aboutsummaryrefslogtreecommitdiffstats
path: root/src/plugins/af_packet/node.c
blob: da01492a232c6e5171b588d54e361b72578543e0 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
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
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
/*
 *------------------------------------------------------------------
 * af_packet.c - linux kernel packet interface
 *
 * Copyright (c) 2016 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 <linux/if_packet.h>

#include <vlib/vlib.h>
#include <vlib/unix/unix.h>
#include <vnet/ip/ip.h>
#include <vnet/ethernet/ethernet.h>
#include <vnet/interface/rx_queue_funcs.h>
#include <vnet/feature/feature.h>
#include <vnet/ethernet/packet.h>

#include <af_packet/af_packet.h>
#include <vnet/devices/virtio/virtio_std.h>

#define foreach_af_packet_input_error                                         \
  _ (PARTIAL_PKT, "partial packet")                                           \
  _ (TIMEDOUT_BLK, "timed out block")                                         \
  _ (TOTAL_RECV_BLK, "total received block")
typedef enum
{
#define _(f,s) AF_PACKET_INPUT_ERROR_##f,
  foreach_af_packet_input_error
#undef _
    AF_PACKET_INPUT_N_ERROR,
} af_packet_input_error_t;

static char *af_packet_input_error_strings[] = {
#define _(n,s) s,
  foreach_af_packet_input_error
#undef _
};

typedef struct
{
  u32 next_index;
  u32 hw_if_index;
  u16 queue_id;
  int block;
  u32 pkt_num;
  void *block_start;
  block_desc_t bd;
  union
  {
    tpacket3_hdr_t tph3;
    tpacket2_hdr_t tph2;
  };
  vnet_virtio_net_hdr_t vnet_hdr;
  u8 is_v3;
} af_packet_input_trace_t;

static u8 *
format_af_packet_input_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 *);
  af_packet_input_trace_t *t = va_arg (*args, af_packet_input_trace_t *);
  u32 indent = format_get_indent (s);

  s = format (s, "af_packet: hw_if_index %d rx-queue %u next-index %d",
	      t->hw_if_index, t->queue_id, t->next_index);

  if (t->is_v3)
    {
      s = format (
	s, "\n%Ublock %u:\n%Uaddress %p version %u seq_num %lu pkt_num %u",
	format_white_space, indent + 2, t->block, format_white_space,
	indent + 4, t->block_start, t->bd.version, t->bd.hdr.bh1.seq_num,
	t->pkt_num);
      s = format (
	s,
	"\n%Utpacket3_hdr:\n%Ustatus 0x%x len %u snaplen %u mac %u net %u"
	"\n%Usec 0x%x nsec 0x%x vlan %U"
#ifdef TP_STATUS_VLAN_TPID_VALID
	" vlan_tpid %u"
#endif
	,
	format_white_space, indent + 2, format_white_space, indent + 4,
	t->tph3.tp_status, t->tph3.tp_len, t->tph3.tp_snaplen, t->tph3.tp_mac,
	t->tph3.tp_net, format_white_space, indent + 4, t->tph3.tp_sec,
	t->tph3.tp_nsec, format_ethernet_vlan_tci, t->tph3.hv1.tp_vlan_tci
#ifdef TP_STATUS_VLAN_TPID_VALID
	,
	t->tph3.hv1.tp_vlan_tpid
#endif
      );
    }
  else
    {
      s = format (
	s,
	"\n%Utpacket2_hdr:\n%Ustatus 0x%x len %u snaplen %u mac %u net %u"
	"\n%Usec 0x%x nsec 0x%x vlan %U"
#ifdef TP_STATUS_VLAN_TPID_VALID
	" vlan_tpid %u"
#endif
	,
	format_white_space, indent + 2, format_white_space, indent + 4,
	t->tph2.tp_status, t->tph2.tp_len, t->tph2.tp_snaplen, t->tph2.tp_mac,
	t->tph2.tp_net, format_white_space, indent + 4, t->tph2.tp_sec,
	t->tph2.tp_nsec, format_ethernet_vlan_tci, t->tph2.tp_vlan_tci
#ifdef TP_STATUS_VLAN_TPID_VALID
	,
	t->tph2.tp_vlan_tpid
#endif
      );
    }

  s = format (s,
	      "\n%Uvnet-hdr:\n%Uflags 0x%02x gso_type 0x%02x hdr_len %u"
	      "\n%Ugso_size %u csum_start %u csum_offset %u",
	      format_white_space, indent + 2, format_white_space, indent + 4,
	      t->vnet_hdr.flags, t->vnet_hdr.gso_type, t->vnet_hdr.hdr_len,
	      format_white_space, indent + 4, t->vnet_hdr.gso_size,
	      t->vnet_hdr.csum_start, t->vnet_hdr.csum_offset);
  return s;
}

always_inline void
buffer_add_to_chain (vlib_buffer_t *b, vlib_buffer_t *first_b,
		     vlib_buffer_t *prev_b, u32 bi)
{
  /* update first buffer */
  first_b->total_length_not_including_first_buffer += b->current_length;

  /* update previous buffer */
  prev_b->next_buffer = bi;
  prev_b->flags |= VLIB_BUFFER_NEXT_PRESENT;

  /* update current buffer */
  b->next_buffer = ~0;
}

static_always_inline void
fill_gso_offload (vlib_buffer_t *b, u32 gso_size, u8 l4_hdr_sz)
{
  b->flags |= VNET_BUFFER_F_GSO;
  vnet_buffer2 (b)->gso_size = gso_size;
  vnet_buffer2 (b)->gso_l4_hdr_sz = l4_hdr_sz;
}

static_always_inline void
fill_cksum_offload (vlib_buffer_t *b, u8 *l4_hdr_sz, u8 is_ip)
{
  vnet_buffer_oflags_t oflags = 0;
  u16 l2hdr_sz = 0;
  u16 ethertype = 0;
  u8 l4_proto = 0;

  if (is_ip)
    {
      switch (b->data[0] & 0xf0)
	{
	case 0x40:
	  ethertype = ETHERNET_TYPE_IP4;
	  break;
	case 0x60:
	  ethertype = ETHERNET_TYPE_IP6;
	  break;
	}
    }
  else
    {
      ethernet_header_t *eth = (ethernet_header_t *) b->data;
      ethertype = clib_net_to_host_u16 (eth->type);
      l2hdr_sz = sizeof (ethernet_header_t);
      if (ethernet_frame_is_tagged (ethertype))
	{
	  ethernet_vlan_header_t *vlan = (ethernet_vlan_header_t *) (eth + 1);

	  ethertype = clib_net_to_host_u16 (vlan->type);
	  l2hdr_sz += sizeof (*vlan);
	  if (ethertype == ETHERNET_TYPE_VLAN)
	    {
	      vlan++;
	      ethertype = clib_net_to_host_u16 (vlan->type);
	      l2hdr_sz += sizeof (*vlan);
	    }
	}
    }

  vnet_buffer (b)->l2_hdr_offset = 0;
  vnet_buffer (b)->l3_hdr_offset = l2hdr_sz;

  if (ethertype == ETHERNET_TYPE_IP4)
    {
      ip4_header_t *ip4 = (ip4_header_t *) (b->data + l2hdr_sz);
      vnet_buffer (b)->l4_hdr_offset = l2hdr_sz + ip4_header_bytes (ip4);
      b->flags |= (VNET_BUFFER_F_IS_IP4 | VNET_BUFFER_F_L2_HDR_OFFSET_VALID |
		   VNET_BUFFER_F_L3_HDR_OFFSET_VALID |
		   VNET_BUFFER_F_L4_HDR_OFFSET_VALID);

      l4_proto = ip4->protocol;
    }
  else if (ethertype == ETHERNET_TYPE_IP6)
    {
      ip6_header_t *ip6 = (ip6_header_t *) (b->data + l2hdr_sz);
      b->flags |= (VNET_BUFFER_F_IS_IP6 | VNET_BUFFER_F_L2_HDR_OFFSET_VALID |
		   VNET_BUFFER_F_L3_HDR_OFFSET_VALID |
		   VNET_BUFFER_F_L4_HDR_OFFSET_VALID);
      u16 ip6_hdr_len = sizeof (ip6_header_t);

      if (ip6_ext_hdr (ip6->protocol))
	{
	  ip6_ext_header_t *p = (void *) (ip6 + 1);
	  ip6_hdr_len += ip6_ext_header_len (p);
	  while (ip6_ext_hdr (p->next_hdr))
	    {
	      ip6_hdr_len += ip6_ext_header_len (p);
	      p = ip6_ext_next_header (p);
	    }
	  l4_proto = p->next_hdr;
	}
      else
	l4_proto = ip6->protocol;
      vnet_buffer (b)->l4_hdr_offset = l2hdr_sz + ip6_hdr_len;
    }

  if (l4_proto == IP_PROTOCOL_TCP)
    {
      oflags |= VNET_BUFFER_OFFLOAD_F_TCP_CKSUM;
      tcp_header_t *tcp =
	(tcp_header_t *) (b->data + vnet_buffer (b)->l4_hdr_offset);
      *l4_hdr_sz = tcp_header_bytes (tcp);
    }
  else if (l4_proto == IP_PROTOCOL_UDP)
    {
      oflags |= VNET_BUFFER_OFFLOAD_F_UDP_CKSUM;
      *l4_hdr_sz = sizeof (udp_header_t);
    }

  if (oflags)
    vnet_buffer_offload_flags_set (b, oflags);
}

always_inline uword
af_packet_v3_device_input_fn (vlib_main_t *vm, vlib_node_runtime_t *node,
			      vlib_frame_t *frame, af_packet_if_t *apif,
			      u16 queue_id, u8 is_cksum_gso_enabled)
{
  af_packet_main_t *apm = &af_packet_main;
  af_packet_queue_t *rx_queue = vec_elt_at_index (apif->rx_queues, queue_id);
  tpacket3_hdr_t *tph;
  u32 next_index;
  u32 n_free_bufs;
  u32 n_rx_packets = 0;
  u32 n_rx_bytes = 0;
  u32 timedout_blk = 0;
  u32 total = 0;
  u32 *to_next = 0;
  u32 block = rx_queue->next_rx_block;
  u32 block_nr = rx_queue->rx_req->req3.tp_block_nr;
  u8 *block_start = 0;
  uword n_trace = vlib_get_trace_count (vm, node);
  u32 thread_index = vm->thread_index;
  u32 n_buffer_bytes = vlib_buffer_get_default_data_size (vm);
  u32 min_bufs = rx_queue->rx_req->req3.tp_frame_size / n_buffer_bytes;
  u32 num_pkts = 0;
  u32 rx_frame_offset = 0;
  block_desc_t *bd = 0;
  vlib_buffer_t bt = {};
  u8 is_ip = (apif->mode == AF_PACKET_IF_MODE_IP);

  if (is_ip)
    next_index = VNET_DEVICE_INPUT_NEXT_IP4_INPUT;
  else
    {
      next_index = VNET_DEVICE_INPUT_NEXT_ETHERNET_INPUT;
      if (PREDICT_FALSE (apif->per_interface_next_index != ~0))
	next_index = apif->per_interface_next_index;

      /* redirect if feature path enabled */
      vnet_feature_start_device_input (apif->sw_if_index, &next_index, &bt);
    }

  if ((((block_desc_t *) (block_start = rx_queue->rx_ring[block]))
	 ->hdr.bh1.block_status &
       TP_STATUS_USER) != 0)
    {
      u32 n_required = 0;
      bd = (block_desc_t *) block_start;

      if (PREDICT_FALSE (rx_queue->is_rx_pending))
	{
	  num_pkts = rx_queue->num_rx_pkts;
	  rx_frame_offset = rx_queue->rx_frame_offset;
	  rx_queue->is_rx_pending = 0;
	}
      else
	{
	  num_pkts = bd->hdr.bh1.num_pkts;
	  rx_frame_offset = bd->hdr.bh1.offset_to_first_pkt;
	  total++;

	  if (TP_STATUS_BLK_TMO & bd->hdr.bh1.block_status)
	    timedout_blk++;
	}

      n_required = clib_max (num_pkts, VLIB_FRAME_SIZE);
      n_free_bufs = vec_len (apm->rx_buffers[thread_index]);
      if (PREDICT_FALSE (n_free_bufs < n_required))
	{
	  vec_validate (apm->rx_buffers[thread_index],
			n_required + n_free_bufs - 1);
	  n_free_bufs += vlib_buffer_alloc (
	    vm, &apm->rx_buffers[thread_index][n_free_bufs], n_required);
	  vec_set_len (apm->rx_buffers[thread_index], n_free_bufs);
	}

      while (num_pkts && (n_free_bufs >= min_bufs))
	{
	  u32 next0 = next_index;
	  u32 n_left_to_next;

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

	  while (num_pkts && n_left_to_next && (n_free_bufs >= min_bufs))
	    {
	      tph = (tpacket3_hdr_t *) (block_start + rx_frame_offset);

	      if (num_pkts > 1)
		CLIB_PREFETCH (block_start + rx_frame_offset +
				 tph->tp_next_offset,
			       2 * CLIB_CACHE_LINE_BYTES, LOAD);

	      vlib_buffer_t *b0 = 0, *first_b0 = 0, *prev_b0 = 0;
	      vnet_virtio_net_hdr_t *vnet_hdr = 0;
	      u32 data_len = tph->tp_snaplen;
	      u32 offset = 0;
	      u32 bi0 = ~0, first_bi0 = ~0;
	      u8 l4_hdr_sz = 0;

	      if (is_cksum_gso_enabled)
		vnet_hdr =
		  (vnet_virtio_net_hdr_t *) ((u8 *) tph + tph->tp_mac -
					     sizeof (vnet_virtio_net_hdr_t));

	      // save current state and return
	      if (PREDICT_FALSE (((data_len / n_buffer_bytes) + 1) >
				 vec_len (apm->rx_buffers[thread_index])))
		{
		  rx_queue->rx_frame_offset = rx_frame_offset;
		  rx_queue->num_rx_pkts = num_pkts;
		  rx_queue->is_rx_pending = 1;
		  vlib_put_next_frame (vm, node, next_index, n_left_to_next);
		  goto done;
		}

	      while (data_len)
		{
		  /* grab free buffer */
		  u32 last_empty_buffer =
		    vec_len (apm->rx_buffers[thread_index]) - 1;
		  bi0 = apm->rx_buffers[thread_index][last_empty_buffer];
		  vec_set_len (apm->rx_buffers[thread_index],
			       last_empty_buffer);
		  n_free_bufs--;

		  /* copy data */
		  u32 bytes_to_copy =
		    data_len > n_buffer_bytes ? n_buffer_bytes : data_len;
		  u32 vlan_len = 0;
		  u32 bytes_copied = 0;

		  b0 = vlib_get_buffer (vm, bi0);
		  b0->current_data = 0;

		  /* Kernel removes VLAN headers, so reconstruct VLAN */
		  if (PREDICT_FALSE (tph->tp_status & TP_STATUS_VLAN_VALID))
		    {
		      if (PREDICT_TRUE (offset == 0))
			{
			  clib_memcpy_fast (vlib_buffer_get_current (b0),
					    (u8 *) tph + tph->tp_mac,
					    sizeof (ethernet_header_t));
			  ethernet_header_t *eth =
			    vlib_buffer_get_current (b0);
			  ethernet_vlan_header_t *vlan =
			    (ethernet_vlan_header_t *) (eth + 1);
			  vlan->priority_cfi_and_id =
			    clib_host_to_net_u16 (tph->hv1.tp_vlan_tci);
			  vlan->type = eth->type;
			  eth->type =
			    clib_host_to_net_u16 (ETHERNET_TYPE_VLAN);
			  vlan_len = sizeof (ethernet_vlan_header_t);
			  bytes_copied = sizeof (ethernet_header_t);
			}
		    }
		  clib_memcpy_fast (((u8 *) vlib_buffer_get_current (b0)) +
				      bytes_copied + vlan_len,
				    (u8 *) tph + tph->tp_mac + offset +
				      bytes_copied,
				    (bytes_to_copy - bytes_copied));

		  /* fill buffer header */
		  b0->current_length = bytes_to_copy + vlan_len;

		  if (offset == 0)
		    {
		      b0->total_length_not_including_first_buffer = 0;
		      b0->flags = VLIB_BUFFER_TOTAL_LENGTH_VALID;
		      vnet_buffer (b0)->sw_if_index[VLIB_RX] =
			apif->sw_if_index;
		      vnet_buffer (b0)->sw_if_index[VLIB_TX] = (u32) ~0;
		      first_b0 = b0;
		      first_bi0 = bi0;
		      if (is_cksum_gso_enabled)
			{
			  if (vnet_hdr->flags & VIRTIO_NET_HDR_F_NEEDS_CSUM)
			    fill_cksum_offload (first_b0, &l4_hdr_sz, is_ip);
			  if (vnet_hdr->gso_type & (VIRTIO_NET_HDR_GSO_TCPV4 |
						    VIRTIO_NET_HDR_GSO_TCPV6))
			    fill_gso_offload (first_b0, vnet_hdr->gso_size,
					      l4_hdr_sz);
			}
		    }
		  else
		    buffer_add_to_chain (b0, first_b0, prev_b0, bi0);

		  prev_b0 = b0;
		  offset += bytes_to_copy;
		  data_len -= bytes_to_copy;
		}
	      n_rx_packets++;
	      n_rx_bytes += tph->tp_snaplen;
	      to_next[0] = first_bi0;
	      to_next += 1;
	      n_left_to_next--;

	      /* drop partial packets */
	      if (PREDICT_FALSE (tph->tp_len != tph->tp_snaplen))
		{
		  next0 = VNET_DEVICE_INPUT_NEXT_DROP;
		  first_b0->error =
		    node->errors[AF_PACKET_INPUT_ERROR_PARTIAL_PKT];
		}
	      else
		{
		  if (PREDICT_FALSE (apif->mode == AF_PACKET_IF_MODE_IP))
		    {
		      switch (first_b0->data[0] & 0xf0)
			{
			case 0x40:
			  next0 = VNET_DEVICE_INPUT_NEXT_IP4_INPUT;
			  break;
			case 0x60:
			  next0 = VNET_DEVICE_INPUT_NEXT_IP6_INPUT;
			  break;
			default:
			  next0 = VNET_DEVICE_INPUT_NEXT_DROP;
			  break;
			}
		      if (PREDICT_FALSE (apif->per_interface_next_index != ~0))
			next0 = apif->per_interface_next_index;
		    }
		  else
		    {
		      next0 = VNET_DEVICE_INPUT_NEXT_ETHERNET_INPUT;
		      if (PREDICT_FALSE (apif->per_interface_next_index != ~0))
			next0 = apif->per_interface_next_index;
		      /* copy feature arc data from template */
		      first_b0->current_config_index = bt.current_config_index;
		      vnet_buffer (first_b0)->feature_arc_index =
			vnet_buffer (&bt)->feature_arc_index;
		    }
		}

	      /* trace */
	      if (PREDICT_FALSE (n_trace > 0 &&
				 vlib_trace_buffer (vm, node, next0, first_b0,
						    /* follow_chain */ 0)))
		{
		  af_packet_input_trace_t *tr;
		  vlib_set_trace_count (vm, node, --n_trace);
		  tr = vlib_add_trace (vm, node, first_b0, sizeof (*tr));
		  tr->is_v3 = 1;
		  tr->next_index = next0;
		  tr->hw_if_index = apif->hw_if_index;
		  tr->queue_id = queue_id;
		  tr->block = block;
		  tr->block_start = bd;
		  tr->pkt_num = bd->hdr.bh1.num_pkts - num_pkts;
		  clib_memcpy_fast (&tr->bd, bd, sizeof (block_desc_t));
		  clib_memcpy_fast (&tr->tph3, tph, sizeof (tpacket3_hdr_t));
		  if (is_cksum_gso_enabled)
		    clib_memcpy_fast (&tr->vnet_hdr, vnet_hdr,
				      sizeof (vnet_virtio_net_hdr_t));
		  else
		    clib_memset_u8 (&tr->vnet_hdr, 0,
				    sizeof (vnet_virtio_net_hdr_t));
		}

	      /* enque and take next packet */
	      vlib_validate_buffer_enqueue_x1 (vm, node, next_index, to_next,
					       n_left_to_next, first_bi0,
					       next0);

	      /* next packet */
	      num_pkts--;
	      rx_frame_offset += tph->tp_next_offset;
	    }

	  vlib_put_next_frame (vm, node, next_index, n_left_to_next);
	}

      if (PREDICT_TRUE (num_pkts == 0))
	{
	  bd->hdr.bh1.block_status = TP_STATUS_KERNEL;
	  block = (block + 1) % block_nr;
	}
      else
	{
	  rx_queue->rx_frame_offset = rx_frame_offset;
	  rx_queue->num_rx_pkts = num_pkts;
	  rx_queue->is_rx_pending = 1;
	}
    }

  rx_queue->next_rx_block = block;

done:

  if (apm->polling_count == 0)
    {
      if ((((block_desc_t *) (block_start = rx_queue->rx_ring[block]))
	     ->hdr.bh1.block_status &
	   TP_STATUS_USER) != 0)
	vlib_node_set_state (vm, node->node_index, VLIB_NODE_STATE_POLLING);
      else
	vlib_node_set_state (vm, node->node_index, VLIB_NODE_STATE_INTERRUPT);
    }

  vlib_error_count (vm, node->node_index, AF_PACKET_INPUT_ERROR_TOTAL_RECV_BLK,
		    total);
  vlib_error_count (vm, node->node_index, AF_PACKET_INPUT_ERROR_TIMEDOUT_BLK,
		    timedout_blk);

  vlib_increment_combined_counter
    (vnet_get_main ()->interface_main.combined_sw_if_counters
     + VNET_INTERFACE_COUNTER_RX,
     vlib_get_thread_index (), apif->hw_if_index, n_rx_packets, n_rx_bytes);

  vnet_device_increment_rx_packets (thread_index, n_rx_packets);
  return n_rx_packets;
}

always_inline uword
af_packet_v2_device_input_fn (vlib_main_t *vm, vlib_node_runtime_t *node,
			      vlib_frame_t *frame, af_packet_if_t *apif,
			      u16 queue_id, u8 is_cksum_gso_enabled)
{
  af_packet_main_t *apm = &af_packet_main;
  af_packet_queue_t *rx_queue = vec_elt_at_index (apif->rx_queues, queue_id);
  tpacket2_hdr_t *tph;
  u32 next_index;
  u32 block = 0;
  u32 rx_frame;
  u32 n_free_bufs;
  u32 n_rx_packets = 0;
  u32 n_rx_bytes = 0;
  u32 *to_next = 0;
  u32 frame_size = rx_queue->rx_req->req.tp_frame_size;
  u32 frame_num = rx_queue->rx_req->req.tp_frame_nr;
  u8 *block_start = rx_queue->rx_ring[block];
  uword n_trace = vlib_get_trace_count (vm, node);
  u32 thread_index = vm->thread_index;
  u32 n_buffer_bytes = vlib_buffer_get_default_data_size (vm);
  u32 min_bufs = rx_queue->rx_req->req.tp_frame_size / n_buffer_bytes;
  u8 is_ip = (apif->mode == AF_PACKET_IF_MODE_IP);
  vlib_buffer_t bt = {};

  if (is_ip)
    {
      next_index = VNET_DEVICE_INPUT_NEXT_IP4_INPUT;
    }
  else
    {
      next_index = VNET_DEVICE_INPUT_NEXT_ETHERNET_INPUT;
      if (PREDICT_FALSE (apif->per_interface_next_index != ~0))
	next_index = apif->per_interface_next_index;

      /* redirect if feature path enabled */
      vnet_feature_start_device_input (apif->sw_if_index, &next_index, &bt);
    }

  n_free_bufs = vec_len (apm->rx_buffers[thread_index]);
  if (PREDICT_FALSE (n_free_bufs < VLIB_FRAME_SIZE))
    {
      vec_validate (apm->rx_buffers[thread_index],
		    VLIB_FRAME_SIZE + n_free_bufs - 1);
      n_free_bufs += vlib_buffer_alloc (
	vm, &apm->rx_buffers[thread_index][n_free_bufs], VLIB_FRAME_SIZE);
      vec_set_len (apm->rx_buffers[thread_index], n_free_bufs);
    }

  rx_frame = rx_queue->next_rx_frame;
  tph = (tpacket2_hdr_t *) (block_start + rx_frame * frame_size);
  while ((tph->tp_status & TP_STATUS_USER) && (n_free_bufs > min_bufs))
    {
      vlib_buffer_t *b0 = 0, *first_b0 = 0, *prev_b0 = 0;
      u32 next0 = next_index;

      u32 n_left_to_next;
      vlib_get_next_frame (vm, node, next_index, to_next, n_left_to_next);
      while ((tph->tp_status & TP_STATUS_USER) && (n_free_bufs > min_bufs) &&
	     n_left_to_next)
	{
	  vnet_virtio_net_hdr_t *vnet_hdr = 0;
	  u32 data_len = tph->tp_snaplen;
	  u32 offset = 0;
	  u32 bi0 = 0, first_bi0 = 0;
	  u8 l4_hdr_sz = 0;

	  if (is_cksum_gso_enabled)
	    vnet_hdr =
	      (vnet_virtio_net_hdr_t *) ((u8 *) tph + tph->tp_mac -
					 sizeof (vnet_virtio_net_hdr_t));
	  while (data_len)
	    {
	      /* grab free buffer */
	      u32 last_empty_buffer =
		vec_len (apm->rx_buffers[thread_index]) - 1;
	      bi0 = apm->rx_buffers[thread_index][last_empty_buffer];
	      b0 = vlib_get_buffer (vm, bi0);
	      vec_set_len (apm->rx_buffers[thread_index], last_empty_buffer);
	      n_free_bufs--;

	      /* copy data */
	      u32 bytes_to_copy =
		data_len > n_buffer_bytes ? n_buffer_bytes : data_len;
	      u32 vlan_len = 0;
	      u32 bytes_copied = 0;
	      b0->current_data = 0;
	      /* Kernel removes VLAN headers, so reconstruct VLAN */
	      if (PREDICT_FALSE (tph->tp_status & TP_STATUS_VLAN_VALID))
		{
		  if (PREDICT_TRUE (offset == 0))
		    {
		      clib_memcpy_fast (vlib_buffer_get_current (b0),
					(u8 *) tph + tph->tp_mac,
					sizeof (ethernet_header_t));
		      ethernet_header_t *eth = vlib_buffer_get_current (b0);
		      ethernet_vlan_header_t *vlan =
			(ethernet_vlan_header_t *) (eth + 1);
		      vlan->priority_cfi_and_id =
			clib_host_to_net_u16 (tph->tp_vlan_tci);
		      vlan->type = eth->type;
		      eth->type = clib_host_to_net_u16 (ETHERNET_TYPE_VLAN);
		      vlan_len = sizeof (ethernet_vlan_header_t);
		      bytes_copied = sizeof (ethernet_header_t);
		    }
		}
	      clib_memcpy_fast (((u8 *) vlib_buffer_get_current (b0)) +
				  bytes_copied + vlan_len,
				(u8 *) tph + tph->tp_mac + offset +
				  bytes_copied,
				(bytes_to_copy - bytes_copied));

	      /* fill buffer header */
	      b0->current_length = bytes_to_copy + vlan_len;

	      if (offset == 0)
		{
		  b0->total_length_not_including_first_buffer = 0;
		  b0->flags = VLIB_BUFFER_TOTAL_LENGTH_VALID;
		  vnet_buffer (b0)->sw_if_index[VLIB_RX] = apif->sw_if_index;
		  vnet_buffer (b0)->sw_if_index[VLIB_TX] = (u32) ~0;
		  first_bi0 = bi0;
		  first_b0 = vlib_get_buffer (vm, first_bi0);

		  if (is_cksum_gso_enabled)
		    {
		      if (vnet_hdr->flags & VIRTIO_NET_HDR_F_NEEDS_CSUM)
			fill_cksum_offload (first_b0, &l4_hdr_sz, is_ip);
		      if (vnet_hdr->gso_type & (VIRTIO_NET_HDR_GSO_TCPV4 |
						VIRTIO_NET_HDR_GSO_TCPV6))
			fill_gso_offload (first_b0, vnet_hdr->gso_size,
					  l4_hdr_sz);
		    }
		}
	      else
		buffer_add_to_chain (b0, first_b0, prev_b0, bi0);

	      prev_b0 = b0;
	      offset += bytes_to_copy;
	      data_len -= bytes_to_copy;
	    }
	  n_rx_packets++;
	  n_rx_bytes += tph->tp_snaplen;
	  to_next[0] = first_bi0;
	  to_next += 1;
	  n_left_to_next--;

	  /* drop partial packets */
	  if (PREDICT_FALSE (tph->tp_len != tph->tp_snaplen))
	    {
	      next0 = VNET_DEVICE_INPUT_NEXT_DROP;
	      first_b0->error =
		node->errors[AF_PACKET_INPUT_ERROR_PARTIAL_PKT];
	    }
	  else
	    {
	      if (PREDICT_FALSE (is_ip))
		{
		  switch (first_b0->data[0] & 0xf0)
		    {
		    case 0x40:
		      next0 = VNET_DEVICE_INPUT_NEXT_IP4_INPUT;
		      break;
		    case 0x60:
		      next0 = VNET_DEVICE_INPUT_NEXT_IP6_INPUT;
		      break;
		    default:
		      next0 = VNET_DEVICE_INPUT_NEXT_DROP;
		      break;
		    }
		  if (PREDICT_FALSE (apif->per_interface_next_index != ~0))
		    next0 = apif->per_interface_next_index;
		}
	      else
		{
		  next0 = VNET_DEVICE_INPUT_NEXT_ETHERNET_INPUT;
		  if (PREDICT_FALSE (apif->per_interface_next_index != ~0))
		    next0 = apif->per_interface_next_index;
		  /* copy feature arc data from template */
		  first_b0->current_config_index = bt.current_config_index;
		  vnet_buffer (first_b0)->feature_arc_index =
		    vnet_buffer (&bt)->feature_arc_index;
		}
	    }

	  /* trace */
	  if (PREDICT_FALSE (n_trace > 0 &&
			     vlib_trace_buffer (vm, node, next0, first_b0,
						/* follow_chain */ 0)))
	    {
	      af_packet_input_trace_t *tr;
	      vlib_set_trace_count (vm, node, --n_trace);
	      tr = vlib_add_trace (vm, node, first_b0, sizeof (*tr));
	      tr->is_v3 = 0;
	      tr->next_index = next0;
	      tr->hw_if_index = apif->hw_if_index;
	      tr->queue_id = queue_id;
	      clib_memcpy_fast (&tr->tph2, tph, sizeof (struct tpacket2_hdr));
	      if (is_cksum_gso_enabled)
		clib_memcpy_fast (&tr->vnet_hdr, vnet_hdr,
				  sizeof (vnet_virtio_net_hdr_t));
	      else
		clib_memset_u8 (&tr->vnet_hdr, 0,
				sizeof (vnet_virtio_net_hdr_t));
	    }

	  /* enque and take next packet */
	  vlib_validate_buffer_enqueue_x1 (vm, node, next_index, to_next,
					   n_left_to_next, first_bi0, next0);

	  /* next packet */
	  tph->tp_status = TP_STATUS_KERNEL;
	  rx_frame = (rx_frame + 1) % frame_num;
	  tph = (struct tpacket2_hdr *) (block_start + rx_frame * frame_size);
	}

      vlib_put_next_frame (vm, node, next_index, n_left_to_next);
    }

  rx_queue->next_rx_frame = rx_frame;

  vlib_increment_combined_counter (
    vnet_get_main ()->interface_main.combined_sw_if_counters +
      VNET_INTERFACE_COUNTER_RX,
    vlib_get_thread_index (), apif->hw_if_index, n_rx_packets, n_rx_bytes);

  vnet_device_increment_rx_packets (thread_index, n_rx_packets);
  return n_rx_packets;
}

always_inline uword
af_packet_device_input_fn (vlib_main_t *vm, vlib_node_runtime_t *node,
			   vlib_frame_t *frame, af_packet_if_t *apif,
			   u16 queue_id, u8 is_cksum_gso_enabled)

{
  if (apif->version == TPACKET_V3)
    return af_packet_v3_device_input_fn (vm, node, frame, apif, queue_id,
					 is_cksum_gso_enabled);
  else
    return af_packet_v2_device_input_fn (vm, node, frame, apif, queue_id,
					 is_cksum_gso_enabled);
}

VLIB_NODE_FN (af_packet_input_node) (vlib_main_t * vm,
				     vlib_node_runtime_t * node,
				     vlib_frame_t * frame)
{
  u32 n_rx_packets = 0;
  af_packet_main_t *apm = &af_packet_main;
  vnet_hw_if_rxq_poll_vector_t *pv;
  pv = vnet_hw_if_get_rxq_poll_vector (vm, node);
  for (int i = 0; i < vec_len (pv); i++)
    {
      af_packet_if_t *apif;
      apif = vec_elt_at_index (apm->interfaces, pv[i].dev_instance);
      if (apif->is_admin_up)
	{
	  if (apif->is_cksum_gso_enabled)
	    n_rx_packets += af_packet_device_input_fn (vm, node, frame, apif,
						       pv[i].queue_id, 1);
	  else
	    n_rx_packets += af_packet_device_input_fn (vm, node, frame, apif,
						       pv[i].queue_id, 0);
	}
    }
  return n_rx_packets;
}

VLIB_REGISTER_NODE (af_packet_input_node) = {
  .name = "af-packet-input",
  .flags = VLIB_NODE_FLAG_TRACE_SUPPORTED,
  .sibling_of = "device-input",
  .format_trace = format_af_packet_input_trace,
  .type = VLIB_NODE_TYPE_INPUT,
  .state = VLIB_NODE_STATE_INTERRUPT,
  .n_errors = AF_PACKET_INPUT_N_ERROR,
  .error_strings = af_packet_input_error_strings,
};


/*
 * fd.io coding-style-patch-verification: ON
 *
 * Local Variables:
 * eval: (c-set-style "gnu")
 * End:
 */