[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/ip_tunnels.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 dst_entry *dst = skb_dst(skb);
	struct net *net = dev_net(dst->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, ip6_dst_idev(dst)->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);

	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;
		break;
	case SEG6_IPTUN_MODE_ENCAP:
		err = iptunnel_handle_offloads(skb, SKB_GSO_IPXIP6);
		if (err)
			return err;

		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_set_inner_transport_header(skb, skb_transport_offset(skb));
		skb_set_inner_protocol(skb, skb->protocol);
		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));

	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_ATOMIC);
	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>
5807b22c	19	#include <net/ip_tunnels.h>
6c8702c6 DL	20	#include <net/lwtunnel.h>
	21	#include <net/netevent.h>
	22	#include <net/netns/generic.h>
	23	#include <net/ip6_fib.h>
	24	#include <net/route.h>
	25	#include <net/seg6.h>
	26	#include <linux/seg6.h>
	27	#include <linux/seg6_iptunnel.h>
	28	#include <net/addrconf.h>
	29	#include <net/ip6_route.h>
6c8702c6	30	#include <net/dst_cache.h>
9baee834 DL	31	#ifdef CONFIG_IPV6_SEG6_HMAC
	32	#include <net/seg6_hmac.h>
	33	#endif
6c8702c6 DL	34
6c8702c6 DL	35	struct seg6_lwt {
6c8702c6	36	struct dst_cache cache;
6c8702c6 DL	37	struct seg6_iptunnel_encap tuninfo[0];
	38	};
	39
	40	static inline struct seg6_lwt seg6_lwt_lwtunnel(struct lwtunnel_state lwt)
	41	{
	42	return (struct seg6_lwt *)lwt->data;
	43	}
	44
	45	static inline struct seg6_iptunnel_encap *
	46	seg6_encap_lwtunnel(struct lwtunnel_state *lwt)
	47	{
	48	return seg6_lwt_lwtunnel(lwt)->tuninfo;
	49	}
	50
	51	static const struct nla_policy seg6_iptunnel_policy[SEG6_IPTUNNEL_MAX + 1] = {
	52	[SEG6_IPTUNNEL_SRH] = { .type = NLA_BINARY },
	53	};
	54
bb4005ba WY	55	static int nla_put_srh(struct sk_buff *skb, int attrtype,
bb4005ba WY	56	struct seg6_iptunnel_encap *tuninfo)
6c8702c6 DL	57	{
	58	struct seg6_iptunnel_encap *data;
	59	struct nlattr *nla;
	60	int len;
	61
	62	len = SEG6_IPTUN_ENCAP_SIZE(tuninfo);
	63
	64	nla = nla_reserve(skb, attrtype, len);
	65	if (!nla)
	66	return -EMSGSIZE;
	67
	68	data = nla_data(nla);
	69	memcpy(data, tuninfo, len);
	70
	71	return 0;
	72	}
	73
	74	static void set_tun_src(struct net net, struct net_device dev,
	75	struct in6_addr daddr, struct in6_addr saddr)
	76	{
	77	struct seg6_pernet_data *sdata = seg6_pernet(net);
	78	struct in6_addr *tun_src;
	79
	80	rcu_read_lock();
	81
	82	tun_src = rcu_dereference(sdata->tun_src);
	83
	84	if (!ipv6_addr_any(tun_src)) {
	85	memcpy(saddr, tun_src, sizeof(struct in6_addr));
	86	} else {
	87	ipv6_dev_get_saddr(net, dev, daddr, IPV6_PREFER_SRC_PUBLIC,
	88	saddr);
	89	}
	90
	91	rcu_read_unlock();
	92	}
	93
	94	/* encapsulate an IPv6 packet within an outer IPv6 header with a given SRH */
32d99d0b	95	int seg6_do_srh_encap(struct sk_buff skb, struct ipv6_sr_hdr osrh, int proto)
6c8702c6	96	{
8936ef76 DL	97	struct dst_entry *dst = skb_dst(skb);
8936ef76 DL	98	struct net *net = dev_net(dst->dev);
6c8702c6 DL	99	struct ipv6hdr hdr, inner_hdr;
	100	struct ipv6_sr_hdr *isrh;
	101	int hdrlen, tot_len, err;
	102
	103	hdrlen = (osrh->hdrlen + 1) << 3;
	104	tot_len = hdrlen + sizeof(*hdr);
	105
19d5a26f	106	err = skb_cow_head(skb, tot_len);
6c8702c6 DL	107	if (unlikely(err))
	108	return err;
	109
	110	inner_hdr = ipv6_hdr(skb);
	111
	112	skb_push(skb, tot_len);
	113	skb_reset_network_header(skb);
	114	skb_mac_header_rebuild(skb);
	115	hdr = ipv6_hdr(skb);
	116
	117	/* inherit tc, flowlabel and hlim
	118	* hlim will be decremented in ip6_forward() afterwards and
	119	* decapsulation will overwrite inner hlim with outer hlim
	120	*/
32d99d0b DL	121
	122	if (skb->protocol == htons(ETH_P_IPV6)) {
	123	ip6_flow_hdr(hdr, ip6_tclass(ip6_flowinfo(inner_hdr)),
	124	ip6_flowlabel(inner_hdr));
	125	hdr->hop_limit = inner_hdr->hop_limit;
	126	} else {
	127	ip6_flow_hdr(hdr, 0, 0);
	128	hdr->hop_limit = ip6_dst_hoplimit(skb_dst(skb));
	129	}
	130
6c8702c6 DL	131	hdr->nexthdr = NEXTHDR_ROUTING;
	132
	133	isrh = (void )hdr + sizeof(hdr);
	134	memcpy(isrh, osrh, hdrlen);
	135
32d99d0b	136	isrh->nexthdr = proto;
6c8702c6 DL	137
6c8702c6 DL	138	hdr->daddr = isrh->segments[isrh->first_segment];
8936ef76	139	set_tun_src(net, ip6_dst_idev(dst)->dev, &hdr->daddr, &hdr->saddr);
6c8702c6	140
9baee834 DL	141	#ifdef CONFIG_IPV6_SEG6_HMAC
	142	if (sr_has_hmac(isrh)) {
	143	err = seg6_push_hmac(net, &hdr->saddr, isrh);
	144	if (unlikely(err))
	145	return err;
	146	}
	147	#endif
	148
6c8702c6 DL	149	skb_postpush_rcsum(skb, hdr, tot_len);
	150
	151	return 0;
	152	}
b04c80d3	153	EXPORT_SYMBOL_GPL(seg6_do_srh_encap);
6c8702c6 DL	154
6c8702c6 DL	155	/* insert an SRH within an IPv6 packet, just after the IPv6 header */
b04c80d3	156	int seg6_do_srh_inline(struct sk_buff skb, struct ipv6_sr_hdr osrh)
6c8702c6 DL	157	{
	158	struct ipv6hdr hdr, oldhdr;
	159	struct ipv6_sr_hdr *isrh;
	160	int hdrlen, err;
	161
	162	hdrlen = (osrh->hdrlen + 1) << 3;
	163
19d5a26f	164	err = skb_cow_head(skb, hdrlen);
6c8702c6 DL	165	if (unlikely(err))
	166	return err;
	167
	168	oldhdr = ipv6_hdr(skb);
	169
	170	skb_pull(skb, sizeof(struct ipv6hdr));
	171	skb_postpull_rcsum(skb, skb_network_header(skb),
	172	sizeof(struct ipv6hdr));
	173
	174	skb_push(skb, sizeof(struct ipv6hdr) + hdrlen);
	175	skb_reset_network_header(skb);
	176	skb_mac_header_rebuild(skb);
	177
	178	hdr = ipv6_hdr(skb);
	179
	180	memmove(hdr, oldhdr, sizeof(*hdr));
	181
	182	isrh = (void )hdr + sizeof(hdr);
	183	memcpy(isrh, osrh, hdrlen);
	184
	185	isrh->nexthdr = hdr->nexthdr;
	186	hdr->nexthdr = NEXTHDR_ROUTING;
	187
	188	isrh->segments[0] = hdr->daddr;
	189	hdr->daddr = isrh->segments[isrh->first_segment];
	190
9baee834 DL	191	#ifdef CONFIG_IPV6_SEG6_HMAC
	192	if (sr_has_hmac(isrh)) {
	193	struct net *net = dev_net(skb_dst(skb)->dev);
	194
	195	err = seg6_push_hmac(net, &hdr->saddr, isrh);
	196	if (unlikely(err))
	197	return err;
	198	}
	199	#endif
	200
6c8702c6 DL	201	skb_postpush_rcsum(skb, hdr, sizeof(struct ipv6hdr) + hdrlen);
	202
	203	return 0;
	204	}
b04c80d3	205	EXPORT_SYMBOL_GPL(seg6_do_srh_inline);
6c8702c6 DL	206
	207	static int seg6_do_srh(struct sk_buff *skb)
	208	{
	209	struct dst_entry *dst = skb_dst(skb);
	210	struct seg6_iptunnel_encap *tinfo;
32d99d0b	211	int proto, err = 0;
6c8702c6 DL	212
	213	tinfo = seg6_encap_lwtunnel(dst->lwtstate);
	214
6c8702c6	215	switch (tinfo->mode) {
6c8702c6	216	case SEG6_IPTUN_MODE_INLINE:
32d99d0b DL	217	if (skb->protocol != htons(ETH_P_IPV6))
	218	return -EINVAL;
	219
6c8702c6	220	err = seg6_do_srh_inline(skb, tinfo->srh);
32d99d0b DL	221	if (err)
32d99d0b DL	222	return err;
6c8702c6	223	break;
6c8702c6	224	case SEG6_IPTUN_MODE_ENCAP:
5807b22c DL	225	err = iptunnel_handle_offloads(skb, SKB_GSO_IPXIP6);
	226	if (err)
	227	return err;
	228
32d99d0b DL	229	if (skb->protocol == htons(ETH_P_IPV6))
	230	proto = IPPROTO_IPV6;
	231	else if (skb->protocol == htons(ETH_P_IP))
	232	proto = IPPROTO_IPIP;
	233	else
	234	return -EINVAL;
	235
	236	err = seg6_do_srh_encap(skb, tinfo->srh, proto);
	237	if (err)
	238	return err;
	239
5807b22c DL	240	skb_set_inner_transport_header(skb, skb_transport_offset(skb));
5807b22c DL	241	skb_set_inner_protocol(skb, skb->protocol);
38ee7f2d DL	242	skb->protocol = htons(ETH_P_IPV6);
	243	break;
	244	case SEG6_IPTUN_MODE_L2ENCAP:
	245	if (!skb_mac_header_was_set(skb))
	246	return -EINVAL;
	247
	248	if (pskb_expand_head(skb, skb->mac_len, 0, GFP_ATOMIC) < 0)
	249	return -ENOMEM;
	250
	251	skb_mac_header_rebuild(skb);
	252	skb_push(skb, skb->mac_len);
	253
	254	err = seg6_do_srh_encap(skb, tinfo->srh, NEXTHDR_NONE);
	255	if (err)
	256	return err;
	257
32d99d0b	258	skb->protocol = htons(ETH_P_IPV6);
6c8702c6 DL	259	break;
	260	}
	261
6c8702c6 DL	262	ipv6_hdr(skb)->payload_len = htons(skb->len - sizeof(struct ipv6hdr));
	263	skb_set_transport_header(skb, sizeof(struct ipv6hdr));
	264
6c8702c6 DL	265	return 0;
	266	}
	267
bb4005ba	268	static int seg6_input(struct sk_buff *skb)
6c8702c6	269	{
af4a2209 DL	270	struct dst_entry *orig_dst = skb_dst(skb);
	271	struct dst_entry *dst = NULL;
	272	struct seg6_lwt *slwt;
6c8702c6 DL	273	int err;
	274
	275	err = seg6_do_srh(skb);
	276	if (unlikely(err)) {
	277	kfree_skb(skb);
	278	return err;
	279	}
	280
af4a2209 DL	281	slwt = seg6_lwt_lwtunnel(orig_dst->lwtstate);
af4a2209 DL	282
af4a2209 DL	283	preempt_disable();
	284	dst = dst_cache_get(&slwt->cache);
	285	preempt_enable();
af4a2209	286
6c8702c6	287	skb_dst_drop(skb);
af4a2209 DL	288
	289	if (!dst) {
	290	ip6_route_input(skb);
af4a2209 DL	291	dst = skb_dst(skb);
	292	if (!dst->error) {
	293	preempt_disable();
	294	dst_cache_set_ip6(&slwt->cache, dst,
	295	&ipv6_hdr(skb)->saddr);
	296	preempt_enable();
	297	}
af4a2209 DL	298	} else {
	299	skb_dst_set(skb, dst);
	300	}
6c8702c6	301
af3b5158 DL	302	err = skb_cow_head(skb, LL_RESERVED_SPACE(dst->dev));
	303	if (unlikely(err))
	304	return err;
	305
6c8702c6 DL	306	return dst_input(skb);
	307	}
	308
bb4005ba	309	static int seg6_output(struct net net, struct sock sk, struct sk_buff *skb)
6c8702c6 DL	310	{
	311	struct dst_entry *orig_dst = skb_dst(skb);
	312	struct dst_entry *dst = NULL;
	313	struct seg6_lwt *slwt;
	314	int err = -EINVAL;
	315
	316	err = seg6_do_srh(skb);
	317	if (unlikely(err))
	318	goto drop;
	319
	320	slwt = seg6_lwt_lwtunnel(orig_dst->lwtstate);
	321
fa79581e	322	preempt_disable();
6c8702c6	323	dst = dst_cache_get(&slwt->cache);
fa79581e	324	preempt_enable();
6c8702c6 DL	325
	326	if (unlikely(!dst)) {
	327	struct ipv6hdr *hdr = ipv6_hdr(skb);
	328	struct flowi6 fl6;
	329
	330	fl6.daddr = hdr->daddr;
	331	fl6.saddr = hdr->saddr;
	332	fl6.flowlabel = ip6_flowinfo(hdr);
	333	fl6.flowi6_mark = skb->mark;
	334	fl6.flowi6_proto = hdr->nexthdr;
	335
	336	dst = ip6_route_output(net, NULL, &fl6);
	337	if (dst->error) {
	338	err = dst->error;
	339	dst_release(dst);
	340	goto drop;
	341	}
	342
fa79581e	343	preempt_disable();
6c8702c6	344	dst_cache_set_ip6(&slwt->cache, dst, &fl6.saddr);
fa79581e	345	preempt_enable();
6c8702c6 DL	346	}
	347
	348	skb_dst_drop(skb);
	349	skb_dst_set(skb, dst);
	350
af3b5158 DL	351	err = skb_cow_head(skb, LL_RESERVED_SPACE(dst->dev));
	352	if (unlikely(err))
	353	goto drop;
	354
6c8702c6 DL	355	return dst_output(net, sk, skb);
	356	drop:
	357	kfree_skb(skb);
	358	return err;
	359	}
	360
30357d7d	361	static int seg6_build_state(struct nlattr *nla,
6c8702c6	362	unsigned int family, const void *cfg,
9ae28727 DA	363	struct lwtunnel_state **ts,
9ae28727 DA	364	struct netlink_ext_ack *extack)
6c8702c6 DL	365	{
	366	struct nlattr *tb[SEG6_IPTUNNEL_MAX + 1];
	367	struct seg6_iptunnel_encap *tuninfo;
	368	struct lwtunnel_state *newts;
	369	int tuninfo_len, min_size;
	370	struct seg6_lwt *slwt;
	371	int err;
	372
32d99d0b DL	373	if (family != AF_INET && family != AF_INET6)
	374	return -EINVAL;
	375
6c8702c6	376	err = nla_parse_nested(tb, SEG6_IPTUNNEL_MAX, nla,
9ae28727	377	seg6_iptunnel_policy, extack);
6c8702c6 DL	378
	379	if (err < 0)
	380	return err;
	381
	382	if (!tb[SEG6_IPTUNNEL_SRH])
	383	return -EINVAL;
	384
	385	tuninfo = nla_data(tb[SEG6_IPTUNNEL_SRH]);
	386	tuninfo_len = nla_len(tb[SEG6_IPTUNNEL_SRH]);
	387
	388	/* tuninfo must contain at least the iptunnel encap structure,
	389	* the SRH and one segment
	390	*/
	391	min_size = sizeof(*tuninfo) + sizeof(struct ipv6_sr_hdr) +
	392	sizeof(struct in6_addr);
	393	if (tuninfo_len < min_size)
	394	return -EINVAL;
	395
	396	switch (tuninfo->mode) {
6c8702c6	397	case SEG6_IPTUN_MODE_INLINE:
32d99d0b DL	398	if (family != AF_INET6)
	399	return -EINVAL;
	400
6c8702c6	401	break;
6c8702c6 DL	402	case SEG6_IPTUN_MODE_ENCAP:
6c8702c6 DL	403	break;
38ee7f2d DL	404	case SEG6_IPTUN_MODE_L2ENCAP:
38ee7f2d DL	405	break;
6c8702c6 DL	406	default:
	407	return -EINVAL;
	408	}
	409
	410	/* verify that SRH is consistent */
	411	if (!seg6_validate_srh(tuninfo->srh, tuninfo_len - sizeof(*tuninfo)))
	412	return -EINVAL;
	413
	414	newts = lwtunnel_state_alloc(tuninfo_len + sizeof(*slwt));
	415	if (!newts)
	416	return -ENOMEM;
	417
	418	slwt = seg6_lwt_lwtunnel(newts);
	419
191f86ca	420	err = dst_cache_init(&slwt->cache, GFP_ATOMIC);
6c8702c6 DL	421	if (err) {
	422	kfree(newts);
	423	return err;
	424	}
6c8702c6 DL	425
	426	memcpy(&slwt->tuninfo, tuninfo, tuninfo_len);
	427
	428	newts->type = LWTUNNEL_ENCAP_SEG6;
38ee7f2d DL	429	newts->flags \|= LWTUNNEL_STATE_INPUT_REDIRECT;
	430
	431	if (tuninfo->mode != SEG6_IPTUN_MODE_L2ENCAP)
	432	newts->flags \|= LWTUNNEL_STATE_OUTPUT_REDIRECT;
	433
6c8702c6 DL	434	newts->headroom = seg6_lwt_headroom(tuninfo);
	435
	436	*ts = newts;
	437
	438	return 0;
	439	}
	440
6c8702c6 DL	441	static void seg6_destroy_state(struct lwtunnel_state *lwt)
	442	{
	443	dst_cache_destroy(&seg6_lwt_lwtunnel(lwt)->cache);
	444	}
6c8702c6 DL	445
	446	static int seg6_fill_encap_info(struct sk_buff *skb,
	447	struct lwtunnel_state *lwtstate)
	448	{
	449	struct seg6_iptunnel_encap *tuninfo = seg6_encap_lwtunnel(lwtstate);
	450
	451	if (nla_put_srh(skb, SEG6_IPTUNNEL_SRH, tuninfo))
	452	return -EMSGSIZE;
	453
	454	return 0;
	455	}
	456
	457	static int seg6_encap_nlsize(struct lwtunnel_state *lwtstate)
	458	{
	459	struct seg6_iptunnel_encap *tuninfo = seg6_encap_lwtunnel(lwtstate);
	460
	461	return nla_total_size(SEG6_IPTUN_ENCAP_SIZE(tuninfo));
	462	}
	463
	464	static int seg6_encap_cmp(struct lwtunnel_state a, struct lwtunnel_state b)
	465	{
	466	struct seg6_iptunnel_encap *a_hdr = seg6_encap_lwtunnel(a);
	467	struct seg6_iptunnel_encap *b_hdr = seg6_encap_lwtunnel(b);
	468	int len = SEG6_IPTUN_ENCAP_SIZE(a_hdr);
	469
	470	if (len != SEG6_IPTUN_ENCAP_SIZE(b_hdr))
	471	return 1;
	472
	473	return memcmp(a_hdr, b_hdr, len);
	474	}
	475
	476	static const struct lwtunnel_encap_ops seg6_iptun_ops = {
	477	.build_state = seg6_build_state,
6c8702c6	478	.destroy_state = seg6_destroy_state,
6c8702c6 DL	479	.output = seg6_output,
	480	.input = seg6_input,
	481	.fill_encap = seg6_fill_encap_info,
	482	.get_encap_size = seg6_encap_nlsize,
	483	.cmp_encap = seg6_encap_cmp,
88ff7334	484	.owner = THIS_MODULE,
6c8702c6 DL	485	};
	486
	487	int __init seg6_iptunnel_init(void)
	488	{
	489	return lwtunnel_encap_add_ops(&seg6_iptun_ops, LWTUNNEL_ENCAP_SEG6);
	490	}
	491
	492	void seg6_iptunnel_exit(void)
	493	{
	494	lwtunnel_encap_del_ops(&seg6_iptun_ops, LWTUNNEL_ENCAP_SEG6);
	495	}