[thirdparty/linux.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();
}

/* Compute flowlabel for outer IPv6 header */
static __be32 seg6_make_flowlabel(struct net *net, struct sk_buff *skb,
				  struct ipv6hdr *inner_hdr)
{
	int do_flowlabel = net->ipv6.sysctl.seg6_flowlabel;
	__be32 flowlabel = 0;
	u32 hash;

	if (do_flowlabel > 0) {
		hash = skb_get_hash(skb);
		hash = rol32(hash, 16);
		flowlabel = (__force __be32)hash & IPV6_FLOWLABEL_MASK;
	} else if (!do_flowlabel && skb->protocol == htons(ETH_P_IPV6)) {
		flowlabel = ip6_flowlabel(inner_hdr);
	}
	return flowlabel;
}

/* 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;
	__be32 flowlabel;

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

	err = skb_cow_head(skb, tot_len + skb->mac_len);
	if (unlikely(err))
		return err;

	inner_hdr = ipv6_hdr(skb);
	flowlabel = seg6_make_flowlabel(net, skb, inner_hdr);

	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)),
			     flowlabel);
		hdr->hop_limit = inner_hdr->hop_limit;
	} else {
		ip6_flow_hdr(hdr, 0, flowlabel);
		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, 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 + skb->mac_len);
	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;

		memset(&fl6, 0, sizeof(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
b5facfdb AA	94	/* Compute flowlabel for outer IPv6 header */
	95	static __be32 seg6_make_flowlabel(struct net net, struct sk_buff skb,
	96	struct ipv6hdr *inner_hdr)
	97	{
	98	int do_flowlabel = net->ipv6.sysctl.seg6_flowlabel;
	99	__be32 flowlabel = 0;
	100	u32 hash;
	101
	102	if (do_flowlabel > 0) {
	103	hash = skb_get_hash(skb);
3ee593ad	104	hash = rol32(hash, 16);
b5facfdb AA	105	flowlabel = (__force __be32)hash & IPV6_FLOWLABEL_MASK;
	106	} else if (!do_flowlabel && skb->protocol == htons(ETH_P_IPV6)) {
	107	flowlabel = ip6_flowlabel(inner_hdr);
	108	}
	109	return flowlabel;
	110	}
	111
6c8702c6	112	/* encapsulate an IPv6 packet within an outer IPv6 header with a given SRH */
32d99d0b	113	int seg6_do_srh_encap(struct sk_buff skb, struct ipv6_sr_hdr osrh, int proto)
6c8702c6	114	{
8936ef76 DL	115	struct dst_entry *dst = skb_dst(skb);
8936ef76 DL	116	struct net *net = dev_net(dst->dev);
6c8702c6 DL	117	struct ipv6hdr hdr, inner_hdr;
	118	struct ipv6_sr_hdr *isrh;
	119	int hdrlen, tot_len, err;
b5facfdb	120	__be32 flowlabel;
6c8702c6 DL	121
	122	hdrlen = (osrh->hdrlen + 1) << 3;
	123	tot_len = hdrlen + sizeof(*hdr);
	124
bbb40a0b	125	err = skb_cow_head(skb, tot_len + skb->mac_len);
6c8702c6 DL	126	if (unlikely(err))
	127	return err;
	128
	129	inner_hdr = ipv6_hdr(skb);
6df93462	130	flowlabel = seg6_make_flowlabel(net, skb, inner_hdr);
6c8702c6 DL	131
	132	skb_push(skb, tot_len);
	133	skb_reset_network_header(skb);
	134	skb_mac_header_rebuild(skb);
	135	hdr = ipv6_hdr(skb);
	136
	137	/* inherit tc, flowlabel and hlim
	138	* hlim will be decremented in ip6_forward() afterwards and
	139	* decapsulation will overwrite inner hlim with outer hlim
	140	*/
32d99d0b DL	141
	142	if (skb->protocol == htons(ETH_P_IPV6)) {
	143	ip6_flow_hdr(hdr, ip6_tclass(ip6_flowinfo(inner_hdr)),
b5facfdb	144	flowlabel);
32d99d0b DL	145	hdr->hop_limit = inner_hdr->hop_limit;
32d99d0b DL	146	} else {
b5facfdb	147	ip6_flow_hdr(hdr, 0, flowlabel);
32d99d0b DL	148	hdr->hop_limit = ip6_dst_hoplimit(skb_dst(skb));
	149	}
	150
6c8702c6 DL	151	hdr->nexthdr = NEXTHDR_ROUTING;
	152
	153	isrh = (void )hdr + sizeof(hdr);
	154	memcpy(isrh, osrh, hdrlen);
	155
32d99d0b	156	isrh->nexthdr = proto;
6c8702c6 DL	157
6c8702c6 DL	158	hdr->daddr = isrh->segments[isrh->first_segment];
a957fa19	159	set_tun_src(net, dst->dev, &hdr->daddr, &hdr->saddr);
6c8702c6	160
9baee834 DL	161	#ifdef CONFIG_IPV6_SEG6_HMAC
	162	if (sr_has_hmac(isrh)) {
	163	err = seg6_push_hmac(net, &hdr->saddr, isrh);
	164	if (unlikely(err))
	165	return err;
	166	}
	167	#endif
	168
6c8702c6 DL	169	skb_postpush_rcsum(skb, hdr, tot_len);
	170
	171	return 0;
	172	}
b04c80d3	173	EXPORT_SYMBOL_GPL(seg6_do_srh_encap);
6c8702c6 DL	174
6c8702c6 DL	175	/* insert an SRH within an IPv6 packet, just after the IPv6 header */
b04c80d3	176	int seg6_do_srh_inline(struct sk_buff skb, struct ipv6_sr_hdr osrh)
6c8702c6 DL	177	{
	178	struct ipv6hdr hdr, oldhdr;
	179	struct ipv6_sr_hdr *isrh;
	180	int hdrlen, err;
	181
	182	hdrlen = (osrh->hdrlen + 1) << 3;
	183
bbb40a0b	184	err = skb_cow_head(skb, hdrlen + skb->mac_len);
6c8702c6 DL	185	if (unlikely(err))
	186	return err;
	187
	188	oldhdr = ipv6_hdr(skb);
	189
	190	skb_pull(skb, sizeof(struct ipv6hdr));
	191	skb_postpull_rcsum(skb, skb_network_header(skb),
	192	sizeof(struct ipv6hdr));
	193
	194	skb_push(skb, sizeof(struct ipv6hdr) + hdrlen);
	195	skb_reset_network_header(skb);
	196	skb_mac_header_rebuild(skb);
	197
	198	hdr = ipv6_hdr(skb);
	199
	200	memmove(hdr, oldhdr, sizeof(*hdr));
	201
	202	isrh = (void )hdr + sizeof(hdr);
	203	memcpy(isrh, osrh, hdrlen);
	204
	205	isrh->nexthdr = hdr->nexthdr;
	206	hdr->nexthdr = NEXTHDR_ROUTING;
	207
	208	isrh->segments[0] = hdr->daddr;
	209	hdr->daddr = isrh->segments[isrh->first_segment];
	210
9baee834 DL	211	#ifdef CONFIG_IPV6_SEG6_HMAC
	212	if (sr_has_hmac(isrh)) {
	213	struct net *net = dev_net(skb_dst(skb)->dev);
	214
	215	err = seg6_push_hmac(net, &hdr->saddr, isrh);
	216	if (unlikely(err))
	217	return err;
	218	}
	219	#endif
	220
6c8702c6 DL	221	skb_postpush_rcsum(skb, hdr, sizeof(struct ipv6hdr) + hdrlen);
	222
	223	return 0;
	224	}
b04c80d3	225	EXPORT_SYMBOL_GPL(seg6_do_srh_inline);
6c8702c6 DL	226
	227	static int seg6_do_srh(struct sk_buff *skb)
	228	{
	229	struct dst_entry *dst = skb_dst(skb);
	230	struct seg6_iptunnel_encap *tinfo;
32d99d0b	231	int proto, err = 0;
6c8702c6 DL	232
	233	tinfo = seg6_encap_lwtunnel(dst->lwtstate);
	234
6c8702c6	235	switch (tinfo->mode) {
6c8702c6	236	case SEG6_IPTUN_MODE_INLINE:
32d99d0b DL	237	if (skb->protocol != htons(ETH_P_IPV6))
	238	return -EINVAL;
	239
6c8702c6	240	err = seg6_do_srh_inline(skb, tinfo->srh);
32d99d0b DL	241	if (err)
32d99d0b DL	242	return err;
6c8702c6	243	break;
6c8702c6	244	case SEG6_IPTUN_MODE_ENCAP:
5807b22c DL	245	err = iptunnel_handle_offloads(skb, SKB_GSO_IPXIP6);
	246	if (err)
	247	return err;
	248
32d99d0b DL	249	if (skb->protocol == htons(ETH_P_IPV6))
	250	proto = IPPROTO_IPV6;
	251	else if (skb->protocol == htons(ETH_P_IP))
	252	proto = IPPROTO_IPIP;
	253	else
	254	return -EINVAL;
	255
	256	err = seg6_do_srh_encap(skb, tinfo->srh, proto);
	257	if (err)
	258	return err;
	259
5807b22c DL	260	skb_set_inner_transport_header(skb, skb_transport_offset(skb));
5807b22c DL	261	skb_set_inner_protocol(skb, skb->protocol);
38ee7f2d DL	262	skb->protocol = htons(ETH_P_IPV6);
	263	break;
	264	case SEG6_IPTUN_MODE_L2ENCAP:
	265	if (!skb_mac_header_was_set(skb))
	266	return -EINVAL;
	267
	268	if (pskb_expand_head(skb, skb->mac_len, 0, GFP_ATOMIC) < 0)
	269	return -ENOMEM;
	270
	271	skb_mac_header_rebuild(skb);
	272	skb_push(skb, skb->mac_len);
	273
	274	err = seg6_do_srh_encap(skb, tinfo->srh, NEXTHDR_NONE);
	275	if (err)
	276	return err;
	277
32d99d0b	278	skb->protocol = htons(ETH_P_IPV6);
6c8702c6 DL	279	break;
	280	}
	281
6c8702c6 DL	282	ipv6_hdr(skb)->payload_len = htons(skb->len - sizeof(struct ipv6hdr));
	283	skb_set_transport_header(skb, sizeof(struct ipv6hdr));
	284
6c8702c6 DL	285	return 0;
	286	}
	287
bb4005ba	288	static int seg6_input(struct sk_buff *skb)
6c8702c6	289	{
af4a2209 DL	290	struct dst_entry *orig_dst = skb_dst(skb);
	291	struct dst_entry *dst = NULL;
	292	struct seg6_lwt *slwt;
6c8702c6 DL	293	int err;
	294
	295	err = seg6_do_srh(skb);
	296	if (unlikely(err)) {
	297	kfree_skb(skb);
	298	return err;
	299	}
	300
af4a2209 DL	301	slwt = seg6_lwt_lwtunnel(orig_dst->lwtstate);
af4a2209 DL	302
af4a2209 DL	303	preempt_disable();
	304	dst = dst_cache_get(&slwt->cache);
	305	preempt_enable();
af4a2209	306
6c8702c6	307	skb_dst_drop(skb);
af4a2209 DL	308
	309	if (!dst) {
	310	ip6_route_input(skb);
af4a2209 DL	311	dst = skb_dst(skb);
	312	if (!dst->error) {
	313	preempt_disable();
	314	dst_cache_set_ip6(&slwt->cache, dst,
	315	&ipv6_hdr(skb)->saddr);
	316	preempt_enable();
	317	}
af4a2209 DL	318	} else {
	319	skb_dst_set(skb, dst);
	320	}
6c8702c6	321
af3b5158 DL	322	err = skb_cow_head(skb, LL_RESERVED_SPACE(dst->dev));
	323	if (unlikely(err))
	324	return err;
	325
6c8702c6 DL	326	return dst_input(skb);
	327	}
	328
bb4005ba	329	static int seg6_output(struct net net, struct sock sk, struct sk_buff *skb)
6c8702c6 DL	330	{
	331	struct dst_entry *orig_dst = skb_dst(skb);
	332	struct dst_entry *dst = NULL;
	333	struct seg6_lwt *slwt;
	334	int err = -EINVAL;
	335
	336	err = seg6_do_srh(skb);
	337	if (unlikely(err))
	338	goto drop;
	339
	340	slwt = seg6_lwt_lwtunnel(orig_dst->lwtstate);
	341
fa79581e	342	preempt_disable();
6c8702c6	343	dst = dst_cache_get(&slwt->cache);
fa79581e	344	preempt_enable();
6c8702c6 DL	345
	346	if (unlikely(!dst)) {
	347	struct ipv6hdr *hdr = ipv6_hdr(skb);
	348	struct flowi6 fl6;
	349
1b4e5ad5	350	memset(&fl6, 0, sizeof(fl6));
6c8702c6 DL	351	fl6.daddr = hdr->daddr;
	352	fl6.saddr = hdr->saddr;
	353	fl6.flowlabel = ip6_flowinfo(hdr);
	354	fl6.flowi6_mark = skb->mark;
	355	fl6.flowi6_proto = hdr->nexthdr;
	356
	357	dst = ip6_route_output(net, NULL, &fl6);
	358	if (dst->error) {
	359	err = dst->error;
	360	dst_release(dst);
	361	goto drop;
	362	}
	363
fa79581e	364	preempt_disable();
6c8702c6	365	dst_cache_set_ip6(&slwt->cache, dst, &fl6.saddr);
fa79581e	366	preempt_enable();
6c8702c6 DL	367	}
	368
	369	skb_dst_drop(skb);
	370	skb_dst_set(skb, dst);
	371
af3b5158 DL	372	err = skb_cow_head(skb, LL_RESERVED_SPACE(dst->dev));
	373	if (unlikely(err))
	374	goto drop;
	375
6c8702c6 DL	376	return dst_output(net, sk, skb);
	377	drop:
	378	kfree_skb(skb);
	379	return err;
	380	}
	381
30357d7d	382	static int seg6_build_state(struct nlattr *nla,
6c8702c6	383	unsigned int family, const void *cfg,
9ae28727 DA	384	struct lwtunnel_state **ts,
9ae28727 DA	385	struct netlink_ext_ack *extack)
6c8702c6 DL	386	{
	387	struct nlattr *tb[SEG6_IPTUNNEL_MAX + 1];
	388	struct seg6_iptunnel_encap *tuninfo;
	389	struct lwtunnel_state *newts;
	390	int tuninfo_len, min_size;
	391	struct seg6_lwt *slwt;
	392	int err;
	393
32d99d0b DL	394	if (family != AF_INET && family != AF_INET6)
	395	return -EINVAL;
	396
6c8702c6	397	err = nla_parse_nested(tb, SEG6_IPTUNNEL_MAX, nla,
9ae28727	398	seg6_iptunnel_policy, extack);
6c8702c6 DL	399
	400	if (err < 0)
	401	return err;
	402
	403	if (!tb[SEG6_IPTUNNEL_SRH])
	404	return -EINVAL;
	405
	406	tuninfo = nla_data(tb[SEG6_IPTUNNEL_SRH]);
	407	tuninfo_len = nla_len(tb[SEG6_IPTUNNEL_SRH]);
	408
	409	/* tuninfo must contain at least the iptunnel encap structure,
	410	* the SRH and one segment
	411	*/
	412	min_size = sizeof(*tuninfo) + sizeof(struct ipv6_sr_hdr) +
	413	sizeof(struct in6_addr);
	414	if (tuninfo_len < min_size)
	415	return -EINVAL;
	416
	417	switch (tuninfo->mode) {
6c8702c6	418	case SEG6_IPTUN_MODE_INLINE:
32d99d0b DL	419	if (family != AF_INET6)
	420	return -EINVAL;
	421
6c8702c6	422	break;
6c8702c6 DL	423	case SEG6_IPTUN_MODE_ENCAP:
6c8702c6 DL	424	break;
38ee7f2d DL	425	case SEG6_IPTUN_MODE_L2ENCAP:
38ee7f2d DL	426	break;
6c8702c6 DL	427	default:
	428	return -EINVAL;
	429	}
	430
	431	/* verify that SRH is consistent */
	432	if (!seg6_validate_srh(tuninfo->srh, tuninfo_len - sizeof(*tuninfo)))
	433	return -EINVAL;
	434
	435	newts = lwtunnel_state_alloc(tuninfo_len + sizeof(*slwt));
	436	if (!newts)
	437	return -ENOMEM;
	438
	439	slwt = seg6_lwt_lwtunnel(newts);
	440
191f86ca	441	err = dst_cache_init(&slwt->cache, GFP_ATOMIC);
6c8702c6 DL	442	if (err) {
	443	kfree(newts);
	444	return err;
	445	}
6c8702c6 DL	446
	447	memcpy(&slwt->tuninfo, tuninfo, tuninfo_len);
	448
	449	newts->type = LWTUNNEL_ENCAP_SEG6;
38ee7f2d DL	450	newts->flags \|= LWTUNNEL_STATE_INPUT_REDIRECT;
	451
	452	if (tuninfo->mode != SEG6_IPTUN_MODE_L2ENCAP)
	453	newts->flags \|= LWTUNNEL_STATE_OUTPUT_REDIRECT;
	454
6c8702c6 DL	455	newts->headroom = seg6_lwt_headroom(tuninfo);
	456
	457	*ts = newts;
	458
	459	return 0;
	460	}
	461
6c8702c6 DL	462	static void seg6_destroy_state(struct lwtunnel_state *lwt)
	463	{
	464	dst_cache_destroy(&seg6_lwt_lwtunnel(lwt)->cache);
	465	}
6c8702c6 DL	466
	467	static int seg6_fill_encap_info(struct sk_buff *skb,
	468	struct lwtunnel_state *lwtstate)
	469	{
	470	struct seg6_iptunnel_encap *tuninfo = seg6_encap_lwtunnel(lwtstate);
	471
	472	if (nla_put_srh(skb, SEG6_IPTUNNEL_SRH, tuninfo))
	473	return -EMSGSIZE;
	474
	475	return 0;
	476	}
	477
	478	static int seg6_encap_nlsize(struct lwtunnel_state *lwtstate)
	479	{
	480	struct seg6_iptunnel_encap *tuninfo = seg6_encap_lwtunnel(lwtstate);
	481
	482	return nla_total_size(SEG6_IPTUN_ENCAP_SIZE(tuninfo));
	483	}
	484
	485	static int seg6_encap_cmp(struct lwtunnel_state a, struct lwtunnel_state b)
	486	{
	487	struct seg6_iptunnel_encap *a_hdr = seg6_encap_lwtunnel(a);
	488	struct seg6_iptunnel_encap *b_hdr = seg6_encap_lwtunnel(b);
	489	int len = SEG6_IPTUN_ENCAP_SIZE(a_hdr);
	490
	491	if (len != SEG6_IPTUN_ENCAP_SIZE(b_hdr))
	492	return 1;
	493
	494	return memcmp(a_hdr, b_hdr, len);
	495	}
	496
	497	static const struct lwtunnel_encap_ops seg6_iptun_ops = {
	498	.build_state = seg6_build_state,
6c8702c6	499	.destroy_state = seg6_destroy_state,
6c8702c6 DL	500	.output = seg6_output,
	501	.input = seg6_input,
	502	.fill_encap = seg6_fill_encap_info,
	503	.get_encap_size = seg6_encap_nlsize,
	504	.cmp_encap = seg6_encap_cmp,
88ff7334	505	.owner = THIS_MODULE,
6c8702c6 DL	506	};
	507
	508	int __init seg6_iptunnel_init(void)
	509	{
	510	return lwtunnel_encap_add_ops(&seg6_iptun_ops, LWTUNNEL_ENCAP_SEG6);
	511	}
	512
	513	void seg6_iptunnel_exit(void)
	514	{
	515	lwtunnel_encap_del_ops(&seg6_iptun_ops, LWTUNNEL_ENCAP_SEG6);
	516	}