err = -EINVAL;
	struct fib4_rule *rule4 = (struct fib4_rule *) rule;

	if (frh->tos & ~IPTOS_TOS_MASK)
		goto errout;

	if (rule->table == RT_TABLE_UNSPEC) {
		if (rule->action == FR_ACT_TO_TBL) {
			struct fib_table *table;

			table = fib_empty_table(net);
			if (table == NULL) {
				err = -ENOBUFS;
				goto errout;
			}

			rule->table = table->tb_id;
		}
	}

	if (frh->src_len)
		rule4->src = nla_get_be32(tb[FRA_SRC]);

	if (frh->dst_len)
		rule4->dst = nla_get_be32(tb[FRA_DST]);

#ifdef CONFIG_IP_ROUTE_CLASSID
	if (tb[FRA_FLOW])
		rule4->tclassid = nla_get_u32(tb[FRA_FLOW]);
#endif

	rule4->src_len = frh->src_len;
	rule4->srcmask = inet_make_mask(rule4->src_len);
	rule4->dst_len = frh->dst_len;
	rule4->dstmask = inet_make_mask(rule4->dst_len);
	rule4->tos = frh->tos;

	err = 0;
errout:
	return err;
}

static int fib4_rule_compare(struct fib_rule *rule, struct fib_rule_hdr *frh,
			     struct nlattr **tb)
{
	struct fib4_rule *rule4 = (struct fib4_rule *) rule;

	if (frh->src_len && (rule4->src_len != frh->src_len))
		return 0;

	if (frh->dst_len && (rule4->dst_len != frh->dst_len))
		return 0;

	if (frh->tos && (rule4->tos != frh->tos))
		return 0;

#ifdef CONFIG_IP_ROUTE_CLASSID
	if (tb[FRA_FLOW] && (rule4->tclassid != nla_get_u32(tb[FRA_FLOW])))
		return 0;
#endif

	if (frh->src_len && (rule4->src != nla_get_be32(tb[FRA_SRC])))
		return 0;

	if (frh->dst_len && (rule4->dst != nla_get_be32(tb[FRA_DST])))
		return 0;

	return 1;
}

static int fib4_rule_fill(struct fib_rule *rule, struct sk_buff *skb,
			  struct fib_rule_hdr *frh)
{
	struct fib4_rule *rule4 = (struct fib4_rule *) rule;

	frh->dst_len = rule4->dst_len;
	frh->src_len = rule4->src_len;
	frh->tos = rule4->tos;

	if (rule4->dst_len)
		NLA_PUT_BE32(skb, FRA_DST, rule4->dst);

	if (rule4->src_len)
		NLA_PUT_BE32(skb, FRA_SRC, rule4->src);

#ifdef CONFIG_IP_ROUTE_CLASSID
	if (rule4->tclassid)
		NLA_PUT_U32(skb, FRA_FLOW, rule4->tclassid);
#endif
	return 0;

nla_put_failure:
	return -ENOBUFS;
}

static size_t fib4_rule_nlmsg_payload(struct fib_rule *rule)
{
	return nla_total_size(4) /* dst */
	       + nla_total_size(4) /* src */
	       + nla_total_size(4); /* flow */
}

static void fib4_rule_flush_cache(struct fib_rules_ops *ops)
{
	rt_cache_flush(ops->fro_net, -1);
}

static const struct fib_rules_ops __net_initdata fib4_rules_ops_template = {
	.family		= AF_INET,
	.rule_size	= sizeof(struct fib4_rule),
	.addr_size	= sizeof(u32),
	.action		= fib4_rule_action,
	.match		= fib4_rule_match,
	.configure	= fib4_rule_configure,
	.compare	= fib4_rule_compare,
	.fill		= fib4_rule_fill,
	.default_pref	= fib_default_rule_pref,
	.nlmsg_payload	= fib4_rule_nlmsg_payload,
	.flush_cache	= fib4_rule_flush_cache,
	.nlgroup	= RTNLGRP_IPV4_RULE,
	.policy		= fib4_rule_policy,
	.owner		= THIS_MODULE,
};

static int fib_default_rules_init(struct fib_rules_ops *ops)
{
	int err;

	err = fib_default_rule_add(ops, 0, RT_TABLE_LOCAL, 0);
	if (err < 0)
		return err;
	err = fib_default_rule_add(ops, 0x7FFE, RT_TABLE_MAIN, 0);
	if (err < 0)
		return err;
	err = fib_default_rule_add(ops, 0x7FFF, RT_TABLE_DEFAULT, 0);
	if (err < 0)
		return err;
	return 0;
}

int __net_init fib4_rules_init(struct net *net)
{
	int err;
	struct fib_rules_ops *ops;

	ops = fib_rules_register(&fib4_rules_ops_template, net);
	if (IS_ERR(ops))
		return PTR_ERR(ops);

	err = fib_default_rules_init(ops);
	if (err < 0)
		goto fail;
	net->ipv4.rules_ops = ops;
	return 0;

fail:
	/* also cleans all rules already added */
	fib_rules_unregister(ops);
	return err;
}

void __net_exit fib4_rules_exit(struct net *net)
{
	fib_rules_unregister(net->ipv4.rules_ops);
}
Commit	Line	Data
	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	* IPv4 Forwarding Information Base: policy rules.
	7	*
	8	* Authors: Alexey Kuznetsov, <kuznet@ms2.inr.ac.ru>
	9	* Thomas Graf <tgraf@suug.ch>
	10	*
	11	* This program is free software; you can redistribute it and/or
	12	* modify it under the terms of the GNU General Public License
	13	* as published by the Free Software Foundation; either version
	14	* 2 of the License, or (at your option) any later version.
	15	*
	16	* Fixes:
	17	* Rani Assaf : local_rule cannot be deleted
	18	* Marc Boucher : routing by fwmark
	19	*/
	20
	21	#include <linux/types.h>
	22	#include <linux/kernel.h>
	23	#include <linux/netdevice.h>
	24	#include <linux/netlink.h>
	25	#include <linux/inetdevice.h>
	26	#include <linux/init.h>
	27	#include <linux/list.h>
	28	#include <linux/rcupdate.h>
	29	#include <net/ip.h>
	30	#include <net/route.h>
	31	#include <net/tcp.h>
	32	#include <net/ip_fib.h>
	33	#include <net/fib_rules.h>
	34
	35	struct fib4_rule {
	36	struct fib_rule common;
	37	u8 dst_len;
	38	u8 src_len;
	39	u8 tos;
	40	__be32 src;
	41	__be32 srcmask;
	42	__be32 dst;
	43	__be32 dstmask;
	44	#ifdef CONFIG_IP_ROUTE_CLASSID
	45	u32 tclassid;
	46	#endif
	47	};
	48
	49	#ifdef CONFIG_IP_ROUTE_CLASSID
	50	u32 fib_rules_tclass(struct fib_result *res)
	51	{
	52	return res->r ? ((struct fib4_rule *) res->r)->tclassid : 0;
	53	}
	54	#endif
	55
	56	int fib_lookup(struct net net, struct flowi flp, struct fib_result *res)
	57	{
	58	struct fib_lookup_arg arg = {
	59	.result = res,
	60	.flags = FIB_LOOKUP_NOREF,
	61	};
	62	int err;
	63
	64	err = fib_rules_lookup(net->ipv4.rules_ops, flp, 0, &arg);
	65	res->r = arg.rule;
	66
	67	return err;
	68	}
	69
	70	static int fib4_rule_action(struct fib_rule rule, struct flowi flp,
	71	int flags, struct fib_lookup_arg *arg)
	72	{
	73	int err = -EAGAIN;
	74	struct fib_table *tbl;
	75
	76	switch (rule->action) {
	77	case FR_ACT_TO_TBL:
	78	break;
	79
	80	case FR_ACT_UNREACHABLE:
	81	err = -ENETUNREACH;
	82	goto errout;
	83
	84	case FR_ACT_PROHIBIT:
	85	err = -EACCES;
	86	goto errout;
	87
	88	case FR_ACT_BLACKHOLE:
	89	default:
	90	err = -EINVAL;
	91	goto errout;
	92	}
	93
	94	tbl = fib_get_table(rule->fr_net, rule->table);
	95	if (!tbl)
	96	goto errout;
	97
	98	err = fib_table_lookup(tbl, flp, (struct fib_result *) arg->result, arg->flags);
	99	if (err > 0)
	100	err = -EAGAIN;
	101	errout:
	102	return err;
	103	}
	104
	105
	106	static int fib4_rule_match(struct fib_rule rule, struct flowi fl, int flags)
	107	{
	108	struct fib4_rule r = (struct fib4_rule ) rule;
	109	__be32 daddr = fl->fl4_dst;
	110	__be32 saddr = fl->fl4_src;
	111
	112	if (((saddr ^ r->src) & r->srcmask) \|\|
	113	((daddr ^ r->dst) & r->dstmask))
	114	return 0;
	115
	116	if (r->tos && (r->tos != fl->fl4_tos))
	117	return 0;
	118
	119	return 1;
	120	}
	121
	122	static struct fib_table fib_empty_table(struct net net)
	123	{
	124	u32 id;
	125
	126	for (id = 1; id <= RT_TABLE_MAX; id++)
	127	if (fib_get_table(net, id) == NULL)
	128	return fib_new_table(net, id);
	129	return NULL;
	130	}
	131
	132	static const struct nla_policy fib4_rule_policy[FRA_MAX+1] = {
	133	FRA_GENERIC_POLICY,
	134	[FRA_FLOW] = { .type = NLA_U32 },
	135	};
	136
	137	static int fib4_rule_configure(struct fib_rule rule, struct sk_buff skb,
	138	struct fib_rule_hdr *frh,
	139	struct nlattr **tb)
	140	{
	141	struct net *net = sock_net(skb->sk);
	142	int err = -EINVAL;
	143	struct fib4_rule rule4 = (struct fib4_rule ) rule;
	144
	145	if (frh->tos & ~IPTOS_TOS_MASK)
	146	goto errout;
	147
	148	if (rule->table == RT_TABLE_UNSPEC) {
	149	if (rule->action == FR_ACT_TO_TBL) {
	150	struct fib_table *table;
	151
	152	table = fib_empty_table(net);
	153	if (table == NULL) {
	154	err = -ENOBUFS;
	155	goto errout;
	156	}
	157
	158	rule->table = table->tb_id;
	159	}
	160	}
	161
	162	if (frh->src_len)
	163	rule4->src = nla_get_be32(tb[FRA_SRC]);
	164
	165	if (frh->dst_len)
	166	rule4->dst = nla_get_be32(tb[FRA_DST]);
	167
	168	#ifdef CONFIG_IP_ROUTE_CLASSID
	169	if (tb[FRA_FLOW])
	170	rule4->tclassid = nla_get_u32(tb[FRA_FLOW]);
	171	#endif
	172
	173	rule4->src_len = frh->src_len;
	174	rule4->srcmask = inet_make_mask(rule4->src_len);
	175	rule4->dst_len = frh->dst_len;
	176	rule4->dstmask = inet_make_mask(rule4->dst_len);
	177	rule4->tos = frh->tos;
	178
	179	err = 0;
	180	errout:
	181	return err;
	182	}
	183
	184	static int fib4_rule_compare(struct fib_rule rule, struct fib_rule_hdr frh,
	185	struct nlattr **tb)
	186	{
	187	struct fib4_rule rule4 = (struct fib4_rule ) rule;
	188
	189	if (frh->src_len && (rule4->src_len != frh->src_len))
	190	return 0;
	191
	192	if (frh->dst_len && (rule4->dst_len != frh->dst_len))
	193	return 0;
	194
	195	if (frh->tos && (rule4->tos != frh->tos))
	196	return 0;
	197
	198	#ifdef CONFIG_IP_ROUTE_CLASSID
	199	if (tb[FRA_FLOW] && (rule4->tclassid != nla_get_u32(tb[FRA_FLOW])))
	200	return 0;
	201	#endif
	202
	203	if (frh->src_len && (rule4->src != nla_get_be32(tb[FRA_SRC])))
	204	return 0;
	205
	206	if (frh->dst_len && (rule4->dst != nla_get_be32(tb[FRA_DST])))
	207	return 0;
	208
	209	return 1;
	210	}
	211
	212	static int fib4_rule_fill(struct fib_rule rule, struct sk_buff skb,
	213	struct fib_rule_hdr *frh)
	214	{
	215	struct fib4_rule rule4 = (struct fib4_rule ) rule;
	216
	217	frh->dst_len = rule4->dst_len;
	218	frh->src_len = rule4->src_len;
	219	frh->tos = rule4->tos;
	220
	221	if (rule4->dst_len)
	222	NLA_PUT_BE32(skb, FRA_DST, rule4->dst);
	223
	224	if (rule4->src_len)
	225	NLA_PUT_BE32(skb, FRA_SRC, rule4->src);
	226
	227	#ifdef CONFIG_IP_ROUTE_CLASSID
	228	if (rule4->tclassid)
	229	NLA_PUT_U32(skb, FRA_FLOW, rule4->tclassid);
	230	#endif
	231	return 0;
	232
	233	nla_put_failure:
	234	return -ENOBUFS;
	235	}
	236
	237	static size_t fib4_rule_nlmsg_payload(struct fib_rule *rule)
	238	{
	239	return nla_total_size(4) /* dst */
	240	+ nla_total_size(4) /* src */
	241	+ nla_total_size(4); /* flow */
	242	}
	243
	244	static void fib4_rule_flush_cache(struct fib_rules_ops *ops)
	245	{
	246	rt_cache_flush(ops->fro_net, -1);
	247	}
	248
	249	static const struct fib_rules_ops __net_initdata fib4_rules_ops_template = {
	250	.family = AF_INET,
	251	.rule_size = sizeof(struct fib4_rule),
	252	.addr_size = sizeof(u32),
	253	.action = fib4_rule_action,
	254	.match = fib4_rule_match,
	255	.configure = fib4_rule_configure,
	256	.compare = fib4_rule_compare,
	257	.fill = fib4_rule_fill,
	258	.default_pref = fib_default_rule_pref,
	259	.nlmsg_payload = fib4_rule_nlmsg_payload,
	260	.flush_cache = fib4_rule_flush_cache,
	261	.nlgroup = RTNLGRP_IPV4_RULE,
	262	.policy = fib4_rule_policy,
	263	.owner = THIS_MODULE,
	264	};
	265
	266	static int fib_default_rules_init(struct fib_rules_ops *ops)
	267	{
	268	int err;
	269
	270	err = fib_default_rule_add(ops, 0, RT_TABLE_LOCAL, 0);
	271	if (err < 0)
	272	return err;
	273	err = fib_default_rule_add(ops, 0x7FFE, RT_TABLE_MAIN, 0);
	274	if (err < 0)
	275	return err;
	276	err = fib_default_rule_add(ops, 0x7FFF, RT_TABLE_DEFAULT, 0);
	277	if (err < 0)
	278	return err;
	279	return 0;
	280	}
	281
	282	int __net_init fib4_rules_init(struct net *net)
	283	{
	284	int err;
	285	struct fib_rules_ops *ops;
	286
	287	ops = fib_rules_register(&fib4_rules_ops_template, net);
	288	if (IS_ERR(ops))
	289	return PTR_ERR(ops);
	290
	291	err = fib_default_rules_init(ops);
	292	if (err < 0)
	293	goto fail;
	294	net->ipv4.rules_ops = ops;
	295	return 0;
	296
	297	fail:
	298	/* also cleans all rules already added */
	299	fib_rules_unregister(ops);
	300	return err;
	301	}
	302
	303	void __net_exit fib4_rules_exit(struct net *net)
	304	{
	305	fib_rules_unregister(net->ipv4.rules_ops);
	306	}