[PATCH net-next 00/10] geneve: introduce double tunnel GSO/GRO

[PATCH net-next 00/10] geneve: introduce double tunnel GSO/GRO

[Paolo Abeni]

This is the [belated] incarnation of topic discussed in the last Neconf
[1].

In container orchestration in virtual environments there is a consistent
usage of double UDP tunneling - specifically geneve. Such setup lack
support of GRO and GSO for inter VM traffic.

After commit b430f6c38da6 ("Merge branch 'virtio_udp_tunnel_08_07_2025'
of https://github.com/pabeni/linux-devel") and the qemu cunter-part, VMs
are able to send/receive GSO over UDP aggregated packets.

This series introduces the missing bit for full end-to-end aggregation
in the above mentioned scenario. Specifically:

- introduces a new netdev feature set to generalize existing per device
driver GSO admission check.1
- adds GSO partial support for the geneve and vxlan drivers
- introduces and use a geneve option to assist double tunnel GRO
- adds some simple functional tests for the above.

The new device features set is not strictly needed for the following
work, but avoids the introduction of trivial `ndo_features_check` to
support GSO partial and thus possible performance regression due to the
additional indirect call. Such feature set could be leveraged by a
number of existing drivers (intel, meta and possibly wangxun) to avoid
duplicate code/tests. Such part has been omitted here to keep the series
small.

Both GSO partial support and double GRO support have some downsides.
With the first in place, GSO partial packets will traverse the network
stack 'downstream' the outer geneve UDP tunnel and will be visible by
the udp/IP/IPv6 and by netfilter. Currently only H/W NICs implement GSO
partial support and such packets are visible only via software taps.

Double UDP tunnel GRO will cook 'GSO partial' like aggregate packets,
i.e. the inner UDP encapsulation headers set will still carry the
wire-level lengths and csum, so that segmentation considering such
headers parts of a giant, constant encapsulation header will yield the
correct result.

The correct GSO packet layout is applied when the packet traverse the
outermost geneve encapsulation.

Both GSO partial and double UDP encap are disabled by default and must
be explicitly enabled via, respectively ethtool and geneve device
configuration.

Finally note that the GSO partial feature could potentially be applied
to all the other UDP tunnels, but this series limits its usage to geneve
and vxlan devices.

Link: https://netdev.bots.linux.dev/netconf/2024/paolo.pdf [1]

Paolo Abeni (10):
  net: introduce mangleid_features
  geneve: expose gso partial features for tunnel offload
  vxlan: expose gso partial features for tunnel  offload
  geneve: add netlink support for GRO hint
  geneve: constify geneve_hlen()
  geneve: pass the geneve device ptr to geneve_build_skb()
  geneve: add GRO hint output path
  geneve: extract hint option at GRO stage
  geneve: use GRO hint option in the RX path
  selftests: net: tests for add double tunneling GRO/GSO

 Documentation/netlink/specs/rt-link.yaml      |   3 +
 drivers/net/geneve.c                          | 520 +++++++++++++++++-
 drivers/net/vxlan/vxlan_core.c                |  16 +-
 include/linux/netdevice.h                     |   5 +-
 include/net/udp_tunnel.h                      |  32 ++
 include/uapi/linux/if_link.h                  |   1 +
 net/core/dev.c                                |   4 +-
 tools/testing/selftests/net/Makefile          |   1 +
 .../testing/selftests/net/double_udp_encap.sh | 340 ++++++++++++
 9 files changed, 888 insertions(+), 34 deletions(-)
 create mode 100755 tools/testing/selftests/net/double_udp_encap.sh

-- 
2.52.0

[PATCH net-next 01/10] net: introduce mangleid_features

[Paolo Abeni]

Some/most devices implementing gso_partial need to disable the GSO partial
features when the IP ID can't be mangled; to that extend each of them
implements something alike the following:

	if (skb->encapsulation && !(features & NETIF_F_TSO_MANGLEID))
		features &= ~NETIF_F_TSO;

in the ndo_features_check() op, which leads to a bit of duplicate code.

Later patch in the series will implement GSO partial support for virtual
device, and the current status quo will require more duplicate code and
a new indirect call in the TX path for such devices.

Introduce the mangleid_features mask, allowing the core to disable NIC
features based on/requiring MANGLEID, without any further intervention
from the driver.

The same functionality could be alternatively implemented adding a single
boolean flag to the struct net_device, but would require an additional
checks in ndo_features_check().

Also note that the above mentioned action is incorrect if the NIC
additionally implements NETIF_F_GSO_UDP_L4, mangleid_features
transparently handle even such a case.

Signed-off-by: Paolo Abeni <pabeni@redhat.com>
---
 include/linux/netdevice.h | 5 ++++-
 net/core/dev.c            | 4 +++-
 2 files changed, 7 insertions(+), 2 deletions(-)

diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index e808071dbb7d..92ff602d8f30 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -1817,7 +1817,9 @@ enum netdev_reg_state {
  *				and drivers will need to set them appropriately.
  *
  *	@mpls_features:	Mask of features inheritable by MPLS
- *	@gso_partial_features: value(s) from NETIF_F_GSO\*
+ *	@gso_partial_features: value(s) from NETIF_F_GSO
+ *	@mangleid_features:	Mask of features requiring MANGLEID, will be
+ *				disabled together with the latter.
  *
  *	@ifindex:	interface index
  *	@group:		The group the device belongs to
@@ -2206,6 +2208,7 @@ struct net_device {
 	netdev_features_t	vlan_features;
 	netdev_features_t	hw_enc_features;
 	netdev_features_t	mpls_features;
+	netdev_features_t	mangleid_features;
 
 	unsigned int		min_mtu;
 	unsigned int		max_mtu;
diff --git a/net/core/dev.c b/net/core/dev.c
index 69515edd17bc..9d3bbc0529d5 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -3789,8 +3789,10 @@ static netdev_features_t gso_features_check(const struct sk_buff *skb,
 		struct iphdr *iph = skb->encapsulation ?
 				    inner_ip_hdr(skb) : ip_hdr(skb);
 
-		if (!(iph->frag_off & htons(IP_DF)))
+		if (!(iph->frag_off & htons(IP_DF))) {
 			features &= ~NETIF_F_TSO_MANGLEID;
+			features &= ~dev->mangleid_features;
+		}
 	}
 
 	/* NETIF_F_IPV6_CSUM does not support IPv6 extension headers,
-- 
2.52.0

[PATCH net-next 02/10] geneve: expose gso partial features for tunnel offload

[Paolo Abeni]

GSO partial features for tunnels do not require any kind of support from
the underlying device: we can safely add them to the geneve UDP tunnel.

The only point of attention is the skb required features propagation in
the device xmit op: partial features must be stripped, except for
UDP_TUNNEL*.

Keep partial features disabled by default.

Signed-off-by: Paolo Abeni <pabeni@redhat.com>
---
 drivers/net/geneve.c     | 11 ++++++++++-
 include/net/udp_tunnel.h | 32 ++++++++++++++++++++++++++++++++
 2 files changed, 42 insertions(+), 1 deletion(-)

diff --git a/drivers/net/geneve.c b/drivers/net/geneve.c
index 77b0c3d52041..64ea4b970376 100644
--- a/drivers/net/geneve.c
+++ b/drivers/net/geneve.c
@@ -774,6 +774,7 @@ static int geneve_build_skb(struct dst_entry *dst, struct sk_buff *skb,
 	bool udp_sum = test_bit(IP_TUNNEL_CSUM_BIT, info->key.tun_flags);
 	struct genevehdr *gnvh;
 	__be16 inner_proto;
+	bool double_encap;
 	int min_headroom;
 	int err;
 
@@ -786,6 +787,7 @@ static int geneve_build_skb(struct dst_entry *dst, struct sk_buff *skb,
 	if (unlikely(err))
 		goto free_dst;
 
+	double_encap = udp_tunnel_handle_partial(skb);
 	err = udp_tunnel_handle_offloads(skb, udp_sum);
 	if (err)
 		goto free_dst;
@@ -793,7 +795,7 @@ static int geneve_build_skb(struct dst_entry *dst, struct sk_buff *skb,
 	gnvh = __skb_push(skb, sizeof(*gnvh) + info->options_len);
 	inner_proto = inner_proto_inherit ? skb->protocol : htons(ETH_P_TEB);
 	geneve_build_header(gnvh, info, inner_proto);
-	skb_set_inner_protocol(skb, inner_proto);
+	udp_tunnel_set_inner_protocol(skb, double_encap, inner_proto);
 	return 0;
 
 free_dst:
@@ -1211,9 +1213,16 @@ static void geneve_setup(struct net_device *dev)
 	dev->features    |= NETIF_F_RXCSUM;
 	dev->features    |= NETIF_F_GSO_SOFTWARE;
 
+	/* Partial features are disabled by default. */
 	dev->hw_features |= NETIF_F_SG | NETIF_F_HW_CSUM | NETIF_F_FRAGLIST;
 	dev->hw_features |= NETIF_F_RXCSUM;
 	dev->hw_features |= NETIF_F_GSO_SOFTWARE;
+	dev->hw_features |= UDP_TUNNEL_PARTIAL_FEATURES;
+	dev->hw_features |= NETIF_F_GSO_PARTIAL;
+
+	dev->hw_enc_features = dev->hw_features;
+	dev->gso_partial_features = UDP_TUNNEL_PARTIAL_FEATURES;
+	dev->mangleid_features = NETIF_F_GSO_PARTIAL;
 
 	dev->pcpu_stat_type = NETDEV_PCPU_STAT_DSTATS;
 	/* MTU range: 68 - (something less than 65535) */
diff --git a/include/net/udp_tunnel.h b/include/net/udp_tunnel.h
index 9acef2fbd2fd..d9c6d04bb3b5 100644
--- a/include/net/udp_tunnel.h
+++ b/include/net/udp_tunnel.h
@@ -10,6 +10,11 @@
 #include <net/ipv6_stubs.h>
 #endif
 
+#define UDP_TUNNEL_PARTIAL_FEATURES	NETIF_F_GSO_ENCAP_ALL
+#define UDP_TUNNEL_STRIPPED_GSO_TYPES	((UDP_TUNNEL_PARTIAL_FEATURES |	\
+					  NETIF_F_GSO_PARTIAL) >>	\
+					 NETIF_F_GSO_SHIFT)
+
 struct udp_port_cfg {
 	u8			family;
 
@@ -145,6 +150,33 @@ void udp_tunnel6_xmit_skb(struct dst_entry *dst, struct sock *sk,
 			  __be16 src_port, __be16 dst_port, bool nocheck,
 			  u16 ip6cb_flags);
 
+static inline bool udp_tunnel_handle_partial(struct sk_buff *skb)
+{
+	bool double_encap = !!(skb_shinfo(skb)->gso_type & SKB_GSO_PARTIAL);
+
+	/*
+	 * If the skb went through partial segmentation, lower devices
+	 * will not need to offload the related features - except for
+	 * UDP_TUNNEL, that will be re-added by the later
+	 * udp_tunnel_handle_offloads().
+	 */
+	if (double_encap)
+		skb_shinfo(skb)->gso_type &= ~UDP_TUNNEL_STRIPPED_GSO_TYPES;
+	return double_encap;
+}
+
+static inline void udp_tunnel_set_inner_protocol(struct sk_buff *skb,
+						 bool double_encap,
+						 __be16 inner_proto)
+{
+	/*
+	 * The inner protocol has been set by the nested tunnel, don't
+	 * overraid it.
+	 */
+	if (!double_encap)
+		skb_set_inner_protocol(skb, inner_proto);
+}
+
 void udp_tunnel_sock_release(struct socket *sock);
 
 struct rtable *udp_tunnel_dst_lookup(struct sk_buff *skb,
-- 
2.52.0

[PATCH net-next 03/10] vxlan: expose gso partial features for tunnel offload

[Paolo Abeni]

Similar to the previous patch, reuse the same helpers to add tunnel GSO
partial capabilities to vxlan devices.

Signed-off-by: Paolo Abeni <pabeni@redhat.com>
---
 drivers/net/vxlan/vxlan_core.c | 16 +++++++++++++---
 1 file changed, 13 insertions(+), 3 deletions(-)

diff --git a/drivers/net/vxlan/vxlan_core.c b/drivers/net/vxlan/vxlan_core.c
index a5c55e7e4d79..d3fed41e390e 100644
--- a/drivers/net/vxlan/vxlan_core.c
+++ b/drivers/net/vxlan/vxlan_core.c
@@ -2183,11 +2183,12 @@ static int vxlan_build_skb(struct sk_buff *skb, struct dst_entry *dst,
 			   struct vxlan_metadata *md, u32 vxflags,
 			   bool udp_sum)
 {
+	int type = udp_sum ? SKB_GSO_UDP_TUNNEL_CSUM : SKB_GSO_UDP_TUNNEL;
+	__be16 inner_protocol = htons(ETH_P_TEB);
 	struct vxlanhdr *vxh;
+	bool double_encap;
 	int min_headroom;
 	int err;
-	int type = udp_sum ? SKB_GSO_UDP_TUNNEL_CSUM : SKB_GSO_UDP_TUNNEL;
-	__be16 inner_protocol = htons(ETH_P_TEB);
 
 	if ((vxflags & VXLAN_F_REMCSUM_TX) &&
 	    skb->ip_summed == CHECKSUM_PARTIAL) {
@@ -2208,6 +2209,7 @@ static int vxlan_build_skb(struct sk_buff *skb, struct dst_entry *dst,
 	if (unlikely(err))
 		return err;
 
+	double_encap = udp_tunnel_handle_partial(skb);
 	err = iptunnel_handle_offloads(skb, type);
 	if (err)
 		return err;
@@ -2238,7 +2240,7 @@ static int vxlan_build_skb(struct sk_buff *skb, struct dst_entry *dst,
 		inner_protocol = skb->protocol;
 	}
 
-	skb_set_inner_protocol(skb, inner_protocol);
+	udp_tunnel_set_inner_protocol(skb, double_encap, inner_protocol);
 	return 0;
 }
 
@@ -3336,10 +3338,18 @@ static void vxlan_setup(struct net_device *dev)
 	dev->features   |= NETIF_F_RXCSUM;
 	dev->features   |= NETIF_F_GSO_SOFTWARE;
 
+	/* Partial features are disabled by default. */
 	dev->vlan_features = dev->features;
 	dev->hw_features |= NETIF_F_SG | NETIF_F_HW_CSUM | NETIF_F_FRAGLIST;
 	dev->hw_features |= NETIF_F_RXCSUM;
 	dev->hw_features |= NETIF_F_GSO_SOFTWARE;
+	dev->hw_features |= UDP_TUNNEL_PARTIAL_FEATURES;
+	dev->hw_features |= NETIF_F_GSO_PARTIAL;
+
+	dev->hw_enc_features = dev->hw_features;
+	dev->gso_partial_features = UDP_TUNNEL_PARTIAL_FEATURES;
+	dev->mangleid_features = NETIF_F_GSO_PARTIAL;
+
 	netif_keep_dst(dev);
 	dev->priv_flags |= IFF_NO_QUEUE;
 	dev->change_proto_down = true;
-- 
2.52.0

[PATCH net-next 04/10] geneve: add netlink support for GRO hint

[Paolo Abeni]

Allow configuring and dumping the new device option, and cache its value
into the geneve socket itself.
The new option is not tie to it any code yet.

Signed-off-by: Paolo Abeni <pabeni@redhat.com>
---
 Documentation/netlink/specs/rt-link.yaml |  3 +++
 drivers/net/geneve.c                     | 29 ++++++++++++++++++++----
 include/uapi/linux/if_link.h             |  1 +
 3 files changed, 29 insertions(+), 4 deletions(-)

diff --git a/Documentation/netlink/specs/rt-link.yaml b/Documentation/netlink/specs/rt-link.yaml
index ca22c68ca691..2dfb4becd6d0 100644
--- a/Documentation/netlink/specs/rt-link.yaml
+++ b/Documentation/netlink/specs/rt-link.yaml
@@ -1913,6 +1913,9 @@ attribute-sets:
         name: port-range
         type: binary
         struct: ifla-geneve-port-range
+      -
+        name: gro-hint
+        type: flag
   -
     name: linkinfo-hsr-attrs
     name-prefix: ifla-hsr-
diff --git a/drivers/net/geneve.c b/drivers/net/geneve.c
index 64ea4b970376..9837df99de65 100644
--- a/drivers/net/geneve.c
+++ b/drivers/net/geneve.c
@@ -56,6 +56,7 @@ struct geneve_config {
 	bool			collect_md;
 	bool			use_udp6_rx_checksums;
 	bool			ttl_inherit;
+	bool			gro_hint;
 	enum ifla_geneve_df	df;
 	bool			inner_proto_inherit;
 	u16			port_min;
@@ -84,6 +85,7 @@ struct geneve_dev {
 
 struct geneve_sock {
 	bool			collect_md;
+	bool			gro_hint;
 	struct list_head	list;
 	struct socket		*sock;
 	struct rcu_head		rcu;
@@ -659,13 +661,15 @@ static void geneve_sock_release(struct geneve_dev *geneve)
 
 static struct geneve_sock *geneve_find_sock(struct geneve_net *gn,
 					    sa_family_t family,
-					    __be16 dst_port)
+					    __be16 dst_port,
+					    bool gro_hint)
 {
 	struct geneve_sock *gs;
 
 	list_for_each_entry(gs, &gn->sock_list, list) {
 		if (inet_sk(gs->sock->sk)->inet_sport == dst_port &&
-		    geneve_get_sk_family(gs) == family) {
+		    geneve_get_sk_family(gs) == family &&
+		    gs->gro_hint == gro_hint) {
 			return gs;
 		}
 	}
@@ -676,12 +680,14 @@ static int geneve_sock_add(struct geneve_dev *geneve, bool ipv6)
 {
 	struct net *net = geneve->net;
 	struct geneve_net *gn = net_generic(net, geneve_net_id);
+	bool gro_hint = geneve->cfg.gro_hint;
 	struct geneve_dev_node *node;
 	struct geneve_sock *gs;
 	__u8 vni[3];
 	__u32 hash;
 
-	gs = geneve_find_sock(gn, ipv6 ? AF_INET6 : AF_INET, geneve->cfg.info.key.tp_dst);
+	gs = geneve_find_sock(gn, ipv6 ? AF_INET6 : AF_INET,
+			      geneve->cfg.info.key.tp_dst, gro_hint);
 	if (gs) {
 		gs->refcnt++;
 		goto out;
@@ -694,6 +700,7 @@ static int geneve_sock_add(struct geneve_dev *geneve, bool ipv6)
 
 out:
 	gs->collect_md = geneve->cfg.collect_md;
+	gs->gro_hint = gro_hint;
 #if IS_ENABLED(CONFIG_IPV6)
 	if (ipv6) {
 		rcu_assign_pointer(geneve->sock6, gs);
@@ -1257,6 +1264,7 @@ static const struct nla_policy geneve_policy[IFLA_GENEVE_MAX + 1] = {
 	[IFLA_GENEVE_DF]		= { .type = NLA_U8 },
 	[IFLA_GENEVE_INNER_PROTO_INHERIT]	= { .type = NLA_FLAG },
 	[IFLA_GENEVE_PORT_RANGE]	= NLA_POLICY_EXACT_LEN(sizeof(struct ifla_geneve_port_range)),
+	[IFLA_GENEVE_GRO_HINT]		= { .type = NLA_FLAG },
 };
 
 static int geneve_validate(struct nlattr *tb[], struct nlattr *data[],
@@ -1607,10 +1615,18 @@ static int geneve_nl2info(struct nlattr *tb[], struct nlattr *data[],
 		cfg->inner_proto_inherit = true;
 	}
 
+	if (data[IFLA_GENEVE_GRO_HINT]) {
+		if (changelink) {
+			attrtype = IFLA_GENEVE_GRO_HINT;
+			goto change_notsup;
+		}
+		cfg->gro_hint = true;
+	}
+
 	return 0;
 change_notsup:
 	NL_SET_ERR_MSG_ATTR(extack, data[attrtype],
-			    "Changing VNI, Port, endpoint IP address family, external, inner_proto_inherit, and UDP checksum attributes are not supported");
+			    "Changing VNI, Port, endpoint IP address family, external, inner_proto_inherit, gro_hint and UDP checksum attributes are not supported");
 	return -EOPNOTSUPP;
 }
 
@@ -1793,6 +1809,7 @@ static size_t geneve_get_size(const struct net_device *dev)
 		nla_total_size(sizeof(__u8)) + /* IFLA_GENEVE_TTL_INHERIT */
 		nla_total_size(0) +	 /* IFLA_GENEVE_INNER_PROTO_INHERIT */
 		nla_total_size(sizeof(struct ifla_geneve_port_range)) + /* IFLA_GENEVE_PORT_RANGE */
+		nla_total_size(0) +	 /* IFLA_GENEVE_GRO_HINT */
 		0;
 }
 
@@ -1865,6 +1882,10 @@ static int geneve_fill_info(struct sk_buff *skb, const struct net_device *dev)
 	if (nla_put(skb, IFLA_GENEVE_PORT_RANGE, sizeof(ports), &ports))
 		goto nla_put_failure;
 
+	if (geneve->cfg.gro_hint &&
+	    nla_put_u8(skb, IFLA_GENEVE_GRO_HINT, geneve->cfg.gro_hint))
+		goto nla_put_failure;
+
 	return 0;
 
 nla_put_failure:
diff --git a/include/uapi/linux/if_link.h b/include/uapi/linux/if_link.h
index 3b491d96e52e..e9b5f79e1ee1 100644
--- a/include/uapi/linux/if_link.h
+++ b/include/uapi/linux/if_link.h
@@ -1443,6 +1443,7 @@ enum {
 	IFLA_GENEVE_DF,
 	IFLA_GENEVE_INNER_PROTO_INHERIT,
 	IFLA_GENEVE_PORT_RANGE,
+	IFLA_GENEVE_GRO_HINT,
 	__IFLA_GENEVE_MAX
 };
 #define IFLA_GENEVE_MAX	(__IFLA_GENEVE_MAX - 1)
-- 
2.52.0

[PATCH net-next 05/10] geneve: constify geneve_hlen()

[Paolo Abeni]

Such helper does not modify the argument; constifying it will additionally
simplify later patches.

Additionally move the definition earlier, still for later's patchesi sake.

Signed-off-by: Paolo Abeni <pabeni@redhat.com>
---
 drivers/net/geneve.c | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

diff --git a/drivers/net/geneve.c b/drivers/net/geneve.c
index 9837df99de65..fc8c41221eb6 100644
--- a/drivers/net/geneve.c
+++ b/drivers/net/geneve.c
@@ -365,6 +365,11 @@ static void geneve_uninit(struct net_device *dev)
 	gro_cells_destroy(&geneve->gro_cells);
 }
 
+static int geneve_hlen(const struct genevehdr *gh)
+{
+	return sizeof(*gh) + gh->opt_len * 4;
+}
+
 /* Callback from net/ipv4/udp.c to receive packets */
 static int geneve_udp_encap_recv(struct sock *sk, struct sk_buff *skb)
 {
@@ -497,11 +502,6 @@ static struct socket *geneve_create_sock(struct net *net, bool ipv6,
 	return sock;
 }
 
-static int geneve_hlen(struct genevehdr *gh)
-{
-	return sizeof(*gh) + gh->opt_len * 4;
-}
-
 static struct sk_buff *geneve_gro_receive(struct sock *sk,
 					  struct list_head *head,
 					  struct sk_buff *skb)
-- 
2.52.0

[PATCH net-next 06/10] geneve: pass the geneve device ptr to geneve_build_skb()

[Paolo Abeni]

Instead of handing to it the geneve configuration in multiple arguments.
This already avoids some code duplication and we are going to pass soon
more arguments to such function.

Signed-off-by: Paolo Abeni <pabeni@redhat.com>
---
 drivers/net/geneve.c | 20 ++++++++------------
 1 file changed, 8 insertions(+), 12 deletions(-)

diff --git a/drivers/net/geneve.c b/drivers/net/geneve.c
index fc8c41221eb6..3287a24ccfc1 100644
--- a/drivers/net/geneve.c
+++ b/drivers/net/geneve.c
@@ -775,10 +775,11 @@ static void geneve_build_header(struct genevehdr *geneveh,
 
 static int geneve_build_skb(struct dst_entry *dst, struct sk_buff *skb,
 			    const struct ip_tunnel_info *info,
-			    bool xnet, int ip_hdr_len,
-			    bool inner_proto_inherit)
+			    const struct geneve_dev *geneve, int ip_hdr_len)
 {
 	bool udp_sum = test_bit(IP_TUNNEL_CSUM_BIT, info->key.tun_flags);
+	bool inner_proto_inherit = geneve->cfg.inner_proto_inherit;
+	bool xnet = !net_eq(geneve->net, dev_net(geneve->dev));
 	struct genevehdr *gnvh;
 	__be16 inner_proto;
 	bool double_encap;
@@ -830,8 +831,6 @@ static int geneve_xmit_skb(struct sk_buff *skb, struct net_device *dev,
 			   struct geneve_dev *geneve,
 			   const struct ip_tunnel_info *info)
 {
-	bool inner_proto_inherit = geneve->cfg.inner_proto_inherit;
-	bool xnet = !net_eq(geneve->net, dev_net(geneve->dev));
 	struct geneve_sock *gs4 = rcu_dereference(geneve->sock4);
 	const struct ip_tunnel_key *key = &info->key;
 	struct rtable *rt;
@@ -842,7 +841,7 @@ static int geneve_xmit_skb(struct sk_buff *skb, struct net_device *dev,
 	__be16 sport;
 	int err;
 
-	if (skb_vlan_inet_prepare(skb, inner_proto_inherit))
+	if (skb_vlan_inet_prepare(skb, geneve->cfg.inner_proto_inherit))
 		return -EINVAL;
 
 	if (!gs4)
@@ -925,8 +924,8 @@ static int geneve_xmit_skb(struct sk_buff *skb, struct net_device *dev,
 		}
 	}
 
-	err = geneve_build_skb(&rt->dst, skb, info, xnet, sizeof(struct iphdr),
-			       inner_proto_inherit);
+	err = geneve_build_skb(&rt->dst, skb, info, geneve,
+			       sizeof(struct iphdr));
 	if (unlikely(err))
 		return err;
 
@@ -943,8 +942,6 @@ static int geneve6_xmit_skb(struct sk_buff *skb, struct net_device *dev,
 			    struct geneve_dev *geneve,
 			    const struct ip_tunnel_info *info)
 {
-	bool inner_proto_inherit = geneve->cfg.inner_proto_inherit;
-	bool xnet = !net_eq(geneve->net, dev_net(geneve->dev));
 	struct geneve_sock *gs6 = rcu_dereference(geneve->sock6);
 	const struct ip_tunnel_key *key = &info->key;
 	struct dst_entry *dst = NULL;
@@ -954,7 +951,7 @@ static int geneve6_xmit_skb(struct sk_buff *skb, struct net_device *dev,
 	__be16 sport;
 	int err;
 
-	if (skb_vlan_inet_prepare(skb, inner_proto_inherit))
+	if (skb_vlan_inet_prepare(skb, geneve->cfg.inner_proto_inherit))
 		return -EINVAL;
 
 	if (!gs6)
@@ -1017,8 +1014,7 @@ static int geneve6_xmit_skb(struct sk_buff *skb, struct net_device *dev,
 			ttl = key->ttl;
 		ttl = ttl ? : ip6_dst_hoplimit(dst);
 	}
-	err = geneve_build_skb(dst, skb, info, xnet, sizeof(struct ipv6hdr),
-			       inner_proto_inherit);
+	err = geneve_build_skb(dst, skb, info, geneve, sizeof(struct ipv6hdr));
 	if (unlikely(err))
 		return err;
 
-- 
2.52.0

[PATCH net-next 07/10] geneve: add GRO hint output path

[Paolo Abeni]

If a geneve egress packet contains nested UDP encap headers, add a geneve
option including the information necessary on the RX side to perform GRO
aggregation of the whole packets: the nested network and transport headers,
and the nested protocol type.

Use geneve option class `netdev`, already registered in the Network
Virtualization Overlay (NVO3) IANA registry:

https://www.iana.org/assignments/nvo3/nvo3.xhtml#Linux-NetDev.

To pass the GRO hint information across the different xmit path functions,
store them in the skb control buffer, to avoid adding additional arguments.

Signed-off-by: Paolo Abeni <pabeni@redhat.com>
---
 drivers/net/geneve.c | 124 +++++++++++++++++++++++++++++++++++++++++--
 1 file changed, 120 insertions(+), 4 deletions(-)

diff --git a/drivers/net/geneve.c b/drivers/net/geneve.c
index 3287a24ccfc1..6571035c8129 100644
--- a/drivers/net/geneve.c
+++ b/drivers/net/geneve.c
@@ -38,6 +38,26 @@ MODULE_PARM_DESC(log_ecn_error, "Log packets received with corrupted ECN");
 #define GENEVE_IPV4_HLEN (ETH_HLEN + sizeof(struct iphdr) + GENEVE_BASE_HLEN)
 #define GENEVE_IPV6_HLEN (ETH_HLEN + sizeof(struct ipv6hdr) + GENEVE_BASE_HLEN)
 
+#define GENEVE_OPT_NETDEV_CLASS		0x100
+#define GENEVE_OPT_GRO_HINT_SIZE	8
+#define GENEVE_OPT_GRO_HINT_TYPE	1
+#define GENEVE_OPT_GRO_HINT_LEN		1
+
+struct geneve_opt_gro_hint {
+	u8	inner_proto_id:2,
+		nested_is_v6:1;
+	u8	nested_nh_offset;
+	u8	nested_tp_offset;
+	u8	nested_hdr_len;
+};
+
+struct geneve_skb_cb {
+	unsigned int	gro_hint_len;
+	struct geneve_opt_gro_hint gro_hint;
+};
+
+#define GENEVE_SKB_CB(__skb)	((struct geneve_skb_cb *)&((__skb)->cb[0]))
+
 /* per-network namespace private data for this module */
 struct geneve_net {
 	struct list_head	geneve_list;
@@ -93,6 +113,21 @@ struct geneve_sock {
 	struct hlist_head	vni_list[VNI_HASH_SIZE];
 };
 
+static const __be16 proto_id_map[] = { htons(ETH_P_TEB),
+				       htons(ETH_P_IPV6),
+				       htons(ETH_P_IP) };
+
+static int proto_to_id(__be16 proto)
+{
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(proto_id_map); i++)
+		if (proto_id_map[i] == proto)
+			return i;
+
+	return ARRAY_SIZE(proto_id_map);
+}
+
 static inline __u32 geneve_net_vni_hash(u8 vni[3])
 {
 	__u32 vnid;
@@ -773,6 +808,78 @@ static void geneve_build_header(struct genevehdr *geneveh,
 		ip_tunnel_info_opts_get(geneveh->options, info);
 }
 
+static int geneve_build_gro_hint_opt(const struct geneve_dev *geneve,
+				     struct sk_buff *skb)
+{
+	struct geneve_skb_cb *cb = GENEVE_SKB_CB(skb);
+	struct geneve_opt_gro_hint *hint;
+	unsigned int nhlen;
+	bool nested_is_v6;
+	int id;
+
+	BUILD_BUG_ON(sizeof(skb->cb) < sizeof(struct geneve_skb_cb));
+	cb->gro_hint_len = 0;
+
+	/* Try to add the GRO hint only in case of double encap. */
+	if (!geneve->cfg.gro_hint || !skb->encapsulation)
+		return 0;
+
+	/*
+	 * The nested headers must fit the geneve opt len fields and the
+	 * nested encap must carry a nested transport (UDP) header.
+	 */
+	nhlen = skb_inner_mac_header(skb) - skb->data;
+	if (nhlen > 255 || !skb_transport_header_was_set(skb) ||
+	    skb->inner_protocol_type != ENCAP_TYPE_ETHER ||
+	    (skb_transport_offset(skb) + sizeof(struct udphdr) > nhlen))
+		return 0;
+
+	id = proto_to_id(skb->inner_protocol);
+	if (id < 0)
+		return 0;
+
+	nested_is_v6 = skb->protocol == htons(ETH_P_IPV6);
+	if (nested_is_v6) {
+		int start = skb_network_offset(skb) + sizeof(struct ipv6hdr);
+		u8 proto = ipv6_hdr(skb)->nexthdr;
+		__be16 foff;
+
+		if (ipv6_skip_exthdr(skb, start, &proto, &foff) < 0 ||
+		    proto != IPPROTO_UDP)
+			return 0;
+	} else {
+		if (ip_hdr(skb)->protocol != IPPROTO_UDP)
+			return 0;
+	}
+
+	hint = &cb->gro_hint;
+	memset(hint, 0, sizeof(*hint));
+	hint->inner_proto_id = id;
+	hint->nested_is_v6 = skb->protocol == htons(ETH_P_IPV6);
+	hint->nested_nh_offset = skb_network_offset(skb);
+	hint->nested_tp_offset = skb_transport_offset(skb);
+	hint->nested_hdr_len = nhlen;
+	cb->gro_hint_len = GENEVE_OPT_GRO_HINT_SIZE;
+	return GENEVE_OPT_GRO_HINT_SIZE;
+}
+
+static void geneve_put_gro_hint_opt(struct genevehdr *gnvh, int opt_size,
+				    const struct geneve_opt_gro_hint *hint)
+{
+	struct geneve_opt *gro_opt;
+
+	/* geneve_build_header() did not took in account the GRO hint. */
+	gnvh->opt_len = (opt_size + GENEVE_OPT_GRO_HINT_SIZE) >> 2;
+
+	gro_opt = (void *)(gnvh + 1) + opt_size;
+	memset(gro_opt, 0, sizeof(*gro_opt));
+
+	gro_opt->opt_class = htons(GENEVE_OPT_NETDEV_CLASS);
+	gro_opt->type = GENEVE_OPT_GRO_HINT_TYPE;
+	gro_opt->length = GENEVE_OPT_GRO_HINT_LEN;
+	memcpy(gro_opt + 1, hint, sizeof(*hint));
+}
+
 static int geneve_build_skb(struct dst_entry *dst, struct sk_buff *skb,
 			    const struct ip_tunnel_info *info,
 			    const struct geneve_dev *geneve, int ip_hdr_len)
@@ -780,17 +887,20 @@ static int geneve_build_skb(struct dst_entry *dst, struct sk_buff *skb,
 	bool udp_sum = test_bit(IP_TUNNEL_CSUM_BIT, info->key.tun_flags);
 	bool inner_proto_inherit = geneve->cfg.inner_proto_inherit;
 	bool xnet = !net_eq(geneve->net, dev_net(geneve->dev));
+	struct geneve_skb_cb *cb = GENEVE_SKB_CB(skb);
 	struct genevehdr *gnvh;
 	__be16 inner_proto;
 	bool double_encap;
 	int min_headroom;
+	int opt_size;
 	int err;
 
 	skb_reset_mac_header(skb);
 	skb_scrub_packet(skb, xnet);
 
+	opt_size =  info->options_len + cb->gro_hint_len;
 	min_headroom = LL_RESERVED_SPACE(dst->dev) + dst->header_len +
-		       GENEVE_BASE_HLEN + info->options_len + ip_hdr_len;
+		       GENEVE_BASE_HLEN + opt_size + ip_hdr_len;
 	err = skb_cow_head(skb, min_headroom);
 	if (unlikely(err))
 		goto free_dst;
@@ -800,9 +910,13 @@ static int geneve_build_skb(struct dst_entry *dst, struct sk_buff *skb,
 	if (err)
 		goto free_dst;
 
-	gnvh = __skb_push(skb, sizeof(*gnvh) + info->options_len);
+	gnvh = __skb_push(skb, sizeof(*gnvh) + opt_size);
 	inner_proto = inner_proto_inherit ? skb->protocol : htons(ETH_P_TEB);
 	geneve_build_header(gnvh, info, inner_proto);
+
+	if (cb->gro_hint_len)
+		geneve_put_gro_hint_opt(gnvh, info->options_len, &cb->gro_hint);
+
 	udp_tunnel_set_inner_protocol(skb, double_encap, inner_proto);
 	return 0;
 
@@ -862,7 +976,8 @@ static int geneve_xmit_skb(struct sk_buff *skb, struct net_device *dev,
 		return PTR_ERR(rt);
 
 	err = skb_tunnel_check_pmtu(skb, &rt->dst,
-				    GENEVE_IPV4_HLEN + info->options_len,
+				    GENEVE_IPV4_HLEN + info->options_len +
+				    geneve_build_gro_hint_opt(geneve, skb),
 				    netif_is_any_bridge_port(dev));
 	if (err < 0) {
 		dst_release(&rt->dst);
@@ -972,7 +1087,8 @@ static int geneve6_xmit_skb(struct sk_buff *skb, struct net_device *dev,
 		return PTR_ERR(dst);
 
 	err = skb_tunnel_check_pmtu(skb, dst,
-				    GENEVE_IPV6_HLEN + info->options_len,
+				    GENEVE_IPV6_HLEN + info->options_len +
+				    geneve_build_gro_hint_opt(geneve, skb),
 				    netif_is_any_bridge_port(dev));
 	if (err < 0) {
 		dst_release(dst);
-- 
2.52.0

[PATCH net-next 08/10] geneve: extract hint option at GRO stage

[Paolo Abeni]

Add helpers for finding a GRO hint option in the geneve header, performing
basic sanitization of the option offsets vs the actual packet layout,
validate the option for GRO aggregation and check the nested header
checksum.

The validation helper closely mirrors similar check performed by the ipv4
and ipv6 gro callbacks, with the additional twist of accessing the
relevant network header via the GRO hint offset.

To validate the nested UDP checksum, leverage the csum completed of the
outer header, similarly to LCO, with the main difference that in this case
we have the outer checksum available.

Use the helpers to extract the hint info at the GRO stage.

Signed-off-by: Paolo Abeni <pabeni@redhat.com>
---
 drivers/net/geneve.c | 172 ++++++++++++++++++++++++++++++++++++++++++-
 1 file changed, 171 insertions(+), 1 deletion(-)

diff --git a/drivers/net/geneve.c b/drivers/net/geneve.c
index 6571035c8129..5030ad0db3fc 100644
--- a/drivers/net/geneve.c
+++ b/drivers/net/geneve.c
@@ -405,6 +405,159 @@ static int geneve_hlen(const struct genevehdr *gh)
 	return sizeof(*gh) + gh->opt_len * 4;
 }
 
+/*
+ * Look for GRO hint in the genenve options; return the extracted
+ * data only if they pass basic sanitization.
+ */
+static struct geneve_opt_gro_hint *
+geneve_opt_gro_hint(const struct genevehdr *gh, __be16 *type,
+		    unsigned int *gh_len)
+{
+	struct geneve_opt *opt = (void *)(gh + 1);
+	struct geneve_opt_gro_hint *gro_hint;
+	int id, opt_len = gh->opt_len;
+
+	while (opt_len >= (GENEVE_OPT_GRO_HINT_SIZE >> 2)) {
+		if (opt->opt_class == htons(GENEVE_OPT_NETDEV_CLASS) &&
+		    opt->type == GENEVE_OPT_GRO_HINT_TYPE &&
+		    opt->length == GENEVE_OPT_GRO_HINT_LEN)
+			goto found;
+
+		/* check for bad opt len */
+		if (opt->length + 1 >= opt_len)
+			return NULL;
+
+		/* next opt */
+		opt_len -= opt->length + 1;
+		opt = ((void *)opt) + ((opt->length + 1) << 2);
+	}
+	return NULL;
+
+found:
+	gro_hint = (struct geneve_opt_gro_hint *)opt->opt_data;
+
+	/*
+	 * Sanitize the hinted hdrs: the nested transport is UDP and must fit
+	 * the overall hinted hdr size.
+	 */
+	if (gro_hint->nested_tp_offset + sizeof(struct udphdr) >
+	    gro_hint->nested_hdr_len)
+		return NULL;
+
+	if (gro_hint->nested_nh_offset +
+	    (gro_hint->nested_is_v6 ? sizeof(struct ipv6hdr) :
+				      sizeof(struct iphdr)) >
+	    gro_hint->nested_tp_offset)
+		return NULL;
+
+	/* Allow only supported L2. */
+	id = gro_hint->inner_proto_id;
+	if (id >= ARRAY_SIZE(proto_id_map))
+		return NULL;
+
+	*type = proto_id_map[id];
+	*gh_len += gro_hint->nested_hdr_len;
+
+	return gro_hint;
+}
+
+static struct geneve_opt_gro_hint *
+geneve_sk_gro_hint(const struct sock *sk, const struct genevehdr *gh,
+		   __be16 *type, unsigned int *gh_len)
+{
+	const struct geneve_sock *gs = rcu_dereference_sk_user_data(sk);
+
+	if (!gs || !gs->gro_hint)
+		return NULL;
+	return geneve_opt_gro_hint(gh, type, gh_len);
+}
+
+/* Validate the packet headers pointed by data WRT the provided hint */
+static bool
+geneve_opt_gro_hint_validate(void *data,
+			     const struct geneve_opt_gro_hint *gro_hint)
+{
+	void *nested_nh = data + gro_hint->nested_nh_offset;
+	struct iphdr *iph;
+
+	if (gro_hint->nested_is_v6) {
+		struct ipv6hdr *ipv6h = nested_nh;
+		struct ipv6_opt_hdr *opth;
+		int offset, len;
+
+		if (ipv6h->nexthdr == IPPROTO_UDP)
+			return true;
+
+		offset = sizeof(*ipv6h) + gro_hint->nested_nh_offset;
+		while (offset + sizeof(*opth) <= gro_hint->nested_tp_offset) {
+			opth = data + offset;
+
+			len = ipv6_optlen(opth);
+			if (len + offset > gro_hint->nested_tp_offset)
+				return false;
+			if (opth->nexthdr == IPPROTO_UDP)
+				return true;
+
+			offset += len;
+		}
+		return false;
+	}
+
+	iph = nested_nh;
+	if (*(u8 *)iph != 0x45 || ip_is_fragment(iph) ||
+	    iph->protocol != IPPROTO_UDP || ip_fast_csum((u8 *)iph, 5))
+		return false;
+
+	return true;
+}
+
+/*
+ * Validate the skb headers following the specified geneve hdr vs the
+ * provided hint, including nested L4 checksum.
+ * The caller already ensured that the relevant amount of data is available
+ * in the linear part.
+ */
+static bool
+geneve_opt_gro_hint_validate_csum(const struct sk_buff *skb,
+				  const struct genevehdr *gh,
+				  const struct geneve_opt_gro_hint *gro_hint)
+{
+	unsigned int plen, gh_len = geneve_hlen(gh);
+	void *nested = (void *)gh + gh_len;
+	struct udphdr *nested_uh;
+	unsigned int nested_len;
+	struct ipv6hdr *ipv6h;
+	struct iphdr *iph;
+	__wsum csum, psum;
+
+	if (!geneve_opt_gro_hint_validate(nested, gro_hint))
+		return false;
+
+	/* Use GRO hints with nested csum only if the outer header has csum. */
+	nested_uh = nested + gro_hint->nested_tp_offset;
+	if (!nested_uh->check || skb->ip_summed == CHECKSUM_PARTIAL)
+		return true;
+
+	if (!NAPI_GRO_CB(skb)->csum_valid)
+		return false;
+
+	/* Compute the complete checksum up to the nested transport. */
+	plen = gh_len + gro_hint->nested_tp_offset;
+	csum = csum_sub(NAPI_GRO_CB(skb)->csum, csum_partial(gh, plen, 0));
+	nested_len = skb_gro_len(skb) - plen;
+
+	/* Compute the nested pseudo header csum. */
+	ipv6h = nested + gro_hint->nested_nh_offset;
+	iph = (struct iphdr *)ipv6h;
+	psum = gro_hint->nested_is_v6 ?
+	       ~csum_unfold(csum_ipv6_magic(&ipv6h->saddr, &ipv6h->daddr,
+					    nested_len, IPPROTO_UDP, 0)) :
+	       csum_tcpudp_nofold(iph->saddr, iph->daddr,
+				  nested_len, IPPROTO_UDP, 0);
+
+	return !csum_fold(csum_add(psum, csum));
+}
+
 /* Callback from net/ipv4/udp.c to receive packets */
 static int geneve_udp_encap_recv(struct sock *sk, struct sk_buff *skb)
 {
@@ -541,6 +694,7 @@ static struct sk_buff *geneve_gro_receive(struct sock *sk,
 					  struct list_head *head,
 					  struct sk_buff *skb)
 {
+	struct geneve_opt_gro_hint *gro_hint;
 	struct sk_buff *pp = NULL;
 	struct sk_buff *p;
 	struct genevehdr *gh, *gh2;
@@ -558,6 +712,7 @@ static struct sk_buff *geneve_gro_receive(struct sock *sk,
 	if (gh->ver != GENEVE_VER || gh->oam)
 		goto out;
 	gh_len = geneve_hlen(gh);
+	type = gh->proto_type;
 
 	hlen = off_gnv + gh_len;
 	if (!skb_gro_may_pull(skb, hlen)) {
@@ -566,6 +721,22 @@ static struct sk_buff *geneve_gro_receive(struct sock *sk,
 			goto out;
 	}
 
+	/* The GRO hint/nested hdr could use a different ethernet type. */
+	gro_hint = geneve_sk_gro_hint(sk, gh, &type, &gh_len);
+	if (gro_hint) {
+		/*
+		 * If the hint is present, and nested hdr validation fails, do
+		 * not attempt plain GRO: it will ignore inner hdrs and cause
+		 * OoO.
+		 */
+		gh = skb_gro_header(skb, off_gnv + gh_len, off_gnv);
+		if (unlikely(!gh))
+			goto out;
+
+		if (!geneve_opt_gro_hint_validate_csum(skb, gh, gro_hint))
+			goto out;
+	}
+
 	list_for_each_entry(p, head, list) {
 		if (!NAPI_GRO_CB(p)->same_flow)
 			continue;
@@ -580,7 +751,6 @@ static struct sk_buff *geneve_gro_receive(struct sock *sk,
 
 	skb_gro_pull(skb, gh_len);
 	skb_gro_postpull_rcsum(skb, gh, gh_len);
-	type = gh->proto_type;
 	if (likely(type == htons(ETH_P_TEB)))
 		return call_gro_receive(eth_gro_receive, head, skb);
 
-- 
2.52.0

[PATCH net-next 09/10] geneve: use GRO hint option in the RX path

[Paolo Abeni]

At the GRO stage, when a valid hint option is found, try match the whole
nested headers and try to aggregate on the inner protocol; in case of hdr
mismatch extract the nested address and port to properly flush on a
per-inner flow basis.

On GRO completion, the (unmodified) nested headers will be considered part
of the (constant) outer geneve encap header so that plain UDP tunnel
segmentation will yield valid wire packets.

In the geneve RX path, when processing a GSO packet carrying a GRO hint
option, update the nested header length fields from the wire packet size to
the GSO-packet one. If the nested header additionally carries a checksum,
convert it to CSUM-partial.

Signed-off-by: Paolo Abeni <pabeni@redhat.com>
---
 drivers/net/geneve.c | 160 +++++++++++++++++++++++++++++++++++++++++--
 1 file changed, 155 insertions(+), 5 deletions(-)

diff --git a/drivers/net/geneve.c b/drivers/net/geneve.c
index 5030ad0db3fc..b33c72538cfb 100644
--- a/drivers/net/geneve.c
+++ b/drivers/net/geneve.c
@@ -259,9 +259,8 @@ static struct geneve_dev *geneve_lookup_skb(struct geneve_sock *gs,
 
 /* geneve receive/decap routine */
 static void geneve_rx(struct geneve_dev *geneve, struct geneve_sock *gs,
-		      struct sk_buff *skb)
+		      struct sk_buff *skb, const struct genevehdr *gnvh)
 {
-	struct genevehdr *gnvh = geneve_hdr(skb);
 	struct metadata_dst *tun_dst = NULL;
 	unsigned int len;
 	int nh, err = 0;
@@ -558,6 +557,76 @@ geneve_opt_gro_hint_validate_csum(const struct sk_buff *skb,
 	return !csum_fold(csum_add(psum, csum));
 }
 
+static int geneve_post_decap_hint(const struct sock *sk, struct sk_buff *skb,
+				  unsigned int gh_len)
+{
+	struct geneve_opt_gro_hint *gro_hint;
+	unsigned int len, total_len;
+	struct ipv6hdr *ipv6h;
+	struct iphdr *iph;
+	struct udphdr *uh;
+	__be16 p;
+
+	gro_hint = geneve_sk_gro_hint(sk, (void *)skb->data - gh_len, &p, &len);
+	if (!gro_hint)
+		return 0;
+
+	if (!skb_is_gso(skb))
+		return 0;
+
+	if (unlikely(!pskb_may_pull(skb, gro_hint->nested_hdr_len)))
+		return -ENOMEM;
+
+	if (unlikely(skb_shinfo(skb)->gso_type & SKB_GSO_DODGY &&
+		     !geneve_opt_gro_hint_validate(skb->data, gro_hint)))
+		return -EINVAL;
+
+	ipv6h = (void *)skb->data + gro_hint->nested_nh_offset;
+	iph = (struct iphdr *)ipv6h;
+	total_len = skb->len - gro_hint->nested_nh_offset;
+	if (total_len > GRO_LEGACY_MAX_SIZE)
+		return -E2BIG;
+
+	/*
+	 * After stripping the outer encap, the packet still carries a
+	 * tunnel encapsulation: the nested one.
+	 */
+	skb->encapsulation = 1;
+
+	/* GSO expect a valid transpor header, move it to the current one. */
+	skb_set_transport_header(skb, gro_hint->nested_tp_offset);
+
+	/* Adjust the nested IP{6} hdr to actual GSO len. */
+	if (gro_hint->nested_is_v6) {
+		ipv6h->payload_len = htons(total_len - sizeof(*ipv6h));
+	} else {
+		__be16 old_len = iph->tot_len;
+
+		iph->tot_len = htons(total_len);
+
+		/* For IPv4 additionally adjust the nested csum. */
+		csum_replace2(&iph->check, old_len, iph->tot_len);
+		ip_send_check(iph);
+	}
+
+	/* Adjust the nested UDP header len and checksum. */
+	uh = udp_hdr(skb);
+	uh->len = htons(skb->len - gro_hint->nested_tp_offset);
+	if (uh->check) {
+		len = skb->len - gro_hint->nested_nh_offset;
+		skb_shinfo(skb)->gso_type |= SKB_GSO_UDP_TUNNEL_CSUM;
+		if (gro_hint->nested_is_v6)
+			uh->check = ~udp_v6_check(len, &ipv6h->saddr,
+						  &ipv6h->daddr, 0);
+		else
+			uh->check = ~udp_v4_check(len, iph->saddr,
+						  iph->daddr, 0);
+	} else {
+		skb_shinfo(skb)->gso_type |= SKB_GSO_UDP_TUNNEL;
+	}
+	return 0;
+}
+
 /* Callback from net/ipv4/udp.c to receive packets */
 static int geneve_udp_encap_recv(struct sock *sk, struct sk_buff *skb)
 {
@@ -599,7 +668,15 @@ static int geneve_udp_encap_recv(struct sock *sk, struct sk_buff *skb)
 		goto drop;
 	}
 
-	geneve_rx(geneve, gs, skb);
+	/* The hint will also set the transport header to the nested one. */
+	geneveh = geneve_hdr(skb);
+	if (geneve_post_decap_hint(sk, skb, sizeof(struct genevehdr) +
+				   opts_len)) {
+		DEV_STATS_INC(geneve->dev, rx_errors);
+		goto drop;
+	}
+
+	geneve_rx(geneve, gs, skb, geneveh);
 	return 0;
 
 drop:
@@ -690,6 +767,79 @@ static struct socket *geneve_create_sock(struct net *net, bool ipv6,
 	return sock;
 }
 
+static bool geneve_hdr_match(struct sk_buff *skb,
+			     const struct genevehdr *gh,
+			     const struct genevehdr *gh2,
+			     const struct geneve_opt_gro_hint *gro_hint)
+{
+	void *nested, *nested2, *nh, *nh2;
+	struct udphdr *udp, *udp2;
+	unsigned int gh_len;
+
+	/* Match the geneve hdr and options */
+	if (gh->opt_len != gh2->opt_len)
+		return false;
+
+	gh_len = geneve_hlen(gh);
+	if (memcmp(gh, gh2, gh_len))
+		return false;
+
+	if (!gro_hint)
+		return true;
+
+	/*
+	 * When gro is present consider the nested headers as part
+	 * of the geneve options
+	 */
+	nested = (void *)gh + gh_len;
+	nested2 = (void *)gh2 + gh_len;
+	if (!memcmp(nested, nested2, gro_hint->nested_hdr_len))
+		return true;
+
+	/*
+	 * The nested headers differ; the packets can still belong to
+	 * the same flow when IPs/proto/ports match; if so flushing is
+	 * required.
+	 */
+	nh = nested + gro_hint->nested_nh_offset;
+	nh2 = nested2 + gro_hint->nested_nh_offset;
+	if (gro_hint->nested_is_v6) {
+		struct ipv6hdr *iph = nh, *iph2 = nh2;
+		unsigned int nested_nlen;
+		__be32 first_word;
+
+		first_word = *(__be32 *)iph ^ *(__be32 *)iph2;
+		if ((first_word & htonl(0xF00FFFFF)) ||
+		    !ipv6_addr_equal(&iph->saddr, &iph2->saddr) ||
+		    !ipv6_addr_equal(&iph->daddr, &iph2->daddr) ||
+		    iph->nexthdr != iph2->nexthdr)
+			return false;
+
+		nested_nlen = gro_hint->nested_tp_offset -
+			      gro_hint->nested_nh_offset;
+		if (nested_nlen > sizeof(struct ipv6hdr) &&
+		    (memcmp(iph + 1, iph2 + 1,
+			    nested_nlen - sizeof(struct ipv6hdr))))
+			return false;
+	} else {
+		struct iphdr *iph = nh, *iph2 = nh2;
+
+		if ((iph->protocol ^ iph2->protocol) |
+		    ((__force u32)iph->saddr ^ (__force u32)iph2->saddr) |
+		    ((__force u32)iph->daddr ^ (__force u32)iph2->daddr))
+			return false;
+	}
+
+	udp = nested + gro_hint->nested_tp_offset;
+	udp2 = nested2 + gro_hint->nested_tp_offset;
+	if (udp->source != udp2->source || udp->dest != udp2->dest ||
+	    udp->check != udp2->check)
+		return false;
+
+	NAPI_GRO_CB(skb)->flush = 1;
+	return true;
+}
+
 static struct sk_buff *geneve_gro_receive(struct sock *sk,
 					  struct list_head *head,
 					  struct sk_buff *skb)
@@ -742,8 +892,7 @@ static struct sk_buff *geneve_gro_receive(struct sock *sk,
 			continue;
 
 		gh2 = (struct genevehdr *)(p->data + off_gnv);
-		if (gh->opt_len != gh2->opt_len ||
-		    memcmp(gh, gh2, gh_len)) {
+		if (!geneve_hdr_match(skb, gh, gh2, gro_hint)) {
 			NAPI_GRO_CB(p)->same_flow = 0;
 			continue;
 		}
@@ -779,6 +928,7 @@ static int geneve_gro_complete(struct sock *sk, struct sk_buff *skb,
 	gh = (struct genevehdr *)(skb->data + nhoff);
 	gh_len = geneve_hlen(gh);
 	type = gh->proto_type;
+	geneve_opt_gro_hint(gh, &type, &gh_len);
 
 	/* since skb->encapsulation is set, eth_gro_complete() sets the inner mac header */
 	if (likely(type == htons(ETH_P_TEB)))
-- 
2.52.0

[PATCH net-next 10/10] selftests: net: tests for add double tunneling GRO/GSO

[Paolo Abeni]

Create a simple, netns-based topology with double, nested UDP tunnels and
perform TSO transfers on top.

Explicitly enable GSO and/or GRO and check the skb layout consistency with
different configuration allowing (or not) GSO frames to be delivered on
the other end.

Signed-off-by: Paolo Abeni <pabeni@redhat.com>
---
 tools/testing/selftests/net/Makefile          |   1 +
 .../testing/selftests/net/double_udp_encap.sh | 340 ++++++++++++++++++
 2 files changed, 341 insertions(+)
 create mode 100755 tools/testing/selftests/net/double_udp_encap.sh

diff --git a/tools/testing/selftests/net/Makefile b/tools/testing/selftests/net/Makefile
index b66ba04f19d9..063155f42cd7 100644
--- a/tools/testing/selftests/net/Makefile
+++ b/tools/testing/selftests/net/Makefile
@@ -22,6 +22,7 @@ TEST_PROGS := \
 	cmsg_so_mark.sh \
 	cmsg_so_priority.sh \
 	cmsg_time.sh \
+	double_udp_encap.sh \
 	drop_monitor_tests.sh \
 	fcnal-ipv4.sh \
 	fcnal-ipv6.sh \
diff --git a/tools/testing/selftests/net/double_udp_encap.sh b/tools/testing/selftests/net/double_udp_encap.sh
new file mode 100755
index 000000000000..99c855b11df2
--- /dev/null
+++ b/tools/testing/selftests/net/double_udp_encap.sh
@@ -0,0 +1,340 @@
+#!/bin/bash
+# SPDX-License-Identifier: GPL-2.0
+
+source lib.sh
+
+readonly CLI=$(dirname $(readlink -f $0))/../../../net/ynl/pyynl/cli.py
+
+readonly SRC=1
+readonly DST=2
+
+readonly NET_V4=192.168.1.
+readonly NET_V6=2001:db8::
+readonly OL1_NET_V4=172.16.1.
+readonly OL1_NET_V6=2001:db8:1::
+readonly OL2_NET_V4=172.16.2.
+readonly OL2_NET_V6=2001:db8:2::
+
+trap cleanup_all_ns EXIT
+
+is_ipv6() {
+	if [[ $1 =~ .*:.* ]]; then
+		return 0
+	fi
+	return 1
+}
+
+create_gnv_endpoint() {
+	local -r netns=$1
+	local -r bm_dev=$2
+	local -r bm_rem_addr=$3
+	local -r gnv_dev=$4
+	local -r gnv_id=$5
+	local gnv_json
+	local opts
+	local rem
+	shift 5
+
+	if is_ipv6 $bm_rem_addr; then
+		rem=remote6
+	else
+		rem=remote
+	fi
+
+	while [ -n "$1" ]; do
+		opts="$opts, $1"
+		shift
+	done
+
+	gnv_json="{ \"id\": $gnv_id, \"$rem\": \"$bm_rem_addr\"$opts }"
+	ip netns exec $netns $CLI --family rt-link --create --excl \
+		--do newlink  --json "{\"ifname\": \"$gnv_dev\",
+				       \"linkinfo\": {\"kind\":\"geneve\",
+				       \"data\": $gnv_json } }" > /dev/null
+	ip -n $netns link set dev $gnv_dev up
+}
+
+create_vxlan_endpoint() {
+	local -r netns=$1
+	local -r bm_dev=$2
+	local -r bm_rem_addr=$3
+	local -r vxlan_dev=$4
+	local -r vxlan_id=$5
+	local opts
+	shift 5
+
+	# convert the arguments from yaml format
+	while [ -n "$1" ]; do
+		local opt=$1
+		local pattern='"port":'
+
+		opts="$opts ${opt/$pattern/dstport }"
+		shift
+	done
+	[ -n "$opts" ] || opts="dstport 4789"
+
+	ip -n $netns link add $vxlan_dev type vxlan id $vxlan_id \
+		remote $bm_rem_addr $opts
+	ip -n $netns link set dev $vxlan_dev up
+}
+
+create_ns() {
+	local create_endpoint
+	local nested_opt="$2"
+	local options="$1"
+	local addr_src
+	local addr_dst
+	local feature
+	local dev
+	local id
+	local ns
+
+	RET=0
+
+	#  +-------------+    +-------------+
+	#  | NS_SRC      |    | NS_NST_DST  |
+	#  |             |    |             |
+	#  |   gnv_nst1  |    |  gnv_nst2   |
+	#  |   +         |    |         +   |
+	#  |   |         |    |         |   |
+	#  |   +         |    |         +   |
+	#  |  gnv1       |    |        gnv2 |
+	#  |   +         |    |         +   |
+	#  |   |         |    |         |   |
+	#  |   + veth1 +--------+ veth2 +   |
+	#  |             |    |             |
+	#  +-------------+    +-------------+
+
+	setup_ns NS_SRC NS_DST
+
+	ip link add name veth$SRC netns $NS_SRC type veth \
+		peer name veth$DST netns $NS_DST
+	case "$ENCAP" in
+	vxlan)
+		create_endpoint=create_vxlan_endpoint
+		dev=vx
+		;;
+	geneve)
+		create_endpoint=create_gnv_endpoint
+		dev=gnv
+		;;
+	esac
+
+	id=1
+	for ns in ${NS_LIST[@]}; do
+		ip -n $ns link set dev veth$id up
+
+		# ensure the sender can do large write just after 3whs
+		ip netns exec $ns \
+			sysctl -qw net.ipv4.tcp_wmem="4096 4194304 4194304"
+
+		# note that 3 - $SRC == $DST and 3 - $DST == $SRC
+		if [ $FAMILY = "4" ]; then
+			ip -n $ns addr add dev veth$id $NET_V4$id/24
+			$create_endpoint $ns veth$id $NET_V4$((3 - $id)) \
+				$dev$id 4 $options
+			ip -n $ns addr add dev $dev$id $OL1_NET_V4$id/24
+
+			# nested tunnel devices
+			# pmtu can't be propagated to upper layer devices;
+			# need manual adjust
+			$create_endpoint $ns $dev$id $OL1_NET_V4$((3 - $id)) \
+				$dev"_nst"$id 40 '"port":6082' $nested_opt
+			ip -n $ns addr add dev $dev"_nst"$id $OL2_NET_V4$id/24
+			ip -n $ns link set dev $dev"_nst"$id mtu 1392
+		else
+			ip -n $ns addr add dev veth$id $NET_V6$id/64 nodad
+			$create_endpoint $ns veth$id $NET_V6$((3 - $id)) \
+				$dev"6"$id 6 $options
+			ip -n $ns addr add dev $dev"6"$id $OL1_NET_V6$id/64 \
+				nodad
+
+			$create_endpoint $ns $dev"6"$id $OL1_NET_V6$((3 - $id))\
+				$dev"6_nst"$id 60 '"port":6082' $nested_opt
+			ip -n $ns addr add dev $dev"6_nst"$id $OL2_NET_V6$id/64\
+				nodad
+			ip -n $ns link set dev $dev"6_nst"$id mtu 1352
+		fi
+		id=$((id+1))
+	done
+
+	# enable GRO heuristic on the veth peer and ensure UDP L4 over tunnel is
+	# actually segmented
+	for feature in tso tx-udp_tnl-segmentation; do
+		ip netns exec $NS_SRC ethtool -K veth$SRC $feature off \
+			2>/dev/null
+	done
+}
+
+create_ns_gso()
+{
+	local dev
+
+	create_ns $*
+	if [ $ENCAP = "geneve" ]; then
+		dev=gnv
+	else
+		dev=vx
+	fi
+	if [ $FAMILY = "4" ]; then
+		ip netns exec $NS_SRC ethtool -K $dev$SRC tx-gso-partial on \
+			tx-udp_tnl-segmentation on \
+			tx-udp_tnl-csum-segmentation on
+	else
+		ip netns exec $NS_SRC ethtool -K $dev"6"$SRC tx-gso-partial on \
+			tx-udp_tnl-segmentation on \
+			tx-udp_tnl-csum-segmentation on
+	fi
+}
+
+create_ns_gso_gro()
+{
+	create_ns_gso $*
+	ip netns exec $NS_DST ethtool -K veth$DST gro on
+	ip netns exec $NS_SRC ethtool -K veth$SRC tx off >/dev/null 2>&1
+}
+
+run_test() {
+	local -r dst=$NET$DST
+	local -r msg=$1
+	local -r total_size=$2
+	local -r encappkts=$3
+	local rx_args=""
+	local rx_family="-4"
+	local filter=IpInReceives
+	local ipt=iptables
+	local wire_pkts
+	local dport
+	local pkts
+
+	if [ $FAMILY = "6" ]; then
+		# rx program does not support '-6' and implies ipv6 usage by
+		# default
+		rx_family=""
+		filter=Ip6InReceives
+		ipt=ip6tables
+	fi
+
+	# The received can only check fixed size packet
+	pkts=$((total_size / $GSO_SIZE))
+	rx_args="$rx_family"
+	if [ -n "$4" ]; then
+		wire_pkts=$4
+	elif [ $((total_size % $GSO_SIZE)) -eq 0 ]; then
+		wire_pkts=1
+		rx_args="$rx_args -l $GSO_SIZE"
+	else
+		wire_pkts=2
+		pkts=$((pkts + 1))
+	fi
+
+	if [ $ENCAP = "geneve" ]; then
+		dport=6081
+	else
+		dport=4789
+	fi
+
+	# ignore shorts packet, to avoid arp/mld induced noise
+	ip netns exec $NS_SRC $ipt -A OUTPUT -p udp --dport $dport \
+		-m length --length 300:65535
+	ip netns exec $NS_DST $ipt -A INPUT -p udp --dport $dport \
+		-m length --length 300:65535
+
+	ip netns exec $NS_DST ./udpgso_bench_rx -C 2000 -t -R 100 -n $pkts \
+		$rx_args &
+	local spid=$!
+	wait_local_port_listen "$NS_DST" 8000 tcp
+	ip netns exec $NS_SRC ./udpgso_bench_tx -$FAMILY -t -M 1 \
+		-s $total_size -D $dst
+	local ret=$?
+	check_err $ret "client failure exit code $ret"
+	wait $spid
+	ret=$?
+	check_err $ret "sever failure exit code $ret"
+
+	local snd=$(ip netns exec $NS_SRC $ipt"-save" -c |
+		    grep "dport $dport" | sed -e 's/\[//' -e 's/:.*//')
+	[ "$snd" = "$wire_pkts" ]
+	check_err $? "send $snd packets on the lowest link, expected $wire_pkts"
+
+	local rcvpkts=$(ip netns exec $NS_DST $ipt"-save" -c | \
+			grep "dport $dport" | sed -e 's/\[//' -e 's/:.*//')
+
+	[ "$rcvpkts" = "$encappkts" ]
+	check_err $? "received $rcvpkts $ENCAP packets, expected $encappkts"
+	log_test "$msg"
+}
+
+# tcp retransmisions will break the accounting
+[ "$KSFT_MACHINE_SLOW" = yes ] && FAIL_TO_XFAIL=yes
+for FAMILY in 4 6; do
+	NET=$OL2_NET_V4
+	IPT=iptables
+	GSO_SIZE=1340 # 1392 - 20 - 32
+
+	if [ $FAMILY = 6 ]; then
+		NET=$OL2_NET_V6
+		IPT=ip6tables
+		GSO_SIZE=1280 # 1352 - 40 - 32
+	fi
+
+	echo "IPv$FAMILY"
+
+	# "geneve" must be last encap in list, so that later
+	# test cases will run on it
+	for ENCAP in "vxlan" "geneve"; do
+		create_ns
+		run_test "No GSO - $ENCAP" $((GSO_SIZE * 4)) 4 4
+		cleanup_all_ns
+
+		create_ns_gso
+		run_test "GSO without GRO - $ENCAP" $((GSO_SIZE * 4)) 4 1
+		cleanup_all_ns
+
+		# IPv4 only test
+		[ $FAMILY = "4" ] || continue
+		create_ns_gso
+		ip netns exec $NS_SRC sysctl -qw net.ipv4.ip_no_pmtu_disc=1
+		run_test "GSO disable due to no fixedid - $ENCAP" \
+			$((GSO_SIZE * 4)) 4 4
+		cleanup_all_ns
+	done
+
+	# GRO tests imply/require geneve encap, the only one providing
+	# GRO hints
+	create_ns_gso_gro
+	run_test "double tunnel GRO, no hints" $((GSO_SIZE * 4)) 4
+	cleanup_all_ns
+
+	create_ns_gso_gro '"gro-hint":1'
+	run_test "double tunnel GRO" $((GSO_SIZE * 4)) 1
+	cleanup_all_ns
+
+	create_ns_gso_gro '"gro-hint":1,"udp-csum":1' '"udp-csum":1'
+	run_test "double tunnel GRO - csum complete" $((GSO_SIZE * 4)) 1
+	cleanup_all_ns
+
+	create_ns_gso_gro '"gro-hint":1' \
+		'"udp-csum":1,"udp-zero-csum6-tx":1,"udp-zero-csum6-rx":1'
+	run_test "double tunnel GRO - no nested csum" $((GSO_SIZE * 4)) 1
+	cleanup_all_ns
+
+	create_ns_gso_gro \
+		'"gro-hint":1,"udp-zero-csum6-tx":1,"udp-zero-csum6-rx":1' \
+		'"udp-csum":1'
+	run_test "double tunnel GRO - skip due nested csum with outer 0-csum" \
+		$((GSO_SIZE * 4)) 4
+	cleanup_all_ns
+
+	create_ns_gso_gro '"gro-hint":1,"udp-csum":1' \
+		'"udp-csum":1 "inner-proto-inherit":1'
+	run_test "double tunnel GRO - nested inherit proto" $((GSO_SIZE * 4)) 1
+	cleanup_all_ns
+
+	create_ns_gso_gro '"gro-hint":1'
+	run_test "double tunnel GRO - short last pkt" \
+		$((GSO_SIZE * 4 + $GSO_SIZE / 2)) 2
+	cleanup_all_ns
+done
+
+exit $EXIT_STATUS
-- 
2.52.0

Re: [PATCH net-next 00/10] geneve: introduce double tunnel GSO/GRO

[Jakub Kicinski]

On Tue, 25 Nov 2025 17:11:05 +0100 Paolo Abeni wrote:
> This is the [belated] incarnation of topic discussed in the last Neconf
> [1].

You mentioned off-list that the ai-reported issues here are real, so:
pw-bot: cr