Merge tag 'ipsec-next-2025-03-24' of git://git.kernel.org/pub/scm/linux/kernel/git/klassert/ipsec-next

Steffen Klassert says:

====================
pull request (net-next): ipsec-next 2025-03-24

1) Prevent setting high order sequence number bits input in
   non-ESN mode. From Leon Romanovsky.

2) Support PMTU handling in tunnel mode for packet offload.
   From Leon Romanovsky.

3) Make xfrm_state_lookup_byaddr lockless.
   From Florian Westphal.

4) Remove unnecessary NULL check in xfrm_lookup_with_ifid().
   From Dan Carpenter.

* tag 'ipsec-next-2025-03-24' of git://git.kernel.org/pub/scm/linux/kernel/git/klassert/ipsec-next:
  xfrm: Remove unnecessary NULL check in xfrm_lookup_with_ifid()
  xfrm: state: make xfrm_state_lookup_byaddr lockless
  xfrm: check for PMTU in tunnel mode for packet offload
  xfrm: provide common xdo_dev_offload_ok callback implementation
  xfrm: rely on XFRM offload
  xfrm: simplify SA initialization routine
  xfrm: delay initialization of offload path till its actually requested
  xfrm: prevent high SEQ input in non-ESN mode
====================

Link: https://patch.msgid.link/20250324061855.4116819-1-steffen.klassert@secunet.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

5 files changed, 76 insertions, 46 deletions

diff --git a/net/xfrm/xfrm_device.c b/net/xfrm/xfrm_device.c
index d1fa94e52cea..d62f76161d83 100644
--- a/net/xfrm/xfrm_device.c
+++ b/net/xfrm/xfrm_device.c
@@ -244,11 +244,6 @@ int xfrm_dev_state_add(struct net *net, struct xfrm_state *x,
 	xfrm_address_t *daddr;
 	bool is_packet_offload;
 
-	if (!x->type_offload) {
-		NL_SET_ERR_MSG(extack, "Type doesn't support offload");
-		return -EINVAL;
-	}
-
 	if (xuo->flags &
 	    ~(XFRM_OFFLOAD_IPV6 | XFRM_OFFLOAD_INBOUND | XFRM_OFFLOAD_PACKET)) {
 		NL_SET_ERR_MSG(extack, "Unrecognized flags in offload request");
@@ -310,6 +305,13 @@ int xfrm_dev_state_add(struct net *net, struct xfrm_state *x,
 		return -EINVAL;
 	}
 
+	xfrm_set_type_offload(x);
+	if (!x->type_offload) {
+		NL_SET_ERR_MSG(extack, "Type doesn't support offload");
+		dev_put(dev);
+		return -EINVAL;
+	}
+
 	xso->dev = dev;
 	netdev_tracker_alloc(dev, &xso->dev_tracker, GFP_ATOMIC);
 	xso->real_dev = dev;
@@ -332,6 +334,7 @@ int xfrm_dev_state_add(struct net *net, struct xfrm_state *x,
 		netdev_put(dev, &xso->dev_tracker);
 		xso->type = XFRM_DEV_OFFLOAD_UNSPECIFIED;
 
+		xfrm_unset_type_offload(x);
 		/* User explicitly requested packet offload mode and configured
 		 * policy in addition to the XFRM state. So be civil to users,
 		 * and return an error instead of taking fallback path.
@@ -415,14 +418,12 @@ bool xfrm_dev_offload_ok(struct sk_buff *skb, struct xfrm_state *x)
 	struct dst_entry *dst = skb_dst(skb);
 	struct xfrm_dst *xdst = (struct xfrm_dst *)dst;
 	struct net_device *dev = x->xso.dev;
+	bool check_tunnel_size;
 
-	if (!x->type_offload ||
-	    (x->xso.type == XFRM_DEV_OFFLOAD_UNSPECIFIED && x->encap))
+	if (x->xso.type == XFRM_DEV_OFFLOAD_UNSPECIFIED)
 		return false;
 
-	if (x->xso.type == XFRM_DEV_OFFLOAD_PACKET ||
-	    ((!dev || (dev == xfrm_dst_path(dst)->dev)) &&
-	     !xdst->child->xfrm)) {
+	if ((dev == xfrm_dst_path(dst)->dev) && !xdst->child->xfrm) {
 		mtu = xfrm_state_mtu(x, xdst->child_mtu_cached);
 		if (skb->len <= mtu)
 			goto ok;
@@ -434,8 +435,29 @@ bool xfrm_dev_offload_ok(struct sk_buff *skb, struct xfrm_state *x)
 	return false;
 
 ok:
-	if (dev && dev->xfrmdev_ops && dev->xfrmdev_ops->xdo_dev_offload_ok)
-		return x->xso.dev->xfrmdev_ops->xdo_dev_offload_ok(skb, x);
+	check_tunnel_size = x->xso.type == XFRM_DEV_OFFLOAD_PACKET &&
+			    x->props.mode == XFRM_MODE_TUNNEL;
+	switch (x->props.family) {
+	case AF_INET:
+		/* Check for IPv4 options */
+		if (ip_hdr(skb)->ihl != 5)
+			return false;
+		if (check_tunnel_size && xfrm4_tunnel_check_size(skb))
+			return false;
+		break;
+	case AF_INET6:
+		/* Check for IPv6 extensions */
+		if (ipv6_ext_hdr(ipv6_hdr(skb)->nexthdr))
+			return false;
+		if (check_tunnel_size && xfrm6_tunnel_check_size(skb))
+			return false;
+		break;
+	default:
+		break;
+	}
+
+	if (dev->xfrmdev_ops->xdo_dev_offload_ok)
+		return dev->xfrmdev_ops->xdo_dev_offload_ok(skb, x);
 
 	return true;
 }
diff --git a/net/xfrm/xfrm_output.c b/net/xfrm/xfrm_output.c
index 3cabc87978dd..9077730ff7d0 100644
--- a/net/xfrm/xfrm_output.c
+++ b/net/xfrm/xfrm_output.c
@@ -827,7 +827,7 @@ out:
 }
 EXPORT_SYMBOL_GPL(xfrm_output);
 
-static int xfrm4_tunnel_check_size(struct sk_buff *skb)
+int xfrm4_tunnel_check_size(struct sk_buff *skb)
 {
 	int mtu, ret = 0;
 
@@ -853,6 +853,7 @@ static int xfrm4_tunnel_check_size(struct sk_buff *skb)
 out:
 	return ret;
 }
+EXPORT_SYMBOL_GPL(xfrm4_tunnel_check_size);
 
 static int xfrm4_extract_output(struct xfrm_state *x, struct sk_buff *skb)
 {
@@ -875,7 +876,7 @@ static int xfrm4_extract_output(struct xfrm_state *x, struct sk_buff *skb)
 }
 
 #if IS_ENABLED(CONFIG_IPV6)
-static int xfrm6_tunnel_check_size(struct sk_buff *skb)
+int xfrm6_tunnel_check_size(struct sk_buff *skb)
 {
 	int mtu, ret = 0;
 	struct dst_entry *dst = skb_dst(skb);
@@ -905,6 +906,7 @@ static int xfrm6_tunnel_check_size(struct sk_buff *skb)
 out:
 	return ret;
 }
+EXPORT_SYMBOL_GPL(xfrm6_tunnel_check_size);
 #endif
 
 static int xfrm6_extract_output(struct xfrm_state *x, struct sk_buff *skb)
diff --git a/net/xfrm/xfrm_policy.c b/net/xfrm/xfrm_policy.c
index 6551e588fe52..30970d40a454 100644
--- a/net/xfrm/xfrm_policy.c
+++ b/net/xfrm/xfrm_policy.c
@@ -3294,7 +3294,7 @@ no_transform:
 
 ok:
 	xfrm_pols_put(pols, drop_pols);
-	if (dst && dst->xfrm &&
+	if (dst->xfrm &&
 	    (dst->xfrm->props.mode == XFRM_MODE_TUNNEL ||
 	     dst->xfrm->props.mode == XFRM_MODE_IPTFS))
 		dst->flags |= DST_XFRM_TUNNEL;
diff --git a/net/xfrm/xfrm_state.c b/net/xfrm/xfrm_state.c
index ad2202fa82f3..07545944a536 100644
--- a/net/xfrm/xfrm_state.c
+++ b/net/xfrm/xfrm_state.c
@@ -424,18 +424,18 @@ void xfrm_unregister_type_offload(const struct xfrm_type_offload *type,
 }
 EXPORT_SYMBOL(xfrm_unregister_type_offload);
 
-static const struct xfrm_type_offload *
-xfrm_get_type_offload(u8 proto, unsigned short family, bool try_load)
+void xfrm_set_type_offload(struct xfrm_state *x)
 {
 	const struct xfrm_type_offload *type = NULL;
 	struct xfrm_state_afinfo *afinfo;
+	bool try_load = true;
 
 retry:
-	afinfo = xfrm_state_get_afinfo(family);
+	afinfo = xfrm_state_get_afinfo(x->props.family);
 	if (unlikely(afinfo == NULL))
-		return NULL;
+		goto out;
 
-	switch (proto) {
+	switch (x->id.proto) {
 	case IPPROTO_ESP:
 		type = afinfo->type_offload_esp;
 		break;
@@ -449,18 +449,16 @@ retry:
 	rcu_read_unlock();
 
 	if (!type && try_load) {
-		request_module("xfrm-offload-%d-%d", family, proto);
+		request_module("xfrm-offload-%d-%d", x->props.family,
+			       x->id.proto);
 		try_load = false;
 		goto retry;
 	}
 
-	return type;
-}
-
-static void xfrm_put_type_offload(const struct xfrm_type_offload *type)
-{
-	module_put(type->owner);
+out:
+	x->type_offload = type;
 }
+EXPORT_SYMBOL(xfrm_set_type_offload);
 
 static const struct xfrm_mode xfrm4_mode_map[XFRM_MODE_MAX] = {
 	[XFRM_MODE_BEET] = {
@@ -609,8 +607,6 @@ static void ___xfrm_state_destroy(struct xfrm_state *x)
 	kfree(x->coaddr);
 	kfree(x->replay_esn);
 	kfree(x->preplay_esn);
-	if (x->type_offload)
-		xfrm_put_type_offload(x->type_offload);
 	if (x->type) {
 		x->type->destructor(x);
 		xfrm_put_type(x->type);
@@ -784,6 +780,8 @@ void xfrm_dev_state_free(struct xfrm_state *x)
 	struct xfrm_dev_offload *xso = &x->xso;
 	struct net_device *dev = READ_ONCE(xso->dev);
 
+	xfrm_unset_type_offload(x);
+
 	if (dev && dev->xfrmdev_ops) {
 		spin_lock_bh(&xfrm_state_dev_gc_lock);
 		if (!hlist_unhashed(&x->dev_gclist))
@@ -2315,12 +2313,12 @@ xfrm_state_lookup_byaddr(struct net *net, u32 mark,
 	struct xfrm_hash_state_ptrs state_ptrs;
 	struct xfrm_state *x;
 
-	spin_lock_bh(&net->xfrm.xfrm_state_lock);
+	rcu_read_lock();
 
 	xfrm_hash_ptrs_get(net, &state_ptrs);
 
 	x = __xfrm_state_lookup_byaddr(&state_ptrs, mark, daddr, saddr, proto, family);
-	spin_unlock_bh(&net->xfrm.xfrm_state_lock);
+	rcu_read_unlock();
 	return x;
 }
 EXPORT_SYMBOL(xfrm_state_lookup_byaddr);
@@ -3122,8 +3120,7 @@ u32 xfrm_state_mtu(struct xfrm_state *x, int mtu)
 }
 EXPORT_SYMBOL_GPL(xfrm_state_mtu);
 
-int __xfrm_init_state(struct xfrm_state *x, bool init_replay, bool offload,
-		      struct netlink_ext_ack *extack)
+int __xfrm_init_state(struct xfrm_state *x, struct netlink_ext_ack *extack)
 {
 	const struct xfrm_mode *inner_mode;
 	const struct xfrm_mode *outer_mode;
@@ -3178,8 +3175,6 @@ int __xfrm_init_state(struct xfrm_state *x, bool init_replay, bool offload,
 		goto error;
 	}
 
-	x->type_offload = xfrm_get_type_offload(x->id.proto, family, offload);
-
 	err = x->type->init_state(x, extack);
 	if (err)
 		goto error;
@@ -3192,12 +3187,6 @@ int __xfrm_init_state(struct xfrm_state *x, bool init_replay, bool offload,
 	}
 
 	x->outer_mode = *outer_mode;
-	if (init_replay) {
-		err = xfrm_init_replay(x, extack);
-		if (err)
-			goto error;
-	}
-
 	if (x->nat_keepalive_interval) {
 		if (x->dir != XFRM_SA_DIR_OUT) {
 			NL_SET_ERR_MSG(extack, "NAT keepalive is only supported for outbound SAs");
@@ -3229,11 +3218,16 @@ int xfrm_init_state(struct xfrm_state *x)
 {
 	int err;
 
-	err = __xfrm_init_state(x, true, false, NULL);
-	if (!err)
-		x->km.state = XFRM_STATE_VALID;
+	err = __xfrm_init_state(x, NULL);
+	if (err)
+		return err;
 
-	return err;
+	err = xfrm_init_replay(x, NULL);
+	if (err)
+		return err;
+
+	x->km.state = XFRM_STATE_VALID;
+	return 0;
 }
 
 EXPORT_SYMBOL(xfrm_init_state);
diff --git a/net/xfrm/xfrm_user.c b/net/xfrm/xfrm_user.c
index 08c6d6f0179f..784a2d124749 100644
--- a/net/xfrm/xfrm_user.c
+++ b/net/xfrm/xfrm_user.c
@@ -178,6 +178,12 @@ static inline int verify_replay(struct xfrm_usersa_info *p,
 				       "Replay seq and seq_hi should be 0 for output SA");
 			return -EINVAL;
 		}
+		if (rs->oseq_hi && !(p->flags & XFRM_STATE_ESN)) {
+			NL_SET_ERR_MSG(
+				extack,
+				"Replay oseq_hi should be 0 in non-ESN mode for output SA");
+			return -EINVAL;
+		}
 		if (rs->bmp_len) {
 			NL_SET_ERR_MSG(extack, "Replay bmp_len should 0 for output SA");
 			return -EINVAL;
@@ -190,6 +196,12 @@ static inline int verify_replay(struct xfrm_usersa_info *p,
 				       "Replay oseq and oseq_hi should be 0 for input SA");
 			return -EINVAL;
 		}
+		if (rs->seq_hi && !(p->flags & XFRM_STATE_ESN)) {
+			NL_SET_ERR_MSG(
+				extack,
+				"Replay seq_hi should be 0 in non-ESN mode for input SA");
+			return -EINVAL;
+		}
 	}
 
 	return 0;
@@ -907,7 +919,7 @@ static struct xfrm_state *xfrm_state_construct(struct net *net,
 			goto error;
 	}
 
-	err = __xfrm_init_state(x, false, attrs[XFRMA_OFFLOAD_DEV], extack);
+	err = __xfrm_init_state(x, extack);
 	if (err)
 		goto error;