Merge branch 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/klassert/ipsec-next

Steffen Klassert says:

====================
pull request (net-next): ipsec-next 2020-05-29

1) Add IPv6 encapsulation support for ESP over UDP and TCP.
   From Sabrina Dubroca.

2) Remove unneeded reference when initializing xfrm interfaces.
   From Nicolas Dichtel.

3) Remove some indirect calls from the state_afinfo.
   From Florian Westphal.

Please note that this pull request has two merge conflicts

between commit:

  0c922a48 ("xfrm: Always set XFRM_TRANSFORMED in xfrm{4,6}_output_finish")

  from Linus' tree and commit:

    2ab6096d ("xfrm: remove output_finish indirection from xfrm_state_afinfo")

    from the ipsec-next tree.

and between commit:

  3986912f ("ipv6: move SIOCADDRT and SIOCDELRT handling into ->compat_ioctl")

  from the net-next tree and commit:

    0146dca7 ("xfrm: add support for UDPv6 encapsulation of ESP")

    from the ipsec-next tree.

Both conflicts can be resolved as done in linux-next.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

include/net/ipv6_stubs.h

@@ -56,6 +56,12 @@ struct ipv6_stub {
 	void (*ndisc_send_na)(struct net_device *dev, const struct in6_addr *daddr,
 			      const struct in6_addr *solicited_addr,
 			      bool router, bool solicited, bool override, bool inc_opt);
+#if IS_ENABLED(CONFIG_XFRM)
+	void (*xfrm6_local_rxpmtu)(struct sk_buff *skb, u32 mtu);
+	int (*xfrm6_udp_encap_rcv)(struct sock *sk, struct sk_buff *skb);
+	int (*xfrm6_rcv_encap)(struct sk_buff *skb, int nexthdr, __be32 spi,
+			       int encap_type);
+#endif
 	struct neigh_table *nd_tbl;
 };
 extern const struct ipv6_stub *ipv6_stub __read_mostly;

include/net/xfrm.h

@@ -361,11 +361,6 @@ struct xfrm_state_afinfo {
 	const struct xfrm_type		*type_dstopts;
 
 	int			(*output)(struct net *net, struct sock *sk, struct sk_buff *skb);
-	int			(*output_finish)(struct sock *sk, struct sk_buff *skb);
-	int			(*extract_input)(struct xfrm_state *x,
-						 struct sk_buff *skb);
-	int			(*extract_output)(struct xfrm_state *x,
-						  struct sk_buff *skb);
 	int			(*transport_finish)(struct sk_buff *skb,
 						    int async);
 	void			(*local_error)(struct sk_buff *skb, u32 mtu);
@@ -1406,6 +1401,8 @@ struct xfrm4_protocol {
 
 struct xfrm6_protocol {
 	int (*handler)(struct sk_buff *skb);
+	int (*input_handler)(struct sk_buff *skb, int nexthdr, __be32 spi,
+			     int encap_type);
 	int (*cb_handler)(struct sk_buff *skb, int err);
 	int (*err_handler)(struct sk_buff *skb, struct inet6_skb_parm *opt,
 			   u8 type, u8 code, int offset, __be32 info);
@@ -1562,7 +1559,6 @@ int pktgen_xfrm_outer_mode_output(struct xfrm_state *x, struct sk_buff *skb);
 #endif
 
 void xfrm_local_error(struct sk_buff *skb, int mtu);
-int xfrm4_extract_header(struct sk_buff *skb);
 int xfrm4_extract_input(struct xfrm_state *x, struct sk_buff *skb);
 int xfrm4_rcv_encap(struct sk_buff *skb, int nexthdr, __be32 spi,
 		    int encap_type);
@@ -1578,7 +1574,6 @@ static inline int xfrm4_rcv_spi(struct sk_buff *skb, int nexthdr, __be32 spi)
 	return xfrm_input(skb, nexthdr, spi, 0);
 }
 
-int xfrm4_extract_output(struct xfrm_state *x, struct sk_buff *skb);
 int xfrm4_output(struct net *net, struct sock *sk, struct sk_buff *skb);
 int xfrm4_output_finish(struct sock *sk, struct sk_buff *skb);
 int xfrm4_protocol_register(struct xfrm4_protocol *handler, unsigned char protocol);
@@ -1586,10 +1581,11 @@ int xfrm4_protocol_deregister(struct xfrm4_protocol *handler, unsigned char prot
 int xfrm4_tunnel_register(struct xfrm_tunnel *handler, unsigned short family);
 int xfrm4_tunnel_deregister(struct xfrm_tunnel *handler, unsigned short family);
 void xfrm4_local_error(struct sk_buff *skb, u32 mtu);
-int xfrm6_extract_header(struct sk_buff *skb);
 int xfrm6_extract_input(struct xfrm_state *x, struct sk_buff *skb);
 int xfrm6_rcv_spi(struct sk_buff *skb, int nexthdr, __be32 spi,
 		  struct ip6_tnl *t);
+int xfrm6_rcv_encap(struct sk_buff *skb, int nexthdr, __be32 spi,
+		    int encap_type);
 int xfrm6_transport_finish(struct sk_buff *skb, int async);
 int xfrm6_rcv_tnl(struct sk_buff *skb, struct ip6_tnl *t);
 int xfrm6_rcv(struct sk_buff *skb);
@@ -1602,14 +1598,15 @@ int xfrm6_tunnel_register(struct xfrm6_tunnel *handler, unsigned short family);
 int xfrm6_tunnel_deregister(struct xfrm6_tunnel *handler, unsigned short family);
 __be32 xfrm6_tunnel_alloc_spi(struct net *net, xfrm_address_t *saddr);
 __be32 xfrm6_tunnel_spi_lookup(struct net *net, const xfrm_address_t *saddr);
-int xfrm6_extract_output(struct xfrm_state *x, struct sk_buff *skb);
 int xfrm6_output(struct net *net, struct sock *sk, struct sk_buff *skb);
 int xfrm6_output_finish(struct sock *sk, struct sk_buff *skb);
 int xfrm6_find_1stfragopt(struct xfrm_state *x, struct sk_buff *skb,
 			  u8 **prevhdr);
 
 #ifdef CONFIG_XFRM
+void xfrm6_local_rxpmtu(struct sk_buff *skb, u32 mtu);
 int xfrm4_udp_encap_rcv(struct sock *sk, struct sk_buff *skb);
+int xfrm6_udp_encap_rcv(struct sock *sk, struct sk_buff *skb);
 int xfrm_user_policy(struct sock *sk, int optname,
 		     u8 __user *optval, int optlen);
 #else
@@ -1992,4 +1989,20 @@ static inline int xfrm_tunnel_check(struct sk_buff *skb, struct xfrm_state *x,
 
 	return 0;
 }
+
+#if IS_ENABLED(CONFIG_IPV6)
+static inline bool xfrm6_local_dontfrag(const struct sock *sk)
+{
+	int proto;
+
+	if (!sk || sk->sk_family != AF_INET6)
+		return false;
+
+	proto = sk->sk_protocol;
+	if (proto == IPPROTO_UDP || proto == IPPROTO_RAW)
+		return inet6_sk(sk)->dontfrag;
+
+	return false;
+}
+#endif
 #endif	/* _NET_XFRM_H */

net/ipv4/Kconfig

@@ -384,6 +384,7 @@ config INET_ESPINTCP
 	depends on XFRM && INET_ESP
 	select STREAM_PARSER
 	select NET_SOCK_MSG
+	select XFRM_ESPINTCP
 	help
 	  Support for RFC 8229 encapsulation of ESP and IKE over
 	  TCP/IPv4 sockets.

net/ipv4/udp.c

@@ -112,6 +112,9 @@
 #include <net/sock_reuseport.h>
 #include <net/addrconf.h>
 #include <net/udp_tunnel.h>
+#if IS_ENABLED(CONFIG_IPV6)
+#include <net/ipv6_stubs.h>
+#endif
 
 struct udp_table udp_table __read_mostly;
 EXPORT_SYMBOL(udp_table);
@@ -2563,6 +2566,11 @@ int udp_lib_setsockopt(struct sock *sk, int level, int optname,
 #ifdef CONFIG_XFRM
 		case UDP_ENCAP_ESPINUDP:
 		case UDP_ENCAP_ESPINUDP_NON_IKE:
+#if IS_ENABLED(CONFIG_IPV6)
+			if (sk->sk_family == AF_INET6)
+				up->encap_rcv = ipv6_stub->xfrm6_udp_encap_rcv;
+			else
+#endif
 				up->encap_rcv = xfrm4_udp_encap_rcv;
 #endif
 			fallthrough;

net/ipv4/xfrm4_input.c

@@ -18,11 +18,6 @@
 #include <net/ip.h>
 #include <net/xfrm.h>
 
-int xfrm4_extract_input(struct xfrm_state *x, struct sk_buff *skb)
-{
-	return xfrm4_extract_header(skb);
-}
-
 static int xfrm4_rcv_encap_finish2(struct net *net, struct sock *sk,
 				   struct sk_buff *skb)
 {

net/ipv4/xfrm4_output.c

@@ -14,77 +14,18 @@
 #include <net/xfrm.h>
 #include <net/icmp.h>
 
-static int xfrm4_tunnel_check_size(struct sk_buff *skb)
-{
-	int mtu, ret = 0;
-
-	if (IPCB(skb)->flags & IPSKB_XFRM_TUNNEL_SIZE)
-		goto out;
-
-	if (!(ip_hdr(skb)->frag_off & htons(IP_DF)) || skb->ignore_df)
-		goto out;
-
-	mtu = dst_mtu(skb_dst(skb));
-	if ((!skb_is_gso(skb) && skb->len > mtu) ||
-	    (skb_is_gso(skb) &&
-	     !skb_gso_validate_network_len(skb, ip_skb_dst_mtu(skb->sk, skb)))) {
-		skb->protocol = htons(ETH_P_IP);
-
-		if (skb->sk)
-			xfrm_local_error(skb, mtu);
-		else
-			icmp_send(skb, ICMP_DEST_UNREACH,
-				  ICMP_FRAG_NEEDED, htonl(mtu));
-		ret = -EMSGSIZE;
-	}
-out:
-	return ret;
-}
-
-int xfrm4_extract_output(struct xfrm_state *x, struct sk_buff *skb)
-{
-	int err;
-
-	err = xfrm4_tunnel_check_size(skb);
-	if (err)
-		return err;
-
-	XFRM_MODE_SKB_CB(skb)->protocol = ip_hdr(skb)->protocol;
-
-	return xfrm4_extract_header(skb);
-}
-
-int xfrm4_output_finish(struct sock *sk, struct sk_buff *skb)
-{
-	memset(IPCB(skb), 0, sizeof(*IPCB(skb)));
-
-	IPCB(skb)->flags |= IPSKB_XFRM_TRANSFORMED;
-
-	return xfrm_output(sk, skb);
-}
-
 static int __xfrm4_output(struct net *net, struct sock *sk, struct sk_buff *skb)
 {
+#ifdef CONFIG_NETFILTER
 	struct xfrm_state *x = skb_dst(skb)->xfrm;
-	const struct xfrm_state_afinfo *afinfo;
-	int ret = -EAFNOSUPPORT;
 
-#ifdef CONFIG_NETFILTER
 	if (!x) {
 		IPCB(skb)->flags |= IPSKB_REROUTED;
 		return dst_output(net, sk, skb);
 	}
 #endif
 
-	rcu_read_lock();
-	afinfo = xfrm_state_afinfo_get_rcu(x->outer_mode.family);
-	if (likely(afinfo))
-		ret = afinfo->output_finish(sk, skb);
-	else
-		kfree_skb(skb);
-	rcu_read_unlock();
-
-	return ret;
+	return xfrm_output(sk, skb);
 }
 
 int xfrm4_output(struct net *net, struct sock *sk, struct sk_buff *skb)

net/ipv4/xfrm4_state.c

@@ -8,36 +8,12 @@
  *
  */
 
-#include <net/ip.h>
 #include <net/xfrm.h>
-#include <linux/pfkeyv2.h>
-#include <linux/ipsec.h>
-#include <linux/netfilter_ipv4.h>
-#include <linux/export.h>
-
-int xfrm4_extract_header(struct sk_buff *skb)
-{
-	const struct iphdr *iph = ip_hdr(skb);
-
-	XFRM_MODE_SKB_CB(skb)->ihl = sizeof(*iph);
-	XFRM_MODE_SKB_CB(skb)->id = iph->id;
-	XFRM_MODE_SKB_CB(skb)->frag_off = iph->frag_off;
-	XFRM_MODE_SKB_CB(skb)->tos = iph->tos;
-	XFRM_MODE_SKB_CB(skb)->ttl = iph->ttl;
-	XFRM_MODE_SKB_CB(skb)->optlen = iph->ihl * 4 - sizeof(*iph);
-	memset(XFRM_MODE_SKB_CB(skb)->flow_lbl, 0,
-	       sizeof(XFRM_MODE_SKB_CB(skb)->flow_lbl));
-
-	return 0;
-}
 
 static struct xfrm_state_afinfo xfrm4_state_afinfo = {
 	.family			= AF_INET,
 	.proto			= IPPROTO_IPIP,
 	.output			= xfrm4_output,
-	.output_finish		= xfrm4_output_finish,
-	.extract_input		= xfrm4_extract_input,
-	.extract_output		= xfrm4_extract_output,
 	.transport_finish	= xfrm4_transport_finish,
 	.local_error		= xfrm4_local_error,
 };

net/ipv6/Kconfig

@@ -88,6 +88,18 @@ config INET6_ESP_OFFLOAD
 
 	  If unsure, say N.
 
+config INET6_ESPINTCP
+	bool "IPv6: ESP in TCP encapsulation (RFC 8229)"
+	depends on XFRM && INET6_ESP
+	select STREAM_PARSER
+	select NET_SOCK_MSG
+	select XFRM_ESPINTCP
+	help
+	  Support for RFC 8229 encapsulation of ESP and IKE over
+	  TCP/IPv6 sockets.
+
+	  If unsure, say N.
+
 config INET6_IPCOMP
 	tristate "IPv6: IPComp transformation"
 	select INET6_XFRM_TUNNEL

net/ipv6/af_inet6.c

@@ -61,6 +61,7 @@
 #include <net/seg6.h>
 #include <net/rpl.h>
 #include <net/compat.h>
+#include <net/xfrm.h>
 
 #include <linux/uaccess.h>
 #include <linux/mroute6.h>
@@ -1023,6 +1024,11 @@ static const struct ipv6_stub ipv6_stub_impl = {
 	.ip6_del_rt	   = ip6_del_rt,
 	.udpv6_encap_enable = udpv6_encap_enable,
 	.ndisc_send_na = ndisc_send_na,
+#if IS_ENABLED(CONFIG_XFRM)
+	.xfrm6_local_rxpmtu = xfrm6_local_rxpmtu,
+	.xfrm6_udp_encap_rcv = xfrm6_udp_encap_rcv,
+	.xfrm6_rcv_encap = xfrm6_rcv_encap,
+#endif
 	.nd_tbl	= &nd_tbl,
 };
 

net/ipv6/ah6.c

@@ -767,6 +767,7 @@ static const struct xfrm_type ah6_type = {
 
 static struct xfrm6_protocol ah6_protocol = {
 	.handler	=	xfrm6_rcv,
+	.input_handler	=	xfrm_input,
 	.cb_handler	=	ah6_rcv_cb,
 	.err_handler	=	ah6_err,
 	.priority	=	0,

net/ipv6/esp6.c

@@ -26,11 +26,16 @@
 #include <linux/random.h>
 #include <linux/slab.h>
 #include <linux/spinlock.h>
+#include <net/ip6_checksum.h>
 #include <net/ip6_route.h>
 #include <net/icmp.h>
 #include <net/ipv6.h>
 #include <net/protocol.h>
+#include <net/udp.h>
 #include <linux/icmpv6.h>
+#include <net/tcp.h>
+#include <net/espintcp.h>
+#include <net/inet6_hashtables.h>
 
 #include <linux/highmem.h>
 
@@ -39,6 +44,11 @@ struct esp_skb_cb {
 	void *tmp;
 };
 
+struct esp_output_extra {
+	__be32 seqhi;
+	u32 esphoff;
+};
+
 #define ESP_SKB_CB(__skb) ((struct esp_skb_cb *)&((__skb)->cb[0]))
 
 /*
@@ -72,9 +82,9 @@ static void *esp_alloc_tmp(struct crypto_aead *aead, int nfrags, int seqihlen)
 	return kmalloc(len, GFP_ATOMIC);
 }
 
-static inline __be32 *esp_tmp_seqhi(void *tmp)
+static inline void *esp_tmp_extra(void *tmp)
 {
-	return PTR_ALIGN((__be32 *)tmp, __alignof__(__be32));
+	return PTR_ALIGN(tmp, __alignof__(struct esp_output_extra));
 }
 
 static inline u8 *esp_tmp_iv(struct crypto_aead *aead, void *tmp, int seqhilen)
@@ -104,16 +114,17 @@ static inline struct scatterlist *esp_req_sg(struct crypto_aead *aead,
 
 static void esp_ssg_unref(struct xfrm_state *x, void *tmp)
 {
+	struct esp_output_extra *extra = esp_tmp_extra(tmp);
 	struct crypto_aead *aead = x->data;
-	int seqhilen = 0;
+	int extralen = 0;
 	u8 *iv;
 	struct aead_request *req;
 	struct scatterlist *sg;
 
 	if (x->props.flags & XFRM_STATE_ESN)
-		seqhilen += sizeof(__be32);
+		extralen += sizeof(*extra);
 
-	iv = esp_tmp_iv(aead, tmp, seqhilen);
+	iv = esp_tmp_iv(aead, tmp, extralen);
 	req = esp_tmp_req(aead, iv);
 
 	/* Unref skb_frag_pages in the src scatterlist if necessary.
@@ -124,6 +135,149 @@ static void esp_ssg_unref(struct xfrm_state *x, void *tmp)
 			put_page(sg_page(sg));
 }
 
+#ifdef CONFIG_INET6_ESPINTCP
+struct esp_tcp_sk {
+	struct sock *sk;
+	struct rcu_head rcu;
+};
+
+static void esp_free_tcp_sk(struct rcu_head *head)
+{
+	struct esp_tcp_sk *esk = container_of(head, struct esp_tcp_sk, rcu);
+
+	sock_put(esk->sk);
+	kfree(esk);
+}
+
+static struct sock *esp6_find_tcp_sk(struct xfrm_state *x)
+{
+	struct xfrm_encap_tmpl *encap = x->encap;
+	struct esp_tcp_sk *esk;
+	__be16 sport, dport;
+	struct sock *nsk;
+	struct sock *sk;
+
+	sk = rcu_dereference(x->encap_sk);
+	if (sk && sk->sk_state == TCP_ESTABLISHED)
+		return sk;
+
+	spin_lock_bh(&x->lock);
+	sport = encap->encap_sport;
+	dport = encap->encap_dport;
+	nsk = rcu_dereference_protected(x->encap_sk,
+					lockdep_is_held(&x->lock));
+	if (sk && sk == nsk) {
+		esk = kmalloc(sizeof(*esk), GFP_ATOMIC);
+		if (!esk) {
+			spin_unlock_bh(&x->lock);
+			return ERR_PTR(-ENOMEM);
+		}
+		RCU_INIT_POINTER(x->encap_sk, NULL);
+		esk->sk = sk;
+		call_rcu(&esk->rcu, esp_free_tcp_sk);
+	}
+	spin_unlock_bh(&x->lock);
+
+	sk = __inet6_lookup_established(xs_net(x), &tcp_hashinfo, &x->id.daddr.in6,
+					dport, &x->props.saddr.in6, ntohs(sport), 0, 0);
+	if (!sk)
+		return ERR_PTR(-ENOENT);
+
+	if (!tcp_is_ulp_esp(sk)) {
+		sock_put(sk);
+		return ERR_PTR(-EINVAL);
+	}
+
+	spin_lock_bh(&x->lock);
+	nsk = rcu_dereference_protected(x->encap_sk,
+					lockdep_is_held(&x->lock));
+	if (encap->encap_sport != sport ||
+	    encap->encap_dport != dport) {
+		sock_put(sk);
+		sk = nsk ?: ERR_PTR(-EREMCHG);
+	} else if (sk == nsk) {
+		sock_put(sk);
+	} else {
+		rcu_assign_pointer(x->encap_sk, sk);
+	}
+	spin_unlock_bh(&x->lock);
+
+	return sk;
+}
+
+static int esp_output_tcp_finish(struct xfrm_state *x, struct sk_buff *skb)
+{
+	struct sock *sk;
+	int err;
+
+	rcu_read_lock();
+
+	sk = esp6_find_tcp_sk(x);
+	err = PTR_ERR_OR_ZERO(sk);
+	if (err)
+		goto out;
+
+	bh_lock_sock(sk);
+	if (sock_owned_by_user(sk))
+		err = espintcp_queue_out(sk, skb);
+	else
+		err = espintcp_push_skb(sk, skb);
+	bh_unlock_sock(sk);
+
+out:
+	rcu_read_unlock();
+	return err;
+}
+
+static int esp_output_tcp_encap_cb(struct net *net, struct sock *sk,
+				   struct sk_buff *skb)
+{
+	struct dst_entry *dst = skb_dst(skb);
+	struct xfrm_state *x = dst->xfrm;
+
+	return esp_output_tcp_finish(x, skb);
+}
+
+static int esp_output_tail_tcp(struct xfrm_state *x, struct sk_buff *skb)
+{
+	int err;
+
+	local_bh_disable();
+	err = xfrm_trans_queue_net(xs_net(x), skb, esp_output_tcp_encap_cb);
+	local_bh_enable();
+
+	/* EINPROGRESS just happens to do the right thing.  It
+	 * actually means that the skb has been consumed and
+	 * isn't coming back.
+	 */
+	return err ?: -EINPROGRESS;
+}
+#else
+static int esp_output_tail_tcp(struct xfrm_state *x, struct sk_buff *skb)
+{
+	kfree_skb(skb);
+
+	return -EOPNOTSUPP;
+}
+#endif
+
+static void esp_output_encap_csum(struct sk_buff *skb)
+{
+	/* UDP encap with IPv6 requires a valid checksum */
+	if (*skb_mac_header(skb) == IPPROTO_UDP) {
+		struct udphdr *uh = udp_hdr(skb);
+		struct ipv6hdr *ip6h = ipv6_hdr(skb);
+		int len = ntohs(uh->len);
+		unsigned int offset = skb_transport_offset(skb);
+		__wsum csum = skb_checksum(skb, offset, skb->len - offset, 0);
+
+		uh->check = csum_ipv6_magic(&ip6h->saddr, &ip6h->daddr,
+					    len, IPPROTO_UDP, csum);
+		if (uh->check == 0)
+			uh->check = CSUM_MANGLED_0;
+	}
+}
+
 static void esp_output_done(struct crypto_async_request *base, int err)
 {
 	struct sk_buff *skb = base->data;
@@ -143,6 +297,8 @@ static void esp_output_done(struct crypto_async_request *base, int err)
 	esp_ssg_unref(x, tmp);
 	kfree(tmp);
 
+	esp_output_encap_csum(skb);
+
 	if (xo && (xo->flags & XFRM_DEV_RESUME)) {
 		if (err) {
 			XFRM_INC_STATS(xs_net(x), LINUX_MIB_XFRMOUTSTATEPROTOERROR);
@@ -154,6 +310,10 @@ static void esp_output_done(struct crypto_async_request *base, int err)
 		secpath_reset(skb);
 		xfrm_dev_resume(skb);
 	} else {
+		if (!err &&
+		    x->encap && x->encap->encap_type == TCP_ENCAP_ESPINTCP)
+			esp_output_tail_tcp(x, skb);
+		else
 			xfrm_output_resume(skb, err);
 	}
 }
@@ -163,7 +323,7 @@ static void esp_restore_header(struct sk_buff *skb, unsigned int offset)
 {
 	struct ip_esp_hdr *esph = (void *)(skb->data + offset);
 	void *tmp = ESP_SKB_CB(skb)->tmp;
-	__be32 *seqhi = esp_tmp_seqhi(tmp);
+	__be32 *seqhi = esp_tmp_extra(tmp);
 
 	esph->seq_no = esph->spi;
 	esph->spi = *seqhi;
@@ -171,27 +331,36 @@ static void esp_restore_header(struct sk_buff *skb, unsigned int offset)
 
 static void esp_output_restore_header(struct sk_buff *skb)
 {
-	esp_restore_header(skb, skb_transport_offset(skb) - sizeof(__be32));
+	void *tmp = ESP_SKB_CB(skb)->tmp;
+	struct esp_output_extra *extra = esp_tmp_extra(tmp);
+
+	esp_restore_header(skb, skb_transport_offset(skb) + extra->esphoff -
+				sizeof(__be32));
 }
 
 static struct ip_esp_hdr *esp_output_set_esn(struct sk_buff *skb,
 					     struct xfrm_state *x,
 					     struct ip_esp_hdr *esph,
-					     __be32 *seqhi)
+					     struct esp_output_extra *extra)
 {
 	/* For ESN we move the header forward by 4 bytes to
 	 * accomodate the high bits.  We will move it back after
 	 * encryption.
 	 */
 	if ((x->props.flags & XFRM_STATE_ESN)) {
+		__u32 seqhi;
 		struct xfrm_offload *xo = xfrm_offload(skb);
 
-		esph = (void *)(skb_transport_header(skb) - sizeof(__be32));
-		*seqhi = esph->spi;
 		if (xo)
-			esph->seq_no = htonl(xo->seq.hi);
+			seqhi = xo->seq.hi;
 		else
-			esph->seq_no = htonl(XFRM_SKB_CB(skb)->seq.output.hi);
+			seqhi = XFRM_SKB_CB(skb)->seq.output.hi;
+
+		extra->esphoff = (unsigned char *)esph -
+				 skb_transport_header(skb);
+		esph = (struct ip_esp_hdr *)((unsigned char *)esph - 4);
+		extra->seqhi = esph->spi;
+		esph->seq_no = htonl(seqhi);
 	}
 
 	esph->spi = x->id.spi;
@@ -207,15 +376,122 @@ static void esp_output_done_esn(struct crypto_async_request *base, int err)
 	esp_output_done(base, err);
 }
 
+static struct ip_esp_hdr *esp6_output_udp_encap(struct sk_buff *skb,
+					       int encap_type,
+					       struct esp_info *esp,
+					       __be16 sport,
+					       __be16 dport)
+{
+	struct udphdr *uh;
+	__be32 *udpdata32;
+	unsigned int len;
+
+	len = skb->len + esp->tailen - skb_transport_offset(skb);
+	if (len > U16_MAX)
+		return ERR_PTR(-EMSGSIZE);
+
+	uh = (struct udphdr *)esp->esph;
+	uh->source = sport;
+	uh->dest = dport;
+	uh->len = htons(len);
+	uh->check = 0;
+
+	*skb_mac_header(skb) = IPPROTO_UDP;
+
+	if (encap_type == UDP_ENCAP_ESPINUDP_NON_IKE) {
+		udpdata32 = (__be32 *)(uh + 1);
+		udpdata32[0] = udpdata32[1] = 0;
+		return (struct ip_esp_hdr *)(udpdata32 + 2);
+	}
+
+	return (struct ip_esp_hdr *)(uh + 1);
+}
+
+#ifdef CONFIG_INET6_ESPINTCP
+static struct ip_esp_hdr *esp6_output_tcp_encap(struct xfrm_state *x,
+						struct sk_buff *skb,
+						struct esp_info *esp)
+{
+	__be16 *lenp = (void *)esp->esph;
+	struct ip_esp_hdr *esph;
+	unsigned int len;
+	struct sock *sk;
+
+	len = skb->len + esp->tailen - skb_transport_offset(skb);
+	if (len > IP_MAX_MTU)
+		return ERR_PTR(-EMSGSIZE);
+
+	rcu_read_lock();
+	sk = esp6_find_tcp_sk(x);
+	rcu_read_unlock();
+
+	if (IS_ERR(sk))
+		return ERR_CAST(sk);
+
+	*lenp = htons(len);
+	esph = (struct ip_esp_hdr *)(lenp + 1);
+
+	return esph;
+}
+#else
+static struct ip_esp_hdr *esp6_output_tcp_encap(struct xfrm_state *x,
+						struct sk_buff *skb,
+						struct esp_info *esp)
+{
+	return ERR_PTR(-EOPNOTSUPP);
+}
+#endif
+
+static int esp6_output_encap(struct xfrm_state *x, struct sk_buff *skb,
+			    struct esp_info *esp)
+{
+	struct xfrm_encap_tmpl *encap = x->encap;
+	struct ip_esp_hdr *esph;
+	__be16 sport, dport;
+	int encap_type;
+
+	spin_lock_bh(&x->lock);
+	sport = encap->encap_sport;
+	dport = encap->encap_dport;
+	encap_type = encap->encap_type;
+	spin_unlock_bh(&x->lock);
+
+	switch (encap_type) {
+	default:
+	case UDP_ENCAP_ESPINUDP:
+	case UDP_ENCAP_ESPINUDP_NON_IKE:
+		esph = esp6_output_udp_encap(skb, encap_type, esp, sport, dport);
+		break;
+	case TCP_ENCAP_ESPINTCP:
+		esph = esp6_output_tcp_encap(x, skb, esp);
+		break;
+	}
+
+	if (IS_ERR(esph))
+		return PTR_ERR(esph);
+
+	esp->esph = esph;
+
+	return 0;
+}
+
 int esp6_output_head(struct xfrm_state *x, struct sk_buff *skb, struct esp_info *esp)
 {
 	u8 *tail;
 	u8 *vaddr;
 	int nfrags;
+	int esph_offset;
 	struct page *page;
 	struct sk_buff *trailer;
 	int tailen = esp->tailen;
 
+	if (x->encap) {
+		int err = esp6_output_encap(x, skb, esp);
+
+		if (err < 0)
+			return err;
+	}
+
 	if (!skb_cloned(skb)) {
 		if (tailen <= skb_tailroom(skb)) {
 			nfrags = 1;
@@ -274,10 +550,13 @@ int esp6_output_head(struct xfrm_state *x, struct sk_buff *skb, struct esp_info
 	}
 
 cow:
+	esph_offset = (unsigned char *)esp->esph - skb_transport_header(skb);
+
 	nfrags = skb_cow_data(skb, tailen, &trailer);
 	if (nfrags < 0)
 		goto out;
 	tail = skb_tail_pointer(trailer);
+	esp->esph = (struct ip_esp_hdr *)(skb_transport_header(skb) + esph_offset);
 
 skip_cow:
 	esp_output_fill_trailer(tail, esp->tfclen, esp->plen, esp->proto);
@@ -295,20 +574,20 @@ int esp6_output_tail(struct xfrm_state *x, struct sk_buff *skb, struct esp_info
 	void *tmp;
 	int ivlen;
 	int assoclen;
-	int seqhilen;
-	__be32 *seqhi;
+	int extralen;
 	struct page *page;
 	struct ip_esp_hdr *esph;
 	struct aead_request *req;
 	struct crypto_aead *aead;
 	struct scatterlist *sg, *dsg;
+	struct esp_output_extra *extra;
 	int err = -ENOMEM;
 
 	assoclen = sizeof(struct ip_esp_hdr);
-	seqhilen = 0;
+	extralen = 0;
 
 	if (x->props.flags & XFRM_STATE_ESN) {
-		seqhilen += sizeof(__be32);
+		extralen += sizeof(*extra);
 		assoclen += sizeof(__be32);
 	}
 
@@ -316,12 +595,12 @@ int esp6_output_tail(struct xfrm_state *x, struct sk_buff *skb, struct esp_info
 	alen = crypto_aead_authsize(aead);
 	ivlen = crypto_aead_ivsize(aead);
 
-	tmp = esp_alloc_tmp(aead, esp->nfrags + 2, seqhilen);
+	tmp = esp_alloc_tmp(aead, esp->nfrags + 2, extralen);
 	if (!tmp)
 		goto error;
 
-	seqhi = esp_tmp_seqhi(tmp);
-	iv = esp_tmp_iv(aead, tmp, seqhilen);
+	extra = esp_tmp_extra(tmp);
+	iv = esp_tmp_iv(aead, tmp, extralen);
 	req = esp_tmp_req(aead, iv);
 	sg = esp_req_sg(aead, req);
 
@@ -330,7 +609,8 @@ int esp6_output_tail(struct xfrm_state *x, struct sk_buff *skb, struct esp_info
 	else
 		dsg = &sg[esp->nfrags];
 
-	esph = esp_output_set_esn(skb, x, ip_esp_hdr(skb), seqhi);
+	esph = esp_output_set_esn(skb, x, esp->esph, extra);
+	esp->esph = esph;
 
 	sg_init_table(sg, esp->nfrags);
 	err = skb_to_sgvec(skb, sg,
@@ -394,11 +674,15 @@ int esp6_output_tail(struct xfrm_state *x, struct sk_buff *skb, struct esp_info
 	case 0:
 		if ((x->props.flags & XFRM_STATE_ESN))
 			esp_output_restore_header(skb);
+		esp_output_encap_csum(skb);
 	}
 
 	if (sg != dsg)
 		esp_ssg_unref(x, tmp);
 
+	if (!err && x->encap && x->encap->encap_type == TCP_ENCAP_ESPINTCP)
+		err = esp_output_tail_tcp(x, skb);
+
 error_free:
 	kfree(tmp);
 error:
@@ -438,11 +722,13 @@ static int esp6_output(struct xfrm_state *x, struct sk_buff *skb)
 	esp.plen = esp.clen - skb->len - esp.tfclen;
 	esp.tailen = esp.tfclen + esp.plen + alen;
 
+	esp.esph = ip_esp_hdr(skb);
+
 	esp.nfrags = esp6_output_head(x, skb, &esp);
 	if (esp.nfrags < 0)
 		return esp.nfrags;
 
-	esph = ip_esp_hdr(skb);
+	esph = esp.esph;
 	esph->spi = x->id.spi;
 
 	esph->seq_no = htonl(XFRM_SKB_CB(skb)->seq.output.low);
@@ -517,6 +803,60 @@ int esp6_input_done2(struct sk_buff *skb, int err)
 	if (unlikely(err < 0))
 		goto out;
 
+	if (x->encap) {
+		const struct ipv6hdr *ip6h = ipv6_hdr(skb);
+		struct xfrm_encap_tmpl *encap = x->encap;
+		struct udphdr *uh = (void *)(skb_network_header(skb) + hdr_len);
+		struct tcphdr *th = (void *)(skb_network_header(skb) + hdr_len);
+		__be16 source;
+
+		switch (x->encap->encap_type) {
+		case TCP_ENCAP_ESPINTCP:
+			source = th->source;
+			break;
+		case UDP_ENCAP_ESPINUDP:
+		case UDP_ENCAP_ESPINUDP_NON_IKE:
+			source = uh->source;
+			break;
+		default:
+			WARN_ON_ONCE(1);
+			err = -EINVAL;
+			goto out;
+		}
+
+		/*
+		 * 1) if the NAT-T peer's IP or port changed then
+		 *    advertize the change to the keying daemon.
+		 *    This is an inbound SA, so just compare
+		 *    SRC ports.
+		 */
+		if (!ipv6_addr_equal(&ip6h->saddr, &x->props.saddr.in6) ||
+		    source != encap->encap_sport) {
+			xfrm_address_t ipaddr;
+
+			memcpy(&ipaddr.a6, &ip6h->saddr.s6_addr, sizeof(ipaddr.a6));
+			km_new_mapping(x, &ipaddr, source);
+
+			/* XXX: perhaps add an extra
+			 * policy check here, to see
+			 * if we should allow or
+			 * reject a packet from a
+			 * different source
+			 * address/port.
+			 */
+		}
+
+		/*
+		 * 2) ignore UDP/TCP checksums in case
+		 *    of NAT-T in Transport Mode, or
+		 *    perform other post-processing fixes
+		 *    as per draft-ietf-ipsec-udp-encaps-06,
+		 *    section 3.1.2
+		 */
+		if (x->props.mode == XFRM_MODE_TRANSPORT)
+			skb->ip_summed = CHECKSUM_UNNECESSARY;
+	}
+
 	skb_postpull_rcsum(skb, skb_network_header(skb),
 			   skb_network_header_len(skb));
 	skb_pull_rcsum(skb, hlen);
@@ -632,7 +972,7 @@ static int esp6_input(struct xfrm_state *x, struct sk_buff *skb)
 		goto out;
 
 	ESP_SKB_CB(skb)->tmp = tmp;
-	seqhi = esp_tmp_seqhi(tmp);
+	seqhi = esp_tmp_extra(tmp);
 	iv = esp_tmp_iv(aead, tmp, seqhilen);
 	req = esp_tmp_req(aead, iv);
 	sg = esp_req_sg(aead, req);
@@ -836,9 +1176,6 @@ static int esp6_init_state(struct xfrm_state *x)
 	u32 align;
 	int err;
 
-	if (x->encap)
-		return -EINVAL;
-
 	x->data = NULL;
 
 	if (x->aead)
@@ -867,6 +1204,30 @@ static int esp6_init_state(struct xfrm_state *x)
 		break;
 	}
 
+	if (x->encap) {
+		struct xfrm_encap_tmpl *encap = x->encap;
+
+		switch (encap->encap_type) {
+		default:
+			err = -EINVAL;
+			goto error;
+		case UDP_ENCAP_ESPINUDP:
+			x->props.header_len += sizeof(struct udphdr);
+			break;
+		case UDP_ENCAP_ESPINUDP_NON_IKE:
+			x->props.header_len += sizeof(struct udphdr) + 2 * sizeof(u32);
+			break;
+#ifdef CONFIG_INET6_ESPINTCP
+		case TCP_ENCAP_ESPINTCP:
+			/* only the length field, TCP encap is done by
+			 * the socket
+			 */
+			x->props.header_len += 2;
+			break;
+#endif
+		}
+	}
+
 	align = ALIGN(crypto_aead_blocksize(aead), 4);
 	x->props.trailer_len = align + 1 + crypto_aead_authsize(aead);
 
@@ -893,6 +1254,7 @@ static const struct xfrm_type esp6_type = {
 
 static struct xfrm6_protocol esp6_protocol = {
 	.handler	=	xfrm6_rcv,
+	.input_handler	=	xfrm_input,
 	.cb_handler	=	esp6_rcv_cb,
 	.err_handler	=	esp6_err,
 	.priority	=	0,

net/ipv6/esp6_offload.c

@@ -271,7 +271,6 @@ static int esp6_xmit(struct xfrm_state *x, struct sk_buff *skb,  netdev_features
 	int alen;
 	int blksize;
 	struct xfrm_offload *xo;
-	struct ip_esp_hdr *esph;
 	struct crypto_aead *aead;
 	struct esp_info esp;
 	bool hw_offload = true;
@@ -312,13 +311,13 @@ static int esp6_xmit(struct xfrm_state *x, struct sk_buff *skb,  netdev_features
 
 	seq = xo->seq.low;
 
-	esph = ip_esp_hdr(skb);
-	esph->spi = x->id.spi;
+	esp.esph = ip_esp_hdr(skb);
+	esp.esph->spi = x->id.spi;
 
 	skb_push(skb, -skb_network_offset(skb));
 
 	if (xo->flags & XFRM_GSO_SEGMENT) {
-		esph->seq_no = htonl(seq);
+		esp.esph->seq_no = htonl(seq);
 
 		if (!skb_is_gso(skb))
 			xo->seq.low++;

net/ipv6/ip6_vti.c

@@ -296,7 +296,8 @@ static void vti6_dev_uninit(struct net_device *dev)
 	dev_put(dev);
 }
 
-static int vti6_rcv(struct sk_buff *skb)
+static int vti6_input_proto(struct sk_buff *skb, int nexthdr, __be32 spi,
+			    int encap_type)
 {
 	struct ip6_tnl *t;
 	const struct ipv6hdr *ipv6h = ipv6_hdr(skb);
@@ -323,7 +324,10 @@ static int vti6_rcv(struct sk_buff *skb)
 
 		rcu_read_unlock();
 
-		return xfrm6_rcv_tnl(skb, t);
+		XFRM_TUNNEL_SKB_CB(skb)->tunnel.ip6 = t;
+		XFRM_SPI_SKB_CB(skb)->family = AF_INET6;
+		XFRM_SPI_SKB_CB(skb)->daddroff = offsetof(struct ipv6hdr, daddr);
+		return xfrm_input(skb, nexthdr, spi, encap_type);
 	}
 	rcu_read_unlock();
 	return -EINVAL;
@@ -332,6 +336,13 @@ static int vti6_rcv(struct sk_buff *skb)
 	return 0;
 }
 
+static int vti6_rcv(struct sk_buff *skb)
+{
+	int nexthdr = skb_network_header(skb)[IP6CB(skb)->nhoff];
+
+	return vti6_input_proto(skb, nexthdr, 0, 0);
+}
+
 static int vti6_rcv_cb(struct sk_buff *skb, int err)
 {
 	unsigned short family;
@@ -1185,6 +1196,7 @@ static struct pernet_operations vti6_net_ops = {
 
 static struct xfrm6_protocol vti_esp6_protocol __read_mostly = {
 	.handler	=	vti6_rcv,
+	.input_handler	=	vti6_input_proto,
 	.cb_handler	=	vti6_rcv_cb,
 	.err_handler	=	vti6_err,
 	.priority	=	100,
@@ -1192,6 +1204,7 @@ static struct xfrm6_protocol vti_esp6_protocol __read_mostly = {
 
 static struct xfrm6_protocol vti_ah6_protocol __read_mostly = {
 	.handler	=	vti6_rcv,
+	.input_handler	=	vti6_input_proto,
 	.cb_handler	=	vti6_rcv_cb,
 	.err_handler	=	vti6_err,
 	.priority	=	100,
@@ -1199,6 +1212,7 @@ static struct xfrm6_protocol vti_ah6_protocol __read_mostly = {
 
 static struct xfrm6_protocol vti_ipcomp6_protocol __read_mostly = {
 	.handler	=	vti6_rcv,
+	.input_handler	=	vti6_input_proto,
 	.cb_handler	=	vti6_rcv_cb,
 	.err_handler	=	vti6_err,
 	.priority	=	100,

net/ipv6/ipcomp6.c

@@ -183,6 +183,7 @@ static const struct xfrm_type ipcomp6_type = {
 
 static struct xfrm6_protocol ipcomp6_protocol = {
 	.handler	= xfrm6_rcv,
+	.input_handler	= xfrm_input,
 	.cb_handler	= ipcomp6_rcv_cb,
 	.err_handler	= ipcomp6_err,
 	.priority	= 0,

net/ipv6/xfrm6_input.c

@@ -17,11 +17,6 @@
 #include <net/ipv6.h>
 #include <net/xfrm.h>
 
-int xfrm6_extract_input(struct xfrm_state *x, struct sk_buff *skb)
-{
-	return xfrm6_extract_header(skb);
-}
-
 int xfrm6_rcv_spi(struct sk_buff *skb, int nexthdr, __be32 spi,
 		  struct ip6_tnl *t)
 {
@@ -35,9 +30,12 @@ EXPORT_SYMBOL(xfrm6_rcv_spi);
 static int xfrm6_transport_finish2(struct net *net, struct sock *sk,
 				   struct sk_buff *skb)
 {
-	if (xfrm_trans_queue(skb, ip6_rcv_finish))
-		__kfree_skb(skb);
-	return -1;
+	if (xfrm_trans_queue(skb, ip6_rcv_finish)) {
+		kfree_skb(skb);
+		return NET_RX_DROP;
+	}
+
+	return 0;
 }
 
 int xfrm6_transport_finish(struct sk_buff *skb, int async)
@@ -60,13 +58,106 @@ int xfrm6_transport_finish(struct sk_buff *skb, int async)
 	if (xo && (xo->flags & XFRM_GRO)) {
 		skb_mac_header_rebuild(skb);
 		skb_reset_transport_header(skb);
-		return -1;
+		return 0;
 	}
 
 	NF_HOOK(NFPROTO_IPV6, NF_INET_PRE_ROUTING,
 		dev_net(skb->dev), NULL, skb, skb->dev, NULL,
 		xfrm6_transport_finish2);
-	return -1;
+	return 0;
+}
+
+/* If it's a keepalive packet, then just eat it.
+ * If it's an encapsulated packet, then pass it to the
+ * IPsec xfrm input.
+ * Returns 0 if skb passed to xfrm or was dropped.
+ * Returns >0 if skb should be passed to UDP.
+ * Returns <0 if skb should be resubmitted (-ret is protocol)
+ */
+int xfrm6_udp_encap_rcv(struct sock *sk, struct sk_buff *skb)
+{
+	struct udp_sock *up = udp_sk(sk);
+	struct udphdr *uh;
+	struct ipv6hdr *ip6h;
+	int len;
+	int ip6hlen = sizeof(struct ipv6hdr);
+
+	__u8 *udpdata;
+	__be32 *udpdata32;
+	__u16 encap_type = up->encap_type;
+
+	/* if this is not encapsulated socket, then just return now */
+	if (!encap_type)
+		return 1;
+
+	/* If this is a paged skb, make sure we pull up
+	 * whatever data we need to look at. */
+	len = skb->len - sizeof(struct udphdr);
+	if (!pskb_may_pull(skb, sizeof(struct udphdr) + min(len, 8)))
+		return 1;
+
+	/* Now we can get the pointers */
+	uh = udp_hdr(skb);
+	udpdata = (__u8 *)uh + sizeof(struct udphdr);
+	udpdata32 = (__be32 *)udpdata;
+
+	switch (encap_type) {
+	default:
+	case UDP_ENCAP_ESPINUDP:
+		/* Check if this is a keepalive packet.  If so, eat it. */
+		if (len == 1 && udpdata[0] == 0xff) {
+			goto drop;
+		} else if (len > sizeof(struct ip_esp_hdr) && udpdata32[0] != 0) {
+			/* ESP Packet without Non-ESP header */
+			len = sizeof(struct udphdr);
+		} else
+			/* Must be an IKE packet.. pass it through */
+			return 1;
+		break;
+	case UDP_ENCAP_ESPINUDP_NON_IKE:
+		/* Check if this is a keepalive packet.  If so, eat it. */
+		if (len == 1 && udpdata[0] == 0xff) {
+			goto drop;
+		} else if (len > 2 * sizeof(u32) + sizeof(struct ip_esp_hdr) &&
+			   udpdata32[0] == 0 && udpdata32[1] == 0) {
+
+			/* ESP Packet with Non-IKE marker */
+			len = sizeof(struct udphdr) + 2 * sizeof(u32);
+		} else
+			/* Must be an IKE packet.. pass it through */
+			return 1;
+		break;
+	}
+
+	/* At this point we are sure that this is an ESPinUDP packet,
+	 * so we need to remove 'len' bytes from the packet (the UDP
+	 * header and optional ESP marker bytes) and then modify the
+	 * protocol to ESP, and then call into the transform receiver.
+	 */
+	if (skb_unclone(skb, GFP_ATOMIC))
+		goto drop;
+
+	/* Now we can update and verify the packet length... */
+	ip6h = ipv6_hdr(skb);
+	ip6h->payload_len = htons(ntohs(ip6h->payload_len) - len);
+	if (skb->len < ip6hlen + len) {
+		/* packet is too small!?! */
+		goto drop;
+	}
+
+	/* pull the data buffer up to the ESP header and set the
+	 * transport header to point to ESP.  Keep UDP on the stack
+	 * for later.
+	 */
+	__skb_pull(skb, len);
+	skb_reset_transport_header(skb);
+
+	/* process ESP */
+	return xfrm6_rcv_encap(skb, IPPROTO_ESP, 0, encap_type);
+
+drop:
+	kfree_skb(skb);
+	return 0;
 }
 
 int xfrm6_rcv_tnl(struct sk_buff *skb, struct ip6_tnl *t)

net/ipv6/xfrm6_output.c

@@ -23,24 +23,7 @@ int xfrm6_find_1stfragopt(struct xfrm_state *x, struct sk_buff *skb,
 }
 EXPORT_SYMBOL(xfrm6_find_1stfragopt);
 
-static int xfrm6_local_dontfrag(struct sk_buff *skb)
-{
-	int proto;
-	struct sock *sk = skb->sk;
-
-	if (sk) {
-		if (sk->sk_family != AF_INET6)
-			return 0;
-
-		proto = sk->sk_protocol;
-		if (proto == IPPROTO_UDP || proto == IPPROTO_RAW)
-			return inet6_sk(sk)->dontfrag;
-	}
-
-	return 0;
-}
-
-static void xfrm6_local_rxpmtu(struct sk_buff *skb, u32 mtu)
+void xfrm6_local_rxpmtu(struct sk_buff *skb, u32 mtu)
 {
 	struct flowi6 fl6;
 	struct sock *sk = skb->sk;
@@ -64,80 +47,9 @@ void xfrm6_local_error(struct sk_buff *skb, u32 mtu)
 	ipv6_local_error(sk, EMSGSIZE, &fl6, mtu);
 }
 
-static int xfrm6_tunnel_check_size(struct sk_buff *skb)
-{
-	int mtu, ret = 0;
-	struct dst_entry *dst = skb_dst(skb);
-
-	if (skb->ignore_df)
-		goto out;
-
-	mtu = dst_mtu(dst);
-	if (mtu < IPV6_MIN_MTU)
-		mtu = IPV6_MIN_MTU;
-
-	if ((!skb_is_gso(skb) && skb->len > mtu) ||
-	    (skb_is_gso(skb) &&
-	     !skb_gso_validate_network_len(skb, ip6_skb_dst_mtu(skb)))) {
-		skb->dev = dst->dev;
-		skb->protocol = htons(ETH_P_IPV6);
-
-		if (xfrm6_local_dontfrag(skb))
-			xfrm6_local_rxpmtu(skb, mtu);
-		else if (skb->sk)
-			xfrm_local_error(skb, mtu);
-		else
-			icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu);
-		ret = -EMSGSIZE;
-	}
-out:
-	return ret;
-}
-
-int xfrm6_extract_output(struct xfrm_state *x, struct sk_buff *skb)
-{
-	int err;
-
-	err = xfrm6_tunnel_check_size(skb);
-	if (err)
-		return err;
-
-	XFRM_MODE_SKB_CB(skb)->protocol = ipv6_hdr(skb)->nexthdr;
-
-	return xfrm6_extract_header(skb);
-}
-
-int xfrm6_output_finish(struct sock *sk, struct sk_buff *skb)
-{
-	memset(IP6CB(skb), 0, sizeof(*IP6CB(skb)));
-
-	IP6CB(skb)->flags |= IP6SKB_XFRM_TRANSFORMED;
-
-	return xfrm_output(sk, skb);
-}
-
-static int __xfrm6_output_state_finish(struct xfrm_state *x, struct sock *sk,
-				       struct sk_buff *skb)
-{
-	const struct xfrm_state_afinfo *afinfo;
-	int ret = -EAFNOSUPPORT;
-
-	rcu_read_lock();
-	afinfo = xfrm_state_afinfo_get_rcu(x->outer_mode.family);
-	if (likely(afinfo))
-		ret = afinfo->output_finish(sk, skb);
-	else
-		kfree_skb(skb);
-	rcu_read_unlock();
-
-	return ret;
-}
-
 static int __xfrm6_output_finish(struct net *net, struct sock *sk, struct sk_buff *skb)
 {
-	struct xfrm_state *x = skb_dst(skb)->xfrm;
-
-	return __xfrm6_output_state_finish(x, sk, skb);
+	return xfrm_output(sk, skb);
 }
 
 static int __xfrm6_output(struct net *net, struct sock *sk, struct sk_buff *skb)
@@ -164,7 +76,7 @@ static int __xfrm6_output(struct net *net, struct sock *sk, struct sk_buff *skb)
 
 	toobig = skb->len > mtu && !skb_is_gso(skb);
 
-	if (toobig && xfrm6_local_dontfrag(skb)) {
+	if (toobig && xfrm6_local_dontfrag(skb->sk)) {
 		xfrm6_local_rxpmtu(skb, mtu);
 		kfree_skb(skb);
 		return -EMSGSIZE;
@@ -179,7 +91,7 @@ static int __xfrm6_output(struct net *net, struct sock *sk, struct sk_buff *skb)
 				    __xfrm6_output_finish);
 
 skip_frag:
-	return __xfrm6_output_state_finish(x, sk, skb);
+	return xfrm_output(sk, skb);
 }
 
 int xfrm6_output(struct net *net, struct sock *sk, struct sk_buff *skb)

net/ipv6/xfrm6_protocol.c

@@ -14,6 +14,7 @@
 #include <linux/mutex.h>
 #include <linux/skbuff.h>
 #include <linux/icmpv6.h>
+#include <net/ip6_route.h>
 #include <net/ipv6.h>
 #include <net/protocol.h>
 #include <net/xfrm.h>
@@ -58,6 +59,53 @@ static int xfrm6_rcv_cb(struct sk_buff *skb, u8 protocol, int err)
 	return 0;
 }
 
+int xfrm6_rcv_encap(struct sk_buff *skb, int nexthdr, __be32 spi,
+		    int encap_type)
+{
+	int ret;
+	struct xfrm6_protocol *handler;
+	struct xfrm6_protocol __rcu **head = proto_handlers(nexthdr);
+
+	XFRM_TUNNEL_SKB_CB(skb)->tunnel.ip6 = NULL;
+	XFRM_SPI_SKB_CB(skb)->family = AF_INET6;
+	XFRM_SPI_SKB_CB(skb)->daddroff = offsetof(struct ipv6hdr, daddr);
+
+	if (!head)
+		goto out;
+
+	if (!skb_dst(skb)) {
+		const struct ipv6hdr *ip6h = ipv6_hdr(skb);
+		int flags = RT6_LOOKUP_F_HAS_SADDR;
+		struct dst_entry *dst;
+		struct flowi6 fl6 = {
+			.flowi6_iif   = skb->dev->ifindex,
+			.daddr        = ip6h->daddr,
+			.saddr        = ip6h->saddr,
+			.flowlabel    = ip6_flowinfo(ip6h),
+			.flowi6_mark  = skb->mark,
+			.flowi6_proto = ip6h->nexthdr,
+		};
+
+		dst = ip6_route_input_lookup(dev_net(skb->dev), skb->dev, &fl6,
+					     skb, flags);
+		if (dst->error)
+			goto drop;
+		skb_dst_set(skb, dst);
+	}
+
+	for_each_protocol_rcu(*head, handler)
+		if ((ret = handler->input_handler(skb, nexthdr, spi, encap_type)) != -EINVAL)
+			return ret;
+
+out:
+	icmpv6_send(skb, ICMPV6_DEST_UNREACH, ICMPV6_PORT_UNREACH, 0);
+
+drop:
+	kfree_skb(skb);
+	return 0;
+}
+EXPORT_SYMBOL(xfrm6_rcv_encap);
+
 static int xfrm6_esp_rcv(struct sk_buff *skb)
 {
 	int ret;

net/ipv6/xfrm6_state.c

@@ -13,37 +13,11 @@
  */
 
 #include <net/xfrm.h>
-#include <linux/pfkeyv2.h>
-#include <linux/ipsec.h>
-#include <linux/netfilter_ipv6.h>
-#include <linux/export.h>
-#include <net/dsfield.h>
-#include <net/ipv6.h>
-#include <net/addrconf.h>
-
-int xfrm6_extract_header(struct sk_buff *skb)
-{
-	struct ipv6hdr *iph = ipv6_hdr(skb);
-
-	XFRM_MODE_SKB_CB(skb)->ihl = sizeof(*iph);
-	XFRM_MODE_SKB_CB(skb)->id = 0;
-	XFRM_MODE_SKB_CB(skb)->frag_off = htons(IP_DF);
-	XFRM_MODE_SKB_CB(skb)->tos = ipv6_get_dsfield(iph);
-	XFRM_MODE_SKB_CB(skb)->ttl = iph->hop_limit;
-	XFRM_MODE_SKB_CB(skb)->optlen = 0;
-	memcpy(XFRM_MODE_SKB_CB(skb)->flow_lbl, iph->flow_lbl,
-	       sizeof(XFRM_MODE_SKB_CB(skb)->flow_lbl));
-
-	return 0;
-}
 
 static struct xfrm_state_afinfo xfrm6_state_afinfo = {
 	.family			= AF_INET6,
 	.proto			= IPPROTO_IPV6,
 	.output			= xfrm6_output,
-	.output_finish		= xfrm6_output_finish,
-	.extract_input		= xfrm6_extract_input,
-	.extract_output		= xfrm6_extract_output,
 	.transport_finish	= xfrm6_transport_finish,
 	.local_error		= xfrm6_local_error,
 };

net/xfrm/Kconfig

@@ -99,4 +99,7 @@ config NET_KEY_MIGRATE
 
 	  If unsure, say N.
 
+config XFRM_ESPINTCP
+	bool
+
 endif # INET

net/xfrm/Makefile

@@ -11,4 +11,4 @@ obj-$(CONFIG_XFRM_ALGO) += xfrm_algo.o
 obj-$(CONFIG_XFRM_USER) += xfrm_user.o
 obj-$(CONFIG_XFRM_IPCOMP) += xfrm_ipcomp.o
 obj-$(CONFIG_XFRM_INTERFACE) += xfrm_interface.o
-obj-$(CONFIG_INET_ESPINTCP) += espintcp.o
+obj-$(CONFIG_XFRM_ESPINTCP) += espintcp.o

net/xfrm/espintcp.c

@@ -6,6 +6,9 @@
 #include <net/espintcp.h>
 #include <linux/skmsg.h>
 #include <net/inet_common.h>
+#if IS_ENABLED(CONFIG_IPV6)
+#include <net/ipv6_stubs.h>
+#endif
 
 static void handle_nonesp(struct espintcp_ctx *ctx, struct sk_buff *skb,
 			  struct sock *sk)
@@ -31,6 +34,11 @@ static void handle_esp(struct sk_buff *skb, struct sock *sk)
 	rcu_read_lock();
 	skb->dev = dev_get_by_index_rcu(sock_net(sk), skb->skb_iif);
 	local_bh_disable();
+#if IS_ENABLED(CONFIG_IPV6)
+	if (sk->sk_family == AF_INET6)
+		ipv6_stub->xfrm6_rcv_encap(skb, IPPROTO_ESP, 0, TCP_ENCAP_ESPINTCP);
+	else
+#endif
 		xfrm4_rcv_encap(skb, IPPROTO_ESP, 0, TCP_ENCAP_ESPINTCP);
 	local_bh_enable();
 	rcu_read_unlock();
@@ -347,6 +355,9 @@ static int espintcp_sendmsg(struct sock *sk, struct msghdr *msg, size_t size)
 
 static struct proto espintcp_prot __ro_after_init;
 static struct proto_ops espintcp_ops __ro_after_init;
+static struct proto espintcp6_prot;
+static struct proto_ops espintcp6_ops;
+static DEFINE_MUTEX(tcpv6_prot_mutex);
 
 static void espintcp_data_ready(struct sock *sk)
 {
@@ -384,10 +395,14 @@ static void espintcp_destruct(struct sock *sk)
 
 bool tcp_is_ulp_esp(struct sock *sk)
 {
-	return sk->sk_prot == &espintcp_prot;
+	return sk->sk_prot == &espintcp_prot || sk->sk_prot == &espintcp6_prot;
 }
 EXPORT_SYMBOL_GPL(tcp_is_ulp_esp);
 
+static void build_protos(struct proto *espintcp_prot,
+			 struct proto_ops *espintcp_ops,
+			 const struct proto *orig_prot,
+			 const struct proto_ops *orig_ops);
 static int espintcp_init_sk(struct sock *sk)
 {
 	struct inet_connection_sock *icsk = inet_csk(sk);
@@ -415,8 +430,19 @@ static int espintcp_init_sk(struct sock *sk)
 	strp_check_rcv(&ctx->strp);
 	skb_queue_head_init(&ctx->ike_queue);
 	skb_queue_head_init(&ctx->out_queue);
+
+	if (sk->sk_family == AF_INET) {
 		sk->sk_prot = &espintcp_prot;
 		sk->sk_socket->ops = &espintcp_ops;
+	} else {
+		mutex_lock(&tcpv6_prot_mutex);
+		if (!espintcp6_prot.recvmsg)
+			build_protos(&espintcp6_prot, &espintcp6_ops, sk->sk_prot, sk->sk_socket->ops);
+		mutex_unlock(&tcpv6_prot_mutex);
+
+		sk->sk_prot = &espintcp6_prot;
+		sk->sk_socket->ops = &espintcp6_ops;
+	}
 	ctx->saved_data_ready = sk->sk_data_ready;
 	ctx->saved_write_space = sk->sk_write_space;
 	sk->sk_data_ready = espintcp_data_ready;
@@ -489,6 +515,20 @@ static __poll_t espintcp_poll(struct file *file, struct socket *sock,
 	return mask;
 }
 
+static void build_protos(struct proto *espintcp_prot,
+			 struct proto_ops *espintcp_ops,
+			 const struct proto *orig_prot,
+			 const struct proto_ops *orig_ops)
+{
+	memcpy(espintcp_prot, orig_prot, sizeof(struct proto));
+	memcpy(espintcp_ops, orig_ops, sizeof(struct proto_ops));
+	espintcp_prot->sendmsg = espintcp_sendmsg;
+	espintcp_prot->recvmsg = espintcp_recvmsg;
+	espintcp_prot->close = espintcp_close;
+	espintcp_prot->release_cb = espintcp_release;
+	espintcp_ops->poll = espintcp_poll;
+}
+
 static struct tcp_ulp_ops espintcp_ulp __read_mostly = {
 	.name = "espintcp",
 	.owner = THIS_MODULE,
@@ -497,13 +537,7 @@ static struct tcp_ulp_ops espintcp_ulp __read_mostly = {
 
 void __init espintcp_init(void)
 {
-	memcpy(&espintcp_prot, &tcp_prot, sizeof(tcp_prot));
-	memcpy(&espintcp_ops, &inet_stream_ops, sizeof(inet_stream_ops));
-	espintcp_prot.sendmsg = espintcp_sendmsg;
-	espintcp_prot.recvmsg = espintcp_recvmsg;
-	espintcp_prot.close = espintcp_close;
-	espintcp_prot.release_cb = espintcp_release;
-	espintcp_ops.poll = espintcp_poll;
+	build_protos(&espintcp_prot, &espintcp_ops, &tcp_prot, &inet_stream_ops);
 
 	tcp_register_ulp(&espintcp_ulp);
 }

net/xfrm/xfrm_inout.h

@@ -6,6 +6,38 @@
 #ifndef XFRM_INOUT_H
 #define XFRM_INOUT_H 1
 
+static inline void xfrm4_extract_header(struct sk_buff *skb)
+{
+	const struct iphdr *iph = ip_hdr(skb);
+
+	XFRM_MODE_SKB_CB(skb)->ihl = sizeof(*iph);
+	XFRM_MODE_SKB_CB(skb)->id = iph->id;
+	XFRM_MODE_SKB_CB(skb)->frag_off = iph->frag_off;
+	XFRM_MODE_SKB_CB(skb)->tos = iph->tos;
+	XFRM_MODE_SKB_CB(skb)->ttl = iph->ttl;
+	XFRM_MODE_SKB_CB(skb)->optlen = iph->ihl * 4 - sizeof(*iph);
+	memset(XFRM_MODE_SKB_CB(skb)->flow_lbl, 0,
+	       sizeof(XFRM_MODE_SKB_CB(skb)->flow_lbl));
+}
+
+static inline void xfrm6_extract_header(struct sk_buff *skb)
+{
+#if IS_ENABLED(CONFIG_IPV6)
+	struct ipv6hdr *iph = ipv6_hdr(skb);
+
+	XFRM_MODE_SKB_CB(skb)->ihl = sizeof(*iph);
+	XFRM_MODE_SKB_CB(skb)->id = 0;
+	XFRM_MODE_SKB_CB(skb)->frag_off = htons(IP_DF);
+	XFRM_MODE_SKB_CB(skb)->tos = ipv6_get_dsfield(iph);
+	XFRM_MODE_SKB_CB(skb)->ttl = iph->hop_limit;
+	XFRM_MODE_SKB_CB(skb)->optlen = 0;
+	memcpy(XFRM_MODE_SKB_CB(skb)->flow_lbl, iph->flow_lbl,
+	       sizeof(XFRM_MODE_SKB_CB(skb)->flow_lbl));
+#else
+	WARN_ON_ONCE(1);
+#endif
+}
+
 static inline void xfrm6_beet_make_header(struct sk_buff *skb)
 {
 	struct ipv6hdr *iph = ipv6_hdr(skb);

net/xfrm/xfrm_input.c

@@ -353,17 +353,18 @@ xfrm_inner_mode_encap_remove(struct xfrm_state *x,
 static int xfrm_prepare_input(struct xfrm_state *x, struct sk_buff *skb)
 {
 	const struct xfrm_mode *inner_mode = &x->inner_mode;
-	const struct xfrm_state_afinfo *afinfo;
-	int err = -EAFNOSUPPORT;
-
-	rcu_read_lock();
-	afinfo = xfrm_state_afinfo_get_rcu(x->outer_mode.family);
-	if (likely(afinfo))
-		err = afinfo->extract_input(x, skb);
-	rcu_read_unlock();
 
-	if (err)
-		return err;
+	switch (x->outer_mode.family) {
+	case AF_INET:
+		xfrm4_extract_header(skb);
+		break;
+	case AF_INET6:
+		xfrm6_extract_header(skb);
+		break;
+	default:
+		WARN_ON_ONCE(1);
+		return -EAFNOSUPPORT;
+	}
 
 	if (x->sel.family == AF_UNSPEC) {
 		inner_mode = xfrm_ip2inner_mode(x, XFRM_MODE_SKB_CB(skb)->protocol);

net/xfrm/xfrm_interface.c

@@ -145,7 +145,6 @@ static int xfrmi_create(struct net_device *dev)
 	if (err < 0)
 		goto out;
 
-	dev_hold(dev);
 	xfrmi_link(xfrmn, xi);
 
 	return 0;
@@ -175,7 +174,6 @@ static void xfrmi_dev_uninit(struct net_device *dev)
 	struct xfrmi_net *xfrmn = net_generic(xi->net, xfrmi_net_id);
 
 	xfrmi_unlink(xfrmn, xi);
-	dev_put(dev);
 }
 
 static void xfrmi_scrub_packet(struct sk_buff *skb, bool xnet)
@@ -757,6 +755,7 @@ static struct pernet_operations xfrmi_net_ops = {
 
 static struct xfrm6_protocol xfrmi_esp6_protocol __read_mostly = {
 	.handler	=	xfrm6_rcv,
+	.input_handler	=	xfrm_input,
 	.cb_handler	=	xfrmi_rcv_cb,
 	.err_handler	=	xfrmi6_err,
 	.priority	=	10,
@@ -764,6 +763,7 @@ static struct xfrm6_protocol xfrmi_esp6_protocol __read_mostly = {
 
 static struct xfrm6_protocol xfrmi_ah6_protocol __read_mostly = {
 	.handler	=	xfrm6_rcv,
+	.input_handler	=	xfrm_input,
 	.cb_handler	=	xfrmi_rcv_cb,
 	.err_handler	=	xfrmi6_err,
 	.priority	=	10,
@@ -771,6 +771,7 @@ static struct xfrm6_protocol xfrmi_ah6_protocol __read_mostly = {
 
 static struct xfrm6_protocol xfrmi_ipcomp6_protocol __read_mostly = {
 	.handler	=	xfrm6_rcv,
+	.input_handler	=	xfrm_input,
 	.cb_handler	=	xfrmi_rcv_cb,
 	.err_handler	=	xfrmi6_err,
 	.priority	=	10,

net/xfrm/xfrm_output.c

@@ -13,9 +13,15 @@
 #include <linux/slab.h>
 #include <linux/spinlock.h>
 #include <net/dst.h>
+#include <net/icmp.h>
 #include <net/inet_ecn.h>
 #include <net/xfrm.h>
 
+#if IS_ENABLED(CONFIG_IPV6)
+#include <net/ip6_route.h>
+#include <net/ipv6_stubs.h>
+#endif
+
 #include "xfrm_inout.h"
 
 static int xfrm_output2(struct net *net, struct sock *sk, struct sk_buff *skb);
@@ -565,6 +571,22 @@ int xfrm_output(struct sock *sk, struct sk_buff *skb)
 	struct xfrm_state *x = skb_dst(skb)->xfrm;
 	int err;
 
+	switch (x->outer_mode.family) {
+	case AF_INET:
+		memset(IPCB(skb), 0, sizeof(*IPCB(skb)));
+#ifdef CONFIG_NETFILTER
+		IPCB(skb)->flags |= IPSKB_XFRM_TRANSFORMED;
+#endif
+		break;
+	case AF_INET6:
+		memset(IP6CB(skb), 0, sizeof(*IP6CB(skb)));
+
+#ifdef CONFIG_NETFILTER
+		IP6CB(skb)->flags |= IP6SKB_XFRM_TRANSFORMED;
+#endif
+		break;
+	}
+
 	secpath_reset(skb);
 
 	if (xfrm_dev_offload_ok(skb, x)) {
@@ -609,11 +631,101 @@ int xfrm_output(struct sock *sk, struct sk_buff *skb)
 }
 EXPORT_SYMBOL_GPL(xfrm_output);
 
+static int xfrm4_tunnel_check_size(struct sk_buff *skb)
+{
+	int mtu, ret = 0;
+
+	if (IPCB(skb)->flags & IPSKB_XFRM_TUNNEL_SIZE)
+		goto out;
+
+	if (!(ip_hdr(skb)->frag_off & htons(IP_DF)) || skb->ignore_df)
+		goto out;
+
+	mtu = dst_mtu(skb_dst(skb));
+	if ((!skb_is_gso(skb) && skb->len > mtu) ||
+	    (skb_is_gso(skb) &&
+	     !skb_gso_validate_network_len(skb, ip_skb_dst_mtu(skb->sk, skb)))) {
+		skb->protocol = htons(ETH_P_IP);
+
+		if (skb->sk)
+			xfrm_local_error(skb, mtu);
+		else
+			icmp_send(skb, ICMP_DEST_UNREACH,
+				  ICMP_FRAG_NEEDED, htonl(mtu));
+		ret = -EMSGSIZE;
+	}
+out:
+	return ret;
+}
+
+static int xfrm4_extract_output(struct xfrm_state *x, struct sk_buff *skb)
+{
+	int err;
+
+	err = xfrm4_tunnel_check_size(skb);
+	if (err)
+		return err;
+
+	XFRM_MODE_SKB_CB(skb)->protocol = ip_hdr(skb)->protocol;
+
+	xfrm4_extract_header(skb);
+	return 0;
+}
+
+#if IS_ENABLED(CONFIG_IPV6)
+static int xfrm6_tunnel_check_size(struct sk_buff *skb)
+{
+	int mtu, ret = 0;
+	struct dst_entry *dst = skb_dst(skb);
+
+	if (skb->ignore_df)
+		goto out;
+
+	mtu = dst_mtu(dst);
+	if (mtu < IPV6_MIN_MTU)
+		mtu = IPV6_MIN_MTU;
+
+	if ((!skb_is_gso(skb) && skb->len > mtu) ||
+	    (skb_is_gso(skb) &&
+	     !skb_gso_validate_network_len(skb, ip6_skb_dst_mtu(skb)))) {
+		skb->dev = dst->dev;
+		skb->protocol = htons(ETH_P_IPV6);
+
+		if (xfrm6_local_dontfrag(skb->sk))
+			ipv6_stub->xfrm6_local_rxpmtu(skb, mtu);
+		else if (skb->sk)
+			xfrm_local_error(skb, mtu);
+		else
+			icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu);
+		ret = -EMSGSIZE;
+	}
+out:
+	return ret;
+}
+#endif
+
+static int xfrm6_extract_output(struct xfrm_state *x, struct sk_buff *skb)
+{
+#if IS_ENABLED(CONFIG_IPV6)
+	int err;
+
+	err = xfrm6_tunnel_check_size(skb);
+	if (err)
+		return err;
+
+	XFRM_MODE_SKB_CB(skb)->protocol = ipv6_hdr(skb)->nexthdr;
+
+	xfrm6_extract_header(skb);
+	return 0;
+#else
+	WARN_ON_ONCE(1);
+	return -EAFNOSUPPORT;
+#endif
+}
+
 static int xfrm_inner_extract_output(struct xfrm_state *x, struct sk_buff *skb)
 {
-	const struct xfrm_state_afinfo *afinfo;
 	const struct xfrm_mode *inner_mode;
-	int err = -EAFNOSUPPORT;
 
 	if (x->sel.family == AF_UNSPEC)
 		inner_mode = xfrm_ip2inner_mode(x,
@@ -624,13 +736,14 @@ static int xfrm_inner_extract_output(struct xfrm_state *x, struct sk_buff *skb)
 	if (inner_mode == NULL)
 		return -EAFNOSUPPORT;
 
-	rcu_read_lock();
-	afinfo = xfrm_state_afinfo_get_rcu(inner_mode->family);
-	if (likely(afinfo))
-		err = afinfo->extract_output(x, skb);
-	rcu_read_unlock();
+	switch (inner_mode->family) {
+	case AF_INET:
+		return xfrm4_extract_output(x, skb);
+	case AF_INET6:
+		return xfrm6_extract_output(x, skb);
+	}
 
-	return err;
+	return -EAFNOSUPPORT;
 }
 
 void xfrm_local_error(struct sk_buff *skb, int mtu)