aboutsummaryrefslogtreecommitdiffstats
path: root/src/plugins/npt66
diff options
context:
space:
mode:
authorAlexander Chernavin <achernavin@netgate.com>2023-11-07 11:25:21 +0000
committerMatthew Smith <mgsmith@netgate.com>2023-11-12 21:52:13 +0000
commit64d6463d2eac0c0fe434f3a7aa56fe4d85c046d9 (patch)
tree89871a62760cb89dfb37993fcf5b6d825cb6d0eb /src/plugins/npt66
parent4aaedaa2ffaf0cee9ceae62c785f940b6fe23535 (diff)
flowprobe: fix tx flows generated for rewritten traffic
Currently, when IPFIX records generation is enabled for an interface in the TX direction, some rewritten traffic is being sent from that interface, and the Ethernet header's location has changed due to rewriting, generated TX flows will contain fields with wrong and zero values. For example, that can be observed when traffic is rewritten from a subinterface to a hardware interface (i.e. when tags are removed). A TX flow generated in this case will have wrong L2 fields because of an incorrectly located Ethernet header. And zero L3/L4 fields because the Ethernet type will match neither IP4 nor IP6. The same code is executed to generate flows for both input and output features. And the same mechanism is applied to identify the Ethernet header in the buffer's data. However, such general code usually works with the buffer's data conditionally based on the direction. For most input features, the buffer's current_data will likely point to the IP header. For most output features, the buffer's current_data will likely point to the Ethernet header. With this fix: - Keep relying on ethernet_buffer_get_header() to locate the Ethernet header for input features. And start using vlib_buffer_get_current() to locate the Ethernet header for output features. The function will account for the Ethernet header's position change in the buffer's data if there is rewriting. - After fixing Ethernet header determination in the buffer's data, L3/L4 fields will contain non-zero but still incorrect data. That is because IP header determination needs to be fixed too. It currently relies on the fact that the Ethernet header is always located at the beginning of the buffer's data and that l2_hdr_sz can be used as an IP header offset. However, this may not be the case after rewriting. So start calculating the actual offset of the IP header in the buffer's data. - Add a unit test to cover the case. Type: fix Change-Id: Icf3f9e6518912d06dff0d5aa48e103b3dc94edb7 Signed-off-by: Alexander Chernavin <achernavin@netgate.com>
Diffstat (limited to 'src/plugins/npt66')
0 files changed, 0 insertions, 0 deletions
href='#n186'>186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560
/*
 * Copyright (c) 2017 Intel 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 <gtpu/gtpu.h>

/* Statistics (not all errors) */
#define foreach_gtpu_encap_error    \
_(ENCAPSULATED, "good packets encapsulated")

static char * gtpu_encap_error_strings[] = {
#define _(sym,string) string,
  foreach_gtpu_encap_error
#undef _
};

typedef enum {
#define _(sym,str) GTPU_ENCAP_ERROR_##sym,
    foreach_gtpu_encap_error
#undef _
    GTPU_ENCAP_N_ERROR,
} gtpu_encap_error_t;

typedef enum {
    GTPU_ENCAP_NEXT_DROP,
    GTPU_ENCAP_N_NEXT,
} gtpu_encap_next_t;

typedef struct {
  u32 tunnel_index;
  u32 teid;
} gtpu_encap_trace_t;

u8 * format_gtpu_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 *);
  gtpu_encap_trace_t * t
      = va_arg (*args, gtpu_encap_trace_t *);

  s = format (s, "GTPU encap to gtpu_tunnel%d teid %d",
	      t->tunnel_index, t->teid);
  return s;
}


#define foreach_fixed_header4_offset            \
    _(0) _(1) _(2) _(3)

#define foreach_fixed_header6_offset            \
    _(0) _(1) _(2) _(3) _(4) _(5) _(6)

always_inline uword
gtpu_encap_inline (vlib_main_t * vm,
		    vlib_node_runtime_t * node,
		    vlib_frame_t * from_frame,
		    u32 is_ip4)
{
  u32 n_left_from, next_index, * from, * to_next;
  gtpu_main_t * gtm = &gtpu_main;
  vnet_main_t * vnm = gtm->vnet_main;
  vnet_interface_main_t * im = &vnm->interface_main;
  u32 pkts_encapsulated = 0;
  u16 old_l0 = 0, old_l1 = 0;
  u32 thread_index = vlib_get_thread_index();
  u32 stats_sw_if_index, stats_n_packets, stats_n_bytes;
  u32 sw_if_index0 = 0, sw_if_index1 = 0;
  u32 next0 = 0, next1 = 0;
  vnet_hw_interface_t * hi0, * hi1;
  gtpu_tunnel_t * t0 = NULL, * t1 = NULL;

  from = vlib_frame_vector_args (from_frame);
  n_left_from = from_frame->n_vectors;

  next_index = node->cached_next_index;
  stats_sw_if_index = node->runtime_data[0];
  stats_n_packets = stats_n_bytes = 0;

  while (n_left_from > 0)
    {
      u32 n_left_to_next;

      vlib_get_next_frame (vm, node, next_index,
			   to_next, n_left_to_next);

      while (n_left_from >= 4 && n_left_to_next >= 2)
	{
          u32 bi0, bi1;
	  vlib_buffer_t * b0, * b1;
          u32 flow_hash0, flow_hash1;
	  u32 len0, len1;
          ip4_header_t * ip4_0, * ip4_1;
          ip6_header_t * ip6_0, * ip6_1;
          udp_header_t * udp0, * udp1;
          gtpu_header_t * gtpu0, * gtpu1;
          u64 * copy_src0, * copy_dst0;
          u64 * copy_src1, * copy_dst1;
          u32 * copy_src_last0, * copy_dst_last0;
          u32 * copy_src_last1, * copy_dst_last1;
          u16 new_l0, new_l1;
          ip_csum_t sum0, sum1;

	  /* 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);

          flow_hash0 = vnet_l2_compute_flow_hash (b0);
          flow_hash1 = vnet_l2_compute_flow_hash (b1);

	  /* Get next node index and adj index from tunnel next_dpo */
	  sw_if_index0 = vnet_buffer(b0)->sw_if_index[VLIB_TX];
	  hi0 = vnet_get_sup_hw_interface (vnm, sw_if_index0);
	  t0 = &gtm->tunnels[hi0->dev_instance];
	  /* Note: change to always set next0 if it may be set to drop */
	  next0 = t0->next_dpo.dpoi_next_node;
          vnet_buffer(b0)->ip.adj_index[VLIB_TX] = t0->next_dpo.dpoi_index;

	  /* Get next node index and adj index from tunnel next_dpo */
	  sw_if_index1 = vnet_buffer(b1)->sw_if_index[VLIB_TX];
	  hi1 = vnet_get_sup_hw_interface (vnm, sw_if_index1);
	  t1 = &gtm->tunnels[hi1->dev_instance];
	  /* Note: change to always set next1 if it may be set to drop */
	  next1 = t1->next_dpo.dpoi_next_node;
          vnet_buffer(b1)->ip.adj_index[VLIB_TX] = t1->next_dpo.dpoi_index;

          /* Apply the rewrite string. $$$$ vnet_rewrite? */
          vlib_buffer_advance (b0, -(word)_vec_len(t0->rewrite));
          vlib_buffer_advance (b1, -(word)_vec_len(t1->rewrite));

	  if (is_ip4)
	    {
	      ip4_0 = vlib_buffer_get_current(b0);
	      ip4_1 = vlib_buffer_get_current(b1);

	      /* Copy the fixed header */
	      copy_dst0 = (u64 *) ip4_0;
	      copy_src0 = (u64 *) t0->rewrite;
	      copy_dst1 = (u64 *) ip4_1;
	      copy_src1 = (u64 *) t1->rewrite;
	      /* Copy first 32 octets 8-bytes at a time */
#define _(offs) copy_dst0[offs] = copy_src0[offs];
	      foreach_fixed_header4_offset;
#undef _
#define _(offs) copy_dst1[offs] = copy_src1[offs];
	      foreach_fixed_header4_offset;
#undef _
	      /* Last 4 octets. Hopefully gcc will be our friend */
              copy_dst_last0 = (u32 *)(&copy_dst0[4]);
              copy_src_last0 = (u32 *)(&copy_src0[4]);
              copy_dst_last0[0] = copy_src_last0[0];
              copy_dst_last1 = (u32 *)(&copy_dst1[4]);
              copy_src_last1 = (u32 *)(&copy_src1[4]);
              copy_dst_last1[0] = copy_src_last1[0];

	      /* Fix the IP4 checksum and length */
	      sum0 = ip4_0->checksum;
	      new_l0 = /* old_l0 always 0, see the rewrite setup */
                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 */);
	      ip4_0->checksum = ip_csum_fold (sum0);
	      ip4_0->length = new_l0;
	      sum1 = ip4_1->checksum;
	      new_l1 = /* old_l1 always 0, see the rewrite setup */
                clib_host_to_net_u16 (vlib_buffer_length_in_chain (vm, b1));
              sum1 = ip_csum_update (sum1, old_l1, new_l1, ip4_header_t,
				     length /* changed member */);
	      ip4_1->checksum = ip_csum_fold (sum1);
	      ip4_1->length = new_l1;

	      /* Fix UDP length and set source port */
	      udp0 = (udp_header_t *)(ip4_0+1);
	      new_l0 = clib_host_to_net_u16 (vlib_buffer_length_in_chain(vm, b0)
					     - sizeof (*ip4_0));
	      udp0->length = new_l0;
	      udp0->src_port = flow_hash0;
	      udp1 = (udp_header_t *)(ip4_1+1);
	      new_l1 = clib_host_to_net_u16 (vlib_buffer_length_in_chain(vm, b1)
					     - sizeof (*ip4_1));
	      udp1->length = new_l1;
	      udp1->src_port = flow_hash1;

	      /* Fix GTPU length */
	      gtpu0 = (gtpu_header_t *)(udp0+1);
	      new_l0 = clib_host_to_net_u16 (vlib_buffer_length_in_chain(vm, b0)
					     - sizeof (*ip4_0) - sizeof(*udp0));
	      gtpu0->length = new_l0;
	      gtpu1 = (gtpu_header_t *)(udp1+1);
	      new_l1 = clib_host_to_net_u16 (vlib_buffer_length_in_chain(vm, b1)
					     - sizeof (*ip4_1) - sizeof(*udp1));
	      gtpu1->length = new_l1;
	    }
	  else /* ipv6 */
	    {
              int bogus = 0;

	      ip6_0 = vlib_buffer_get_current(b0);
	      ip6_1 = vlib_buffer_get_current(b1);

	      /* Copy the fixed header */
	      copy_dst0 = (u64 *) ip6_0;
	      copy_src0 = (u64 *) t0->rewrite;
	      copy_dst1 = (u64 *) ip6_1;
	      copy_src1 = (u64 *) t1->rewrite;
	      /* Copy first 56 (ip6) octets 8-bytes at a time */
#define _(offs) copy_dst0[offs] = copy_src0[offs];
	      foreach_fixed_header6_offset;
#undef _
#define _(offs) copy_dst1[offs] = copy_src1[offs];
	      foreach_fixed_header6_offset;
#undef _
	      /* Fix IP6 payload length */
	      new_l0 =
                clib_host_to_net_u16 (vlib_buffer_length_in_chain (vm, b0)
				      - sizeof(*ip6_0));
	      ip6_0->payload_length = new_l0;
	      new_l1 =
                clib_host_to_net_u16 (vlib_buffer_length_in_chain (vm, b1)
				      - sizeof(*ip6_1));
	      ip6_1->payload_length = new_l1;

	      /* Fix UDP length  and set source port */
	      udp0 = (udp_header_t *)(ip6_0+1);
	      udp0->length = new_l0;
	      udp0->src_port = flow_hash0;
	      udp1 = (udp_header_t *)(ip6_1+1);
	      udp1->length = new_l1;
	      udp1->src_port = flow_hash1;

	      /* IPv6 UDP checksum is mandatory */
	      udp0->checksum = ip6_tcp_udp_icmp_compute_checksum(vm, b0,
								 ip6_0, &bogus);
	      if (udp0->checksum == 0)
		udp0->checksum = 0xffff;
	      udp1->checksum = ip6_tcp_udp_icmp_compute_checksum(vm, b1,
                                                        ip6_1, &bogus);
	      if (udp1->checksum == 0)
		udp1->checksum = 0xffff;

	      /* Fix GTPU length */
	      gtpu0 = (gtpu_header_t *)(udp0+1);
	      new_l0 = clib_host_to_net_u16 (vlib_buffer_length_in_chain(vm, b0)
					     - sizeof (*ip4_0) - sizeof(*udp0));
	      gtpu0->length = new_l0;
	      gtpu1 = (gtpu_header_t *)(udp1+1);
	      new_l1 = clib_host_to_net_u16 (vlib_buffer_length_in_chain(vm, b1)
					     - sizeof (*ip4_1) - sizeof(*udp1));
	      gtpu1->length = new_l1;
	    }

          pkts_encapsulated += 2;
 	  len0 = vlib_buffer_length_in_chain (vm, b0);
 	  len1 = vlib_buffer_length_in_chain (vm, b1);
	  stats_n_packets += 2;
	  stats_n_bytes += len0 + len1;

	  /* Batch stats increment on the same gtpu tunnel so counter is not
	     incremented per packet. Note stats are still incremented for deleted
	     and admin-down tunnel where packets are dropped. It is not worthwhile
	     to check for this rare case and affect normal path performance. */
	  if (PREDICT_FALSE ((sw_if_index0 != stats_sw_if_index) ||
			     (sw_if_index1 != stats_sw_if_index)))
	    {
	      stats_n_packets -= 2;
	      stats_n_bytes -= len0 + len1;
	      if (sw_if_index0 == sw_if_index1)
	        {
		  if (stats_n_packets)
		    vlib_increment_combined_counter
		      (im->combined_sw_if_counters + VNET_INTERFACE_COUNTER_TX,
		       thread_index, stats_sw_if_index,
		       stats_n_packets, stats_n_bytes);
		  stats_sw_if_index = sw_if_index0;
		  stats_n_packets = 2;
		  stats_n_bytes = len0 + len1;
	        }
	      else
	        {
		  vlib_increment_combined_counter
		      (im->combined_sw_if_counters + VNET_INTERFACE_COUNTER_TX,
		       thread_index, sw_if_index0, 1, len0);
		  vlib_increment_combined_counter
		      (im->combined_sw_if_counters + VNET_INTERFACE_COUNTER_TX,
		       thread_index, sw_if_index1, 1, len1);
		}
	    }

	  if (PREDICT_FALSE(b0->flags & VLIB_BUFFER_IS_TRACED))
            {
              gtpu_encap_trace_t *tr =
                vlib_add_trace (vm, node, b0, sizeof (*tr));
              tr->tunnel_index = t0 - gtm->tunnels;
              tr->teid = t0->teid;
           }

          if (PREDICT_FALSE(b1->flags & VLIB_BUFFER_IS_TRACED))
            {
              gtpu_encap_trace_t *tr =
                vlib_add_trace (vm, node, b1, sizeof (*tr));
              tr->tunnel_index = t1 - gtm->tunnels;
              tr->teid = t1->teid;
            }

	  vlib_validate_buffer_enqueue_x2 (vm, node, next_index,
					   to_next, n_left_to_next,
					   bi0, bi1, next0, next1);
	}

      while (n_left_from > 0 && n_left_to_next > 0)
	{
	  u32 bi0;
	  vlib_buffer_t * b0;
          u32 flow_hash0;
	  u32 len0;
          ip4_header_t * ip4_0;
          ip6_header_t * ip6_0;
          udp_header_t * udp0;
          gtpu_header_t * gtpu0;
          u64 * copy_src0, * copy_dst0;
          u32 * copy_src_last0, * copy_dst_last0;
          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);

          flow_hash0 = vnet_l2_compute_flow_hash(b0);

	  /* Get next node index and adj index from tunnel next_dpo */
	  sw_if_index0 = vnet_buffer(b0)->sw_if_index[VLIB_TX];
	  hi0 = vnet_get_sup_hw_interface (vnm, sw_if_index0);
	  t0 = &gtm->tunnels[hi0->dev_instance];
	  /* Note: change to always set next0 if it may be set to drop */
	  next0 = t0->next_dpo.dpoi_next_node;
	  vnet_buffer(b0)->ip.adj_index[VLIB_TX] = t0->next_dpo.dpoi_index;

          /* Apply the rewrite string. $$$$ vnet_rewrite? */
          vlib_buffer_advance (b0, -(word)_vec_len(t0->rewrite));

	  if (is_ip4)
	    {
	      ip4_0 = vlib_buffer_get_current(b0);

	      /* Copy the fixed header */
	      copy_dst0 = (u64 *) ip4_0;
	      copy_src0 = (u64 *) t0->rewrite;
	      /* Copy first 32 octets 8-bytes at a time */
#define _(offs) copy_dst0[offs] = copy_src0[offs];
	      foreach_fixed_header4_offset;
#undef _
	      /* Last 4 octets. Hopefully gcc will be our friend */
              copy_dst_last0 = (u32 *)(&copy_dst0[4]);
              copy_src_last0 = (u32 *)(&copy_src0[4]);
              copy_dst_last0[0] = copy_src_last0[0];

	      /* Fix the IP4 checksum and length */
	      sum0 = ip4_0->checksum;
	      new_l0 = /* old_l0 always 0, see the rewrite setup */
                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 */);
	      ip4_0->checksum = ip_csum_fold (sum0);
	      ip4_0->length = new_l0;

	      /* Fix UDP length and set source port */
	      udp0 = (udp_header_t *)(ip4_0+1);
	      new_l0 = clib_host_to_net_u16 (vlib_buffer_length_in_chain(vm, b0)
					     - sizeof (*ip4_0));
	      udp0->length = new_l0;
	      udp0->src_port = flow_hash0;

	      /* Fix GTPU length */
	      gtpu0 = (gtpu_header_t *)(udp0+1);
	      new_l0 = clib_host_to_net_u16 (vlib_buffer_length_in_chain(vm, b0)
					     - sizeof (*ip4_0) - sizeof(*udp0));
	      gtpu0->length = new_l0;
	    }

	  else /* ip6 path */
	    {
              int bogus = 0;

	      ip6_0 = vlib_buffer_get_current(b0);
	      /* Copy the fixed header */
	      copy_dst0 = (u64 *) ip6_0;
	      copy_src0 = (u64 *) t0->rewrite;
	      /* Copy first 56 (ip6) octets 8-bytes at a time */
#define _(offs) copy_dst0[offs] = copy_src0[offs];
	      foreach_fixed_header6_offset;
#undef _
	      /* Fix IP6 payload length */
	      new_l0 =
                clib_host_to_net_u16 (vlib_buffer_length_in_chain (vm, b0)
				      - sizeof(*ip6_0));
	      ip6_0->payload_length = new_l0;

	      /* Fix UDP length  and set source port */
	      udp0 = (udp_header_t *)(ip6_0+1);
	      udp0->length = new_l0;
	      udp0->src_port = flow_hash0;

	      /* IPv6 UDP checksum is mandatory */
	      udp0->checksum = ip6_tcp_udp_icmp_compute_checksum(vm, b0,
								 ip6_0, &bogus);
	      if (udp0->checksum == 0)
		udp0->checksum = 0xffff;

	      /* Fix GTPU length */
	      gtpu0 = (gtpu_header_t *)(udp0+1);
	      new_l0 = clib_host_to_net_u16 (vlib_buffer_length_in_chain(vm, b0)
					     - sizeof (*ip4_0) - sizeof(*udp0));
	      gtpu0->length = new_l0;
	    }

          pkts_encapsulated ++;
	  len0 = vlib_buffer_length_in_chain (vm, b0);
	  stats_n_packets += 1;
	  stats_n_bytes += len0;

	  /* Batch stats increment on the same gtpu tunnel so counter is not
	     incremented per packet. Note stats are still incremented for deleted
	     and admin-down tunnel where packets are dropped. It is not worthwhile
	     to check for this rare case and affect normal path performance. */
	  if (PREDICT_FALSE (sw_if_index0 != stats_sw_if_index))
	    {
	      stats_n_packets -= 1;
	      stats_n_bytes -= len0;
	      if (stats_n_packets)
		vlib_increment_combined_counter
		  (im->combined_sw_if_counters + VNET_INTERFACE_COUNTER_TX,
		   thread_index, stats_sw_if_index,
		   stats_n_packets, stats_n_bytes);
	      stats_n_packets = 1;
	      stats_n_bytes = len0;
	      stats_sw_if_index = sw_if_index0;
	    }

          if (PREDICT_FALSE(b0->flags & VLIB_BUFFER_IS_TRACED))
            {
              gtpu_encap_trace_t *tr =
                vlib_add_trace (vm, node, b0, sizeof (*tr));
              tr->tunnel_index = t0 - gtm->tunnels;
              tr->teid = t0->teid;
            }
	  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);
    }

  /* Do we still need this now that tunnel tx stats is kept? */
  vlib_node_increment_counter (vm, node->node_index,
                               GTPU_ENCAP_ERROR_ENCAPSULATED,
                               pkts_encapsulated);

  /* Increment any remaining batch stats */
  if (stats_n_packets)
    {
      vlib_increment_combined_counter
	(im->combined_sw_if_counters + VNET_INTERFACE_COUNTER_TX,
	 thread_index, stats_sw_if_index, stats_n_packets, stats_n_bytes);
      node->runtime_data[0] = stats_sw_if_index;
    }

  return from_frame->n_vectors;
}

static uword
gtpu4_encap (vlib_main_t * vm,
	      vlib_node_runtime_t * node,
	      vlib_frame_t * from_frame)
{
  return gtpu_encap_inline (vm, node, from_frame, /* is_ip4 */ 1);
}

static uword
gtpu6_encap (vlib_main_t * vm,
	      vlib_node_runtime_t * node,
	      vlib_frame_t * from_frame)
{
  return gtpu_encap_inline (vm, node, from_frame, /* is_ip4 */ 0);
}

VLIB_REGISTER_NODE (gtpu4_encap_node) = {
  .function = gtpu4_encap,
  .name = "gtpu4-encap",
  .vector_size = sizeof (u32),
  .format_trace = format_gtpu_encap_trace,
  .type = VLIB_NODE_TYPE_INTERNAL,
  .n_errors = ARRAY_LEN(gtpu_encap_error_strings),
  .error_strings = gtpu_encap_error_strings,
  .n_next_nodes = GTPU_ENCAP_N_NEXT,
  .next_nodes = {
        [GTPU_ENCAP_NEXT_DROP] = "error-drop",
  },
};

VLIB_NODE_FUNCTION_MULTIARCH (gtpu4_encap_node, gtpu4_encap)

VLIB_REGISTER_NODE (gtpu6_encap_node) = {
  .function = gtpu6_encap,
  .name = "gtpu6-encap",
  .vector_size = sizeof (u32),
  .format_trace = format_gtpu_encap_trace,
  .type = VLIB_NODE_TYPE_INTERNAL,
  .n_errors = ARRAY_LEN(gtpu_encap_error_strings),
  .error_strings = gtpu_encap_error_strings,
  .n_next_nodes = GTPU_ENCAP_N_NEXT,
  .next_nodes = {
        [GTPU_ENCAP_NEXT_DROP] = "error-drop",
  },
};

VLIB_NODE_FUNCTION_MULTIARCH (gtpu6_encap_node, gtpu6_encap)