[deliverable/linux.git] / net / ipv4 / ip_tunnel_core.c

/*
 * Copyright (c) 2013 Nicira, Inc.
 *
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of version 2 of the GNU General Public
 * License as published by the Free Software Foundation.
 *
 * This program is distributed in the hope that it will be useful, but
 * WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
 * General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program; if not, write to the Free Software
 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
 * 02110-1301, USA
 */

#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt

#include <linux/types.h>
#include <linux/kernel.h>
#include <linux/skbuff.h>
#include <linux/netdevice.h>
#include <linux/in.h>
#include <linux/if_arp.h>
#include <linux/init.h>
#include <linux/in6.h>
#include <linux/inetdevice.h>
#include <linux/netfilter_ipv4.h>
#include <linux/etherdevice.h>
#include <linux/if_ether.h>
#include <linux/if_vlan.h>
#include <linux/static_key.h>

#include <net/ip.h>
#include <net/icmp.h>
#include <net/protocol.h>
#include <net/ip_tunnels.h>
#include <net/arp.h>
#include <net/checksum.h>
#include <net/dsfield.h>
#include <net/inet_ecn.h>
#include <net/xfrm.h>
#include <net/net_namespace.h>
#include <net/netns/generic.h>
#include <net/rtnetlink.h>
#include <net/dst_metadata.h>

void iptunnel_xmit(struct sock *sk, struct rtable *rt, struct sk_buff *skb,
		   __be32 src, __be32 dst, __u8 proto,
		   __u8 tos, __u8 ttl, __be16 df, bool xnet)
{
	int pkt_len = skb->len - skb_inner_network_offset(skb);
	struct net *net = dev_net(rt->dst.dev);
	struct net_device *dev = skb->dev;
	struct iphdr *iph;
	int err;

	skb_scrub_packet(skb, xnet);

	skb_clear_hash(skb);
	skb_dst_set(skb, &rt->dst);
	memset(IPCB(skb), 0, sizeof(*IPCB(skb)));

	/* Push down and install the IP header. */
	skb_push(skb, sizeof(struct iphdr));
	skb_reset_network_header(skb);

	iph = ip_hdr(skb);

	iph->version	=	4;
	iph->ihl	=	sizeof(struct iphdr) >> 2;
	iph->frag_off	=	df;
	iph->protocol	=	proto;
	iph->tos	=	tos;
	iph->daddr	=	dst;
	iph->saddr	=	src;
	iph->ttl	=	ttl;
	__ip_select_ident(net, iph, skb_shinfo(skb)->gso_segs ?: 1);

	err = ip_local_out(net, sk, skb);
	if (unlikely(net_xmit_eval(err)))
		pkt_len = 0;
	iptunnel_xmit_stats(dev, pkt_len);
}
EXPORT_SYMBOL_GPL(iptunnel_xmit);

int iptunnel_pull_header(struct sk_buff *skb, int hdr_len, __be16 inner_proto,
			 bool xnet)
{
	if (unlikely(!pskb_may_pull(skb, hdr_len)))
		return -ENOMEM;

	skb_pull_rcsum(skb, hdr_len);

	if (inner_proto == htons(ETH_P_TEB)) {
		struct ethhdr *eh;

		if (unlikely(!pskb_may_pull(skb, ETH_HLEN)))
			return -ENOMEM;

		eh = (struct ethhdr *)skb->data;
		if (likely(eth_proto_is_802_3(eh->h_proto)))
			skb->protocol = eh->h_proto;
		else
			skb->protocol = htons(ETH_P_802_2);

	} else {
		skb->protocol = inner_proto;
	}

	skb_clear_hash_if_not_l4(skb);
	skb->vlan_tci = 0;
	skb_set_queue_mapping(skb, 0);
	skb_scrub_packet(skb, xnet);
	return 0;
}
EXPORT_SYMBOL_GPL(iptunnel_pull_header);

struct metadata_dst *iptunnel_metadata_reply(struct metadata_dst *md,
					     gfp_t flags)
{
	struct metadata_dst *res;
	struct ip_tunnel_info *dst, *src;

	if (!md || md->u.tun_info.mode & IP_TUNNEL_INFO_TX)
		return NULL;

	res = metadata_dst_alloc(0, flags);
	if (!res)
		return NULL;

	dst = &res->u.tun_info;
	src = &md->u.tun_info;
	dst->key.tun_id = src->key.tun_id;
	if (src->mode & IP_TUNNEL_INFO_IPV6)
		memcpy(&dst->key.u.ipv6.dst, &src->key.u.ipv6.src,
		       sizeof(struct in6_addr));
	else
		dst->key.u.ipv4.dst = src->key.u.ipv4.src;
	dst->mode = src->mode | IP_TUNNEL_INFO_TX;

	return res;
}
EXPORT_SYMBOL_GPL(iptunnel_metadata_reply);

struct sk_buff *iptunnel_handle_offloads(struct sk_buff *skb,
					 int gso_type_mask)
{
	int err;

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

	if (skb_is_gso(skb)) {
		err = skb_unclone(skb, GFP_ATOMIC);
		if (unlikely(err))
			goto error;
		skb_shinfo(skb)->gso_type |= gso_type_mask;
		return skb;
	}

	if (skb->ip_summed != CHECKSUM_PARTIAL) {
		skb->ip_summed = CHECKSUM_NONE;
		/* We clear encapsulation here to prevent badly-written
		 * drivers potentially deciding to offload an inner checksum
		 * if we set CHECKSUM_PARTIAL on the outer header.
		 * This should go away when the drivers are all fixed.
		 */
		skb->encapsulation = 0;
	}

	return skb;
error:
	kfree_skb(skb);
	return ERR_PTR(err);
}
EXPORT_SYMBOL_GPL(iptunnel_handle_offloads);

/* Often modified stats are per cpu, other are shared (netdev->stats) */
struct rtnl_link_stats64 *ip_tunnel_get_stats64(struct net_device *dev,
						struct rtnl_link_stats64 *tot)
{
	int i;

	netdev_stats_to_stats64(tot, &dev->stats);

	for_each_possible_cpu(i) {
		const struct pcpu_sw_netstats *tstats =
						   per_cpu_ptr(dev->tstats, i);
		u64 rx_packets, rx_bytes, tx_packets, tx_bytes;
		unsigned int start;

		do {
			start = u64_stats_fetch_begin_irq(&tstats->syncp);
			rx_packets = tstats->rx_packets;
			tx_packets = tstats->tx_packets;
			rx_bytes = tstats->rx_bytes;
			tx_bytes = tstats->tx_bytes;
		} while (u64_stats_fetch_retry_irq(&tstats->syncp, start));

		tot->rx_packets += rx_packets;
		tot->tx_packets += tx_packets;
		tot->rx_bytes   += rx_bytes;
		tot->tx_bytes   += tx_bytes;
	}

	return tot;
}
EXPORT_SYMBOL_GPL(ip_tunnel_get_stats64);

static const struct nla_policy ip_tun_policy[LWTUNNEL_IP_MAX + 1] = {
	[LWTUNNEL_IP_ID]	= { .type = NLA_U64 },
	[LWTUNNEL_IP_DST]	= { .type = NLA_U32 },
	[LWTUNNEL_IP_SRC]	= { .type = NLA_U32 },
	[LWTUNNEL_IP_TTL]	= { .type = NLA_U8 },
	[LWTUNNEL_IP_TOS]	= { .type = NLA_U8 },
	[LWTUNNEL_IP_FLAGS]	= { .type = NLA_U16 },
};

static int ip_tun_build_state(struct net_device *dev, struct nlattr *attr,
			      unsigned int family, const void *cfg,
			      struct lwtunnel_state **ts)
{
	struct ip_tunnel_info *tun_info;
	struct lwtunnel_state *new_state;
	struct nlattr *tb[LWTUNNEL_IP_MAX + 1];
	int err;

	err = nla_parse_nested(tb, LWTUNNEL_IP_MAX, attr, ip_tun_policy);
	if (err < 0)
		return err;

	new_state = lwtunnel_state_alloc(sizeof(*tun_info));
	if (!new_state)
		return -ENOMEM;

	new_state->type = LWTUNNEL_ENCAP_IP;

	tun_info = lwt_tun_info(new_state);

	if (tb[LWTUNNEL_IP_ID])
		tun_info->key.tun_id = nla_get_be64(tb[LWTUNNEL_IP_ID]);

	if (tb[LWTUNNEL_IP_DST])
		tun_info->key.u.ipv4.dst = nla_get_be32(tb[LWTUNNEL_IP_DST]);

	if (tb[LWTUNNEL_IP_SRC])
		tun_info->key.u.ipv4.src = nla_get_be32(tb[LWTUNNEL_IP_SRC]);

	if (tb[LWTUNNEL_IP_TTL])
		tun_info->key.ttl = nla_get_u8(tb[LWTUNNEL_IP_TTL]);

	if (tb[LWTUNNEL_IP_TOS])
		tun_info->key.tos = nla_get_u8(tb[LWTUNNEL_IP_TOS]);

	if (tb[LWTUNNEL_IP_FLAGS])
		tun_info->key.tun_flags = nla_get_be16(tb[LWTUNNEL_IP_FLAGS]);

	tun_info->mode = IP_TUNNEL_INFO_TX;
	tun_info->options_len = 0;

	*ts = new_state;

	return 0;
}

static int ip_tun_fill_encap_info(struct sk_buff *skb,
				  struct lwtunnel_state *lwtstate)
{
	struct ip_tunnel_info *tun_info = lwt_tun_info(lwtstate);

	if (nla_put_be64(skb, LWTUNNEL_IP_ID, tun_info->key.tun_id) ||
	    nla_put_be32(skb, LWTUNNEL_IP_DST, tun_info->key.u.ipv4.dst) ||
	    nla_put_be32(skb, LWTUNNEL_IP_SRC, tun_info->key.u.ipv4.src) ||
	    nla_put_u8(skb, LWTUNNEL_IP_TOS, tun_info->key.tos) ||
	    nla_put_u8(skb, LWTUNNEL_IP_TTL, tun_info->key.ttl) ||
	    nla_put_be16(skb, LWTUNNEL_IP_FLAGS, tun_info->key.tun_flags))
		return -ENOMEM;

	return 0;
}

static int ip_tun_encap_nlsize(struct lwtunnel_state *lwtstate)
{
	return nla_total_size(8)	/* LWTUNNEL_IP_ID */
		+ nla_total_size(4)	/* LWTUNNEL_IP_DST */
		+ nla_total_size(4)	/* LWTUNNEL_IP_SRC */
		+ nla_total_size(1)	/* LWTUNNEL_IP_TOS */
		+ nla_total_size(1)	/* LWTUNNEL_IP_TTL */
		+ nla_total_size(2);	/* LWTUNNEL_IP_FLAGS */
}

static int ip_tun_cmp_encap(struct lwtunnel_state *a, struct lwtunnel_state *b)
{
	return memcmp(lwt_tun_info(a), lwt_tun_info(b),
		      sizeof(struct ip_tunnel_info));
}

static const struct lwtunnel_encap_ops ip_tun_lwt_ops = {
	.build_state = ip_tun_build_state,
	.fill_encap = ip_tun_fill_encap_info,
	.get_encap_size = ip_tun_encap_nlsize,
	.cmp_encap = ip_tun_cmp_encap,
};

static const struct nla_policy ip6_tun_policy[LWTUNNEL_IP6_MAX + 1] = {
	[LWTUNNEL_IP6_ID]		= { .type = NLA_U64 },
	[LWTUNNEL_IP6_DST]		= { .len = sizeof(struct in6_addr) },
	[LWTUNNEL_IP6_SRC]		= { .len = sizeof(struct in6_addr) },
	[LWTUNNEL_IP6_HOPLIMIT]		= { .type = NLA_U8 },
	[LWTUNNEL_IP6_TC]		= { .type = NLA_U8 },
	[LWTUNNEL_IP6_FLAGS]		= { .type = NLA_U16 },
};

static int ip6_tun_build_state(struct net_device *dev, struct nlattr *attr,
			       unsigned int family, const void *cfg,
			       struct lwtunnel_state **ts)
{
	struct ip_tunnel_info *tun_info;
	struct lwtunnel_state *new_state;
	struct nlattr *tb[LWTUNNEL_IP6_MAX + 1];
	int err;

	err = nla_parse_nested(tb, LWTUNNEL_IP6_MAX, attr, ip6_tun_policy);
	if (err < 0)
		return err;

	new_state = lwtunnel_state_alloc(sizeof(*tun_info));
	if (!new_state)
		return -ENOMEM;

	new_state->type = LWTUNNEL_ENCAP_IP6;

	tun_info = lwt_tun_info(new_state);

	if (tb[LWTUNNEL_IP6_ID])
		tun_info->key.tun_id = nla_get_be64(tb[LWTUNNEL_IP6_ID]);

	if (tb[LWTUNNEL_IP6_DST])
		tun_info->key.u.ipv6.dst = nla_get_in6_addr(tb[LWTUNNEL_IP6_DST]);

	if (tb[LWTUNNEL_IP6_SRC])
		tun_info->key.u.ipv6.src = nla_get_in6_addr(tb[LWTUNNEL_IP6_SRC]);

	if (tb[LWTUNNEL_IP6_HOPLIMIT])
		tun_info->key.ttl = nla_get_u8(tb[LWTUNNEL_IP6_HOPLIMIT]);

	if (tb[LWTUNNEL_IP6_TC])
		tun_info->key.tos = nla_get_u8(tb[LWTUNNEL_IP6_TC]);

	if (tb[LWTUNNEL_IP6_FLAGS])
		tun_info->key.tun_flags = nla_get_be16(tb[LWTUNNEL_IP6_FLAGS]);

	tun_info->mode = IP_TUNNEL_INFO_TX | IP_TUNNEL_INFO_IPV6;
	tun_info->options_len = 0;

	*ts = new_state;

	return 0;
}

static int ip6_tun_fill_encap_info(struct sk_buff *skb,
				   struct lwtunnel_state *lwtstate)
{
	struct ip_tunnel_info *tun_info = lwt_tun_info(lwtstate);

	if (nla_put_be64(skb, LWTUNNEL_IP6_ID, tun_info->key.tun_id) ||
	    nla_put_in6_addr(skb, LWTUNNEL_IP6_DST, &tun_info->key.u.ipv6.dst) ||
	    nla_put_in6_addr(skb, LWTUNNEL_IP6_SRC, &tun_info->key.u.ipv6.src) ||
	    nla_put_u8(skb, LWTUNNEL_IP6_HOPLIMIT, tun_info->key.tos) ||
	    nla_put_u8(skb, LWTUNNEL_IP6_TC, tun_info->key.ttl) ||
	    nla_put_be16(skb, LWTUNNEL_IP6_FLAGS, tun_info->key.tun_flags))
		return -ENOMEM;

	return 0;
}

static int ip6_tun_encap_nlsize(struct lwtunnel_state *lwtstate)
{
	return nla_total_size(8)	/* LWTUNNEL_IP6_ID */
		+ nla_total_size(16)	/* LWTUNNEL_IP6_DST */
		+ nla_total_size(16)	/* LWTUNNEL_IP6_SRC */
		+ nla_total_size(1)	/* LWTUNNEL_IP6_HOPLIMIT */
		+ nla_total_size(1)	/* LWTUNNEL_IP6_TC */
		+ nla_total_size(2);	/* LWTUNNEL_IP6_FLAGS */
}

static const struct lwtunnel_encap_ops ip6_tun_lwt_ops = {
	.build_state = ip6_tun_build_state,
	.fill_encap = ip6_tun_fill_encap_info,
	.get_encap_size = ip6_tun_encap_nlsize,
	.cmp_encap = ip_tun_cmp_encap,
};

void __init ip_tunnel_core_init(void)
{
	/* If you land here, make sure whether increasing ip_tunnel_info's
	 * options_len is a reasonable choice with its usage in front ends
	 * (f.e., it's part of flow keys, etc).
	 */
	BUILD_BUG_ON(IP_TUNNEL_OPTS_MAX != 255);

	lwtunnel_encap_add_ops(&ip_tun_lwt_ops, LWTUNNEL_ENCAP_IP);
	lwtunnel_encap_add_ops(&ip6_tun_lwt_ops, LWTUNNEL_ENCAP_IP6);
}

struct static_key ip_tunnel_metadata_cnt = STATIC_KEY_INIT_FALSE;
EXPORT_SYMBOL(ip_tunnel_metadata_cnt);

void ip_tunnel_need_metadata(void)
{
	static_key_slow_inc(&ip_tunnel_metadata_cnt);
}
EXPORT_SYMBOL_GPL(ip_tunnel_need_metadata);

void ip_tunnel_unneed_metadata(void)
{
	static_key_slow_dec(&ip_tunnel_metadata_cnt);
}
EXPORT_SYMBOL_GPL(ip_tunnel_unneed_metadata);
Commit	Line	Data
0e6fbc5b PS	1	/*
	2	* Copyright (c) 2013 Nicira, Inc.
	3	*
	4	* This program is free software; you can redistribute it and/or
	5	* modify it under the terms of version 2 of the GNU General Public
	6	* License as published by the Free Software Foundation.
	7	*
	8	* This program is distributed in the hope that it will be useful, but
	9	* WITHOUT ANY WARRANTY; without even the implied warranty of
	10	* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
	11	* General Public License for more details.
	12	*
	13	* You should have received a copy of the GNU General Public License
	14	* along with this program; if not, write to the Free Software
	15	* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
	16	* 02110-1301, USA
	17	*/
	18
	19	#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
	20
	21	#include <linux/types.h>
	22	#include <linux/kernel.h>
	23	#include <linux/skbuff.h>
	24	#include <linux/netdevice.h>
	25	#include <linux/in.h>
	26	#include <linux/if_arp.h>
0e6fbc5b PS	27	#include <linux/init.h>
	28	#include <linux/in6.h>
	29	#include <linux/inetdevice.h>
	30	#include <linux/netfilter_ipv4.h>
	31	#include <linux/etherdevice.h>
	32	#include <linux/if_ether.h>
	33	#include <linux/if_vlan.h>
e7030878	34	#include <linux/static_key.h>
0e6fbc5b PS	35
	36	#include <net/ip.h>
	37	#include <net/icmp.h>
	38	#include <net/protocol.h>
	39	#include <net/ip_tunnels.h>
	40	#include <net/arp.h>
	41	#include <net/checksum.h>
	42	#include <net/dsfield.h>
	43	#include <net/inet_ecn.h>
	44	#include <net/xfrm.h>
	45	#include <net/net_namespace.h>
	46	#include <net/netns/generic.h>
	47	#include <net/rtnetlink.h>
63d008a4	48	#include <net/dst_metadata.h>
0e6fbc5b	49
039f5062 PS	50	void iptunnel_xmit(struct sock sk, struct rtable rt, struct sk_buff *skb,
	51	__be32 src, __be32 dst, __u8 proto,
	52	__u8 tos, __u8 ttl, __be16 df, bool xnet)
0e6fbc5b	53	{
bc22a0e2	54	int pkt_len = skb->len - skb_inner_network_offset(skb);
f859b0f6	55	struct net *net = dev_net(rt->dst.dev);
039f5062	56	struct net_device *dev = skb->dev;
0e6fbc5b PS	57	struct iphdr *iph;
	58	int err;
	59
963a88b3 ND	60	skb_scrub_packet(skb, xnet);
963a88b3 ND	61
7539fadc	62	skb_clear_hash(skb);
0e6fbc5b PS	63	skb_dst_set(skb, &rt->dst);
	64	memset(IPCB(skb), 0, sizeof(*IPCB(skb)));
	65
	66	/* Push down and install the IP header. */
78a3694d	67	skb_push(skb, sizeof(struct iphdr));
0e6fbc5b PS	68	skb_reset_network_header(skb);
	69
	70	iph = ip_hdr(skb);
	71
	72	iph->version = 4;
	73	iph->ihl = sizeof(struct iphdr) >> 2;
	74	iph->frag_off = df;
	75	iph->protocol = proto;
	76	iph->tos = tos;
	77	iph->daddr = dst;
	78	iph->saddr = src;
	79	iph->ttl = ttl;
f859b0f6	80	__ip_select_ident(net, iph, skb_shinfo(skb)->gso_segs ?: 1);
0e6fbc5b	81
33224b16	82	err = ip_local_out(net, sk, skb);
0e6fbc5b PS	83	if (unlikely(net_xmit_eval(err)))
0e6fbc5b PS	84	pkt_len = 0;
039f5062	85	iptunnel_xmit_stats(dev, pkt_len);
0e6fbc5b PS	86	}
0e6fbc5b PS	87	EXPORT_SYMBOL_GPL(iptunnel_xmit);
3d7b46cd	88
7f290c94 JB	89	int iptunnel_pull_header(struct sk_buff *skb, int hdr_len, __be16 inner_proto,
7f290c94 JB	90	bool xnet)
3d7b46cd PS	91	{
	92	if (unlikely(!pskb_may_pull(skb, hdr_len)))
	93	return -ENOMEM;
	94
	95	skb_pull_rcsum(skb, hdr_len);
	96
	97	if (inner_proto == htons(ETH_P_TEB)) {
1245dfc8	98	struct ethhdr *eh;
3d7b46cd PS	99
	100	if (unlikely(!pskb_may_pull(skb, ETH_HLEN)))
	101	return -ENOMEM;
	102
1245dfc8	103	eh = (struct ethhdr *)skb->data;
d181ddca	104	if (likely(eth_proto_is_802_3(eh->h_proto)))
3d7b46cd PS	105	skb->protocol = eh->h_proto;
	106	else
	107	skb->protocol = htons(ETH_P_802_2);
	108
	109	} else {
	110	skb->protocol = inner_proto;
	111	}
	112
7539fadc	113	skb_clear_hash_if_not_l4(skb);
3d7b46cd PS	114	skb->vlan_tci = 0;
3d7b46cd PS	115	skb_set_queue_mapping(skb, 0);
7f290c94	116	skb_scrub_packet(skb, xnet);
3d7b46cd PS	117	return 0;
	118	}
	119	EXPORT_SYMBOL_GPL(iptunnel_pull_header);
2d26f0a3	120
63d008a4 JB	121	struct metadata_dst iptunnel_metadata_reply(struct metadata_dst md,
	122	gfp_t flags)
	123	{
	124	struct metadata_dst *res;
	125	struct ip_tunnel_info dst, src;
	126
	127	if (!md \|\| md->u.tun_info.mode & IP_TUNNEL_INFO_TX)
	128	return NULL;
	129
	130	res = metadata_dst_alloc(0, flags);
	131	if (!res)
	132	return NULL;
	133
	134	dst = &res->u.tun_info;
	135	src = &md->u.tun_info;
	136	dst->key.tun_id = src->key.tun_id;
	137	if (src->mode & IP_TUNNEL_INFO_IPV6)
	138	memcpy(&dst->key.u.ipv6.dst, &src->key.u.ipv6.src,
	139	sizeof(struct in6_addr));
	140	else
	141	dst->key.u.ipv4.dst = src->key.u.ipv4.src;
	142	dst->mode = src->mode \| IP_TUNNEL_INFO_TX;
	143
	144	return res;
	145	}
	146	EXPORT_SYMBOL_GPL(iptunnel_metadata_reply);
	147
2d26f0a3	148	struct sk_buff iptunnel_handle_offloads(struct sk_buff skb,
2d26f0a3 ED	149	int gso_type_mask)
	150	{
	151	int err;
	152
	153	if (likely(!skb->encapsulation)) {
	154	skb_reset_inner_headers(skb);
	155	skb->encapsulation = 1;
	156	}
	157
	158	if (skb_is_gso(skb)) {
	159	err = skb_unclone(skb, GFP_ATOMIC);
	160	if (unlikely(err))
	161	goto error;
	162	skb_shinfo(skb)->gso_type \|= gso_type_mask;
	163	return skb;
	164	}
	165
6fa79666	166	if (skb->ip_summed != CHECKSUM_PARTIAL) {
2d26f0a3	167	skb->ip_summed = CHECKSUM_NONE;
6fa79666 EC	168	/* We clear encapsulation here to prevent badly-written
	169	* drivers potentially deciding to offload an inner checksum
	170	* if we set CHECKSUM_PARTIAL on the outer header.
	171	* This should go away when the drivers are all fixed.
	172	*/
179bc67f EC	173	skb->encapsulation = 0;
179bc67f EC	174	}
2d26f0a3 ED	175
	176	return skb;
	177	error:
	178	kfree_skb(skb);
	179	return ERR_PTR(err);
	180	}
	181	EXPORT_SYMBOL_GPL(iptunnel_handle_offloads);
ebe44f35 DM	182
	183	/* Often modified stats are per cpu, other are shared (netdev->stats) */
	184	struct rtnl_link_stats64 ip_tunnel_get_stats64(struct net_device dev,
	185	struct rtnl_link_stats64 *tot)
	186	{
	187	int i;
	188
c24a5964 AD	189	netdev_stats_to_stats64(tot, &dev->stats);
c24a5964 AD	190
ebe44f35 DM	191	for_each_possible_cpu(i) {
	192	const struct pcpu_sw_netstats *tstats =
	193	per_cpu_ptr(dev->tstats, i);
	194	u64 rx_packets, rx_bytes, tx_packets, tx_bytes;
	195	unsigned int start;
	196
	197	do {
57a7744e	198	start = u64_stats_fetch_begin_irq(&tstats->syncp);
ebe44f35 DM	199	rx_packets = tstats->rx_packets;
	200	tx_packets = tstats->tx_packets;
	201	rx_bytes = tstats->rx_bytes;
	202	tx_bytes = tstats->tx_bytes;
57a7744e	203	} while (u64_stats_fetch_retry_irq(&tstats->syncp, start));
ebe44f35 DM	204
	205	tot->rx_packets += rx_packets;
	206	tot->tx_packets += tx_packets;
	207	tot->rx_bytes += rx_bytes;
	208	tot->tx_bytes += tx_bytes;
	209	}
	210
ebe44f35 DM	211	return tot;
	212	}
	213	EXPORT_SYMBOL_GPL(ip_tunnel_get_stats64);
3093fbe7	214
a1c234f9 JB	215	static const struct nla_policy ip_tun_policy[LWTUNNEL_IP_MAX + 1] = {
	216	[LWTUNNEL_IP_ID] = { .type = NLA_U64 },
	217	[LWTUNNEL_IP_DST] = { .type = NLA_U32 },
	218	[LWTUNNEL_IP_SRC] = { .type = NLA_U32 },
	219	[LWTUNNEL_IP_TTL] = { .type = NLA_U8 },
	220	[LWTUNNEL_IP_TOS] = { .type = NLA_U8 },
a1c234f9	221	[LWTUNNEL_IP_FLAGS] = { .type = NLA_U16 },
3093fbe7 TG	222	};
	223
	224	static int ip_tun_build_state(struct net_device dev, struct nlattr attr,
127eb7cd	225	unsigned int family, const void *cfg,
3093fbe7 TG	226	struct lwtunnel_state **ts)
	227	{
	228	struct ip_tunnel_info *tun_info;
	229	struct lwtunnel_state *new_state;
a1c234f9	230	struct nlattr *tb[LWTUNNEL_IP_MAX + 1];
3093fbe7 TG	231	int err;
3093fbe7 TG	232
a1c234f9	233	err = nla_parse_nested(tb, LWTUNNEL_IP_MAX, attr, ip_tun_policy);
3093fbe7 TG	234	if (err < 0)
	235	return err;
	236
	237	new_state = lwtunnel_state_alloc(sizeof(*tun_info));
	238	if (!new_state)
	239	return -ENOMEM;
	240
	241	new_state->type = LWTUNNEL_ENCAP_IP;
	242
	243	tun_info = lwt_tun_info(new_state);
	244
a1c234f9	245	if (tb[LWTUNNEL_IP_ID])
30d3d83a	246	tun_info->key.tun_id = nla_get_be64(tb[LWTUNNEL_IP_ID]);
3093fbe7	247
a1c234f9	248	if (tb[LWTUNNEL_IP_DST])
c1ea5d67	249	tun_info->key.u.ipv4.dst = nla_get_be32(tb[LWTUNNEL_IP_DST]);
3093fbe7	250
a1c234f9	251	if (tb[LWTUNNEL_IP_SRC])
c1ea5d67	252	tun_info->key.u.ipv4.src = nla_get_be32(tb[LWTUNNEL_IP_SRC]);
3093fbe7	253
a1c234f9	254	if (tb[LWTUNNEL_IP_TTL])
7c383fb2	255	tun_info->key.ttl = nla_get_u8(tb[LWTUNNEL_IP_TTL]);
3093fbe7	256
a1c234f9	257	if (tb[LWTUNNEL_IP_TOS])
7c383fb2	258	tun_info->key.tos = nla_get_u8(tb[LWTUNNEL_IP_TOS]);
3093fbe7	259
a1c234f9	260	if (tb[LWTUNNEL_IP_FLAGS])
30d3d83a	261	tun_info->key.tun_flags = nla_get_be16(tb[LWTUNNEL_IP_FLAGS]);
3093fbe7 TG	262
3093fbe7 TG	263	tun_info->mode = IP_TUNNEL_INFO_TX;
3093fbe7 TG	264	tun_info->options_len = 0;
	265
	266	*ts = new_state;
	267
	268	return 0;
	269	}
	270
	271	static int ip_tun_fill_encap_info(struct sk_buff *skb,
	272	struct lwtunnel_state *lwtstate)
	273	{
	274	struct ip_tunnel_info *tun_info = lwt_tun_info(lwtstate);
	275
30d3d83a	276	if (nla_put_be64(skb, LWTUNNEL_IP_ID, tun_info->key.tun_id) \|\|
c1ea5d67 JB	277	nla_put_be32(skb, LWTUNNEL_IP_DST, tun_info->key.u.ipv4.dst) \|\|
c1ea5d67 JB	278	nla_put_be32(skb, LWTUNNEL_IP_SRC, tun_info->key.u.ipv4.src) \|\|
7c383fb2 JB	279	nla_put_u8(skb, LWTUNNEL_IP_TOS, tun_info->key.tos) \|\|
7c383fb2 JB	280	nla_put_u8(skb, LWTUNNEL_IP_TTL, tun_info->key.ttl) \|\|
30d3d83a	281	nla_put_be16(skb, LWTUNNEL_IP_FLAGS, tun_info->key.tun_flags))
3093fbe7 TG	282	return -ENOMEM;
	283
	284	return 0;
	285	}
	286
	287	static int ip_tun_encap_nlsize(struct lwtunnel_state *lwtstate)
	288	{
a1c234f9 JB	289	return nla_total_size(8) /* LWTUNNEL_IP_ID */
	290	+ nla_total_size(4) /* LWTUNNEL_IP_DST */
	291	+ nla_total_size(4) /* LWTUNNEL_IP_SRC */
	292	+ nla_total_size(1) /* LWTUNNEL_IP_TOS */
	293	+ nla_total_size(1) /* LWTUNNEL_IP_TTL */
a1c234f9	294	+ nla_total_size(2); /* LWTUNNEL_IP_FLAGS */
3093fbe7 TG	295	}
3093fbe7 TG	296
2d798499 JB	297	static int ip_tun_cmp_encap(struct lwtunnel_state a, struct lwtunnel_state b)
	298	{
	299	return memcmp(lwt_tun_info(a), lwt_tun_info(b),
	300	sizeof(struct ip_tunnel_info));
	301	}
	302
3093fbe7 TG	303	static const struct lwtunnel_encap_ops ip_tun_lwt_ops = {
	304	.build_state = ip_tun_build_state,
	305	.fill_encap = ip_tun_fill_encap_info,
	306	.get_encap_size = ip_tun_encap_nlsize,
2d798499	307	.cmp_encap = ip_tun_cmp_encap,
3093fbe7 TG	308	};
3093fbe7 TG	309
32a2b002 JB	310	static const struct nla_policy ip6_tun_policy[LWTUNNEL_IP6_MAX + 1] = {
	311	[LWTUNNEL_IP6_ID] = { .type = NLA_U64 },
	312	[LWTUNNEL_IP6_DST] = { .len = sizeof(struct in6_addr) },
	313	[LWTUNNEL_IP6_SRC] = { .len = sizeof(struct in6_addr) },
	314	[LWTUNNEL_IP6_HOPLIMIT] = { .type = NLA_U8 },
	315	[LWTUNNEL_IP6_TC] = { .type = NLA_U8 },
32a2b002 JB	316	[LWTUNNEL_IP6_FLAGS] = { .type = NLA_U16 },
	317	};
	318
	319	static int ip6_tun_build_state(struct net_device dev, struct nlattr attr,
127eb7cd	320	unsigned int family, const void *cfg,
32a2b002 JB	321	struct lwtunnel_state **ts)
	322	{
	323	struct ip_tunnel_info *tun_info;
	324	struct lwtunnel_state *new_state;
	325	struct nlattr *tb[LWTUNNEL_IP6_MAX + 1];
	326	int err;
	327
	328	err = nla_parse_nested(tb, LWTUNNEL_IP6_MAX, attr, ip6_tun_policy);
	329	if (err < 0)
	330	return err;
	331
	332	new_state = lwtunnel_state_alloc(sizeof(*tun_info));
	333	if (!new_state)
	334	return -ENOMEM;
	335
	336	new_state->type = LWTUNNEL_ENCAP_IP6;
	337
	338	tun_info = lwt_tun_info(new_state);
	339
	340	if (tb[LWTUNNEL_IP6_ID])
30d3d83a	341	tun_info->key.tun_id = nla_get_be64(tb[LWTUNNEL_IP6_ID]);
32a2b002 JB	342
	343	if (tb[LWTUNNEL_IP6_DST])
	344	tun_info->key.u.ipv6.dst = nla_get_in6_addr(tb[LWTUNNEL_IP6_DST]);
	345
	346	if (tb[LWTUNNEL_IP6_SRC])
	347	tun_info->key.u.ipv6.src = nla_get_in6_addr(tb[LWTUNNEL_IP6_SRC]);
	348
	349	if (tb[LWTUNNEL_IP6_HOPLIMIT])
	350	tun_info->key.ttl = nla_get_u8(tb[LWTUNNEL_IP6_HOPLIMIT]);
	351
	352	if (tb[LWTUNNEL_IP6_TC])
	353	tun_info->key.tos = nla_get_u8(tb[LWTUNNEL_IP6_TC]);
	354
32a2b002	355	if (tb[LWTUNNEL_IP6_FLAGS])
30d3d83a	356	tun_info->key.tun_flags = nla_get_be16(tb[LWTUNNEL_IP6_FLAGS]);
32a2b002	357
7f9562a1	358	tun_info->mode = IP_TUNNEL_INFO_TX \| IP_TUNNEL_INFO_IPV6;
32a2b002 JB	359	tun_info->options_len = 0;
	360
	361	*ts = new_state;
	362
	363	return 0;
	364	}
	365
	366	static int ip6_tun_fill_encap_info(struct sk_buff *skb,
	367	struct lwtunnel_state *lwtstate)
	368	{
	369	struct ip_tunnel_info *tun_info = lwt_tun_info(lwtstate);
	370
30d3d83a	371	if (nla_put_be64(skb, LWTUNNEL_IP6_ID, tun_info->key.tun_id) \|\|
32a2b002 JB	372	nla_put_in6_addr(skb, LWTUNNEL_IP6_DST, &tun_info->key.u.ipv6.dst) \|\|
	373	nla_put_in6_addr(skb, LWTUNNEL_IP6_SRC, &tun_info->key.u.ipv6.src) \|\|
	374	nla_put_u8(skb, LWTUNNEL_IP6_HOPLIMIT, tun_info->key.tos) \|\|
	375	nla_put_u8(skb, LWTUNNEL_IP6_TC, tun_info->key.ttl) \|\|
30d3d83a	376	nla_put_be16(skb, LWTUNNEL_IP6_FLAGS, tun_info->key.tun_flags))
32a2b002 JB	377	return -ENOMEM;
	378
	379	return 0;
	380	}
	381
	382	static int ip6_tun_encap_nlsize(struct lwtunnel_state *lwtstate)
	383	{
	384	return nla_total_size(8) /* LWTUNNEL_IP6_ID */
	385	+ nla_total_size(16) /* LWTUNNEL_IP6_DST */
	386	+ nla_total_size(16) /* LWTUNNEL_IP6_SRC */
	387	+ nla_total_size(1) /* LWTUNNEL_IP6_HOPLIMIT */
	388	+ nla_total_size(1) /* LWTUNNEL_IP6_TC */
32a2b002 JB	389	+ nla_total_size(2); /* LWTUNNEL_IP6_FLAGS */
	390	}
	391
	392	static const struct lwtunnel_encap_ops ip6_tun_lwt_ops = {
	393	.build_state = ip6_tun_build_state,
	394	.fill_encap = ip6_tun_fill_encap_info,
	395	.get_encap_size = ip6_tun_encap_nlsize,
	396	.cmp_encap = ip_tun_cmp_encap,
	397	};
	398
045a0fa0	399	void __init ip_tunnel_core_init(void)
3093fbe7	400	{
fca5fdf6 DB	401	/* If you land here, make sure whether increasing ip_tunnel_info's
	402	* options_len is a reasonable choice with its usage in front ends
	403	* (f.e., it's part of flow keys, etc).
	404	*/
	405	BUILD_BUG_ON(IP_TUNNEL_OPTS_MAX != 255);
	406
3093fbe7	407	lwtunnel_encap_add_ops(&ip_tun_lwt_ops, LWTUNNEL_ENCAP_IP);
32a2b002	408	lwtunnel_encap_add_ops(&ip6_tun_lwt_ops, LWTUNNEL_ENCAP_IP6);
3093fbe7	409	}
e7030878 TG	410
	411	struct static_key ip_tunnel_metadata_cnt = STATIC_KEY_INIT_FALSE;
	412	EXPORT_SYMBOL(ip_tunnel_metadata_cnt);
	413
	414	void ip_tunnel_need_metadata(void)
	415	{
	416	static_key_slow_inc(&ip_tunnel_metadata_cnt);
	417	}
	418	EXPORT_SYMBOL_GPL(ip_tunnel_need_metadata);
	419
	420	void ip_tunnel_unneed_metadata(void)
	421	{
	422	static_key_slow_dec(&ip_tunnel_metadata_cnt);
	423	}
	424	EXPORT_SYMBOL_GPL(ip_tunnel_unneed_metadata);