[net-next-2.6.git] / net / core / fib_rules.c

/*
 * net/core/fib_rules.c		Generic Routing Rules
 *
 *	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, version 2.
 *
 * Authors:	Thomas Graf <tgraf@suug.ch>
 */

#include <linux/types.h>
#include <linux/kernel.h>
#include <linux/list.h>
#include <net/fib_rules.h>

static LIST_HEAD(rules_ops);
static DEFINE_SPINLOCK(rules_mod_lock);

static void notify_rule_change(int event, struct fib_rule *rule,
			       struct fib_rules_ops *ops, struct nlmsghdr *nlh,
			       u32 pid);

static struct fib_rules_ops *lookup_rules_ops(int family)
{
	struct fib_rules_ops *ops;

	rcu_read_lock();
	list_for_each_entry_rcu(ops, &rules_ops, list) {
		if (ops->family == family) {
			if (!try_module_get(ops->owner))
				ops = NULL;
			rcu_read_unlock();
			return ops;
		}
	}
	rcu_read_unlock();

	return NULL;
}

static void rules_ops_put(struct fib_rules_ops *ops)
{
	if (ops)
		module_put(ops->owner);
}

int fib_rules_register(struct fib_rules_ops *ops)
{
	int err = -EEXIST;
	struct fib_rules_ops *o;

	if (ops->rule_size < sizeof(struct fib_rule))
		return -EINVAL;

	if (ops->match == NULL || ops->configure == NULL ||
	    ops->compare == NULL || ops->fill == NULL ||
	    ops->action == NULL)
		return -EINVAL;

	spin_lock(&rules_mod_lock);
	list_for_each_entry(o, &rules_ops, list)
		if (ops->family == o->family)
			goto errout;

	list_add_tail_rcu(&ops->list, &rules_ops);
	err = 0;
errout:
	spin_unlock(&rules_mod_lock);

	return err;
}

EXPORT_SYMBOL_GPL(fib_rules_register);

static void cleanup_ops(struct fib_rules_ops *ops)
{
	struct fib_rule *rule, *tmp;

	list_for_each_entry_safe(rule, tmp, ops->rules_list, list) {
		list_del_rcu(&rule->list);
		fib_rule_put(rule);
	}
}

int fib_rules_unregister(struct fib_rules_ops *ops)
{
	int err = 0;
	struct fib_rules_ops *o;

	spin_lock(&rules_mod_lock);
	list_for_each_entry(o, &rules_ops, list) {
		if (o == ops) {
			list_del_rcu(&o->list);
			cleanup_ops(ops);
			goto out;
		}
	}

	err = -ENOENT;
out:
	spin_unlock(&rules_mod_lock);

	synchronize_rcu();

	return err;
}

EXPORT_SYMBOL_GPL(fib_rules_unregister);

static int fib_rule_match(struct fib_rule *rule, struct fib_rules_ops *ops,
			  struct flowi *fl, int flags)
{
	int ret = 0;

	if (rule->ifindex && (rule->ifindex != fl->iif))
		goto out;

	if ((rule->mark ^ fl->mark) & rule->mark_mask)
		goto out;

	ret = ops->match(rule, fl, flags);
out:
	return (rule->flags & FIB_RULE_INVERT) ? !ret : ret;
}

int fib_rules_lookup(struct fib_rules_ops *ops, struct flowi *fl,
		     int flags, struct fib_lookup_arg *arg)
{
	struct fib_rule *rule;
	int err;

	rcu_read_lock();

	list_for_each_entry_rcu(rule, ops->rules_list, list) {
jumped:
		if (!fib_rule_match(rule, ops, fl, flags))
			continue;

		if (rule->action == FR_ACT_GOTO) {
			struct fib_rule *target;

			target = rcu_dereference(rule->ctarget);
			if (target == NULL) {
				continue;
			} else {
				rule = target;
				goto jumped;
			}
		} else
			err = ops->action(rule, fl, flags, arg);

		if (err != -EAGAIN) {
			fib_rule_get(rule);
			arg->rule = rule;
			goto out;
		}
	}

	err = -ESRCH;
out:
	rcu_read_unlock();

	return err;
}

EXPORT_SYMBOL_GPL(fib_rules_lookup);

static int validate_rulemsg(struct fib_rule_hdr *frh, struct nlattr **tb,
			    struct fib_rules_ops *ops)
{
	int err = -EINVAL;

	if (frh->src_len)
		if (tb[FRA_SRC] == NULL ||
		    frh->src_len > (ops->addr_size * 8) ||
		    nla_len(tb[FRA_SRC]) != ops->addr_size)
			goto errout;

	if (frh->dst_len)
		if (tb[FRA_DST] == NULL ||
		    frh->dst_len > (ops->addr_size * 8) ||
		    nla_len(tb[FRA_DST]) != ops->addr_size)
			goto errout;

	err = 0;
errout:
	return err;
}

static int fib_nl_newrule(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg)
{
	struct fib_rule_hdr *frh = nlmsg_data(nlh);
	struct fib_rules_ops *ops = NULL;
	struct fib_rule *rule, *r, *last = NULL;
	struct nlattr *tb[FRA_MAX+1];
	int err = -EINVAL, unresolved = 0;

	if (nlh->nlmsg_len < nlmsg_msg_size(sizeof(*frh)))
		goto errout;

	ops = lookup_rules_ops(frh->family);
	if (ops == NULL) {
		err = EAFNOSUPPORT;
		goto errout;
	}

	err = nlmsg_parse(nlh, sizeof(*frh), tb, FRA_MAX, ops->policy);
	if (err < 0)
		goto errout;

	err = validate_rulemsg(frh, tb, ops);
	if (err < 0)
		goto errout;

	rule = kzalloc(ops->rule_size, GFP_KERNEL);
	if (rule == NULL) {
		err = -ENOMEM;
		goto errout;
	}

	if (tb[FRA_PRIORITY])
		rule->pref = nla_get_u32(tb[FRA_PRIORITY]);

	if (tb[FRA_IFNAME]) {
		struct net_device *dev;

		rule->ifindex = -1;
		nla_strlcpy(rule->ifname, tb[FRA_IFNAME], IFNAMSIZ);
		dev = __dev_get_by_name(rule->ifname);
		if (dev)
			rule->ifindex = dev->ifindex;
	}

	if (tb[FRA_FWMARK]) {
		rule->mark = nla_get_u32(tb[FRA_FWMARK]);
		if (rule->mark)
			/* compatibility: if the mark value is non-zero all bits
			 * are compared unless a mask is explicitly specified.
			 */
			rule->mark_mask = 0xFFFFFFFF;
	}

	if (tb[FRA_FWMASK])
		rule->mark_mask = nla_get_u32(tb[FRA_FWMASK]);

	rule->action = frh->action;
	rule->flags = frh->flags;
	rule->table = frh_get_table(frh, tb);

	if (!rule->pref && ops->default_pref)
		rule->pref = ops->default_pref();

	err = -EINVAL;
	if (tb[FRA_GOTO]) {
		if (rule->action != FR_ACT_GOTO)
			goto errout_free;

		rule->target = nla_get_u32(tb[FRA_GOTO]);
		/* Backward jumps are prohibited to avoid endless loops */
		if (rule->target <= rule->pref)
			goto errout_free;

		list_for_each_entry(r, ops->rules_list, list) {
			if (r->pref == rule->target) {
				rule->ctarget = r;
				break;
			}
		}

		if (rule->ctarget == NULL)
			unresolved = 1;
	} else if (rule->action == FR_ACT_GOTO)
		goto errout_free;

	err = ops->configure(rule, skb, nlh, frh, tb);
	if (err < 0)
		goto errout_free;

	list_for_each_entry(r, ops->rules_list, list) {
		if (r->pref > rule->pref)
			break;
		last = r;
	}

	fib_rule_get(rule);

	if (ops->unresolved_rules) {
		/*
		 * There are unresolved goto rules in the list, check if
		 * any of them are pointing to this new rule.
		 */
		list_for_each_entry(r, ops->rules_list, list) {
			if (r->action == FR_ACT_GOTO &&
			    r->target == rule->pref) {
				BUG_ON(r->ctarget != NULL);
				rcu_assign_pointer(r->ctarget, rule);
				if (--ops->unresolved_rules == 0)
					break;
			}
		}
	}

	if (rule->action == FR_ACT_GOTO)
		ops->nr_goto_rules++;

	if (unresolved)
		ops->unresolved_rules++;

	if (last)
		list_add_rcu(&rule->list, &last->list);
	else
		list_add_rcu(&rule->list, ops->rules_list);

	notify_rule_change(RTM_NEWRULE, rule, ops, nlh, NETLINK_CB(skb).pid);
	rules_ops_put(ops);
	return 0;

errout_free:
	kfree(rule);
errout:
	rules_ops_put(ops);
	return err;
}

static int fib_nl_delrule(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg)
{
	struct fib_rule_hdr *frh = nlmsg_data(nlh);
	struct fib_rules_ops *ops = NULL;
	struct fib_rule *rule, *tmp;
	struct nlattr *tb[FRA_MAX+1];
	int err = -EINVAL;

	if (nlh->nlmsg_len < nlmsg_msg_size(sizeof(*frh)))
		goto errout;

	ops = lookup_rules_ops(frh->family);
	if (ops == NULL) {
		err = EAFNOSUPPORT;
		goto errout;
	}

	err = nlmsg_parse(nlh, sizeof(*frh), tb, FRA_MAX, ops->policy);
	if (err < 0)
		goto errout;

	err = validate_rulemsg(frh, tb, ops);
	if (err < 0)
		goto errout;

	list_for_each_entry(rule, ops->rules_list, list) {
		if (frh->action && (frh->action != rule->action))
			continue;

		if (frh->table && (frh_get_table(frh, tb) != rule->table))
			continue;

		if (tb[FRA_PRIORITY] &&
		    (rule->pref != nla_get_u32(tb[FRA_PRIORITY])))
			continue;

		if (tb[FRA_IFNAME] &&
		    nla_strcmp(tb[FRA_IFNAME], rule->ifname))
			continue;

		if (tb[FRA_FWMARK] &&
		    (rule->mark != nla_get_u32(tb[FRA_FWMARK])))
			continue;

		if (tb[FRA_FWMASK] &&
		    (rule->mark_mask != nla_get_u32(tb[FRA_FWMASK])))
			continue;

		if (!ops->compare(rule, frh, tb))
			continue;

		if (rule->flags & FIB_RULE_PERMANENT) {
			err = -EPERM;
			goto errout;
		}

		list_del_rcu(&rule->list);

		if (rule->action == FR_ACT_GOTO)
			ops->nr_goto_rules--;

		/*
		 * Check if this rule is a target to any of them. If so,
		 * disable them. As this operation is eventually very
		 * expensive, it is only performed if goto rules have
		 * actually been added.
		 */
		if (ops->nr_goto_rules > 0) {
			list_for_each_entry(tmp, ops->rules_list, list) {
				if (tmp->ctarget == rule) {
					rcu_assign_pointer(tmp->ctarget, NULL);
					ops->unresolved_rules++;
				}
			}
		}

		synchronize_rcu();
		notify_rule_change(RTM_DELRULE, rule, ops, nlh,
				   NETLINK_CB(skb).pid);
		fib_rule_put(rule);
		rules_ops_put(ops);
		return 0;
	}

	err = -ENOENT;
errout:
	rules_ops_put(ops);
	return err;
}

static inline size_t fib_rule_nlmsg_size(struct fib_rules_ops *ops,
					 struct fib_rule *rule)
{
	size_t payload = NLMSG_ALIGN(sizeof(struct fib_rule_hdr))
			 + nla_total_size(IFNAMSIZ) /* FRA_IFNAME */
			 + nla_total_size(4) /* FRA_PRIORITY */
			 + nla_total_size(4) /* FRA_TABLE */
			 + nla_total_size(4) /* FRA_FWMARK */
			 + nla_total_size(4); /* FRA_FWMASK */

	if (ops->nlmsg_payload)
		payload += ops->nlmsg_payload(rule);

	return payload;
}

static int fib_nl_fill_rule(struct sk_buff *skb, struct fib_rule *rule,
			    u32 pid, u32 seq, int type, int flags,
			    struct fib_rules_ops *ops)
{
	struct nlmsghdr *nlh;
	struct fib_rule_hdr *frh;

	nlh = nlmsg_put(skb, pid, seq, type, sizeof(*frh), flags);
	if (nlh == NULL)
		return -EMSGSIZE;

	frh = nlmsg_data(nlh);
	frh->table = rule->table;
	NLA_PUT_U32(skb, FRA_TABLE, rule->table);
	frh->res1 = 0;
	frh->res2 = 0;
	frh->action = rule->action;
	frh->flags = rule->flags;

	if (rule->action == FR_ACT_GOTO && rule->ctarget == NULL)
		frh->flags |= FIB_RULE_UNRESOLVED;

	if (rule->ifname[0])
		NLA_PUT_STRING(skb, FRA_IFNAME, rule->ifname);

	if (rule->pref)
		NLA_PUT_U32(skb, FRA_PRIORITY, rule->pref);

	if (rule->mark)
		NLA_PUT_U32(skb, FRA_FWMARK, rule->mark);

	if (rule->mark_mask || rule->mark)
		NLA_PUT_U32(skb, FRA_FWMASK, rule->mark_mask);

	if (rule->target)
		NLA_PUT_U32(skb, FRA_GOTO, rule->target);

	if (ops->fill(rule, skb, nlh, frh) < 0)
		goto nla_put_failure;

	return nlmsg_end(skb, nlh);

nla_put_failure:
	nlmsg_cancel(skb, nlh);
	return -EMSGSIZE;
}

static int dump_rules(struct sk_buff *skb, struct netlink_callback *cb,
		      struct fib_rules_ops *ops)
{
	int idx = 0;
	struct fib_rule *rule;

	rcu_read_lock();
	list_for_each_entry_rcu(rule, ops->rules_list, list) {
		if (idx < cb->args[1])
			goto skip;

		if (fib_nl_fill_rule(skb, rule, NETLINK_CB(cb->skb).pid,
				     cb->nlh->nlmsg_seq, RTM_NEWRULE,
				     NLM_F_MULTI, ops) < 0)
			break;
skip:
		idx++;
	}
	rcu_read_unlock();
	cb->args[1] = idx;
	rules_ops_put(ops);

	return skb->len;
}

static int fib_nl_dumprule(struct sk_buff *skb, struct netlink_callback *cb)
{
	struct fib_rules_ops *ops;
	int idx = 0, family;

	family = rtnl_msg_family(cb->nlh);
	if (family != AF_UNSPEC) {
		/* Protocol specific dump request */
		ops = lookup_rules_ops(family);
		if (ops == NULL)
			return -EAFNOSUPPORT;

		return dump_rules(skb, cb, ops);
	}

	rcu_read_lock();
	list_for_each_entry_rcu(ops, &rules_ops, list) {
		if (idx < cb->args[0] || !try_module_get(ops->owner))
			goto skip;

		if (dump_rules(skb, cb, ops) < 0)
			break;

		cb->args[1] = 0;
	skip:
		idx++;
	}
	rcu_read_unlock();
	cb->args[0] = idx;

	return skb->len;
}

static void notify_rule_change(int event, struct fib_rule *rule,
			       struct fib_rules_ops *ops, struct nlmsghdr *nlh,
			       u32 pid)
{
	struct sk_buff *skb;
	int err = -ENOBUFS;

	skb = nlmsg_new(fib_rule_nlmsg_size(ops, rule), GFP_KERNEL);
	if (skb == NULL)
		goto errout;

	err = fib_nl_fill_rule(skb, rule, pid, nlh->nlmsg_seq, event, 0, ops);
	if (err < 0) {
		/* -EMSGSIZE implies BUG in fib_rule_nlmsg_size() */
		WARN_ON(err == -EMSGSIZE);
		kfree_skb(skb);
		goto errout;
	}
	err = rtnl_notify(skb, pid, ops->nlgroup, nlh, GFP_KERNEL);
errout:
	if (err < 0)
		rtnl_set_sk_err(ops->nlgroup, err);
}

static void attach_rules(struct list_head *rules, struct net_device *dev)
{
	struct fib_rule *rule;

	list_for_each_entry(rule, rules, list) {
		if (rule->ifindex == -1 &&
		    strcmp(dev->name, rule->ifname) == 0)
			rule->ifindex = dev->ifindex;
	}
}

static void detach_rules(struct list_head *rules, struct net_device *dev)
{
	struct fib_rule *rule;

	list_for_each_entry(rule, rules, list)
		if (rule->ifindex == dev->ifindex)
			rule->ifindex = -1;
}


static int fib_rules_event(struct notifier_block *this, unsigned long event,
			    void *ptr)
{
	struct net_device *dev = ptr;
	struct fib_rules_ops *ops;

	ASSERT_RTNL();
	rcu_read_lock();

	switch (event) {
	case NETDEV_REGISTER:
		list_for_each_entry(ops, &rules_ops, list)
			attach_rules(ops->rules_list, dev);
		break;

	case NETDEV_UNREGISTER:
		list_for_each_entry(ops, &rules_ops, list)
			detach_rules(ops->rules_list, dev);
		break;
	}

	rcu_read_unlock();

	return NOTIFY_DONE;
}

static struct notifier_block fib_rules_notifier = {
	.notifier_call = fib_rules_event,
};

static int __init fib_rules_init(void)
{
	rtnl_register(PF_UNSPEC, RTM_NEWRULE, fib_nl_newrule, NULL);
	rtnl_register(PF_UNSPEC, RTM_DELRULE, fib_nl_delrule, NULL);
	rtnl_register(PF_UNSPEC, RTM_GETRULE, NULL, fib_nl_dumprule);

	return register_netdevice_notifier(&fib_rules_notifier);
}

subsys_initcall(fib_rules_init);
Commit	Line	Data
14c0b97d TG	1	/*
	2	* net/core/fib_rules.c Generic Routing Rules
	3	*
	4	* This program is free software; you can redistribute it and/or
	5	* modify it under the terms of the GNU General Public License as
	6	* published by the Free Software Foundation, version 2.
	7	*
	8	* Authors: Thomas Graf <tgraf@suug.ch>
	9	*/
	10
14c0b97d TG	11	#include <linux/types.h>
	12	#include <linux/kernel.h>
	13	#include <linux/list.h>
	14	#include <net/fib_rules.h>
	15
	16	static LIST_HEAD(rules_ops);
	17	static DEFINE_SPINLOCK(rules_mod_lock);
	18
	19	static void notify_rule_change(int event, struct fib_rule *rule,
c17084d2 TG	20	struct fib_rules_ops ops, struct nlmsghdr nlh,
c17084d2 TG	21	u32 pid);
14c0b97d TG	22
	23	static struct fib_rules_ops *lookup_rules_ops(int family)
	24	{
	25	struct fib_rules_ops *ops;
	26
	27	rcu_read_lock();
	28	list_for_each_entry_rcu(ops, &rules_ops, list) {
	29	if (ops->family == family) {
	30	if (!try_module_get(ops->owner))
	31	ops = NULL;
	32	rcu_read_unlock();
	33	return ops;
	34	}
	35	}
	36	rcu_read_unlock();
	37
	38	return NULL;
	39	}
	40
	41	static void rules_ops_put(struct fib_rules_ops *ops)
	42	{
	43	if (ops)
	44	module_put(ops->owner);
	45	}
	46
	47	int fib_rules_register(struct fib_rules_ops *ops)
	48	{
	49	int err = -EEXIST;
	50	struct fib_rules_ops *o;
	51
	52	if (ops->rule_size < sizeof(struct fib_rule))
	53	return -EINVAL;
	54
	55	if (ops->match == NULL \|\| ops->configure == NULL \|\|
	56	ops->compare == NULL \|\| ops->fill == NULL \|\|
	57	ops->action == NULL)
	58	return -EINVAL;
	59
	60	spin_lock(&rules_mod_lock);
	61	list_for_each_entry(o, &rules_ops, list)
	62	if (ops->family == o->family)
	63	goto errout;
	64
	65	list_add_tail_rcu(&ops->list, &rules_ops);
	66	err = 0;
	67	errout:
	68	spin_unlock(&rules_mod_lock);
	69
	70	return err;
	71	}
	72
	73	EXPORT_SYMBOL_GPL(fib_rules_register);
	74
	75	static void cleanup_ops(struct fib_rules_ops *ops)
	76	{
	77	struct fib_rule rule, tmp;
	78
	79	list_for_each_entry_safe(rule, tmp, ops->rules_list, list) {
	80	list_del_rcu(&rule->list);
	81	fib_rule_put(rule);
	82	}
	83	}
	84
	85	int fib_rules_unregister(struct fib_rules_ops *ops)
86	{
87	int err = 0;
88	struct fib_rules_ops *o;
89
90	spin_lock(&rules_mod_lock);
91	list_for_each_entry(o, &rules_ops, list) {
92	if (o == ops) {
93	list_del_rcu(&o->list);
94	cleanup_ops(ops);
95	goto out;
96	}
97	}
98
99	err = -ENOENT;
100	out:
101	spin_unlock(&rules_mod_lock);
102
103	synchronize_rcu();
104
105	return err;
106	}
107
108	EXPORT_SYMBOL_GPL(fib_rules_unregister);
109
3dfbcc41 TG	110	static int fib_rule_match(struct fib_rule rule, struct fib_rules_ops ops,
	111	struct flowi *fl, int flags)
	112	{
	113	int ret = 0;
	114
	115	if (rule->ifindex && (rule->ifindex != fl->iif))
	116	goto out;
	117
	118	if ((rule->mark ^ fl->mark) & rule->mark_mask)
	119	goto out;
	120
	121	ret = ops->match(rule, fl, flags);
	122	out:
	123	return (rule->flags & FIB_RULE_INVERT) ? !ret : ret;
	124	}
	125
14c0b97d TG	126	int fib_rules_lookup(struct fib_rules_ops ops, struct flowi fl,
	127	int flags, struct fib_lookup_arg *arg)
	128	{
	129	struct fib_rule *rule;
	130	int err;
	131
	132	rcu_read_lock();
	133
	134	list_for_each_entry_rcu(rule, ops->rules_list, list) {
0947c9fe	135	jumped:
3dfbcc41	136	if (!fib_rule_match(rule, ops, fl, flags))
14c0b97d TG	137	continue;
14c0b97d TG	138
0947c9fe TG	139	if (rule->action == FR_ACT_GOTO) {
	140	struct fib_rule *target;
	141
	142	target = rcu_dereference(rule->ctarget);
	143	if (target == NULL) {
	144	continue;
	145	} else {
	146	rule = target;
	147	goto jumped;
	148	}
	149	} else
	150	err = ops->action(rule, fl, flags, arg);
	151
14c0b97d TG	152	if (err != -EAGAIN) {
	153	fib_rule_get(rule);
	154	arg->rule = rule;
	155	goto out;
	156	}
	157	}
	158
83886b6b	159	err = -ESRCH;
14c0b97d TG	160	out:
	161	rcu_read_unlock();
	162
	163	return err;
	164	}
	165
	166	EXPORT_SYMBOL_GPL(fib_rules_lookup);
	167
e1701c68 TG	168	static int validate_rulemsg(struct fib_rule_hdr frh, struct nlattr *tb,
	169	struct fib_rules_ops *ops)
	170	{
	171	int err = -EINVAL;
	172
	173	if (frh->src_len)
	174	if (tb[FRA_SRC] == NULL \|\|
	175	frh->src_len > (ops->addr_size * 8) \|\|
	176	nla_len(tb[FRA_SRC]) != ops->addr_size)
	177	goto errout;
	178
	179	if (frh->dst_len)
	180	if (tb[FRA_DST] == NULL \|\|
	181	frh->dst_len > (ops->addr_size * 8) \|\|
	182	nla_len(tb[FRA_DST]) != ops->addr_size)
	183	goto errout;
	184
	185	err = 0;
	186	errout:
	187	return err;
	188	}
	189
9d9e6a58	190	static int fib_nl_newrule(struct sk_buff skb, struct nlmsghdr nlh, void *arg)
14c0b97d TG	191	{
	192	struct fib_rule_hdr *frh = nlmsg_data(nlh);
	193	struct fib_rules_ops *ops = NULL;
	194	struct fib_rule rule, r, *last = NULL;
	195	struct nlattr *tb[FRA_MAX+1];
0947c9fe	196	int err = -EINVAL, unresolved = 0;
14c0b97d TG	197
	198	if (nlh->nlmsg_len < nlmsg_msg_size(sizeof(*frh)))
	199	goto errout;
	200
	201	ops = lookup_rules_ops(frh->family);
	202	if (ops == NULL) {
	203	err = EAFNOSUPPORT;
	204	goto errout;
	205	}
	206
	207	err = nlmsg_parse(nlh, sizeof(*frh), tb, FRA_MAX, ops->policy);
	208	if (err < 0)
	209	goto errout;
	210
e1701c68 TG	211	err = validate_rulemsg(frh, tb, ops);
	212	if (err < 0)
	213	goto errout;
	214
14c0b97d TG	215	rule = kzalloc(ops->rule_size, GFP_KERNEL);
	216	if (rule == NULL) {
	217	err = -ENOMEM;
	218	goto errout;
	219	}
	220
	221	if (tb[FRA_PRIORITY])
	222	rule->pref = nla_get_u32(tb[FRA_PRIORITY]);
	223
	224	if (tb[FRA_IFNAME]) {
	225	struct net_device *dev;
	226
	227	rule->ifindex = -1;
5176f91e	228	nla_strlcpy(rule->ifname, tb[FRA_IFNAME], IFNAMSIZ);
14c0b97d TG	229	dev = __dev_get_by_name(rule->ifname);
	230	if (dev)
	231	rule->ifindex = dev->ifindex;
	232	}
	233
b8964ed9 TG	234	if (tb[FRA_FWMARK]) {
	235	rule->mark = nla_get_u32(tb[FRA_FWMARK]);
	236	if (rule->mark)
	237	/* compatibility: if the mark value is non-zero all bits
	238	* are compared unless a mask is explicitly specified.
	239	*/
	240	rule->mark_mask = 0xFFFFFFFF;
	241	}
	242
	243	if (tb[FRA_FWMASK])
	244	rule->mark_mask = nla_get_u32(tb[FRA_FWMASK]);
	245
14c0b97d TG	246	rule->action = frh->action;
14c0b97d TG	247	rule->flags = frh->flags;
9e762a4a	248	rule->table = frh_get_table(frh, tb);
14c0b97d TG	249
	250	if (!rule->pref && ops->default_pref)
	251	rule->pref = ops->default_pref();
	252
0947c9fe TG	253	err = -EINVAL;
	254	if (tb[FRA_GOTO]) {
	255	if (rule->action != FR_ACT_GOTO)
	256	goto errout_free;
	257
	258	rule->target = nla_get_u32(tb[FRA_GOTO]);
	259	/* Backward jumps are prohibited to avoid endless loops */
	260	if (rule->target <= rule->pref)
	261	goto errout_free;
	262
	263	list_for_each_entry(r, ops->rules_list, list) {
	264	if (r->pref == rule->target) {
	265	rule->ctarget = r;
	266	break;
	267	}
	268	}
	269
	270	if (rule->ctarget == NULL)
	271	unresolved = 1;
	272	} else if (rule->action == FR_ACT_GOTO)
	273	goto errout_free;
	274
14c0b97d TG	275	err = ops->configure(rule, skb, nlh, frh, tb);
	276	if (err < 0)
	277	goto errout_free;
	278
	279	list_for_each_entry(r, ops->rules_list, list) {
	280	if (r->pref > rule->pref)
	281	break;
	282	last = r;
	283	}
	284
	285	fib_rule_get(rule);
	286
0947c9fe TG	287	if (ops->unresolved_rules) {
	288	/*
	289	* There are unresolved goto rules in the list, check if
	290	* any of them are pointing to this new rule.
	291	*/
	292	list_for_each_entry(r, ops->rules_list, list) {
	293	if (r->action == FR_ACT_GOTO &&
	294	r->target == rule->pref) {
	295	BUG_ON(r->ctarget != NULL);
	296	rcu_assign_pointer(r->ctarget, rule);
	297	if (--ops->unresolved_rules == 0)
	298	break;
	299	}
	300	}
	301	}
	302
	303	if (rule->action == FR_ACT_GOTO)
	304	ops->nr_goto_rules++;
	305
	306	if (unresolved)
	307	ops->unresolved_rules++;
	308
14c0b97d TG	309	if (last)
	310	list_add_rcu(&rule->list, &last->list);
	311	else
	312	list_add_rcu(&rule->list, ops->rules_list);
	313
c17084d2	314	notify_rule_change(RTM_NEWRULE, rule, ops, nlh, NETLINK_CB(skb).pid);
14c0b97d TG	315	rules_ops_put(ops);
	316	return 0;
	317
	318	errout_free:
	319	kfree(rule);
	320	errout:
	321	rules_ops_put(ops);
	322	return err;
	323	}
	324
9d9e6a58	325	static int fib_nl_delrule(struct sk_buff skb, struct nlmsghdr nlh, void *arg)
14c0b97d TG	326	{
	327	struct fib_rule_hdr *frh = nlmsg_data(nlh);
	328	struct fib_rules_ops *ops = NULL;
0947c9fe	329	struct fib_rule rule, tmp;
14c0b97d TG	330	struct nlattr *tb[FRA_MAX+1];
	331	int err = -EINVAL;
	332
	333	if (nlh->nlmsg_len < nlmsg_msg_size(sizeof(*frh)))
	334	goto errout;
	335
	336	ops = lookup_rules_ops(frh->family);
	337	if (ops == NULL) {
	338	err = EAFNOSUPPORT;
	339	goto errout;
	340	}
	341
	342	err = nlmsg_parse(nlh, sizeof(*frh), tb, FRA_MAX, ops->policy);
	343	if (err < 0)
	344	goto errout;
	345
e1701c68 TG	346	err = validate_rulemsg(frh, tb, ops);
	347	if (err < 0)
	348	goto errout;
	349
14c0b97d TG	350	list_for_each_entry(rule, ops->rules_list, list) {
	351	if (frh->action && (frh->action != rule->action))
	352	continue;
	353
9e762a4a	354	if (frh->table && (frh_get_table(frh, tb) != rule->table))
14c0b97d TG	355	continue;
	356
	357	if (tb[FRA_PRIORITY] &&
	358	(rule->pref != nla_get_u32(tb[FRA_PRIORITY])))
	359	continue;
	360
	361	if (tb[FRA_IFNAME] &&
	362	nla_strcmp(tb[FRA_IFNAME], rule->ifname))
	363	continue;
	364
b8964ed9 TG	365	if (tb[FRA_FWMARK] &&
	366	(rule->mark != nla_get_u32(tb[FRA_FWMARK])))
	367	continue;
	368
	369	if (tb[FRA_FWMASK] &&
	370	(rule->mark_mask != nla_get_u32(tb[FRA_FWMASK])))
	371	continue;
	372
14c0b97d TG	373	if (!ops->compare(rule, frh, tb))
	374	continue;
	375
	376	if (rule->flags & FIB_RULE_PERMANENT) {
	377	err = -EPERM;
	378	goto errout;
	379	}
	380
	381	list_del_rcu(&rule->list);
0947c9fe TG	382
	383	if (rule->action == FR_ACT_GOTO)
	384	ops->nr_goto_rules--;
	385
	386	/*
	387	* Check if this rule is a target to any of them. If so,
	388	* disable them. As this operation is eventually very
	389	* expensive, it is only performed if goto rules have
	390	* actually been added.
	391	*/
	392	if (ops->nr_goto_rules > 0) {
	393	list_for_each_entry(tmp, ops->rules_list, list) {
	394	if (tmp->ctarget == rule) {
	395	rcu_assign_pointer(tmp->ctarget, NULL);
	396	ops->unresolved_rules++;
	397	}
	398	}
	399	}
	400
14c0b97d	401	synchronize_rcu();
c17084d2 TG	402	notify_rule_change(RTM_DELRULE, rule, ops, nlh,
c17084d2 TG	403	NETLINK_CB(skb).pid);
14c0b97d TG	404	fib_rule_put(rule);
	405	rules_ops_put(ops);
	406	return 0;
	407	}
	408
	409	err = -ENOENT;
	410	errout:
	411	rules_ops_put(ops);
	412	return err;
	413	}
	414
339bf98f TG	415	static inline size_t fib_rule_nlmsg_size(struct fib_rules_ops *ops,
	416	struct fib_rule *rule)
	417	{
	418	size_t payload = NLMSG_ALIGN(sizeof(struct fib_rule_hdr))
	419	+ nla_total_size(IFNAMSIZ) /* FRA_IFNAME */
	420	+ nla_total_size(4) /* FRA_PRIORITY */
	421	+ nla_total_size(4) /* FRA_TABLE */
	422	+ nla_total_size(4) /* FRA_FWMARK */
	423	+ nla_total_size(4); /* FRA_FWMASK */
	424
	425	if (ops->nlmsg_payload)
	426	payload += ops->nlmsg_payload(rule);
	427
	428	return payload;
	429	}
	430
14c0b97d TG	431	static int fib_nl_fill_rule(struct sk_buff skb, struct fib_rule rule,
	432	u32 pid, u32 seq, int type, int flags,
	433	struct fib_rules_ops *ops)
	434	{
	435	struct nlmsghdr *nlh;
	436	struct fib_rule_hdr *frh;
	437
	438	nlh = nlmsg_put(skb, pid, seq, type, sizeof(*frh), flags);
	439	if (nlh == NULL)
26932566	440	return -EMSGSIZE;
14c0b97d TG	441
	442	frh = nlmsg_data(nlh);
	443	frh->table = rule->table;
9e762a4a	444	NLA_PUT_U32(skb, FRA_TABLE, rule->table);
14c0b97d TG	445	frh->res1 = 0;
	446	frh->res2 = 0;
	447	frh->action = rule->action;
	448	frh->flags = rule->flags;
	449
0947c9fe TG	450	if (rule->action == FR_ACT_GOTO && rule->ctarget == NULL)
	451	frh->flags \|= FIB_RULE_UNRESOLVED;
	452
14c0b97d TG	453	if (rule->ifname[0])
	454	NLA_PUT_STRING(skb, FRA_IFNAME, rule->ifname);
	455
	456	if (rule->pref)
	457	NLA_PUT_U32(skb, FRA_PRIORITY, rule->pref);
	458
b8964ed9 TG	459	if (rule->mark)
	460	NLA_PUT_U32(skb, FRA_FWMARK, rule->mark);
	461
	462	if (rule->mark_mask \|\| rule->mark)
	463	NLA_PUT_U32(skb, FRA_FWMASK, rule->mark_mask);
	464
0947c9fe TG	465	if (rule->target)
	466	NLA_PUT_U32(skb, FRA_GOTO, rule->target);
	467
14c0b97d TG	468	if (ops->fill(rule, skb, nlh, frh) < 0)
	469	goto nla_put_failure;
	470
	471	return nlmsg_end(skb, nlh);
	472
	473	nla_put_failure:
26932566 PM	474	nlmsg_cancel(skb, nlh);
26932566 PM	475	return -EMSGSIZE;
14c0b97d TG	476	}
14c0b97d TG	477
c454673d TG	478	static int dump_rules(struct sk_buff skb, struct netlink_callback cb,
c454673d TG	479	struct fib_rules_ops *ops)
14c0b97d TG	480	{
	481	int idx = 0;
	482	struct fib_rule *rule;
14c0b97d TG	483
14c0b97d TG	484	rcu_read_lock();
ec25615b	485	list_for_each_entry_rcu(rule, ops->rules_list, list) {
c454673d	486	if (idx < cb->args[1])
14c0b97d TG	487	goto skip;
	488
	489	if (fib_nl_fill_rule(skb, rule, NETLINK_CB(cb->skb).pid,
	490	cb->nlh->nlmsg_seq, RTM_NEWRULE,
	491	NLM_F_MULTI, ops) < 0)
	492	break;
	493	skip:
	494	idx++;
	495	}
	496	rcu_read_unlock();
c454673d	497	cb->args[1] = idx;
14c0b97d TG	498	rules_ops_put(ops);
	499
	500	return skb->len;
	501	}
	502
c454673d TG	503	static int fib_nl_dumprule(struct sk_buff skb, struct netlink_callback cb)
	504	{
	505	struct fib_rules_ops *ops;
	506	int idx = 0, family;
	507
	508	family = rtnl_msg_family(cb->nlh);
	509	if (family != AF_UNSPEC) {
	510	/* Protocol specific dump request */
	511	ops = lookup_rules_ops(family);
	512	if (ops == NULL)
	513	return -EAFNOSUPPORT;
	514
	515	return dump_rules(skb, cb, ops);
	516	}
	517
	518	rcu_read_lock();
	519	list_for_each_entry_rcu(ops, &rules_ops, list) {
	520	if (idx < cb->args[0] \|\| !try_module_get(ops->owner))
	521	goto skip;
	522
	523	if (dump_rules(skb, cb, ops) < 0)
	524	break;
	525
	526	cb->args[1] = 0;
	527	skip:
	528	idx++;
	529	}
	530	rcu_read_unlock();
	531	cb->args[0] = idx;
	532
	533	return skb->len;
	534	}
14c0b97d TG	535
14c0b97d TG	536	static void notify_rule_change(int event, struct fib_rule *rule,
c17084d2 TG	537	struct fib_rules_ops ops, struct nlmsghdr nlh,
c17084d2 TG	538	u32 pid)
14c0b97d	539	{
c17084d2 TG	540	struct sk_buff *skb;
c17084d2 TG	541	int err = -ENOBUFS;
14c0b97d	542
339bf98f	543	skb = nlmsg_new(fib_rule_nlmsg_size(ops, rule), GFP_KERNEL);
14c0b97d	544	if (skb == NULL)
c17084d2 TG	545	goto errout;
	546
	547	err = fib_nl_fill_rule(skb, rule, pid, nlh->nlmsg_seq, event, 0, ops);
26932566 PM	548	if (err < 0) {
	549	/* -EMSGSIZE implies BUG in fib_rule_nlmsg_size() */
	550	WARN_ON(err == -EMSGSIZE);
	551	kfree_skb(skb);
	552	goto errout;
	553	}
c17084d2 TG	554	err = rtnl_notify(skb, pid, ops->nlgroup, nlh, GFP_KERNEL);
	555	errout:
	556	if (err < 0)
	557	rtnl_set_sk_err(ops->nlgroup, err);
14c0b97d TG	558	}
	559
	560	static void attach_rules(struct list_head rules, struct net_device dev)
	561	{
	562	struct fib_rule *rule;
	563
	564	list_for_each_entry(rule, rules, list) {
	565	if (rule->ifindex == -1 &&
	566	strcmp(dev->name, rule->ifname) == 0)
	567	rule->ifindex = dev->ifindex;
	568	}
	569	}
	570
	571	static void detach_rules(struct list_head rules, struct net_device dev)
	572	{
	573	struct fib_rule *rule;
	574
	575	list_for_each_entry(rule, rules, list)
	576	if (rule->ifindex == dev->ifindex)
	577	rule->ifindex = -1;
	578	}
	579
	580
	581	static int fib_rules_event(struct notifier_block *this, unsigned long event,
	582	void *ptr)
	583	{
	584	struct net_device *dev = ptr;
	585	struct fib_rules_ops *ops;
	586
	587	ASSERT_RTNL();
	588	rcu_read_lock();
	589
	590	switch (event) {
	591	case NETDEV_REGISTER:
	592	list_for_each_entry(ops, &rules_ops, list)
	593	attach_rules(ops->rules_list, dev);
	594	break;
	595
	596	case NETDEV_UNREGISTER:
	597	list_for_each_entry(ops, &rules_ops, list)
	598	detach_rules(ops->rules_list, dev);
	599	break;
	600	}
	601
	602	rcu_read_unlock();
	603
	604	return NOTIFY_DONE;
	605	}
	606
	607	static struct notifier_block fib_rules_notifier = {
	608	.notifier_call = fib_rules_event,
	609	};
	610
	611	static int __init fib_rules_init(void)
	612	{
9d9e6a58 TG	613	rtnl_register(PF_UNSPEC, RTM_NEWRULE, fib_nl_newrule, NULL);
9d9e6a58 TG	614	rtnl_register(PF_UNSPEC, RTM_DELRULE, fib_nl_delrule, NULL);
c454673d	615	rtnl_register(PF_UNSPEC, RTM_GETRULE, NULL, fib_nl_dumprule);
9d9e6a58	616
14c0b97d TG	617	return register_netdevice_notifier(&fib_rules_notifier);
	618	}
	619
	620	subsys_initcall(fib_rules_init);