This new field allows you to restrict the metadata template for a given
tunnel driver. This is convenient in scenarios that combine different
tunneling drivers, eg. vxlan and erspan. This helps you deal with
possible incorrect configurations. Default value is TUNNEL_TYPE_UNSPEC,
to retain the existing behaviour. This also implicitly exposes what
drivers are currently supported in the TUNNEL_INFO_TX mode.
Signed-off-by: Pablo Neira Ayuso <pablo@xxxxxxxxxxxxx>
---
drivers/net/geneve.c | 3 ++-
drivers/net/vxlan.c | 13 +++++++------
include/net/dst_metadata.h | 1 +
include/net/ip_tunnels.h | 8 ++++++++
include/uapi/linux/if_tunnel.h | 13 ++++++++++++-
net/core/filter.c | 1 +
net/ipv4/ip_gre.c | 2 ++
net/ipv4/ip_tunnel.c | 3 ++-
net/ipv6/ip6_gre.c | 2 ++
net/ipv6/ip6_tunnel.c | 6 ++++--
net/openvswitch/flow_netlink.c | 1 +
11 files changed, 42 insertions(+), 11 deletions(-)
diff --git a/drivers/net/geneve.c b/drivers/net/geneve.c
index 82eccc930c5c..e4fd2acb6732 100644
--- a/drivers/net/geneve.c
+++ b/drivers/net/geneve.c
@@ -920,7 +920,8 @@ static netdev_tx_t geneve_xmit(struct sk_buff *skb, struct net_device *dev)
if (geneve->collect_md) {
info = skb_tunnel_info(skb);
- if (unlikely(!info || !(info->mode & IP_TUNNEL_INFO_TX))) {
+ if (unlikely(!info || !(info->mode & IP_TUNNEL_INFO_TX) ||
+ !ip_tunnel_type(info, TUNNEL_TYPE_GENEVE))) {
err = -EINVAL;
netdev_dbg(dev, "no tunnel metadata\n");
goto tx_error;
diff --git a/drivers/net/vxlan.c b/drivers/net/vxlan.c
index fb0cdbba8d76..c279c50816cf 100644
--- a/drivers/net/vxlan.c
+++ b/drivers/net/vxlan.c
@@ -2296,14 +2296,15 @@ static netdev_tx_t vxlan_xmit(struct sk_buff *skb, struct net_device *dev)
skb_reset_mac_header(skb);
if (vxlan->cfg.flags & VXLAN_F_COLLECT_METADATA) {
- if (info && info->mode & IP_TUNNEL_INFO_BRIDGE &&
- info->mode & IP_TUNNEL_INFO_TX) {
+ if (unlikely(!info || !(info->mode & IP_TUNNEL_INFO_TX) ||
+ !ip_tunnel_type(info, TUNNEL_TYPE_VXLAN))) {
+ kfree_skb(skb);
+ return NETDEV_TX_OK;
+ }
+ if (info->mode & IP_TUNNEL_INFO_BRIDGE) {
vni = tunnel_id_to_key32(info->key.tun_id);
} else {
- if (info && info->mode & IP_TUNNEL_INFO_TX)
- vxlan_xmit_one(skb, dev, vni, NULL, false);
- else
- kfree_skb(skb);
+ vxlan_xmit_one(skb, dev, vni, NULL, false);
return NETDEV_TX_OK;
}
}
diff --git a/include/net/dst_metadata.h b/include/net/dst_metadata.h
index 56cb3c38569a..fb3865b2f038 100644
--- a/include/net/dst_metadata.h
+++ b/include/net/dst_metadata.h
@@ -100,6 +100,7 @@ static inline struct metadata_dst *tun_rx_dst(int md_size)
if (!tun_dst)
return NULL;
+ tun_dst->u.tun_info.type = TUNNEL_TYPE_UNSPEC;
tun_dst->u.tun_info.options_len = 0;
tun_dst->u.tun_info.mode = 0;
return tun_dst;
diff --git a/include/net/ip_tunnels.h b/include/net/ip_tunnels.h
index b0d022ff6ea1..34d748ca8b30 100644
--- a/include/net/ip_tunnels.h
+++ b/include/net/ip_tunnels.h
@@ -67,6 +67,7 @@ struct ip_tunnel_key {
options_len) * BITS_PER_BYTE) - 1, 0)
struct ip_tunnel_info {
+ enum tunnel_type type;
struct ip_tunnel_key key;
#ifdef CONFIG_DST_CACHE
struct dst_cache dst_cache;
@@ -75,6 +76,13 @@ struct ip_tunnel_info {
u8 mode;
};
+static inline bool ip_tunnel_type(const struct ip_tunnel_info *tun_info,
+ enum tunnel_type type)
+{
+ return tun_info->type == TUNNEL_TYPE_UNSPEC ||
+ tun_info->type == type;
+}
+
/* 6rd prefix/relay information */
#ifdef CONFIG_IPV6_SIT_6RD
struct ip_tunnel_6rd_parm {
diff --git a/include/uapi/linux/if_tunnel.h b/include/uapi/linux/if_tunnel.h
index 1b3d148c4560..1659b81fbae6 100644
--- a/include/uapi/linux/if_tunnel.h
+++ b/include/uapi/linux/if_tunnel.h
@@ -158,6 +158,17 @@ enum {
IFLA_VTI_FWMARK,
__IFLA_VTI_MAX,
};
-
#define IFLA_VTI_MAX (__IFLA_VTI_MAX - 1)
+
+enum tunnel_type {
+ TUNNEL_TYPE_UNSPEC = 0,
+ TUNNEL_TYPE_GRE,
+ TUNNEL_TYPE_VXLAN,
+ TUNNEL_TYPE_GENEVE,
+ TUNNEL_TYPE_ERSPAN,
+ TUNNEL_TYPE_IPIP,
+ TUNNEL_TYPE_IPIP6,
+ TUNNEL_TYPE_IP6IP6,
+};
+
#endif /* _UAPI_IF_TUNNEL_H_ */
diff --git a/net/core/filter.c b/net/core/filter.c
index 4bbc6567fcb8..2e75e1b014df 100644
--- a/net/core/filter.c
+++ b/net/core/filter.c
@@ -3654,6 +3654,7 @@ BPF_CALL_4(bpf_skb_set_tunnel_key, struct sk_buff *, skb,
info = &md->u.tun_info;
memset(info, 0, sizeof(*info));
info->mode = IP_TUNNEL_INFO_TX;
+ info->type = TUNNEL_TYPE_UNSPEC;
info->key.tun_flags = TUNNEL_KEY | TUNNEL_CSUM | TUNNEL_NOCACHE;
if (flags & BPF_F_DONT_FRAGMENT)
diff --git a/net/ipv4/ip_gre.c b/net/ipv4/ip_gre.c
index 38befe829caf..ab5de3901bc3 100644
--- a/net/ipv4/ip_gre.c
+++ b/net/ipv4/ip_gre.c
@@ -534,6 +534,7 @@ static void gre_fb_xmit(struct sk_buff *skb, struct net_device *dev,
tun_info = skb_tunnel_info(skb);
if (unlikely(!tun_info || !(tun_info->mode & IP_TUNNEL_INFO_TX) ||
+ !ip_tunnel_type(tun_info, TUNNEL_TYPE_GRE) ||
ip_tunnel_info_af(tun_info) != AF_INET))
goto err_free_skb;
@@ -585,6 +586,7 @@ static void erspan_fb_xmit(struct sk_buff *skb, struct net_device *dev,
tun_info = skb_tunnel_info(skb);
if (unlikely(!tun_info || !(tun_info->mode & IP_TUNNEL_INFO_TX) ||
+ !ip_tunnel_type(tun_info, TUNNEL_TYPE_ERSPAN) ||
ip_tunnel_info_af(tun_info) != AF_INET))
goto err_free_skb;
diff --git a/net/ipv4/ip_tunnel.c b/net/ipv4/ip_tunnel.c
index 284a22154b4e..b62b424183b4 100644
--- a/net/ipv4/ip_tunnel.c
+++ b/net/ipv4/ip_tunnel.c
@@ -562,7 +562,8 @@ void ip_md_tunnel_xmit(struct sk_buff *skb, struct net_device *dev, u8 proto)
tun_info = skb_tunnel_info(skb);
if (unlikely(!tun_info || !(tun_info->mode & IP_TUNNEL_INFO_TX) ||
- ip_tunnel_info_af(tun_info) != AF_INET))
+ ip_tunnel_info_af(tun_info) != AF_INET) ||
+ !ip_tunnel_type(tun_info, TUNNEL_TYPE_IPIP))
goto tx_error;
key = &tun_info->key;
memset(&(IPCB(skb)->opt), 0, sizeof(IPCB(skb)->opt));
diff --git a/net/ipv6/ip6_gre.c b/net/ipv6/ip6_gre.c
index 515adbdba1d2..7ff469652e6f 100644
--- a/net/ipv6/ip6_gre.c
+++ b/net/ipv6/ip6_gre.c
@@ -732,6 +732,7 @@ static netdev_tx_t __gre6_xmit(struct sk_buff *skb,
tun_info = skb_tunnel_info(skb);
if (unlikely(!tun_info ||
!(tun_info->mode & IP_TUNNEL_INFO_TX) ||
+ !ip_tunnel_type(tun_info, TUNNEL_TYPE_GRE) ||
ip_tunnel_info_af(tun_info) != AF_INET6))
return -EINVAL;
@@ -960,6 +961,7 @@ static netdev_tx_t ip6erspan_tunnel_xmit(struct sk_buff *skb,
tun_info = skb_tunnel_info(skb);
if (unlikely(!tun_info ||
!(tun_info->mode & IP_TUNNEL_INFO_TX) ||
+ !ip_tunnel_type(tun_info, TUNNEL_TYPE_ERSPAN) ||
ip_tunnel_info_af(tun_info) != AF_INET6))
return -EINVAL;
diff --git a/net/ipv6/ip6_tunnel.c b/net/ipv6/ip6_tunnel.c
index a0b6932c3afd..819abbcd2ebe 100644
--- a/net/ipv6/ip6_tunnel.c
+++ b/net/ipv6/ip6_tunnel.c
@@ -1259,7 +1259,8 @@ ip4ip6_tnl_xmit(struct sk_buff *skb, struct net_device *dev)
tun_info = skb_tunnel_info(skb);
if (unlikely(!tun_info || !(tun_info->mode & IP_TUNNEL_INFO_TX) ||
- ip_tunnel_info_af(tun_info) != AF_INET6))
+ ip_tunnel_info_af(tun_info) != AF_INET6) ||
+ !ip_tunnel_type(tun_info, TUNNEL_TYPE_IPIP6))
return -1;
key = &tun_info->key;
memset(&fl6, 0, sizeof(fl6));
@@ -1335,7 +1336,8 @@ ip6ip6_tnl_xmit(struct sk_buff *skb, struct net_device *dev)
tun_info = skb_tunnel_info(skb);
if (unlikely(!tun_info || !(tun_info->mode & IP_TUNNEL_INFO_TX) ||
- ip_tunnel_info_af(tun_info) != AF_INET6))
+ ip_tunnel_info_af(tun_info) != AF_INET6 ||
+ !ip_tunnel_type(tun_info, TUNNEL_TYPE_IP6IP6)))
return -1;
key = &tun_info->key;
memset(&fl6, 0, sizeof(fl6));
diff --git a/net/openvswitch/flow_netlink.c b/net/openvswitch/flow_netlink.c
index a70097ecf33c..29623ad54611 100644
--- a/net/openvswitch/flow_netlink.c
+++ b/net/openvswitch/flow_netlink.c
@@ -2602,6 +2602,7 @@ static int validate_and_copy_set_tun(const struct nlattr *attr,
ovs_tun->tun_dst = tun_dst;
tun_info = &tun_dst->u.tun_info;
+ tun_info->type = TUNNEL_TYPE_UNSPEC;
tun_info->mode = IP_TUNNEL_INFO_TX;
if (key.tun_proto == AF_INET6)
tun_info->mode |= IP_TUNNEL_INFO_IPV6;
--
2.11.0
