[mirror_ubuntu-zesty-kernel.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>

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

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

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

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

	return NULL;
}

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

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;

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

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

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

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

	return err;
}

EXPORT_SYMBOL_GPL(fib_rules_register);

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

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

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

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

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

	synchronize_rcu();

	return err;
}

EXPORT_SYMBOL_GPL(fib_rules_unregister);

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

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

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

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

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

	rcu_read_lock();

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

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

			target = rcu_dereference(rule->ctarget);
			if (target == NULL) {
				continue;
			} else {
				rule = target;
				goto jumped;
			}
		} else 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(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();

	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 fib_rule_hdr *frh = nlmsg_data(nlh);
	struct fib_rules_ops *ops = NULL;
	struct fib_rule *rule, *tmp;
	struct nlattr *tb[FRA_MAX+1];
	int err = -EINVAL;

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

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

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

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

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

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

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

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

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

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

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

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

		list_del_rcu(&rule->list);

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

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

		synchronize_rcu();
		notify_rule_change(RTM_DELRULE, rule, ops, nlh,
				   NETLINK_CB(skb).pid);
		fib_rule_put(rule);
		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 fib_rules_ops *ops;
	int idx = 0, family;

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

		return dump_rules(skb, cb, ops);
	}

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

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

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

	return skb->len;
}

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

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

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

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

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

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

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


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

	if (dev->nd_net != &init_net)
		return NOTIFY_DONE;

	ASSERT_RTNL();
	rcu_read_lock();

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

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

	rcu_read_unlock();

	return NOTIFY_DONE;
}

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

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

	return register_netdevice_notifier(&fib_rules_notifier);
}

subsys_initcall(fib_rules_init);
Commit	Line	Data
14c0b97d TG	1	/*
	2	* net/core/fib_rules.c Generic Routing Rules
	3	*
	4	* This program is free software; you can redistribute it and/or
	5	* modify it under the terms of the GNU General Public License as
	6	* published by the Free Software Foundation, version 2.
	7	*
	8	* Authors: Thomas Graf <tgraf@suug.ch>
	9	*/
	10
14c0b97d TG	11	#include <linux/types.h>
	12	#include <linux/kernel.h>
	13	#include <linux/list.h>
e9dc8653	14	#include <net/net_namespace.h>
881d966b	15	#include <net/sock.h>
14c0b97d TG	16	#include <net/fib_rules.h>
	17
	18	static LIST_HEAD(rules_ops);
	19	static DEFINE_SPINLOCK(rules_mod_lock);
	20
	21	static void notify_rule_change(int event, struct fib_rule *rule,
c17084d2 TG	22	struct fib_rules_ops ops, struct nlmsghdr nlh,
c17084d2 TG	23	u32 pid);
14c0b97d TG	24
	25	static struct fib_rules_ops *lookup_rules_ops(int family)
	26	{
	27	struct fib_rules_ops *ops;
	28
	29	rcu_read_lock();
	30	list_for_each_entry_rcu(ops, &rules_ops, list) {
	31	if (ops->family == family) {
	32	if (!try_module_get(ops->owner))
	33	ops = NULL;
	34	rcu_read_unlock();
	35	return ops;
	36	}
	37	}
	38	rcu_read_unlock();
	39
	40	return NULL;
	41	}
	42
	43	static void rules_ops_put(struct fib_rules_ops *ops)
	44	{
	45	if (ops)
	46	module_put(ops->owner);
	47	}
	48
73417f61 TG	49	static void flush_route_cache(struct fib_rules_ops *ops)
	50	{
	51	if (ops->flush_cache)
	52	ops->flush_cache();
	53	}
	54
14c0b97d TG	55	int fib_rules_register(struct fib_rules_ops *ops)
	56	{
	57	int err = -EEXIST;
	58	struct fib_rules_ops *o;
	59
	60	if (ops->rule_size < sizeof(struct fib_rule))
	61	return -EINVAL;
	62
	63	if (ops->match == NULL \|\| ops->configure == NULL \|\|
	64	ops->compare == NULL \|\| ops->fill == NULL \|\|
	65	ops->action == NULL)
	66	return -EINVAL;
	67
	68	spin_lock(&rules_mod_lock);
	69	list_for_each_entry(o, &rules_ops, list)
	70	if (ops->family == o->family)
	71	goto errout;
	72
	73	list_add_tail_rcu(&ops->list, &rules_ops);
	74	err = 0;
	75	errout:
	76	spin_unlock(&rules_mod_lock);
	77
	78	return err;
	79	}
	80
	81	EXPORT_SYMBOL_GPL(fib_rules_register);
	82
	83	static void cleanup_ops(struct fib_rules_ops *ops)
	84	{
	85	struct fib_rule rule, tmp;
	86
	87	list_for_each_entry_safe(rule, tmp, ops->rules_list, list) {
	88	list_del_rcu(&rule->list);
	89	fib_rule_put(rule);
	90	}
	91	}
	92
	93	int fib_rules_unregister(struct fib_rules_ops *ops)
	94	{
	95	int err = 0;
	96	struct fib_rules_ops *o;
	97
	98	spin_lock(&rules_mod_lock);
	99	list_for_each_entry(o, &rules_ops, list) {
	100	if (o == ops) {
	101	list_del_rcu(&o->list);
	102	cleanup_ops(ops);
	103	goto out;
	104	}
	105	}
	106
	107	err = -ENOENT;
	108	out:
	109	spin_unlock(&rules_mod_lock);
	110
	111	synchronize_rcu();
	112
	113	return err;
	114	}
	115
	116	EXPORT_SYMBOL_GPL(fib_rules_unregister);
	117
3dfbcc41 TG	118	static int fib_rule_match(struct fib_rule rule, struct fib_rules_ops ops,
	119	struct flowi *fl, int flags)
	120	{
	121	int ret = 0;
	122
	123	if (rule->ifindex && (rule->ifindex != fl->iif))
	124	goto out;
	125
	126	if ((rule->mark ^ fl->mark) & rule->mark_mask)
	127	goto out;
	128
	129	ret = ops->match(rule, fl, flags);
	130	out:
	131	return (rule->flags & FIB_RULE_INVERT) ? !ret : ret;
	132	}
	133
14c0b97d TG	134	int fib_rules_lookup(struct fib_rules_ops ops, struct flowi fl,
	135	int flags, struct fib_lookup_arg *arg)
	136	{
	137	struct fib_rule *rule;
	138	int err;
	139
	140	rcu_read_lock();
	141
	142	list_for_each_entry_rcu(rule, ops->rules_list, list) {
0947c9fe	143	jumped:
3dfbcc41	144	if (!fib_rule_match(rule, ops, fl, flags))
14c0b97d TG	145	continue;
14c0b97d TG	146
0947c9fe TG	147	if (rule->action == FR_ACT_GOTO) {
	148	struct fib_rule *target;
	149
	150	target = rcu_dereference(rule->ctarget);
	151	if (target == NULL) {
	152	continue;
	153	} else {
	154	rule = target;
	155	goto jumped;
	156	}
fa0b2d1d TG	157	} else if (rule->action == FR_ACT_NOP)
	158	continue;
	159	else
0947c9fe TG	160	err = ops->action(rule, fl, flags, arg);
0947c9fe TG	161
14c0b97d TG	162	if (err != -EAGAIN) {
	163	fib_rule_get(rule);
	164	arg->rule = rule;
	165	goto out;
	166	}
	167	}
	168
83886b6b	169	err = -ESRCH;
14c0b97d TG	170	out:
	171	rcu_read_unlock();
	172
	173	return err;
	174	}
	175
	176	EXPORT_SYMBOL_GPL(fib_rules_lookup);
	177
e1701c68 TG	178	static int validate_rulemsg(struct fib_rule_hdr frh, struct nlattr *tb,
	179	struct fib_rules_ops *ops)
	180	{
	181	int err = -EINVAL;
	182
	183	if (frh->src_len)
	184	if (tb[FRA_SRC] == NULL \|\|
	185	frh->src_len > (ops->addr_size * 8) \|\|
	186	nla_len(tb[FRA_SRC]) != ops->addr_size)
	187	goto errout;
	188
	189	if (frh->dst_len)
	190	if (tb[FRA_DST] == NULL \|\|
	191	frh->dst_len > (ops->addr_size * 8) \|\|
	192	nla_len(tb[FRA_DST]) != ops->addr_size)
	193	goto errout;
	194
	195	err = 0;
	196	errout:
	197	return err;
	198	}
	199
9d9e6a58	200	static int fib_nl_newrule(struct sk_buff skb, struct nlmsghdr nlh, void *arg)
14c0b97d	201	{
881d966b	202	struct net *net = skb->sk->sk_net;
14c0b97d TG	203	struct fib_rule_hdr *frh = nlmsg_data(nlh);
	204	struct fib_rules_ops *ops = NULL;
	205	struct fib_rule rule, r, *last = NULL;
	206	struct nlattr *tb[FRA_MAX+1];
0947c9fe	207	int err = -EINVAL, unresolved = 0;
14c0b97d TG	208
	209	if (nlh->nlmsg_len < nlmsg_msg_size(sizeof(*frh)))
	210	goto errout;
	211
	212	ops = lookup_rules_ops(frh->family);
	213	if (ops == NULL) {
	214	err = EAFNOSUPPORT;
	215	goto errout;
	216	}
	217
	218	err = nlmsg_parse(nlh, sizeof(*frh), tb, FRA_MAX, ops->policy);
	219	if (err < 0)
	220	goto errout;
	221
e1701c68 TG	222	err = validate_rulemsg(frh, tb, ops);
	223	if (err < 0)
	224	goto errout;
	225
14c0b97d TG	226	rule = kzalloc(ops->rule_size, GFP_KERNEL);
	227	if (rule == NULL) {
	228	err = -ENOMEM;
	229	goto errout;
	230	}
	231
	232	if (tb[FRA_PRIORITY])
	233	rule->pref = nla_get_u32(tb[FRA_PRIORITY]);
	234
	235	if (tb[FRA_IFNAME]) {
	236	struct net_device *dev;
	237
	238	rule->ifindex = -1;
5176f91e	239	nla_strlcpy(rule->ifname, tb[FRA_IFNAME], IFNAMSIZ);
881d966b	240	dev = __dev_get_by_name(net, rule->ifname);
14c0b97d TG	241	if (dev)
	242	rule->ifindex = dev->ifindex;
	243	}
	244
b8964ed9 TG	245	if (tb[FRA_FWMARK]) {
	246	rule->mark = nla_get_u32(tb[FRA_FWMARK]);
	247	if (rule->mark)
	248	/* compatibility: if the mark value is non-zero all bits
	249	* are compared unless a mask is explicitly specified.
	250	*/
	251	rule->mark_mask = 0xFFFFFFFF;
	252	}
	253
	254	if (tb[FRA_FWMASK])
	255	rule->mark_mask = nla_get_u32(tb[FRA_FWMASK]);
	256
14c0b97d TG	257	rule->action = frh->action;
14c0b97d TG	258	rule->flags = frh->flags;
9e762a4a	259	rule->table = frh_get_table(frh, tb);
14c0b97d TG	260
	261	if (!rule->pref && ops->default_pref)
	262	rule->pref = ops->default_pref();
	263
0947c9fe TG	264	err = -EINVAL;
	265	if (tb[FRA_GOTO]) {
	266	if (rule->action != FR_ACT_GOTO)
	267	goto errout_free;
	268
	269	rule->target = nla_get_u32(tb[FRA_GOTO]);
	270	/* Backward jumps are prohibited to avoid endless loops */
	271	if (rule->target <= rule->pref)
	272	goto errout_free;
	273
	274	list_for_each_entry(r, ops->rules_list, list) {
	275	if (r->pref == rule->target) {
	276	rule->ctarget = r;
	277	break;
	278	}
	279	}
	280
	281	if (rule->ctarget == NULL)
	282	unresolved = 1;
	283	} else if (rule->action == FR_ACT_GOTO)
	284	goto errout_free;
	285
14c0b97d TG	286	err = ops->configure(rule, skb, nlh, frh, tb);
	287	if (err < 0)
	288	goto errout_free;
	289
	290	list_for_each_entry(r, ops->rules_list, list) {
	291	if (r->pref > rule->pref)
	292	break;
	293	last = r;
	294	}
	295
	296	fib_rule_get(rule);
	297
0947c9fe TG	298	if (ops->unresolved_rules) {
	299	/*
	300	* There are unresolved goto rules in the list, check if
	301	* any of them are pointing to this new rule.
	302	*/
	303	list_for_each_entry(r, ops->rules_list, list) {
	304	if (r->action == FR_ACT_GOTO &&
	305	r->target == rule->pref) {
	306	BUG_ON(r->ctarget != NULL);
	307	rcu_assign_pointer(r->ctarget, rule);
	308	if (--ops->unresolved_rules == 0)
	309	break;
	310	}
	311	}
	312	}
	313
	314	if (rule->action == FR_ACT_GOTO)
	315	ops->nr_goto_rules++;
	316
	317	if (unresolved)
	318	ops->unresolved_rules++;
	319
14c0b97d TG	320	if (last)
	321	list_add_rcu(&rule->list, &last->list);
	322	else
	323	list_add_rcu(&rule->list, ops->rules_list);
	324
c17084d2	325	notify_rule_change(RTM_NEWRULE, rule, ops, nlh, NETLINK_CB(skb).pid);
73417f61	326	flush_route_cache(ops);
14c0b97d TG	327	rules_ops_put(ops);
	328	return 0;
	329
	330	errout_free:
	331	kfree(rule);
	332	errout:
	333	rules_ops_put(ops);
	334	return err;
	335	}
	336
9d9e6a58	337	static int fib_nl_delrule(struct sk_buff skb, struct nlmsghdr nlh, void *arg)
14c0b97d TG	338	{
	339	struct fib_rule_hdr *frh = nlmsg_data(nlh);
	340	struct fib_rules_ops *ops = NULL;
0947c9fe	341	struct fib_rule rule, tmp;
14c0b97d TG	342	struct nlattr *tb[FRA_MAX+1];
	343	int err = -EINVAL;
	344
	345	if (nlh->nlmsg_len < nlmsg_msg_size(sizeof(*frh)))
	346	goto errout;
	347
	348	ops = lookup_rules_ops(frh->family);
	349	if (ops == NULL) {
	350	err = EAFNOSUPPORT;
	351	goto errout;
	352	}
	353
	354	err = nlmsg_parse(nlh, sizeof(*frh), tb, FRA_MAX, ops->policy);
	355	if (err < 0)
	356	goto errout;
	357
e1701c68 TG	358	err = validate_rulemsg(frh, tb, ops);
	359	if (err < 0)
	360	goto errout;
	361
14c0b97d TG	362	list_for_each_entry(rule, ops->rules_list, list) {
	363	if (frh->action && (frh->action != rule->action))
	364	continue;
	365
9e762a4a	366	if (frh->table && (frh_get_table(frh, tb) != rule->table))
14c0b97d TG	367	continue;
	368
	369	if (tb[FRA_PRIORITY] &&
	370	(rule->pref != nla_get_u32(tb[FRA_PRIORITY])))
	371	continue;
	372
	373	if (tb[FRA_IFNAME] &&
	374	nla_strcmp(tb[FRA_IFNAME], rule->ifname))
	375	continue;
	376
b8964ed9 TG	377	if (tb[FRA_FWMARK] &&
	378	(rule->mark != nla_get_u32(tb[FRA_FWMARK])))
	379	continue;
	380
	381	if (tb[FRA_FWMASK] &&
	382	(rule->mark_mask != nla_get_u32(tb[FRA_FWMASK])))
	383	continue;
	384
14c0b97d TG	385	if (!ops->compare(rule, frh, tb))
	386	continue;
	387
	388	if (rule->flags & FIB_RULE_PERMANENT) {
	389	err = -EPERM;
	390	goto errout;
	391	}
	392
	393	list_del_rcu(&rule->list);
0947c9fe TG	394
	395	if (rule->action == FR_ACT_GOTO)
	396	ops->nr_goto_rules--;
	397
	398	/*
	399	* Check if this rule is a target to any of them. If so,
	400	* disable them. As this operation is eventually very
	401	* expensive, it is only performed if goto rules have
	402	* actually been added.
	403	*/
	404	if (ops->nr_goto_rules > 0) {
	405	list_for_each_entry(tmp, ops->rules_list, list) {
	406	if (tmp->ctarget == rule) {
	407	rcu_assign_pointer(tmp->ctarget, NULL);
	408	ops->unresolved_rules++;
	409	}
	410	}
	411	}
	412
14c0b97d	413	synchronize_rcu();
c17084d2 TG	414	notify_rule_change(RTM_DELRULE, rule, ops, nlh,
c17084d2 TG	415	NETLINK_CB(skb).pid);
14c0b97d	416	fib_rule_put(rule);
73417f61	417	flush_route_cache(ops);
14c0b97d TG	418	rules_ops_put(ops);
	419	return 0;
	420	}
	421
	422	err = -ENOENT;
	423	errout:
	424	rules_ops_put(ops);
	425	return err;
	426	}
	427
339bf98f TG	428	static inline size_t fib_rule_nlmsg_size(struct fib_rules_ops *ops,
	429	struct fib_rule *rule)
	430	{
	431	size_t payload = NLMSG_ALIGN(sizeof(struct fib_rule_hdr))
	432	+ nla_total_size(IFNAMSIZ) /* FRA_IFNAME */
	433	+ nla_total_size(4) /* FRA_PRIORITY */
	434	+ nla_total_size(4) /* FRA_TABLE */
	435	+ nla_total_size(4) /* FRA_FWMARK */
	436	+ nla_total_size(4); /* FRA_FWMASK */
	437
	438	if (ops->nlmsg_payload)
	439	payload += ops->nlmsg_payload(rule);
	440
	441	return payload;
	442	}
	443
14c0b97d TG	444	static int fib_nl_fill_rule(struct sk_buff skb, struct fib_rule rule,
	445	u32 pid, u32 seq, int type, int flags,
	446	struct fib_rules_ops *ops)
	447	{
	448	struct nlmsghdr *nlh;
	449	struct fib_rule_hdr *frh;
	450
	451	nlh = nlmsg_put(skb, pid, seq, type, sizeof(*frh), flags);
	452	if (nlh == NULL)
26932566	453	return -EMSGSIZE;
14c0b97d TG	454
	455	frh = nlmsg_data(nlh);
	456	frh->table = rule->table;
9e762a4a	457	NLA_PUT_U32(skb, FRA_TABLE, rule->table);
14c0b97d TG	458	frh->res1 = 0;
	459	frh->res2 = 0;
	460	frh->action = rule->action;
	461	frh->flags = rule->flags;
	462
0947c9fe TG	463	if (rule->action == FR_ACT_GOTO && rule->ctarget == NULL)
	464	frh->flags \|= FIB_RULE_UNRESOLVED;
	465
2b443683	466	if (rule->ifname[0]) {
14c0b97d TG	467	NLA_PUT_STRING(skb, FRA_IFNAME, rule->ifname);
14c0b97d TG	468
2b443683 TG	469	if (rule->ifindex == -1)
	470	frh->flags \|= FIB_RULE_DEV_DETACHED;
	471	}
	472
14c0b97d TG	473	if (rule->pref)
	474	NLA_PUT_U32(skb, FRA_PRIORITY, rule->pref);
	475
b8964ed9 TG	476	if (rule->mark)
	477	NLA_PUT_U32(skb, FRA_FWMARK, rule->mark);
	478
	479	if (rule->mark_mask \|\| rule->mark)
	480	NLA_PUT_U32(skb, FRA_FWMASK, rule->mark_mask);
	481
0947c9fe TG	482	if (rule->target)
	483	NLA_PUT_U32(skb, FRA_GOTO, rule->target);
	484
14c0b97d TG	485	if (ops->fill(rule, skb, nlh, frh) < 0)
	486	goto nla_put_failure;
	487
	488	return nlmsg_end(skb, nlh);
	489
	490	nla_put_failure:
26932566 PM	491	nlmsg_cancel(skb, nlh);
26932566 PM	492	return -EMSGSIZE;
14c0b97d TG	493	}
14c0b97d TG	494
c454673d TG	495	static int dump_rules(struct sk_buff skb, struct netlink_callback cb,
c454673d TG	496	struct fib_rules_ops *ops)
14c0b97d TG	497	{
	498	int idx = 0;
	499	struct fib_rule *rule;
14c0b97d	500
6313c1e0	501	list_for_each_entry(rule, ops->rules_list, list) {
c454673d	502	if (idx < cb->args[1])
14c0b97d TG	503	goto skip;
	504
	505	if (fib_nl_fill_rule(skb, rule, NETLINK_CB(cb->skb).pid,
	506	cb->nlh->nlmsg_seq, RTM_NEWRULE,
	507	NLM_F_MULTI, ops) < 0)
	508	break;
	509	skip:
	510	idx++;
	511	}
c454673d	512	cb->args[1] = idx;
14c0b97d TG	513	rules_ops_put(ops);
	514
	515	return skb->len;
	516	}
	517
c454673d TG	518	static int fib_nl_dumprule(struct sk_buff skb, struct netlink_callback cb)
	519	{
	520	struct fib_rules_ops *ops;
	521	int idx = 0, family;
	522
	523	family = rtnl_msg_family(cb->nlh);
	524	if (family != AF_UNSPEC) {
	525	/* Protocol specific dump request */
	526	ops = lookup_rules_ops(family);
	527	if (ops == NULL)
	528	return -EAFNOSUPPORT;
	529
	530	return dump_rules(skb, cb, ops);
	531	}
	532
	533	rcu_read_lock();
	534	list_for_each_entry_rcu(ops, &rules_ops, list) {
	535	if (idx < cb->args[0] \|\| !try_module_get(ops->owner))
	536	goto skip;
	537
	538	if (dump_rules(skb, cb, ops) < 0)
	539	break;
	540
	541	cb->args[1] = 0;
	542	skip:
	543	idx++;
	544	}
	545	rcu_read_unlock();
	546	cb->args[0] = idx;
	547
	548	return skb->len;
	549	}
14c0b97d TG	550
14c0b97d TG	551	static void notify_rule_change(int event, struct fib_rule *rule,
c17084d2 TG	552	struct fib_rules_ops ops, struct nlmsghdr nlh,
c17084d2 TG	553	u32 pid)
14c0b97d	554	{
c17084d2 TG	555	struct sk_buff *skb;
c17084d2 TG	556	int err = -ENOBUFS;
14c0b97d	557
339bf98f	558	skb = nlmsg_new(fib_rule_nlmsg_size(ops, rule), GFP_KERNEL);
14c0b97d	559	if (skb == NULL)
c17084d2 TG	560	goto errout;
	561
	562	err = fib_nl_fill_rule(skb, rule, pid, nlh->nlmsg_seq, event, 0, ops);
26932566 PM	563	if (err < 0) {
	564	/* -EMSGSIZE implies BUG in fib_rule_nlmsg_size() */
	565	WARN_ON(err == -EMSGSIZE);
	566	kfree_skb(skb);
	567	goto errout;
	568	}
c17084d2 TG	569	err = rtnl_notify(skb, pid, ops->nlgroup, nlh, GFP_KERNEL);
	570	errout:
	571	if (err < 0)
	572	rtnl_set_sk_err(ops->nlgroup, err);
14c0b97d TG	573	}
	574
	575	static void attach_rules(struct list_head rules, struct net_device dev)
	576	{
	577	struct fib_rule *rule;
	578
	579	list_for_each_entry(rule, rules, list) {
	580	if (rule->ifindex == -1 &&
	581	strcmp(dev->name, rule->ifname) == 0)
	582	rule->ifindex = dev->ifindex;
	583	}
	584	}
	585
	586	static void detach_rules(struct list_head rules, struct net_device dev)
	587	{
	588	struct fib_rule *rule;
	589
	590	list_for_each_entry(rule, rules, list)
	591	if (rule->ifindex == dev->ifindex)
	592	rule->ifindex = -1;
	593	}
	594
	595
	596	static int fib_rules_event(struct notifier_block *this, unsigned long event,
	597	void *ptr)
	598	{
	599	struct net_device *dev = ptr;
	600	struct fib_rules_ops *ops;
	601
e9dc8653 EB	602	if (dev->nd_net != &init_net)
	603	return NOTIFY_DONE;
	604
14c0b97d TG	605	ASSERT_RTNL();
	606	rcu_read_lock();
	607
	608	switch (event) {
	609	case NETDEV_REGISTER:
	610	list_for_each_entry(ops, &rules_ops, list)
	611	attach_rules(ops->rules_list, dev);
	612	break;
	613
	614	case NETDEV_UNREGISTER:
	615	list_for_each_entry(ops, &rules_ops, list)
	616	detach_rules(ops->rules_list, dev);
	617	break;
	618	}
	619
	620	rcu_read_unlock();
	621
	622	return NOTIFY_DONE;
	623	}
	624
	625	static struct notifier_block fib_rules_notifier = {
	626	.notifier_call = fib_rules_event,
	627	};
	628
	629	static int __init fib_rules_init(void)
	630	{
9d9e6a58 TG	631	rtnl_register(PF_UNSPEC, RTM_NEWRULE, fib_nl_newrule, NULL);
9d9e6a58 TG	632	rtnl_register(PF_UNSPEC, RTM_DELRULE, fib_nl_delrule, NULL);
c454673d	633	rtnl_register(PF_UNSPEC, RTM_GETRULE, NULL, fib_nl_dumprule);
9d9e6a58	634
14c0b97d TG	635	return register_netdevice_notifier(&fib_rules_notifier);
	636	}
	637
	638	subsys_initcall(fib_rules_init);