diff options
author | Ed Warnicke <eaw@cisco.com> | 2015-12-08 15:45:58 -0700 |
---|---|---|
committer | Ed Warnicke <eaw@cisco.com> | 2015-12-08 15:47:27 -0700 |
commit | cb9cadad578297ffd78fa8a33670bdf1ab669e7e (patch) | |
tree | 6ac2be912482cc7849a26f0ab845561c3d7f4e26 /vnet/vnet/lisp-gpe | |
parent | fb0815d4ae4bb0fe27bd9313f34b45c8593b907e (diff) |
Initial commit of vpp code.
Change-Id: Ib246f1fbfce93274020ee93ce461e3d8bd8b9f17
Signed-off-by: Ed Warnicke <eaw@cisco.com>
Diffstat (limited to 'vnet/vnet/lisp-gpe')
-rw-r--r-- | vnet/vnet/lisp-gpe/decap.c | 298 | ||||
-rw-r--r-- | vnet/vnet/lisp-gpe/encap.c | 299 | ||||
-rw-r--r-- | vnet/vnet/lisp-gpe/lisp_gpe.c | 498 | ||||
-rw-r--r-- | vnet/vnet/lisp-gpe/lisp_gpe.h | 137 | ||||
-rw-r--r-- | vnet/vnet/lisp-gpe/lisp_gpe_error.def | 16 | ||||
-rw-r--r-- | vnet/vnet/lisp-gpe/lisp_gpe_packet.h | 123 | ||||
-rw-r--r-- | vnet/vnet/lisp-gpe/rfc.txt | 826 |
7 files changed, 2197 insertions, 0 deletions
diff --git a/vnet/vnet/lisp-gpe/decap.c b/vnet/vnet/lisp-gpe/decap.c new file mode 100644 index 00000000..e10f1f2e --- /dev/null +++ b/vnet/vnet/lisp-gpe/decap.c @@ -0,0 +1,298 @@ +/* + * decap.c: lisp-gpe decap processing + * + * Copyright (c) 2014 Cisco and/or its affiliates. + * Licensed under the Apache License, Version 2.0 (the "License"); + * you may not use this file except in compliance with the License. + * You may obtain a copy of the License at: + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + +#include <vlib/vlib.h> +#include <vnet/pg/pg.h> +#include <vnet/lisp-gpe/lisp_gpe.h> + +typedef struct { + u32 next_index; + u32 tunnel_index; + u32 error; + lisp_gpe_header_t h; +} lisp_gpe_rx_trace_t; + +static u8 * format_lisp_gpe_rx_trace (u8 * s, va_list * args) +{ + CLIB_UNUSED (vlib_main_t * vm) = va_arg (*args, vlib_main_t *); + CLIB_UNUSED (vlib_node_t * node) = va_arg (*args, vlib_node_t *); + lisp_gpe_rx_trace_t * t = va_arg (*args, lisp_gpe_rx_trace_t *); + + if (t->tunnel_index != ~0) + { + s = format (s, "NSH-VXLAN: tunnel %d next %d error %d", t->tunnel_index, + t->next_index, t->error); + } + else + { + s = format (s, "NSH-VXLAN: no tunnel next %d error %d\n", t->next_index, + t->error); + } + s = format (s, "\n %U", format_lisp_gpe_header_with_length, &t->h, + (u32) sizeof (t->h) /* max size */); + return s; +} + +static uword +lisp_gpe_input (vlib_main_t * vm, + vlib_node_runtime_t * node, + vlib_frame_t * from_frame) +{ + u32 n_left_from, next_index, * from, * to_next; + lisp_gpe_main_t * ngm = &lisp_gpe_main; + u32 last_tunnel_index = ~0; + lisp_gpe_tunnel_key_t last_key; + u32 pkts_decapsulated = 0; + + memset (&last_key, 0xff, sizeof (last_key)); + + from = vlib_frame_vector_args (from_frame); + n_left_from = from_frame->n_vectors; + + next_index = node->cached_next_index; + + while (n_left_from > 0) + { + u32 n_left_to_next; + + vlib_get_next_frame (vm, node, next_index, + to_next, n_left_to_next); + +#if 0 + while (n_left_from >= 4 && n_left_to_next >= 2) + { + u32 bi0, bi1; + vlib_buffer_t * b0, * b1; + nsh_unicast_header_t * h0, * h1; + u32 label0, label1; + u32 next0, next1; + uword * p0, * p1; + + /* Prefetch next iteration. */ + { + vlib_buffer_t * p2, * p3; + + p2 = vlib_get_buffer (vm, from[2]); + p3 = vlib_get_buffer (vm, from[3]); + + vlib_prefetch_buffer_header (p2, LOAD); + vlib_prefetch_buffer_header (p3, LOAD); + + CLIB_PREFETCH (p2->data, 2*CLIB_CACHE_LINE_BYTES, LOAD); + CLIB_PREFETCH (p3->data, 2*CLIB_CACHE_LINE_BYTES, LOAD); + } + + bi0 = from[0]; + bi1 = from[1]; + to_next[0] = bi0; + to_next[1] = bi1; + from += 2; + to_next += 2; + n_left_to_next -= 2; + n_left_from -= 2; + + b0 = vlib_get_buffer (vm, bi0); + b1 = vlib_get_buffer (vm, bi1); + + h0 = vlib_buffer_get_current (b0); + h1 = vlib_buffer_get_current (b1); + + next0 = next1 = LISP_GPE_INPUT_NEXT_IP4_INPUT; + + label0 = clib_net_to_host_u32 (h0->label_exp_s_ttl); + label1 = clib_net_to_host_u32 (h1->label_exp_s_ttl); + + /* + * Translate label contents into a fib index. + * This is a decent sanity check, and guarantees + * a sane FIB for the downstream lookup + */ + label0 = vnet_nsh_uc_get_label (label0); + label1 = vnet_nsh_uc_get_label (label1); + + /* If 2xlabels match, and match the 1-wide cache, use it */ + if (label0 == label1 && rt->last_label == label0) + { + vnet_buffer(b0)->sw_if_index[VLIB_TX] = rt->last_fib_index; + vnet_buffer(b1)->sw_if_index[VLIB_TX] = rt->last_fib_index; + } + else + { + p0 = hash_get (rt->mm->fib_index_by_nsh_label, label0); + if (PREDICT_FALSE (p0 == 0)) + { + next0 = LISP_GPE_INPUT_NEXT_DROP; + b0->error = node->errors[NSH_ERROR_BAD_LABEL]; + } + else + vnet_buffer(b0)->sw_if_index[VLIB_TX] = p0[0]; + + p1 = hash_get (rt->mm->fib_index_by_nsh_label, label1); + if (PREDICT_FALSE (p1 == 0)) + { + next1 = LISP_GPE_INPUT_NEXT_DROP; + b1->error = node->errors[NSH_ERROR_BAD_LABEL]; + } + else + { + vnet_buffer(b1)->sw_if_index[VLIB_TX] = p1[0]; + rt->last_fib_index = p1[0]; + rt->last_label = label1; + } + } + + if (PREDICT_FALSE(b0->flags & VLIB_BUFFER_IS_TRACED)) + { + nsh_rx_trace_t *tr = vlib_add_trace (vm, node, + b0, sizeof (*tr)); + tr->label_exp_s_ttl = label0; + } + if (PREDICT_FALSE(b1->flags & VLIB_BUFFER_IS_TRACED)) + { + nsh_rx_trace_t *tr = vlib_add_trace (vm, node, + b1, sizeof (*tr)); + tr->label_exp_s_ttl = label1; + } + + vlib_buffer_advance (b0, sizeof (*h0)); + vlib_buffer_advance (b1, sizeof (*h1)); + + vlib_validate_buffer_enqueue_x2 (vm, node, next_index, + to_next, n_left_to_next, + bi0, bi1, next0, next1); + } +#endif + + while (n_left_from > 0 && n_left_to_next > 0) + { + u32 bi0; + vlib_buffer_t * b0; + u32 next0; + ip4_udp_lisp_gpe_header_t * iul0; + uword * p0; + u32 tunnel_index0; + lisp_gpe_tunnel_t * t0; + lisp_gpe_tunnel_key_t key0; + u32 error0; + + bi0 = from[0]; + to_next[0] = bi0; + from += 1; + to_next += 1; + n_left_from -= 1; + n_left_to_next -= 1; + + b0 = vlib_get_buffer (vm, bi0); + + /* udp leaves current_data pointing at the lisp header */ + vlib_buffer_advance + (b0, -(word)(sizeof(udp_header_t)+sizeof(ip4_header_t))); + + iul0 = vlib_buffer_get_current (b0); + + /* pop (ip, udp, lisp-gpe) */ + vlib_buffer_advance (b0, sizeof (*iul0)); + + tunnel_index0 = ~0; + error0 = 0; + next0 = LISP_GPE_INPUT_NEXT_DROP; + + key0.src = iul0->ip4.src_address.as_u32; + key0.iid = iul0->lisp.iid; + + if (PREDICT_FALSE ((key0.as_u64[0] != last_key.as_u64[0]))) + { + p0 = hash_get_mem (ngm->lisp_gpe_tunnel_by_key, &key0); + + if (p0 == 0) + { + error0 = LISP_GPE_ERROR_NO_SUCH_TUNNEL; + goto trace0; + } + + last_key.as_u64[0] = key0.as_u64[0]; + tunnel_index0 = last_tunnel_index = p0[0]; + } + else + tunnel_index0 = last_tunnel_index; + + t0 = pool_elt_at_index (ngm->tunnels, tunnel_index0); + + next0 = t0->decap_next_index; + + /* Required to make the l2 tag push / pop code work on l2 subifs */ + vnet_update_l2_len (b0); + + /* + * ip[46] lookup in the configured FIB + * lisp-gpe-encap, here's the encap tunnel sw_if_index + */ + vnet_buffer(b0)->sw_if_index[VLIB_TX] = t0->decap_fib_index; + pkts_decapsulated ++; + + trace0: + b0->error = error0 ? node->errors[error0] : 0; + + if (PREDICT_FALSE(b0->flags & VLIB_BUFFER_IS_TRACED)) + { + lisp_gpe_rx_trace_t *tr + = vlib_add_trace (vm, node, b0, sizeof (*tr)); + tr->next_index = next0; + tr->error = error0; + tr->tunnel_index = tunnel_index0; + tr->h = iul0->lisp; + } + vlib_validate_buffer_enqueue_x1 (vm, node, next_index, + to_next, n_left_to_next, + bi0, next0); + } + + vlib_put_next_frame (vm, node, next_index, n_left_to_next); + } + vlib_node_increment_counter (vm, lisp_gpe_input_node.index, + LISP_GPE_ERROR_DECAPSULATED, + pkts_decapsulated); + return from_frame->n_vectors; +} + +static char * lisp_gpe_error_strings[] = { +#define lisp_gpe_error(n,s) s, +#include <vnet/lisp-gpe/lisp_gpe_error.def> +#undef lisp_gpe_error +#undef _ +}; + +VLIB_REGISTER_NODE (lisp_gpe_input_node) = { + .function = lisp_gpe_input, + .name = "lisp-gpe-input", + /* Takes a vector of packets. */ + .vector_size = sizeof (u32), + + .n_errors = LISP_GPE_N_ERROR, + .error_strings = lisp_gpe_error_strings, + + .n_next_nodes = LISP_GPE_INPUT_N_NEXT, + .next_nodes = { +#define _(s,n) [LISP_GPE_INPUT_NEXT_##s] = n, + foreach_lisp_gpe_input_next +#undef _ + }, + + .format_buffer = format_lisp_gpe_header_with_length, + .format_trace = format_lisp_gpe_rx_trace, + // $$$$ .unformat_buffer = unformat_lisp_gpe_header, +}; diff --git a/vnet/vnet/lisp-gpe/encap.c b/vnet/vnet/lisp-gpe/encap.c new file mode 100644 index 00000000..b3a52c46 --- /dev/null +++ b/vnet/vnet/lisp-gpe/encap.c @@ -0,0 +1,299 @@ +/* + * Copyright (c) 2015 Cisco and/or its affiliates. + * Licensed under the Apache License, Version 2.0 (the "License"); + * you may not use this file except in compliance with the License. + * You may obtain a copy of the License at: + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ +#include <vppinfra/error.h> +#include <vppinfra/hash.h> +#include <vnet/vnet.h> +#include <vnet/ip/ip.h> +#include <vnet/ethernet/ethernet.h> +#include <vnet/lisp-gpe/lisp_gpe.h> + +/* Statistics (not really errors) */ +#define foreach_lisp_gpe_encap_error \ +_(ENCAPSULATED, "good packets encapsulated") + +static char * lisp_gpe_encap_error_strings[] = { +#define _(sym,string) string, + foreach_lisp_gpe_encap_error +#undef _ +}; + +typedef enum { +#define _(sym,str) LISP_GPE_ENCAP_ERROR_##sym, + foreach_lisp_gpe_encap_error +#undef _ + LISP_GPE_ENCAP_N_ERROR, +} lisp_gpe_encap_error_t; + +typedef enum { + LISP_GPE_ENCAP_NEXT_IP4_LOOKUP, + LISP_GPE_ENCAP_NEXT_DROP, + LISP_GPE_ENCAP_N_NEXT, +} lisp_gpe_encap_next_t; + +typedef struct { + u32 tunnel_index; +} lisp_gpe_encap_trace_t; + +u8 * format_lisp_gpe_encap_trace (u8 * s, va_list * args) +{ + CLIB_UNUSED (vlib_main_t * vm) = va_arg (*args, vlib_main_t *); + CLIB_UNUSED (vlib_node_t * node) = va_arg (*args, vlib_node_t *); + lisp_gpe_encap_trace_t * t + = va_arg (*args, lisp_gpe_encap_trace_t *); + + s = format (s, "LISP-GPE-ENCAP: tunnel %d", t->tunnel_index); + return s; +} + +#define foreach_fixed_header_offset \ +_(0) _(1) _(2) _(3) + +static uword +lisp_gpe_encap (vlib_main_t * vm, + vlib_node_runtime_t * node, + vlib_frame_t * from_frame) +{ + u32 n_left_from, next_index, * from, * to_next; + lisp_gpe_main_t * ngm = &lisp_gpe_main; + vnet_main_t * vnm = ngm->vnet_main; + u32 pkts_encapsulated = 0; + u16 old_l0 = 0; + + from = vlib_frame_vector_args (from_frame); + n_left_from = from_frame->n_vectors; + + next_index = node->cached_next_index; + + while (n_left_from > 0) + { + u32 n_left_to_next; + + vlib_get_next_frame (vm, node, next_index, + to_next, n_left_to_next); + +#if 0 + while (n_left_from >= 4 && n_left_to_next >= 2) + { + u32 bi0, bi1; + vlib_buffer_t * b0, * b1; + nsh_unicast_header_t * h0, * h1; + u32 label0, label1; + u32 next0, next1; + uword * p0, * p1; + + /* Prefetch next iteration. */ + { + vlib_buffer_t * p2, * p3; + + p2 = vlib_get_buffer (vm, from[2]); + p3 = vlib_get_buffer (vm, from[3]); + + vlib_prefetch_buffer_header (p2, LOAD); + vlib_prefetch_buffer_header (p3, LOAD); + + CLIB_PREFETCH (p2->data, 2*CLIB_CACHE_LINE_BYTES, LOAD); + CLIB_PREFETCH (p3->data, 2*CLIB_CACHE_LINE_BYTES, LOAD); + } + + bi0 = from[0]; + bi1 = from[1]; + to_next[0] = bi0; + to_next[1] = bi1; + from += 2; + to_next += 2; + n_left_to_next -= 2; + n_left_from -= 2; + + b0 = vlib_get_buffer (vm, bi0); + b1 = vlib_get_buffer (vm, bi1); + + h0 = vlib_buffer_get_current (b0); + h1 = vlib_buffer_get_current (b1); + + next0 = next1 = NSH_INPUT_NEXT_IP4_INPUT; + + label0 = clib_net_to_host_u32 (h0->label_exp_s_ttl); + label1 = clib_net_to_host_u32 (h1->label_exp_s_ttl); + + /* + * Translate label contents into a fib index. + * This is a decent sanity check, and guarantees + * a sane FIB for the downstream lookup + */ + label0 = vnet_nsh_uc_get_label (label0); + label1 = vnet_nsh_uc_get_label (label1); + + /* If 2xlabels match, and match the 1-wide cache, use it */ + if (label0 == label1 && rt->last_label == label0) + { + vnet_buffer(b0)->sw_if_index[VLIB_TX] = rt->last_fib_index; + vnet_buffer(b1)->sw_if_index[VLIB_TX] = rt->last_fib_index; + } + else + { + p0 = hash_get (rt->mm->fib_index_by_nsh_label, label0); + if (PREDICT_FALSE (p0 == 0)) + { + next0 = NSH_INPUT_NEXT_DROP; + b0->error = node->errors[NSH_ERROR_BAD_LABEL]; + } + else + vnet_buffer(b0)->sw_if_index[VLIB_TX] = p0[0]; + + p1 = hash_get (rt->mm->fib_index_by_nsh_label, label1); + if (PREDICT_FALSE (p1 == 0)) + { + next1 = NSH_INPUT_NEXT_DROP; + b1->error = node->errors[NSH_ERROR_BAD_LABEL]; + } + else + { + vnet_buffer(b1)->sw_if_index[VLIB_TX] = p1[0]; + rt->last_fib_index = p1[0]; + rt->last_label = label1; + } + } + + if (PREDICT_FALSE(b0->flags & VLIB_BUFFER_IS_TRACED)) + { + nsh_rx_trace_t *tr = vlib_add_trace (vm, node, + b0, sizeof (*tr)); + tr->label_exp_s_ttl = label0; + } + if (PREDICT_FALSE(b1->flags & VLIB_BUFFER_IS_TRACED)) + { + nsh_rx_trace_t *tr = vlib_add_trace (vm, node, + b1, sizeof (*tr)); + tr->label_exp_s_ttl = label1; + } + + vlib_buffer_advance (b0, sizeof (*h0)); + vlib_buffer_advance (b1, sizeof (*h1)); + + vlib_validate_buffer_enqueue_x2 (vm, node, next_index, + to_next, n_left_to_next, + bi0, bi1, next0, next1); + } +#endif + + while (n_left_from > 0 && n_left_to_next > 0) + { + u32 bi0; + vlib_buffer_t * b0; + u32 next0 = LISP_GPE_ENCAP_NEXT_IP4_LOOKUP; + vnet_hw_interface_t * hi0; + ip4_header_t * ip0; + udp_header_t * udp0; + u64 * copy_src0, * copy_dst0; + u32 * copy_src_last0, * copy_dst_last0; + lisp_gpe_tunnel_t * t0; + u16 new_l0; + ip_csum_t sum0; + + bi0 = from[0]; + to_next[0] = bi0; + from += 1; + to_next += 1; + n_left_from -= 1; + n_left_to_next -= 1; + + b0 = vlib_get_buffer (vm, bi0); + + /* 1-wide cache? */ + hi0 = vnet_get_sup_hw_interface + (vnm, vnet_buffer(b0)->sw_if_index[VLIB_TX]); + + t0 = pool_elt_at_index (ngm->tunnels, hi0->dev_instance); + + ASSERT(vec_len(t0->rewrite) >= 24); + + /* Apply the rewrite string. $$$$ vnet_rewrite? */ + vlib_buffer_advance (b0, -(word)_vec_len(t0->rewrite)); + + ip0 = vlib_buffer_get_current(b0); + /* Copy the fixed header */ + copy_dst0 = (u64 *) ip0; + copy_src0 = (u64 *) t0->rewrite; + + ASSERT (sizeof (ip4_udp_lisp_gpe_header_t) == 36); + + /* Copy first 32 octets 8-bytes at a time */ +#define _(offs) copy_dst0[offs] = copy_src0[offs]; + foreach_fixed_header_offset; +#undef _ + /* Last 4 octets. Hopefully gcc will be our friend */ + copy_dst_last0 = (u32 *)(©_dst0[4]); + copy_src_last0 = (u32 *)(©_src0[4]); + + copy_dst_last0[0] = copy_src_last0[0]; + + /* fix the <bleep>ing outer-IP checksum */ + sum0 = ip0->checksum; + /* old_l0 always 0, see the rewrite setup */ + new_l0 = + clib_host_to_net_u16 (vlib_buffer_length_in_chain (vm, b0)); + + sum0 = ip_csum_update (sum0, old_l0, new_l0, ip4_header_t, + length /* changed member */); + ip0->checksum = ip_csum_fold (sum0); + ip0->length = new_l0; + + /* Fix UDP length */ + udp0 = (udp_header_t *)(ip0+1); + new_l0 = clib_host_to_net_u16 (vlib_buffer_length_in_chain (vm, b0) + - sizeof (*ip0)); + + udp0->length = new_l0; + + /* Reset to look up tunnel partner in the configured FIB */ + vnet_buffer(b0)->sw_if_index[VLIB_TX] = t0->encap_fib_index; + pkts_encapsulated ++; + + if (PREDICT_FALSE(b0->flags & VLIB_BUFFER_IS_TRACED)) + { + lisp_gpe_encap_trace_t *tr = + vlib_add_trace (vm, node, b0, sizeof (*tr)); + tr->tunnel_index = t0 - ngm->tunnels; + } + vlib_validate_buffer_enqueue_x1 (vm, node, next_index, + to_next, n_left_to_next, + bi0, next0); + } + + vlib_put_next_frame (vm, node, next_index, n_left_to_next); + } + vlib_node_increment_counter (vm, node->node_index, + LISP_GPE_ENCAP_ERROR_ENCAPSULATED, + pkts_encapsulated); + return from_frame->n_vectors; +} + +VLIB_REGISTER_NODE (lisp_gpe_encap_node) = { + .function = lisp_gpe_encap, + .name = "lisp-gpe-encap", + .vector_size = sizeof (u32), + .format_trace = format_lisp_gpe_encap_trace, + .type = VLIB_NODE_TYPE_INTERNAL, + + .n_errors = ARRAY_LEN(lisp_gpe_encap_error_strings), + .error_strings = lisp_gpe_encap_error_strings, + + .n_next_nodes = LISP_GPE_ENCAP_N_NEXT, + + .next_nodes = { + [LISP_GPE_ENCAP_NEXT_IP4_LOOKUP] = "ip4-lookup", + [LISP_GPE_ENCAP_NEXT_DROP] = "error-drop", + }, +}; diff --git a/vnet/vnet/lisp-gpe/lisp_gpe.c b/vnet/vnet/lisp-gpe/lisp_gpe.c new file mode 100644 index 00000000..eb4ca919 --- /dev/null +++ b/vnet/vnet/lisp-gpe/lisp_gpe.c @@ -0,0 +1,498 @@ +/* + * Copyright (c) 2015 Cisco and/or its affiliates. + * Licensed under the Apache License, Version 2.0 (the "License"); + * you may not use this file except in compliance with the License. + * You may obtain a copy of the License at: + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ +#include <vnet/lisp-gpe/lisp_gpe.h> + +lisp_gpe_main_t lisp_gpe_main; + +static u8 * format_decap_next (u8 * s, va_list * args) +{ + u32 next_index = va_arg (*args, u32); + + switch (next_index) + { + case LISP_GPE_INPUT_NEXT_DROP: + return format (s, "drop"); + case LISP_GPE_INPUT_NEXT_IP4_INPUT: + return format (s, "ip4"); + case LISP_GPE_INPUT_NEXT_IP6_INPUT: + return format (s, "ip6"); + case LISP_GPE_INPUT_NEXT_LISP_GPE_ENCAP: + return format (s, "nsh-lisp-gpe"); + default: + return format (s, "unknown %d", next_index); + } + return s; +} + +u8 * format_lisp_gpe_tunnel (u8 * s, va_list * args) +{ + lisp_gpe_tunnel_t * t = va_arg (*args, lisp_gpe_tunnel_t *); + lisp_gpe_main_t * ngm = &lisp_gpe_main; + + s = format (s, + "[%d] %U (src) %U (dst) fibs: encap %d, decap %d", + t - ngm->tunnels, + format_ip4_address, &t->src, + format_ip4_address, &t->dst, + t->encap_fib_index, + t->decap_fib_index); + + s = format (s, " decap next %U\n", format_decap_next, t->decap_next_index); + s = format (s, "lisp ver %d ", (t->ver_res>>6)); + +#define _(n,v) if (t->flags & v) s = format (s, "%s-bit ", #n); + foreach_lisp_gpe_flag_bit; +#undef _ + + s = format (s, "next_protocol %d ver_res %x res %x\n", + t->next_protocol, t->ver_res, t->res); + + s = format (s, "iid %d (0x%x)\n", t->iid, t->iid); + return s; +} + +static u8 * format_lisp_gpe_name (u8 * s, va_list * args) +{ + u32 dev_instance = va_arg (*args, u32); + return format (s, "lisp_gpe_tunnel%d", dev_instance); +} + +static uword dummy_interface_tx (vlib_main_t * vm, + vlib_node_runtime_t * node, + vlib_frame_t * frame) +{ + clib_warning ("you shouldn't be here, leaking buffers..."); + return frame->n_vectors; +} + +VNET_DEVICE_CLASS (lisp_gpe_device_class,static) = { + .name = "LISP_GPE", + .format_device_name = format_lisp_gpe_name, + .format_tx_trace = format_lisp_gpe_encap_trace, + .tx_function = dummy_interface_tx, +}; + +static uword dummy_set_rewrite (vnet_main_t * vnm, + u32 sw_if_index, + u32 l3_type, + void * dst_address, + void * rewrite, + uword max_rewrite_bytes) +{ + return 0; +} + +u8 * format_lisp_gpe_header_with_length (u8 * s, va_list * args) +{ + lisp_gpe_header_t * h = va_arg (*args, lisp_gpe_header_t *); + u32 max_header_bytes = va_arg (*args, u32); + u32 header_bytes; + + header_bytes = sizeof (h[0]); + if (max_header_bytes != 0 && header_bytes > max_header_bytes) + return format (s, "gre-nsh header truncated"); + + s = format (s, "flags: "); +#define _(n,v) if (h->flags & v) s = format (s, "%s ", #n); + foreach_lisp_gpe_flag_bit; +#undef _ + + s = format (s, "\n ver_res %d res %d next_protocol %d iid %d(%x)", + h->ver_res, h->res, h->next_protocol, + clib_net_to_host_u32 (h->iid), + clib_net_to_host_u32 (h->iid)); + return s; +} + +VNET_HW_INTERFACE_CLASS (lisp_gpe_hw_class) = { + .name = "LISP_GPE", + .format_header = format_lisp_gpe_header_with_length, + .set_rewrite = dummy_set_rewrite, +}; + +#define foreach_copy_field \ +_(src.as_u32) \ +_(dst.as_u32) \ +_(encap_fib_index) \ +_(decap_fib_index) \ +_(decap_next_index) \ +_(flags) \ +_(next_protocol) \ +_(ver_res) \ +_(res) \ +_(iid) + +static int lisp_gpe_rewrite (lisp_gpe_tunnel_t * t) +{ + u8 *rw = 0; + ip4_header_t * ip0; + lisp_gpe_header_t * lisp0; + ip4_udp_lisp_gpe_header_t * h0; + int len; + + len = sizeof (*h0); + + vec_validate_aligned (rw, len-1, CLIB_CACHE_LINE_BYTES); + + h0 = (ip4_udp_lisp_gpe_header_t *) rw; + + /* Fixed portion of the (outer) ip4 header */ + ip0 = &h0->ip4; + ip0->ip_version_and_header_length = 0x45; + ip0->ttl = 254; + ip0->protocol = IP_PROTOCOL_UDP; + + /* we fix up the ip4 header length and checksum after-the-fact */ + ip0->src_address.as_u32 = t->src.as_u32; + ip0->dst_address.as_u32 = t->dst.as_u32; + ip0->checksum = ip4_header_checksum (ip0); + + /* UDP header, randomize src port on something, maybe? */ + h0->udp.src_port = clib_host_to_net_u16 (4341); + h0->udp.dst_port = clib_host_to_net_u16 (UDP_DST_PORT_lisp_gpe); + + /* LISP-gpe header */ + lisp0 = &h0->lisp; + + lisp0->flags = t->flags; + lisp0->ver_res = t->ver_res; + lisp0->res = t->res; + lisp0->next_protocol = t->next_protocol; + lisp0->iid = clib_host_to_net_u32 (t->iid); + + t->rewrite = rw; + return (0); +} + +int vnet_lisp_gpe_add_del_tunnel +(vnet_lisp_gpe_add_del_tunnel_args_t *a, u32 * sw_if_indexp) +{ + lisp_gpe_main_t * ngm = &lisp_gpe_main; + lisp_gpe_tunnel_t *t = 0; + vnet_main_t * vnm = ngm->vnet_main; + vnet_hw_interface_t * hi; + uword * p; + u32 hw_if_index = ~0; + u32 sw_if_index = ~0; + int rv; + lisp_gpe_tunnel_key_t key, *key_copy; + hash_pair_t *hp; + + key.src = a->src.as_u32; + key.iid = clib_host_to_net_u32(a->iid); + + p = hash_get_mem (ngm->lisp_gpe_tunnel_by_key, &key); + + if (a->is_add) + { + /* adding a tunnel: tunnel must not already exist */ + if (p) + return VNET_API_ERROR_INVALID_VALUE; + + if (a->decap_next_index >= LISP_GPE_INPUT_N_NEXT) + return VNET_API_ERROR_INVALID_DECAP_NEXT; + + pool_get_aligned (ngm->tunnels, t, CLIB_CACHE_LINE_BYTES); + memset (t, 0, sizeof (*t)); + + /* copy from arg structure */ +#define _(x) t->x = a->x; + foreach_copy_field; +#undef _ + + rv = lisp_gpe_rewrite (t); + + if (rv) + { + pool_put (ngm->tunnels, t); + return rv; + } + + key_copy = clib_mem_alloc (sizeof (*key_copy)); + memcpy (key_copy, &key, sizeof (*key_copy)); + + hash_set_mem (ngm->lisp_gpe_tunnel_by_key, key_copy, + t - ngm->tunnels); + + if (vec_len (ngm->free_lisp_gpe_tunnel_hw_if_indices) > 0) + { + hw_if_index = ngm->free_lisp_gpe_tunnel_hw_if_indices + [vec_len (ngm->free_lisp_gpe_tunnel_hw_if_indices)-1]; + _vec_len (ngm->free_lisp_gpe_tunnel_hw_if_indices) -= 1; + + hi = vnet_get_hw_interface (vnm, hw_if_index); + hi->dev_instance = t - ngm->tunnels; + hi->hw_instance = hi->dev_instance; + } + else + { + hw_if_index = vnet_register_interface + (vnm, lisp_gpe_device_class.index, t - ngm->tunnels, + lisp_gpe_hw_class.index, t - ngm->tunnels); + hi = vnet_get_hw_interface (vnm, hw_if_index); + hi->output_node_index = lisp_gpe_encap_node.index; + } + + t->hw_if_index = hw_if_index; + t->sw_if_index = sw_if_index = hi->sw_if_index; + + vnet_sw_interface_set_flags (vnm, hi->sw_if_index, + VNET_SW_INTERFACE_FLAG_ADMIN_UP); + } + else + { + /* deleting a tunnel: tunnel must exist */ + if (!p) + return VNET_API_ERROR_NO_SUCH_ENTRY; + + t = pool_elt_at_index (ngm->tunnels, p[0]); + + vnet_sw_interface_set_flags (vnm, t->sw_if_index, 0 /* down */); + vec_add1 (ngm->free_lisp_gpe_tunnel_hw_if_indices, t->hw_if_index); + + hp = hash_get_pair (ngm->lisp_gpe_tunnel_by_key, &key); + key_copy = (void *)(hp->key); + hash_unset_mem (ngm->lisp_gpe_tunnel_by_key, &key); + clib_mem_free (key_copy); + + vec_free (t->rewrite); + pool_put (ngm->tunnels, t); + } + + if (sw_if_indexp) + *sw_if_indexp = sw_if_index; + + return 0; +} + +static u32 fib_index_from_fib_id (u32 fib_id) +{ + ip4_main_t * im = &ip4_main; + uword * p; + + p = hash_get (im->fib_index_by_table_id, fib_id); + if (!p) + return ~0; + + return p[0]; +} + +static uword unformat_decap_next (unformat_input_t * input, va_list * args) +{ + u32 * result = va_arg (*args, u32 *); + u32 tmp; + + if (unformat (input, "drop")) + *result = LISP_GPE_INPUT_NEXT_DROP; + else if (unformat (input, "ip4")) + *result = LISP_GPE_INPUT_NEXT_IP4_INPUT; + else if (unformat (input, "ip6")) + *result = LISP_GPE_INPUT_NEXT_IP6_INPUT; + else if (unformat (input, "ethernet")) + *result = LISP_GPE_INPUT_NEXT_IP6_INPUT; + else if (unformat (input, "lisp-gpe")) + *result = LISP_GPE_INPUT_NEXT_LISP_GPE_ENCAP; + else if (unformat (input, "%d", &tmp)) + *result = tmp; + else + return 0; + return 1; +} + +static clib_error_t * +lisp_gpe_add_del_tunnel_command_fn (vlib_main_t * vm, + unformat_input_t * input, + vlib_cli_command_t * cmd) +{ + unformat_input_t _line_input, * line_input = &_line_input; + ip4_address_t src, dst; + u8 is_add = 1; + u8 src_set = 0; + u8 dst_set = 0; + u32 encap_fib_index = 0; + u32 decap_fib_index = 0; + u8 next_protocol = LISP_GPE_NEXT_PROTOCOL_IP4; + u32 decap_next_index = LISP_GPE_INPUT_NEXT_IP4_INPUT; + u8 flags = LISP_GPE_FLAGS_P; + u8 ver_res = 0; + u8 res = 0; + u32 iid = 0; + u8 iid_set = 0; + u32 tmp; + int rv; + vnet_lisp_gpe_add_del_tunnel_args_t _a, * a = &_a; + + /* Get a line of input. */ + if (! unformat_user (input, unformat_line_input, line_input)) + return 0; + + while (unformat_check_input (line_input) != UNFORMAT_END_OF_INPUT) { + if (unformat (line_input, "del")) + is_add = 0; + else if (unformat (line_input, "src %U", + unformat_ip4_address, &src)) + src_set = 1; + else if (unformat (line_input, "dst %U", + unformat_ip4_address, &dst)) + dst_set = 1; + else if (unformat (line_input, "encap-vrf-id %d", &tmp)) + { + encap_fib_index = fib_index_from_fib_id (tmp); + if (encap_fib_index == ~0) + return clib_error_return (0, "nonexistent encap fib id %d", tmp); + } + else if (unformat (line_input, "decap-vrf-id %d", &tmp)) + { + decap_fib_index = fib_index_from_fib_id (tmp); + if (decap_fib_index == ~0) + return clib_error_return (0, "nonexistent decap fib id %d", tmp); + } + else if (unformat (line_input, "decap-next %U", unformat_decap_next, + &decap_next_index)) + ; + else if (unformat(line_input, "next-ip4")) + next_protocol = 1; + else if (unformat(line_input, "next-ip6")) + next_protocol = 2; + else if (unformat(line_input, "next-ethernet")) + next_protocol = 3; + else if (unformat(line_input, "next-nsh")) + next_protocol = 4; + /* Allow the user to specify anything they want in the LISP hdr */ + else if (unformat (line_input, "ver_res %x", &tmp)) + ver_res = tmp; + else if (unformat (line_input, "res %x", &tmp)) + res = tmp; + else if (unformat (line_input, "flags %x", &tmp)) + flags = tmp; + else if (unformat (line_input, "n-bit")) + flags |= LISP_GPE_FLAGS_N; + else if (unformat (line_input, "l-bit")) + flags |= LISP_GPE_FLAGS_L; + else if (unformat (line_input, "e-bit")) + flags |= LISP_GPE_FLAGS_E; + else if (unformat (line_input, "v-bit")) + flags |= LISP_GPE_FLAGS_V; + else if (unformat (line_input, "i-bit")) + flags |= LISP_GPE_FLAGS_V; + else if (unformat (line_input, "not-p-bit")) + flags &= ~LISP_GPE_FLAGS_P; + else if (unformat (line_input, "p-bit")) + flags |= LISP_GPE_FLAGS_P; + else if (unformat (line_input, "o-bit")) + flags |= LISP_GPE_FLAGS_O; + else if (unformat (line_input, "iidx %x", &iid)) + iid_set = 1; + else if (unformat (line_input, "iid %d", &iid)) + iid_set = 1; + else + return clib_error_return (0, "parse error: '%U'", + format_unformat_error, line_input); + } + + unformat_free (line_input); + + if (src_set == 0) + return clib_error_return (0, "tunnel src address not specified"); + + if (dst_set == 0) + return clib_error_return (0, "tunnel dst address not specified"); + + if (iid_set == 0) + return clib_error_return (0, "iid not specified"); + + memset (a, 0, sizeof (*a)); + + a->is_add = is_add; + +#define _(x) a->x = x; + foreach_copy_field; +#undef _ + + rv = vnet_lisp_gpe_add_del_tunnel (a, 0 /* hw_if_indexp */); + + switch(rv) + { + case 0: + break; + case VNET_API_ERROR_INVALID_DECAP_NEXT: + return clib_error_return (0, "invalid decap-next..."); + + case VNET_API_ERROR_TUNNEL_EXIST: + return clib_error_return (0, "tunnel already exists..."); + + case VNET_API_ERROR_NO_SUCH_ENTRY: + return clib_error_return (0, "tunnel does not exist..."); + + default: + return clib_error_return + (0, "vnet_lisp_gpe_add_del_tunnel returned %d", rv); + } + + return 0; +} + +VLIB_CLI_COMMAND (create_lisp_gpe_tunnel_command, static) = { + .path = "lisp gpe tunnel", + .short_help = + "lisp gpe tunnel src <ip4-addr> dst <ip4-addr> iidx <0xnn> | iid <nn>\n" + " [encap-fib-id <nn>] [decap-fib-id <nn>]\n" + " [n-bit][l-bit][e-bit][v-bit][i-bit][p-bit][not-p-bit][o-bit]\n" + " [next-ip4][next-ip6][next-ethernet][next-nsh]\n" + " [decap-next [ip4|ip6|ethernet|nsh-encap|<nn>]][del]\n", + .function = lisp_gpe_add_del_tunnel_command_fn, +}; + +static clib_error_t * +show_lisp_gpe_tunnel_command_fn (vlib_main_t * vm, + unformat_input_t * input, + vlib_cli_command_t * cmd) +{ + lisp_gpe_main_t * ngm = &lisp_gpe_main; + lisp_gpe_tunnel_t * t; + + if (pool_elts (ngm->tunnels) == 0) + vlib_cli_output (vm, "No lisp-gpe tunnels configured..."); + + pool_foreach (t, ngm->tunnels, + ({ + vlib_cli_output (vm, "%U", format_lisp_gpe_tunnel, t); + })); + + return 0; +} + +VLIB_CLI_COMMAND (show_lisp_gpe_tunnel_command, static) = { + .path = "show lisp gpe tunnel", + .function = show_lisp_gpe_tunnel_command_fn, +}; + +clib_error_t *lisp_gpe_init (vlib_main_t *vm) +{ + lisp_gpe_main_t *ngm = &lisp_gpe_main; + + ngm->vnet_main = vnet_get_main(); + ngm->vlib_main = vm; + + ngm->lisp_gpe_tunnel_by_key + = hash_create_mem (0, sizeof(lisp_gpe_tunnel_key_t), sizeof (uword)); + + udp_register_dst_port (vm, UDP_DST_PORT_lisp_gpe, + lisp_gpe_input_node.index, 1 /* is_ip4 */); + return 0; +} + +VLIB_INIT_FUNCTION(lisp_gpe_init); + diff --git a/vnet/vnet/lisp-gpe/lisp_gpe.h b/vnet/vnet/lisp-gpe/lisp_gpe.h new file mode 100644 index 00000000..8ca721dd --- /dev/null +++ b/vnet/vnet/lisp-gpe/lisp_gpe.h @@ -0,0 +1,137 @@ +/* + * Copyright (c) 2015 Cisco and/or its affiliates. + * Licensed under the Apache License, Version 2.0 (the "License"); + * you may not use this file except in compliance with the License. + * You may obtain a copy of the License at: + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ +#ifndef included_vnet_lisp_gpe_h +#define included_vnet_lisp_gpe_h + +#include <vppinfra/error.h> +#include <vppinfra/hash.h> +#include <vnet/vnet.h> +#include <vnet/ip/ip.h> +#include <vnet/l2/l2_input.h> +#include <vnet/ethernet/ethernet.h> +#include <vnet/lisp-gpe/lisp_gpe_packet.h> +#include <vnet/ip/ip4_packet.h> +#include <vnet/ip/udp.h> + +typedef CLIB_PACKED (struct { + ip4_header_t ip4; /* 20 bytes */ + udp_header_t udp; /* 8 bytes */ + lisp_gpe_header_t lisp; /* 8 bytes */ +}) ip4_udp_lisp_gpe_header_t; + +typedef CLIB_PACKED(struct { + /* + * Key fields: ip src, LISP iid, ??? $$$$$$$$$ correct answer ??? + * all fields in NET byte order + */ + union { + struct { + u32 src; + u32 iid; + }; + u64 as_u64[1]; + }; +}) lisp_gpe_tunnel_key_t; + +typedef struct { + /* Rewrite string. $$$$ embed vnet_rewrite header */ + u8 * rewrite; + + /* decap next index */ + u32 decap_next_index; + + /* tunnel src and dst addresses */ + ip4_address_t src; + ip4_address_t dst; + + /* FIB indices */ + u32 encap_fib_index; /* tunnel partner lookup here */ + u32 decap_fib_index; /* inner IP lookup here */ + + /* vnet intfc hw/sw_if_index */ + u32 hw_if_index; + u32 sw_if_index; + + /* LISP header fields in HOST byte order */ + u8 flags; + u8 ver_res; + u8 res; + u8 next_protocol; + u32 iid; +} lisp_gpe_tunnel_t; + +#define foreach_lisp_gpe_input_next \ +_(DROP, "error-drop") \ +_(IP4_INPUT, "ip4-input") \ +_(IP6_INPUT, "ip6-input") \ +_(ETHERNET_INPUT, "ethernet-input") \ +_(LISP_GPE_ENCAP, "lisp-gpe-encap") + +typedef enum { +#define _(s,n) LISP_GPE_INPUT_NEXT_##s, + foreach_lisp_gpe_input_next +#undef _ + LISP_GPE_INPUT_N_NEXT, +} lisp_gpe_input_next_t; + +typedef enum { +#define lisp_gpe_error(n,s) LISP_GPE_ERROR_##n, +#include <vnet/lisp-gpe/lisp_gpe_error.def> +#undef lisp_gpe_error + LISP_GPE_N_ERROR, +} lisp_gpe_input_error_t; + +typedef struct { + /* vector of encap tunnel instances */ + lisp_gpe_tunnel_t *tunnels; + + /* lookup tunnel by key */ + uword * lisp_gpe_tunnel_by_key; + + /* Free vlib hw_if_indices */ + u32 * free_lisp_gpe_tunnel_hw_if_indices; + + /* convenience */ + vlib_main_t * vlib_main; + vnet_main_t * vnet_main; +} lisp_gpe_main_t; + +lisp_gpe_main_t lisp_gpe_main; + +vlib_node_registration_t lisp_gpe_input_node; +vlib_node_registration_t lisp_gpe_encap_node; + +u8 * format_lisp_gpe_encap_trace (u8 * s, va_list * args); +u8 * format_lisp_gpe_header_with_length (u8 * s, va_list * args); + +typedef struct { + u8 is_add; + ip4_address_t src, dst; + u32 encap_fib_index; + u32 decap_fib_index; + u32 decap_next_index; + u8 flags; + u8 ver_res; + u8 res; + u8 next_protocol; + u32 iid; /* host byte order */ +} vnet_lisp_gpe_add_del_tunnel_args_t; + +int vnet_lisp_gpe_add_del_tunnel +(vnet_lisp_gpe_add_del_tunnel_args_t *a, u32 * sw_if_indexp); + +u8 * format_lisp_gpe_header_with_length (u8 * s, va_list * args); + +#endif /* included_vnet_lisp_gpe_h */ diff --git a/vnet/vnet/lisp-gpe/lisp_gpe_error.def b/vnet/vnet/lisp-gpe/lisp_gpe_error.def new file mode 100644 index 00000000..6ef894f4 --- /dev/null +++ b/vnet/vnet/lisp-gpe/lisp_gpe_error.def @@ -0,0 +1,16 @@ +/* + * Copyright (c) 2015 Cisco and/or its affiliates. + * Licensed under the Apache License, Version 2.0 (the "License"); + * you may not use this file except in compliance with the License. + * You may obtain a copy of the License at: + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ +lisp_gpe_error (DECAPSULATED, "good packets decapsulated") +lisp_gpe_error (NO_SUCH_TUNNEL, "no such tunnel packets") diff --git a/vnet/vnet/lisp-gpe/lisp_gpe_packet.h b/vnet/vnet/lisp-gpe/lisp_gpe_packet.h new file mode 100644 index 00000000..b3d96ed9 --- /dev/null +++ b/vnet/vnet/lisp-gpe/lisp_gpe_packet.h @@ -0,0 +1,123 @@ +/* + * Copyright (c) 2015 Cisco and/or its affiliates. + * Licensed under the Apache License, Version 2.0 (the "License"); + * you may not use this file except in compliance with the License. + * You may obtain a copy of the License at: + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ +#ifndef included_lisp_gpe_packet_h +#define included_lisp_gpe_packet_h + +/* + * From draft-lewis-lisp-gpe-02.txt + * + * 0 1 2 3 + * 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 + * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + * |N|L|E|V|I|P|R|O|Ver| Reserved | Next Protocol | + * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + * | Instance ID/Locator-Status-Bits | + * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + * + * N: The N-bit is the nonce-present bit. When this bit is set to 1, + * the low-order 24 bits of the first 32 bits of the LISP header + * contain a Nonce. See Section 6.3.1 for details. Both N- and + * V-bits MUST NOT be set in the same packet. If they are, a + * decapsulating ETR MUST treat the 'Nonce/Map-Version' field as + * having a Nonce value present. + * + * L: The L-bit is the 'Locator-Status-Bits' field enabled bit. When + * this bit is set to 1, the Locator-Status-Bits in the second + * 32 bits of the LISP header are in use. + * + * E: The E-bit is the echo-nonce-request bit. This bit MUST be ignored + * and has no meaning when the N-bit is set to 0. When the N-bit is + * set to 1 and this bit is set to 1, an ITR is requesting that the + * nonce value in the 'Nonce' field be echoed back in LISP- + * encapsulated packets when the ITR is also an ETR. See + * Section 6.3.1 for details. + * + * V: The V-bit is the Map-Version present bit. When this bit is set to + * 1, the N-bit MUST be 0. Refer to Section 6.6.3 for more details. + * + * I: The I-bit is the Instance ID bit. See Section 5.5 for more + * details. When this bit is set to 1, the 'Locator-Status-Bits' + * field is reduced to 8 bits and the high-order 24 bits are used as + * an Instance ID. If the L-bit is set to 0, then the low-order + * 8 bits are transmitted as zero and ignored on receipt. + * + * P Bit: Flag bit 5 is defined as the Next Protocol bit. The P bit + * MUST be set to 1 to indicate the presence of the 8 bit next + * protocol field. + * + * P = 0 indicates that the payload MUST conform to LISP as defined + * in [RFC6830]. + * + * Flag bit 5 was chosen as the P bit because this flag bit is + * currently unallocated in LISP [RFC6830]. + * + * O: Flag bit 7 is defined as the O bit. When the O bit is set to 1, the + * packet is an OAM packet and OAM processing MUST occur. The OAM + * protocol details are out of scope for this document. As with the + * P-bit, bit 7 is currently a reserved flag in [RFC6830]. + * + * Next Protocol Field: The lower 8 bits of the first word are used to + * carry a next protocol. This next protocol field contains the + * protocol of the encapsulated payload packet. + * + * LISP [RFC6830] uses the lower 16 bits of the first word for either + * a nonce, an echo-nonce ([RFC6830]) or to support map-versioning + * ([RFC6834]). These are all optional capabilities that are + * indicated by setting the N, E, and the V bit respectively. + * + * To maintain the desired data plane compatibility, when the P bit + * is set, the N, E, and V bits MUST be set to zero. + * + * A new protocol registry will be requested from IANA for the Next + * Protocol field. This draft defines the following Next Protocol + * values: + * + * 0x1 : IPv4 + * 0x2 : IPv6 + * 0x3 : Ethernet + * 0x4: Network Service Header + */ + +typedef struct { + u8 flags; + u8 ver_res; + u8 res; + u8 next_protocol; + u32 iid; +} lisp_gpe_header_t; + +#define foreach_lisp_gpe_flag_bit \ +_(N, 0x80) \ +_(L, 0x40) \ +_(E, 0x20) \ +_(V, 0x10) \ +_(I, 0x08) \ +_(P, 0x04) \ +_(O, 0x01) + +typedef enum { +#define _(n,v) LISP_GPE_FLAGS_##n = v, +foreach_lisp_gpe_flag_bit +#undef _ +} vnet_lisp_gpe_flag_bit_t; + +#define LISP_GPE_VERSION 0x0 + +#define LISP_GPE_NEXT_PROTOCOL_IP4 0x1 +#define LISP_GPE_NEXT_PROTOCOL_IP6 0x2 +#define LISP_GPE_NEXT_PROTOCOL_ETHERNET 0x3 +#define LISP_GPE_NEXT_PROTOCOL_NSH 0x4 + +#endif /* included_lisp_gpe_packet_h */ diff --git a/vnet/vnet/lisp-gpe/rfc.txt b/vnet/vnet/lisp-gpe/rfc.txt new file mode 100644 index 00000000..5e3da150 --- /dev/null +++ b/vnet/vnet/lisp-gpe/rfc.txt @@ -0,0 +1,826 @@ +Network Working Group D. Lewis +Internet-Draft Cisco Systems, Inc. +Intended status: Informational P. Agarwal +Expires: January 5, 2015 Broadcom + L. Kreeger + F. Maino + P. Quinn + M. Smith + N. Yadav + Cisco Systems, Inc. + July 4, 2014 + + + LISP Generic Protocol Extension + draft-lewis-lisp-gpe-02.txt + +Abstract + + This draft describes extending the Locator/ID Separation Protocol + (LISP) [RFC6830], via changes to the LISP header, with three new + capabilities: support for multi-protocol encapsulation, operations, + administration and management (OAM) signaling, and explicit + versioning. + +Status of this Memo + + This Internet-Draft is submitted in full conformance with the + provisions of BCP 78 and BCP 79. + + Internet-Drafts are working documents of the Internet Engineering + Task Force (IETF). Note that other groups may also distribute + working documents as Internet-Drafts. The list of current Internet- + Drafts is at http://datatracker.ietf.org/drafts/current/. + + Internet-Drafts are draft documents valid for a maximum of six months + and may be updated, replaced, or obsoleted by other documents at any + time. It is inappropriate to use Internet-Drafts as reference + material or to cite them other than as "work in progress." + + This Internet-Draft will expire on January 5, 2015. + +Copyright Notice + + Copyright (c) 2014 IETF Trust and the persons identified as the + document authors. All rights reserved. + + This document is subject to BCP 78 and the IETF Trust's Legal + Provisions Relating to IETF Documents + + + +Lewis, et al. Expires January 5, 2015 [Page 1] + +Internet-Draft LISP Generic Protocol Extension July 2014 + + + (http://trustee.ietf.org/license-info) in effect on the date of + publication of this document. Please review these documents + carefully, as they describe your rights and restrictions with respect + to this document. Code Components extracted from this document must + include Simplified BSD License text as described in Section 4.e of + the Trust Legal Provisions and are provided without warranty as + described in the Simplified BSD License. + + +Table of Contents + + 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 3 + 2. LISP Header Without Protocol Extensions . . . . . . . . . . . 4 + 3. Generic Protocol Extension for LISP (LISP-gpe) . . . . . . . . 5 + 3.1. Multi Protocol Support . . . . . . . . . . . . . . . . . . 5 + 3.2. OAM Support . . . . . . . . . . . . . . . . . . . . . . . 6 + 3.3. Version Bits . . . . . . . . . . . . . . . . . . . . . . . 6 + 4. Backward Compatibility . . . . . . . . . . . . . . . . . . . . 8 + 4.1. LISP-gpe Routers to (legacy) LISP Routers . . . . . . . . 8 + 4.2. (legacy) LISP Routers to LISP-gpe Routers . . . . . . . . 8 + 4.3. Type of Service . . . . . . . . . . . . . . . . . . . . . 8 + 4.4. VLAN Identifier (VID) . . . . . . . . . . . . . . . . . . 8 + 5. LISP-gpe Examples . . . . . . . . . . . . . . . . . . . . . . 9 + 6. Security Considerations . . . . . . . . . . . . . . . . . . . 11 + 7. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 12 + 8. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 13 + 9. References . . . . . . . . . . . . . . . . . . . . . . . . . . 14 + 9.1. Normative References . . . . . . . . . . . . . . . . . . . 14 + 9.2. Informative References . . . . . . . . . . . . . . . . . . 14 + Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 15 + + + + + + + + + + + + + + + + + + + + + +Lewis, et al. Expires January 5, 2015 [Page 2] + +Internet-Draft LISP Generic Protocol Extension July 2014 + + +1. Introduction + + LISP [RFC6830] defines an encapsulation format that carries IPv4 or + IPv6 (henceforth referred to as IP) packets in a LISP header and + outer UDP/IP transport. + + The LISP header does not specify the protocol being encapsulated and + therefore is currently limited to encapsulating only IP packet + payloads. Other protocols, most notably VXLAN [VXLAN] (which defines + a similar header format to LISP), are used to encapsulate L2 + protocols such as Ethernet. LISP [RFC6830] can be extended to + indicate the inner protocol, enabling the encapsulation of Ethernet, + IP or any other desired protocol all the while ensuring compatibility + with existing LISP [RFC6830] deployments. + + As LISP is deployed, there's also the need to provide increased + visibility and diagnostic capabilities within the overlay. + + This document describes extending LISP ([RFC6830]) via the following + changes: + + Next Protocol Bit (P bit): A reserved flag bit is allocated, and set + in the LISP-gpe header to indicate that a next protocol field is + present. + + OAM Flag Bit (O bit): A reserved flag bit is allocated, and set in + the LISP-gpe header, to indicate that the packet is an OAM packet. + + Version: Two reserved bits are allocated, and set in the LISP-gpe + header, to indicate LISP-gpe protocol version. + + Next protocol: An 8 bit next protocol field is present in the LISP- + gpe header. + + + + + + + + + + + + + + + + + + +Lewis, et al. Expires January 5, 2015 [Page 3] + +Internet-Draft LISP Generic Protocol Extension July 2014 + + +2. LISP Header Without Protocol Extensions + + As described in the introduction, the LISP header has no protocol + identifier that indicates the type of payload being carried by LISP. + Because of this, LISP is limited to an IP payload. Furthermore, the + LISP header has no mechanism to signal OAM packets. + + The LISP header contains flags (some defined, some reserved), a + Nonce/Map-version field and an instance ID/Locator-status-bit field. + The flags provide flexibility to define how the reserved bits can be + used to change the definition of the LISP header. + + + +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + |N|L|E|V|I|flags| Nonce/Map-Version | + +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + | Instance ID/Locator-Status-Bits | + +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + + + Figure 1: LISP Header + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +Lewis, et al. Expires January 5, 2015 [Page 4] + +Internet-Draft LISP Generic Protocol Extension July 2014 + + +3. Generic Protocol Extension for LISP (LISP-gpe) + +3.1. Multi Protocol Support + + This draft defines the following changes to the LISP header in order + to support multi-protocol encapsulation. + + P Bit: Flag bit 5 is defined as the Next Protocol bit. The P bit + MUST be set to 1 to indicate the presence of the 8 bit next + protocol field. + + P = 0 indicates that the payload MUST conform to LISP as defined + in [RFC6830]. + + Flag bit 5 was chosen as the P bit because this flag bit is + currently unallocated in LISP [RFC6830]. + + Next Protocol Field: The lower 8 bits of the first word are used to + carry a next protocol. This next protocol field contains the + protocol of the encapsulated payload packet. + + LISP [RFC6830] uses the lower 16 bits of the first word for either + a nonce, an echo-nonce ([RFC6830]) or to support map-versioning + ([RFC6834]). These are all optional capabilities that are + indicated by setting the N, E, and the V bit respectively. + + To maintain the desired data plane compatibility, when the P bit + is set, the N, E, and V bits MUST be set to zero. + + A new protocol registry will be requested from IANA for the Next + Protocol field. This draft defines the following Next Protocol + values: + + 0x1 : IPv4 + + 0x2 : IPv6 + + 0x3 : Ethernet + + 0x4: Network Service Header + + + + + + + + + + + +Lewis, et al. Expires January 5, 2015 [Page 5] + +Internet-Draft LISP Generic Protocol Extension July 2014 + + + 0 1 2 3 + 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 + +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + |N|L|E|V|I|P|R|R| Reserved | Next Protocol | + +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + | Instance ID/Locator-Status-Bits | + +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + + + + Figure 2: LISP-gpe Next Protocol (P=1) + +3.2. OAM Support + + Flag bit 7 is defined as the O bit. When the O bit is set to 1, the + packet is an OAM packet and OAM processing MUST occur. The OAM + protocol details are out of scope for this document. As with the + P-bit, bit 7 is currently a reserved flag in [RFC6830]. + + + + + 0 1 2 3 + 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 + +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + |N|L|E|V|I|P|R|O| Reserved | Next Protocol | + +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + | Instance ID/Locator-Status-Bits | + +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + + + + Figure 3: LISP-gpe OAM bit (P=1) + +3.3. Version Bits + + LISP-gpe bits8 and 9 are defined as version bits. The version field + is used to ensure backward compatibility going forward with future + LISP-gpe updates. + + The initial version for LISP-gpe is 0. + + + + + + + + + + +Lewis, et al. Expires January 5, 2015 [Page 6] + +Internet-Draft LISP Generic Protocol Extension July 2014 + + + 0 1 2 3 + 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 + +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + |N|L|E|V|I|P|R|O|Ver| Reserved | Next Protocol | + +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + | Instance ID/Locator-Status-Bits | + +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + + + + Figure 4: LISP-gpe Version bits (P=1) + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +Lewis, et al. Expires January 5, 2015 [Page 7] + +Internet-Draft LISP Generic Protocol Extension July 2014 + + +4. Backward Compatibility + + Undefined (in RFC6830) flag bits 5 and 7, LISP-gpe P and O bits, were + selected to ensure compatibility with existing LISP [RFC6830] + deployments. + + Similarly, using P = 0 to indicate that the format of the header and + payload conforms to [RFC6830] ensures compatibility with existing + LISP hardware forwarding platforms. + +4.1. LISP-gpe Routers to (legacy) LISP Routers + + A LISP-gpe router MUST not encapsulate non-IP packet nor OAM packets + to a LISP router. A method for determining the capabilities of a + LISP router (gpe or "legacy") is out of the scope of this draft. + + When encapsulating IP packets to a LISP router the P bit SHOULD be + set to 1 and the UDP port MUST be set to 4341. OAM bit MUST be set + to 0. The Next Protocol field SHOULD be 0x1 (IPv4) or 0x2 (IPv6). + The (legacy) LISP router will ignore the P bit and the protocol type + field. The (legacy) LISP router will treat the packet as a LISP + packet and inspect the first nibble of the payload to determine the + IP version. + + When the P bit is set, the N, E, and V bits MUST be set to zero. The + receiving (legacy) LISP router will ignore N, E and V bits, when the + P bit is set. + +4.2. (legacy) LISP Routers to LISP-gpe Routers + + When a LISP-gpe router receives a packet from a (legacy) LISP router, + the P bit MUST not be set and the UDP port MUST be 4341. The payload + MUST be IP, and the LISP-gpe router will inspect the first nibble of + the payload to determine IP version. + +4.3. Type of Service + + When a LISP-gpe router performs Ethernet encapsulation, the inner + 802.1Q [IEEE8021Q] priority code point (PCP) field MAY be mapped from + the encapsulated frame to the Type of Service field in the outer IPv4 + header, or in the case of IPv6 the 'Traffic Class' field. + +4.4. VLAN Identifier (VID) + + When a LISP-gpe router performs Ethernet encapsulation, the inner + header 802.1Q [IEEE8021Q] VLAN Identifier (VID) MAY be mapped to, or + used to determine the LISP Instance ID field. + + + + +Lewis, et al. Expires January 5, 2015 [Page 8] + +Internet-Draft LISP Generic Protocol Extension July 2014 + + +5. LISP-gpe Examples + + This section provides two examples of IP protocols, and one example + of Ethernet encapsulated LISP-gpe using the generic extension + described in this document. + + + + 0 1 2 3 + 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 + +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + |N|L|E|V|I|1|0|0|0| Reserved | NP = IPv4 | + +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + | Instance ID/Locator-Status-Bits | + +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + | Original IPv4 Packet | + +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + + + + Figure 5: IPv4 and LISP-gpe + + + + + 0 1 2 3 + 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 + +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + |N|L|E|V|I|1|0|0|0| Reserved | NP = IPv6 | + +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + | Instance ID/Locator-Status-Bits | + +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + | Original IPv6 Packet | + +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + + + + Figure 6: IPv6 and LISP-gpe + + + + + + + + + + + + + +Lewis, et al. Expires January 5, 2015 [Page 9] + +Internet-Draft LISP Generic Protocol Extension July 2014 + + + 0 1 2 3 + 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 + +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + |N|L|E|V|I|1|0|0|0| Reserved | NP = Ethernet | + +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + | Instance ID/Locator-Status-Bits | + +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + | Original Ethernet Frame | + +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + + + + Figure 7: Ethernet and LISP-gpe + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +Lewis, et al. Expires January 5, 2015 [Page 10] + +Internet-Draft LISP Generic Protocol Extension July 2014 + + +6. Security Considerations + + LISP-gpe security considerations are similar to the LISP security + considerations documented at length in LISP [RFC6830]. With LISP- + gpe, issues such as dataplane spoofing, flooding, and traffic + redirection are dependent on the particular protocol payload + encapsulated. + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +Lewis, et al. Expires January 5, 2015 [Page 11] + +Internet-Draft LISP Generic Protocol Extension July 2014 + + +7. Acknowledgments + + A special thank you goes to Dino Farinacci for his guidance and + detailed review. + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +Lewis, et al. Expires January 5, 2015 [Page 12] + +Internet-Draft LISP Generic Protocol Extension July 2014 + + +8. IANA Considerations + + IANA is requested to set up a registry of "Next Protocol". These are + 8-bit values. Next Protocol values 0, 1, 2, 3 and 4 are defined in + this draft. New values are assigned via Standards Action [RFC5226]. + + +---------------+-------------+---------------+ + | Next Protocol | Description | Reference | + +---------------+-------------+---------------+ + | 0 | Reserved | This document | + | | | | + | 1 | IPv4 | This document | + | | | | + | 2 | IPv6 | This document | + | | | | + | 3 | Ethernet | This document | + | | | | + | 4 | NSH | This document | + | | | | + | 5..253 | Unassigned | | + +---------------+-------------+---------------+ + + Table 1 + + There are ten bits at the beginning of the LISP-gpe header. New + bits are assigned via Standards Action [RFC5226]. + + Bits 0-3 - Assigned by LISP [RFC6830] + Bit 4 - Instance ID (I bit) + Bit 5 - Next Protocol (P bit) + Bit 6 - Reserved + Bit 7 - OAM (O bit) + Bits 8-9 - Version + + + + + + + + + + + + + + + + + + +Lewis, et al. Expires January 5, 2015 [Page 13] + +Internet-Draft LISP Generic Protocol Extension July 2014 + + +9. References + +9.1. Normative References + + [RFC0768] Postel, J., "User Datagram Protocol", STD 6, RFC 768, + August 1980. + + [RFC0791] Postel, J., "Internet Protocol", STD 5, RFC 791, + September 1981. + + [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate + Requirement Levels", BCP 14, RFC 2119, March 1997. + + [RFC5226] Narten, T. and H. Alvestrand, "Guidelines for Writing an + IANA Considerations Section in RFCs", BCP 26, RFC 5226, + May 2008. + +9.2. Informative References + + [ETYPES] The IEEE Registration Authority, "IEEE 802 Numbers", 2012, + <http://www.iana.org/assignments/ieee-802-numbers/ + ieee-802-numbers.xml>. + + [IEEE8021Q] + The IEEE Computer Society, "Media Access Control (MAC) + Bridges and Virtual Bridge Local Area Networks", August + 2012, <http://standards.ieee.org/getieee802/download/ + 802.1Q-2011.pdf>. + + [RFC1700] Reynolds, J. and J. Postel, "Assigned Numbers", RFC 1700, + October 1994. + + [RFC6830] Farinacci, D., Fuller, V., Meyer, D., and D. Lewis, "The + Locator/ID Separation Protocol (LISP)", RFC 6830, + January 2013. + + [RFC6834] Iannone, L., Saucez, D., and O. Bonaventure, "Locator/ID + Separation Protocol (LISP) Map-Versioning", RFC 6834, + January 2013. + + [VXLAN] Dutt, D., Mahalingam, M., Duda, K., Agarwal, P., Kreeger, + L., Sridhar, T., Bursell, M., and C. Wright, "VXLAN: A + Framework for Overlaying Virtualized Layer 2 Networks over + Layer 3 Networks", 2013. + + + + + + + +Lewis, et al. Expires January 5, 2015 [Page 14] + +Internet-Draft LISP Generic Protocol Extension July 2014 + + +Authors' Addresses + + Darrel Lewis + Cisco Systems, Inc. + + Email: darlewis@cisco.com + + + Puneet Agarwal + Broadcom + + Email: pagarwal@broadcom.com + + + Larry Kreeger + Cisco Systems, Inc. + + Email: kreeger@cisco.com + + + Fabio Maino + Cisco Systems, Inc. + + Email: fmaino@cisco.com + + + Paul Quinn + Cisco Systems, Inc. + + Email: paulq@cisco.com + + + Michael Smith + Cisco Systems, Inc. + + Email: michsmit@cisco.com + + + Navindra Yadav + Cisco Systems, Inc. + + Email: nyadav@cisco.com |