#define GENEVE_DST_PORT 6081
#define VXLAN_DST_PORT 4789
#define LISP_DST_PORT 4341
+#define STT_DST_PORT 7471
#define VXLAN_HLEN (sizeof(struct eth_header) + \
sizeof(struct ip_header) + \
sizeof(struct udp_header) + \
sizeof(struct vxlanhdr))
+#define GENEVE_BASE_HLEN (sizeof(struct eth_header) + \
+ sizeof(struct ip_header) + \
+ sizeof(struct udp_header) + \
+ sizeof(struct genevehdr))
+
#define DEFAULT_TTL 64
struct netdev_vport {
return (class->get_config == get_tunnel_config &&
(!strcmp("geneve", type) || !strcmp("vxlan", type) ||
- !strcmp("lisp", type)));
+ !strcmp("lisp", type) || !strcmp("stt", type)) );
}
const char *
dev->tnl_cfg.dst_port = htons(VXLAN_DST_PORT);
} else if (!strcmp(type, "lisp")) {
dev->tnl_cfg.dst_port = htons(LISP_DST_PORT);
+ } else if (!strcmp(type, "stt")) {
+ dev->tnl_cfg.dst_port = htons(STT_DST_PORT);
}
+ dev->tnl_cfg.dont_fragment = true;
+ dev->tnl_cfg.ttl = DEFAULT_TTL;
return 0;
}
struct smap_node *node;
has_csum = strstr(type, "gre") || strstr(type, "geneve") ||
- strstr(type, "vxlan");
+ strstr(type, "stt") || strstr(type, "vxlan");
ipsec_mech_set = false;
memset(&tnl_cfg, 0, sizeof tnl_cfg);
tnl_cfg.dst_port = htons(LISP_DST_PORT);
}
+ if (!strcmp(type, "stt")) {
+ tnl_cfg.dst_port = htons(STT_DST_PORT);
+ }
+
needs_dst_port = netdev_vport_needs_dst_port(dev_);
tnl_cfg.ipsec = strstr(type, "ipsec");
tnl_cfg.dont_fragment = true;
if ((!strcmp("geneve", type) && dst_port != GENEVE_DST_PORT) ||
(!strcmp("vxlan", type) && dst_port != VXLAN_DST_PORT) ||
- (!strcmp("lisp", type) && dst_port != LISP_DST_PORT)) {
+ (!strcmp("lisp", type) && dst_port != LISP_DST_PORT) ||
+ (!strcmp("stt", type) && dst_port != STT_DST_PORT)) {
smap_add_format(args, "dst_port", "%d", dst_port);
}
}
return ip + 1;
}
+static void *
+udp_extract_tnl_md(struct dp_packet *packet, struct flow_tnl *tnl)
+{
+ struct udp_header *udp;
+
+ udp = ip_extract_tnl_md(packet, tnl);
+ if (!udp) {
+ return NULL;
+ }
+
+ if (udp->udp_csum) {
+ uint32_t csum = packet_csum_pseudoheader(dp_packet_l3(packet));
+
+ csum = csum_continue(csum, udp, dp_packet_size(packet) -
+ ((const unsigned char *)udp -
+ (const unsigned char *)dp_packet_l2(packet)));
+ if (csum_finish(csum)) {
+ return NULL;
+ }
+ tnl->flags |= FLOW_TNL_F_CSUM;
+ }
+
+ tnl->tp_src = udp->udp_src;
+ tnl->tp_dst = udp->udp_dst;
+
+ return udp + 1;
+}
+
+static ovs_be16
+get_src_port(struct dp_packet *packet)
+{
+ uint32_t hash;
+
+ hash = dp_packet_get_rss_hash(packet);
+
+ return htons((((uint64_t) hash * (tnl_udp_port_max - tnl_udp_port_min)) >> 32) +
+ tnl_udp_port_min);
+}
+
+static void
+push_udp_header(struct dp_packet *packet,
+ const struct ovs_action_push_tnl *data)
+{
+ struct udp_header *udp;
+ int ip_tot_size;
+
+ udp = push_ip_header(packet, data->header, data->header_len, &ip_tot_size);
+
+ /* set udp src port */
+ udp->udp_src = get_src_port(packet);
+ udp->udp_len = htons(ip_tot_size - sizeof (struct ip_header));
+
+ if (udp->udp_csum) {
+ uint32_t csum = packet_csum_pseudoheader(ip_hdr(dp_packet_data(packet)));
+
+ csum = csum_continue(csum, udp,
+ ip_tot_size - sizeof (struct ip_header));
+ udp->udp_csum = csum_finish(csum);
+
+ if (!udp->udp_csum) {
+ udp->udp_csum = htons(0xffff);
+ }
+ }
+}
+
+static void *
+udp_build_header(struct netdev_tunnel_config *tnl_cfg,
+ const struct flow *tnl_flow,
+ struct ovs_action_push_tnl *data)
+{
+ struct ip_header *ip;
+ struct udp_header *udp;
+
+ ip = ip_hdr(data->header);
+ ip->ip_proto = IPPROTO_UDP;
+
+ udp = (struct udp_header *) (ip + 1);
+ udp->udp_dst = tnl_cfg->dst_port;
+
+ if (tnl_flow->tunnel.flags & FLOW_TNL_F_CSUM) {
+ /* Write a value in now to mark that we should compute the checksum
+ * later. 0xffff is handy because it is transparent to the
+ * calculation. */
+ udp->udp_csum = htons(0xffff);
+ }
+
+ return udp + 1;
+}
+
static int
gre_header_len(ovs_be16 flags)
{
}
static void
-reset_tnl_md(struct pkt_metadata *md)
+pkt_metadata_init_tnl(struct pkt_metadata *md)
{
- memset(&md->tunnel, 0, sizeof(md->tunnel));
+ memset(md, 0, offsetof(struct pkt_metadata, tunnel.metadata));
+
+ /* If 'opt_map' is zero then none of the rest of the tunnel metadata
+ * will be read, so we can skip clearing it. */
+ md->tunnel.metadata.opt_map = 0;
}
-static void
-gre_extract_md(struct dp_packet *packet)
+static int
+netdev_gre_pop_header(struct dp_packet *packet)
{
struct pkt_metadata *md = &packet->md;
struct flow_tnl *tnl = &md->tunnel;
int hlen = sizeof(struct eth_header) +
sizeof(struct ip_header) + 4;
- memset(md, 0, sizeof *md);
+ pkt_metadata_init_tnl(md);
if (hlen > dp_packet_size(packet)) {
- return;
+ return EINVAL;
}
hlen = parse_gre_header(packet, tnl);
if (hlen < 0) {
- reset_tnl_md(md);
+ return -hlen;
}
dp_packet_reset_packet(packet, hlen);
-}
-static int
-netdev_gre_pop_header(struct netdev *netdev_ OVS_UNUSED,
- struct dp_packet **pkt, int cnt)
-{
- int i;
-
- for (i = 0; i < cnt; i++) {
- gre_extract_md(pkt[i]);
- }
return 0;
}
static void
-netdev_gre_push_header__(struct dp_packet *packet,
- const void *header, int size)
+netdev_gre_push_header(struct dp_packet *packet,
+ const struct ovs_action_push_tnl *data)
{
struct gre_base_hdr *greh;
int ip_tot_size;
- greh = push_ip_header(packet, header, size, &ip_tot_size);
+ greh = push_ip_header(packet, data->header, data->header_len, &ip_tot_size);
if (greh->flags & htons(GRE_CSUM)) {
ovs_be16 *csum_opt = (ovs_be16 *) (greh + 1);
}
}
-static int
-netdev_gre_push_header(const struct netdev *netdev OVS_UNUSED,
- struct dp_packet **packets, int cnt,
- const struct ovs_action_push_tnl *data)
-{
- int i;
-
- for (i = 0; i < cnt; i++) {
- netdev_gre_push_header__(packets[i], data->header, data->header_len);
- packets[i]->md = PKT_METADATA_INITIALIZER(u32_to_odp(data->out_port));
- }
- return 0;
-}
-
-
static int
netdev_gre_build_header(const struct netdev *netdev,
struct ovs_action_push_tnl *data,
return 0;
}
-static void
-vxlan_extract_md(struct dp_packet *packet)
+static int
+netdev_vxlan_pop_header(struct dp_packet *packet)
{
struct pkt_metadata *md = &packet->md;
struct flow_tnl *tnl = &md->tunnel;
- struct udp_header *udp;
struct vxlanhdr *vxh;
- memset(md, 0, sizeof *md);
+ pkt_metadata_init_tnl(md);
if (VXLAN_HLEN > dp_packet_size(packet)) {
- return;
+ return EINVAL;
}
- udp = ip_extract_tnl_md(packet, tnl);
- if (!udp) {
- return;
+ vxh = udp_extract_tnl_md(packet, tnl);
+ if (!vxh) {
+ return EINVAL;
}
- vxh = (struct vxlanhdr *) (udp + 1);
if (get_16aligned_be32(&vxh->vx_flags) != htonl(VXLAN_FLAGS) ||
(get_16aligned_be32(&vxh->vx_vni) & htonl(0xff))) {
VLOG_WARN_RL(&err_rl, "invalid vxlan flags=%#x vni=%#x\n",
ntohl(get_16aligned_be32(&vxh->vx_flags)),
ntohl(get_16aligned_be32(&vxh->vx_vni)));
- reset_tnl_md(md);
- return;
+ return EINVAL;
}
- tnl->tp_src = udp->udp_src;
- tnl->tp_dst = udp->udp_dst;
tnl->tun_id = htonll(ntohl(get_16aligned_be32(&vxh->vx_vni)) >> 8);
+ tnl->flags |= FLOW_TNL_F_KEY;
dp_packet_reset_packet(packet, VXLAN_HLEN);
-}
-static int
-netdev_vxlan_pop_header(struct netdev *netdev_ OVS_UNUSED,
- struct dp_packet **pkt, int cnt)
-{
- int i;
-
- for (i = 0; i < cnt; i++) {
- vxlan_extract_md(pkt[i]);
- }
return 0;
}
{
struct netdev_vport *dev = netdev_vport_cast(netdev);
struct netdev_tunnel_config *tnl_cfg;
- struct ip_header *ip;
- struct udp_header *udp;
struct vxlanhdr *vxh;
/* XXX: RCUfy tnl_cfg. */
ovs_mutex_lock(&dev->mutex);
tnl_cfg = &dev->tnl_cfg;
- ip = ip_hdr(data->header);
- ip->ip_proto = IPPROTO_UDP;
+ vxh = udp_build_header(tnl_cfg, tnl_flow, data);
- udp = (struct udp_header *) (ip + 1);
- udp->udp_dst = tnl_cfg->dst_port;
-
- vxh = (struct vxlanhdr *) (udp + 1);
put_16aligned_be32(&vxh->vx_flags, htonl(VXLAN_FLAGS));
put_16aligned_be32(&vxh->vx_vni, htonl(ntohll(tnl_flow->tunnel.tun_id) << 8));
return 0;
}
-static ovs_be16
-get_src_port(struct dp_packet *packet)
+static int
+netdev_geneve_pop_header(struct dp_packet *packet)
{
- uint32_t hash;
+ struct pkt_metadata *md = &packet->md;
+ struct flow_tnl *tnl = &md->tunnel;
+ struct genevehdr *gnh;
+ unsigned int hlen;
+ int err;
+
+ pkt_metadata_init_tnl(md);
+ if (GENEVE_BASE_HLEN > dp_packet_size(packet)) {
+ VLOG_WARN_RL(&err_rl, "geneve packet too small: min header=%u packet size=%u\n",
+ (unsigned int)GENEVE_BASE_HLEN, dp_packet_size(packet));
+ return EINVAL;
+ }
- hash = dp_packet_get_dp_hash(packet);
+ gnh = udp_extract_tnl_md(packet, tnl);
+ if (!gnh) {
+ return EINVAL;
+ }
- return htons((((uint64_t) hash * (tnl_udp_port_max - tnl_udp_port_min)) >> 32) +
- tnl_udp_port_min);
-}
+ hlen = GENEVE_BASE_HLEN + gnh->opt_len * 4;
+ if (hlen > dp_packet_size(packet)) {
+ VLOG_WARN_RL(&err_rl, "geneve packet too small: header len=%u packet size=%u\n",
+ hlen, dp_packet_size(packet));
+ return EINVAL;
+ }
-static void
-netdev_vxlan_push_header__(struct dp_packet *packet,
- const void *header, int size)
-{
- struct udp_header *udp;
- int ip_tot_size;
+ if (gnh->ver != 0) {
+ VLOG_WARN_RL(&err_rl, "unknown geneve version: %"PRIu8"\n", gnh->ver);
+ return EINVAL;
+ }
- udp = push_ip_header(packet, header, size, &ip_tot_size);
+ if (gnh->proto_type != htons(ETH_TYPE_TEB)) {
+ VLOG_WARN_RL(&err_rl, "unknown geneve encapsulated protocol: %#x\n",
+ ntohs(gnh->proto_type));
+ return EINVAL;
+ }
- /* set udp src port */
- udp->udp_src = get_src_port(packet);
- udp->udp_len = htons(ip_tot_size - sizeof (struct ip_header));
- /* udp_csum is zero */
+ tnl->flags |= gnh->oam ? FLOW_TNL_F_OAM : 0;
+ tnl->tun_id = htonll(ntohl(get_16aligned_be32(&gnh->vni)) >> 8);
+ tnl->flags |= FLOW_TNL_F_KEY;
+
+ err = tun_metadata_from_geneve_header(gnh->options, gnh->opt_len * 4,
+ &tnl->metadata);
+ if (err) {
+ VLOG_WARN_RL(&err_rl, "invalid geneve options");
+ return err;
+ }
+
+ dp_packet_reset_packet(packet, hlen);
+
+ return 0;
}
static int
-netdev_vxlan_push_header(const struct netdev *netdev OVS_UNUSED,
- struct dp_packet **packets, int cnt,
- const struct ovs_action_push_tnl *data)
+netdev_geneve_build_header(const struct netdev *netdev,
+ struct ovs_action_push_tnl *data,
+ const struct flow *tnl_flow)
{
- int i;
+ struct netdev_vport *dev = netdev_vport_cast(netdev);
+ struct netdev_tunnel_config *tnl_cfg;
+ struct genevehdr *gnh;
+ int opt_len;
+ bool crit_opt;
- for (i = 0; i < cnt; i++) {
- netdev_vxlan_push_header__(packets[i],
- data->header, VXLAN_HLEN);
- packets[i]->md = PKT_METADATA_INITIALIZER(u32_to_odp(data->out_port));
- }
+ /* XXX: RCUfy tnl_cfg. */
+ ovs_mutex_lock(&dev->mutex);
+ tnl_cfg = &dev->tnl_cfg;
+
+ gnh = udp_build_header(tnl_cfg, tnl_flow, data);
+
+ put_16aligned_be32(&gnh->vni, htonl(ntohll(tnl_flow->tunnel.tun_id) << 8));
+
+ ovs_mutex_unlock(&dev->mutex);
+
+ opt_len = tun_metadata_to_geneve_header(&tnl_flow->tunnel.metadata,
+ gnh->options, &crit_opt);
+
+ gnh->opt_len = opt_len / 4;
+ gnh->oam = !!(tnl_flow->tunnel.flags & FLOW_TNL_F_OAM);
+ gnh->critical = crit_opt ? 1 : 0;
+ gnh->proto_type = htons(ETH_TYPE_TEB);
+
+ data->header_len = GENEVE_BASE_HLEN + opt_len;
+ data->tnl_type = OVS_VPORT_TYPE_GENEVE;
return 0;
}
/* The name of the dpif_port should be short enough to accomodate adding
* a port number to the end if one is necessary. */
static const struct vport_class vport_classes[] = {
- TUNNEL_CLASS("geneve", "genev_sys", NULL, NULL, NULL),
+ TUNNEL_CLASS("geneve", "genev_sys", netdev_geneve_build_header,
+ push_udp_header,
+ netdev_geneve_pop_header),
TUNNEL_CLASS("gre", "gre_sys", netdev_gre_build_header,
netdev_gre_push_header,
netdev_gre_pop_header),
TUNNEL_CLASS("gre64", "gre64_sys", NULL, NULL, NULL),
TUNNEL_CLASS("ipsec_gre64", "gre64_sys", NULL, NULL, NULL),
TUNNEL_CLASS("vxlan", "vxlan_sys", netdev_vxlan_build_header,
- netdev_vxlan_push_header,
+ push_udp_header,
netdev_vxlan_pop_header),
- TUNNEL_CLASS("lisp", "lisp_sys", NULL, NULL, NULL)
+ TUNNEL_CLASS("lisp", "lisp_sys", NULL, NULL, NULL),
+ TUNNEL_CLASS("stt", "stt_sys", NULL, NULL, NULL),
};
static struct ovsthread_once once = OVSTHREAD_ONCE_INITIALIZER;
projects / cascardo / ovs.git / blobdiff