[deliverable/linux.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/net_namespace.h>
#include <net/sock.h>
#include <net/fib_rules.h>

int fib_default_rule_add(struct fib_rules_ops *ops,
			 u32 pref, u32 table, u32 flags)
{
	struct fib_rule *r;

	r = kzalloc(ops->rule_size, GFP_KERNEL);
	if (r == NULL)
		return -ENOMEM;

	atomic_set(&r->refcnt, 1);
	r->action = FR_ACT_TO_TBL;
	r->pref = pref;
	r->table = table;
	r->flags = flags;

	/* The lock is not required here, the list in unreacheable
	 * at the moment this function is called */
	list_add_tail(&r->list, &ops->rules_list);
	return 0;
}
EXPORT_SYMBOL(fib_default_rule_add);

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(struct net *net, int family)
{
	struct fib_rules_ops *ops;

	rcu_read_lock();
	list_for_each_entry_rcu(ops, &net->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);
}

static void flush_route_cache(struct fib_rules_ops *ops)
{
	if (ops->flush_cache)
		ops->flush_cache();
}

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

	net = ops->fro_net;

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

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

	return err;
}

EXPORT_SYMBOL_GPL(fib_rules_register);

void fib_rules_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);
	}
}
EXPORT_SYMBOL_GPL(fib_rules_cleanup_ops);

void fib_rules_unregister(struct fib_rules_ops *ops)
{
	struct net *net = ops->fro_net;

	spin_lock(&net->rules_mod_lock);
	list_del_rcu(&ops->list);
	fib_rules_cleanup_ops(ops);
	spin_unlock(&net->rules_mod_lock);

	synchronize_rcu();
	release_net(net);
}

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 if (rule->action == FR_ACT_NOP)
			continue;
		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 net *net = skb->sk->sk_net;
	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(net, 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(net, 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(ops);

	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);
	flush_route_cache(ops);
	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 net *net = skb->sk->sk_net;
	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(net, 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);
		flush_route_cache(ops);
		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->ifindex == -1)
			frh->flags |= FIB_RULE_DEV_DETACHED;
	}

	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;

	list_for_each_entry(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++;
	}
	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 net *net = skb->sk->sk_net;
	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(net, family);
		if (ops == NULL)
			return -EAFNOSUPPORT;

		return dump_rules(skb, cb, ops);
	}

	rcu_read_lock();
	list_for_each_entry_rcu(ops, &net->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 net *net;
	struct sk_buff *skb;
	int err = -ENOBUFS;

	net = ops->fro_net;
	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, net, pid, ops->nlgroup, nlh, GFP_KERNEL);
errout:
	if (err < 0)
		rtnl_set_sk_err(net, 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 net *net = dev->nd_net;
	struct fib_rules_ops *ops;

	ASSERT_RTNL();
	rcu_read_lock();

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

	case NETDEV_UNREGISTER:
		list_for_each_entry(ops, &net->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 fib_rules_net_init(struct net *net)
{
	INIT_LIST_HEAD(&net->rules_ops);
	spin_lock_init(&net->rules_mod_lock);
	return 0;
}

static struct pernet_operations fib_rules_net_ops = {
	.init = fib_rules_net_init,
};

static int __init fib_rules_init(void)
{
	int err;
	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);

	err = register_netdevice_notifier(&fib_rules_notifier);
	if (err < 0)
		goto fail;

	err = register_pernet_subsys(&fib_rules_net_ops);
	if (err < 0)
		goto fail_unregister;
	return 0;

fail_unregister:
	unregister_netdevice_notifier(&fib_rules_notifier);
fail:
	rtnl_unregister(PF_UNSPEC, RTM_NEWRULE);
	rtnl_unregister(PF_UNSPEC, RTM_DELRULE);
	rtnl_unregister(PF_UNSPEC, RTM_GETRULE);
	return err;
}

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>
e9dc8653	14	#include <net/net_namespace.h>
881d966b	15	#include <net/sock.h>
14c0b97d TG	16	#include <net/fib_rules.h>
14c0b97d TG	17
2994c638 DL	18	int fib_default_rule_add(struct fib_rules_ops *ops,
	19	u32 pref, u32 table, u32 flags)
	20	{
	21	struct fib_rule *r;
	22
	23	r = kzalloc(ops->rule_size, GFP_KERNEL);
	24	if (r == NULL)
	25	return -ENOMEM;
	26
	27	atomic_set(&r->refcnt, 1);
	28	r->action = FR_ACT_TO_TBL;
	29	r->pref = pref;
	30	r->table = table;
	31	r->flags = flags;
	32
	33	/* The lock is not required here, the list in unreacheable
	34	* at the moment this function is called */
	35	list_add_tail(&r->list, &ops->rules_list);
	36	return 0;
	37	}
	38	EXPORT_SYMBOL(fib_default_rule_add);
	39
9e3a5487	40	static void notify_rule_change(int event, struct fib_rule *rule,
c17084d2 TG	41	struct fib_rules_ops ops, struct nlmsghdr nlh,
c17084d2 TG	42	u32 pid);
14c0b97d	43
5fd30ee7	44	static struct fib_rules_ops lookup_rules_ops(struct net net, int family)
14c0b97d TG	45	{
	46	struct fib_rules_ops *ops;
	47
	48	rcu_read_lock();
5fd30ee7	49	list_for_each_entry_rcu(ops, &net->rules_ops, list) {
14c0b97d TG	50	if (ops->family == family) {
	51	if (!try_module_get(ops->owner))
	52	ops = NULL;
	53	rcu_read_unlock();
	54	return ops;
	55	}
	56	}
	57	rcu_read_unlock();
	58
	59	return NULL;
	60	}
	61
	62	static void rules_ops_put(struct fib_rules_ops *ops)
	63	{
	64	if (ops)
	65	module_put(ops->owner);
	66	}
	67
73417f61 TG	68	static void flush_route_cache(struct fib_rules_ops *ops)
	69	{
	70	if (ops->flush_cache)
	71	ops->flush_cache();
	72	}
	73
9e3a5487	74	int fib_rules_register(struct fib_rules_ops *ops)
14c0b97d TG	75	{
	76	int err = -EEXIST;
	77	struct fib_rules_ops *o;
9e3a5487 DL	78	struct net *net;
	79
	80	net = ops->fro_net;
14c0b97d TG	81
	82	if (ops->rule_size < sizeof(struct fib_rule))
	83	return -EINVAL;
	84
	85	if (ops->match == NULL \|\| ops->configure == NULL \|\|
	86	ops->compare == NULL \|\| ops->fill == NULL \|\|
	87	ops->action == NULL)
	88	return -EINVAL;
	89
5fd30ee7 DL	90	spin_lock(&net->rules_mod_lock);
5fd30ee7 DL	91	list_for_each_entry(o, &net->rules_ops, list)
14c0b97d TG	92	if (ops->family == o->family)
	93	goto errout;
	94
5fd30ee7 DL	95	hold_net(net);
5fd30ee7 DL	96	list_add_tail_rcu(&ops->list, &net->rules_ops);
14c0b97d TG	97	err = 0;
14c0b97d TG	98	errout:
5fd30ee7	99	spin_unlock(&net->rules_mod_lock);
14c0b97d TG	100
	101	return err;
	102	}
	103
	104	EXPORT_SYMBOL_GPL(fib_rules_register);
	105
9eb87f3f	106	void fib_rules_cleanup_ops(struct fib_rules_ops *ops)
14c0b97d TG	107	{
	108	struct fib_rule rule, tmp;
	109
76c72d4f	110	list_for_each_entry_safe(rule, tmp, &ops->rules_list, list) {
14c0b97d TG	111	list_del_rcu(&rule->list);
	112	fib_rule_put(rule);
	113	}
	114	}
9eb87f3f	115	EXPORT_SYMBOL_GPL(fib_rules_cleanup_ops);
14c0b97d	116
9e3a5487	117	void fib_rules_unregister(struct fib_rules_ops *ops)
14c0b97d	118	{
9e3a5487	119	struct net *net = ops->fro_net;
14c0b97d	120
5fd30ee7	121	spin_lock(&net->rules_mod_lock);
72132c1b DL	122	list_del_rcu(&ops->list);
72132c1b DL	123	fib_rules_cleanup_ops(ops);
5fd30ee7	124	spin_unlock(&net->rules_mod_lock);
14c0b97d TG	125
14c0b97d TG	126	synchronize_rcu();
72132c1b	127	release_net(net);
14c0b97d TG	128	}
	129
	130	EXPORT_SYMBOL_GPL(fib_rules_unregister);
	131
3dfbcc41 TG	132	static int fib_rule_match(struct fib_rule rule, struct fib_rules_ops ops,
	133	struct flowi *fl, int flags)
	134	{
	135	int ret = 0;
	136
	137	if (rule->ifindex && (rule->ifindex != fl->iif))
	138	goto out;
	139
	140	if ((rule->mark ^ fl->mark) & rule->mark_mask)
	141	goto out;
	142
	143	ret = ops->match(rule, fl, flags);
	144	out:
	145	return (rule->flags & FIB_RULE_INVERT) ? !ret : ret;
	146	}
	147
14c0b97d TG	148	int fib_rules_lookup(struct fib_rules_ops ops, struct flowi fl,
	149	int flags, struct fib_lookup_arg *arg)
	150	{
	151	struct fib_rule *rule;
	152	int err;
	153
	154	rcu_read_lock();
	155
76c72d4f	156	list_for_each_entry_rcu(rule, &ops->rules_list, list) {
0947c9fe	157	jumped:
3dfbcc41	158	if (!fib_rule_match(rule, ops, fl, flags))
14c0b97d TG	159	continue;
14c0b97d TG	160
0947c9fe TG	161	if (rule->action == FR_ACT_GOTO) {
	162	struct fib_rule *target;
	163
	164	target = rcu_dereference(rule->ctarget);
	165	if (target == NULL) {
	166	continue;
	167	} else {
	168	rule = target;
	169	goto jumped;
	170	}
fa0b2d1d TG	171	} else if (rule->action == FR_ACT_NOP)
	172	continue;
	173	else
0947c9fe TG	174	err = ops->action(rule, fl, flags, arg);
0947c9fe TG	175
14c0b97d TG	176	if (err != -EAGAIN) {
	177	fib_rule_get(rule);
	178	arg->rule = rule;
	179	goto out;
	180	}
	181	}
	182
83886b6b	183	err = -ESRCH;
14c0b97d TG	184	out:
	185	rcu_read_unlock();
	186
	187	return err;
	188	}
	189
	190	EXPORT_SYMBOL_GPL(fib_rules_lookup);
	191
e1701c68 TG	192	static int validate_rulemsg(struct fib_rule_hdr frh, struct nlattr *tb,
	193	struct fib_rules_ops *ops)
	194	{
	195	int err = -EINVAL;
	196
	197	if (frh->src_len)
	198	if (tb[FRA_SRC] == NULL \|\|
	199	frh->src_len > (ops->addr_size * 8) \|\|
	200	nla_len(tb[FRA_SRC]) != ops->addr_size)
	201	goto errout;
	202
	203	if (frh->dst_len)
	204	if (tb[FRA_DST] == NULL \|\|
	205	frh->dst_len > (ops->addr_size * 8) \|\|
	206	nla_len(tb[FRA_DST]) != ops->addr_size)
	207	goto errout;
	208
	209	err = 0;
	210	errout:
	211	return err;
	212	}
	213
9d9e6a58	214	static int fib_nl_newrule(struct sk_buff skb, struct nlmsghdr nlh, void *arg)
14c0b97d	215	{
881d966b	216	struct net *net = skb->sk->sk_net;
14c0b97d TG	217	struct fib_rule_hdr *frh = nlmsg_data(nlh);
	218	struct fib_rules_ops *ops = NULL;
	219	struct fib_rule rule, r, *last = NULL;
	220	struct nlattr *tb[FRA_MAX+1];
0947c9fe	221	int err = -EINVAL, unresolved = 0;
14c0b97d TG	222
	223	if (nlh->nlmsg_len < nlmsg_msg_size(sizeof(*frh)))
	224	goto errout;
	225
5fd30ee7	226	ops = lookup_rules_ops(net, frh->family);
14c0b97d TG	227	if (ops == NULL) {
	228	err = EAFNOSUPPORT;
	229	goto errout;
	230	}
	231
	232	err = nlmsg_parse(nlh, sizeof(*frh), tb, FRA_MAX, ops->policy);
	233	if (err < 0)
	234	goto errout;
	235
e1701c68 TG	236	err = validate_rulemsg(frh, tb, ops);
	237	if (err < 0)
	238	goto errout;
	239
14c0b97d TG	240	rule = kzalloc(ops->rule_size, GFP_KERNEL);
	241	if (rule == NULL) {
	242	err = -ENOMEM;
	243	goto errout;
	244	}
	245
	246	if (tb[FRA_PRIORITY])
	247	rule->pref = nla_get_u32(tb[FRA_PRIORITY]);
	248
	249	if (tb[FRA_IFNAME]) {
	250	struct net_device *dev;
	251
	252	rule->ifindex = -1;
5176f91e	253	nla_strlcpy(rule->ifname, tb[FRA_IFNAME], IFNAMSIZ);
881d966b	254	dev = __dev_get_by_name(net, rule->ifname);
14c0b97d TG	255	if (dev)
	256	rule->ifindex = dev->ifindex;
	257	}
	258
b8964ed9 TG	259	if (tb[FRA_FWMARK]) {
	260	rule->mark = nla_get_u32(tb[FRA_FWMARK]);
	261	if (rule->mark)
	262	/* compatibility: if the mark value is non-zero all bits
	263	* are compared unless a mask is explicitly specified.
	264	*/
	265	rule->mark_mask = 0xFFFFFFFF;
	266	}
	267
	268	if (tb[FRA_FWMASK])
	269	rule->mark_mask = nla_get_u32(tb[FRA_FWMASK]);
	270
14c0b97d TG	271	rule->action = frh->action;
14c0b97d TG	272	rule->flags = frh->flags;
9e762a4a	273	rule->table = frh_get_table(frh, tb);
14c0b97d TG	274
14c0b97d TG	275	if (!rule->pref && ops->default_pref)
868d13ac	276	rule->pref = ops->default_pref(ops);
14c0b97d	277
0947c9fe TG	278	err = -EINVAL;
	279	if (tb[FRA_GOTO]) {
	280	if (rule->action != FR_ACT_GOTO)
	281	goto errout_free;
	282
	283	rule->target = nla_get_u32(tb[FRA_GOTO]);
	284	/* Backward jumps are prohibited to avoid endless loops */
	285	if (rule->target <= rule->pref)
	286	goto errout_free;
	287
76c72d4f	288	list_for_each_entry(r, &ops->rules_list, list) {
0947c9fe TG	289	if (r->pref == rule->target) {
	290	rule->ctarget = r;
	291	break;
	292	}
	293	}
	294
	295	if (rule->ctarget == NULL)
	296	unresolved = 1;
	297	} else if (rule->action == FR_ACT_GOTO)
	298	goto errout_free;
	299
14c0b97d TG	300	err = ops->configure(rule, skb, nlh, frh, tb);
	301	if (err < 0)
	302	goto errout_free;
	303
76c72d4f	304	list_for_each_entry(r, &ops->rules_list, list) {
14c0b97d TG	305	if (r->pref > rule->pref)
	306	break;
	307	last = r;
	308	}
	309
	310	fib_rule_get(rule);
	311
0947c9fe TG	312	if (ops->unresolved_rules) {
	313	/*
	314	* There are unresolved goto rules in the list, check if
	315	* any of them are pointing to this new rule.
	316	*/
76c72d4f	317	list_for_each_entry(r, &ops->rules_list, list) {
0947c9fe TG	318	if (r->action == FR_ACT_GOTO &&
	319	r->target == rule->pref) {
	320	BUG_ON(r->ctarget != NULL);
	321	rcu_assign_pointer(r->ctarget, rule);
	322	if (--ops->unresolved_rules == 0)
	323	break;
	324	}
	325	}
	326	}
	327
	328	if (rule->action == FR_ACT_GOTO)
	329	ops->nr_goto_rules++;
	330
	331	if (unresolved)
	332	ops->unresolved_rules++;
	333
14c0b97d TG	334	if (last)
	335	list_add_rcu(&rule->list, &last->list);
	336	else
76c72d4f	337	list_add_rcu(&rule->list, &ops->rules_list);
14c0b97d	338
9e3a5487	339	notify_rule_change(RTM_NEWRULE, rule, ops, nlh, NETLINK_CB(skb).pid);
73417f61	340	flush_route_cache(ops);
14c0b97d TG	341	rules_ops_put(ops);
	342	return 0;
	343
	344	errout_free:
	345	kfree(rule);
	346	errout:
	347	rules_ops_put(ops);
	348	return err;
	349	}
	350
9d9e6a58	351	static int fib_nl_delrule(struct sk_buff skb, struct nlmsghdr nlh, void *arg)
14c0b97d	352	{
b854272b	353	struct net *net = skb->sk->sk_net;
14c0b97d TG	354	struct fib_rule_hdr *frh = nlmsg_data(nlh);
14c0b97d TG	355	struct fib_rules_ops *ops = NULL;
0947c9fe	356	struct fib_rule rule, tmp;
14c0b97d TG	357	struct nlattr *tb[FRA_MAX+1];
	358	int err = -EINVAL;
	359
	360	if (nlh->nlmsg_len < nlmsg_msg_size(sizeof(*frh)))
	361	goto errout;
	362
5fd30ee7	363	ops = lookup_rules_ops(net, frh->family);
14c0b97d TG	364	if (ops == NULL) {
	365	err = EAFNOSUPPORT;
	366	goto errout;
	367	}
	368
	369	err = nlmsg_parse(nlh, sizeof(*frh), tb, FRA_MAX, ops->policy);
	370	if (err < 0)
	371	goto errout;
	372
e1701c68 TG	373	err = validate_rulemsg(frh, tb, ops);
	374	if (err < 0)
	375	goto errout;
	376
76c72d4f	377	list_for_each_entry(rule, &ops->rules_list, list) {
14c0b97d TG	378	if (frh->action && (frh->action != rule->action))
	379	continue;
	380
9e762a4a	381	if (frh->table && (frh_get_table(frh, tb) != rule->table))
14c0b97d TG	382	continue;
	383
	384	if (tb[FRA_PRIORITY] &&
	385	(rule->pref != nla_get_u32(tb[FRA_PRIORITY])))
	386	continue;
	387
	388	if (tb[FRA_IFNAME] &&
	389	nla_strcmp(tb[FRA_IFNAME], rule->ifname))
	390	continue;
	391
b8964ed9 TG	392	if (tb[FRA_FWMARK] &&
	393	(rule->mark != nla_get_u32(tb[FRA_FWMARK])))
	394	continue;
	395
	396	if (tb[FRA_FWMASK] &&
	397	(rule->mark_mask != nla_get_u32(tb[FRA_FWMASK])))
	398	continue;
	399
14c0b97d TG	400	if (!ops->compare(rule, frh, tb))
	401	continue;
	402
	403	if (rule->flags & FIB_RULE_PERMANENT) {
	404	err = -EPERM;
	405	goto errout;
	406	}
	407
	408	list_del_rcu(&rule->list);
0947c9fe TG	409
	410	if (rule->action == FR_ACT_GOTO)
	411	ops->nr_goto_rules--;
	412
	413	/*
	414	* Check if this rule is a target to any of them. If so,
	415	* disable them. As this operation is eventually very
	416	* expensive, it is only performed if goto rules have
	417	* actually been added.
	418	*/
	419	if (ops->nr_goto_rules > 0) {
76c72d4f	420	list_for_each_entry(tmp, &ops->rules_list, list) {
0947c9fe TG	421	if (tmp->ctarget == rule) {
	422	rcu_assign_pointer(tmp->ctarget, NULL);
	423	ops->unresolved_rules++;
	424	}
	425	}
	426	}
	427
14c0b97d	428	synchronize_rcu();
9e3a5487	429	notify_rule_change(RTM_DELRULE, rule, ops, nlh,
c17084d2	430	NETLINK_CB(skb).pid);
14c0b97d	431	fib_rule_put(rule);
73417f61	432	flush_route_cache(ops);
14c0b97d TG	433	rules_ops_put(ops);
	434	return 0;
	435	}
	436
	437	err = -ENOENT;
	438	errout:
	439	rules_ops_put(ops);
	440	return err;
	441	}
	442
339bf98f TG	443	static inline size_t fib_rule_nlmsg_size(struct fib_rules_ops *ops,
	444	struct fib_rule *rule)
	445	{
	446	size_t payload = NLMSG_ALIGN(sizeof(struct fib_rule_hdr))
	447	+ nla_total_size(IFNAMSIZ) /* FRA_IFNAME */
	448	+ nla_total_size(4) /* FRA_PRIORITY */
	449	+ nla_total_size(4) /* FRA_TABLE */
	450	+ nla_total_size(4) /* FRA_FWMARK */
	451	+ nla_total_size(4); /* FRA_FWMASK */
	452
	453	if (ops->nlmsg_payload)
	454	payload += ops->nlmsg_payload(rule);
	455
	456	return payload;
	457	}
	458
14c0b97d TG	459	static int fib_nl_fill_rule(struct sk_buff skb, struct fib_rule rule,
	460	u32 pid, u32 seq, int type, int flags,
	461	struct fib_rules_ops *ops)
	462	{
	463	struct nlmsghdr *nlh;
	464	struct fib_rule_hdr *frh;
	465
	466	nlh = nlmsg_put(skb, pid, seq, type, sizeof(*frh), flags);
	467	if (nlh == NULL)
26932566	468	return -EMSGSIZE;
14c0b97d TG	469
	470	frh = nlmsg_data(nlh);
	471	frh->table = rule->table;
9e762a4a	472	NLA_PUT_U32(skb, FRA_TABLE, rule->table);
14c0b97d TG	473	frh->res1 = 0;
	474	frh->res2 = 0;
	475	frh->action = rule->action;
	476	frh->flags = rule->flags;
	477
0947c9fe TG	478	if (rule->action == FR_ACT_GOTO && rule->ctarget == NULL)
	479	frh->flags \|= FIB_RULE_UNRESOLVED;
	480
2b443683	481	if (rule->ifname[0]) {
14c0b97d TG	482	NLA_PUT_STRING(skb, FRA_IFNAME, rule->ifname);
14c0b97d TG	483
2b443683 TG	484	if (rule->ifindex == -1)
	485	frh->flags \|= FIB_RULE_DEV_DETACHED;
	486	}
	487
14c0b97d TG	488	if (rule->pref)
	489	NLA_PUT_U32(skb, FRA_PRIORITY, rule->pref);
	490
b8964ed9 TG	491	if (rule->mark)
	492	NLA_PUT_U32(skb, FRA_FWMARK, rule->mark);
	493
	494	if (rule->mark_mask \|\| rule->mark)
	495	NLA_PUT_U32(skb, FRA_FWMASK, rule->mark_mask);
	496
0947c9fe TG	497	if (rule->target)
	498	NLA_PUT_U32(skb, FRA_GOTO, rule->target);
	499
14c0b97d TG	500	if (ops->fill(rule, skb, nlh, frh) < 0)
	501	goto nla_put_failure;
	502
	503	return nlmsg_end(skb, nlh);
	504
	505	nla_put_failure:
26932566 PM	506	nlmsg_cancel(skb, nlh);
26932566 PM	507	return -EMSGSIZE;
14c0b97d TG	508	}
14c0b97d TG	509
c454673d TG	510	static int dump_rules(struct sk_buff skb, struct netlink_callback cb,
c454673d TG	511	struct fib_rules_ops *ops)
14c0b97d TG	512	{
	513	int idx = 0;
	514	struct fib_rule *rule;
14c0b97d	515
76c72d4f	516	list_for_each_entry(rule, &ops->rules_list, list) {
c454673d	517	if (idx < cb->args[1])
14c0b97d TG	518	goto skip;
	519
	520	if (fib_nl_fill_rule(skb, rule, NETLINK_CB(cb->skb).pid,
	521	cb->nlh->nlmsg_seq, RTM_NEWRULE,
	522	NLM_F_MULTI, ops) < 0)
	523	break;
	524	skip:
	525	idx++;
	526	}
c454673d	527	cb->args[1] = idx;
14c0b97d TG	528	rules_ops_put(ops);
	529
	530	return skb->len;
	531	}
	532
c454673d TG	533	static int fib_nl_dumprule(struct sk_buff skb, struct netlink_callback cb)
c454673d TG	534	{
b854272b	535	struct net *net = skb->sk->sk_net;
c454673d TG	536	struct fib_rules_ops *ops;
	537	int idx = 0, family;
	538
	539	family = rtnl_msg_family(cb->nlh);
	540	if (family != AF_UNSPEC) {
	541	/* Protocol specific dump request */
5fd30ee7	542	ops = lookup_rules_ops(net, family);
c454673d TG	543	if (ops == NULL)
	544	return -EAFNOSUPPORT;
	545
	546	return dump_rules(skb, cb, ops);
	547	}
	548
	549	rcu_read_lock();
5fd30ee7	550	list_for_each_entry_rcu(ops, &net->rules_ops, list) {
c454673d TG	551	if (idx < cb->args[0] \|\| !try_module_get(ops->owner))
	552	goto skip;
	553
	554	if (dump_rules(skb, cb, ops) < 0)
	555	break;
	556
	557	cb->args[1] = 0;
	558	skip:
	559	idx++;
	560	}
	561	rcu_read_unlock();
	562	cb->args[0] = idx;
	563
	564	return skb->len;
	565	}
14c0b97d	566
9e3a5487	567	static void notify_rule_change(int event, struct fib_rule *rule,
c17084d2 TG	568	struct fib_rules_ops ops, struct nlmsghdr nlh,
c17084d2 TG	569	u32 pid)
14c0b97d	570	{
9e3a5487	571	struct net *net;
c17084d2 TG	572	struct sk_buff *skb;
c17084d2 TG	573	int err = -ENOBUFS;
14c0b97d	574
9e3a5487	575	net = ops->fro_net;
339bf98f	576	skb = nlmsg_new(fib_rule_nlmsg_size(ops, rule), GFP_KERNEL);
14c0b97d	577	if (skb == NULL)
c17084d2 TG	578	goto errout;
	579
	580	err = fib_nl_fill_rule(skb, rule, pid, nlh->nlmsg_seq, event, 0, ops);
26932566 PM	581	if (err < 0) {
	582	/* -EMSGSIZE implies BUG in fib_rule_nlmsg_size() */
	583	WARN_ON(err == -EMSGSIZE);
	584	kfree_skb(skb);
	585	goto errout;
	586	}
9e3a5487	587
5fd30ee7	588	err = rtnl_notify(skb, net, pid, ops->nlgroup, nlh, GFP_KERNEL);
c17084d2 TG	589	errout:
c17084d2 TG	590	if (err < 0)
5fd30ee7	591	rtnl_set_sk_err(net, ops->nlgroup, err);
14c0b97d TG	592	}
	593
	594	static void attach_rules(struct list_head rules, struct net_device dev)
	595	{
	596	struct fib_rule *rule;
	597
	598	list_for_each_entry(rule, rules, list) {
	599	if (rule->ifindex == -1 &&
	600	strcmp(dev->name, rule->ifname) == 0)
	601	rule->ifindex = dev->ifindex;
	602	}
	603	}
	604
	605	static void detach_rules(struct list_head rules, struct net_device dev)
	606	{
	607	struct fib_rule *rule;
	608
	609	list_for_each_entry(rule, rules, list)
	610	if (rule->ifindex == dev->ifindex)
	611	rule->ifindex = -1;
	612	}
	613
	614
	615	static int fib_rules_event(struct notifier_block *this, unsigned long event,
	616	void *ptr)
	617	{
	618	struct net_device *dev = ptr;
5fd30ee7	619	struct net *net = dev->nd_net;
14c0b97d TG	620	struct fib_rules_ops *ops;
	621
	622	ASSERT_RTNL();
	623	rcu_read_lock();
	624
	625	switch (event) {
	626	case NETDEV_REGISTER:
5fd30ee7	627	list_for_each_entry(ops, &net->rules_ops, list)
76c72d4f	628	attach_rules(&ops->rules_list, dev);
14c0b97d TG	629	break;
	630
	631	case NETDEV_UNREGISTER:
5fd30ee7	632	list_for_each_entry(ops, &net->rules_ops, list)
76c72d4f	633	detach_rules(&ops->rules_list, dev);
14c0b97d TG	634	break;
	635	}
	636
	637	rcu_read_unlock();
	638
	639	return NOTIFY_DONE;
	640	}
	641
	642	static struct notifier_block fib_rules_notifier = {
	643	.notifier_call = fib_rules_event,
	644	};
	645
5fd30ee7 DL	646	static int fib_rules_net_init(struct net *net)
	647	{
	648	INIT_LIST_HEAD(&net->rules_ops);
	649	spin_lock_init(&net->rules_mod_lock);
	650	return 0;
	651	}
	652
	653	static struct pernet_operations fib_rules_net_ops = {
	654	.init = fib_rules_net_init,
	655	};
	656
14c0b97d TG	657	static int __init fib_rules_init(void)
14c0b97d TG	658	{
5fd30ee7	659	int err;
9d9e6a58 TG	660	rtnl_register(PF_UNSPEC, RTM_NEWRULE, fib_nl_newrule, NULL);
9d9e6a58 TG	661	rtnl_register(PF_UNSPEC, RTM_DELRULE, fib_nl_delrule, NULL);
c454673d	662	rtnl_register(PF_UNSPEC, RTM_GETRULE, NULL, fib_nl_dumprule);
9d9e6a58	663
5fd30ee7 DL	664	err = register_netdevice_notifier(&fib_rules_notifier);
	665	if (err < 0)
	666	goto fail;
	667
	668	err = register_pernet_subsys(&fib_rules_net_ops);
	669	if (err < 0)
	670	goto fail_unregister;
	671	return 0;
	672
	673	fail_unregister:
	674	unregister_netdevice_notifier(&fib_rules_notifier);
	675	fail:
	676	rtnl_unregister(PF_UNSPEC, RTM_NEWRULE);
	677	rtnl_unregister(PF_UNSPEC, RTM_DELRULE);
	678	rtnl_unregister(PF_UNSPEC, RTM_GETRULE);
	679	return err;
14c0b97d TG	680	}
	681
	682	subsys_initcall(fib_rules_init);