Contributors: 55
Author Tokens Token Proportion Commits Commit Proportion
Herbert Xu 1082 27.08% 37 19.47%
Steffen Klassert 724 18.12% 18 9.47%
Florian Westphal 426 10.66% 16 8.42%
Alexey Kuznetsov 332 8.31% 4 2.11%
Diego Beltrami 179 4.48% 1 0.53%
Arnaldo Carvalho de Melo 112 2.80% 11 5.79%
Masahide Nakamura 105 2.63% 2 1.05%
Linus Torvalds (pre-git) 90 2.25% 16 8.42%
Alexander Duyck 84 2.10% 1 0.53%
Alexey Dobriyan 63 1.58% 7 3.68%
LiuJian 62 1.55% 1 0.53%
Kazunori Miyazawa 62 1.55% 4 2.11%
Xin Long 61 1.53% 2 1.05%
Hideaki Yoshifuji / 吉藤英明 60 1.50% 5 2.63%
Antony Antony 47 1.18% 2 1.05%
Sabrina Dubroca 44 1.10% 2 1.05%
Yossi Kuperman 41 1.03% 3 1.58%
Paul Davey 40 1.00% 1 0.53%
Aviv Heller 39 0.98% 1 0.53%
David S. Miller 38 0.95% 6 3.16%
Patrick McHardy 32 0.80% 6 3.16%
Linus Torvalds 22 0.55% 1 0.53%
Paul Moore 20 0.50% 1 0.53%
Fan Du 17 0.43% 1 0.53%
Eyal Birger 17 0.43% 1 0.53%
Eric Dumazet 16 0.40% 5 2.63%
Subash Abhinov Kasiviswanathan 14 0.35% 1 0.53%
Xiaodong Xu 14 0.35% 1 0.53%
Mitsuru Kanda 13 0.33% 1 0.53%
Ilan Tayari 12 0.30% 1 0.53%
Adrian Bunk 10 0.25% 1 0.53%
Alexey Kodanev 10 0.25% 2 1.05%
Mike Yu 10 0.25% 2 1.05%
Jamal Hadi Salim 10 0.25% 2 1.05%
Derek Atkins 8 0.20% 1 0.53%
Jean-Francois Dive 7 0.18% 1 0.53%
James Chapman 6 0.15% 1 0.53%
Allen Pais 6 0.15% 1 0.53%
Li RongQing 6 0.15% 2 1.05%
Benedict Wong 6 0.15% 1 0.53%
Al Viro 5 0.13% 1 0.53%
Eric W. Biedermann 5 0.13% 1 0.53%
James Morris 5 0.13% 1 0.53%
Miika Komu 5 0.13% 2 1.05%
Sowmini Varadhan 4 0.10% 1 0.53%
Stephen Hemminger 4 0.10% 1 0.53%
Thomas Egerer 4 0.10% 1 0.53%
Yury Polyanskiy 3 0.08% 1 0.53%
Christian Hopps 3 0.08% 1 0.53%
Kuniyuki Iwashima 3 0.08% 1 0.53%
Jiri Benc 2 0.05% 1 0.53%
Andi Kleen 2 0.05% 2 1.05%
Pravin B Shelar 2 0.05% 1 0.53%
Greg Kroah-Hartman 1 0.03% 1 0.53%
Arnd Bergmann 1 0.03% 1 0.53%
Total 3996 190


// SPDX-License-Identifier: GPL-2.0
/*
 * xfrm_input.c
 *
 * Changes:
 * 	YOSHIFUJI Hideaki @USAGI
 * 		Split up af-specific portion
 *
 */

#include <linux/bottom_half.h>
#include <linux/cache.h>
#include <linux/interrupt.h>
#include <linux/slab.h>
#include <linux/module.h>
#include <linux/netdevice.h>
#include <linux/percpu.h>
#include <net/dst.h>
#include <net/ip.h>
#include <net/xfrm.h>
#include <net/ip_tunnels.h>
#include <net/ip6_tunnel.h>
#include <net/dst_metadata.h>
#include <net/hotdata.h>

#include "xfrm_inout.h"

struct xfrm_trans_tasklet {
	struct work_struct work;
	spinlock_t queue_lock;
	struct sk_buff_head queue;
};

struct xfrm_trans_cb {
	union {
		struct inet_skb_parm	h4;
#if IS_ENABLED(CONFIG_IPV6)
		struct inet6_skb_parm	h6;
#endif
	} header;
	int (*finish)(struct net *net, struct sock *sk, struct sk_buff *skb);
	struct net *net;
};

#define XFRM_TRANS_SKB_CB(__skb) ((struct xfrm_trans_cb *)&((__skb)->cb[0]))

static DEFINE_SPINLOCK(xfrm_input_afinfo_lock);
static struct xfrm_input_afinfo const __rcu *xfrm_input_afinfo[2][AF_INET6 + 1];

static struct gro_cells gro_cells;
static struct net_device xfrm_napi_dev;

static DEFINE_PER_CPU(struct xfrm_trans_tasklet, xfrm_trans_tasklet);

int xfrm_input_register_afinfo(const struct xfrm_input_afinfo *afinfo)
{
	int err = 0;

	if (WARN_ON(afinfo->family > AF_INET6))
		return -EAFNOSUPPORT;

	spin_lock_bh(&xfrm_input_afinfo_lock);
	if (unlikely(xfrm_input_afinfo[afinfo->is_ipip][afinfo->family]))
		err = -EEXIST;
	else
		rcu_assign_pointer(xfrm_input_afinfo[afinfo->is_ipip][afinfo->family], afinfo);
	spin_unlock_bh(&xfrm_input_afinfo_lock);
	return err;
}
EXPORT_SYMBOL(xfrm_input_register_afinfo);

int xfrm_input_unregister_afinfo(const struct xfrm_input_afinfo *afinfo)
{
	int err = 0;

	spin_lock_bh(&xfrm_input_afinfo_lock);
	if (likely(xfrm_input_afinfo[afinfo->is_ipip][afinfo->family])) {
		if (unlikely(xfrm_input_afinfo[afinfo->is_ipip][afinfo->family] != afinfo))
			err = -EINVAL;
		else
			RCU_INIT_POINTER(xfrm_input_afinfo[afinfo->is_ipip][afinfo->family], NULL);
	}
	spin_unlock_bh(&xfrm_input_afinfo_lock);
	synchronize_rcu();
	return err;
}
EXPORT_SYMBOL(xfrm_input_unregister_afinfo);

static const struct xfrm_input_afinfo *xfrm_input_get_afinfo(u8 family, bool is_ipip)
{
	const struct xfrm_input_afinfo *afinfo;

	if (WARN_ON_ONCE(family > AF_INET6))
		return NULL;

	rcu_read_lock();
	afinfo = rcu_dereference(xfrm_input_afinfo[is_ipip][family]);
	if (unlikely(!afinfo))
		rcu_read_unlock();
	return afinfo;
}

static int xfrm_rcv_cb(struct sk_buff *skb, unsigned int family, u8 protocol,
		       int err)
{
	bool is_ipip = (protocol == IPPROTO_IPIP || protocol == IPPROTO_IPV6);
	const struct xfrm_input_afinfo *afinfo;
	int ret;

	afinfo = xfrm_input_get_afinfo(family, is_ipip);
	if (!afinfo)
		return -EAFNOSUPPORT;

	ret = afinfo->callback(skb, protocol, err);
	rcu_read_unlock();

	return ret;
}

struct sec_path *secpath_set(struct sk_buff *skb)
{
	struct sec_path *sp, *tmp = skb_ext_find(skb, SKB_EXT_SEC_PATH);

	sp = skb_ext_add(skb, SKB_EXT_SEC_PATH);
	if (!sp)
		return NULL;

	if (tmp) /* reused existing one (was COW'd if needed) */
		return sp;

	/* allocated new secpath */
	memset(sp->ovec, 0, sizeof(sp->ovec));
	sp->olen = 0;
	sp->len = 0;
	sp->verified_cnt = 0;

	return sp;
}
EXPORT_SYMBOL(secpath_set);

/* Fetch spi and seq from ipsec header */

int xfrm_parse_spi(struct sk_buff *skb, u8 nexthdr, __be32 *spi, __be32 *seq)
{
	int offset, offset_seq;
	int hlen;

	switch (nexthdr) {
	case IPPROTO_AH:
		hlen = sizeof(struct ip_auth_hdr);
		offset = offsetof(struct ip_auth_hdr, spi);
		offset_seq = offsetof(struct ip_auth_hdr, seq_no);
		break;
	case IPPROTO_ESP:
		hlen = sizeof(struct ip_esp_hdr);
		offset = offsetof(struct ip_esp_hdr, spi);
		offset_seq = offsetof(struct ip_esp_hdr, seq_no);
		break;
	case IPPROTO_COMP:
		if (!pskb_may_pull(skb, sizeof(struct ip_comp_hdr)))
			return -EINVAL;
		*spi = htonl(ntohs(*(__be16 *)(skb_transport_header(skb) + 2)));
		*seq = 0;
		return 0;
	default:
		return 1;
	}

	if (!pskb_may_pull(skb, hlen))
		return -EINVAL;

	*spi = *(__be32 *)(skb_transport_header(skb) + offset);
	*seq = *(__be32 *)(skb_transport_header(skb) + offset_seq);
	return 0;
}
EXPORT_SYMBOL(xfrm_parse_spi);

static int xfrm4_remove_beet_encap(struct xfrm_state *x, struct sk_buff *skb)
{
	struct iphdr *iph;
	int optlen = 0;
	int err = -EINVAL;

	skb->protocol = htons(ETH_P_IP);

	if (unlikely(XFRM_MODE_SKB_CB(skb)->protocol == IPPROTO_BEETPH)) {
		struct ip_beet_phdr *ph;
		int phlen;

		if (!pskb_may_pull(skb, sizeof(*ph)))
			goto out;

		ph = (struct ip_beet_phdr *)skb->data;

		phlen = sizeof(*ph) + ph->padlen;
		optlen = ph->hdrlen * 8 + (IPV4_BEET_PHMAXLEN - phlen);
		if (optlen < 0 || optlen & 3 || optlen > 250)
			goto out;

		XFRM_MODE_SKB_CB(skb)->protocol = ph->nexthdr;

		if (!pskb_may_pull(skb, phlen))
			goto out;
		__skb_pull(skb, phlen);
	}

	skb_push(skb, sizeof(*iph));
	skb_reset_network_header(skb);
	skb_mac_header_rebuild(skb);

	xfrm4_beet_make_header(skb);

	iph = ip_hdr(skb);

	iph->ihl += optlen / 4;
	iph->tot_len = htons(skb->len);
	iph->daddr = x->sel.daddr.a4;
	iph->saddr = x->sel.saddr.a4;
	iph->check = 0;
	iph->check = ip_fast_csum(skb_network_header(skb), iph->ihl);
	err = 0;
out:
	return err;
}

static void ipip_ecn_decapsulate(struct sk_buff *skb)
{
	struct iphdr *inner_iph = ipip_hdr(skb);

	if (INET_ECN_is_ce(XFRM_MODE_SKB_CB(skb)->tos))
		IP_ECN_set_ce(inner_iph);
}

static int xfrm4_remove_tunnel_encap(struct xfrm_state *x, struct sk_buff *skb)
{
	int err = -EINVAL;

	skb->protocol = htons(ETH_P_IP);

	if (!pskb_may_pull(skb, sizeof(struct iphdr)))
		goto out;

	err = skb_unclone(skb, GFP_ATOMIC);
	if (err)
		goto out;

	if (x->props.flags & XFRM_STATE_DECAP_DSCP)
		ipv4_copy_dscp(XFRM_MODE_SKB_CB(skb)->tos, ipip_hdr(skb));
	if (!(x->props.flags & XFRM_STATE_NOECN))
		ipip_ecn_decapsulate(skb);

	skb_reset_network_header(skb);
	skb_mac_header_rebuild(skb);
	if (skb->mac_len)
		eth_hdr(skb)->h_proto = skb->protocol;

	err = 0;

out:
	return err;
}

static void ipip6_ecn_decapsulate(struct sk_buff *skb)
{
	struct ipv6hdr *inner_iph = ipipv6_hdr(skb);

	if (INET_ECN_is_ce(XFRM_MODE_SKB_CB(skb)->tos))
		IP6_ECN_set_ce(skb, inner_iph);
}

static int xfrm6_remove_tunnel_encap(struct xfrm_state *x, struct sk_buff *skb)
{
	int err = -EINVAL;

	skb->protocol = htons(ETH_P_IPV6);

	if (!pskb_may_pull(skb, sizeof(struct ipv6hdr)))
		goto out;

	err = skb_unclone(skb, GFP_ATOMIC);
	if (err)
		goto out;

	if (x->props.flags & XFRM_STATE_DECAP_DSCP)
		ipv6_copy_dscp(XFRM_MODE_SKB_CB(skb)->tos, ipipv6_hdr(skb));
	if (!(x->props.flags & XFRM_STATE_NOECN))
		ipip6_ecn_decapsulate(skb);

	skb_reset_network_header(skb);
	skb_mac_header_rebuild(skb);
	if (skb->mac_len)
		eth_hdr(skb)->h_proto = skb->protocol;

	err = 0;

out:
	return err;
}

static int xfrm6_remove_beet_encap(struct xfrm_state *x, struct sk_buff *skb)
{
	struct ipv6hdr *ip6h;
	int size = sizeof(struct ipv6hdr);
	int err;

	skb->protocol = htons(ETH_P_IPV6);

	err = skb_cow_head(skb, size + skb->mac_len);
	if (err)
		goto out;

	__skb_push(skb, size);
	skb_reset_network_header(skb);
	skb_mac_header_rebuild(skb);

	xfrm6_beet_make_header(skb);

	ip6h = ipv6_hdr(skb);
	ip6h->payload_len = htons(skb->len - size);
	ip6h->daddr = x->sel.daddr.in6;
	ip6h->saddr = x->sel.saddr.in6;
	err = 0;
out:
	return err;
}

/* Remove encapsulation header.
 *
 * The IP header will be moved over the top of the encapsulation
 * header.
 *
 * On entry, the transport header shall point to where the IP header
 * should be and the network header shall be set to where the IP
 * header currently is.  skb->data shall point to the start of the
 * payload.
 */
static int
xfrm_inner_mode_encap_remove(struct xfrm_state *x,
			     struct sk_buff *skb)
{
	switch (x->props.mode) {
	case XFRM_MODE_BEET:
		switch (x->sel.family) {
		case AF_INET:
			return xfrm4_remove_beet_encap(x, skb);
		case AF_INET6:
			return xfrm6_remove_beet_encap(x, skb);
		}
		break;
	case XFRM_MODE_TUNNEL:
		switch (XFRM_MODE_SKB_CB(skb)->protocol) {
		case IPPROTO_IPIP:
			return xfrm4_remove_tunnel_encap(x, skb);
		case IPPROTO_IPV6:
			return xfrm6_remove_tunnel_encap(x, skb);
		break;
		}
		return -EINVAL;
	}

	WARN_ON_ONCE(1);
	return -EOPNOTSUPP;
}

static int xfrm_prepare_input(struct xfrm_state *x, struct sk_buff *skb)
{
	switch (x->props.family) {
	case AF_INET:
		xfrm4_extract_header(skb);
		break;
	case AF_INET6:
		xfrm6_extract_header(skb);
		break;
	default:
		WARN_ON_ONCE(1);
		return -EAFNOSUPPORT;
	}

	return xfrm_inner_mode_encap_remove(x, skb);
}

/* Remove encapsulation header.
 *
 * The IP header will be moved over the top of the encapsulation header.
 *
 * On entry, skb_transport_header() shall point to where the IP header
 * should be and skb_network_header() shall be set to where the IP header
 * currently is.  skb->data shall point to the start of the payload.
 */
static int xfrm4_transport_input(struct xfrm_state *x, struct sk_buff *skb)
{
	struct xfrm_offload *xo = xfrm_offload(skb);
	int ihl = skb->data - skb_transport_header(skb);

	if (skb->transport_header != skb->network_header) {
		memmove(skb_transport_header(skb),
			skb_network_header(skb), ihl);
		if (xo)
			xo->orig_mac_len =
				skb_mac_header_was_set(skb) ? skb_mac_header_len(skb) : 0;
		skb->network_header = skb->transport_header;
	}
	ip_hdr(skb)->tot_len = htons(skb->len + ihl);
	skb_reset_transport_header(skb);
	return 0;
}

static int xfrm6_transport_input(struct xfrm_state *x, struct sk_buff *skb)
{
#if IS_ENABLED(CONFIG_IPV6)
	struct xfrm_offload *xo = xfrm_offload(skb);
	int ihl = skb->data - skb_transport_header(skb);

	if (skb->transport_header != skb->network_header) {
		memmove(skb_transport_header(skb),
			skb_network_header(skb), ihl);
		if (xo)
			xo->orig_mac_len =
				skb_mac_header_was_set(skb) ? skb_mac_header_len(skb) : 0;
		skb->network_header = skb->transport_header;
	}
	ipv6_hdr(skb)->payload_len = htons(skb->len + ihl -
					   sizeof(struct ipv6hdr));
	skb_reset_transport_header(skb);
	return 0;
#else
	WARN_ON_ONCE(1);
	return -EAFNOSUPPORT;
#endif
}

static int xfrm_inner_mode_input(struct xfrm_state *x,
				 struct sk_buff *skb)
{
	switch (x->props.mode) {
	case XFRM_MODE_BEET:
	case XFRM_MODE_TUNNEL:
		return xfrm_prepare_input(x, skb);
	case XFRM_MODE_TRANSPORT:
		if (x->props.family == AF_INET)
			return xfrm4_transport_input(x, skb);
		if (x->props.family == AF_INET6)
			return xfrm6_transport_input(x, skb);
		break;
	case XFRM_MODE_ROUTEOPTIMIZATION:
		WARN_ON_ONCE(1);
		break;
	default:
		WARN_ON_ONCE(1);
		break;
	}

	return -EOPNOTSUPP;
}

int xfrm_input(struct sk_buff *skb, int nexthdr, __be32 spi, int encap_type)
{
	const struct xfrm_state_afinfo *afinfo;
	struct net *net = dev_net(skb->dev);
	int err;
	__be32 seq;
	__be32 seq_hi;
	struct xfrm_state *x = NULL;
	xfrm_address_t *daddr;
	u32 mark = skb->mark;
	unsigned int family = AF_UNSPEC;
	int decaps = 0;
	int async = 0;
	bool xfrm_gro = false;
	bool crypto_done = false;
	struct xfrm_offload *xo = xfrm_offload(skb);
	struct sec_path *sp;

	if (encap_type < 0 || (xo && (xo->flags & XFRM_GRO || encap_type == 0 ||
				      encap_type == UDP_ENCAP_ESPINUDP))) {
		x = xfrm_input_state(skb);

		if (unlikely(x->km.state != XFRM_STATE_VALID)) {
			if (x->km.state == XFRM_STATE_ACQ)
				XFRM_INC_STATS(net, LINUX_MIB_XFRMACQUIREERROR);
			else
				XFRM_INC_STATS(net,
					       LINUX_MIB_XFRMINSTATEINVALID);

			if (encap_type == -1)
				dev_put(skb->dev);
			goto drop;
		}

		family = x->props.family;

		/* An encap_type of -1 indicates async resumption. */
		if (encap_type == -1) {
			async = 1;
			seq = XFRM_SKB_CB(skb)->seq.input.low;
			goto resume;
		}
		/* GRO call */
		seq = XFRM_SPI_SKB_CB(skb)->seq;

		if (xo && (xo->flags & CRYPTO_DONE)) {
			crypto_done = true;
			family = XFRM_SPI_SKB_CB(skb)->family;

			if (!(xo->status & CRYPTO_SUCCESS)) {
				if (xo->status &
				    (CRYPTO_TRANSPORT_AH_AUTH_FAILED |
				     CRYPTO_TRANSPORT_ESP_AUTH_FAILED |
				     CRYPTO_TUNNEL_AH_AUTH_FAILED |
				     CRYPTO_TUNNEL_ESP_AUTH_FAILED)) {

					xfrm_audit_state_icvfail(x, skb,
								 x->type->proto);
					x->stats.integrity_failed++;
					XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEPROTOERROR);
					goto drop;
				}

				if (xo->status & CRYPTO_INVALID_PROTOCOL) {
					XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEPROTOERROR);
					goto drop;
				}

				XFRM_INC_STATS(net, LINUX_MIB_XFRMINBUFFERERROR);
				goto drop;
			}

			if (xfrm_parse_spi(skb, nexthdr, &spi, &seq)) {
				XFRM_INC_STATS(net, LINUX_MIB_XFRMINHDRERROR);
				goto drop;
			}
		}

		goto lock;
	}

	family = XFRM_SPI_SKB_CB(skb)->family;

	/* if tunnel is present override skb->mark value with tunnel i_key */
	switch (family) {
	case AF_INET:
		if (XFRM_TUNNEL_SKB_CB(skb)->tunnel.ip4)
			mark = be32_to_cpu(XFRM_TUNNEL_SKB_CB(skb)->tunnel.ip4->parms.i_key);
		break;
	case AF_INET6:
		if (XFRM_TUNNEL_SKB_CB(skb)->tunnel.ip6)
			mark = be32_to_cpu(XFRM_TUNNEL_SKB_CB(skb)->tunnel.ip6->parms.i_key);
		break;
	}

	sp = secpath_set(skb);
	if (!sp) {
		XFRM_INC_STATS(net, LINUX_MIB_XFRMINERROR);
		goto drop;
	}

	seq = 0;
	if (!spi && xfrm_parse_spi(skb, nexthdr, &spi, &seq)) {
		secpath_reset(skb);
		XFRM_INC_STATS(net, LINUX_MIB_XFRMINHDRERROR);
		goto drop;
	}

	daddr = (xfrm_address_t *)(skb_network_header(skb) +
				   XFRM_SPI_SKB_CB(skb)->daddroff);
	do {
		sp = skb_sec_path(skb);

		if (sp->len == XFRM_MAX_DEPTH) {
			secpath_reset(skb);
			XFRM_INC_STATS(net, LINUX_MIB_XFRMINBUFFERERROR);
			goto drop;
		}

		x = xfrm_state_lookup(net, mark, daddr, spi, nexthdr, family);
		if (x == NULL) {
			secpath_reset(skb);
			XFRM_INC_STATS(net, LINUX_MIB_XFRMINNOSTATES);
			xfrm_audit_state_notfound(skb, family, spi, seq);
			goto drop;
		}

		if (unlikely(x->dir && x->dir != XFRM_SA_DIR_IN)) {
			secpath_reset(skb);
			XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEDIRERROR);
			xfrm_audit_state_notfound(skb, family, spi, seq);
			xfrm_state_put(x);
			x = NULL;
			goto drop;
		}

		skb->mark = xfrm_smark_get(skb->mark, x);

		sp->xvec[sp->len++] = x;

		skb_dst_force(skb);
		if (!skb_dst(skb)) {
			XFRM_INC_STATS(net, LINUX_MIB_XFRMINERROR);
			goto drop;
		}

lock:
		spin_lock(&x->lock);

		if (unlikely(x->km.state != XFRM_STATE_VALID)) {
			if (x->km.state == XFRM_STATE_ACQ)
				XFRM_INC_STATS(net, LINUX_MIB_XFRMACQUIREERROR);
			else
				XFRM_INC_STATS(net,
					       LINUX_MIB_XFRMINSTATEINVALID);
			goto drop_unlock;
		}

		if ((x->encap ? x->encap->encap_type : 0) != encap_type) {
			XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEMISMATCH);
			goto drop_unlock;
		}

		if (xfrm_replay_check(x, skb, seq)) {
			XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATESEQERROR);
			goto drop_unlock;
		}

		if (xfrm_state_check_expire(x)) {
			XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEEXPIRED);
			goto drop_unlock;
		}

		spin_unlock(&x->lock);

		if (xfrm_tunnel_check(skb, x, family)) {
			XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEMODEERROR);
			goto drop;
		}

		seq_hi = htonl(xfrm_replay_seqhi(x, seq));

		XFRM_SKB_CB(skb)->seq.input.low = seq;
		XFRM_SKB_CB(skb)->seq.input.hi = seq_hi;

		dev_hold(skb->dev);

		if (crypto_done)
			nexthdr = x->type_offload->input_tail(x, skb);
		else
			nexthdr = x->type->input(x, skb);

		if (nexthdr == -EINPROGRESS)
			return 0;
resume:
		dev_put(skb->dev);

		spin_lock(&x->lock);
		if (nexthdr < 0) {
			if (nexthdr == -EBADMSG) {
				xfrm_audit_state_icvfail(x, skb,
							 x->type->proto);
				x->stats.integrity_failed++;
			}
			XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEPROTOERROR);
			goto drop_unlock;
		}

		/* only the first xfrm gets the encap type */
		encap_type = 0;

		if (xfrm_replay_recheck(x, skb, seq)) {
			XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATESEQERROR);
			goto drop_unlock;
		}

		xfrm_replay_advance(x, seq);

		x->curlft.bytes += skb->len;
		x->curlft.packets++;
		x->lastused = ktime_get_real_seconds();

		spin_unlock(&x->lock);

		XFRM_MODE_SKB_CB(skb)->protocol = nexthdr;

		if (xfrm_inner_mode_input(x, skb)) {
			XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEMODEERROR);
			goto drop;
		}

		if (x->outer_mode.flags & XFRM_MODE_FLAG_TUNNEL) {
			decaps = 1;
			break;
		}

		/*
		 * We need the inner address.  However, we only get here for
		 * transport mode so the outer address is identical.
		 */
		daddr = &x->id.daddr;
		family = x->props.family;

		err = xfrm_parse_spi(skb, nexthdr, &spi, &seq);
		if (err < 0) {
			XFRM_INC_STATS(net, LINUX_MIB_XFRMINHDRERROR);
			goto drop;
		}
		crypto_done = false;
	} while (!err);

	err = xfrm_rcv_cb(skb, family, x->type->proto, 0);
	if (err)
		goto drop;

	nf_reset_ct(skb);

	if (decaps) {
		sp = skb_sec_path(skb);
		if (sp)
			sp->olen = 0;
		if (skb_valid_dst(skb))
			skb_dst_drop(skb);
		gro_cells_receive(&gro_cells, skb);
		return 0;
	} else {
		xo = xfrm_offload(skb);
		if (xo)
			xfrm_gro = xo->flags & XFRM_GRO;

		err = -EAFNOSUPPORT;
		rcu_read_lock();
		afinfo = xfrm_state_afinfo_get_rcu(x->props.family);
		if (likely(afinfo))
			err = afinfo->transport_finish(skb, xfrm_gro || async);
		rcu_read_unlock();
		if (xfrm_gro) {
			sp = skb_sec_path(skb);
			if (sp)
				sp->olen = 0;
			if (skb_valid_dst(skb))
				skb_dst_drop(skb);
			gro_cells_receive(&gro_cells, skb);
			return err;
		}

		return err;
	}

drop_unlock:
	spin_unlock(&x->lock);
drop:
	xfrm_rcv_cb(skb, family, x && x->type ? x->type->proto : nexthdr, -1);
	kfree_skb(skb);
	return 0;
}
EXPORT_SYMBOL(xfrm_input);

int xfrm_input_resume(struct sk_buff *skb, int nexthdr)
{
	return xfrm_input(skb, nexthdr, 0, -1);
}
EXPORT_SYMBOL(xfrm_input_resume);

static void xfrm_trans_reinject(struct work_struct *work)
{
	struct xfrm_trans_tasklet *trans = container_of(work, struct xfrm_trans_tasklet, work);
	struct sk_buff_head queue;
	struct sk_buff *skb;

	__skb_queue_head_init(&queue);
	spin_lock_bh(&trans->queue_lock);
	skb_queue_splice_init(&trans->queue, &queue);
	spin_unlock_bh(&trans->queue_lock);

	local_bh_disable();
	while ((skb = __skb_dequeue(&queue)))
		XFRM_TRANS_SKB_CB(skb)->finish(XFRM_TRANS_SKB_CB(skb)->net,
					       NULL, skb);
	local_bh_enable();
}

int xfrm_trans_queue_net(struct net *net, struct sk_buff *skb,
			 int (*finish)(struct net *, struct sock *,
				       struct sk_buff *))
{
	struct xfrm_trans_tasklet *trans;

	trans = this_cpu_ptr(&xfrm_trans_tasklet);

	if (skb_queue_len(&trans->queue) >= READ_ONCE(net_hotdata.max_backlog))
		return -ENOBUFS;

	BUILD_BUG_ON(sizeof(struct xfrm_trans_cb) > sizeof(skb->cb));

	XFRM_TRANS_SKB_CB(skb)->finish = finish;
	XFRM_TRANS_SKB_CB(skb)->net = net;
	spin_lock_bh(&trans->queue_lock);
	__skb_queue_tail(&trans->queue, skb);
	spin_unlock_bh(&trans->queue_lock);
	schedule_work(&trans->work);
	return 0;
}
EXPORT_SYMBOL(xfrm_trans_queue_net);

int xfrm_trans_queue(struct sk_buff *skb,
		     int (*finish)(struct net *, struct sock *,
				   struct sk_buff *))
{
	return xfrm_trans_queue_net(dev_net(skb->dev), skb, finish);
}
EXPORT_SYMBOL(xfrm_trans_queue);

void __init xfrm_input_init(void)
{
	int err;
	int i;

	init_dummy_netdev(&xfrm_napi_dev);
	err = gro_cells_init(&gro_cells, &xfrm_napi_dev);
	if (err)
		gro_cells.cells = NULL;

	for_each_possible_cpu(i) {
		struct xfrm_trans_tasklet *trans;

		trans = &per_cpu(xfrm_trans_tasklet, i);
		spin_lock_init(&trans->queue_lock);
		__skb_queue_head_init(&trans->queue);
		INIT_WORK(&trans->work, xfrm_trans_reinject);
	}
}