[PATCH RFC ipsec-next 0/3] Virtual xfrm interfaces

[PATCH RFC ipsec-next 0/3] Virtual xfrm interfaces

[Steffen Klassert]

This patchset introduces new virtual xfrm interfaces.
The design of virtual xfrm interfaces interfaces was
discussed at the Linux IPsec workshop 2018. This patchset
implements these interfaces as the IPsec userspace and
kernel developers agreed. The purpose of these interfaces
is to overcome the design limitations that the existing
VTI devices have.

We had two presentations about xfrm interfaces at
the workshop. Slides with further informations
can be found at the workshop homepage:

https://workshop.linux-ipsec.org/2018/

[PATCH RFC ipsec-next 1/3] flow: Extend flow informations with xfrm interface id.

[Steffen Klassert]

Add a new flowi_xfrm structure with informations needed to do
a xfrm lookup. At the moment it keeps the informations about
the xfrm interface id.

Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
Acked-by: Shannon Nelson <shannon.nelson@oracle.com>
Acked-by: Benedict Wong <benedictwong@google.com>
Tested-by: Benedict Wong <benedictwong@google.com>
Tested-by: Antony Antony <antony@phenome.org>
Reviewed-by: Eyal Birger <eyal.birger@gmail.com>
---
 include/net/flow.h | 9 +++++++++
 1 file changed, 9 insertions(+)

diff --git a/include/net/flow.h b/include/net/flow.h
index 8ce21793094e..187c9bef672f 100644
--- a/include/net/flow.h
+++ b/include/net/flow.h
@@ -26,6 +26,10 @@ struct flowi_tunnel {
 	__be64			tun_id;
 };
 
+struct flowi_xfrm {
+	__u32			if_id;
+};
+
 struct flowi_common {
 	int	flowic_oif;
 	int	flowic_iif;
@@ -39,6 +43,7 @@ struct flowi_common {
 #define FLOWI_FLAG_SKIP_NH_OIF		0x04
 	__u32	flowic_secid;
 	struct flowi_tunnel flowic_tun_key;
+	struct flowi_xfrm xfrm;
 	kuid_t  flowic_uid;
 };
 
@@ -78,6 +83,7 @@ struct flowi4 {
 #define flowi4_secid		__fl_common.flowic_secid
 #define flowi4_tun_key		__fl_common.flowic_tun_key
 #define flowi4_uid		__fl_common.flowic_uid
+#define flowi4_xfrm		__fl_common.xfrm
 
 	/* (saddr,daddr) must be grouped, same order as in IP header */
 	__be32			saddr;
@@ -109,6 +115,7 @@ static inline void flowi4_init_output(struct flowi4 *fl4, int oif,
 	fl4->flowi4_flags = flags;
 	fl4->flowi4_secid = 0;
 	fl4->flowi4_tun_key.tun_id = 0;
+	fl4->flowi4_xfrm.if_id = 0;
 	fl4->flowi4_uid = uid;
 	fl4->daddr = daddr;
 	fl4->saddr = saddr;
@@ -138,6 +145,7 @@ struct flowi6 {
 #define flowi6_secid		__fl_common.flowic_secid
 #define flowi6_tun_key		__fl_common.flowic_tun_key
 #define flowi6_uid		__fl_common.flowic_uid
+#define flowi6_xfrm		__fl_common.xfrm
 	struct in6_addr		daddr;
 	struct in6_addr		saddr;
 	/* Note: flowi6_tos is encoded in flowlabel, too. */
@@ -185,6 +193,7 @@ struct flowi {
 #define flowi_secid	u.__fl_common.flowic_secid
 #define flowi_tun_key	u.__fl_common.flowic_tun_key
 #define flowi_uid	u.__fl_common.flowic_uid
+#define flowi_xfrm	u.__fl_common.xfrm
 } __attribute__((__aligned__(BITS_PER_LONG/8)));
 
 static inline struct flowi *flowi4_to_flowi(struct flowi4 *fl4)
-- 
2.14.1

[PATCH RFC ipsec-next 2/3] xfrm: Add a new lookup key to match xfrm interfaces.

[Steffen Klassert]

This patch adds the xfrm interface id as a lookup key
for xfrm states and policies. With this we can assign
states and policies to virtual xfrm interfaces.

Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
Acked-by: Shannon Nelson <shannon.nelson@oracle.com>
Acked-by: Benedict Wong <benedictwong@google.com>
Tested-by: Benedict Wong <benedictwong@google.com>
Tested-by: Antony Antony <antony@phenome.org>
Reviewed-by: Eyal Birger <eyal.birger@gmail.com>
---
 include/net/xfrm.h        | 21 ++++++++++++++-----
 include/uapi/linux/xfrm.h |  1 +
 net/core/pktgen.c         |  2 +-
 net/key/af_key.c          |  6 +++---
 net/xfrm/xfrm_policy.c    | 18 +++++++++++-----
 net/xfrm/xfrm_state.c     | 19 ++++++++++++-----
 net/xfrm/xfrm_user.c      | 53 ++++++++++++++++++++++++++++++++++++++++++-----
 7 files changed, 96 insertions(+), 24 deletions(-)

diff --git a/include/net/xfrm.h b/include/net/xfrm.h
index 45e75c36b738..26df8d6d507c 100644
--- a/include/net/xfrm.h
+++ b/include/net/xfrm.h
@@ -147,6 +147,7 @@ struct xfrm_state {
 	struct xfrm_id		id;
 	struct xfrm_selector	sel;
 	struct xfrm_mark	mark;
+	u32			if_id;
 	u32			tfcpad;
 
 	u32			genid;
@@ -574,6 +575,7 @@ struct xfrm_policy {
 	atomic_t		genid;
 	u32			priority;
 	u32			index;
+	u32			if_id;
 	struct xfrm_mark	mark;
 	struct xfrm_selector	selector;
 	struct xfrm_lifetime_cfg lft;
@@ -1533,7 +1535,7 @@ struct xfrm_state *xfrm_state_find(const xfrm_address_t *daddr,
 				   struct xfrm_tmpl *tmpl,
 				   struct xfrm_policy *pol, int *err,
 				   unsigned short family);
-struct xfrm_state *xfrm_stateonly_find(struct net *net, u32 mark,
+struct xfrm_state *xfrm_stateonly_find(struct net *net, u32 mark, u32 if_id,
 				       xfrm_address_t *daddr,
 				       xfrm_address_t *saddr,
 				       unsigned short family,
@@ -1690,20 +1692,20 @@ int xfrm_policy_walk(struct net *net, struct xfrm_policy_walk *walk,
 		     void *);
 void xfrm_policy_walk_done(struct xfrm_policy_walk *walk, struct net *net);
 int xfrm_policy_insert(int dir, struct xfrm_policy *policy, int excl);
-struct xfrm_policy *xfrm_policy_bysel_ctx(struct net *net, u32 mark,
+struct xfrm_policy *xfrm_policy_bysel_ctx(struct net *net, u32 mark, u32 if_id,
 					  u8 type, int dir,
 					  struct xfrm_selector *sel,
 					  struct xfrm_sec_ctx *ctx, int delete,
 					  int *err);
-struct xfrm_policy *xfrm_policy_byid(struct net *net, u32 mark, u8, int dir,
-				     u32 id, int delete, int *err);
+struct xfrm_policy *xfrm_policy_byid(struct net *net, u32 mark, u32 if_id, u8,
+				     int dir, u32 id, int delete, int *err);
 int xfrm_policy_flush(struct net *net, u8 type, bool task_valid);
 void xfrm_policy_hash_rebuild(struct net *net);
 u32 xfrm_get_acqseq(void);
 int verify_spi_info(u8 proto, u32 min, u32 max);
 int xfrm_alloc_spi(struct xfrm_state *x, u32 minspi, u32 maxspi);
 struct xfrm_state *xfrm_find_acq(struct net *net, const struct xfrm_mark *mark,
-				 u8 mode, u32 reqid, u8 proto,
+				 u8 mode, u32 reqid, u32 if_id, u8 proto,
 				 const xfrm_address_t *daddr,
 				 const xfrm_address_t *saddr, int create,
 				 unsigned short family);
@@ -2012,6 +2014,15 @@ static inline int xfrm_mark_put(struct sk_buff *skb, const struct xfrm_mark *m)
 	return ret;
 }
 
+static inline int xfrm_if_id_put(struct sk_buff *skb, __u32 if_id)
+{
+	int ret = 0;
+
+	if (if_id)
+		ret = nla_put_u32(skb, XFRMA_IF_ID, if_id);
+	return ret;
+}
+
 static inline int xfrm_tunnel_check(struct sk_buff *skb, struct xfrm_state *x,
 				    unsigned int family)
 {
diff --git a/include/uapi/linux/xfrm.h b/include/uapi/linux/xfrm.h
index e3af2859188b..690df6e7580b 100644
--- a/include/uapi/linux/xfrm.h
+++ b/include/uapi/linux/xfrm.h
@@ -306,6 +306,7 @@ enum xfrm_attr_type_t {
 	XFRMA_PAD,
 	XFRMA_OFFLOAD_DEV,	/* struct xfrm_state_offload */
 	XFRMA_OUTPUT_MARK,	/* __u32 */
+	XFRMA_IF_ID,		/* __u32 */
 	__XFRMA_MAX
 
 #define XFRMA_MAX (__XFRMA_MAX - 1)
diff --git a/net/core/pktgen.c b/net/core/pktgen.c
index 7e4ede34cc52..a1b2d769fa36 100644
--- a/net/core/pktgen.c
+++ b/net/core/pktgen.c
@@ -2255,7 +2255,7 @@ static void get_ipsec_sa(struct pktgen_dev *pkt_dev, int flow)
 			x = xfrm_state_lookup_byspi(pn->net, htonl(pkt_dev->spi), AF_INET);
 		} else {
 			/* slow path: we dont already have xfrm_state */
-			x = xfrm_stateonly_find(pn->net, DUMMY_MARK,
+			x = xfrm_stateonly_find(pn->net, DUMMY_MARK, 0,
 						(xfrm_address_t *)&pkt_dev->cur_daddr,
 						(xfrm_address_t *)&pkt_dev->cur_saddr,
 						AF_INET,
diff --git a/net/key/af_key.c b/net/key/af_key.c
index e62e52e8f141..bf38b358cf38 100644
--- a/net/key/af_key.c
+++ b/net/key/af_key.c
@@ -1383,7 +1383,7 @@ static int pfkey_getspi(struct sock *sk, struct sk_buff *skb, const struct sadb_
 	}
 
 	if (!x)
-		x = xfrm_find_acq(net, &dummy_mark, mode, reqid, proto, xdaddr, xsaddr, 1, family);
+		x = xfrm_find_acq(net, &dummy_mark, mode, reqid, 0, proto, xdaddr, xsaddr, 1, family);
 
 	if (x == NULL)
 		return -ENOENT;
@@ -2414,7 +2414,7 @@ static int pfkey_spddelete(struct sock *sk, struct sk_buff *skb, const struct sa
 			return err;
 	}
 
-	xp = xfrm_policy_bysel_ctx(net, DUMMY_MARK, XFRM_POLICY_TYPE_MAIN,
+	xp = xfrm_policy_bysel_ctx(net, DUMMY_MARK, 0, XFRM_POLICY_TYPE_MAIN,
 				   pol->sadb_x_policy_dir - 1, &sel, pol_ctx,
 				   1, &err);
 	security_xfrm_policy_free(pol_ctx);
@@ -2663,7 +2663,7 @@ static int pfkey_spdget(struct sock *sk, struct sk_buff *skb, const struct sadb_
 		return -EINVAL;
 
 	delete = (hdr->sadb_msg_type == SADB_X_SPDDELETE2);
-	xp = xfrm_policy_byid(net, DUMMY_MARK, XFRM_POLICY_TYPE_MAIN,
+	xp = xfrm_policy_byid(net, DUMMY_MARK, 0, XFRM_POLICY_TYPE_MAIN,
 			      dir, pol->sadb_x_policy_id, delete, &err);
 	if (xp == NULL)
 		return -ENOENT;
diff --git a/net/xfrm/xfrm_policy.c b/net/xfrm/xfrm_policy.c
index 40b54cc64243..08b745aade43 100644
--- a/net/xfrm/xfrm_policy.c
+++ b/net/xfrm/xfrm_policy.c
@@ -747,6 +747,7 @@ int xfrm_policy_insert(int dir, struct xfrm_policy *policy, int excl)
 	newpos = NULL;
 	hlist_for_each_entry(pol, chain, bydst) {
 		if (pol->type == policy->type &&
+		    pol->if_id == policy->if_id &&
 		    !selector_cmp(&pol->selector, &policy->selector) &&
 		    xfrm_policy_mark_match(policy, pol) &&
 		    xfrm_sec_ctx_match(pol->security, policy->security) &&
@@ -798,8 +799,9 @@ int xfrm_policy_insert(int dir, struct xfrm_policy *policy, int excl)
 }
 EXPORT_SYMBOL(xfrm_policy_insert);
 
-struct xfrm_policy *xfrm_policy_bysel_ctx(struct net *net, u32 mark, u8 type,
-					  int dir, struct xfrm_selector *sel,
+struct xfrm_policy *xfrm_policy_bysel_ctx(struct net *net, u32 mark, u32 if_id,
+					  u8 type, int dir,
+					  struct xfrm_selector *sel,
 					  struct xfrm_sec_ctx *ctx, int delete,
 					  int *err)
 {
@@ -812,6 +814,7 @@ struct xfrm_policy *xfrm_policy_bysel_ctx(struct net *net, u32 mark, u8 type,
 	ret = NULL;
 	hlist_for_each_entry(pol, chain, bydst) {
 		if (pol->type == type &&
+		    pol->if_id == if_id &&
 		    (mark & pol->mark.m) == pol->mark.v &&
 		    !selector_cmp(sel, &pol->selector) &&
 		    xfrm_sec_ctx_match(ctx, pol->security)) {
@@ -837,8 +840,9 @@ struct xfrm_policy *xfrm_policy_bysel_ctx(struct net *net, u32 mark, u8 type,
 }
 EXPORT_SYMBOL(xfrm_policy_bysel_ctx);
 
-struct xfrm_policy *xfrm_policy_byid(struct net *net, u32 mark, u8 type,
-				     int dir, u32 id, int delete, int *err)
+struct xfrm_policy *xfrm_policy_byid(struct net *net, u32 mark, u32 if_id,
+				     u8 type, int dir, u32 id, int delete,
+				     int *err)
 {
 	struct xfrm_policy *pol, *ret;
 	struct hlist_head *chain;
@@ -853,6 +857,7 @@ struct xfrm_policy *xfrm_policy_byid(struct net *net, u32 mark, u8 type,
 	ret = NULL;
 	hlist_for_each_entry(pol, chain, byidx) {
 		if (pol->type == type && pol->index == id &&
+		    pol->if_id == if_id &&
 		    (mark & pol->mark.m) == pol->mark.v) {
 			xfrm_pol_hold(pol);
 			if (delete) {
@@ -1063,6 +1068,7 @@ static int xfrm_policy_match(const struct xfrm_policy *pol,
 	bool match;
 
 	if (pol->family != family ||
+	    pol->if_id != fl->flowi_xfrm.if_id ||
 	    (fl->flowi_mark & pol->mark.m) != pol->mark.v ||
 	    pol->type != type)
 		return ret;
@@ -1177,7 +1183,8 @@ static struct xfrm_policy *xfrm_sk_policy_lookup(const struct sock *sk, int dir,
 
 		match = xfrm_selector_match(&pol->selector, fl, family);
 		if (match) {
-			if ((sk->sk_mark & pol->mark.m) != pol->mark.v) {
+			if ((sk->sk_mark & pol->mark.m) != pol->mark.v ||
+			    pol->if_id != fl->flowi_xfrm.if_id) {
 				pol = NULL;
 				goto out;
 			}
@@ -1305,6 +1312,7 @@ static struct xfrm_policy *clone_policy(const struct xfrm_policy *old, int dir)
 		newp->lft = old->lft;
 		newp->curlft = old->curlft;
 		newp->mark = old->mark;
+		newp->if_id = old->if_id;
 		newp->action = old->action;
 		newp->flags = old->flags;
 		newp->xfrm_nr = old->xfrm_nr;
diff --git a/net/xfrm/xfrm_state.c b/net/xfrm/xfrm_state.c
index 8308281f3253..3803b6813fc5 100644
--- a/net/xfrm/xfrm_state.c
+++ b/net/xfrm/xfrm_state.c
@@ -941,6 +941,7 @@ xfrm_state_find(const xfrm_address_t *daddr, const xfrm_address_t *saddr,
 	int error = 0;
 	struct xfrm_state *best = NULL;
 	u32 mark = pol->mark.v & pol->mark.m;
+	u32 if_id = fl->flowi_xfrm.if_id;
 	unsigned short encap_family = tmpl->encap_family;
 	unsigned int sequence;
 	struct km_event c;
@@ -955,6 +956,7 @@ xfrm_state_find(const xfrm_address_t *daddr, const xfrm_address_t *saddr,
 		if (x->props.family == encap_family &&
 		    x->props.reqid == tmpl->reqid &&
 		    (mark & x->mark.m) == x->mark.v &&
+		    x->if_id == if_id &&
 		    !(x->props.flags & XFRM_STATE_WILDRECV) &&
 		    xfrm_state_addr_check(x, daddr, saddr, encap_family) &&
 		    tmpl->mode == x->props.mode &&
@@ -971,6 +973,7 @@ xfrm_state_find(const xfrm_address_t *daddr, const xfrm_address_t *saddr,
 		if (x->props.family == encap_family &&
 		    x->props.reqid == tmpl->reqid &&
 		    (mark & x->mark.m) == x->mark.v &&
+		    x->if_id == if_id &&
 		    !(x->props.flags & XFRM_STATE_WILDRECV) &&
 		    xfrm_addr_equal(&x->id.daddr, daddr, encap_family) &&
 		    tmpl->mode == x->props.mode &&
@@ -1010,6 +1013,7 @@ xfrm_state_find(const xfrm_address_t *daddr, const xfrm_address_t *saddr,
 		 * to current session. */
 		xfrm_init_tempstate(x, fl, tmpl, daddr, saddr, family);
 		memcpy(&x->mark, &pol->mark, sizeof(x->mark));
+		x->if_id = if_id;
 
 		error = security_xfrm_state_alloc_acquire(x, pol->security, fl->flowi_secid);
 		if (error) {
@@ -1067,7 +1071,7 @@ xfrm_state_find(const xfrm_address_t *daddr, const xfrm_address_t *saddr,
 }
 
 struct xfrm_state *
-xfrm_stateonly_find(struct net *net, u32 mark,
+xfrm_stateonly_find(struct net *net, u32 mark, u32 if_id,
 		    xfrm_address_t *daddr, xfrm_address_t *saddr,
 		    unsigned short family, u8 mode, u8 proto, u32 reqid)
 {
@@ -1080,6 +1084,7 @@ xfrm_stateonly_find(struct net *net, u32 mark,
 		if (x->props.family == family &&
 		    x->props.reqid == reqid &&
 		    (mark & x->mark.m) == x->mark.v &&
+		    x->if_id == if_id &&
 		    !(x->props.flags & XFRM_STATE_WILDRECV) &&
 		    xfrm_state_addr_check(x, daddr, saddr, family) &&
 		    mode == x->props.mode &&
@@ -1160,11 +1165,13 @@ static void __xfrm_state_bump_genids(struct xfrm_state *xnew)
 	struct xfrm_state *x;
 	unsigned int h;
 	u32 mark = xnew->mark.v & xnew->mark.m;
+	u32 if_id = xnew->if_id;
 
 	h = xfrm_dst_hash(net, &xnew->id.daddr, &xnew->props.saddr, reqid, family);
 	hlist_for_each_entry(x, net->xfrm.state_bydst+h, bydst) {
 		if (x->props.family	== family &&
 		    x->props.reqid	== reqid &&
+		    x->if_id		== if_id &&
 		    (mark & x->mark.m) == x->mark.v &&
 		    xfrm_addr_equal(&x->id.daddr, &xnew->id.daddr, family) &&
 		    xfrm_addr_equal(&x->props.saddr, &xnew->props.saddr, family))
@@ -1187,7 +1194,7 @@ EXPORT_SYMBOL(xfrm_state_insert);
 static struct xfrm_state *__find_acq_core(struct net *net,
 					  const struct xfrm_mark *m,
 					  unsigned short family, u8 mode,
-					  u32 reqid, u8 proto,
+					  u32 reqid, u32 if_id, u8 proto,
 					  const xfrm_address_t *daddr,
 					  const xfrm_address_t *saddr,
 					  int create)
@@ -1242,6 +1249,7 @@ static struct xfrm_state *__find_acq_core(struct net *net,
 		x->props.family = family;
 		x->props.mode = mode;
 		x->props.reqid = reqid;
+		x->if_id = if_id;
 		x->mark.v = m->v;
 		x->mark.m = m->m;
 		x->lft.hard_add_expires_seconds = net->xfrm.sysctl_acq_expires;
@@ -1296,7 +1304,7 @@ int xfrm_state_add(struct xfrm_state *x)
 
 	if (use_spi && !x1)
 		x1 = __find_acq_core(net, &x->mark, family, x->props.mode,
-				     x->props.reqid, x->id.proto,
+				     x->props.reqid, x->if_id, x->id.proto,
 				     &x->id.daddr, &x->props.saddr, 0);
 
 	__xfrm_state_bump_genids(x);
@@ -1395,6 +1403,7 @@ static struct xfrm_state *xfrm_state_clone(struct xfrm_state *orig,
 	x->props.flags = orig->props.flags;
 	x->props.extra_flags = orig->props.extra_flags;
 
+	x->if_id = orig->if_id;
 	x->tfcpad = orig->tfcpad;
 	x->replay_maxdiff = orig->replay_maxdiff;
 	x->replay_maxage = orig->replay_maxage;
@@ -1619,13 +1628,13 @@ EXPORT_SYMBOL(xfrm_state_lookup_byaddr);
 
 struct xfrm_state *
 xfrm_find_acq(struct net *net, const struct xfrm_mark *mark, u8 mode, u32 reqid,
-	      u8 proto, const xfrm_address_t *daddr,
+	      u32 if_id, u8 proto, const xfrm_address_t *daddr,
 	      const xfrm_address_t *saddr, int create, unsigned short family)
 {
 	struct xfrm_state *x;
 
 	spin_lock_bh(&net->xfrm.xfrm_state_lock);
-	x = __find_acq_core(net, mark, family, mode, reqid, proto, daddr, saddr, create);
+	x = __find_acq_core(net, mark, family, mode, reqid, if_id, proto, daddr, saddr, create);
 	spin_unlock_bh(&net->xfrm.xfrm_state_lock);
 
 	return x;
diff --git a/net/xfrm/xfrm_user.c b/net/xfrm/xfrm_user.c
index 080035f056d9..08ec1ab8e36c 100644
--- a/net/xfrm/xfrm_user.c
+++ b/net/xfrm/xfrm_user.c
@@ -582,6 +582,9 @@ static struct xfrm_state *xfrm_state_construct(struct net *net,
 	if (attrs[XFRMA_OUTPUT_MARK])
 		x->props.output_mark = nla_get_u32(attrs[XFRMA_OUTPUT_MARK]);
 
+	if (attrs[XFRMA_IF_ID])
+		x->if_id = nla_get_u32(attrs[XFRMA_IF_ID]);
+
 	err = __xfrm_init_state(x, false, attrs[XFRMA_OFFLOAD_DEV]);
 	if (err)
 		goto error;
@@ -905,6 +908,11 @@ static int copy_to_user_state_extra(struct xfrm_state *x,
 		if (ret)
 			goto out;
 	}
+	if (x->if_id) {
+		ret = nla_put_u32(skb, XFRMA_IF_ID, x->if_id);
+		if (ret)
+			goto out;
+	}
 	if (x->security)
 		ret = copy_sec_ctx(x->security, skb);
 out:
@@ -1253,6 +1261,7 @@ static int xfrm_alloc_userspi(struct sk_buff *skb, struct nlmsghdr *nlh,
 	int err;
 	u32 mark;
 	struct xfrm_mark m;
+	u32 if_id = 0;
 
 	p = nlmsg_data(nlh);
 	err = verify_spi_info(p->info.id.proto, p->min, p->max);
@@ -1265,6 +1274,10 @@ static int xfrm_alloc_userspi(struct sk_buff *skb, struct nlmsghdr *nlh,
 	x = NULL;
 
 	mark = xfrm_mark_get(attrs, &m);
+
+	if (attrs[XFRMA_IF_ID])
+		if_id = nla_get_u32(attrs[XFRMA_IF_ID]);
+
 	if (p->info.seq) {
 		x = xfrm_find_acq_byseq(net, mark, p->info.seq);
 		if (x && !xfrm_addr_equal(&x->id.daddr, daddr, family)) {
@@ -1275,7 +1288,7 @@ static int xfrm_alloc_userspi(struct sk_buff *skb, struct nlmsghdr *nlh,
 
 	if (!x)
 		x = xfrm_find_acq(net, &m, p->info.mode, p->info.reqid,
-				  p->info.id.proto, daddr,
+				  if_id, p->info.id.proto, daddr,
 				  &p->info.saddr, 1,
 				  family);
 	err = -ENOENT;
@@ -1563,6 +1576,9 @@ static struct xfrm_policy *xfrm_policy_construct(struct net *net, struct xfrm_us
 
 	xfrm_mark_get(attrs, &xp->mark);
 
+	if (attrs[XFRMA_IF_ID])
+		xp->if_id = nla_get_u32(attrs[XFRMA_IF_ID]);
+
 	return xp;
  error:
 	*errp = err;
@@ -1708,6 +1724,8 @@ static int dump_one_policy(struct xfrm_policy *xp, int dir, int count, void *ptr
 		err = copy_to_user_policy_type(xp->type, skb);
 	if (!err)
 		err = xfrm_mark_put(skb, &xp->mark);
+	if (!err)
+		err = xfrm_if_id_put(skb, xp->if_id);
 	if (err) {
 		nlmsg_cancel(skb, nlh);
 		return err;
@@ -1789,6 +1807,7 @@ static int xfrm_get_policy(struct sk_buff *skb, struct nlmsghdr *nlh,
 	int delete;
 	struct xfrm_mark m;
 	u32 mark = xfrm_mark_get(attrs, &m);
+	u32 if_id = 0;
 
 	p = nlmsg_data(nlh);
 	delete = nlh->nlmsg_type == XFRM_MSG_DELPOLICY;
@@ -1801,8 +1820,11 @@ static int xfrm_get_policy(struct sk_buff *skb, struct nlmsghdr *nlh,
 	if (err)
 		return err;
 
+	if (attrs[XFRMA_IF_ID])
+		if_id = nla_get_u32(attrs[XFRMA_IF_ID]);
+
 	if (p->index)
-		xp = xfrm_policy_byid(net, mark, type, p->dir, p->index, delete, &err);
+		xp = xfrm_policy_byid(net, mark, if_id, type, p->dir, p->index, delete, &err);
 	else {
 		struct nlattr *rt = attrs[XFRMA_SEC_CTX];
 		struct xfrm_sec_ctx *ctx;
@@ -1819,7 +1841,7 @@ static int xfrm_get_policy(struct sk_buff *skb, struct nlmsghdr *nlh,
 			if (err)
 				return err;
 		}
-		xp = xfrm_policy_bysel_ctx(net, mark, type, p->dir, &p->sel,
+		xp = xfrm_policy_bysel_ctx(net, mark, if_id, type, p->dir, &p->sel,
 					   ctx, delete, &err);
 		security_xfrm_policy_free(ctx);
 	}
@@ -1942,6 +1964,10 @@ static int build_aevent(struct sk_buff *skb, struct xfrm_state *x, const struct
 	if (err)
 		goto out_cancel;
 
+	err = xfrm_if_id_put(skb, x->if_id);
+	if (err)
+		goto out_cancel;
+
 	nlmsg_end(skb, nlh);
 	return 0;
 
@@ -2084,6 +2110,7 @@ static int xfrm_add_pol_expire(struct sk_buff *skb, struct nlmsghdr *nlh,
 	int err = -ENOENT;
 	struct xfrm_mark m;
 	u32 mark = xfrm_mark_get(attrs, &m);
+	u32 if_id = 0;
 
 	err = copy_from_user_policy_type(&type, attrs);
 	if (err)
@@ -2093,8 +2120,11 @@ static int xfrm_add_pol_expire(struct sk_buff *skb, struct nlmsghdr *nlh,
 	if (err)
 		return err;
 
+	if (attrs[XFRMA_IF_ID])
+		if_id = nla_get_u32(attrs[XFRMA_IF_ID]);
+
 	if (p->index)
-		xp = xfrm_policy_byid(net, mark, type, p->dir, p->index, 0, &err);
+		xp = xfrm_policy_byid(net, mark, if_id, type, p->dir, p->index, 0, &err);
 	else {
 		struct nlattr *rt = attrs[XFRMA_SEC_CTX];
 		struct xfrm_sec_ctx *ctx;
@@ -2111,7 +2141,7 @@ static int xfrm_add_pol_expire(struct sk_buff *skb, struct nlmsghdr *nlh,
 			if (err)
 				return err;
 		}
-		xp = xfrm_policy_bysel_ctx(net, mark, type, p->dir,
+		xp = xfrm_policy_bysel_ctx(net, mark, if_id, type, p->dir,
 					   &p->sel, ctx, 0, &err);
 		security_xfrm_policy_free(ctx);
 	}
@@ -2494,6 +2524,7 @@ static const struct nla_policy xfrma_policy[XFRMA_MAX+1] = {
 	[XFRMA_ADDRESS_FILTER]	= { .len = sizeof(struct xfrm_address_filter) },
 	[XFRMA_OFFLOAD_DEV]	= { .len = sizeof(struct xfrm_user_offload) },
 	[XFRMA_OUTPUT_MARK]	= { .type = NLA_U32 },
+	[XFRMA_IF_ID]		= { .type = NLA_U32 },
 };
 
 static const struct nla_policy xfrma_spd_policy[XFRMA_SPD_MAX+1] = {
@@ -2625,6 +2656,10 @@ static int build_expire(struct sk_buff *skb, struct xfrm_state *x, const struct
 	if (err)
 		return err;
 
+	err = xfrm_if_id_put(skb, x->if_id);
+	if (err)
+		return err;
+
 	nlmsg_end(skb, nlh);
 	return 0;
 }
@@ -2721,6 +2756,8 @@ static inline unsigned int xfrm_sa_len(struct xfrm_state *x)
 		 l += nla_total_size(sizeof(x->xso));
 	if (x->props.output_mark)
 		l += nla_total_size(sizeof(x->props.output_mark));
+	if (x->if_id)
+		l += nla_total_size(sizeof(x->if_id));
 
 	/* Must count x->lastused as it may become non-zero behind our back. */
 	l += nla_total_size_64bit(sizeof(u64));
@@ -2850,6 +2887,8 @@ static int build_acquire(struct sk_buff *skb, struct xfrm_state *x,
 		err = copy_to_user_policy_type(xp->type, skb);
 	if (!err)
 		err = xfrm_mark_put(skb, &xp->mark);
+	if (!err)
+		err = xfrm_if_id_put(skb, xp->if_id);
 	if (err) {
 		nlmsg_cancel(skb, nlh);
 		return err;
@@ -2966,6 +3005,8 @@ static int build_polexpire(struct sk_buff *skb, struct xfrm_policy *xp,
 		err = copy_to_user_policy_type(xp->type, skb);
 	if (!err)
 		err = xfrm_mark_put(skb, &xp->mark);
+	if (!err)
+		err = xfrm_if_id_put(skb, xp->if_id);
 	if (err) {
 		nlmsg_cancel(skb, nlh);
 		return err;
@@ -3047,6 +3088,8 @@ static int xfrm_notify_policy(struct xfrm_policy *xp, int dir, const struct km_e
 		err = copy_to_user_policy_type(xp->type, skb);
 	if (!err)
 		err = xfrm_mark_put(skb, &xp->mark);
+	if (!err)
+		err = xfrm_if_id_put(skb, xp->if_id);
 	if (err)
 		goto out_free_skb;
 
-- 
2.14.1

[PATCH RFC ipsec-next 3/3] xfrm: Add virtual xfrm interfaces

[Steffen Klassert]

This patch adds support for virtual xfrm interfaces.
Packets that are routed through such an interface
are guaranteed to be IPsec transformed or dropped.

Co-developed-by: Lorenzo Colitti <lorenzo@google.com>
Co-developed-by: Benedict Wong <benedictwong@google.com>
Signed-off-by: Lorenzo Colitti <lorenzo@google.com>
Signed-off-by: Benedict Wong <benedictwong@google.com>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
Acked-by: Shannon Nelson <shannon.nelson@oracle.com>
Tested-by: Benedict Wong <benedictwong@google.com>
Tested-by: Antony Antony <antony@phenome.org>
Reviewed-by: Eyal Birger <eyal.birger@gmail.com>
---
 include/net/xfrm.h           |  24 ++
 include/uapi/linux/if_link.h |  10 +
 net/xfrm/Kconfig             |   8 +
 net/xfrm/Makefile            |   1 +
 net/xfrm/xfrm_input.c        |   3 +
 net/xfrm/xfrm_interface.c    | 971 +++++++++++++++++++++++++++++++++++++++++++
 net/xfrm/xfrm_policy.c       |  43 ++
 7 files changed, 1060 insertions(+)
 create mode 100644 net/xfrm/xfrm_interface.c

diff --git a/include/net/xfrm.h b/include/net/xfrm.h
index 26df8d6d507c..211b15c55a92 100644
--- a/include/net/xfrm.h
+++ b/include/net/xfrm.h
@@ -23,6 +23,7 @@
 #include <net/ipv6.h>
 #include <net/ip6_fib.h>
 #include <net/flow.h>
+#include <net/gro_cells.h>
 
 #include <linux/interrupt.h>
 
@@ -293,6 +294,13 @@ struct xfrm_replay {
 	int	(*overflow)(struct xfrm_state *x, struct sk_buff *skb);
 };
 
+struct xfrm_if_cb {
+	struct xfrm_if	*(*decode_session)(struct sk_buff *skb);
+};
+
+void xfrm_if_register_cb(const struct xfrm_if_cb *ifcb);
+void xfrm_if_unregister_cb(void);
+
 struct net_device;
 struct xfrm_type;
 struct xfrm_dst;
@@ -1039,6 +1047,22 @@ static inline void xfrm_dst_destroy(struct xfrm_dst *xdst)
 
 void xfrm_dst_ifdown(struct dst_entry *dst, struct net_device *dev);
 
+struct xfrm_if_parms {
+	char name[IFNAMSIZ];	/* name of XFRM device */
+	int link;		/* ifindex of underlying L2 interface */
+	u32 if_id;		/* interface identifyer */
+};
+
+struct xfrm_if {
+	struct xfrm_if __rcu *next;	/* next interface in list */
+	struct net_device *dev;		/* virtual device associated with interface */
+	struct net_device *phydev;	/* physical device */
+	struct net *net;		/* netns for packet i/o */
+	struct xfrm_if_parms p;		/* interface parms */
+
+	struct gro_cells gro_cells;
+};
+
 struct xfrm_offload {
 	/* Output sequence number for replay protection on offloading. */
 	struct {
diff --git a/include/uapi/linux/if_link.h b/include/uapi/linux/if_link.h
index cf01b6824244..bff0af507b32 100644
--- a/include/uapi/linux/if_link.h
+++ b/include/uapi/linux/if_link.h
@@ -459,6 +459,16 @@ enum {
 
 #define IFLA_MACSEC_MAX (__IFLA_MACSEC_MAX - 1)
 
+/* XFRM section */
+enum {
+	IFLA_XFRM_UNSPEC,
+	IFLA_XFRM_LINK,
+	IFLA_XFRM_IF_ID,
+	__IFLA_XFRM_MAX
+};
+
+#define IFLA_XFRM_MAX (__IFLA_XFRM_MAX - 1)
+
 enum macsec_validation_type {
 	MACSEC_VALIDATE_DISABLED = 0,
 	MACSEC_VALIDATE_CHECK = 1,
diff --git a/net/xfrm/Kconfig b/net/xfrm/Kconfig
index 286ed25c1a69..53381888a7b3 100644
--- a/net/xfrm/Kconfig
+++ b/net/xfrm/Kconfig
@@ -25,6 +25,14 @@ config XFRM_USER
 
 	  If unsure, say Y.
 
+config XFRM_INTERFACE
+	tristate "Transformation virtual interface"
+	depends on XFRM && IPV6
+	---help---
+	  This provides a virtual interface to route IPsec traffic.
+
+	  If unsure, say N.
+
 config XFRM_SUB_POLICY
 	bool "Transformation sub policy support"
 	depends on XFRM
diff --git a/net/xfrm/Makefile b/net/xfrm/Makefile
index 0bd2465a8c5a..fbc4552d17b8 100644
--- a/net/xfrm/Makefile
+++ b/net/xfrm/Makefile
@@ -10,3 +10,4 @@ obj-$(CONFIG_XFRM_STATISTICS) += xfrm_proc.o
 obj-$(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
diff --git a/net/xfrm/xfrm_input.c b/net/xfrm/xfrm_input.c
index 352abca2605f..b104724d1caa 100644
--- a/net/xfrm/xfrm_input.c
+++ b/net/xfrm/xfrm_input.c
@@ -320,6 +320,7 @@ int xfrm_input(struct sk_buff *skb, int nexthdr, __be32 spi, int encap_type)
 
 	seq = 0;
 	if (!spi && (err = xfrm_parse_spi(skb, nexthdr, &spi, &seq)) != 0) {
+		secpath_reset(skb);
 		XFRM_INC_STATS(net, LINUX_MIB_XFRMINHDRERROR);
 		goto drop;
 	}
@@ -328,12 +329,14 @@ int xfrm_input(struct sk_buff *skb, int nexthdr, __be32 spi, int encap_type)
 				   XFRM_SPI_SKB_CB(skb)->daddroff);
 	do {
 		if (skb->sp->len == XFRM_MAX_DEPTH) {
+			secpath_reset(skb);
 			XFRM_INC_STATS(net, LINUX_MIB_XFRMINBUFFERERROR);
 			goto drop;
 		}
 
 		x = xfrm_state_lookup(net, mark, daddr, spi, nexthdr, family);
 		if (x == NULL) {
+			secpath_reset(skb);
 			XFRM_INC_STATS(net, LINUX_MIB_XFRMINNOSTATES);
 			xfrm_audit_state_notfound(skb, family, spi, seq);
 			goto drop;
diff --git a/net/xfrm/xfrm_interface.c b/net/xfrm/xfrm_interface.c
new file mode 100644
index 000000000000..57df8f087132
--- /dev/null
+++ b/net/xfrm/xfrm_interface.c
@@ -0,0 +1,971 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ *	XFRM virtual interface
+ *
+ *	Copyright (C) 2018 secunet Security Networks AG
+ *
+ *	Author:
+ *	Steffen Klassert <steffen.klassert@secunet.com>
+ */
+
+#include <linux/module.h>
+#include <linux/capability.h>
+#include <linux/errno.h>
+#include <linux/types.h>
+#include <linux/sockios.h>
+#include <linux/icmp.h>
+#include <linux/if.h>
+#include <linux/in.h>
+#include <linux/ip.h>
+#include <linux/net.h>
+#include <linux/in6.h>
+#include <linux/netdevice.h>
+#include <linux/if_link.h>
+#include <linux/if_arp.h>
+#include <linux/icmpv6.h>
+#include <linux/init.h>
+#include <linux/route.h>
+#include <linux/rtnetlink.h>
+#include <linux/netfilter_ipv6.h>
+#include <linux/slab.h>
+#include <linux/hash.h>
+
+#include <linux/uaccess.h>
+#include <linux/atomic.h>
+
+#include <net/icmp.h>
+#include <net/ip.h>
+#include <net/ipv6.h>
+#include <net/ip6_route.h>
+#include <net/addrconf.h>
+#include <net/xfrm.h>
+#include <net/net_namespace.h>
+#include <net/netns/generic.h>
+#include <linux/etherdevice.h>
+
+static int xfrmi_dev_init(struct net_device *dev);
+static void xfrmi_dev_setup(struct net_device *dev);
+static struct rtnl_link_ops xfrmi_link_ops __read_mostly;
+static unsigned int xfrmi_net_id __read_mostly;
+
+struct xfrmi_net {
+	/* lists for storing interfaces in use */
+	struct xfrm_if __rcu *xfrmi[1];
+};
+
+#define for_each_xfrmi_rcu(start, xi) \
+	for (xi = rcu_dereference(start); xi; xi = rcu_dereference(xi->next))
+
+static struct xfrm_if *xfrmi_lookup(struct net *net, struct xfrm_state *x)
+{
+	struct xfrmi_net *xfrmn = net_generic(net, xfrmi_net_id);
+	struct xfrm_if *xi;
+
+	for_each_xfrmi_rcu(xfrmn->xfrmi[0], xi) {
+		if (x->if_id == xi->p.if_id &&
+		    (xi->dev->flags & IFF_UP))
+			return xi;
+	}
+
+	return NULL;
+}
+
+static struct xfrm_if *xfrmi_decode_session(struct sk_buff *skb)
+{
+	struct xfrmi_net *xfrmn;
+	int ifindex;
+	struct xfrm_if *xi;
+
+	if (!skb->dev)
+		return NULL;
+
+	xfrmn = net_generic(dev_net(skb->dev), xfrmi_net_id);
+	ifindex = skb->dev->ifindex;
+
+	for_each_xfrmi_rcu(xfrmn->xfrmi[0], xi) {
+		if (ifindex == xi->dev->ifindex &&
+			(xi->dev->flags & IFF_UP))
+				return xi;
+	}
+
+	return NULL;
+}
+
+static void xfrmi_link(struct xfrmi_net *xfrmn, struct xfrm_if *xi)
+{
+	struct xfrm_if __rcu **xip = &xfrmn->xfrmi[0];
+
+	rcu_assign_pointer(xi->next , rtnl_dereference(*xip));
+	rcu_assign_pointer(*xip, xi);
+}
+
+static void xfrmi_unlink(struct xfrmi_net *xfrmn, struct xfrm_if *xi)
+{
+	struct xfrm_if __rcu **xip;
+	struct xfrm_if *iter;
+
+	for (xip = &xfrmn->xfrmi[0];
+	     (iter = rtnl_dereference(*xip)) != NULL;
+	     xip = &iter->next) {
+		if (xi == iter) {
+			rcu_assign_pointer(*xip, xi->next);
+			break;
+		}
+	}
+}
+
+static void xfrmi_dev_free(struct net_device *dev)
+{
+	free_percpu(dev->tstats);
+}
+
+static int xfrmi_create2(struct net_device *dev)
+{
+	struct xfrm_if *xi = netdev_priv(dev);
+	struct net *net = dev_net(dev);
+	struct xfrmi_net *xfrmn = net_generic(net, xfrmi_net_id);
+	int err;
+
+	dev->rtnl_link_ops = &xfrmi_link_ops;
+	err = register_netdevice(dev);
+	if (err < 0)
+		goto out;
+
+	strcpy(xi->p.name, dev->name);
+
+	dev_hold(dev);
+	xfrmi_link(xfrmn, xi);
+
+	return 0;
+
+out:
+	return err;
+}
+
+static struct xfrm_if *xfrmi_create(struct net *net, struct xfrm_if_parms *p)
+{
+	struct net_device *dev;
+	struct xfrm_if *xi;
+	char name[IFNAMSIZ];
+	int err;
+
+	if (p->name[0])
+		strlcpy(name, p->name, IFNAMSIZ);
+	else
+		goto failed;
+
+	dev = alloc_netdev(sizeof(*xi), name, NET_NAME_UNKNOWN, xfrmi_dev_setup);
+	if (!dev)
+		goto failed;
+
+	dev_net_set(dev, net);
+
+	xi = netdev_priv(dev);
+	xi->p = *p;
+	xi->net = net;
+	xi->dev = dev;
+	xi->phydev = dev_get_by_index(net, p->link);
+	if (!xi->phydev)
+		goto failed_free;
+
+	err = xfrmi_create2(dev);
+	if (err < 0)
+		goto failed_dev_put;
+
+	return xi;
+
+failed_dev_put:
+	dev_put(xi->phydev);
+failed_free:
+	free_netdev(dev);
+failed:
+	return NULL;
+}
+
+static struct xfrm_if *xfrmi_locate(struct net *net, struct xfrm_if_parms *p,
+				   int create)
+{
+	struct xfrm_if __rcu **xip;
+	struct xfrm_if *xi;
+	struct xfrmi_net *xfrmn = net_generic(net, xfrmi_net_id);
+
+	for (xip = &xfrmn->xfrmi[0];
+	     (xi = rtnl_dereference(*xip)) != NULL;
+	     xip = &xi->next) {
+		if (xi->p.if_id == p->if_id) {
+			if (create)
+				return NULL;
+
+			return xi;
+		}
+	}
+	if (!create)
+		return NULL;
+	return xfrmi_create(net, p);
+}
+
+static void xfrmi_dev_uninit(struct net_device *dev)
+{
+	struct xfrm_if *xi = netdev_priv(dev);
+	struct xfrmi_net *xfrmn = net_generic(xi->net, xfrmi_net_id);
+
+	xfrmi_unlink(xfrmn, xi);
+	dev_put(xi->phydev);
+	dev_put(dev);
+}
+
+static void xfrmi_scrub_packet(struct sk_buff *skb, bool xnet)
+{
+	skb->tstamp = 0;
+	skb->pkt_type = PACKET_HOST;
+	skb->skb_iif = 0;
+	skb->ignore_df = 0;
+	skb_dst_drop(skb);
+	nf_reset(skb);
+	nf_reset_trace(skb);
+
+	if (!xnet)
+		return;
+
+	ipvs_reset(skb);
+	secpath_reset(skb);
+	skb_orphan(skb);
+	skb->mark = 0;
+}
+
+static int xfrmi_rcv_cb(struct sk_buff *skb, int err)
+{
+	struct pcpu_sw_netstats *tstats;
+	struct xfrm_mode *inner_mode;
+	struct net_device *dev;
+	struct xfrm_state *x;
+	struct xfrm_if *xi;
+	bool xnet;
+
+	if (err && !skb->sp)
+		return 0;
+
+	x = xfrm_input_state(skb);
+
+	xi = xfrmi_lookup(xs_net(x), x);
+	if (!xi)
+		return 1;
+
+	dev = xi->dev;
+	skb->dev = dev;
+
+	if (err) {
+		dev->stats.rx_errors++;
+		dev->stats.rx_dropped++;
+
+		return 0;
+	}
+
+	xnet = !net_eq(xi->net, dev_net(skb->dev));
+
+	if (xnet) {
+		inner_mode = x->inner_mode;
+
+		if (x->sel.family == AF_UNSPEC) {
+			inner_mode = xfrm_ip2inner_mode(x, XFRM_MODE_SKB_CB(skb)->protocol);
+			if (inner_mode == NULL) {
+				XFRM_INC_STATS(dev_net(skb->dev),
+					       LINUX_MIB_XFRMINSTATEMODEERROR);
+				return -EINVAL;
+			}
+		}
+
+		if (!xfrm_policy_check(NULL, XFRM_POLICY_IN, skb,
+				       inner_mode->afinfo->family))
+			return -EPERM;
+	}
+
+	xfrmi_scrub_packet(skb, xnet);
+
+	tstats = this_cpu_ptr(dev->tstats);
+
+	u64_stats_update_begin(&tstats->syncp);
+	tstats->rx_packets++;
+	tstats->rx_bytes += skb->len;
+	u64_stats_update_end(&tstats->syncp);
+
+	return 0;
+}
+
+static int
+xfrmi_xmit2(struct sk_buff *skb, struct net_device *dev, struct flowi *fl)
+{
+	struct xfrm_if *xi = netdev_priv(dev);
+	struct net_device_stats *stats = &xi->dev->stats;
+	struct dst_entry *dst = skb_dst(skb);
+	struct net_device *tdev;
+	struct xfrm_state *x;
+	int err = -1;
+	int mtu;
+
+	if (!dst)
+		goto tx_err_link_failure;
+
+	fl->flowi_xfrm.if_id = xi->p.if_id;
+
+	dst_hold(dst);
+	dst = xfrm_lookup(xi->net, dst, fl, NULL, 0);
+	if (IS_ERR(dst)) {
+		err = PTR_ERR(dst);
+		dst = NULL;
+		goto tx_err_link_failure;
+	}
+
+	x = dst->xfrm;
+	if (!x)
+		goto tx_err_link_failure;
+
+	if (x->if_id != xi->p.if_id)
+		goto tx_err_link_failure;
+
+	tdev = dst->dev;
+
+	if (tdev == dev) {
+		stats->collisions++;
+		net_warn_ratelimited("%s: Local routing loop detected!\n",
+				     xi->p.name);
+		goto tx_err_dst_release;
+	}
+
+	mtu = dst_mtu(dst);
+	if (!skb->ignore_df && skb->len > mtu) {
+		skb_dst_update_pmtu(skb, mtu);
+
+		if (skb->protocol == htons(ETH_P_IPV6)) {
+			if (mtu < IPV6_MIN_MTU)
+				mtu = IPV6_MIN_MTU;
+
+			icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu);
+		} else {
+			icmp_send(skb, ICMP_DEST_UNREACH, ICMP_FRAG_NEEDED,
+				  htonl(mtu));
+		}
+
+		dst_release(dst);
+		return -EMSGSIZE;
+	}
+
+	xfrmi_scrub_packet(skb, !net_eq(xi->net, dev_net(dev)));
+	skb_dst_set(skb, dst);
+	skb->dev = tdev;
+
+	err = dst_output(xi->net, skb->sk, skb);
+	if (net_xmit_eval(err) == 0) {
+		struct pcpu_sw_netstats *tstats = this_cpu_ptr(dev->tstats);
+
+		u64_stats_update_begin(&tstats->syncp);
+		tstats->tx_bytes += skb->len;
+		tstats->tx_packets++;
+		u64_stats_update_end(&tstats->syncp);
+	} else {
+		stats->tx_errors++;
+		stats->tx_aborted_errors++;
+	}
+
+	return 0;
+tx_err_link_failure:
+	stats->tx_carrier_errors++;
+	dst_link_failure(skb);
+tx_err_dst_release:
+	dst_release(dst);
+	return err;
+}
+
+static netdev_tx_t xfrmi_xmit(struct sk_buff *skb, struct net_device *dev)
+{
+	struct xfrm_if *xi = netdev_priv(dev);
+	struct net_device_stats *stats = &xi->dev->stats;
+	struct flowi fl;
+	int ret;
+
+	memset(&fl, 0, sizeof(fl));
+
+	switch (skb->protocol) {
+	case htons(ETH_P_IPV6):
+		xfrm_decode_session(skb, &fl, AF_INET6);
+		memset(IP6CB(skb), 0, sizeof(*IP6CB(skb)));
+		break;
+	case htons(ETH_P_IP):
+		xfrm_decode_session(skb, &fl, AF_INET);
+		memset(IPCB(skb), 0, sizeof(*IPCB(skb)));
+		break;
+	default:
+		goto tx_err;
+	}
+
+	fl.flowi_oif = xi->phydev->ifindex;
+
+	ret = xfrmi_xmit2(skb, dev, &fl);
+	if (ret < 0)
+		goto tx_err;
+
+	return NETDEV_TX_OK;
+
+tx_err:
+	stats->tx_errors++;
+	stats->tx_dropped++;
+	kfree_skb(skb);
+	return NETDEV_TX_OK;
+}
+
+static int xfrmi4_err(struct sk_buff *skb, u32 info)
+{
+	const struct iphdr *iph = (const struct iphdr *)skb->data;
+	struct net *net = dev_net(skb->dev);
+	int protocol = iph->protocol;
+	struct ip_comp_hdr *ipch;
+	struct ip_esp_hdr *esph;
+	struct ip_auth_hdr *ah ;
+	struct xfrm_state *x;
+	struct xfrm_if *xi;
+	__be32 spi;
+
+	switch (protocol) {
+	case IPPROTO_ESP:
+		esph = (struct ip_esp_hdr *)(skb->data+(iph->ihl<<2));
+		spi = esph->spi;
+		break;
+	case IPPROTO_AH:
+		ah = (struct ip_auth_hdr *)(skb->data+(iph->ihl<<2));
+		spi = ah->spi;
+		break;
+	case IPPROTO_COMP:
+		ipch = (struct ip_comp_hdr *)(skb->data+(iph->ihl<<2));
+		spi = htonl(ntohs(ipch->cpi));
+		break;
+	default:
+		return 0;
+	}
+
+	switch (icmp_hdr(skb)->type) {
+	case ICMP_DEST_UNREACH:
+		if (icmp_hdr(skb)->code != ICMP_FRAG_NEEDED)
+			return 0;
+	case ICMP_REDIRECT:
+		break;
+	default:
+		return 0;
+	}
+
+	x = xfrm_state_lookup(net, skb->mark, (const xfrm_address_t *)&iph->daddr,
+			      spi, protocol, AF_INET);
+	if (!x)
+		return 0;
+
+	xi = xfrmi_lookup(net, x);
+	if (!xi) {
+		xfrm_state_put(x);
+		return -1;
+	}
+
+	if (icmp_hdr(skb)->type == ICMP_DEST_UNREACH)
+		ipv4_update_pmtu(skb, net, info, 0, 0, protocol, 0);
+	else
+		ipv4_redirect(skb, net, 0, 0, protocol, 0);
+	xfrm_state_put(x);
+
+	return 0;
+}
+
+static int xfrmi6_err(struct sk_buff *skb, struct inet6_skb_parm *opt,
+		    u8 type, u8 code, int offset, __be32 info)
+{
+	const struct ipv6hdr *iph = (const struct ipv6hdr *)skb->data;
+	struct net *net = dev_net(skb->dev);
+	int protocol = iph->nexthdr;
+	struct ip_comp_hdr *ipch;
+	struct ip_esp_hdr *esph;
+	struct ip_auth_hdr *ah;
+	struct xfrm_state *x;
+	struct xfrm_if *xi;
+	__be32 spi;
+
+	switch (protocol) {
+	case IPPROTO_ESP:
+		esph = (struct ip_esp_hdr *)(skb->data + offset);
+		spi = esph->spi;
+		break;
+	case IPPROTO_AH:
+		ah = (struct ip_auth_hdr *)(skb->data + offset);
+		spi = ah->spi;
+		break;
+	case IPPROTO_COMP:
+		ipch = (struct ip_comp_hdr *)(skb->data + offset);
+		spi = htonl(ntohs(ipch->cpi));
+		break;
+	default:
+		return 0;
+	}
+
+	if (type != ICMPV6_PKT_TOOBIG &&
+	    type != NDISC_REDIRECT)
+		return 0;
+
+	x = xfrm_state_lookup(net, skb->mark, (const xfrm_address_t *)&iph->daddr,
+			      spi, protocol, AF_INET6);
+	if (!x)
+		return 0;
+
+	xi = xfrmi_lookup(net, x);
+	if (!xi) {
+		xfrm_state_put(x);
+		return -1;
+	}
+
+	if (type == NDISC_REDIRECT)
+		ip6_redirect(skb, net, skb->dev->ifindex, 0,
+			     sock_net_uid(net, NULL));
+	else
+		ip6_update_pmtu(skb, net, info, 0, 0, sock_net_uid(net, NULL));
+	xfrm_state_put(x);
+
+	return 0;
+}
+
+static int xfrmi_change(struct xfrm_if *xi, const struct xfrm_if_parms *p)
+{
+	if (xi->p.link != p->link)
+		return -EINVAL;
+
+	xi->p.if_id = p->if_id;
+
+	return 0;
+}
+
+static int xfrmi_update(struct xfrm_if *xi, struct xfrm_if_parms *p)
+{
+	struct net *net = dev_net(xi->dev);
+	struct xfrmi_net *xfrmn = net_generic(net, xfrmi_net_id);
+	int err;
+
+	xfrmi_unlink(xfrmn, xi);
+	synchronize_net();
+	err = xfrmi_change(xi, p);
+	xfrmi_link(xfrmn, xi);
+	netdev_state_change(xi->dev);
+	return err;
+}
+
+static void xfrmi_get_stats64(struct net_device *dev,
+			       struct rtnl_link_stats64 *s)
+{
+	int cpu;
+
+	if (!dev->tstats)
+		return;
+
+	for_each_possible_cpu(cpu) {
+		struct pcpu_sw_netstats *stats;
+		struct pcpu_sw_netstats tmp;
+		int start;
+
+		stats = per_cpu_ptr(dev->tstats, cpu);
+		do {
+			start = u64_stats_fetch_begin_irq(&stats->syncp);
+			tmp.rx_packets = stats->rx_packets;
+			tmp.rx_bytes   = stats->rx_bytes;
+			tmp.tx_packets = stats->tx_packets;
+			tmp.tx_bytes   = stats->tx_bytes;
+		} while (u64_stats_fetch_retry_irq(&stats->syncp, start));
+
+		s->rx_packets += tmp.rx_packets;
+		s->rx_bytes   += tmp.rx_bytes;
+		s->tx_packets += tmp.tx_packets;
+		s->tx_bytes   += tmp.tx_bytes;
+	}
+
+	s->rx_dropped = dev->stats.rx_dropped;
+	s->tx_dropped = dev->stats.tx_dropped;
+}
+
+static int xfrmi_get_iflink(const struct net_device *dev)
+{
+	struct xfrm_if *xi = netdev_priv(dev);
+
+	return xi->phydev->ifindex;
+}
+
+
+static const struct net_device_ops xfrmi_netdev_ops = {
+	.ndo_init	= xfrmi_dev_init,
+	.ndo_uninit	= xfrmi_dev_uninit,
+	.ndo_start_xmit = xfrmi_xmit,
+	.ndo_get_stats64 = xfrmi_get_stats64,
+	.ndo_get_iflink = xfrmi_get_iflink,
+};
+
+static void xfrmi_dev_setup(struct net_device *dev)
+{
+	dev->netdev_ops 	= &xfrmi_netdev_ops;
+	dev->type		= ARPHRD_NONE;
+	dev->hard_header_len 	= ETH_HLEN;
+	dev->min_header_len	= ETH_HLEN;
+	dev->mtu		= ETH_DATA_LEN;
+	dev->min_mtu		= ETH_MIN_MTU;
+	dev->max_mtu		= ETH_DATA_LEN;
+	dev->addr_len		= ETH_ALEN;
+	dev->flags 		= IFF_NOARP;
+	dev->needs_free_netdev	= true;
+	dev->priv_destructor	= xfrmi_dev_free;
+	netif_keep_dst(dev);
+}
+
+static int xfrmi_dev_init(struct net_device *dev)
+{
+	struct xfrm_if *xi = netdev_priv(dev);
+	struct net_device *phydev = xi->phydev;
+	int err;
+
+	dev->tstats = netdev_alloc_pcpu_stats(struct pcpu_sw_netstats);
+	if (!dev->tstats)
+		return -ENOMEM;
+
+	err = gro_cells_init(&xi->gro_cells, dev);
+	if (err) {
+		free_percpu(dev->tstats);
+		return err;
+	}
+
+	dev->features |= NETIF_F_LLTX;
+
+	dev->needed_headroom = phydev->needed_headroom;
+	dev->needed_tailroom = phydev->needed_tailroom;
+
+	if (is_zero_ether_addr(dev->dev_addr))
+		eth_hw_addr_inherit(dev, phydev);
+	if (is_zero_ether_addr(dev->broadcast))
+		memcpy(dev->broadcast, phydev->broadcast, dev->addr_len);
+
+	return 0;
+}
+
+static int xfrmi_validate(struct nlattr *tb[], struct nlattr *data[],
+			 struct netlink_ext_ack *extack)
+{
+	return 0;
+}
+
+static void xfrmi_netlink_parms(struct nlattr *data[],
+			       struct xfrm_if_parms *parms)
+{
+	memset(parms, 0, sizeof(*parms));
+
+	if (!data)
+		return;
+
+	if (data[IFLA_XFRM_LINK])
+		parms->link = nla_get_u32(data[IFLA_XFRM_LINK]);
+
+	if (data[IFLA_XFRM_IF_ID])
+		parms->if_id = nla_get_u32(data[IFLA_XFRM_IF_ID]);
+}
+
+static int xfrmi_newlink(struct net *src_net, struct net_device *dev,
+			struct nlattr *tb[], struct nlattr *data[],
+			struct netlink_ext_ack *extack)
+{
+	struct net *net = dev_net(dev);
+	struct xfrm_if_parms *p;
+	struct xfrm_if *xi;
+
+	xi = netdev_priv(dev);
+	p = &xi->p;
+
+	xfrmi_netlink_parms(data, p);
+
+	if (!tb[IFLA_IFNAME])
+		return -EINVAL;
+
+	nla_strlcpy(p->name, tb[IFLA_IFNAME], IFNAMSIZ);
+
+	if (!xfrmi_locate(net, p, 1))
+		return -EEXIST;
+
+	return 0;
+}
+
+static void xfrmi_dellink(struct net_device *dev, struct list_head *head)
+{
+	unregister_netdevice_queue(dev, head);
+}
+
+static int xfrmi_changelink(struct net_device *dev, struct nlattr *tb[],
+			   struct nlattr *data[],
+			   struct netlink_ext_ack *extack)
+{
+	struct xfrm_if *xi = netdev_priv(dev);
+	struct net *net = dev_net(dev);
+
+	xfrmi_netlink_parms(data, &xi->p);
+
+	xi = xfrmi_locate(net, &xi->p, 0);
+
+	if (xi) {
+		if (xi->dev != dev)
+			return -EEXIST;
+	} else
+		xi = netdev_priv(dev);
+
+	return xfrmi_update(xi, &xi->p);
+}
+
+static size_t xfrmi_get_size(const struct net_device *dev)
+{
+	return
+		/* IFLA_XFRM_LINK */
+		nla_total_size(4) +
+		/* IFLA_XFRM_IF_ID */
+		nla_total_size(4) +
+		0;
+}
+
+static int xfrmi_fill_info(struct sk_buff *skb, const struct net_device *dev)
+{
+	struct xfrm_if *xi = netdev_priv(dev);
+	struct xfrm_if_parms *parm = &xi->p;
+
+	if (nla_put_u32(skb, IFLA_XFRM_LINK, parm->link) ||
+	    nla_put_u32(skb, IFLA_XFRM_IF_ID, parm->if_id))
+		goto nla_put_failure;
+	return 0;
+
+nla_put_failure:
+	return -EMSGSIZE;
+}
+
+struct net *xfrmi_get_link_net(const struct net_device *dev)
+{
+	struct xfrm_if *xi = netdev_priv(dev);
+
+	return dev_net(xi->phydev);
+}
+
+static const struct nla_policy xfrmi_policy[IFLA_XFRM_MAX + 1] = {
+	[IFLA_XFRM_LINK]	= { .type = NLA_U32 },
+	[IFLA_XFRM_IF_ID]	= { .type = NLA_U32 },
+};
+
+static struct rtnl_link_ops xfrmi_link_ops __read_mostly = {
+	.kind		= "xfrm",
+	.maxtype	= IFLA_XFRM_MAX,
+	.policy		= xfrmi_policy,
+	.priv_size	= sizeof(struct xfrm_if),
+	.setup		= xfrmi_dev_setup,
+	.validate	= xfrmi_validate,
+	.newlink	= xfrmi_newlink,
+	.dellink	= xfrmi_dellink,
+	.changelink	= xfrmi_changelink,
+	.get_size	= xfrmi_get_size,
+	.fill_info	= xfrmi_fill_info,
+	.get_link_net	= xfrmi_get_link_net,
+};
+
+static void __net_exit xfrmi_destroy_interfaces(struct xfrmi_net *xfrmn)
+{
+	struct xfrm_if *xi;
+	LIST_HEAD(list);
+
+	xi = rtnl_dereference(xfrmn->xfrmi[0]);
+	if (!xi)
+		return;
+
+	unregister_netdevice_queue(xi->dev, &list);
+	unregister_netdevice_many(&list);
+}
+
+static int __net_init xfrmi_init_net(struct net *net)
+{
+	return 0;
+}
+
+static void __net_exit xfrmi_exit_net(struct net *net)
+{
+	struct xfrmi_net *xfrmn = net_generic(net, xfrmi_net_id);
+
+	rtnl_lock();
+	xfrmi_destroy_interfaces(xfrmn);
+	rtnl_unlock();
+}
+
+static struct pernet_operations xfrmi_net_ops = {
+	.init = xfrmi_init_net,
+	.exit = xfrmi_exit_net,
+	.id   = &xfrmi_net_id,
+	.size = sizeof(struct xfrmi_net),
+};
+
+static struct xfrm6_protocol xfrmi_esp6_protocol __read_mostly = {
+	.handler	=	xfrm6_rcv,
+	.cb_handler	=	xfrmi_rcv_cb,
+	.err_handler	=	xfrmi6_err,
+	.priority	=	10,
+};
+
+static struct xfrm6_protocol xfrmi_ah6_protocol __read_mostly = {
+	.handler	=	xfrm6_rcv,
+	.cb_handler	=	xfrmi_rcv_cb,
+	.err_handler	=	xfrmi6_err,
+	.priority	=	10,
+};
+
+static struct xfrm6_protocol xfrmi_ipcomp6_protocol __read_mostly = {
+	.handler	=	xfrm6_rcv,
+	.cb_handler	=	xfrmi_rcv_cb,
+	.err_handler	=	xfrmi6_err,
+	.priority	=	10,
+};
+
+static struct xfrm4_protocol xfrmi_esp4_protocol __read_mostly = {
+	.handler	=	xfrm4_rcv,
+	.input_handler	=	xfrm_input,
+	.cb_handler	=	xfrmi_rcv_cb,
+	.err_handler	=	xfrmi4_err,
+	.priority	=	10,
+};
+
+static struct xfrm4_protocol xfrmi_ah4_protocol __read_mostly = {
+	.handler	=	xfrm4_rcv,
+	.input_handler	=	xfrm_input,
+	.cb_handler	=	xfrmi_rcv_cb,
+	.err_handler	=	xfrmi4_err,
+	.priority	=	10,
+};
+
+static struct xfrm4_protocol xfrmi_ipcomp4_protocol __read_mostly = {
+	.handler	=	xfrm4_rcv,
+	.input_handler	=	xfrm_input,
+	.cb_handler	=	xfrmi_rcv_cb,
+	.err_handler	=	xfrmi4_err,
+	.priority	=	10,
+};
+
+static int __init xfrmi4_init(void)
+{
+	int err;
+
+	err = xfrm4_protocol_register(&xfrmi_esp4_protocol, IPPROTO_ESP);
+	if (err < 0)
+		goto xfrm_proto_esp_failed;
+	err = xfrm4_protocol_register(&xfrmi_ah4_protocol, IPPROTO_AH);
+	if (err < 0)
+		goto xfrm_proto_ah_failed;
+	err = xfrm4_protocol_register(&xfrmi_ipcomp4_protocol, IPPROTO_COMP);
+	if (err < 0)
+		goto xfrm_proto_comp_failed;
+
+	return 0;
+
+xfrm_proto_comp_failed:
+	xfrm4_protocol_deregister(&xfrmi_ah4_protocol, IPPROTO_AH);
+xfrm_proto_ah_failed:
+	xfrm4_protocol_deregister(&xfrmi_esp4_protocol, IPPROTO_ESP);
+xfrm_proto_esp_failed:
+	return err;
+}
+
+static void xfrmi4_fini(void)
+{
+	xfrm4_protocol_deregister(&xfrmi_ipcomp4_protocol, IPPROTO_COMP);
+	xfrm4_protocol_deregister(&xfrmi_ah4_protocol, IPPROTO_AH);
+	xfrm4_protocol_deregister(&xfrmi_esp4_protocol, IPPROTO_ESP);
+}
+
+static int __init xfrmi6_init(void)
+{
+	int err;
+
+	err = xfrm6_protocol_register(&xfrmi_esp6_protocol, IPPROTO_ESP);
+	if (err < 0)
+		goto xfrm_proto_esp_failed;
+	err = xfrm6_protocol_register(&xfrmi_ah6_protocol, IPPROTO_AH);
+	if (err < 0)
+		goto xfrm_proto_ah_failed;
+	err = xfrm6_protocol_register(&xfrmi_ipcomp6_protocol, IPPROTO_COMP);
+	if (err < 0)
+		goto xfrm_proto_comp_failed;
+
+	return 0;
+
+xfrm_proto_comp_failed:
+	xfrm6_protocol_deregister(&xfrmi_ah6_protocol, IPPROTO_AH);
+xfrm_proto_ah_failed:
+	xfrm6_protocol_deregister(&xfrmi_esp6_protocol, IPPROTO_ESP);
+xfrm_proto_esp_failed:
+	return err;
+}
+
+static void xfrmi6_fini(void)
+{
+	xfrm6_protocol_deregister(&xfrmi_ipcomp6_protocol, IPPROTO_COMP);
+	xfrm6_protocol_deregister(&xfrmi_ah6_protocol, IPPROTO_AH);
+	xfrm6_protocol_deregister(&xfrmi_esp6_protocol, IPPROTO_ESP);
+}
+
+static const struct xfrm_if_cb xfrm_if_cb = {
+	.decode_session =	xfrmi_decode_session,
+};
+
+static int __init xfrmi_init(void)
+{
+	const char *msg;
+	int err;
+
+	pr_info("IPsec XFRM device driver\n");
+
+	msg = "tunnel device";
+	err = register_pernet_device(&xfrmi_net_ops);
+	if (err < 0)
+		goto pernet_dev_failed;
+
+	msg = "xfrm4 protocols";
+	err = xfrmi4_init();
+	if (err < 0)
+		goto xfrmi4_failed;
+
+	msg = "xfrm6 protocols";
+	err = xfrmi6_init();
+	if (err < 0)
+		goto xfrmi6_failed;
+
+
+	msg = "netlink interface";
+	err = rtnl_link_register(&xfrmi_link_ops);
+	if (err < 0)
+		goto rtnl_link_failed;
+
+	xfrm_if_register_cb(&xfrm_if_cb);
+
+	return err;
+
+rtnl_link_failed:
+	xfrmi6_fini();
+xfrmi6_failed:
+	xfrmi4_fini();
+xfrmi4_failed:
+	unregister_pernet_device(&xfrmi_net_ops);
+pernet_dev_failed:
+	pr_err("xfrmi init: failed to register %s\n", msg);
+	return err;
+}
+
+static void __exit xfrmi_fini(void)
+{
+	xfrm_if_unregister_cb();
+	rtnl_link_unregister(&xfrmi_link_ops);
+	xfrmi4_fini();
+	xfrmi6_fini();
+	unregister_pernet_device(&xfrmi_net_ops);
+}
+
+module_init(xfrmi_init);
+module_exit(xfrmi_fini);
+MODULE_LICENSE("GPL");
+MODULE_ALIAS_RTNL_LINK("xfrm");
+MODULE_ALIAS_NETDEV("xfrm0");
+MODULE_AUTHOR("Steffen Klassert");
+MODULE_DESCRIPTION("XFRM virtual interface");
diff --git a/net/xfrm/xfrm_policy.c b/net/xfrm/xfrm_policy.c
index 08b745aade43..47f776840df1 100644
--- a/net/xfrm/xfrm_policy.c
+++ b/net/xfrm/xfrm_policy.c
@@ -47,6 +47,9 @@ struct xfrm_flo {
 
 static DEFINE_PER_CPU(struct xfrm_dst *, xfrm_last_dst);
 static struct work_struct *xfrm_pcpu_work __read_mostly;
+static DEFINE_SPINLOCK(xfrm_if_cb_lock);
+static struct xfrm_if_cb const __rcu *xfrm_if_cb __read_mostly;
+
 static DEFINE_SPINLOCK(xfrm_policy_afinfo_lock);
 static struct xfrm_policy_afinfo const __rcu *xfrm_policy_afinfo[AF_INET6 + 1]
 						__read_mostly;
@@ -119,6 +122,12 @@ static const struct xfrm_policy_afinfo *xfrm_policy_get_afinfo(unsigned short fa
 	return afinfo;
 }
 
+/* Called with rcu_read_lock(). */
+static const struct xfrm_if_cb *xfrm_if_get_cb(void)
+{
+	return rcu_dereference(xfrm_if_cb);
+}
+
 struct dst_entry *__xfrm_dst_lookup(struct net *net, int tos, int oif,
 				    const xfrm_address_t *saddr,
 				    const xfrm_address_t *daddr,
@@ -2083,6 +2092,11 @@ xfrm_bundle_lookup(struct net *net, const struct flowi *fl, u16 family, u8 dir,
 
 	if (IS_ERR(xdst)) {
 		err = PTR_ERR(xdst);
+		if (err == -EREMOTE) {
+			xfrm_pols_put(pols, num_pols);
+			return NULL;
+		}
+
 		if (err != -EAGAIN)
 			goto error;
 		goto make_dummy_bundle;
@@ -2176,6 +2190,9 @@ struct dst_entry *xfrm_lookup(struct net *net, struct dst_entry *dst_orig,
 			if (IS_ERR(xdst)) {
 				xfrm_pols_put(pols, num_pols);
 				err = PTR_ERR(xdst);
+				if (err == -EREMOTE)
+					goto nopol;
+
 				goto dropdst;
 			} else if (xdst == NULL) {
 				num_xfrms = 0;
@@ -2368,12 +2385,20 @@ 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);
+	const struct xfrm_if_cb *ifcb = xfrm_if_get_cb();
+	struct xfrm_if *xi;
 	int err;
 
 	if (unlikely(afinfo == NULL))
 		return -EAFNOSUPPORT;
 
 	afinfo->decode_session(skb, fl, reverse);
+	if (ifcb) {
+		xi = ifcb->decode_session(skb);
+		if (xi)
+			fl->flowi_xfrm.if_id = xi->p.if_id;
+	}
+
 	err = security_xfrm_decode_session(skb, &fl->flowi_secid);
 	rcu_read_unlock();
 	return err;
@@ -2828,6 +2853,21 @@ void xfrm_policy_unregister_afinfo(const struct xfrm_policy_afinfo *afinfo)
 }
 EXPORT_SYMBOL(xfrm_policy_unregister_afinfo);
 
+void xfrm_if_register_cb(const struct xfrm_if_cb *ifcb)
+{
+	spin_lock(&xfrm_if_cb_lock);
+	rcu_assign_pointer(xfrm_if_cb, ifcb);
+	spin_unlock(&xfrm_if_cb_lock);
+}
+EXPORT_SYMBOL(xfrm_if_register_cb);
+
+void xfrm_if_unregister_cb(void)
+{
+	RCU_INIT_POINTER(xfrm_if_cb, NULL);
+	synchronize_rcu();
+}
+EXPORT_SYMBOL(xfrm_if_unregister_cb);
+
 #ifdef CONFIG_XFRM_STATISTICS
 static int __net_init xfrm_statistics_init(struct net *net)
 {
@@ -3008,6 +3048,9 @@ void __init xfrm_init(void)
 	xfrm_dev_init();
 	seqcount_init(&xfrm_policy_hash_generation);
 	xfrm_input_init();
+
+	RCU_INIT_POINTER(xfrm_if_cb, NULL);
+	synchronize_rcu();
 }
 
 #ifdef CONFIG_AUDITSYSCALL
-- 
2.14.1

Re: [PATCH RFC ipsec-next 0/3] Virtual xfrm interfaces

[David Miller]

From: Steffen Klassert <steffen.klassert@secunet.com>
Date: Mon, 4 Jun 2018 08:09:07 +0200

> This patchset introduces new virtual xfrm interfaces.
> The design of virtual xfrm interfaces interfaces was
> discussed at the Linux IPsec workshop 2018. This patchset
> implements these interfaces as the IPsec userspace and
> kernel developers agreed. The purpose of these interfaces
> is to overcome the design limitations that the existing
> VTI devices have.
> 
> We had two presentations about xfrm interfaces at
> the workshop. Slides with further informations
> can be found at the workshop homepage:
> 
> https://workshop.linux-ipsec.org/2018/

First off, you will have to describe in detail what the VTI
limitations are and how these new devices overcome them in this commit
message.

You can't just say "we discussed this over there, go take a look".

The place people "take a look" is your text here.

Second, since you didn't explain things, I have to ask.  Why is a new
special ID even necessary?  It makes the flowi bigger, and adds all of
this new logic.

All netdevs have an ifindex and you should be able to find a way to
use the ifindex of these new devices in the key somehow.

Thanks.

Re: [PATCH RFC ipsec-next 0/3] Virtual xfrm interfaces

[Steffen Klassert]

On Mon, Jun 04, 2018 at 08:58:22AM -0400, David Miller wrote:
> From: Steffen Klassert <steffen.klassert@secunet.com>
> Date: Mon, 4 Jun 2018 08:09:07 +0200
> 
> > This patchset introduces new virtual xfrm interfaces.
> > The design of virtual xfrm interfaces interfaces was
> > discussed at the Linux IPsec workshop 2018. This patchset
> > implements these interfaces as the IPsec userspace and
> > kernel developers agreed. The purpose of these interfaces
> > is to overcome the design limitations that the existing
> > VTI devices have.
> > 
> > We had two presentations about xfrm interfaces at
> > the workshop. Slides with further informations
> > can be found at the workshop homepage:
> > 
> > https://workshop.linux-ipsec.org/2018/
> 
> First off, you will have to describe in detail what the VTI
> limitations are and how these new devices overcome them in this commit
> message.

Ok, no problem. Sorry for not doing this in the initial mail.

The main limitations that we see with the current VTI are the following:

- VTI interfaces are L3 tunnels with configurable endpoints.
  For xfrm, the tunnel endpoint are already determined by the SA.
  So the VTI tunnel endpoints must be either the same as on the
  SA or wildcards. In case VTI tunnel endpoints are same as on
  the SA, we get a one to one correlation between the SA and
  the tunnel. So each SA needs its own tunnel interface.

  On the other hand, we can have only one VTI tunnel with
  wildcard src/dst tunnel endpoints in the system because the
  lookup is based on the tunnel endpoints. The existing tunnel
  lookup won't work with multiple tunnels with wildcard
  tunnel endpoints. Some usecases require more than on
  VTI tunnel of this type, for example if somebody has multiple
  namespaces and every namespace requires such a VTI.

- VTI needs separate interfaces for IPv4 and IPv6 tunnels.
  So when routing to a VTI, we have to know to which address
  family this traffic class is going to be encapsulated.
  This is a lmitation because it makes routing more complex
  and it is not always possible to know what happens behind the
  VTI, e.g. when the VTI is move to some namespace.

- VTI works just with tunnel mode SAs. We need generic interfaces
  that ensures transfomation, regardless of the xfrm mode and
  the encapsulated address family.

- VTI is configured with a combination GRE keys and xfrm marks.
  With this we have to deal with some extra cases in the generic
  tunnel lookup because the GRE keys on the VTI are actually
  not GRE keys, the GRE keys were just reused for something else.
  All extensions to the VTI interfaces would require to add
  even more complexity to the generic tunnel lookup.

So to overcome this, we started with the following design goal:

- It should be possible to tunnel IPv4 and IPv6 through the same
  interface.

- No limitation on xfrm mode (tunnel, transport and beet).

- Should be a generic virtual interface that ensures IPsec
  transformation, no need to know what happens behind the
  interface.

- Interfaces should be configured with a new key that must match a
  new policy/SA lookup key.

- The lookup logic should stay in the xfrm codebase, no need to
  change or extend generic routing and tunnel lookups.

- Should be possible to use IPsec hardware offloads of the underlying
  interface.

> 
> You can't just say "we discussed this over there, go take a look".
> 
> The place people "take a look" is your text here.
> 
> Second, since you didn't explain things, I have to ask.  Why is a new
> special ID even necessary?  It makes the flowi bigger, and adds all of
> this new logic.
> 
> All netdevs have an ifindex and you should be able to find a way to
> use the ifindex of these new devices in the key somehow.

The ifindex is already part of the xfrm selector and used as a key
to enforce the output device after the transformation in the
policy/state lookup. So we need some other identifier as the
ifindex is already in use for xfrm lookups.

We tried to keep the lookup logic in xfrm as much as possible,
the only thing that we currently need is a field in the flowi
for this to pass the information down to the lookups.