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 2019-04-30

1) A lot of work to remove indirections from the xfrm code.
   From Florian Westphal.

2) Support ESP offload in combination with gso partial.
   From Boris Pismenny.

3) Remove some duplicated code from vti4.
   From Jeremy Sowden.

Please note that there is merge conflict

between commit:

8742dc86d0c7 ("xfrm4: Fix uninitialized memory read in _decode_session4")

from the ipsec tree and commit:

c53ac41e3720 ("xfrm: remove decode_session indirection from afinfo_policy")

from the ipsec-next tree. The merge conflict will appear
when those trees get merged during the merge window.
The conflict can be solved as it is done in linux-next:

https://lkml.org/lkml/2019/4/25/1207

Please pull or let me know if there are problems.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

8 files changed, 1078 insertions, 181 deletions

diff --git a/net/xfrm/Kconfig b/net/xfrm/Kconfig
index 5d43aaa17027..1ec8071226b2 100644
--- a/net/xfrm/Kconfig
+++ b/net/xfrm/Kconfig
@@ -3,7 +3,7 @@
 #
 config XFRM
        bool
-       depends on NET
+       depends on INET
        select GRO_CELLS
        select SKB_EXTENSIONS
 
@@ -15,9 +15,9 @@ config XFRM_ALGO
 	select XFRM
 	select CRYPTO
 
+if INET
 config XFRM_USER
 	tristate "Transformation user configuration interface"
-	depends on INET
 	select XFRM_ALGO
 	---help---
 	  Support for Transformation(XFRM) user configuration interface
@@ -56,7 +56,7 @@ config XFRM_MIGRATE
 
 config XFRM_STATISTICS
 	bool "Transformation statistics"
-	depends on INET && XFRM && PROC_FS
+	depends on XFRM && PROC_FS
 	---help---
 	  This statistics is not a SNMP/MIB specification but shows
 	  statistics about transformation error (or almost error) factor
@@ -95,3 +95,5 @@ config NET_KEY_MIGRATE
 	  <draft-sugimoto-mip6-pfkey-migrate>.
 
 	  If unsure, say N.
+
+endif # INET
diff --git a/net/xfrm/xfrm_device.c b/net/xfrm/xfrm_device.c
index 2db1626557c5..b24cd86a02c3 100644
--- a/net/xfrm/xfrm_device.c
+++ b/net/xfrm/xfrm_device.c
@@ -23,6 +23,60 @@
 #include <linux/notifier.h>
 
 #ifdef CONFIG_XFRM_OFFLOAD
+static void __xfrm_transport_prep(struct xfrm_state *x, struct sk_buff *skb,
+				  unsigned int hsize)
+{
+	struct xfrm_offload *xo = xfrm_offload(skb);
+
+	skb_reset_mac_len(skb);
+	pskb_pull(skb, skb->mac_len + hsize + x->props.header_len);
+
+	if (xo->flags & XFRM_GSO_SEGMENT) {
+		skb_reset_transport_header(skb);
+		skb->transport_header -= x->props.header_len;
+	}
+}
+
+static void __xfrm_mode_tunnel_prep(struct xfrm_state *x, struct sk_buff *skb,
+				    unsigned int hsize)
+
+{
+	struct xfrm_offload *xo = xfrm_offload(skb);
+
+	if (xo->flags & XFRM_GSO_SEGMENT)
+		skb->transport_header = skb->network_header + hsize;
+
+	skb_reset_mac_len(skb);
+	pskb_pull(skb, skb->mac_len + x->props.header_len);
+}
+
+/* Adjust pointers into the packet when IPsec is done at layer2 */
+static void xfrm_outer_mode_prep(struct xfrm_state *x, struct sk_buff *skb)
+{
+	switch (x->outer_mode.encap) {
+	case XFRM_MODE_TUNNEL:
+		if (x->outer_mode.family == AF_INET)
+			return __xfrm_mode_tunnel_prep(x, skb,
+						       sizeof(struct iphdr));
+		if (x->outer_mode.family == AF_INET6)
+			return __xfrm_mode_tunnel_prep(x, skb,
+						       sizeof(struct ipv6hdr));
+		break;
+	case XFRM_MODE_TRANSPORT:
+		if (x->outer_mode.family == AF_INET)
+			return __xfrm_transport_prep(x, skb,
+						     sizeof(struct iphdr));
+		if (x->outer_mode.family == AF_INET6)
+			return __xfrm_transport_prep(x, skb,
+						     sizeof(struct ipv6hdr));
+		break;
+	case XFRM_MODE_ROUTEOPTIMIZATION:
+	case XFRM_MODE_IN_TRIGGER:
+	case XFRM_MODE_BEET:
+		break;
+	}
+}
+
 struct sk_buff *validate_xmit_xfrm(struct sk_buff *skb, netdev_features_t features, bool *again)
 {
 	int err;
@@ -78,7 +132,8 @@ struct sk_buff *validate_xmit_xfrm(struct sk_buff *skb, netdev_features_t featur
 	}
 
 	if (!skb->next) {
-		x->outer_mode->xmit(x, skb);
+		esp_features |= skb->dev->gso_partial_features;
+		xfrm_outer_mode_prep(x, skb);
 
 		xo->flags |= XFRM_DEV_RESUME;
 
@@ -101,12 +156,14 @@ struct sk_buff *validate_xmit_xfrm(struct sk_buff *skb, netdev_features_t featur
 
 	do {
 		struct sk_buff *nskb = skb2->next;
+
+		esp_features |= skb->dev->gso_partial_features;
 		skb_mark_not_on_list(skb2);
 
 		xo = xfrm_offload(skb2);
 		xo->flags |= XFRM_DEV_RESUME;
 
-		x->outer_mode->xmit(x, skb2);
+		xfrm_outer_mode_prep(x, skb2);
 
 		err = x->type_offload->xmit(x, skb2, esp_features);
 		if (!err) {
diff --git a/net/xfrm/xfrm_inout.h b/net/xfrm/xfrm_inout.h
new file mode 100644
index 000000000000..c7b0318938e2
--- /dev/null
+++ b/net/xfrm/xfrm_inout.h
@@ -0,0 +1,38 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#include <linux/ipv6.h>
+#include <net/dsfield.h>
+#include <net/xfrm.h>
+
+#ifndef XFRM_INOUT_H
+#define XFRM_INOUT_H 1
+
+static inline void xfrm6_beet_make_header(struct sk_buff *skb)
+{
+	struct ipv6hdr *iph = ipv6_hdr(skb);
+
+	iph->version = 6;
+
+	memcpy(iph->flow_lbl, XFRM_MODE_SKB_CB(skb)->flow_lbl,
+	       sizeof(iph->flow_lbl));
+	iph->nexthdr = XFRM_MODE_SKB_CB(skb)->protocol;
+
+	ipv6_change_dsfield(iph, 0, XFRM_MODE_SKB_CB(skb)->tos);
+	iph->hop_limit = XFRM_MODE_SKB_CB(skb)->ttl;
+}
+
+static inline void xfrm4_beet_make_header(struct sk_buff *skb)
+{
+	struct iphdr *iph = ip_hdr(skb);
+
+	iph->ihl = 5;
+	iph->version = 4;
+
+	iph->protocol = XFRM_MODE_SKB_CB(skb)->protocol;
+	iph->tos = XFRM_MODE_SKB_CB(skb)->tos;
+
+	iph->id = XFRM_MODE_SKB_CB(skb)->id;
+	iph->frag_off = XFRM_MODE_SKB_CB(skb)->frag_off;
+	iph->ttl = XFRM_MODE_SKB_CB(skb)->ttl;
+}
+
+#endif
diff --git a/net/xfrm/xfrm_input.c b/net/xfrm/xfrm_input.c
index b3b613660d44..314973aaa414 100644
--- a/net/xfrm/xfrm_input.c
+++ b/net/xfrm/xfrm_input.c
@@ -21,6 +21,8 @@
 #include <net/ip_tunnels.h>
 #include <net/ip6_tunnel.h>
 
+#include "xfrm_inout.h"
+
 struct xfrm_trans_tasklet {
 	struct tasklet_struct tasklet;
 	struct sk_buff_head queue;
@@ -166,35 +168,299 @@ int xfrm_parse_spi(struct sk_buff *skb, u8 nexthdr, __be32 *spi, __be32 *seq)
 }
 EXPORT_SYMBOL(xfrm_parse_spi);
 
-int xfrm_prepare_input(struct xfrm_state *x, struct sk_buff *skb)
+static int xfrm4_remove_beet_encap(struct xfrm_state *x, struct sk_buff *skb)
+{
+	struct iphdr *iph;
+	int optlen = 0;
+	int err = -EINVAL;
+
+	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)
 {
-	struct xfrm_mode *inner_mode = x->inner_mode;
+	int err = -EINVAL;
+
+	if (XFRM_MODE_SKB_CB(skb)->protocol != IPPROTO_IPIP)
+		goto out;
+
+	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;
+
+	if (XFRM_MODE_SKB_CB(skb)->protocol != IPPROTO_IPV6)
+		goto out;
+	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(ipv6_get_dsfield(ipv6_hdr(skb)),
+			       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;
 
-	err = x->outer_mode->afinfo->extract_input(x, skb);
+	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,
+			     const struct xfrm_mode *inner_mode,
+			     struct sk_buff *skb)
+{
+	switch (inner_mode->encap) {
+	case XFRM_MODE_BEET:
+		if (inner_mode->family == AF_INET)
+			return xfrm4_remove_beet_encap(x, skb);
+		if (inner_mode->family == AF_INET6)
+			return xfrm6_remove_beet_encap(x, skb);
+		break;
+	case XFRM_MODE_TUNNEL:
+		if (inner_mode->family == AF_INET)
+			return xfrm4_remove_tunnel_encap(x, skb);
+		if (inner_mode->family == AF_INET6)
+			return xfrm6_remove_tunnel_encap(x, skb);
+		break;
+	}
+
+	WARN_ON_ONCE(1);
+	return -EOPNOTSUPP;
+}
+
+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);
+
+	if (err) {
+		rcu_read_unlock();
 		return err;
+	}
 
 	if (x->sel.family == AF_UNSPEC) {
 		inner_mode = xfrm_ip2inner_mode(x, XFRM_MODE_SKB_CB(skb)->protocol);
-		if (inner_mode == NULL)
+		if (!inner_mode) {
+			rcu_read_unlock();
 			return -EAFNOSUPPORT;
+		}
 	}
 
-	skb->protocol = inner_mode->afinfo->eth_proto;
-	return inner_mode->input2(x, skb);
+	afinfo = xfrm_state_afinfo_get_rcu(inner_mode->family);
+	if (unlikely(!afinfo)) {
+		rcu_read_unlock();
+		return -EAFNOSUPPORT;
+	}
+
+	skb->protocol = afinfo->eth_proto;
+	rcu_read_unlock();
+	return xfrm_inner_mode_encap_remove(x, inner_mode, 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)
+{
+	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);
+		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)
+	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);
+		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,
+				 const struct xfrm_mode *inner_mode,
+				 struct sk_buff *skb)
+{
+	switch (inner_mode->encap) {
+	case XFRM_MODE_BEET:
+	case XFRM_MODE_TUNNEL:
+		return xfrm_prepare_input(x, skb);
+	case XFRM_MODE_TRANSPORT:
+		if (inner_mode->family == AF_INET)
+			return xfrm4_transport_input(x, skb);
+		if (inner_mode->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;
 }
-EXPORT_SYMBOL(xfrm_prepare_input);
 
 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);
+	const struct xfrm_mode *inner_mode;
 	int err;
 	__be32 seq;
 	__be32 seq_hi;
 	struct xfrm_state *x = NULL;
 	xfrm_address_t *daddr;
-	struct xfrm_mode *inner_mode;
 	u32 mark = skb->mark;
 	unsigned int family = AF_UNSPEC;
 	int decaps = 0;
@@ -216,7 +482,7 @@ int xfrm_input(struct sk_buff *skb, int nexthdr, __be32 spi, int encap_type)
 			goto drop;
 		}
 
-		family = x->outer_mode->afinfo->family;
+		family = x->outer_mode.family;
 
 		/* An encap_type of -1 indicates async resumption. */
 		if (encap_type == -1) {
@@ -400,7 +666,7 @@ resume:
 
 		XFRM_MODE_SKB_CB(skb)->protocol = nexthdr;
 
-		inner_mode = x->inner_mode;
+		inner_mode = &x->inner_mode;
 
 		if (x->sel.family == AF_UNSPEC) {
 			inner_mode = xfrm_ip2inner_mode(x, XFRM_MODE_SKB_CB(skb)->protocol);
@@ -410,12 +676,12 @@ resume:
 			}
 		}
 
-		if (inner_mode->input(x, skb)) {
+		if (xfrm_inner_mode_input(x, inner_mode, skb)) {
 			XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEMODEERROR);
 			goto drop;
 		}
 
-		if (x->outer_mode->flags & XFRM_MODE_FLAG_TUNNEL) {
+		if (x->outer_mode.flags & XFRM_MODE_FLAG_TUNNEL) {
 			decaps = 1;
 			break;
 		}
@@ -425,7 +691,7 @@ resume:
 		 * transport mode so the outer address is identical.
 		 */
 		daddr = &x->id.daddr;
-		family = x->outer_mode->afinfo->family;
+		family = x->outer_mode.family;
 
 		err = xfrm_parse_spi(skb, nexthdr, &spi, &seq);
 		if (err < 0) {
@@ -453,7 +719,12 @@ resume:
 		if (xo)
 			xfrm_gro = xo->flags & XFRM_GRO;
 
-		err = x->inner_mode->afinfo->transport_finish(skb, xfrm_gro || async);
+		err = -EAFNOSUPPORT;
+		rcu_read_lock();
+		afinfo = xfrm_state_afinfo_get_rcu(x->inner_mode.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)
diff --git a/net/xfrm/xfrm_interface.c b/net/xfrm/xfrm_interface.c
index dbb3c1945b5c..b9f118530db6 100644
--- a/net/xfrm/xfrm_interface.c
+++ b/net/xfrm/xfrm_interface.c
@@ -244,8 +244,8 @@ static void xfrmi_scrub_packet(struct sk_buff *skb, bool xnet)
 
 static int xfrmi_rcv_cb(struct sk_buff *skb, int err)
 {
+	const struct xfrm_mode *inner_mode;
 	struct pcpu_sw_netstats *tstats;
-	struct xfrm_mode *inner_mode;
 	struct net_device *dev;
 	struct xfrm_state *x;
 	struct xfrm_if *xi;
@@ -273,7 +273,7 @@ static int xfrmi_rcv_cb(struct sk_buff *skb, int err)
 	xnet = !net_eq(xi->net, dev_net(skb->dev));
 
 	if (xnet) {
-		inner_mode = x->inner_mode;
+		inner_mode = &x->inner_mode;
 
 		if (x->sel.family == AF_UNSPEC) {
 			inner_mode = xfrm_ip2inner_mode(x, XFRM_MODE_SKB_CB(skb)->protocol);
@@ -285,7 +285,7 @@ static int xfrmi_rcv_cb(struct sk_buff *skb, int err)
 		}
 
 		if (!xfrm_policy_check(NULL, XFRM_POLICY_IN, skb,
-				       inner_mode->afinfo->family))
+				       inner_mode->family))
 			return -EPERM;
 	}
 
diff --git a/net/xfrm/xfrm_output.c b/net/xfrm/xfrm_output.c
index 9333153bafda..a55510f9ff35 100644
--- a/net/xfrm/xfrm_output.c
+++ b/net/xfrm/xfrm_output.c
@@ -17,9 +17,13 @@
 #include <linux/slab.h>
 #include <linux/spinlock.h>
 #include <net/dst.h>
+#include <net/inet_ecn.h>
 #include <net/xfrm.h>
 
+#include "xfrm_inout.h"
+
 static int xfrm_output2(struct net *net, struct sock *sk, struct sk_buff *skb);
+static int xfrm_inner_extract_output(struct xfrm_state *x, struct sk_buff *skb);
 
 static int xfrm_skb_check_space(struct sk_buff *skb)
 {
@@ -50,6 +54,360 @@ static struct dst_entry *skb_dst_pop(struct sk_buff *skb)
 	return child;
 }
 
+/* Add encapsulation header.
+ *
+ * The IP header will be moved forward to make space for the encapsulation
+ * header.
+ */
+static int xfrm4_transport_output(struct xfrm_state *x, struct sk_buff *skb)
+{
+	struct iphdr *iph = ip_hdr(skb);
+	int ihl = iph->ihl * 4;
+
+	skb_set_inner_transport_header(skb, skb_transport_offset(skb));
+
+	skb_set_network_header(skb, -x->props.header_len);
+	skb->mac_header = skb->network_header +
+			  offsetof(struct iphdr, protocol);
+	skb->transport_header = skb->network_header + ihl;
+	__skb_pull(skb, ihl);
+	memmove(skb_network_header(skb), iph, ihl);
+	return 0;
+}
+
+/* Add encapsulation header.
+ *
+ * The IP header and mutable extension headers will be moved forward to make
+ * space for the encapsulation header.
+ */
+static int xfrm6_transport_output(struct xfrm_state *x, struct sk_buff *skb)
+{
+#if IS_ENABLED(CONFIG_IPV6)
+	struct ipv6hdr *iph;
+	u8 *prevhdr;
+	int hdr_len;
+
+	iph = ipv6_hdr(skb);
+	skb_set_inner_transport_header(skb, skb_transport_offset(skb));
+
+	hdr_len = x->type->hdr_offset(x, skb, &prevhdr);
+	if (hdr_len < 0)
+		return hdr_len;
+	skb_set_mac_header(skb,
+			   (prevhdr - x->props.header_len) - skb->data);
+	skb_set_network_header(skb, -x->props.header_len);
+	skb->transport_header = skb->network_header + hdr_len;
+	__skb_pull(skb, hdr_len);
+	memmove(ipv6_hdr(skb), iph, hdr_len);
+	return 0;
+#else
+	WARN_ON_ONCE(1);
+	return -EAFNOSUPPORT;
+#endif
+}
+
+/* Add route optimization header space.
+ *
+ * The IP header and mutable extension headers will be moved forward to make
+ * space for the route optimization header.
+ */
+static int xfrm6_ro_output(struct xfrm_state *x, struct sk_buff *skb)
+{
+#if IS_ENABLED(CONFIG_IPV6)
+	struct ipv6hdr *iph;
+	u8 *prevhdr;
+	int hdr_len;
+
+	iph = ipv6_hdr(skb);
+
+	hdr_len = x->type->hdr_offset(x, skb, &prevhdr);
+	if (hdr_len < 0)
+		return hdr_len;
+	skb_set_mac_header(skb,
+			   (prevhdr - x->props.header_len) - skb->data);
+	skb_set_network_header(skb, -x->props.header_len);
+	skb->transport_header = skb->network_header + hdr_len;
+	__skb_pull(skb, hdr_len);
+	memmove(ipv6_hdr(skb), iph, hdr_len);
+
+	x->lastused = ktime_get_real_seconds();
+
+	return 0;
+#else
+	WARN_ON_ONCE(1);
+	return -EAFNOSUPPORT;
+#endif
+}
+
+/* Add encapsulation header.
+ *
+ * The top IP header will be constructed per draft-nikander-esp-beet-mode-06.txt.
+ */
+static int xfrm4_beet_encap_add(struct xfrm_state *x, struct sk_buff *skb)
+{
+	struct ip_beet_phdr *ph;
+	struct iphdr *top_iph;
+	int hdrlen, optlen;
+
+	hdrlen = 0;
+	optlen = XFRM_MODE_SKB_CB(skb)->optlen;
+	if (unlikely(optlen))
+		hdrlen += IPV4_BEET_PHMAXLEN - (optlen & 4);
+
+	skb_set_network_header(skb, -x->props.header_len - hdrlen +
+			       (XFRM_MODE_SKB_CB(skb)->ihl - sizeof(*top_iph)));
+	if (x->sel.family != AF_INET6)
+		skb->network_header += IPV4_BEET_PHMAXLEN;
+	skb->mac_header = skb->network_header +
+			  offsetof(struct iphdr, protocol);
+	skb->transport_header = skb->network_header + sizeof(*top_iph);
+
+	xfrm4_beet_make_header(skb);
+
+	ph = __skb_pull(skb, XFRM_MODE_SKB_CB(skb)->ihl - hdrlen);
+
+	top_iph = ip_hdr(skb);
+
+	if (unlikely(optlen)) {
+		if (WARN_ON(optlen < 0))
+			return -EINVAL;
+
+		ph->padlen = 4 - (optlen & 4);
+		ph->hdrlen = optlen / 8;
+		ph->nexthdr = top_iph->protocol;
+		if (ph->padlen)
+			memset(ph + 1, IPOPT_NOP, ph->padlen);
+
+		top_iph->protocol = IPPROTO_BEETPH;
+		top_iph->ihl = sizeof(struct iphdr) / 4;
+	}
+
+	top_iph->saddr = x->props.saddr.a4;
+	top_iph->daddr = x->id.daddr.a4;
+
+	return 0;
+}
+
+/* Add encapsulation header.
+ *
+ * The top IP header will be constructed per RFC 2401.
+ */
+static int xfrm4_tunnel_encap_add(struct xfrm_state *x, struct sk_buff *skb)
+{
+	struct dst_entry *dst = skb_dst(skb);
+	struct iphdr *top_iph;
+	int flags;
+
+	skb_set_inner_network_header(skb, skb_network_offset(skb));
+	skb_set_inner_transport_header(skb, skb_transport_offset(skb));
+
+	skb_set_network_header(skb, -x->props.header_len);
+	skb->mac_header = skb->network_header +
+			  offsetof(struct iphdr, protocol);
+	skb->transport_header = skb->network_header + sizeof(*top_iph);
+	top_iph = ip_hdr(skb);
+
+	top_iph->ihl = 5;
+	top_iph->version = 4;
+
+	top_iph->protocol = xfrm_af2proto(skb_dst(skb)->ops->family);
+
+	/* DS disclosing depends on XFRM_SA_XFLAG_DONT_ENCAP_DSCP */
+	if (x->props.extra_flags & XFRM_SA_XFLAG_DONT_ENCAP_DSCP)
+		top_iph->tos = 0;
+	else
+		top_iph->tos = XFRM_MODE_SKB_CB(skb)->tos;
+	top_iph->tos = INET_ECN_encapsulate(top_iph->tos,
+					    XFRM_MODE_SKB_CB(skb)->tos);
+
+	flags = x->props.flags;
+	if (flags & XFRM_STATE_NOECN)
+		IP_ECN_clear(top_iph);
+
+	top_iph->frag_off = (flags & XFRM_STATE_NOPMTUDISC) ?
+		0 : (XFRM_MODE_SKB_CB(skb)->frag_off & htons(IP_DF));
+
+	top_iph->ttl = ip4_dst_hoplimit(xfrm_dst_child(dst));
+
+	top_iph->saddr = x->props.saddr.a4;
+	top_iph->daddr = x->id.daddr.a4;
+	ip_select_ident(dev_net(dst->dev), skb, NULL);
+
+	return 0;
+}
+
+#if IS_ENABLED(CONFIG_IPV6)
+static int xfrm6_tunnel_encap_add(struct xfrm_state *x, struct sk_buff *skb)
+{
+	struct dst_entry *dst = skb_dst(skb);
+	struct ipv6hdr *top_iph;
+	int dsfield;
+
+	skb_set_inner_network_header(skb, skb_network_offset(skb));
+	skb_set_inner_transport_header(skb, skb_transport_offset(skb));
+
+	skb_set_network_header(skb, -x->props.header_len);
+	skb->mac_header = skb->network_header +
+			  offsetof(struct ipv6hdr, nexthdr);
+	skb->transport_header = skb->network_header + sizeof(*top_iph);
+	top_iph = ipv6_hdr(skb);
+
+	top_iph->version = 6;
+
+	memcpy(top_iph->flow_lbl, XFRM_MODE_SKB_CB(skb)->flow_lbl,
+	       sizeof(top_iph->flow_lbl));
+	top_iph->nexthdr = xfrm_af2proto(skb_dst(skb)->ops->family);
+
+	if (x->props.extra_flags & XFRM_SA_XFLAG_DONT_ENCAP_DSCP)
+		dsfield = 0;
+	else
+		dsfield = XFRM_MODE_SKB_CB(skb)->tos;
+	dsfield = INET_ECN_encapsulate(dsfield, XFRM_MODE_SKB_CB(skb)->tos);
+	if (x->props.flags & XFRM_STATE_NOECN)
+		dsfield &= ~INET_ECN_MASK;
+	ipv6_change_dsfield(top_iph, 0, dsfield);
+	top_iph->hop_limit = ip6_dst_hoplimit(xfrm_dst_child(dst));
+	top_iph->saddr = *(struct in6_addr *)&x->props.saddr;
+	top_iph->daddr = *(struct in6_addr *)&x->id.daddr;
+	return 0;
+}
+
+static int xfrm6_beet_encap_add(struct xfrm_state *x, struct sk_buff *skb)
+{
+	struct ipv6hdr *top_iph;
+	struct ip_beet_phdr *ph;
+	int optlen, hdr_len;
+
+	hdr_len = 0;
+	optlen = XFRM_MODE_SKB_CB(skb)->optlen;
+	if (unlikely(optlen))
+		hdr_len += IPV4_BEET_PHMAXLEN - (optlen & 4);
+
+	skb_set_network_header(skb, -x->props.header_len - hdr_len);
+	if (x->sel.family != AF_INET6)
+		skb->network_header += IPV4_BEET_PHMAXLEN;
+	skb->mac_header = skb->network_header +
+			  offsetof(struct ipv6hdr, nexthdr);
+	skb->transport_header = skb->network_header + sizeof(*top_iph);
+	ph = __skb_pull(skb, XFRM_MODE_SKB_CB(skb)->ihl - hdr_len);
+
+	xfrm6_beet_make_header(skb);
+
+	top_iph = ipv6_hdr(skb);
+	if (unlikely(optlen)) {
+		if (WARN_ON(optlen < 0))
+			return -EINVAL;
+
+		ph->padlen = 4 - (optlen & 4);
+		ph->hdrlen = optlen / 8;
+		ph->nexthdr = top_iph->nexthdr;
+		if (ph->padlen)
+			memset(ph + 1, IPOPT_NOP, ph->padlen);
+
+		top_iph->nexthdr = IPPROTO_BEETPH;
+	}
+
+	top_iph->saddr = *(struct in6_addr *)&x->props.saddr;
+	top_iph->daddr = *(struct in6_addr *)&x->id.daddr;
+	return 0;
+}
+#endif
+
+/* Add encapsulation header.
+ *
+ * On exit, the transport header will be set to the start of the
+ * encapsulation header to be filled in by x->type->output and the mac
+ * header will be set to the nextheader (protocol for IPv4) field of the
+ * extension header directly preceding the encapsulation header, or in
+ * its absence, that of the top IP header.
+ * The value of the network header will always point to the top IP header
+ * while skb->data will point to the payload.
+ */
+static int xfrm4_prepare_output(struct xfrm_state *x, struct sk_buff *skb)
+{
+	int err;
+
+	err = xfrm_inner_extract_output(x, skb);
+	if (err)
+		return err;
+
+	IPCB(skb)->flags |= IPSKB_XFRM_TUNNEL_SIZE;
+	skb->protocol = htons(ETH_P_IP);
+
+	switch (x->outer_mode.encap) {
+	case XFRM_MODE_BEET:
+		return xfrm4_beet_encap_add(x, skb);
+	case XFRM_MODE_TUNNEL:
+		return xfrm4_tunnel_encap_add(x, skb);
+	}
+
+	WARN_ON_ONCE(1);
+	return -EOPNOTSUPP;
+}
+
+static int xfrm6_prepare_output(struct xfrm_state *x, struct sk_buff *skb)
+{
+#if IS_ENABLED(CONFIG_IPV6)
+	int err;
+
+	err = xfrm_inner_extract_output(x, skb);
+	if (err)
+		return err;
+
+	skb->ignore_df = 1;
+	skb->protocol = htons(ETH_P_IPV6);
+
+	switch (x->outer_mode.encap) {
+	case XFRM_MODE_BEET:
+		return xfrm6_beet_encap_add(x, skb);
+	case XFRM_MODE_TUNNEL:
+		return xfrm6_tunnel_encap_add(x, skb);
+	default:
+		WARN_ON_ONCE(1);
+		return -EOPNOTSUPP;
+	}
+#endif
+	WARN_ON_ONCE(1);
+	return -EAFNOSUPPORT;
+}
+
+static int xfrm_outer_mode_output(struct xfrm_state *x, struct sk_buff *skb)
+{
+	switch (x->outer_mode.encap) {
+	case XFRM_MODE_BEET:
+	case XFRM_MODE_TUNNEL:
+		if (x->outer_mode.family == AF_INET)
+			return xfrm4_prepare_output(x, skb);
+		if (x->outer_mode.family == AF_INET6)
+			return xfrm6_prepare_output(x, skb);
+		break;
+	case XFRM_MODE_TRANSPORT:
+		if (x->outer_mode.family == AF_INET)
+			return xfrm4_transport_output(x, skb);
+		if (x->outer_mode.family == AF_INET6)
+			return xfrm6_transport_output(x, skb);
+		break;
+	case XFRM_MODE_ROUTEOPTIMIZATION:
+		if (x->outer_mode.family == AF_INET6)
+			return xfrm6_ro_output(x, skb);
+		WARN_ON_ONCE(1);
+		break;
+	default:
+		WARN_ON_ONCE(1);
+		break;
+	}
+
+	return -EOPNOTSUPP;
+}
+
+#if IS_ENABLED(CONFIG_NET_PKTGEN)
+int pktgen_xfrm_outer_mode_output(struct xfrm_state *x, struct sk_buff *skb)
+{
+	return xfrm_outer_mode_output(x, skb);
+}
+EXPORT_SYMBOL_GPL(pktgen_xfrm_outer_mode_output);
+#endif
+
 static int xfrm_output_one(struct sk_buff *skb, int err)
 {
 	struct dst_entry *dst = skb_dst(skb);
@@ -68,7 +426,7 @@ static int xfrm_output_one(struct sk_buff *skb, int err)
 
 		skb->mark = xfrm_smark_get(skb->mark, x);
 
-		err = x->outer_mode->output(x, skb);
+		err = xfrm_outer_mode_output(x, skb);
 		if (err) {
 			XFRM_INC_STATS(net, LINUX_MIB_XFRMOUTSTATEMODEERROR);
 			goto error_nolock;
@@ -131,7 +489,7 @@ resume:
 		}
 		skb_dst_set(skb, dst);
 		x = dst->xfrm;
-	} while (x && !(x->outer_mode->flags & XFRM_MODE_FLAG_TUNNEL));
+	} while (x && !(x->outer_mode.flags & XFRM_MODE_FLAG_TUNNEL));
 
 	return 0;
 
@@ -258,20 +616,29 @@ out:
 }
 EXPORT_SYMBOL_GPL(xfrm_output);
 
-int xfrm_inner_extract_output(struct xfrm_state *x, struct sk_buff *skb)
+static int xfrm_inner_extract_output(struct xfrm_state *x, struct sk_buff *skb)
 {
-	struct xfrm_mode *inner_mode;
+	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,
 				xfrm_af2proto(skb_dst(skb)->ops->family));
 	else
-		inner_mode = x->inner_mode;
+		inner_mode = &x->inner_mode;
 
 	if (inner_mode == NULL)
 		return -EAFNOSUPPORT;
-	return inner_mode->afinfo->extract_output(x, skb);
+
+	rcu_read_lock();
+	afinfo = xfrm_state_afinfo_get_rcu(inner_mode->family);
+	if (likely(afinfo))
+		err = afinfo->extract_output(x, skb);
+	rcu_read_unlock();
+
+	return err;
 }
-EXPORT_SYMBOL_GPL(xfrm_inner_extract_output);
 
 void xfrm_local_error(struct sk_buff *skb, int mtu)
 {
diff --git a/net/xfrm/xfrm_policy.c b/net/xfrm/xfrm_policy.c
index 8d1a898d0ba5..03b6bf85d70b 100644
--- a/net/xfrm/xfrm_policy.c
+++ b/net/xfrm/xfrm_policy.c
@@ -27,10 +27,14 @@
 #include <linux/cpu.h>
 #include <linux/audit.h>
 #include <linux/rhashtable.h>
+#include <linux/if_tunnel.h>
 #include <net/dst.h>
 #include <net/flow.h>
 #include <net/xfrm.h>
 #include <net/ip.h>
+#if IS_ENABLED(CONFIG_IPV6_MIP6)
+#include <net/mip6.h>
+#endif
 #ifdef CONFIG_XFRM_STATISTICS
 #include <net/snmp.h>
 #endif
@@ -2450,18 +2454,10 @@ xfrm_tmpl_resolve(struct xfrm_policy **pols, int npols, const struct flowi *fl,
 
 static int xfrm_get_tos(const struct flowi *fl, int family)
 {
-	const struct xfrm_policy_afinfo *afinfo;
-	int tos;
+	if (family == AF_INET)
+		return IPTOS_RT_MASK & fl->u.ip4.flowi4_tos;
 
-	afinfo = xfrm_policy_get_afinfo(family);
-	if (!afinfo)
-		return 0;
-
-	tos = afinfo->get_tos(fl);
-
-	rcu_read_unlock();
-
-	return tos;
+	return 0;
 }
 
 static inline struct xfrm_dst *xfrm_alloc_dst(struct net *net, int family)
@@ -2499,21 +2495,14 @@ static inline struct xfrm_dst *xfrm_alloc_dst(struct net *net, int family)
 	return xdst;
 }
 
-static inline int xfrm_init_path(struct xfrm_dst *path, struct dst_entry *dst,
-				 int nfheader_len)
+static void xfrm_init_path(struct xfrm_dst *path, struct dst_entry *dst,
+			   int nfheader_len)
 {
-	const struct xfrm_policy_afinfo *afinfo =
-		xfrm_policy_get_afinfo(dst->ops->family);
-	int err;
-
-	if (!afinfo)
-		return -EINVAL;
-
-	err = afinfo->init_path(path, dst, nfheader_len);
-
-	rcu_read_unlock();
-
-	return err;
+	if (dst->ops->family == AF_INET6) {
+		struct rt6_info *rt = (struct rt6_info *)dst;
+		path->path_cookie = rt6_get_cookie(rt);
+		path->u.rt6.rt6i_nfheader_len = nfheader_len;
+	}
 }
 
 static inline int xfrm_fill_dst(struct xfrm_dst *xdst, struct net_device *dev,
@@ -2545,10 +2534,11 @@ static struct dst_entry *xfrm_bundle_create(struct xfrm_policy *policy,
 					    const struct flowi *fl,
 					    struct dst_entry *dst)
 {
+	const struct xfrm_state_afinfo *afinfo;
+	const struct xfrm_mode *inner_mode;
 	struct net *net = xp_net(policy);
 	unsigned long now = jiffies;
 	struct net_device *dev;
-	struct xfrm_mode *inner_mode;
 	struct xfrm_dst *xdst_prev = NULL;
 	struct xfrm_dst *xdst0 = NULL;
 	int i = 0;
@@ -2594,7 +2584,7 @@ static struct dst_entry *xfrm_bundle_create(struct xfrm_policy *policy,
 				goto put_states;
 			}
 		} else
-			inner_mode = xfrm[i]->inner_mode;
+			inner_mode = &xfrm[i]->inner_mode;
 
 		xdst->route = dst;
 		dst_copy_metrics(dst1, dst);
@@ -2622,7 +2612,14 @@ static struct dst_entry *xfrm_bundle_create(struct xfrm_policy *policy,
 		dst1->lastuse = now;
 
 		dst1->input = dst_discard;
-		dst1->output = inner_mode->afinfo->output;
+
+		rcu_read_lock();
+		afinfo = xfrm_state_afinfo_get_rcu(inner_mode->family);
+		if (likely(afinfo))
+			dst1->output = afinfo->output;
+		else
+			dst1->output = dst_discard_out;
+		rcu_read_unlock();
 
 		xdst_prev = xdst;
 
@@ -3263,20 +3260,229 @@ xfrm_policy_ok(const struct xfrm_tmpl *tmpl, const struct sec_path *sp, int star
 	return start;
 }
 
+static void
+decode_session4(struct sk_buff *skb, struct flowi *fl, bool reverse)
+{
+	const struct iphdr *iph = ip_hdr(skb);
+	u8 *xprth = skb_network_header(skb) + iph->ihl * 4;
+	struct flowi4 *fl4 = &fl->u.ip4;
+	int oif = 0;
+
+	if (skb_dst(skb))
+		oif = skb_dst(skb)->dev->ifindex;
+
+	memset(fl4, 0, sizeof(struct flowi4));
+	fl4->flowi4_mark = skb->mark;
+	fl4->flowi4_oif = reverse ? skb->skb_iif : oif;
+
+	if (!ip_is_fragment(iph)) {
+		switch (iph->protocol) {
+		case IPPROTO_UDP:
+		case IPPROTO_UDPLITE:
+		case IPPROTO_TCP:
+		case IPPROTO_SCTP:
+		case IPPROTO_DCCP:
+			if (xprth + 4 < skb->data ||
+			    pskb_may_pull(skb, xprth + 4 - skb->data)) {
+				__be16 *ports;
+
+				xprth = skb_network_header(skb) + iph->ihl * 4;
+				ports = (__be16 *)xprth;
+
+				fl4->fl4_sport = ports[!!reverse];
+				fl4->fl4_dport = ports[!reverse];
+			}
+			break;
+		case IPPROTO_ICMP:
+			if (xprth + 2 < skb->data ||
+			    pskb_may_pull(skb, xprth + 2 - skb->data)) {
+				u8 *icmp;
+
+				xprth = skb_network_header(skb) + iph->ihl * 4;
+				icmp = xprth;
+
+				fl4->fl4_icmp_type = icmp[0];
+				fl4->fl4_icmp_code = icmp[1];
+			}
+			break;
+		case IPPROTO_ESP:
+			if (xprth + 4 < skb->data ||
+			    pskb_may_pull(skb, xprth + 4 - skb->data)) {
+				__be32 *ehdr;
+
+				xprth = skb_network_header(skb) + iph->ihl * 4;
+				ehdr = (__be32 *)xprth;
+
+				fl4->fl4_ipsec_spi = ehdr[0];
+			}
+			break;
+		case IPPROTO_AH:
+			if (xprth + 8 < skb->data ||
+			    pskb_may_pull(skb, xprth + 8 - skb->data)) {
+				__be32 *ah_hdr;
+
+				xprth = skb_network_header(skb) + iph->ihl * 4;
+				ah_hdr = (__be32 *)xprth;
+
+				fl4->fl4_ipsec_spi = ah_hdr[1];
+			}
+			break;
+		case IPPROTO_COMP:
+			if (xprth + 4 < skb->data ||
+			    pskb_may_pull(skb, xprth + 4 - skb->data)) {
+				__be16 *ipcomp_hdr;
+
+				xprth = skb_network_header(skb) + iph->ihl * 4;
+				ipcomp_hdr = (__be16 *)xprth;
+
+				fl4->fl4_ipsec_spi = htonl(ntohs(ipcomp_hdr[1]));
+			}
+			break;
+		case IPPROTO_GRE:
+			if (xprth + 12 < skb->data ||
+			    pskb_may_pull(skb, xprth + 12 - skb->data)) {
+				__be16 *greflags;
+				__be32 *gre_hdr;
+
+				xprth = skb_network_header(skb) + iph->ihl * 4;
+				greflags = (__be16 *)xprth;
+				gre_hdr = (__be32 *)xprth;
+
+				if (greflags[0] & GRE_KEY) {
+					if (greflags[0] & GRE_CSUM)
+						gre_hdr++;
+					fl4->fl4_gre_key = gre_hdr[1];
+				}
+			}
+			break;
+		default:
+			fl4->fl4_ipsec_spi = 0;
+			break;
+		}
+	}
+	fl4->flowi4_proto = iph->protocol;
+	fl4->daddr = reverse ? iph->saddr : iph->daddr;
+	fl4->saddr = reverse ? iph->daddr : iph->saddr;
+	fl4->flowi4_tos = iph->tos;
+}
+
+#if IS_ENABLED(CONFIG_IPV6)
+static void
+decode_session6(struct sk_buff *skb, struct flowi *fl, bool reverse)
+{
+	struct flowi6 *fl6 = &fl->u.ip6;
+	int onlyproto = 0;
+	const struct ipv6hdr *hdr = ipv6_hdr(skb);
+	u32 offset = sizeof(*hdr);
+	struct ipv6_opt_hdr *exthdr;
+	const unsigned char *nh = skb_network_header(skb);
+	u16 nhoff = IP6CB(skb)->nhoff;
+	int oif = 0;
+	u8 nexthdr;
+
+	if (!nhoff)
+		nhoff = offsetof(struct ipv6hdr, nexthdr);
+
+	nexthdr = nh[nhoff];
+
+	if (skb_dst(skb))
+		oif = skb_dst(skb)->dev->ifindex;
+
+	memset(fl6, 0, sizeof(struct flowi6));
+	fl6->flowi6_mark = skb->mark;
+	fl6->flowi6_oif = reverse ? skb->skb_iif : oif;
+
+	fl6->daddr = reverse ? hdr->saddr : hdr->daddr;
+	fl6->saddr = reverse ? hdr->daddr : hdr->saddr;
+
+	while (nh + offset + sizeof(*exthdr) < skb->data ||
+	       pskb_may_pull(skb, nh + offset + sizeof(*exthdr) - skb->data)) {
+		nh = skb_network_header(skb);
+		exthdr = (struct ipv6_opt_hdr *)(nh + offset);
+
+		switch (nexthdr) {
+		case NEXTHDR_FRAGMENT:
+			onlyproto = 1;
+			/* fall through */
+		case NEXTHDR_ROUTING:
+		case NEXTHDR_HOP:
+		case NEXTHDR_DEST:
+			offset += ipv6_optlen(exthdr);
+			nexthdr = exthdr->nexthdr;
+			exthdr = (struct ipv6_opt_hdr *)(nh + offset);
+			break;
+		case IPPROTO_UDP:
+		case IPPROTO_UDPLITE:
+		case IPPROTO_TCP:
+		case IPPROTO_SCTP:
+		case IPPROTO_DCCP:
+			if (!onlyproto && (nh + offset + 4 < skb->data ||
+			     pskb_may_pull(skb, nh + offset + 4 - skb->data))) {
+				__be16 *ports;
+
+				nh = skb_network_header(skb);
+				ports = (__be16 *)(nh + offset);
+				fl6->fl6_sport = ports[!!reverse];
+				fl6->fl6_dport = ports[!reverse];
+			}
+			fl6->flowi6_proto = nexthdr;
+			return;
+		case IPPROTO_ICMPV6:
+			if (!onlyproto && (nh + offset + 2 < skb->data ||
+			    pskb_may_pull(skb, nh + offset + 2 - skb->data))) {
+				u8 *icmp;
+
+				nh = skb_network_header(skb);
+				icmp = (u8 *)(nh + offset);
+				fl6->fl6_icmp_type = icmp[0];
+				fl6->fl6_icmp_code = icmp[1];
+			}
+			fl6->flowi6_proto = nexthdr;
+			return;
+#if IS_ENABLED(CONFIG_IPV6_MIP6)
+		case IPPROTO_MH:
+			offset += ipv6_optlen(exthdr);
+			if (!onlyproto && (nh + offset + 3 < skb->data ||
+			    pskb_may_pull(skb, nh + offset + 3 - skb->data))) {
+				struct ip6_mh *mh;
+
+				nh = skb_network_header(skb);
+				mh = (struct ip6_mh *)(nh + offset);
+				fl6->fl6_mh_type = mh->ip6mh_type;
+			}
+			fl6->flowi6_proto = nexthdr;
+			return;
+#endif
+		/* XXX Why are there these headers? */
+		case IPPROTO_AH:
+		case IPPROTO_ESP:
+		case IPPROTO_COMP:
+		default:
+			fl6->fl6_ipsec_spi = 0;
+			fl6->flowi6_proto = nexthdr;
+			return;
+		}
+	}
+}
+#endif
+
 int __xfrm_decode_session(struct sk_buff *skb, struct flowi *fl,
 			  unsigned int family, int reverse)
 {
-	const struct xfrm_policy_afinfo *afinfo = xfrm_policy_get_afinfo(family);
-	int err;
-
-	if (unlikely(afinfo == NULL))
+	switch (family) {
+	case AF_INET:
+		decode_session4(skb, fl, reverse);
+		break;
+#if IS_ENABLED(CONFIG_IPV6)
+	case AF_INET6:
+		decode_session6(skb, fl, reverse);
+		break;
+#endif
+	default:
 		return -EAFNOSUPPORT;
+	}
 
-	afinfo->decode_session(skb, fl, reverse);
-
-	err = security_xfrm_decode_session(skb, &fl->flowi_secid);
-	rcu_read_unlock();
-	return err;
+	return security_xfrm_decode_session(skb, &fl->flowi_secid);
 }
 EXPORT_SYMBOL(__xfrm_decode_session);
 
diff --git a/net/xfrm/xfrm_state.c b/net/xfrm/xfrm_state.c
index 1bb971f46fc6..ed25eb81aabe 100644
--- a/net/xfrm/xfrm_state.c
+++ b/net/xfrm/xfrm_state.c
@@ -173,7 +173,7 @@ static DEFINE_SPINLOCK(xfrm_state_gc_lock);
 int __xfrm_state_delete(struct xfrm_state *x);
 
 int km_query(struct xfrm_state *x, struct xfrm_tmpl *t, struct xfrm_policy *pol);
-bool km_is_alive(const struct km_event *c);
+static bool km_is_alive(const struct km_event *c);
 void km_state_expired(struct xfrm_state *x, int hard, u32 portid);
 
 static DEFINE_SPINLOCK(xfrm_type_lock);
@@ -330,100 +330,67 @@ static void xfrm_put_type_offload(const struct xfrm_type_offload *type)
 	module_put(type->owner);
 }
 
-static DEFINE_SPINLOCK(xfrm_mode_lock);
-int xfrm_register_mode(struct xfrm_mode *mode, int family)
-{
-	struct xfrm_state_afinfo *afinfo;
-	struct xfrm_mode **modemap;
-	int err;
-
-	if (unlikely(mode->encap >= XFRM_MODE_MAX))
-		return -EINVAL;
-
-	afinfo = xfrm_state_get_afinfo(family);
-	if (unlikely(afinfo == NULL))
-		return -EAFNOSUPPORT;
-
-	err = -EEXIST;
-	modemap = afinfo->mode_map;
-	spin_lock_bh(&xfrm_mode_lock);
-	if (modemap[mode->encap])
-		goto out;
-
-	err = -ENOENT;
-	if (!try_module_get(afinfo->owner))
-		goto out;
-
-	mode->afinfo = afinfo;
-	modemap[mode->encap] = mode;
-	err = 0;
-
-out:
-	spin_unlock_bh(&xfrm_mode_lock);
-	rcu_read_unlock();
-	return err;
-}
-EXPORT_SYMBOL(xfrm_register_mode);
-
-int xfrm_unregister_mode(struct xfrm_mode *mode, int family)
-{
-	struct xfrm_state_afinfo *afinfo;
-	struct xfrm_mode **modemap;
-	int err;
-
-	if (unlikely(mode->encap >= XFRM_MODE_MAX))
-		return -EINVAL;
-
-	afinfo = xfrm_state_get_afinfo(family);
-	if (unlikely(afinfo == NULL))
-		return -EAFNOSUPPORT;
-
-	err = -ENOENT;
-	modemap = afinfo->mode_map;
-	spin_lock_bh(&xfrm_mode_lock);
-	if (likely(modemap[mode->encap] == mode)) {
-		modemap[mode->encap] = NULL;
-		module_put(mode->afinfo->owner);
-		err = 0;
-	}
-
-	spin_unlock_bh(&xfrm_mode_lock);
-	rcu_read_unlock();
-	return err;
-}
-EXPORT_SYMBOL(xfrm_unregister_mode);
-
-static struct xfrm_mode *xfrm_get_mode(unsigned int encap, int family)
-{
-	struct xfrm_state_afinfo *afinfo;
-	struct xfrm_mode *mode;
-	int modload_attempted = 0;
+static const struct xfrm_mode xfrm4_mode_map[XFRM_MODE_MAX] = {
+	[XFRM_MODE_BEET] = {
+		.encap = XFRM_MODE_BEET,
+		.flags = XFRM_MODE_FLAG_TUNNEL,
+		.family = AF_INET,
+	},
+	[XFRM_MODE_TRANSPORT] = {
+		.encap = XFRM_MODE_TRANSPORT,
+		.family = AF_INET,
+	},
+	[XFRM_MODE_TUNNEL] = {
+		.encap = XFRM_MODE_TUNNEL,
+		.flags = XFRM_MODE_FLAG_TUNNEL,
+		.family = AF_INET,
+	},
+};
+
+static const struct xfrm_mode xfrm6_mode_map[XFRM_MODE_MAX] = {
+	[XFRM_MODE_BEET] = {
+		.encap = XFRM_MODE_BEET,
+		.flags = XFRM_MODE_FLAG_TUNNEL,
+		.family = AF_INET6,
+	},
+	[XFRM_MODE_ROUTEOPTIMIZATION] = {
+		.encap = XFRM_MODE_ROUTEOPTIMIZATION,
+		.family = AF_INET6,
+	},
+	[XFRM_MODE_TRANSPORT] = {
+		.encap = XFRM_MODE_TRANSPORT,
+		.family = AF_INET6,
+	},
+	[XFRM_MODE_TUNNEL] = {
+		.encap = XFRM_MODE_TUNNEL,
+		.flags = XFRM_MODE_FLAG_TUNNEL,
+		.family = AF_INET6,
+	},
+};
+
+static const struct xfrm_mode *xfrm_get_mode(unsigned int encap, int family)
+{
+	const struct xfrm_mode *mode;
 
 	if (unlikely(encap >= XFRM_MODE_MAX))
 		return NULL;
 
-retry:
-	afinfo = xfrm_state_get_afinfo(family);
-	if (unlikely(afinfo == NULL))
-		return NULL;
-
-	mode = READ_ONCE(afinfo->mode_map[encap]);
-	if (unlikely(mode && !try_module_get(mode->owner)))
-		mode = NULL;
-
-	rcu_read_unlock();
-	if (!mode && !modload_attempted) {
-		request_module("xfrm-mode-%d-%d", family, encap);
-		modload_attempted = 1;
-		goto retry;
+	switch (family) {
+	case AF_INET:
+		mode = &xfrm4_mode_map[encap];
+		if (mode->family == family)
+			return mode;
+		break;
+	case AF_INET6:
+		mode = &xfrm6_mode_map[encap];
+		if (mode->family == family)
+			return mode;
+		break;
+	default:
+		break;
 	}
 
-	return mode;
-}
-
-static void xfrm_put_mode(struct xfrm_mode *mode)
-{
-	module_put(mode->owner);
+	return NULL;
 }
 
 void xfrm_state_free(struct xfrm_state *x)
@@ -444,12 +411,6 @@ static void ___xfrm_state_destroy(struct xfrm_state *x)
 	kfree(x->coaddr);
 	kfree(x->replay_esn);
 	kfree(x->preplay_esn);
-	if (x->inner_mode)
-		xfrm_put_mode(x->inner_mode);
-	if (x->inner_mode_iaf)
-		xfrm_put_mode(x->inner_mode_iaf);
-	if (x->outer_mode)
-		xfrm_put_mode(x->outer_mode);
 	if (x->type_offload)
 		xfrm_put_type_offload(x->type_offload);
 	if (x->type) {
@@ -590,8 +551,6 @@ struct xfrm_state *xfrm_state_alloc(struct net *net)
 		x->lft.hard_packet_limit = XFRM_INF;
 		x->replay_maxage = 0;
 		x->replay_maxdiff = 0;
-		x->inner_mode = NULL;
-		x->inner_mode_iaf = NULL;
 		spin_lock_init(&x->lock);
 	}
 	return x;
@@ -2066,7 +2025,7 @@ int km_report(struct net *net, u8 proto, struct xfrm_selector *sel, xfrm_address
 }
 EXPORT_SYMBOL(km_report);
 
-bool km_is_alive(const struct km_event *c)
+static bool km_is_alive(const struct km_event *c)
 {
 	struct xfrm_mgr *km;
 	bool is_alive = false;
@@ -2082,7 +2041,6 @@ bool km_is_alive(const struct km_event *c)
 
 	return is_alive;
 }
-EXPORT_SYMBOL(km_is_alive);
 
 int xfrm_user_policy(struct sock *sk, int optname, u8 __user *optval, int optlen)
 {
@@ -2195,6 +2153,7 @@ struct xfrm_state_afinfo *xfrm_state_afinfo_get_rcu(unsigned int family)
 
 	return rcu_dereference(xfrm_state_afinfo[family]);
 }
+EXPORT_SYMBOL_GPL(xfrm_state_afinfo_get_rcu);
 
 struct xfrm_state_afinfo *xfrm_state_get_afinfo(unsigned int family)
 {
@@ -2242,8 +2201,9 @@ int xfrm_state_mtu(struct xfrm_state *x, int mtu)
 
 int __xfrm_init_state(struct xfrm_state *x, bool init_replay, bool offload)
 {
-	struct xfrm_state_afinfo *afinfo;
-	struct xfrm_mode *inner_mode;
+	const struct xfrm_state_afinfo *afinfo;
+	const struct xfrm_mode *inner_mode;
+	const struct xfrm_mode *outer_mode;
 	int family = x->props.family;
 	int err;
 
@@ -2269,25 +2229,22 @@ int __xfrm_init_state(struct xfrm_state *x, bool init_replay, bool offload)
 			goto error;
 
 		if (!(inner_mode->flags & XFRM_MODE_FLAG_TUNNEL) &&
-		    family != x->sel.family) {
-			xfrm_put_mode(inner_mode);
+		    family != x->sel.family)
 			goto error;
-		}
 
-		x->inner_mode = inner_mode;
+		x->inner_mode = *inner_mode;
 	} else {
-		struct xfrm_mode *inner_mode_iaf;
+		const struct xfrm_mode *inner_mode_iaf;
 		int iafamily = AF_INET;
 
 		inner_mode = xfrm_get_mode(x->props.mode, x->props.family);
 		if (inner_mode == NULL)
 			goto error;
 
-		if (!(inner_mode->flags & XFRM_MODE_FLAG_TUNNEL)) {
-			xfrm_put_mode(inner_mode);
+		if (!(inner_mode->flags & XFRM_MODE_FLAG_TUNNEL))
 			goto error;
-		}
-		x->inner_mode = inner_mode;
+
+		x->inner_mode = *inner_mode;
 
 		if (x->props.family == AF_INET)
 			iafamily = AF_INET6;
@@ -2295,9 +2252,7 @@ int __xfrm_init_state(struct xfrm_state *x, bool init_replay, bool offload)
 		inner_mode_iaf = xfrm_get_mode(x->props.mode, iafamily);
 		if (inner_mode_iaf) {
 			if (inner_mode_iaf->flags & XFRM_MODE_FLAG_TUNNEL)
-				x->inner_mode_iaf = inner_mode_iaf;
-			else
-				xfrm_put_mode(inner_mode_iaf);
+				x->inner_mode_iaf = *inner_mode_iaf;
 		}
 	}
 
@@ -2311,12 +2266,13 @@ int __xfrm_init_state(struct xfrm_state *x, bool init_replay, bool offload)
 	if (err)
 		goto error;
 
-	x->outer_mode = xfrm_get_mode(x->props.mode, family);
-	if (x->outer_mode == NULL) {
+	outer_mode = xfrm_get_mode(x->props.mode, family);
+	if (!outer_mode) {
 		err = -EPROTONOSUPPORT;
 		goto error;
 	}
 
+	x->outer_mode = *outer_mode;
 	if (init_replay) {
 		err = xfrm_init_replay(x);
 		if (err)