vxlan: Add support for UDP checksums (v4 sending, v6 zero csums)
Added VXLAN link configuration for sending UDP checksums, and allowing
TX and RX of UDP6 checksums.
Also, call common iptunnel_handle_offloads and added GSO support for
checksums.
Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
diff --git a/drivers/net/vxlan.c b/drivers/net/vxlan.c
index e68c8eb..4e2caaf 100644
--- a/drivers/net/vxlan.c
+++ b/drivers/net/vxlan.c
@@ -135,7 +135,7 @@
__u16 port_max;
__u8 tos; /* TOS override */
__u8 ttl;
- u32 flags; /* VXLAN_F_* below */
+ u32 flags; /* VXLAN_F_* in vxlan.h */
struct work_struct sock_work;
struct work_struct igmp_join;
@@ -150,13 +150,6 @@
struct hlist_head fdb_head[FDB_HASH_SIZE];
};
-#define VXLAN_F_LEARN 0x01
-#define VXLAN_F_PROXY 0x02
-#define VXLAN_F_RSC 0x04
-#define VXLAN_F_L2MISS 0x08
-#define VXLAN_F_L3MISS 0x10
-#define VXLAN_F_IPV6 0x20 /* internal flag */
-
/* salt for hash table */
static u32 vxlan_salt __read_mostly;
static struct workqueue_struct *vxlan_wq;
@@ -1601,18 +1594,11 @@
}
EXPORT_SYMBOL_GPL(vxlan_src_port);
-static int handle_offloads(struct sk_buff *skb)
+static inline struct sk_buff *vxlan_handle_offloads(struct sk_buff *skb,
+ bool udp_csum)
{
- if (skb_is_gso(skb)) {
- int err = skb_unclone(skb, GFP_ATOMIC);
- if (unlikely(err))
- return err;
-
- skb_shinfo(skb)->gso_type |= SKB_GSO_UDP_TUNNEL;
- } else if (skb->ip_summed != CHECKSUM_PARTIAL)
- skb->ip_summed = CHECKSUM_NONE;
-
- return 0;
+ int type = udp_csum ? SKB_GSO_UDP_TUNNEL_CSUM : SKB_GSO_UDP_TUNNEL;
+ return iptunnel_handle_offloads(skb, udp_csum, type);
}
#if IS_ENABLED(CONFIG_IPV6)
@@ -1629,10 +1615,9 @@
int min_headroom;
int err;
- if (!skb->encapsulation) {
- skb_reset_inner_headers(skb);
- skb->encapsulation = 1;
- }
+ skb = vxlan_handle_offloads(skb, !udp_get_no_check6_tx(vs->sock->sk));
+ if (IS_ERR(skb))
+ return -EINVAL;
skb_scrub_packet(skb, xnet);
@@ -1666,27 +1651,14 @@
uh->source = src_port;
uh->len = htons(skb->len);
- uh->check = 0;
memset(&(IPCB(skb)->opt), 0, sizeof(IPCB(skb)->opt));
IPCB(skb)->flags &= ~(IPSKB_XFRM_TUNNEL_SIZE | IPSKB_XFRM_TRANSFORMED |
IPSKB_REROUTED);
skb_dst_set(skb, dst);
- if (!skb_is_gso(skb) && !(dst->dev->features & NETIF_F_IPV6_CSUM)) {
- __wsum csum = skb_checksum(skb, 0, skb->len, 0);
- skb->ip_summed = CHECKSUM_UNNECESSARY;
- uh->check = csum_ipv6_magic(saddr, daddr, skb->len,
- IPPROTO_UDP, csum);
- if (uh->check == 0)
- uh->check = CSUM_MANGLED_0;
- } else {
- skb->ip_summed = CHECKSUM_PARTIAL;
- skb->csum_start = skb_transport_header(skb) - skb->head;
- skb->csum_offset = offsetof(struct udphdr, check);
- uh->check = ~csum_ipv6_magic(saddr, daddr,
- skb->len, IPPROTO_UDP, 0);
- }
+ udp6_set_csum(udp_get_no_check6_tx(vs->sock->sk), skb,
+ saddr, daddr, skb->len);
__skb_push(skb, sizeof(*ip6h));
skb_reset_network_header(skb);
@@ -1702,10 +1674,6 @@
ip6h->daddr = *daddr;
ip6h->saddr = *saddr;
- err = handle_offloads(skb);
- if (err)
- return err;
-
ip6tunnel_xmit(skb, dev);
return 0;
}
@@ -1721,10 +1689,9 @@
int min_headroom;
int err;
- if (!skb->encapsulation) {
- skb_reset_inner_headers(skb);
- skb->encapsulation = 1;
- }
+ skb = vxlan_handle_offloads(skb, !vs->sock->sk->sk_no_check_tx);
+ if (IS_ERR(skb))
+ return -EINVAL;
min_headroom = LL_RESERVED_SPACE(rt->dst.dev) + rt->dst.header_len
+ VXLAN_HLEN + sizeof(struct iphdr)
@@ -1756,11 +1723,9 @@
uh->source = src_port;
uh->len = htons(skb->len);
- uh->check = 0;
- err = handle_offloads(skb);
- if (err)
- return err;
+ udp_set_csum(vs->sock->sk->sk_no_check_tx, skb,
+ src, dst, skb->len);
return iptunnel_xmit(vs->sock->sk, rt, skb, src, dst, IPPROTO_UDP,
tos, ttl, df, xnet);
@@ -2405,7 +2370,7 @@
* could be used for both IPv4 and IPv6 communications, but
* users may set bindv6only=1.
*/
-static struct socket *create_v6_sock(struct net *net, __be16 port)
+static struct socket *create_v6_sock(struct net *net, __be16 port, u32 flags)
{
struct sock *sk;
struct socket *sock;
@@ -2442,18 +2407,25 @@
/* Disable multicast loopback */
inet_sk(sk)->mc_loop = 0;
+
+ if (flags & VXLAN_F_UDP_ZERO_CSUM6_TX)
+ udp_set_no_check6_tx(sk, true);
+
+ if (flags & VXLAN_F_UDP_ZERO_CSUM6_RX)
+ udp_set_no_check6_rx(sk, true);
+
return sock;
}
#else
-static struct socket *create_v6_sock(struct net *net, __be16 port)
+static struct socket *create_v6_sock(struct net *net, __be16 port, u32 flags)
{
return ERR_PTR(-EPFNOSUPPORT);
}
#endif
-static struct socket *create_v4_sock(struct net *net, __be16 port)
+static struct socket *create_v4_sock(struct net *net, __be16 port, u32 flags)
{
struct sock *sk;
struct socket *sock;
@@ -2486,18 +2458,24 @@
/* Disable multicast loopback */
inet_sk(sk)->mc_loop = 0;
+
+ if (!(flags & VXLAN_F_UDP_CSUM))
+ sock->sk->sk_no_check_tx = 1;
+
return sock;
}
/* Create new listen socket if needed */
static struct vxlan_sock *vxlan_socket_create(struct net *net, __be16 port,
- vxlan_rcv_t *rcv, void *data, bool ipv6)
+ vxlan_rcv_t *rcv, void *data,
+ u32 flags)
{
struct vxlan_net *vn = net_generic(net, vxlan_net_id);
struct vxlan_sock *vs;
struct socket *sock;
struct sock *sk;
unsigned int h;
+ bool ipv6 = !!(flags & VXLAN_F_IPV6);
vs = kzalloc(sizeof(*vs), GFP_KERNEL);
if (!vs)
@@ -2509,9 +2487,9 @@
INIT_WORK(&vs->del_work, vxlan_del_work);
if (ipv6)
- sock = create_v6_sock(net, port);
+ sock = create_v6_sock(net, port, flags);
else
- sock = create_v4_sock(net, port);
+ sock = create_v4_sock(net, port, flags);
if (IS_ERR(sock)) {
kfree(vs);
return ERR_CAST(sock);
@@ -2549,12 +2527,12 @@
struct vxlan_sock *vxlan_sock_add(struct net *net, __be16 port,
vxlan_rcv_t *rcv, void *data,
- bool no_share, bool ipv6)
+ bool no_share, u32 flags)
{
struct vxlan_net *vn = net_generic(net, vxlan_net_id);
struct vxlan_sock *vs;
- vs = vxlan_socket_create(net, port, rcv, data, ipv6);
+ vs = vxlan_socket_create(net, port, rcv, data, flags);
if (!IS_ERR(vs))
return vs;
@@ -2587,7 +2565,7 @@
__be16 port = vxlan->dst_port;
struct vxlan_sock *nvs;
- nvs = vxlan_sock_add(net, port, vxlan_rcv, NULL, false, vxlan->flags & VXLAN_F_IPV6);
+ nvs = vxlan_sock_add(net, port, vxlan_rcv, NULL, false, vxlan->flags);
spin_lock(&vn->sock_lock);
if (!IS_ERR(nvs))
vxlan_vs_add_dev(nvs, vxlan);
@@ -2711,6 +2689,17 @@
if (data[IFLA_VXLAN_PORT])
vxlan->dst_port = nla_get_be16(data[IFLA_VXLAN_PORT]);
+ if (data[IFLA_VXLAN_UDP_CSUM] && nla_get_u8(data[IFLA_VXLAN_UDP_CSUM]))
+ vxlan->flags |= VXLAN_F_UDP_CSUM;
+
+ if (data[IFLA_VXLAN_UDP_ZERO_CSUM6_TX] &&
+ nla_get_u8(data[IFLA_VXLAN_UDP_ZERO_CSUM6_TX]))
+ vxlan->flags |= VXLAN_F_UDP_ZERO_CSUM6_TX;
+
+ if (data[IFLA_VXLAN_UDP_ZERO_CSUM6_RX] &&
+ nla_get_u8(data[IFLA_VXLAN_UDP_ZERO_CSUM6_RX]))
+ vxlan->flags |= VXLAN_F_UDP_ZERO_CSUM6_RX;
+
if (vxlan_find_vni(net, vni, vxlan->dst_port)) {
pr_info("duplicate VNI %u\n", vni);
return -EEXIST;
@@ -2774,7 +2763,10 @@
nla_total_size(sizeof(__u32)) + /* IFLA_VXLAN_AGEING */
nla_total_size(sizeof(__u32)) + /* IFLA_VXLAN_LIMIT */
nla_total_size(sizeof(struct ifla_vxlan_port_range)) +
- nla_total_size(sizeof(__be16))+ /* IFLA_VXLAN_PORT */
+ nla_total_size(sizeof(__be16)) + /* IFLA_VXLAN_PORT */
+ nla_total_size(sizeof(__u8)) + /* IFLA_VXLAN_UDP_CSUM */
+ nla_total_size(sizeof(__u8)) + /* IFLA_VXLAN_UDP_ZERO_CSUM6_TX */
+ nla_total_size(sizeof(__u8)) + /* IFLA_VXLAN_UDP_ZERO_CSUM6_RX */
0;
}
@@ -2834,7 +2826,13 @@
!!(vxlan->flags & VXLAN_F_L3MISS)) ||
nla_put_u32(skb, IFLA_VXLAN_AGEING, vxlan->age_interval) ||
nla_put_u32(skb, IFLA_VXLAN_LIMIT, vxlan->addrmax) ||
- nla_put_be16(skb, IFLA_VXLAN_PORT, vxlan->dst_port))
+ nla_put_be16(skb, IFLA_VXLAN_PORT, vxlan->dst_port) ||
+ nla_put_u8(skb, IFLA_VXLAN_UDP_CSUM,
+ !!(vxlan->flags & VXLAN_F_UDP_CSUM)) ||
+ nla_put_u8(skb, IFLA_VXLAN_UDP_ZERO_CSUM6_TX,
+ !!(vxlan->flags & VXLAN_F_UDP_ZERO_CSUM6_TX)) ||
+ nla_put_u8(skb, IFLA_VXLAN_UDP_ZERO_CSUM6_RX,
+ !!(vxlan->flags & VXLAN_F_UDP_ZERO_CSUM6_RX)))
goto nla_put_failure;
if (nla_put(skb, IFLA_VXLAN_PORT_RANGE, sizeof(ports), &ports))
diff --git a/include/net/vxlan.h b/include/net/vxlan.h
index 7bb4084..12196ce6 100644
--- a/include/net/vxlan.h
+++ b/include/net/vxlan.h
@@ -24,9 +24,19 @@
struct udp_offload udp_offloads;
};
+#define VXLAN_F_LEARN 0x01
+#define VXLAN_F_PROXY 0x02
+#define VXLAN_F_RSC 0x04
+#define VXLAN_F_L2MISS 0x08
+#define VXLAN_F_L3MISS 0x10
+#define VXLAN_F_IPV6 0x20
+#define VXLAN_F_UDP_CSUM 0x40
+#define VXLAN_F_UDP_ZERO_CSUM6_TX 0x80
+#define VXLAN_F_UDP_ZERO_CSUM6_RX 0x100
+
struct vxlan_sock *vxlan_sock_add(struct net *net, __be16 port,
vxlan_rcv_t *rcv, void *data,
- bool no_share, bool ipv6);
+ bool no_share, u32 flags);
void vxlan_sock_release(struct vxlan_sock *vs);
diff --git a/include/uapi/linux/if_link.h b/include/uapi/linux/if_link.h
index 622e7910..b385348 100644
--- a/include/uapi/linux/if_link.h
+++ b/include/uapi/linux/if_link.h
@@ -319,6 +319,9 @@
IFLA_VXLAN_PORT, /* destination port */
IFLA_VXLAN_GROUP6,
IFLA_VXLAN_LOCAL6,
+ IFLA_VXLAN_UDP_CSUM,
+ IFLA_VXLAN_UDP_ZERO_CSUM6_TX,
+ IFLA_VXLAN_UDP_ZERO_CSUM6_RX,
__IFLA_VXLAN_MAX
};
#define IFLA_VXLAN_MAX (__IFLA_VXLAN_MAX - 1)
diff --git a/net/openvswitch/vport-vxlan.c b/net/openvswitch/vport-vxlan.c
index a93efa3..0edbd95 100644
--- a/net/openvswitch/vport-vxlan.c
+++ b/net/openvswitch/vport-vxlan.c
@@ -122,7 +122,7 @@
vxlan_port = vxlan_vport(vport);
strncpy(vxlan_port->name, parms->name, IFNAMSIZ);
- vs = vxlan_sock_add(net, htons(dst_port), vxlan_rcv, vport, true, false);
+ vs = vxlan_sock_add(net, htons(dst_port), vxlan_rcv, vport, true, 0);
if (IS_ERR(vs)) {
ovs_vport_free(vport);
return (void *)vs;