[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/slab.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 = hold_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);

u32 fib_default_rule_pref(struct fib_rules_ops *ops)
{
	struct list_head *pos;
	struct fib_rule *rule;

	if (!list_empty(&ops->rules_list)) {
		pos = ops->rules_list.next;
		if (pos->next != &ops->rules_list) {
			rule = list_entry(pos->next, struct fib_rule, list);
			if (rule->pref)
				return rule->pref - 1;
		}
	}

	return 0;
}
EXPORT_SYMBOL(fib_default_rule_pref);

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(ops);
}

static 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;
}

struct fib_rules_ops *
fib_rules_register(struct fib_rules_ops *tmpl, struct net *net)
{
	struct fib_rules_ops *ops;
	int err;

	ops = kmemdup(tmpl, sizeof(*ops), GFP_KERNEL);
	if (ops == NULL)
		return ERR_PTR(-ENOMEM);

	INIT_LIST_HEAD(&ops->rules_list);
	ops->fro_net = net;

	err = __fib_rules_register(ops);
	if (err) {
		kfree(ops);
		ops = ERR_PTR(err);
	}

	return ops;
}
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);

static void fib_rules_put_rcu(struct rcu_head *head)
{
	struct fib_rules_ops *ops = container_of(head, struct fib_rules_ops, rcu);
	struct net *net = ops->fro_net;

	release_net(net);
	kfree(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);

	call_rcu(&ops->rcu, fib_rules_put_rcu);
}
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->iifindex && (rule->iifindex != fl->iif))
		goto out;

	if (rule->oifindex && (rule->oifindex != fl->oif))
		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 = hold_net(net);

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

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

		rule->iifindex = -1;
		nla_strlcpy(rule->iifname, tb[FRA_IIFNAME], IFNAMSIZ);
		dev = __dev_get_by_name(net, rule->iifname);
		if (dev)
			rule->iifindex = dev->ifindex;
	}

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

		rule->oifindex = -1;
		nla_strlcpy(rule->oifname, tb[FRA_OIFNAME], IFNAMSIZ);
		dev = __dev_get_by_name(net, rule->oifname);
		if (dev)
			rule->oifindex = 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 (!tb[FRA_PRIORITY] && 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, 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:
	release_net(rule->fr_net);
	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_IIFNAME] &&
		    nla_strcmp(tb[FRA_IIFNAME], rule->iifname))
			continue;

		if (tb[FRA_OIFNAME] &&
		    nla_strcmp(tb[FRA_OIFNAME], rule->oifname))
			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_IIFNAME */
			 + nla_total_size(IFNAMSIZ) /* FRA_OIFNAME */
			 + 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->family = ops->family;
	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->iifname[0]) {
		NLA_PUT_STRING(skb, FRA_IIFNAME, rule->iifname);

		if (rule->iifindex == -1)
			frh->flags |= FIB_RULE_IIF_DETACHED;
	}

	if (rule->oifname[0]) {
		NLA_PUT_STRING(skb, FRA_OIFNAME, rule->oifname);

		if (rule->oifindex == -1)
			frh->flags |= FIB_RULE_OIF_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, 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;
	}

	rtnl_notify(skb, net, pid, ops->nlgroup, nlh, GFP_KERNEL);
	return;
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->iifindex == -1 &&
		    strcmp(dev->name, rule->iifname) == 0)
			rule->iifindex = dev->ifindex;
		if (rule->oifindex == -1 &&
		    strcmp(dev->name, rule->oifname) == 0)
			rule->oifindex = 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->iifindex == dev->ifindex)
			rule->iifindex = -1;
		if (rule->oifindex == dev->ifindex)
			rule->oifindex = -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();

	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;
	}

	return NOTIFY_DONE;
}

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

static int __net_init 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_pernet_subsys(&fib_rules_net_ops);
	if (err < 0)
		goto fail;

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

	return 0;

fail_unregister:
	unregister_pernet_subsys(&fib_rules_net_ops);
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>
14c0b97d TG	12	#include <linux/kernel.h>
5a0e3ad6	13	#include <linux/slab.h>
14c0b97d	14	#include <linux/list.h>
e9dc8653	15	#include <net/net_namespace.h>
881d966b	16	#include <net/sock.h>
14c0b97d TG	17	#include <net/fib_rules.h>
14c0b97d TG	18
2994c638 DL	19	int fib_default_rule_add(struct fib_rules_ops *ops,
	20	u32 pref, u32 table, u32 flags)
	21	{
	22	struct fib_rule *r;
	23
	24	r = kzalloc(ops->rule_size, GFP_KERNEL);
	25	if (r == NULL)
	26	return -ENOMEM;
	27
	28	atomic_set(&r->refcnt, 1);
	29	r->action = FR_ACT_TO_TBL;
	30	r->pref = pref;
	31	r->table = table;
	32	r->flags = flags;
3661a910	33	r->fr_net = hold_net(ops->fro_net);
2994c638 DL	34
	35	/* The lock is not required here, the list in unreacheable
	36	* at the moment this function is called */
	37	list_add_tail(&r->list, &ops->rules_list);
	38	return 0;
	39	}
	40	EXPORT_SYMBOL(fib_default_rule_add);
	41
d8a566be PM	42	u32 fib_default_rule_pref(struct fib_rules_ops *ops)
	43	{
	44	struct list_head *pos;
	45	struct fib_rule *rule;
	46
	47	if (!list_empty(&ops->rules_list)) {
	48	pos = ops->rules_list.next;
	49	if (pos->next != &ops->rules_list) {
	50	rule = list_entry(pos->next, struct fib_rule, list);
	51	if (rule->pref)
	52	return rule->pref - 1;
	53	}
	54	}
	55
	56	return 0;
	57	}
	58	EXPORT_SYMBOL(fib_default_rule_pref);
	59
9e3a5487	60	static void notify_rule_change(int event, struct fib_rule *rule,
c17084d2 TG	61	struct fib_rules_ops ops, struct nlmsghdr nlh,
c17084d2 TG	62	u32 pid);
14c0b97d	63
5fd30ee7	64	static struct fib_rules_ops lookup_rules_ops(struct net net, int family)
14c0b97d TG	65	{
	66	struct fib_rules_ops *ops;
	67
	68	rcu_read_lock();
5fd30ee7	69	list_for_each_entry_rcu(ops, &net->rules_ops, list) {
14c0b97d TG	70	if (ops->family == family) {
	71	if (!try_module_get(ops->owner))
	72	ops = NULL;
	73	rcu_read_unlock();
	74	return ops;
	75	}
	76	}
	77	rcu_read_unlock();
	78
	79	return NULL;
	80	}
	81
	82	static void rules_ops_put(struct fib_rules_ops *ops)
	83	{
	84	if (ops)
	85	module_put(ops->owner);
	86	}
	87
73417f61 TG	88	static void flush_route_cache(struct fib_rules_ops *ops)
	89	{
	90	if (ops->flush_cache)
ae299fc0	91	ops->flush_cache(ops);
73417f61 TG	92	}
73417f61 TG	93
e9c5158a	94	static int __fib_rules_register(struct fib_rules_ops *ops)
14c0b97d TG	95	{
	96	int err = -EEXIST;
	97	struct fib_rules_ops *o;
9e3a5487 DL	98	struct net *net;
	99
	100	net = ops->fro_net;
14c0b97d TG	101
	102	if (ops->rule_size < sizeof(struct fib_rule))
	103	return -EINVAL;
	104
	105	if (ops->match == NULL \|\| ops->configure == NULL \|\|
	106	ops->compare == NULL \|\| ops->fill == NULL \|\|
	107	ops->action == NULL)
	108	return -EINVAL;
	109
5fd30ee7 DL	110	spin_lock(&net->rules_mod_lock);
5fd30ee7 DL	111	list_for_each_entry(o, &net->rules_ops, list)
14c0b97d TG	112	if (ops->family == o->family)
	113	goto errout;
	114
5fd30ee7 DL	115	hold_net(net);
5fd30ee7 DL	116	list_add_tail_rcu(&ops->list, &net->rules_ops);
14c0b97d TG	117	err = 0;
14c0b97d TG	118	errout:
5fd30ee7	119	spin_unlock(&net->rules_mod_lock);
14c0b97d TG	120
	121	return err;
	122	}
	123
e9c5158a EB	124	struct fib_rules_ops *
	125	fib_rules_register(struct fib_rules_ops tmpl, struct net net)
	126	{
	127	struct fib_rules_ops *ops;
	128	int err;
	129
2fb3573d	130	ops = kmemdup(tmpl, sizeof(*ops), GFP_KERNEL);
e9c5158a EB	131	if (ops == NULL)
	132	return ERR_PTR(-ENOMEM);
	133
	134	INIT_LIST_HEAD(&ops->rules_list);
	135	ops->fro_net = net;
	136
	137	err = __fib_rules_register(ops);
	138	if (err) {
	139	kfree(ops);
	140	ops = ERR_PTR(err);
	141	}
	142
	143	return ops;
	144	}
14c0b97d TG	145	EXPORT_SYMBOL_GPL(fib_rules_register);
14c0b97d TG	146
9eb87f3f	147	void fib_rules_cleanup_ops(struct fib_rules_ops *ops)
14c0b97d TG	148	{
	149	struct fib_rule rule, tmp;
	150
76c72d4f	151	list_for_each_entry_safe(rule, tmp, &ops->rules_list, list) {
14c0b97d TG	152	list_del_rcu(&rule->list);
	153	fib_rule_put(rule);
	154	}
	155	}
9eb87f3f	156	EXPORT_SYMBOL_GPL(fib_rules_cleanup_ops);
14c0b97d	157
e9c5158a EB	158	static void fib_rules_put_rcu(struct rcu_head *head)
	159	{
	160	struct fib_rules_ops *ops = container_of(head, struct fib_rules_ops, rcu);
	161	struct net *net = ops->fro_net;
	162
	163	release_net(net);
	164	kfree(ops);
	165	}
	166
9e3a5487	167	void fib_rules_unregister(struct fib_rules_ops *ops)
14c0b97d	168	{
9e3a5487	169	struct net *net = ops->fro_net;
14c0b97d	170
5fd30ee7	171	spin_lock(&net->rules_mod_lock);
72132c1b DL	172	list_del_rcu(&ops->list);
72132c1b DL	173	fib_rules_cleanup_ops(ops);
5fd30ee7	174	spin_unlock(&net->rules_mod_lock);
14c0b97d	175
e9c5158a	176	call_rcu(&ops->rcu, fib_rules_put_rcu);
14c0b97d	177	}
14c0b97d TG	178	EXPORT_SYMBOL_GPL(fib_rules_unregister);
14c0b97d TG	179
3dfbcc41 TG	180	static int fib_rule_match(struct fib_rule rule, struct fib_rules_ops ops,
	181	struct flowi *fl, int flags)
	182	{
	183	int ret = 0;
	184
491deb24	185	if (rule->iifindex && (rule->iifindex != fl->iif))
3dfbcc41 TG	186	goto out;
3dfbcc41 TG	187
1b038a5e PM	188	if (rule->oifindex && (rule->oifindex != fl->oif))
	189	goto out;
	190
3dfbcc41 TG	191	if ((rule->mark ^ fl->mark) & rule->mark_mask)
	192	goto out;
	193
	194	ret = ops->match(rule, fl, flags);
	195	out:
	196	return (rule->flags & FIB_RULE_INVERT) ? !ret : ret;
	197	}
	198
14c0b97d TG	199	int fib_rules_lookup(struct fib_rules_ops ops, struct flowi fl,
	200	int flags, struct fib_lookup_arg *arg)
	201	{
	202	struct fib_rule *rule;
	203	int err;
	204
	205	rcu_read_lock();
	206
76c72d4f	207	list_for_each_entry_rcu(rule, &ops->rules_list, list) {
0947c9fe	208	jumped:
3dfbcc41	209	if (!fib_rule_match(rule, ops, fl, flags))
14c0b97d TG	210	continue;
14c0b97d TG	211
0947c9fe TG	212	if (rule->action == FR_ACT_GOTO) {
	213	struct fib_rule *target;
	214
	215	target = rcu_dereference(rule->ctarget);
	216	if (target == NULL) {
	217	continue;
	218	} else {
	219	rule = target;
	220	goto jumped;
	221	}
fa0b2d1d TG	222	} else if (rule->action == FR_ACT_NOP)
	223	continue;
	224	else
0947c9fe TG	225	err = ops->action(rule, fl, flags, arg);
0947c9fe TG	226
14c0b97d TG	227	if (err != -EAGAIN) {
	228	fib_rule_get(rule);
	229	arg->rule = rule;
	230	goto out;
	231	}
	232	}
	233
83886b6b	234	err = -ESRCH;
14c0b97d TG	235	out:
	236	rcu_read_unlock();
	237
	238	return err;
	239	}
14c0b97d TG	240	EXPORT_SYMBOL_GPL(fib_rules_lookup);
14c0b97d TG	241
e1701c68 TG	242	static int validate_rulemsg(struct fib_rule_hdr frh, struct nlattr *tb,
	243	struct fib_rules_ops *ops)
	244	{
	245	int err = -EINVAL;
	246
	247	if (frh->src_len)
	248	if (tb[FRA_SRC] == NULL \|\|
	249	frh->src_len > (ops->addr_size * 8) \|\|
	250	nla_len(tb[FRA_SRC]) != ops->addr_size)
	251	goto errout;
	252
	253	if (frh->dst_len)
	254	if (tb[FRA_DST] == NULL \|\|
	255	frh->dst_len > (ops->addr_size * 8) \|\|
	256	nla_len(tb[FRA_DST]) != ops->addr_size)
	257	goto errout;
	258
	259	err = 0;
	260	errout:
	261	return err;
	262	}
	263
9d9e6a58	264	static int fib_nl_newrule(struct sk_buff skb, struct nlmsghdr nlh, void *arg)
14c0b97d	265	{
3b1e0a65	266	struct net *net = sock_net(skb->sk);
14c0b97d TG	267	struct fib_rule_hdr *frh = nlmsg_data(nlh);
	268	struct fib_rules_ops *ops = NULL;
	269	struct fib_rule rule, r, *last = NULL;
	270	struct nlattr *tb[FRA_MAX+1];
0947c9fe	271	int err = -EINVAL, unresolved = 0;
14c0b97d TG	272
	273	if (nlh->nlmsg_len < nlmsg_msg_size(sizeof(*frh)))
	274	goto errout;
	275
5fd30ee7	276	ops = lookup_rules_ops(net, frh->family);
14c0b97d	277	if (ops == NULL) {
2fe195cf	278	err = -EAFNOSUPPORT;
14c0b97d TG	279	goto errout;
	280	}
	281
	282	err = nlmsg_parse(nlh, sizeof(*frh), tb, FRA_MAX, ops->policy);
	283	if (err < 0)
	284	goto errout;
	285
e1701c68 TG	286	err = validate_rulemsg(frh, tb, ops);
	287	if (err < 0)
	288	goto errout;
	289
14c0b97d TG	290	rule = kzalloc(ops->rule_size, GFP_KERNEL);
	291	if (rule == NULL) {
	292	err = -ENOMEM;
	293	goto errout;
	294	}
3661a910	295	rule->fr_net = hold_net(net);
14c0b97d TG	296
	297	if (tb[FRA_PRIORITY])
	298	rule->pref = nla_get_u32(tb[FRA_PRIORITY]);
	299
491deb24	300	if (tb[FRA_IIFNAME]) {
14c0b97d TG	301	struct net_device *dev;
14c0b97d TG	302
491deb24 PM	303	rule->iifindex = -1;
	304	nla_strlcpy(rule->iifname, tb[FRA_IIFNAME], IFNAMSIZ);
	305	dev = __dev_get_by_name(net, rule->iifname);
14c0b97d	306	if (dev)
491deb24	307	rule->iifindex = dev->ifindex;
14c0b97d TG	308	}
14c0b97d TG	309
1b038a5e PM	310	if (tb[FRA_OIFNAME]) {
	311	struct net_device *dev;
	312
	313	rule->oifindex = -1;
	314	nla_strlcpy(rule->oifname, tb[FRA_OIFNAME], IFNAMSIZ);
	315	dev = __dev_get_by_name(net, rule->oifname);
	316	if (dev)
	317	rule->oifindex = dev->ifindex;
	318	}
	319
b8964ed9 TG	320	if (tb[FRA_FWMARK]) {
	321	rule->mark = nla_get_u32(tb[FRA_FWMARK]);
	322	if (rule->mark)
	323	/* compatibility: if the mark value is non-zero all bits
	324	* are compared unless a mask is explicitly specified.
	325	*/
	326	rule->mark_mask = 0xFFFFFFFF;
	327	}
	328
	329	if (tb[FRA_FWMASK])
	330	rule->mark_mask = nla_get_u32(tb[FRA_FWMASK]);
	331
14c0b97d TG	332	rule->action = frh->action;
14c0b97d TG	333	rule->flags = frh->flags;
9e762a4a	334	rule->table = frh_get_table(frh, tb);
14c0b97d	335
5adef180	336	if (!tb[FRA_PRIORITY] && ops->default_pref)
868d13ac	337	rule->pref = ops->default_pref(ops);
14c0b97d	338
0947c9fe TG	339	err = -EINVAL;
	340	if (tb[FRA_GOTO]) {
	341	if (rule->action != FR_ACT_GOTO)
	342	goto errout_free;
	343
	344	rule->target = nla_get_u32(tb[FRA_GOTO]);
	345	/* Backward jumps are prohibited to avoid endless loops */
	346	if (rule->target <= rule->pref)
	347	goto errout_free;
	348
76c72d4f	349	list_for_each_entry(r, &ops->rules_list, list) {
0947c9fe TG	350	if (r->pref == rule->target) {
	351	rule->ctarget = r;
	352	break;
	353	}
	354	}
	355
	356	if (rule->ctarget == NULL)
	357	unresolved = 1;
	358	} else if (rule->action == FR_ACT_GOTO)
	359	goto errout_free;
	360
8b3521ee	361	err = ops->configure(rule, skb, frh, tb);
14c0b97d TG	362	if (err < 0)
	363	goto errout_free;
	364
76c72d4f	365	list_for_each_entry(r, &ops->rules_list, list) {
14c0b97d TG	366	if (r->pref > rule->pref)
	367	break;
	368	last = r;
	369	}
	370
	371	fib_rule_get(rule);
	372
0947c9fe TG	373	if (ops->unresolved_rules) {
	374	/*
	375	* There are unresolved goto rules in the list, check if
	376	* any of them are pointing to this new rule.
	377	*/
76c72d4f	378	list_for_each_entry(r, &ops->rules_list, list) {
0947c9fe TG	379	if (r->action == FR_ACT_GOTO &&
	380	r->target == rule->pref) {
	381	BUG_ON(r->ctarget != NULL);
	382	rcu_assign_pointer(r->ctarget, rule);
	383	if (--ops->unresolved_rules == 0)
	384	break;
	385	}
	386	}
	387	}
	388
	389	if (rule->action == FR_ACT_GOTO)
	390	ops->nr_goto_rules++;
	391
	392	if (unresolved)
	393	ops->unresolved_rules++;
	394
14c0b97d TG	395	if (last)
	396	list_add_rcu(&rule->list, &last->list);
	397	else
76c72d4f	398	list_add_rcu(&rule->list, &ops->rules_list);
14c0b97d	399
9e3a5487	400	notify_rule_change(RTM_NEWRULE, rule, ops, nlh, NETLINK_CB(skb).pid);
73417f61	401	flush_route_cache(ops);
14c0b97d TG	402	rules_ops_put(ops);
	403	return 0;
	404
	405	errout_free:
3661a910	406	release_net(rule->fr_net);
14c0b97d TG	407	kfree(rule);
	408	errout:
	409	rules_ops_put(ops);
	410	return err;
	411	}
	412
9d9e6a58	413	static int fib_nl_delrule(struct sk_buff skb, struct nlmsghdr nlh, void *arg)
14c0b97d	414	{
3b1e0a65	415	struct net *net = sock_net(skb->sk);
14c0b97d TG	416	struct fib_rule_hdr *frh = nlmsg_data(nlh);
14c0b97d TG	417	struct fib_rules_ops *ops = NULL;
0947c9fe	418	struct fib_rule rule, tmp;
14c0b97d TG	419	struct nlattr *tb[FRA_MAX+1];
	420	int err = -EINVAL;
	421
	422	if (nlh->nlmsg_len < nlmsg_msg_size(sizeof(*frh)))
	423	goto errout;
	424
5fd30ee7	425	ops = lookup_rules_ops(net, frh->family);
14c0b97d	426	if (ops == NULL) {
2fe195cf	427	err = -EAFNOSUPPORT;
14c0b97d TG	428	goto errout;
	429	}
	430
	431	err = nlmsg_parse(nlh, sizeof(*frh), tb, FRA_MAX, ops->policy);
	432	if (err < 0)
	433	goto errout;
	434
e1701c68 TG	435	err = validate_rulemsg(frh, tb, ops);
	436	if (err < 0)
	437	goto errout;
	438
76c72d4f	439	list_for_each_entry(rule, &ops->rules_list, list) {
14c0b97d TG	440	if (frh->action && (frh->action != rule->action))
	441	continue;
	442
9e762a4a	443	if (frh->table && (frh_get_table(frh, tb) != rule->table))
14c0b97d TG	444	continue;
	445
	446	if (tb[FRA_PRIORITY] &&
	447	(rule->pref != nla_get_u32(tb[FRA_PRIORITY])))
	448	continue;
	449
491deb24 PM	450	if (tb[FRA_IIFNAME] &&
491deb24 PM	451	nla_strcmp(tb[FRA_IIFNAME], rule->iifname))
14c0b97d TG	452	continue;
14c0b97d TG	453
1b038a5e PM	454	if (tb[FRA_OIFNAME] &&
	455	nla_strcmp(tb[FRA_OIFNAME], rule->oifname))
	456	continue;
	457
b8964ed9 TG	458	if (tb[FRA_FWMARK] &&
	459	(rule->mark != nla_get_u32(tb[FRA_FWMARK])))
	460	continue;
	461
	462	if (tb[FRA_FWMASK] &&
	463	(rule->mark_mask != nla_get_u32(tb[FRA_FWMASK])))
	464	continue;
	465
14c0b97d TG	466	if (!ops->compare(rule, frh, tb))
	467	continue;
	468
	469	if (rule->flags & FIB_RULE_PERMANENT) {
	470	err = -EPERM;
	471	goto errout;
	472	}
	473
	474	list_del_rcu(&rule->list);
0947c9fe TG	475
	476	if (rule->action == FR_ACT_GOTO)
	477	ops->nr_goto_rules--;
	478
	479	/*
	480	* Check if this rule is a target to any of them. If so,
	481	* disable them. As this operation is eventually very
	482	* expensive, it is only performed if goto rules have
	483	* actually been added.
	484	*/
	485	if (ops->nr_goto_rules > 0) {
76c72d4f	486	list_for_each_entry(tmp, &ops->rules_list, list) {
0947c9fe TG	487	if (tmp->ctarget == rule) {
	488	rcu_assign_pointer(tmp->ctarget, NULL);
	489	ops->unresolved_rules++;
	490	}
	491	}
	492	}
	493
14c0b97d	494	synchronize_rcu();
9e3a5487	495	notify_rule_change(RTM_DELRULE, rule, ops, nlh,
c17084d2	496	NETLINK_CB(skb).pid);
14c0b97d	497	fib_rule_put(rule);
73417f61	498	flush_route_cache(ops);
14c0b97d TG	499	rules_ops_put(ops);
	500	return 0;
	501	}
	502
	503	err = -ENOENT;
	504	errout:
	505	rules_ops_put(ops);
	506	return err;
	507	}
	508
339bf98f TG	509	static inline size_t fib_rule_nlmsg_size(struct fib_rules_ops *ops,
	510	struct fib_rule *rule)
	511	{
	512	size_t payload = NLMSG_ALIGN(sizeof(struct fib_rule_hdr))
491deb24	513	+ nla_total_size(IFNAMSIZ) /* FRA_IIFNAME */
1b038a5e	514	+ nla_total_size(IFNAMSIZ) /* FRA_OIFNAME */
339bf98f TG	515	+ nla_total_size(4) /* FRA_PRIORITY */
	516	+ nla_total_size(4) /* FRA_TABLE */
	517	+ nla_total_size(4) /* FRA_FWMARK */
	518	+ nla_total_size(4); /* FRA_FWMASK */
	519
	520	if (ops->nlmsg_payload)
	521	payload += ops->nlmsg_payload(rule);
	522
	523	return payload;
	524	}
	525
14c0b97d TG	526	static int fib_nl_fill_rule(struct sk_buff skb, struct fib_rule rule,
	527	u32 pid, u32 seq, int type, int flags,
	528	struct fib_rules_ops *ops)
	529	{
	530	struct nlmsghdr *nlh;
	531	struct fib_rule_hdr *frh;
	532
	533	nlh = nlmsg_put(skb, pid, seq, type, sizeof(*frh), flags);
	534	if (nlh == NULL)
26932566	535	return -EMSGSIZE;
14c0b97d TG	536
14c0b97d TG	537	frh = nlmsg_data(nlh);
28bb1726	538	frh->family = ops->family;
14c0b97d	539	frh->table = rule->table;
9e762a4a	540	NLA_PUT_U32(skb, FRA_TABLE, rule->table);
14c0b97d TG	541	frh->res1 = 0;
	542	frh->res2 = 0;
	543	frh->action = rule->action;
	544	frh->flags = rule->flags;
	545
0947c9fe TG	546	if (rule->action == FR_ACT_GOTO && rule->ctarget == NULL)
	547	frh->flags \|= FIB_RULE_UNRESOLVED;
	548
491deb24 PM	549	if (rule->iifname[0]) {
491deb24 PM	550	NLA_PUT_STRING(skb, FRA_IIFNAME, rule->iifname);
14c0b97d	551
491deb24 PM	552	if (rule->iifindex == -1)
491deb24 PM	553	frh->flags \|= FIB_RULE_IIF_DETACHED;
2b443683 TG	554	}
2b443683 TG	555
1b038a5e PM	556	if (rule->oifname[0]) {
	557	NLA_PUT_STRING(skb, FRA_OIFNAME, rule->oifname);
	558
	559	if (rule->oifindex == -1)
	560	frh->flags \|= FIB_RULE_OIF_DETACHED;
	561	}
	562
14c0b97d TG	563	if (rule->pref)
	564	NLA_PUT_U32(skb, FRA_PRIORITY, rule->pref);
	565
b8964ed9 TG	566	if (rule->mark)
	567	NLA_PUT_U32(skb, FRA_FWMARK, rule->mark);
	568
	569	if (rule->mark_mask \|\| rule->mark)
	570	NLA_PUT_U32(skb, FRA_FWMASK, rule->mark_mask);
	571
0947c9fe TG	572	if (rule->target)
	573	NLA_PUT_U32(skb, FRA_GOTO, rule->target);
	574
04af8cf6	575	if (ops->fill(rule, skb, frh) < 0)
14c0b97d TG	576	goto nla_put_failure;
	577
	578	return nlmsg_end(skb, nlh);
	579
	580	nla_put_failure:
26932566 PM	581	nlmsg_cancel(skb, nlh);
26932566 PM	582	return -EMSGSIZE;
14c0b97d TG	583	}
14c0b97d TG	584
c454673d TG	585	static int dump_rules(struct sk_buff skb, struct netlink_callback cb,
c454673d TG	586	struct fib_rules_ops *ops)
14c0b97d TG	587	{
	588	int idx = 0;
	589	struct fib_rule *rule;
14c0b97d	590
76c72d4f	591	list_for_each_entry(rule, &ops->rules_list, list) {
c454673d	592	if (idx < cb->args[1])
14c0b97d TG	593	goto skip;
	594
	595	if (fib_nl_fill_rule(skb, rule, NETLINK_CB(cb->skb).pid,
	596	cb->nlh->nlmsg_seq, RTM_NEWRULE,
	597	NLM_F_MULTI, ops) < 0)
	598	break;
	599	skip:
	600	idx++;
	601	}
c454673d	602	cb->args[1] = idx;
14c0b97d TG	603	rules_ops_put(ops);
	604
	605	return skb->len;
	606	}
	607
c454673d TG	608	static int fib_nl_dumprule(struct sk_buff skb, struct netlink_callback cb)
c454673d TG	609	{
3b1e0a65	610	struct net *net = sock_net(skb->sk);
c454673d TG	611	struct fib_rules_ops *ops;
	612	int idx = 0, family;
	613
	614	family = rtnl_msg_family(cb->nlh);
	615	if (family != AF_UNSPEC) {
	616	/* Protocol specific dump request */
5fd30ee7	617	ops = lookup_rules_ops(net, family);
c454673d TG	618	if (ops == NULL)
	619	return -EAFNOSUPPORT;
	620
	621	return dump_rules(skb, cb, ops);
	622	}
	623
	624	rcu_read_lock();
5fd30ee7	625	list_for_each_entry_rcu(ops, &net->rules_ops, list) {
c454673d TG	626	if (idx < cb->args[0] \|\| !try_module_get(ops->owner))
	627	goto skip;
	628
	629	if (dump_rules(skb, cb, ops) < 0)
	630	break;
	631
	632	cb->args[1] = 0;
2fb3573d	633	skip:
c454673d TG	634	idx++;
	635	}
	636	rcu_read_unlock();
	637	cb->args[0] = idx;
	638
	639	return skb->len;
	640	}
14c0b97d	641
9e3a5487	642	static void notify_rule_change(int event, struct fib_rule *rule,
c17084d2 TG	643	struct fib_rules_ops ops, struct nlmsghdr nlh,
c17084d2 TG	644	u32 pid)
14c0b97d	645	{
9e3a5487	646	struct net *net;
c17084d2 TG	647	struct sk_buff *skb;
c17084d2 TG	648	int err = -ENOBUFS;
14c0b97d	649
9e3a5487	650	net = ops->fro_net;
339bf98f	651	skb = nlmsg_new(fib_rule_nlmsg_size(ops, rule), GFP_KERNEL);
14c0b97d	652	if (skb == NULL)
c17084d2 TG	653	goto errout;
	654
	655	err = fib_nl_fill_rule(skb, rule, pid, nlh->nlmsg_seq, event, 0, ops);
26932566 PM	656	if (err < 0) {
	657	/* -EMSGSIZE implies BUG in fib_rule_nlmsg_size() */
	658	WARN_ON(err == -EMSGSIZE);
	659	kfree_skb(skb);
	660	goto errout;
	661	}
9e3a5487	662
1ce85fe4 PNA	663	rtnl_notify(skb, net, pid, ops->nlgroup, nlh, GFP_KERNEL);
1ce85fe4 PNA	664	return;
c17084d2 TG	665	errout:
c17084d2 TG	666	if (err < 0)
5fd30ee7	667	rtnl_set_sk_err(net, ops->nlgroup, err);
14c0b97d TG	668	}
	669
	670	static void attach_rules(struct list_head rules, struct net_device dev)
	671	{
	672	struct fib_rule *rule;
	673
	674	list_for_each_entry(rule, rules, list) {
491deb24 PM	675	if (rule->iifindex == -1 &&
	676	strcmp(dev->name, rule->iifname) == 0)
	677	rule->iifindex = dev->ifindex;
1b038a5e PM	678	if (rule->oifindex == -1 &&
	679	strcmp(dev->name, rule->oifname) == 0)
	680	rule->oifindex = dev->ifindex;
14c0b97d TG	681	}
	682	}
	683
	684	static void detach_rules(struct list_head rules, struct net_device dev)
	685	{
	686	struct fib_rule *rule;
	687
1b038a5e	688	list_for_each_entry(rule, rules, list) {
491deb24 PM	689	if (rule->iifindex == dev->ifindex)
491deb24 PM	690	rule->iifindex = -1;
1b038a5e PM	691	if (rule->oifindex == dev->ifindex)
	692	rule->oifindex = -1;
	693	}
14c0b97d TG	694	}
	695
	696
	697	static int fib_rules_event(struct notifier_block *this, unsigned long event,
	698	void *ptr)
	699	{
	700	struct net_device *dev = ptr;
c346dca1	701	struct net *net = dev_net(dev);
14c0b97d TG	702	struct fib_rules_ops *ops;
	703
	704	ASSERT_RTNL();
14c0b97d TG	705
	706	switch (event) {
	707	case NETDEV_REGISTER:
5fd30ee7	708	list_for_each_entry(ops, &net->rules_ops, list)
76c72d4f	709	attach_rules(&ops->rules_list, dev);
14c0b97d TG	710	break;
	711
	712	case NETDEV_UNREGISTER:
5fd30ee7	713	list_for_each_entry(ops, &net->rules_ops, list)
76c72d4f	714	detach_rules(&ops->rules_list, dev);
14c0b97d TG	715	break;
	716	}
	717
14c0b97d TG	718	return NOTIFY_DONE;
	719	}
	720
	721	static struct notifier_block fib_rules_notifier = {
	722	.notifier_call = fib_rules_event,
	723	};
	724
2c8c1e72	725	static int __net_init fib_rules_net_init(struct net *net)
5fd30ee7 DL	726	{
	727	INIT_LIST_HEAD(&net->rules_ops);
	728	spin_lock_init(&net->rules_mod_lock);
	729	return 0;
	730	}
	731
	732	static struct pernet_operations fib_rules_net_ops = {
	733	.init = fib_rules_net_init,
	734	};
	735
14c0b97d TG	736	static int __init fib_rules_init(void)
14c0b97d TG	737	{
5fd30ee7	738	int err;
9d9e6a58 TG	739	rtnl_register(PF_UNSPEC, RTM_NEWRULE, fib_nl_newrule, NULL);
9d9e6a58 TG	740	rtnl_register(PF_UNSPEC, RTM_DELRULE, fib_nl_delrule, NULL);
c454673d	741	rtnl_register(PF_UNSPEC, RTM_GETRULE, NULL, fib_nl_dumprule);
9d9e6a58	742
5d6d4809	743	err = register_pernet_subsys(&fib_rules_net_ops);
5fd30ee7 DL	744	if (err < 0)
	745	goto fail;
	746
5d6d4809	747	err = register_netdevice_notifier(&fib_rules_notifier);
5fd30ee7 DL	748	if (err < 0)
5fd30ee7 DL	749	goto fail_unregister;
5d6d4809	750
5fd30ee7 DL	751	return 0;
	752
	753	fail_unregister:
5d6d4809	754	unregister_pernet_subsys(&fib_rules_net_ops);
5fd30ee7 DL	755	fail:
	756	rtnl_unregister(PF_UNSPEC, RTM_NEWRULE);
	757	rtnl_unregister(PF_UNSPEC, RTM_DELRULE);
	758	rtnl_unregister(PF_UNSPEC, RTM_GETRULE);
	759	return err;
14c0b97d TG	760	}
	761
	762	subsys_initcall(fib_rules_init);