[mirror_ubuntu-jammy-kernel.git] / net / ipv6 / seg6_iptunnel.c

/*
 *  SR-IPv6 implementation
 *
 *  Author:
 *  David Lebrun <david.lebrun@uclouvain.be>
 *
 *
 *  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; either version
 *        2 of the License, or (at your option) any later version.
 */

#include <linux/types.h>
#include <linux/skbuff.h>
#include <linux/net.h>
#include <linux/module.h>
#include <net/ip.h>
#include <net/lwtunnel.h>
#include <net/netevent.h>
#include <net/netns/generic.h>
#include <net/ip6_fib.h>
#include <net/route.h>
#include <net/seg6.h>
#include <linux/seg6.h>
#include <linux/seg6_iptunnel.h>
#include <net/addrconf.h>
#include <net/ip6_route.h>
#include <net/dst_cache.h>
#ifdef CONFIG_IPV6_SEG6_HMAC
#include <net/seg6_hmac.h>
#endif

struct seg6_lwt {
	struct dst_cache cache;
	struct seg6_iptunnel_encap tuninfo[0];
};

static inline struct seg6_lwt *seg6_lwt_lwtunnel(struct lwtunnel_state *lwt)
{
	return (struct seg6_lwt *)lwt->data;
}

static inline struct seg6_iptunnel_encap *
seg6_encap_lwtunnel(struct lwtunnel_state *lwt)
{
	return seg6_lwt_lwtunnel(lwt)->tuninfo;
}

static const struct nla_policy seg6_iptunnel_policy[SEG6_IPTUNNEL_MAX + 1] = {
	[SEG6_IPTUNNEL_SRH]	= { .type = NLA_BINARY },
};

static int nla_put_srh(struct sk_buff *skb, int attrtype,
		       struct seg6_iptunnel_encap *tuninfo)
{
	struct seg6_iptunnel_encap *data;
	struct nlattr *nla;
	int len;

	len = SEG6_IPTUN_ENCAP_SIZE(tuninfo);

	nla = nla_reserve(skb, attrtype, len);
	if (!nla)
		return -EMSGSIZE;

	data = nla_data(nla);
	memcpy(data, tuninfo, len);

	return 0;
}

static void set_tun_src(struct net *net, struct net_device *dev,
			struct in6_addr *daddr, struct in6_addr *saddr)
{
	struct seg6_pernet_data *sdata = seg6_pernet(net);
	struct in6_addr *tun_src;

	rcu_read_lock();

	tun_src = rcu_dereference(sdata->tun_src);

	if (!ipv6_addr_any(tun_src)) {
		memcpy(saddr, tun_src, sizeof(struct in6_addr));
	} else {
		ipv6_dev_get_saddr(net, dev, daddr, IPV6_PREFER_SRC_PUBLIC,
				   saddr);
	}

	rcu_read_unlock();
}

/* encapsulate an IPv6 packet within an outer IPv6 header with a given SRH */
int seg6_do_srh_encap(struct sk_buff *skb, struct ipv6_sr_hdr *osrh, int proto)
{
	struct net *net = dev_net(skb_dst(skb)->dev);
	struct ipv6hdr *hdr, *inner_hdr;
	struct ipv6_sr_hdr *isrh;
	int hdrlen, tot_len, err;

	hdrlen = (osrh->hdrlen + 1) << 3;
	tot_len = hdrlen + sizeof(*hdr);

	err = skb_cow_head(skb, tot_len);
	if (unlikely(err))
		return err;

	inner_hdr = ipv6_hdr(skb);

	skb_push(skb, tot_len);
	skb_reset_network_header(skb);
	skb_mac_header_rebuild(skb);
	hdr = ipv6_hdr(skb);

	/* inherit tc, flowlabel and hlim
	 * hlim will be decremented in ip6_forward() afterwards and
	 * decapsulation will overwrite inner hlim with outer hlim
	 */

	if (skb->protocol == htons(ETH_P_IPV6)) {
		ip6_flow_hdr(hdr, ip6_tclass(ip6_flowinfo(inner_hdr)),
			     ip6_flowlabel(inner_hdr));
		hdr->hop_limit = inner_hdr->hop_limit;
	} else {
		ip6_flow_hdr(hdr, 0, 0);
		hdr->hop_limit = ip6_dst_hoplimit(skb_dst(skb));
	}

	hdr->nexthdr = NEXTHDR_ROUTING;

	isrh = (void *)hdr + sizeof(*hdr);
	memcpy(isrh, osrh, hdrlen);

	isrh->nexthdr = proto;

	hdr->daddr = isrh->segments[isrh->first_segment];
	set_tun_src(net, skb->dev, &hdr->daddr, &hdr->saddr);

#ifdef CONFIG_IPV6_SEG6_HMAC
	if (sr_has_hmac(isrh)) {
		err = seg6_push_hmac(net, &hdr->saddr, isrh);
		if (unlikely(err))
			return err;
	}
#endif

	skb_postpush_rcsum(skb, hdr, tot_len);

	return 0;
}
EXPORT_SYMBOL_GPL(seg6_do_srh_encap);

/* insert an SRH within an IPv6 packet, just after the IPv6 header */
int seg6_do_srh_inline(struct sk_buff *skb, struct ipv6_sr_hdr *osrh)
{
	struct ipv6hdr *hdr, *oldhdr;
	struct ipv6_sr_hdr *isrh;
	int hdrlen, err;

	hdrlen = (osrh->hdrlen + 1) << 3;

	err = skb_cow_head(skb, hdrlen);
	if (unlikely(err))
		return err;

	oldhdr = ipv6_hdr(skb);

	skb_pull(skb, sizeof(struct ipv6hdr));
	skb_postpull_rcsum(skb, skb_network_header(skb),
			   sizeof(struct ipv6hdr));

	skb_push(skb, sizeof(struct ipv6hdr) + hdrlen);
	skb_reset_network_header(skb);
	skb_mac_header_rebuild(skb);

	hdr = ipv6_hdr(skb);

	memmove(hdr, oldhdr, sizeof(*hdr));

	isrh = (void *)hdr + sizeof(*hdr);
	memcpy(isrh, osrh, hdrlen);

	isrh->nexthdr = hdr->nexthdr;
	hdr->nexthdr = NEXTHDR_ROUTING;

	isrh->segments[0] = hdr->daddr;
	hdr->daddr = isrh->segments[isrh->first_segment];

#ifdef CONFIG_IPV6_SEG6_HMAC
	if (sr_has_hmac(isrh)) {
		struct net *net = dev_net(skb_dst(skb)->dev);

		err = seg6_push_hmac(net, &hdr->saddr, isrh);
		if (unlikely(err))
			return err;
	}
#endif

	skb_postpush_rcsum(skb, hdr, sizeof(struct ipv6hdr) + hdrlen);

	return 0;
}
EXPORT_SYMBOL_GPL(seg6_do_srh_inline);

static int seg6_do_srh(struct sk_buff *skb)
{
	struct dst_entry *dst = skb_dst(skb);
	struct seg6_iptunnel_encap *tinfo;
	int proto, err = 0;

	tinfo = seg6_encap_lwtunnel(dst->lwtstate);

	if (likely(!skb->encapsulation)) {
		skb_reset_inner_headers(skb);
		skb->encapsulation = 1;
	}

	switch (tinfo->mode) {
	case SEG6_IPTUN_MODE_INLINE:
		if (skb->protocol != htons(ETH_P_IPV6))
			return -EINVAL;

		err = seg6_do_srh_inline(skb, tinfo->srh);
		if (err)
			return err;

		skb_reset_inner_headers(skb);
		break;
	case SEG6_IPTUN_MODE_ENCAP:
		if (skb->protocol == htons(ETH_P_IPV6))
			proto = IPPROTO_IPV6;
		else if (skb->protocol == htons(ETH_P_IP))
			proto = IPPROTO_IPIP;
		else
			return -EINVAL;

		err = seg6_do_srh_encap(skb, tinfo->srh, proto);
		if (err)
			return err;

		skb->protocol = htons(ETH_P_IPV6);
		break;
	case SEG6_IPTUN_MODE_L2ENCAP:
		if (!skb_mac_header_was_set(skb))
			return -EINVAL;

		if (pskb_expand_head(skb, skb->mac_len, 0, GFP_ATOMIC) < 0)
			return -ENOMEM;

		skb_mac_header_rebuild(skb);
		skb_push(skb, skb->mac_len);

		err = seg6_do_srh_encap(skb, tinfo->srh, NEXTHDR_NONE);
		if (err)
			return err;

		skb->protocol = htons(ETH_P_IPV6);
		break;
	}

	ipv6_hdr(skb)->payload_len = htons(skb->len - sizeof(struct ipv6hdr));
	skb_set_transport_header(skb, sizeof(struct ipv6hdr));

	skb_set_inner_protocol(skb, skb->protocol);

	return 0;
}

static int seg6_input(struct sk_buff *skb)
{
	struct dst_entry *orig_dst = skb_dst(skb);
	struct dst_entry *dst = NULL;
	struct seg6_lwt *slwt;
	int err;

	err = seg6_do_srh(skb);
	if (unlikely(err)) {
		kfree_skb(skb);
		return err;
	}

	slwt = seg6_lwt_lwtunnel(orig_dst->lwtstate);

	preempt_disable();
	dst = dst_cache_get(&slwt->cache);
	preempt_enable();

	skb_dst_drop(skb);

	if (!dst) {
		ip6_route_input(skb);
		dst = skb_dst(skb);
		if (!dst->error) {
			preempt_disable();
			dst_cache_set_ip6(&slwt->cache, dst,
					  &ipv6_hdr(skb)->saddr);
			preempt_enable();
		}
	} else {
		skb_dst_set(skb, dst);
	}

	err = skb_cow_head(skb, LL_RESERVED_SPACE(dst->dev));
	if (unlikely(err))
		return err;

	return dst_input(skb);
}

static int seg6_output(struct net *net, struct sock *sk, struct sk_buff *skb)
{
	struct dst_entry *orig_dst = skb_dst(skb);
	struct dst_entry *dst = NULL;
	struct seg6_lwt *slwt;
	int err = -EINVAL;

	err = seg6_do_srh(skb);
	if (unlikely(err))
		goto drop;

	slwt = seg6_lwt_lwtunnel(orig_dst->lwtstate);

	preempt_disable();
	dst = dst_cache_get(&slwt->cache);
	preempt_enable();

	if (unlikely(!dst)) {
		struct ipv6hdr *hdr = ipv6_hdr(skb);
		struct flowi6 fl6;

		fl6.daddr = hdr->daddr;
		fl6.saddr = hdr->saddr;
		fl6.flowlabel = ip6_flowinfo(hdr);
		fl6.flowi6_mark = skb->mark;
		fl6.flowi6_proto = hdr->nexthdr;

		dst = ip6_route_output(net, NULL, &fl6);
		if (dst->error) {
			err = dst->error;
			dst_release(dst);
			goto drop;
		}

		preempt_disable();
		dst_cache_set_ip6(&slwt->cache, dst, &fl6.saddr);
		preempt_enable();
	}

	skb_dst_drop(skb);
	skb_dst_set(skb, dst);

	err = skb_cow_head(skb, LL_RESERVED_SPACE(dst->dev));
	if (unlikely(err))
		goto drop;

	return dst_output(net, sk, skb);
drop:
	kfree_skb(skb);
	return err;
}

static int seg6_build_state(struct nlattr *nla,
			    unsigned int family, const void *cfg,
			    struct lwtunnel_state **ts,
			    struct netlink_ext_ack *extack)
{
	struct nlattr *tb[SEG6_IPTUNNEL_MAX + 1];
	struct seg6_iptunnel_encap *tuninfo;
	struct lwtunnel_state *newts;
	int tuninfo_len, min_size;
	struct seg6_lwt *slwt;
	int err;

	if (family != AF_INET && family != AF_INET6)
		return -EINVAL;

	err = nla_parse_nested(tb, SEG6_IPTUNNEL_MAX, nla,
			       seg6_iptunnel_policy, extack);

	if (err < 0)
		return err;

	if (!tb[SEG6_IPTUNNEL_SRH])
		return -EINVAL;

	tuninfo = nla_data(tb[SEG6_IPTUNNEL_SRH]);
	tuninfo_len = nla_len(tb[SEG6_IPTUNNEL_SRH]);

	/* tuninfo must contain at least the iptunnel encap structure,
	 * the SRH and one segment
	 */
	min_size = sizeof(*tuninfo) + sizeof(struct ipv6_sr_hdr) +
		   sizeof(struct in6_addr);
	if (tuninfo_len < min_size)
		return -EINVAL;

	switch (tuninfo->mode) {
	case SEG6_IPTUN_MODE_INLINE:
		if (family != AF_INET6)
			return -EINVAL;

		break;
	case SEG6_IPTUN_MODE_ENCAP:
		break;
	case SEG6_IPTUN_MODE_L2ENCAP:
		break;
	default:
		return -EINVAL;
	}

	/* verify that SRH is consistent */
	if (!seg6_validate_srh(tuninfo->srh, tuninfo_len - sizeof(*tuninfo)))
		return -EINVAL;

	newts = lwtunnel_state_alloc(tuninfo_len + sizeof(*slwt));
	if (!newts)
		return -ENOMEM;

	slwt = seg6_lwt_lwtunnel(newts);

	err = dst_cache_init(&slwt->cache, GFP_KERNEL);
	if (err) {
		kfree(newts);
		return err;
	}

	memcpy(&slwt->tuninfo, tuninfo, tuninfo_len);

	newts->type = LWTUNNEL_ENCAP_SEG6;
	newts->flags |= LWTUNNEL_STATE_INPUT_REDIRECT;

	if (tuninfo->mode != SEG6_IPTUN_MODE_L2ENCAP)
		newts->flags |= LWTUNNEL_STATE_OUTPUT_REDIRECT;

	newts->headroom = seg6_lwt_headroom(tuninfo);

	*ts = newts;

	return 0;
}

static void seg6_destroy_state(struct lwtunnel_state *lwt)
{
	dst_cache_destroy(&seg6_lwt_lwtunnel(lwt)->cache);
}

static int seg6_fill_encap_info(struct sk_buff *skb,
				struct lwtunnel_state *lwtstate)
{
	struct seg6_iptunnel_encap *tuninfo = seg6_encap_lwtunnel(lwtstate);

	if (nla_put_srh(skb, SEG6_IPTUNNEL_SRH, tuninfo))
		return -EMSGSIZE;

	return 0;
}

static int seg6_encap_nlsize(struct lwtunnel_state *lwtstate)
{
	struct seg6_iptunnel_encap *tuninfo = seg6_encap_lwtunnel(lwtstate);

	return nla_total_size(SEG6_IPTUN_ENCAP_SIZE(tuninfo));
}

static int seg6_encap_cmp(struct lwtunnel_state *a, struct lwtunnel_state *b)
{
	struct seg6_iptunnel_encap *a_hdr = seg6_encap_lwtunnel(a);
	struct seg6_iptunnel_encap *b_hdr = seg6_encap_lwtunnel(b);
	int len = SEG6_IPTUN_ENCAP_SIZE(a_hdr);

	if (len != SEG6_IPTUN_ENCAP_SIZE(b_hdr))
		return 1;

	return memcmp(a_hdr, b_hdr, len);
}

static const struct lwtunnel_encap_ops seg6_iptun_ops = {
	.build_state = seg6_build_state,
	.destroy_state = seg6_destroy_state,
	.output = seg6_output,
	.input = seg6_input,
	.fill_encap = seg6_fill_encap_info,
	.get_encap_size = seg6_encap_nlsize,
	.cmp_encap = seg6_encap_cmp,
	.owner = THIS_MODULE,
};

int __init seg6_iptunnel_init(void)
{
	return lwtunnel_encap_add_ops(&seg6_iptun_ops, LWTUNNEL_ENCAP_SEG6);
}

void seg6_iptunnel_exit(void)
{
	lwtunnel_encap_del_ops(&seg6_iptun_ops, LWTUNNEL_ENCAP_SEG6);
}
Commit	Line	Data
6c8702c6 DL	1	/*
	2	* SR-IPv6 implementation
	3	*
	4	* Author:
	5	* David Lebrun <david.lebrun@uclouvain.be>
	6	*
	7	*
	8	* This program is free software; you can redistribute it and/or
	9	* modify it under the terms of the GNU General Public License
	10	* as published by the Free Software Foundation; either version
	11	* 2 of the License, or (at your option) any later version.
	12	*/
	13
	14	#include <linux/types.h>
	15	#include <linux/skbuff.h>
	16	#include <linux/net.h>
	17	#include <linux/module.h>
	18	#include <net/ip.h>
	19	#include <net/lwtunnel.h>
	20	#include <net/netevent.h>
	21	#include <net/netns/generic.h>
	22	#include <net/ip6_fib.h>
	23	#include <net/route.h>
	24	#include <net/seg6.h>
	25	#include <linux/seg6.h>
	26	#include <linux/seg6_iptunnel.h>
	27	#include <net/addrconf.h>
	28	#include <net/ip6_route.h>
6c8702c6	29	#include <net/dst_cache.h>
9baee834 DL	30	#ifdef CONFIG_IPV6_SEG6_HMAC
	31	#include <net/seg6_hmac.h>
	32	#endif
6c8702c6 DL	33
6c8702c6 DL	34	struct seg6_lwt {
6c8702c6	35	struct dst_cache cache;
6c8702c6 DL	36	struct seg6_iptunnel_encap tuninfo[0];
	37	};
	38
	39	static inline struct seg6_lwt seg6_lwt_lwtunnel(struct lwtunnel_state lwt)
	40	{
	41	return (struct seg6_lwt *)lwt->data;
	42	}
	43
	44	static inline struct seg6_iptunnel_encap *
	45	seg6_encap_lwtunnel(struct lwtunnel_state *lwt)
	46	{
	47	return seg6_lwt_lwtunnel(lwt)->tuninfo;
	48	}
	49
	50	static const struct nla_policy seg6_iptunnel_policy[SEG6_IPTUNNEL_MAX + 1] = {
	51	[SEG6_IPTUNNEL_SRH] = { .type = NLA_BINARY },
	52	};
	53
bb4005ba WY	54	static int nla_put_srh(struct sk_buff *skb, int attrtype,
bb4005ba WY	55	struct seg6_iptunnel_encap *tuninfo)
6c8702c6 DL	56	{
	57	struct seg6_iptunnel_encap *data;
	58	struct nlattr *nla;
	59	int len;
	60
	61	len = SEG6_IPTUN_ENCAP_SIZE(tuninfo);
	62
	63	nla = nla_reserve(skb, attrtype, len);
	64	if (!nla)
	65	return -EMSGSIZE;
	66
	67	data = nla_data(nla);
	68	memcpy(data, tuninfo, len);
	69
	70	return 0;
	71	}
	72
	73	static void set_tun_src(struct net net, struct net_device dev,
	74	struct in6_addr daddr, struct in6_addr saddr)
	75	{
	76	struct seg6_pernet_data *sdata = seg6_pernet(net);
	77	struct in6_addr *tun_src;
	78
	79	rcu_read_lock();
	80
	81	tun_src = rcu_dereference(sdata->tun_src);
	82
	83	if (!ipv6_addr_any(tun_src)) {
	84	memcpy(saddr, tun_src, sizeof(struct in6_addr));
	85	} else {
	86	ipv6_dev_get_saddr(net, dev, daddr, IPV6_PREFER_SRC_PUBLIC,
	87	saddr);
	88	}
	89
	90	rcu_read_unlock();
	91	}
	92
	93	/* encapsulate an IPv6 packet within an outer IPv6 header with a given SRH */
32d99d0b	94	int seg6_do_srh_encap(struct sk_buff skb, struct ipv6_sr_hdr osrh, int proto)
6c8702c6 DL	95	{
	96	struct net *net = dev_net(skb_dst(skb)->dev);
	97	struct ipv6hdr hdr, inner_hdr;
	98	struct ipv6_sr_hdr *isrh;
	99	int hdrlen, tot_len, err;
	100
	101	hdrlen = (osrh->hdrlen + 1) << 3;
	102	tot_len = hdrlen + sizeof(*hdr);
	103
19d5a26f	104	err = skb_cow_head(skb, tot_len);
6c8702c6 DL	105	if (unlikely(err))
	106	return err;
	107
	108	inner_hdr = ipv6_hdr(skb);
	109
	110	skb_push(skb, tot_len);
	111	skb_reset_network_header(skb);
	112	skb_mac_header_rebuild(skb);
	113	hdr = ipv6_hdr(skb);
	114
	115	/* inherit tc, flowlabel and hlim
	116	* hlim will be decremented in ip6_forward() afterwards and
	117	* decapsulation will overwrite inner hlim with outer hlim
	118	*/
32d99d0b DL	119
	120	if (skb->protocol == htons(ETH_P_IPV6)) {
	121	ip6_flow_hdr(hdr, ip6_tclass(ip6_flowinfo(inner_hdr)),
	122	ip6_flowlabel(inner_hdr));
	123	hdr->hop_limit = inner_hdr->hop_limit;
	124	} else {
	125	ip6_flow_hdr(hdr, 0, 0);
	126	hdr->hop_limit = ip6_dst_hoplimit(skb_dst(skb));
	127	}
	128
6c8702c6 DL	129	hdr->nexthdr = NEXTHDR_ROUTING;
	130
	131	isrh = (void )hdr + sizeof(hdr);
	132	memcpy(isrh, osrh, hdrlen);
	133
32d99d0b	134	isrh->nexthdr = proto;
6c8702c6 DL	135
	136	hdr->daddr = isrh->segments[isrh->first_segment];
	137	set_tun_src(net, skb->dev, &hdr->daddr, &hdr->saddr);
	138
9baee834 DL	139	#ifdef CONFIG_IPV6_SEG6_HMAC
	140	if (sr_has_hmac(isrh)) {
	141	err = seg6_push_hmac(net, &hdr->saddr, isrh);
	142	if (unlikely(err))
	143	return err;
	144	}
	145	#endif
	146
6c8702c6 DL	147	skb_postpush_rcsum(skb, hdr, tot_len);
	148
	149	return 0;
	150	}
b04c80d3	151	EXPORT_SYMBOL_GPL(seg6_do_srh_encap);
6c8702c6 DL	152
6c8702c6 DL	153	/* insert an SRH within an IPv6 packet, just after the IPv6 header */
b04c80d3	154	int seg6_do_srh_inline(struct sk_buff skb, struct ipv6_sr_hdr osrh)
6c8702c6 DL	155	{
	156	struct ipv6hdr hdr, oldhdr;
	157	struct ipv6_sr_hdr *isrh;
	158	int hdrlen, err;
	159
	160	hdrlen = (osrh->hdrlen + 1) << 3;
	161
19d5a26f	162	err = skb_cow_head(skb, hdrlen);
6c8702c6 DL	163	if (unlikely(err))
	164	return err;
	165
	166	oldhdr = ipv6_hdr(skb);
	167
	168	skb_pull(skb, sizeof(struct ipv6hdr));
	169	skb_postpull_rcsum(skb, skb_network_header(skb),
	170	sizeof(struct ipv6hdr));
	171
	172	skb_push(skb, sizeof(struct ipv6hdr) + hdrlen);
	173	skb_reset_network_header(skb);
	174	skb_mac_header_rebuild(skb);
	175
	176	hdr = ipv6_hdr(skb);
	177
	178	memmove(hdr, oldhdr, sizeof(*hdr));
	179
	180	isrh = (void )hdr + sizeof(hdr);
	181	memcpy(isrh, osrh, hdrlen);
	182
	183	isrh->nexthdr = hdr->nexthdr;
	184	hdr->nexthdr = NEXTHDR_ROUTING;
	185
	186	isrh->segments[0] = hdr->daddr;
	187	hdr->daddr = isrh->segments[isrh->first_segment];
	188
9baee834 DL	189	#ifdef CONFIG_IPV6_SEG6_HMAC
	190	if (sr_has_hmac(isrh)) {
	191	struct net *net = dev_net(skb_dst(skb)->dev);
	192
	193	err = seg6_push_hmac(net, &hdr->saddr, isrh);
	194	if (unlikely(err))
	195	return err;
	196	}
	197	#endif
	198
6c8702c6 DL	199	skb_postpush_rcsum(skb, hdr, sizeof(struct ipv6hdr) + hdrlen);
	200
	201	return 0;
	202	}
b04c80d3	203	EXPORT_SYMBOL_GPL(seg6_do_srh_inline);
6c8702c6 DL	204
	205	static int seg6_do_srh(struct sk_buff *skb)
	206	{
	207	struct dst_entry *dst = skb_dst(skb);
	208	struct seg6_iptunnel_encap *tinfo;
32d99d0b	209	int proto, err = 0;
6c8702c6 DL	210
	211	tinfo = seg6_encap_lwtunnel(dst->lwtstate);
	212
	213	if (likely(!skb->encapsulation)) {
	214	skb_reset_inner_headers(skb);
	215	skb->encapsulation = 1;
	216	}
	217
	218	switch (tinfo->mode) {
6c8702c6	219	case SEG6_IPTUN_MODE_INLINE:
32d99d0b DL	220	if (skb->protocol != htons(ETH_P_IPV6))
	221	return -EINVAL;
	222
6c8702c6	223	err = seg6_do_srh_inline(skb, tinfo->srh);
32d99d0b DL	224	if (err)
	225	return err;
	226
6c8702c6 DL	227	skb_reset_inner_headers(skb);
6c8702c6 DL	228	break;
6c8702c6	229	case SEG6_IPTUN_MODE_ENCAP:
32d99d0b DL	230	if (skb->protocol == htons(ETH_P_IPV6))
	231	proto = IPPROTO_IPV6;
	232	else if (skb->protocol == htons(ETH_P_IP))
	233	proto = IPPROTO_IPIP;
	234	else
	235	return -EINVAL;
	236
	237	err = seg6_do_srh_encap(skb, tinfo->srh, proto);
	238	if (err)
	239	return err;
	240
38ee7f2d DL	241	skb->protocol = htons(ETH_P_IPV6);
	242	break;
	243	case SEG6_IPTUN_MODE_L2ENCAP:
	244	if (!skb_mac_header_was_set(skb))
	245	return -EINVAL;
	246
	247	if (pskb_expand_head(skb, skb->mac_len, 0, GFP_ATOMIC) < 0)
	248	return -ENOMEM;
	249
	250	skb_mac_header_rebuild(skb);
	251	skb_push(skb, skb->mac_len);
	252
	253	err = seg6_do_srh_encap(skb, tinfo->srh, NEXTHDR_NONE);
	254	if (err)
	255	return err;
	256
32d99d0b	257	skb->protocol = htons(ETH_P_IPV6);
6c8702c6 DL	258	break;
	259	}
	260
6c8702c6 DL	261	ipv6_hdr(skb)->payload_len = htons(skb->len - sizeof(struct ipv6hdr));
	262	skb_set_transport_header(skb, sizeof(struct ipv6hdr));
	263
	264	skb_set_inner_protocol(skb, skb->protocol);
	265
	266	return 0;
	267	}
	268
bb4005ba	269	static int seg6_input(struct sk_buff *skb)
6c8702c6	270	{
af4a2209 DL	271	struct dst_entry *orig_dst = skb_dst(skb);
	272	struct dst_entry *dst = NULL;
	273	struct seg6_lwt *slwt;
6c8702c6 DL	274	int err;
	275
	276	err = seg6_do_srh(skb);
	277	if (unlikely(err)) {
	278	kfree_skb(skb);
	279	return err;
	280	}
	281
af4a2209 DL	282	slwt = seg6_lwt_lwtunnel(orig_dst->lwtstate);
af4a2209 DL	283
af4a2209 DL	284	preempt_disable();
	285	dst = dst_cache_get(&slwt->cache);
	286	preempt_enable();
af4a2209	287
6c8702c6	288	skb_dst_drop(skb);
af4a2209 DL	289
	290	if (!dst) {
	291	ip6_route_input(skb);
af4a2209 DL	292	dst = skb_dst(skb);
	293	if (!dst->error) {
	294	preempt_disable();
	295	dst_cache_set_ip6(&slwt->cache, dst,
	296	&ipv6_hdr(skb)->saddr);
	297	preempt_enable();
	298	}
af4a2209 DL	299	} else {
	300	skb_dst_set(skb, dst);
	301	}
6c8702c6	302
af3b5158 DL	303	err = skb_cow_head(skb, LL_RESERVED_SPACE(dst->dev));
	304	if (unlikely(err))
	305	return err;
	306
6c8702c6 DL	307	return dst_input(skb);
	308	}
	309
bb4005ba	310	static int seg6_output(struct net net, struct sock sk, struct sk_buff *skb)
6c8702c6 DL	311	{
	312	struct dst_entry *orig_dst = skb_dst(skb);
	313	struct dst_entry *dst = NULL;
	314	struct seg6_lwt *slwt;
	315	int err = -EINVAL;
	316
	317	err = seg6_do_srh(skb);
	318	if (unlikely(err))
	319	goto drop;
	320
	321	slwt = seg6_lwt_lwtunnel(orig_dst->lwtstate);
	322
fa79581e	323	preempt_disable();
6c8702c6	324	dst = dst_cache_get(&slwt->cache);
fa79581e	325	preempt_enable();
6c8702c6 DL	326
	327	if (unlikely(!dst)) {
	328	struct ipv6hdr *hdr = ipv6_hdr(skb);
	329	struct flowi6 fl6;
	330
	331	fl6.daddr = hdr->daddr;
	332	fl6.saddr = hdr->saddr;
	333	fl6.flowlabel = ip6_flowinfo(hdr);
	334	fl6.flowi6_mark = skb->mark;
	335	fl6.flowi6_proto = hdr->nexthdr;
	336
	337	dst = ip6_route_output(net, NULL, &fl6);
	338	if (dst->error) {
	339	err = dst->error;
	340	dst_release(dst);
	341	goto drop;
	342	}
	343
fa79581e	344	preempt_disable();
6c8702c6	345	dst_cache_set_ip6(&slwt->cache, dst, &fl6.saddr);
fa79581e	346	preempt_enable();
6c8702c6 DL	347	}
	348
	349	skb_dst_drop(skb);
	350	skb_dst_set(skb, dst);
	351
af3b5158 DL	352	err = skb_cow_head(skb, LL_RESERVED_SPACE(dst->dev));
	353	if (unlikely(err))
	354	goto drop;
	355
6c8702c6 DL	356	return dst_output(net, sk, skb);
	357	drop:
	358	kfree_skb(skb);
	359	return err;
	360	}
	361
30357d7d	362	static int seg6_build_state(struct nlattr *nla,
6c8702c6	363	unsigned int family, const void *cfg,
9ae28727 DA	364	struct lwtunnel_state **ts,
9ae28727 DA	365	struct netlink_ext_ack *extack)
6c8702c6 DL	366	{
	367	struct nlattr *tb[SEG6_IPTUNNEL_MAX + 1];
	368	struct seg6_iptunnel_encap *tuninfo;
	369	struct lwtunnel_state *newts;
	370	int tuninfo_len, min_size;
	371	struct seg6_lwt *slwt;
	372	int err;
	373
32d99d0b DL	374	if (family != AF_INET && family != AF_INET6)
	375	return -EINVAL;
	376
6c8702c6	377	err = nla_parse_nested(tb, SEG6_IPTUNNEL_MAX, nla,
9ae28727	378	seg6_iptunnel_policy, extack);
6c8702c6 DL	379
	380	if (err < 0)
	381	return err;
	382
	383	if (!tb[SEG6_IPTUNNEL_SRH])
	384	return -EINVAL;
	385
	386	tuninfo = nla_data(tb[SEG6_IPTUNNEL_SRH]);
	387	tuninfo_len = nla_len(tb[SEG6_IPTUNNEL_SRH]);
	388
	389	/* tuninfo must contain at least the iptunnel encap structure,
	390	* the SRH and one segment
	391	*/
	392	min_size = sizeof(*tuninfo) + sizeof(struct ipv6_sr_hdr) +
	393	sizeof(struct in6_addr);
	394	if (tuninfo_len < min_size)
	395	return -EINVAL;
	396
	397	switch (tuninfo->mode) {
6c8702c6	398	case SEG6_IPTUN_MODE_INLINE:
32d99d0b DL	399	if (family != AF_INET6)
	400	return -EINVAL;
	401
6c8702c6	402	break;
6c8702c6 DL	403	case SEG6_IPTUN_MODE_ENCAP:
6c8702c6 DL	404	break;
38ee7f2d DL	405	case SEG6_IPTUN_MODE_L2ENCAP:
38ee7f2d DL	406	break;
6c8702c6 DL	407	default:
	408	return -EINVAL;
	409	}
	410
	411	/* verify that SRH is consistent */
	412	if (!seg6_validate_srh(tuninfo->srh, tuninfo_len - sizeof(*tuninfo)))
	413	return -EINVAL;
	414
	415	newts = lwtunnel_state_alloc(tuninfo_len + sizeof(*slwt));
	416	if (!newts)
	417	return -ENOMEM;
	418
	419	slwt = seg6_lwt_lwtunnel(newts);
	420
6c8702c6 DL	421	err = dst_cache_init(&slwt->cache, GFP_KERNEL);
	422	if (err) {
	423	kfree(newts);
	424	return err;
	425	}
6c8702c6 DL	426
	427	memcpy(&slwt->tuninfo, tuninfo, tuninfo_len);
	428
	429	newts->type = LWTUNNEL_ENCAP_SEG6;
38ee7f2d DL	430	newts->flags \|= LWTUNNEL_STATE_INPUT_REDIRECT;
	431
	432	if (tuninfo->mode != SEG6_IPTUN_MODE_L2ENCAP)
	433	newts->flags \|= LWTUNNEL_STATE_OUTPUT_REDIRECT;
	434
6c8702c6 DL	435	newts->headroom = seg6_lwt_headroom(tuninfo);
	436
	437	*ts = newts;
	438
	439	return 0;
	440	}
	441
6c8702c6 DL	442	static void seg6_destroy_state(struct lwtunnel_state *lwt)
	443	{
	444	dst_cache_destroy(&seg6_lwt_lwtunnel(lwt)->cache);
	445	}
6c8702c6 DL	446
	447	static int seg6_fill_encap_info(struct sk_buff *skb,
	448	struct lwtunnel_state *lwtstate)
	449	{
	450	struct seg6_iptunnel_encap *tuninfo = seg6_encap_lwtunnel(lwtstate);
	451
	452	if (nla_put_srh(skb, SEG6_IPTUNNEL_SRH, tuninfo))
	453	return -EMSGSIZE;
	454
	455	return 0;
	456	}
	457
	458	static int seg6_encap_nlsize(struct lwtunnel_state *lwtstate)
	459	{
	460	struct seg6_iptunnel_encap *tuninfo = seg6_encap_lwtunnel(lwtstate);
	461
	462	return nla_total_size(SEG6_IPTUN_ENCAP_SIZE(tuninfo));
	463	}
	464
	465	static int seg6_encap_cmp(struct lwtunnel_state a, struct lwtunnel_state b)
	466	{
	467	struct seg6_iptunnel_encap *a_hdr = seg6_encap_lwtunnel(a);
	468	struct seg6_iptunnel_encap *b_hdr = seg6_encap_lwtunnel(b);
	469	int len = SEG6_IPTUN_ENCAP_SIZE(a_hdr);
	470
	471	if (len != SEG6_IPTUN_ENCAP_SIZE(b_hdr))
	472	return 1;
	473
	474	return memcmp(a_hdr, b_hdr, len);
	475	}
	476
	477	static const struct lwtunnel_encap_ops seg6_iptun_ops = {
	478	.build_state = seg6_build_state,
6c8702c6	479	.destroy_state = seg6_destroy_state,
6c8702c6 DL	480	.output = seg6_output,
	481	.input = seg6_input,
	482	.fill_encap = seg6_fill_encap_info,
	483	.get_encap_size = seg6_encap_nlsize,
	484	.cmp_encap = seg6_encap_cmp,
88ff7334	485	.owner = THIS_MODULE,
6c8702c6 DL	486	};
	487
	488	int __init seg6_iptunnel_init(void)
	489	{
	490	return lwtunnel_encap_add_ops(&seg6_iptun_ops, LWTUNNEL_ENCAP_SEG6);
	491	}
	492
	493	void seg6_iptunnel_exit(void)
	494	{
	495	lwtunnel_encap_del_ops(&seg6_iptun_ops, LWTUNNEL_ENCAP_SEG6);
	496	}