[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/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;
	r->fr_net = ops->fro_net;

	/* 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 = sock_net(skb->sk);
	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;
	}
	rule->fr_net = net;

	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 = sock_net(skb->sk);
	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 = sock_net(skb->sk);
	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_net(dev);
	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;
51314a17	32	r->fr_net = ops->fro_net;
2994c638 DL	33
	34	/* The lock is not required here, the list in unreacheable
	35	* at the moment this function is called */
	36	list_add_tail(&r->list, &ops->rules_list);
	37	return 0;
	38	}
	39	EXPORT_SYMBOL(fib_default_rule_add);
	40
9e3a5487	41	static void notify_rule_change(int event, struct fib_rule *rule,
c17084d2 TG	42	struct fib_rules_ops ops, struct nlmsghdr nlh,
c17084d2 TG	43	u32 pid);
14c0b97d	44
5fd30ee7	45	static struct fib_rules_ops lookup_rules_ops(struct net net, int family)
14c0b97d TG	46	{
	47	struct fib_rules_ops *ops;
	48
	49	rcu_read_lock();
5fd30ee7	50	list_for_each_entry_rcu(ops, &net->rules_ops, list) {
14c0b97d TG	51	if (ops->family == family) {
	52	if (!try_module_get(ops->owner))
	53	ops = NULL;
	54	rcu_read_unlock();
	55	return ops;
	56	}
	57	}
	58	rcu_read_unlock();
	59
	60	return NULL;
	61	}
	62
	63	static void rules_ops_put(struct fib_rules_ops *ops)
	64	{
	65	if (ops)
	66	module_put(ops->owner);
	67	}
	68
73417f61 TG	69	static void flush_route_cache(struct fib_rules_ops *ops)
	70	{
	71	if (ops->flush_cache)
	72	ops->flush_cache();
	73	}
	74
9e3a5487	75	int fib_rules_register(struct fib_rules_ops *ops)
14c0b97d TG	76	{
	77	int err = -EEXIST;
	78	struct fib_rules_ops *o;
9e3a5487 DL	79	struct net *net;
	80
	81	net = ops->fro_net;
14c0b97d TG	82
	83	if (ops->rule_size < sizeof(struct fib_rule))
	84	return -EINVAL;
	85
	86	if (ops->match == NULL \|\| ops->configure == NULL \|\|
	87	ops->compare == NULL \|\| ops->fill == NULL \|\|
	88	ops->action == NULL)
	89	return -EINVAL;
	90
5fd30ee7 DL	91	spin_lock(&net->rules_mod_lock);
5fd30ee7 DL	92	list_for_each_entry(o, &net->rules_ops, list)
14c0b97d TG	93	if (ops->family == o->family)
	94	goto errout;
	95
5fd30ee7 DL	96	hold_net(net);
5fd30ee7 DL	97	list_add_tail_rcu(&ops->list, &net->rules_ops);
14c0b97d TG	98	err = 0;
14c0b97d TG	99	errout:
5fd30ee7	100	spin_unlock(&net->rules_mod_lock);
14c0b97d TG	101
	102	return err;
	103	}
	104
	105	EXPORT_SYMBOL_GPL(fib_rules_register);
	106
9eb87f3f	107	void fib_rules_cleanup_ops(struct fib_rules_ops *ops)
14c0b97d TG	108	{
	109	struct fib_rule rule, tmp;
	110
76c72d4f	111	list_for_each_entry_safe(rule, tmp, &ops->rules_list, list) {
14c0b97d TG	112	list_del_rcu(&rule->list);
	113	fib_rule_put(rule);
	114	}
	115	}
9eb87f3f	116	EXPORT_SYMBOL_GPL(fib_rules_cleanup_ops);
14c0b97d	117
9e3a5487	118	void fib_rules_unregister(struct fib_rules_ops *ops)
14c0b97d	119	{
9e3a5487	120	struct net *net = ops->fro_net;
14c0b97d	121
5fd30ee7	122	spin_lock(&net->rules_mod_lock);
72132c1b DL	123	list_del_rcu(&ops->list);
72132c1b DL	124	fib_rules_cleanup_ops(ops);
5fd30ee7	125	spin_unlock(&net->rules_mod_lock);
14c0b97d TG	126
14c0b97d TG	127	synchronize_rcu();
72132c1b	128	release_net(net);
14c0b97d TG	129	}
	130
	131	EXPORT_SYMBOL_GPL(fib_rules_unregister);
	132
3dfbcc41 TG	133	static int fib_rule_match(struct fib_rule rule, struct fib_rules_ops ops,
	134	struct flowi *fl, int flags)
	135	{
	136	int ret = 0;
	137
	138	if (rule->ifindex && (rule->ifindex != fl->iif))
	139	goto out;
	140
	141	if ((rule->mark ^ fl->mark) & rule->mark_mask)
	142	goto out;
	143
	144	ret = ops->match(rule, fl, flags);
	145	out:
	146	return (rule->flags & FIB_RULE_INVERT) ? !ret : ret;
	147	}
	148
14c0b97d TG	149	int fib_rules_lookup(struct fib_rules_ops ops, struct flowi fl,
	150	int flags, struct fib_lookup_arg *arg)
	151	{
	152	struct fib_rule *rule;
	153	int err;
	154
	155	rcu_read_lock();
	156
76c72d4f	157	list_for_each_entry_rcu(rule, &ops->rules_list, list) {
0947c9fe	158	jumped:
3dfbcc41	159	if (!fib_rule_match(rule, ops, fl, flags))
14c0b97d TG	160	continue;
14c0b97d TG	161
0947c9fe TG	162	if (rule->action == FR_ACT_GOTO) {
	163	struct fib_rule *target;
	164
	165	target = rcu_dereference(rule->ctarget);
	166	if (target == NULL) {
	167	continue;
	168	} else {
	169	rule = target;
	170	goto jumped;
	171	}
fa0b2d1d TG	172	} else if (rule->action == FR_ACT_NOP)
	173	continue;
	174	else
0947c9fe TG	175	err = ops->action(rule, fl, flags, arg);
0947c9fe TG	176
14c0b97d TG	177	if (err != -EAGAIN) {
	178	fib_rule_get(rule);
	179	arg->rule = rule;
	180	goto out;
	181	}
	182	}
	183
83886b6b	184	err = -ESRCH;
14c0b97d TG	185	out:
	186	rcu_read_unlock();
	187
	188	return err;
	189	}
	190
	191	EXPORT_SYMBOL_GPL(fib_rules_lookup);
	192
e1701c68 TG	193	static int validate_rulemsg(struct fib_rule_hdr frh, struct nlattr *tb,
	194	struct fib_rules_ops *ops)
	195	{
	196	int err = -EINVAL;
	197
	198	if (frh->src_len)
	199	if (tb[FRA_SRC] == NULL \|\|
	200	frh->src_len > (ops->addr_size * 8) \|\|
	201	nla_len(tb[FRA_SRC]) != ops->addr_size)
	202	goto errout;
	203
	204	if (frh->dst_len)
	205	if (tb[FRA_DST] == NULL \|\|
	206	frh->dst_len > (ops->addr_size * 8) \|\|
	207	nla_len(tb[FRA_DST]) != ops->addr_size)
	208	goto errout;
	209
	210	err = 0;
	211	errout:
	212	return err;
	213	}
	214
9d9e6a58	215	static int fib_nl_newrule(struct sk_buff skb, struct nlmsghdr nlh, void *arg)
14c0b97d	216	{
3b1e0a65	217	struct net *net = sock_net(skb->sk);
14c0b97d TG	218	struct fib_rule_hdr *frh = nlmsg_data(nlh);
	219	struct fib_rules_ops *ops = NULL;
	220	struct fib_rule rule, r, *last = NULL;
	221	struct nlattr *tb[FRA_MAX+1];
0947c9fe	222	int err = -EINVAL, unresolved = 0;
14c0b97d TG	223
	224	if (nlh->nlmsg_len < nlmsg_msg_size(sizeof(*frh)))
	225	goto errout;
	226
5fd30ee7	227	ops = lookup_rules_ops(net, frh->family);
14c0b97d TG	228	if (ops == NULL) {
	229	err = EAFNOSUPPORT;
	230	goto errout;
	231	}
	232
	233	err = nlmsg_parse(nlh, sizeof(*frh), tb, FRA_MAX, ops->policy);
	234	if (err < 0)
	235	goto errout;
	236
e1701c68 TG	237	err = validate_rulemsg(frh, tb, ops);
	238	if (err < 0)
	239	goto errout;
	240
14c0b97d TG	241	rule = kzalloc(ops->rule_size, GFP_KERNEL);
	242	if (rule == NULL) {
	243	err = -ENOMEM;
	244	goto errout;
	245	}
51314a17	246	rule->fr_net = net;
14c0b97d TG	247
	248	if (tb[FRA_PRIORITY])
	249	rule->pref = nla_get_u32(tb[FRA_PRIORITY]);
	250
	251	if (tb[FRA_IFNAME]) {
	252	struct net_device *dev;
	253
	254	rule->ifindex = -1;
5176f91e	255	nla_strlcpy(rule->ifname, tb[FRA_IFNAME], IFNAMSIZ);
881d966b	256	dev = __dev_get_by_name(net, rule->ifname);
14c0b97d TG	257	if (dev)
	258	rule->ifindex = dev->ifindex;
	259	}
	260
b8964ed9 TG	261	if (tb[FRA_FWMARK]) {
	262	rule->mark = nla_get_u32(tb[FRA_FWMARK]);
	263	if (rule->mark)
	264	/* compatibility: if the mark value is non-zero all bits
	265	* are compared unless a mask is explicitly specified.
	266	*/
	267	rule->mark_mask = 0xFFFFFFFF;
	268	}
	269
	270	if (tb[FRA_FWMASK])
	271	rule->mark_mask = nla_get_u32(tb[FRA_FWMASK]);
	272
14c0b97d TG	273	rule->action = frh->action;
14c0b97d TG	274	rule->flags = frh->flags;
9e762a4a	275	rule->table = frh_get_table(frh, tb);
14c0b97d TG	276
14c0b97d TG	277	if (!rule->pref && ops->default_pref)
868d13ac	278	rule->pref = ops->default_pref(ops);
14c0b97d	279
0947c9fe TG	280	err = -EINVAL;
	281	if (tb[FRA_GOTO]) {
	282	if (rule->action != FR_ACT_GOTO)
	283	goto errout_free;
	284
	285	rule->target = nla_get_u32(tb[FRA_GOTO]);
	286	/* Backward jumps are prohibited to avoid endless loops */
	287	if (rule->target <= rule->pref)
	288	goto errout_free;
	289
76c72d4f	290	list_for_each_entry(r, &ops->rules_list, list) {
0947c9fe TG	291	if (r->pref == rule->target) {
	292	rule->ctarget = r;
	293	break;
	294	}
	295	}
	296
	297	if (rule->ctarget == NULL)
	298	unresolved = 1;
	299	} else if (rule->action == FR_ACT_GOTO)
	300	goto errout_free;
	301
14c0b97d TG	302	err = ops->configure(rule, skb, nlh, frh, tb);
	303	if (err < 0)
	304	goto errout_free;
	305
76c72d4f	306	list_for_each_entry(r, &ops->rules_list, list) {
14c0b97d TG	307	if (r->pref > rule->pref)
	308	break;
	309	last = r;
	310	}
	311
	312	fib_rule_get(rule);
	313
0947c9fe TG	314	if (ops->unresolved_rules) {
	315	/*
	316	* There are unresolved goto rules in the list, check if
	317	* any of them are pointing to this new rule.
	318	*/
76c72d4f	319	list_for_each_entry(r, &ops->rules_list, list) {
0947c9fe TG	320	if (r->action == FR_ACT_GOTO &&
	321	r->target == rule->pref) {
	322	BUG_ON(r->ctarget != NULL);
	323	rcu_assign_pointer(r->ctarget, rule);
	324	if (--ops->unresolved_rules == 0)
	325	break;
	326	}
	327	}
	328	}
	329
	330	if (rule->action == FR_ACT_GOTO)
	331	ops->nr_goto_rules++;
	332
	333	if (unresolved)
	334	ops->unresolved_rules++;
	335
14c0b97d TG	336	if (last)
	337	list_add_rcu(&rule->list, &last->list);
	338	else
76c72d4f	339	list_add_rcu(&rule->list, &ops->rules_list);
14c0b97d	340
9e3a5487	341	notify_rule_change(RTM_NEWRULE, rule, ops, nlh, NETLINK_CB(skb).pid);
73417f61	342	flush_route_cache(ops);
14c0b97d TG	343	rules_ops_put(ops);
	344	return 0;
	345
	346	errout_free:
	347	kfree(rule);
	348	errout:
	349	rules_ops_put(ops);
	350	return err;
	351	}
	352
9d9e6a58	353	static int fib_nl_delrule(struct sk_buff skb, struct nlmsghdr nlh, void *arg)
14c0b97d	354	{
3b1e0a65	355	struct net *net = sock_net(skb->sk);
14c0b97d TG	356	struct fib_rule_hdr *frh = nlmsg_data(nlh);
14c0b97d TG	357	struct fib_rules_ops *ops = NULL;
0947c9fe	358	struct fib_rule rule, tmp;
14c0b97d TG	359	struct nlattr *tb[FRA_MAX+1];
	360	int err = -EINVAL;
	361
	362	if (nlh->nlmsg_len < nlmsg_msg_size(sizeof(*frh)))
	363	goto errout;
	364
5fd30ee7	365	ops = lookup_rules_ops(net, frh->family);
14c0b97d TG	366	if (ops == NULL) {
	367	err = EAFNOSUPPORT;
	368	goto errout;
	369	}
	370
	371	err = nlmsg_parse(nlh, sizeof(*frh), tb, FRA_MAX, ops->policy);
	372	if (err < 0)
	373	goto errout;
	374
e1701c68 TG	375	err = validate_rulemsg(frh, tb, ops);
	376	if (err < 0)
	377	goto errout;
	378
76c72d4f	379	list_for_each_entry(rule, &ops->rules_list, list) {
14c0b97d TG	380	if (frh->action && (frh->action != rule->action))
	381	continue;
	382
9e762a4a	383	if (frh->table && (frh_get_table(frh, tb) != rule->table))
14c0b97d TG	384	continue;
	385
	386	if (tb[FRA_PRIORITY] &&
	387	(rule->pref != nla_get_u32(tb[FRA_PRIORITY])))
	388	continue;
	389
	390	if (tb[FRA_IFNAME] &&
	391	nla_strcmp(tb[FRA_IFNAME], rule->ifname))
	392	continue;
	393
b8964ed9 TG	394	if (tb[FRA_FWMARK] &&
	395	(rule->mark != nla_get_u32(tb[FRA_FWMARK])))
	396	continue;
	397
	398	if (tb[FRA_FWMASK] &&
	399	(rule->mark_mask != nla_get_u32(tb[FRA_FWMASK])))
	400	continue;
	401
14c0b97d TG	402	if (!ops->compare(rule, frh, tb))
	403	continue;
	404
	405	if (rule->flags & FIB_RULE_PERMANENT) {
	406	err = -EPERM;
	407	goto errout;
	408	}
	409
	410	list_del_rcu(&rule->list);
0947c9fe TG	411
	412	if (rule->action == FR_ACT_GOTO)
	413	ops->nr_goto_rules--;
	414
	415	/*
	416	* Check if this rule is a target to any of them. If so,
	417	* disable them. As this operation is eventually very
	418	* expensive, it is only performed if goto rules have
	419	* actually been added.
	420	*/
	421	if (ops->nr_goto_rules > 0) {
76c72d4f	422	list_for_each_entry(tmp, &ops->rules_list, list) {
0947c9fe TG	423	if (tmp->ctarget == rule) {
	424	rcu_assign_pointer(tmp->ctarget, NULL);
	425	ops->unresolved_rules++;
	426	}
	427	}
	428	}
	429
14c0b97d	430	synchronize_rcu();
9e3a5487	431	notify_rule_change(RTM_DELRULE, rule, ops, nlh,
c17084d2	432	NETLINK_CB(skb).pid);
14c0b97d	433	fib_rule_put(rule);
73417f61	434	flush_route_cache(ops);
14c0b97d TG	435	rules_ops_put(ops);
	436	return 0;
	437	}
	438
	439	err = -ENOENT;
	440	errout:
	441	rules_ops_put(ops);
	442	return err;
	443	}
	444
339bf98f TG	445	static inline size_t fib_rule_nlmsg_size(struct fib_rules_ops *ops,
	446	struct fib_rule *rule)
	447	{
	448	size_t payload = NLMSG_ALIGN(sizeof(struct fib_rule_hdr))
	449	+ nla_total_size(IFNAMSIZ) /* FRA_IFNAME */
	450	+ nla_total_size(4) /* FRA_PRIORITY */
	451	+ nla_total_size(4) /* FRA_TABLE */
	452	+ nla_total_size(4) /* FRA_FWMARK */
	453	+ nla_total_size(4); /* FRA_FWMASK */
	454
	455	if (ops->nlmsg_payload)
	456	payload += ops->nlmsg_payload(rule);
	457
	458	return payload;
	459	}
	460
14c0b97d TG	461	static int fib_nl_fill_rule(struct sk_buff skb, struct fib_rule rule,
	462	u32 pid, u32 seq, int type, int flags,
	463	struct fib_rules_ops *ops)
	464	{
	465	struct nlmsghdr *nlh;
	466	struct fib_rule_hdr *frh;
	467
	468	nlh = nlmsg_put(skb, pid, seq, type, sizeof(*frh), flags);
	469	if (nlh == NULL)
26932566	470	return -EMSGSIZE;
14c0b97d TG	471
	472	frh = nlmsg_data(nlh);
	473	frh->table = rule->table;
9e762a4a	474	NLA_PUT_U32(skb, FRA_TABLE, rule->table);
14c0b97d TG	475	frh->res1 = 0;
	476	frh->res2 = 0;
	477	frh->action = rule->action;
	478	frh->flags = rule->flags;
	479
0947c9fe TG	480	if (rule->action == FR_ACT_GOTO && rule->ctarget == NULL)
	481	frh->flags \|= FIB_RULE_UNRESOLVED;
	482
2b443683	483	if (rule->ifname[0]) {
14c0b97d TG	484	NLA_PUT_STRING(skb, FRA_IFNAME, rule->ifname);
14c0b97d TG	485
2b443683 TG	486	if (rule->ifindex == -1)
	487	frh->flags \|= FIB_RULE_DEV_DETACHED;
	488	}
	489
14c0b97d TG	490	if (rule->pref)
	491	NLA_PUT_U32(skb, FRA_PRIORITY, rule->pref);
	492
b8964ed9 TG	493	if (rule->mark)
	494	NLA_PUT_U32(skb, FRA_FWMARK, rule->mark);
	495
	496	if (rule->mark_mask \|\| rule->mark)
	497	NLA_PUT_U32(skb, FRA_FWMASK, rule->mark_mask);
	498
0947c9fe TG	499	if (rule->target)
	500	NLA_PUT_U32(skb, FRA_GOTO, rule->target);
	501
14c0b97d TG	502	if (ops->fill(rule, skb, nlh, frh) < 0)
	503	goto nla_put_failure;
	504
	505	return nlmsg_end(skb, nlh);
	506
	507	nla_put_failure:
26932566 PM	508	nlmsg_cancel(skb, nlh);
26932566 PM	509	return -EMSGSIZE;
14c0b97d TG	510	}
14c0b97d TG	511
c454673d TG	512	static int dump_rules(struct sk_buff skb, struct netlink_callback cb,
c454673d TG	513	struct fib_rules_ops *ops)
14c0b97d TG	514	{
	515	int idx = 0;
	516	struct fib_rule *rule;
14c0b97d	517
76c72d4f	518	list_for_each_entry(rule, &ops->rules_list, list) {
c454673d	519	if (idx < cb->args[1])
14c0b97d TG	520	goto skip;
	521
	522	if (fib_nl_fill_rule(skb, rule, NETLINK_CB(cb->skb).pid,
	523	cb->nlh->nlmsg_seq, RTM_NEWRULE,
	524	NLM_F_MULTI, ops) < 0)
	525	break;
	526	skip:
	527	idx++;
	528	}
c454673d	529	cb->args[1] = idx;
14c0b97d TG	530	rules_ops_put(ops);
	531
	532	return skb->len;
	533	}
	534
c454673d TG	535	static int fib_nl_dumprule(struct sk_buff skb, struct netlink_callback cb)
c454673d TG	536	{
3b1e0a65	537	struct net *net = sock_net(skb->sk);
c454673d TG	538	struct fib_rules_ops *ops;
	539	int idx = 0, family;
	540
	541	family = rtnl_msg_family(cb->nlh);
	542	if (family != AF_UNSPEC) {
	543	/* Protocol specific dump request */
5fd30ee7	544	ops = lookup_rules_ops(net, family);
c454673d TG	545	if (ops == NULL)
	546	return -EAFNOSUPPORT;
	547
	548	return dump_rules(skb, cb, ops);
	549	}
	550
	551	rcu_read_lock();
5fd30ee7	552	list_for_each_entry_rcu(ops, &net->rules_ops, list) {
c454673d TG	553	if (idx < cb->args[0] \|\| !try_module_get(ops->owner))
	554	goto skip;
	555
	556	if (dump_rules(skb, cb, ops) < 0)
	557	break;
	558
	559	cb->args[1] = 0;
	560	skip:
	561	idx++;
	562	}
	563	rcu_read_unlock();
	564	cb->args[0] = idx;
	565
	566	return skb->len;
	567	}
14c0b97d	568
9e3a5487	569	static void notify_rule_change(int event, struct fib_rule *rule,
c17084d2 TG	570	struct fib_rules_ops ops, struct nlmsghdr nlh,
c17084d2 TG	571	u32 pid)
14c0b97d	572	{
9e3a5487	573	struct net *net;
c17084d2 TG	574	struct sk_buff *skb;
c17084d2 TG	575	int err = -ENOBUFS;
14c0b97d	576
9e3a5487	577	net = ops->fro_net;
339bf98f	578	skb = nlmsg_new(fib_rule_nlmsg_size(ops, rule), GFP_KERNEL);
14c0b97d	579	if (skb == NULL)
c17084d2 TG	580	goto errout;
	581
	582	err = fib_nl_fill_rule(skb, rule, pid, nlh->nlmsg_seq, event, 0, ops);
26932566 PM	583	if (err < 0) {
	584	/* -EMSGSIZE implies BUG in fib_rule_nlmsg_size() */
	585	WARN_ON(err == -EMSGSIZE);
	586	kfree_skb(skb);
	587	goto errout;
	588	}
9e3a5487	589
5fd30ee7	590	err = rtnl_notify(skb, net, pid, ops->nlgroup, nlh, GFP_KERNEL);
c17084d2 TG	591	errout:
c17084d2 TG	592	if (err < 0)
5fd30ee7	593	rtnl_set_sk_err(net, ops->nlgroup, err);
14c0b97d TG	594	}
	595
	596	static void attach_rules(struct list_head rules, struct net_device dev)
	597	{
	598	struct fib_rule *rule;
	599
	600	list_for_each_entry(rule, rules, list) {
	601	if (rule->ifindex == -1 &&
	602	strcmp(dev->name, rule->ifname) == 0)
	603	rule->ifindex = dev->ifindex;
	604	}
	605	}
	606
	607	static void detach_rules(struct list_head rules, struct net_device dev)
	608	{
	609	struct fib_rule *rule;
	610
	611	list_for_each_entry(rule, rules, list)
	612	if (rule->ifindex == dev->ifindex)
	613	rule->ifindex = -1;
	614	}
	615
	616
	617	static int fib_rules_event(struct notifier_block *this, unsigned long event,
	618	void *ptr)
	619	{
	620	struct net_device *dev = ptr;
c346dca1	621	struct net *net = dev_net(dev);
14c0b97d TG	622	struct fib_rules_ops *ops;
	623
	624	ASSERT_RTNL();
	625	rcu_read_lock();
	626
	627	switch (event) {
	628	case NETDEV_REGISTER:
5fd30ee7	629	list_for_each_entry(ops, &net->rules_ops, list)
76c72d4f	630	attach_rules(&ops->rules_list, dev);
14c0b97d TG	631	break;
	632
	633	case NETDEV_UNREGISTER:
5fd30ee7	634	list_for_each_entry(ops, &net->rules_ops, list)
76c72d4f	635	detach_rules(&ops->rules_list, dev);
14c0b97d TG	636	break;
	637	}
	638
	639	rcu_read_unlock();
	640
	641	return NOTIFY_DONE;
	642	}
	643
	644	static struct notifier_block fib_rules_notifier = {
	645	.notifier_call = fib_rules_event,
	646	};
	647
5fd30ee7 DL	648	static int fib_rules_net_init(struct net *net)
	649	{
	650	INIT_LIST_HEAD(&net->rules_ops);
	651	spin_lock_init(&net->rules_mod_lock);
	652	return 0;
	653	}
	654
	655	static struct pernet_operations fib_rules_net_ops = {
	656	.init = fib_rules_net_init,
	657	};
	658
14c0b97d TG	659	static int __init fib_rules_init(void)
14c0b97d TG	660	{
5fd30ee7	661	int err;
9d9e6a58 TG	662	rtnl_register(PF_UNSPEC, RTM_NEWRULE, fib_nl_newrule, NULL);
9d9e6a58 TG	663	rtnl_register(PF_UNSPEC, RTM_DELRULE, fib_nl_delrule, NULL);
c454673d	664	rtnl_register(PF_UNSPEC, RTM_GETRULE, NULL, fib_nl_dumprule);
9d9e6a58	665
5fd30ee7 DL	666	err = register_netdevice_notifier(&fib_rules_notifier);
	667	if (err < 0)
	668	goto fail;
	669
	670	err = register_pernet_subsys(&fib_rules_net_ops);
	671	if (err < 0)
	672	goto fail_unregister;
	673	return 0;
	674
	675	fail_unregister:
	676	unregister_netdevice_notifier(&fib_rules_notifier);
	677	fail:
	678	rtnl_unregister(PF_UNSPEC, RTM_NEWRULE);
	679	rtnl_unregister(PF_UNSPEC, RTM_DELRULE);
	680	rtnl_unregister(PF_UNSPEC, RTM_GETRULE);
	681	return err;
14c0b97d TG	682	}
	683
	684	subsys_initcall(fib_rules_init);