[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();
}

/* 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));

		memset(IP6CB(skb), 0, sizeof(*IP6CB(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	148	hdr->hop_limit = ip6_dst_hoplimit(skb_dst(skb));
ef489749 YK	149
ef489749 YK	150	memset(IP6CB(skb), 0, sizeof(*IP6CB(skb)));
32d99d0b DL	151	}
32d99d0b DL	152
6c8702c6 DL	153	hdr->nexthdr = NEXTHDR_ROUTING;
	154
	155	isrh = (void )hdr + sizeof(hdr);
	156	memcpy(isrh, osrh, hdrlen);
	157
32d99d0b	158	isrh->nexthdr = proto;
6c8702c6 DL	159
6c8702c6 DL	160	hdr->daddr = isrh->segments[isrh->first_segment];
a957fa19	161	set_tun_src(net, dst->dev, &hdr->daddr, &hdr->saddr);
6c8702c6	162
9baee834 DL	163	#ifdef CONFIG_IPV6_SEG6_HMAC
	164	if (sr_has_hmac(isrh)) {
	165	err = seg6_push_hmac(net, &hdr->saddr, isrh);
	166	if (unlikely(err))
	167	return err;
	168	}
	169	#endif
	170
6c8702c6 DL	171	skb_postpush_rcsum(skb, hdr, tot_len);
	172
	173	return 0;
	174	}
b04c80d3	175	EXPORT_SYMBOL_GPL(seg6_do_srh_encap);
6c8702c6 DL	176
6c8702c6 DL	177	/* insert an SRH within an IPv6 packet, just after the IPv6 header */
b04c80d3	178	int seg6_do_srh_inline(struct sk_buff skb, struct ipv6_sr_hdr osrh)
6c8702c6 DL	179	{
	180	struct ipv6hdr hdr, oldhdr;
	181	struct ipv6_sr_hdr *isrh;
	182	int hdrlen, err;
	183
	184	hdrlen = (osrh->hdrlen + 1) << 3;
	185
bbb40a0b	186	err = skb_cow_head(skb, hdrlen + skb->mac_len);
6c8702c6 DL	187	if (unlikely(err))
	188	return err;
	189
	190	oldhdr = ipv6_hdr(skb);
	191
	192	skb_pull(skb, sizeof(struct ipv6hdr));
	193	skb_postpull_rcsum(skb, skb_network_header(skb),
	194	sizeof(struct ipv6hdr));
	195
	196	skb_push(skb, sizeof(struct ipv6hdr) + hdrlen);
	197	skb_reset_network_header(skb);
	198	skb_mac_header_rebuild(skb);
	199
	200	hdr = ipv6_hdr(skb);
	201
	202	memmove(hdr, oldhdr, sizeof(*hdr));
	203
	204	isrh = (void )hdr + sizeof(hdr);
	205	memcpy(isrh, osrh, hdrlen);
	206
	207	isrh->nexthdr = hdr->nexthdr;
	208	hdr->nexthdr = NEXTHDR_ROUTING;
	209
	210	isrh->segments[0] = hdr->daddr;
	211	hdr->daddr = isrh->segments[isrh->first_segment];
	212
9baee834 DL	213	#ifdef CONFIG_IPV6_SEG6_HMAC
	214	if (sr_has_hmac(isrh)) {
	215	struct net *net = dev_net(skb_dst(skb)->dev);
	216
	217	err = seg6_push_hmac(net, &hdr->saddr, isrh);
	218	if (unlikely(err))
	219	return err;
	220	}
	221	#endif
	222
6c8702c6 DL	223	skb_postpush_rcsum(skb, hdr, sizeof(struct ipv6hdr) + hdrlen);
	224
	225	return 0;
	226	}
b04c80d3	227	EXPORT_SYMBOL_GPL(seg6_do_srh_inline);
6c8702c6 DL	228
	229	static int seg6_do_srh(struct sk_buff *skb)
	230	{
	231	struct dst_entry *dst = skb_dst(skb);
	232	struct seg6_iptunnel_encap *tinfo;
32d99d0b	233	int proto, err = 0;
6c8702c6 DL	234
	235	tinfo = seg6_encap_lwtunnel(dst->lwtstate);
	236
6c8702c6	237	switch (tinfo->mode) {
6c8702c6	238	case SEG6_IPTUN_MODE_INLINE:
32d99d0b DL	239	if (skb->protocol != htons(ETH_P_IPV6))
	240	return -EINVAL;
	241
6c8702c6	242	err = seg6_do_srh_inline(skb, tinfo->srh);
32d99d0b DL	243	if (err)
32d99d0b DL	244	return err;
6c8702c6	245	break;
6c8702c6	246	case SEG6_IPTUN_MODE_ENCAP:
5807b22c DL	247	err = iptunnel_handle_offloads(skb, SKB_GSO_IPXIP6);
	248	if (err)
	249	return err;
	250
32d99d0b DL	251	if (skb->protocol == htons(ETH_P_IPV6))
	252	proto = IPPROTO_IPV6;
	253	else if (skb->protocol == htons(ETH_P_IP))
	254	proto = IPPROTO_IPIP;
	255	else
	256	return -EINVAL;
	257
	258	err = seg6_do_srh_encap(skb, tinfo->srh, proto);
	259	if (err)
	260	return err;
	261
5807b22c DL	262	skb_set_inner_transport_header(skb, skb_transport_offset(skb));
5807b22c DL	263	skb_set_inner_protocol(skb, skb->protocol);
38ee7f2d DL	264	skb->protocol = htons(ETH_P_IPV6);
	265	break;
	266	case SEG6_IPTUN_MODE_L2ENCAP:
	267	if (!skb_mac_header_was_set(skb))
	268	return -EINVAL;
	269
	270	if (pskb_expand_head(skb, skb->mac_len, 0, GFP_ATOMIC) < 0)
	271	return -ENOMEM;
	272
	273	skb_mac_header_rebuild(skb);
	274	skb_push(skb, skb->mac_len);
	275
	276	err = seg6_do_srh_encap(skb, tinfo->srh, NEXTHDR_NONE);
	277	if (err)
	278	return err;
	279
32d99d0b	280	skb->protocol = htons(ETH_P_IPV6);
6c8702c6 DL	281	break;
	282	}
	283
6c8702c6 DL	284	ipv6_hdr(skb)->payload_len = htons(skb->len - sizeof(struct ipv6hdr));
	285	skb_set_transport_header(skb, sizeof(struct ipv6hdr));
	286
6c8702c6 DL	287	return 0;
	288	}
	289
bb4005ba	290	static int seg6_input(struct sk_buff *skb)
6c8702c6	291	{
af4a2209 DL	292	struct dst_entry *orig_dst = skb_dst(skb);
	293	struct dst_entry *dst = NULL;
	294	struct seg6_lwt *slwt;
6c8702c6 DL	295	int err;
	296
	297	err = seg6_do_srh(skb);
	298	if (unlikely(err)) {
	299	kfree_skb(skb);
	300	return err;
	301	}
	302
af4a2209 DL	303	slwt = seg6_lwt_lwtunnel(orig_dst->lwtstate);
af4a2209 DL	304
af4a2209 DL	305	preempt_disable();
	306	dst = dst_cache_get(&slwt->cache);
	307	preempt_enable();
af4a2209	308
6c8702c6	309	skb_dst_drop(skb);
af4a2209 DL	310
	311	if (!dst) {
	312	ip6_route_input(skb);
af4a2209 DL	313	dst = skb_dst(skb);
	314	if (!dst->error) {
	315	preempt_disable();
	316	dst_cache_set_ip6(&slwt->cache, dst,
	317	&ipv6_hdr(skb)->saddr);
	318	preempt_enable();
	319	}
af4a2209 DL	320	} else {
	321	skb_dst_set(skb, dst);
	322	}
6c8702c6	323
af3b5158 DL	324	err = skb_cow_head(skb, LL_RESERVED_SPACE(dst->dev));
	325	if (unlikely(err))
	326	return err;
	327
6c8702c6 DL	328	return dst_input(skb);
	329	}
	330
bb4005ba	331	static int seg6_output(struct net net, struct sock sk, struct sk_buff *skb)
6c8702c6 DL	332	{
	333	struct dst_entry *orig_dst = skb_dst(skb);
	334	struct dst_entry *dst = NULL;
	335	struct seg6_lwt *slwt;
	336	int err = -EINVAL;
	337
	338	err = seg6_do_srh(skb);
	339	if (unlikely(err))
	340	goto drop;
	341
	342	slwt = seg6_lwt_lwtunnel(orig_dst->lwtstate);
	343
fa79581e	344	preempt_disable();
6c8702c6	345	dst = dst_cache_get(&slwt->cache);
fa79581e	346	preempt_enable();
6c8702c6 DL	347
	348	if (unlikely(!dst)) {
	349	struct ipv6hdr *hdr = ipv6_hdr(skb);
	350	struct flowi6 fl6;
	351
1b4e5ad5	352	memset(&fl6, 0, sizeof(fl6));
6c8702c6 DL	353	fl6.daddr = hdr->daddr;
	354	fl6.saddr = hdr->saddr;
	355	fl6.flowlabel = ip6_flowinfo(hdr);
	356	fl6.flowi6_mark = skb->mark;
	357	fl6.flowi6_proto = hdr->nexthdr;
	358
	359	dst = ip6_route_output(net, NULL, &fl6);
	360	if (dst->error) {
	361	err = dst->error;
	362	dst_release(dst);
	363	goto drop;
	364	}
	365
fa79581e	366	preempt_disable();
6c8702c6	367	dst_cache_set_ip6(&slwt->cache, dst, &fl6.saddr);
fa79581e	368	preempt_enable();
6c8702c6 DL	369	}
	370
	371	skb_dst_drop(skb);
	372	skb_dst_set(skb, dst);
	373
af3b5158 DL	374	err = skb_cow_head(skb, LL_RESERVED_SPACE(dst->dev));
	375	if (unlikely(err))
	376	goto drop;
	377
6c8702c6 DL	378	return dst_output(net, sk, skb);
	379	drop:
	380	kfree_skb(skb);
	381	return err;
	382	}
	383
30357d7d	384	static int seg6_build_state(struct nlattr *nla,
6c8702c6	385	unsigned int family, const void *cfg,
9ae28727 DA	386	struct lwtunnel_state **ts,
9ae28727 DA	387	struct netlink_ext_ack *extack)
6c8702c6 DL	388	{
	389	struct nlattr *tb[SEG6_IPTUNNEL_MAX + 1];
	390	struct seg6_iptunnel_encap *tuninfo;
	391	struct lwtunnel_state *newts;
	392	int tuninfo_len, min_size;
	393	struct seg6_lwt *slwt;
	394	int err;
	395
32d99d0b DL	396	if (family != AF_INET && family != AF_INET6)
	397	return -EINVAL;
	398
6c8702c6	399	err = nla_parse_nested(tb, SEG6_IPTUNNEL_MAX, nla,
9ae28727	400	seg6_iptunnel_policy, extack);
6c8702c6 DL	401
	402	if (err < 0)
	403	return err;
	404
	405	if (!tb[SEG6_IPTUNNEL_SRH])
	406	return -EINVAL;
	407
	408	tuninfo = nla_data(tb[SEG6_IPTUNNEL_SRH]);
	409	tuninfo_len = nla_len(tb[SEG6_IPTUNNEL_SRH]);
	410
	411	/* tuninfo must contain at least the iptunnel encap structure,
	412	* the SRH and one segment
	413	*/
	414	min_size = sizeof(*tuninfo) + sizeof(struct ipv6_sr_hdr) +
	415	sizeof(struct in6_addr);
	416	if (tuninfo_len < min_size)
	417	return -EINVAL;
	418
	419	switch (tuninfo->mode) {
6c8702c6	420	case SEG6_IPTUN_MODE_INLINE:
32d99d0b DL	421	if (family != AF_INET6)
	422	return -EINVAL;
	423
6c8702c6	424	break;
6c8702c6 DL	425	case SEG6_IPTUN_MODE_ENCAP:
6c8702c6 DL	426	break;
38ee7f2d DL	427	case SEG6_IPTUN_MODE_L2ENCAP:
38ee7f2d DL	428	break;
6c8702c6 DL	429	default:
	430	return -EINVAL;
	431	}
	432
	433	/* verify that SRH is consistent */
	434	if (!seg6_validate_srh(tuninfo->srh, tuninfo_len - sizeof(*tuninfo)))
	435	return -EINVAL;
	436
	437	newts = lwtunnel_state_alloc(tuninfo_len + sizeof(*slwt));
	438	if (!newts)
	439	return -ENOMEM;
	440
	441	slwt = seg6_lwt_lwtunnel(newts);
	442
191f86ca	443	err = dst_cache_init(&slwt->cache, GFP_ATOMIC);
6c8702c6 DL	444	if (err) {
	445	kfree(newts);
	446	return err;
	447	}
6c8702c6 DL	448
	449	memcpy(&slwt->tuninfo, tuninfo, tuninfo_len);
	450
	451	newts->type = LWTUNNEL_ENCAP_SEG6;
38ee7f2d DL	452	newts->flags \|= LWTUNNEL_STATE_INPUT_REDIRECT;
	453
	454	if (tuninfo->mode != SEG6_IPTUN_MODE_L2ENCAP)
	455	newts->flags \|= LWTUNNEL_STATE_OUTPUT_REDIRECT;
	456
6c8702c6 DL	457	newts->headroom = seg6_lwt_headroom(tuninfo);
	458
	459	*ts = newts;
	460
	461	return 0;
	462	}
	463
6c8702c6 DL	464	static void seg6_destroy_state(struct lwtunnel_state *lwt)
	465	{
	466	dst_cache_destroy(&seg6_lwt_lwtunnel(lwt)->cache);
	467	}
6c8702c6 DL	468
	469	static int seg6_fill_encap_info(struct sk_buff *skb,
	470	struct lwtunnel_state *lwtstate)
	471	{
	472	struct seg6_iptunnel_encap *tuninfo = seg6_encap_lwtunnel(lwtstate);
	473
	474	if (nla_put_srh(skb, SEG6_IPTUNNEL_SRH, tuninfo))
	475	return -EMSGSIZE;
	476
	477	return 0;
	478	}
	479
	480	static int seg6_encap_nlsize(struct lwtunnel_state *lwtstate)
	481	{
	482	struct seg6_iptunnel_encap *tuninfo = seg6_encap_lwtunnel(lwtstate);
	483
	484	return nla_total_size(SEG6_IPTUN_ENCAP_SIZE(tuninfo));
	485	}
	486
	487	static int seg6_encap_cmp(struct lwtunnel_state a, struct lwtunnel_state b)
	488	{
	489	struct seg6_iptunnel_encap *a_hdr = seg6_encap_lwtunnel(a);
	490	struct seg6_iptunnel_encap *b_hdr = seg6_encap_lwtunnel(b);
	491	int len = SEG6_IPTUN_ENCAP_SIZE(a_hdr);
	492
	493	if (len != SEG6_IPTUN_ENCAP_SIZE(b_hdr))
	494	return 1;
	495
	496	return memcmp(a_hdr, b_hdr, len);
	497	}
	498
	499	static const struct lwtunnel_encap_ops seg6_iptun_ops = {
	500	.build_state = seg6_build_state,
6c8702c6	501	.destroy_state = seg6_destroy_state,
6c8702c6 DL	502	.output = seg6_output,
	503	.input = seg6_input,
	504	.fill_encap = seg6_fill_encap_info,
	505	.get_encap_size = seg6_encap_nlsize,
	506	.cmp_encap = seg6_encap_cmp,
88ff7334	507	.owner = THIS_MODULE,
6c8702c6 DL	508	};
	509
	510	int __init seg6_iptunnel_init(void)
	511	{
	512	return lwtunnel_encap_add_ops(&seg6_iptun_ops, LWTUNNEL_ENCAP_SEG6);
	513	}
	514
	515	void seg6_iptunnel_exit(void)
	516	{
	517	lwtunnel_encap_del_ops(&seg6_iptun_ops, LWTUNNEL_ENCAP_SEG6);
	518	}