summaryrefslogtreecommitdiffstats
path: root/src/vnet/devices/virtio/device.c
blob: fe943a9b6d7830c0ff377970a9e3e546ecd024b4 (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
852
853
854
855
856
857
858
859
860
/*
 *------------------------------------------------------------------
 * 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 <sys/types.h>
#include <sys/stat.h>
#include <fcntl.h>

#include <vlib/vlib.h>
#include <vlib/unix/unix.h>
#include <vnet/vnet.h>
#include <vnet/ethernet/ethernet.h>
#include <vnet/gso/gro_func.h>
#include <vnet/gso/hdr_offset_parser.h>
#include <vnet/ip/ip4_packet.h>
#include <vnet/ip/ip6_packet.h>
#include <vnet/tcp/tcp_packet.h>
#include <vnet/udp/udp_packet.h>
#include <vnet/devices/virtio/virtio.h>

#define foreach_virtio_tx_func_error	       \
_(NO_FREE_SLOTS, "no free tx slots")           \
_(TRUNC_PACKET, "packet > buffer size -- truncated in tx ring") \
_(PENDING_MSGS, "pending msgs in tx ring") \
_(INDIRECT_DESC_ALLOC_FAILED, "indirect descriptor allocation failed - packet drop") \
_(OUT_OF_ORDER, "out-of-order buffers in used ring") \
_(GSO_PACKET_DROP, "gso disabled on itf  -- gso packet drop") \
_(CSUM_OFFLOAD_PACKET_DROP, "checksum offload disabled on itf -- csum offload packet drop")

typedef enum
{
#define _(f,s) VIRTIO_TX_ERROR_##f,
  foreach_virtio_tx_func_error
#undef _
    VIRTIO_TX_N_ERROR,
} virtio_tx_func_error_t;

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

static u8 *
format_virtio_device (u8 * s, va_list * args)
{
  u32 dev_instance = va_arg (*args, u32);
  int verbose = va_arg (*args, int);
  u32 indent = format_get_indent (s);

  s = format (s, "VIRTIO interface");
  if (verbose)
    {
      s = format (s, "\n%U instance %u", format_white_space, indent + 2,
		  dev_instance);
    }
  return s;
}

typedef struct
{
  u32 buffer_index;
  u32 sw_if_index;
  generic_header_offset_t gho;
  vlib_buffer_t buffer;
} virtio_tx_trace_t;

static u8 *
format_virtio_tx_trace (u8 * s, va_list * va)
{
  CLIB_UNUSED (vlib_main_t * vm) = va_arg (*va, vlib_main_t *);
  CLIB_UNUSED (vlib_node_t * node) = va_arg (*va, vlib_node_t *);
  virtio_tx_trace_t *t = va_arg (*va, virtio_tx_trace_t *);
  u32 indent = format_get_indent (s);

  s = format (s, "%Ubuffer 0x%x: %U\n",
	      format_white_space, indent,
	      t->buffer_index, format_vnet_buffer, &t->buffer);
  s =
    format (s, "%U%U\n", format_white_space, indent,
	    format_generic_header_offset, &t->gho);
  s =
    format (s, "%U%U", format_white_space, indent,
	    format_ethernet_header_with_length, t->buffer.pre_data,
	    sizeof (t->buffer.pre_data));
  return s;
}

static_always_inline void
virtio_tx_trace (vlib_main_t * vm, vlib_node_runtime_t * node,
		 virtio_if_type_t type, vlib_buffer_t * b0, u32 bi)
{
  virtio_tx_trace_t *t;
  t = vlib_add_trace (vm, node, b0, sizeof (t[0]));
  t->sw_if_index = vnet_buffer (b0)->sw_if_index[VLIB_TX];
  t->buffer_index = bi;
  if (type == VIRTIO_IF_TYPE_TUN)
    {
      int is_ip4 = 0, is_ip6 = 0;

      switch (((u8 *) vlib_buffer_get_current (b0))[0] & 0xf0)
	{
	case 0x40:
	  is_ip4 = 1;
	  break;
	case 0x60:
	  is_ip6 = 1;
	  break;
	default:
	  break;
	}
      vnet_generic_header_offset_parser (b0, &t->gho, 0, is_ip4, is_ip6);
    }
  else
    vnet_generic_header_offset_parser (b0, &t->gho, 1,
				       b0->flags &
				       VNET_BUFFER_F_IS_IP4,
				       b0->flags & VNET_BUFFER_F_IS_IP6);

  clib_memcpy_fast (&t->buffer, b0, sizeof (*b0) - sizeof (b0->pre_data));
  clib_memcpy_fast (t->buffer.pre_data, vlib_buffer_get_current (b0),
		    sizeof (t->buffer.pre_data));
}

static_always_inline void
virtio_interface_drop_inline (vlib_main_t * vm, uword node_index,
			      u32 * buffers, u16 n,
			      virtio_tx_func_error_t error)
{
  vlib_error_count (vm, node_index, error, n);
  vlib_buffer_free (vm, buffers, n);
}

static_always_inline void
virtio_memset_ring_u32 (u32 * ring, u32 start, u32 ring_size, u32 n_buffers)
{
  ASSERT (n_buffers <= ring_size);

  if (PREDICT_TRUE (start + n_buffers <= ring_size))
    {
      clib_memset_u32 (ring + start, ~0, n_buffers);
    }
  else
    {
      clib_memset_u32 (ring + start, ~0, ring_size - start);
      clib_memset_u32 (ring, ~0, n_buffers - (ring_size - start));
    }
}

static_always_inline void
virtio_free_used_device_desc (vlib_main_t * vm, virtio_vring_t * vring,
			      uword node_index)
{
  u16 used = vring->desc_in_use;
  u16 sz = vring->size;
  u16 mask = sz - 1;
  u16 last = vring->last_used_idx;
  u16 n_left = vring->used->idx - last;
  u16 out_of_order_count = 0;

  if (n_left == 0)
    return;

  while (n_left)
    {
      vring_used_elem_t *e = &vring->used->ring[last & mask];
      u16 slot, n_buffers;
      slot = n_buffers = e->id;

      while (e->id == (n_buffers & mask))
	{
	  n_left--;
	  last++;
	  n_buffers++;
	  vring_desc_t *d = &vring->desc[e->id];
	  u16 next;
	  while (d->flags & VRING_DESC_F_NEXT)
	    {
	      n_buffers++;
	      next = d->next;
	      d = &vring->desc[next];
	    }
	  if (n_left == 0)
	    break;
	  e = &vring->used->ring[last & mask];
	}
      vlib_buffer_free_from_ring (vm, vring->buffers, slot,
				  sz, (n_buffers - slot));
      virtio_memset_ring_u32 (vring->buffers, slot, sz, (n_buffers - slot));
      used -= (n_buffers - slot);

      if (n_left > 0)
	{
	  vlib_buffer_free (vm, &vring->buffers[e->id], 1);
	  vring->buffers[e->id] = ~0;
	  used--;
	  last++;
	  n_left--;
	  out_of_order_count++;
	  vring->flags |= VRING_TX_OUT_OF_ORDER;
	}
    }

  /*
   * Some vhost-backends give buffers back in out-of-order fashion in used ring.
   * It impacts the overall virtio-performance.
   */
  if (out_of_order_count)
    vlib_error_count (vm, node_index, VIRTIO_TX_ERROR_OUT_OF_ORDER,
		      out_of_order_count);

  vring->desc_in_use = used;
  vring->last_used_idx = last;
}

static_always_inline void
set_checksum_offsets (vlib_buffer_t * b, virtio_net_hdr_v1_t * hdr, int is_l2)
{
  if (b->flags & VNET_BUFFER_F_IS_IP4)
    {
      ip4_header_t *ip4;
      generic_header_offset_t gho = { 0 };
      vnet_generic_header_offset_parser (b, &gho, is_l2, 1 /* ip4 */ ,
					 0 /* ip6 */ );
      hdr->flags = VIRTIO_NET_HDR_F_NEEDS_CSUM;
      hdr->csum_start = gho.l4_hdr_offset;	// 0x22;
      if (b->flags & VNET_BUFFER_F_OFFLOAD_TCP_CKSUM)
	{
	  hdr->csum_offset = STRUCT_OFFSET_OF (tcp_header_t, checksum);
	}
      else if (b->flags & VNET_BUFFER_F_OFFLOAD_UDP_CKSUM)
	{
	  hdr->csum_offset = STRUCT_OFFSET_OF (udp_header_t, checksum);
	}

      /*
       * virtio devices do not support IP4 checksum offload. So driver takes care
       * of it while doing tx.
       */
      ip4 =
	(ip4_header_t *) (vlib_buffer_get_current (b) + gho.l3_hdr_offset);
      if (b->flags & VNET_BUFFER_F_OFFLOAD_IP_CKSUM)
	ip4->checksum = ip4_header_checksum (ip4);
    }
  else if (b->flags & VNET_BUFFER_F_IS_IP6)
    {
      generic_header_offset_t gho = { 0 };
      vnet_generic_header_offset_parser (b, &gho, is_l2, 0 /* ip4 */ ,
					 1 /* ip6 */ );
      hdr->flags = VIRTIO_NET_HDR_F_NEEDS_CSUM;
      hdr->csum_start = gho.l4_hdr_offset;	// 0x36;
      if (b->flags & VNET_BUFFER_F_OFFLOAD_TCP_CKSUM)
	{
	  hdr->csum_offset = STRUCT_OFFSET_OF (tcp_header_t, checksum);
	}
      else if (b->flags & VNET_BUFFER_F_OFFLOAD_UDP_CKSUM)
	{
	  hdr->csum_offset = STRUCT_OFFSET_OF (udp_header_t, checksum);
	}
    }
}

static_always_inline void
set_gso_offsets (vlib_buffer_t * b, virtio_net_hdr_v1_t * hdr, int is_l2)
{
  if (b->flags & VNET_BUFFER_F_IS_IP4)
    {
      ip4_header_t *ip4;
      generic_header_offset_t gho = { 0 };
      vnet_generic_header_offset_parser (b, &gho, is_l2, 1 /* ip4 */ ,
					 0 /* ip6 */ );
      hdr->gso_type = VIRTIO_NET_HDR_GSO_TCPV4;
      hdr->gso_size = vnet_buffer2 (b)->gso_size;
      hdr->hdr_len = gho.hdr_sz;
      hdr->flags = VIRTIO_NET_HDR_F_NEEDS_CSUM;
      hdr->csum_start = gho.l4_hdr_offset;	// 0x22;
      hdr->csum_offset = STRUCT_OFFSET_OF (tcp_header_t, checksum);
      ip4 =
	(ip4_header_t *) (vlib_buffer_get_current (b) + gho.l3_hdr_offset);
      /*
       * virtio devices do not support IP4 checksum offload. So driver takes care
       * of it while doing tx.
       */
      if (b->flags & VNET_BUFFER_F_OFFLOAD_IP_CKSUM)
	ip4->checksum = ip4_header_checksum (ip4);
    }
  else if (b->flags & VNET_BUFFER_F_IS_IP6)
    {
      generic_header_offset_t gho = { 0 };
      vnet_generic_header_offset_parser (b, &gho, is_l2, 0 /* ip4 */ ,
					 1 /* ip6 */ );
      hdr->gso_type = VIRTIO_NET_HDR_GSO_TCPV6;
      hdr->gso_size = vnet_buffer2 (b)->gso_size;
      hdr->hdr_len = gho.hdr_sz;
      hdr->flags = VIRTIO_NET_HDR_F_NEEDS_CSUM;
      hdr->csum_start = gho.l4_hdr_offset;	// 0x36;
      hdr->csum_offset = STRUCT_OFFSET_OF (tcp_header_t, checksum);
    }
}

static_always_inline u16
add_buffer_to_slot (vlib_main_t * vm, vlib_node_runtime_t * node,
		    virtio_if_t * vif, virtio_if_type_t type,
		    virtio_vring_t * vring, u32 bi, u16 free_desc_count,
		    u16 avail, u16 next, u16 mask, int do_gso,
		    int csum_offload)
{
  u16 n_added = 0;
  int hdr_sz = vif->virtio_net_hdr_sz;
  vring_desc_t *d;
  d = &vring->desc[next];
  vlib_buffer_t *b = vlib_get_buffer (vm, bi);
  virtio_net_hdr_v1_t *hdr = vlib_buffer_get_current (b) - hdr_sz;
  int is_l2 = (type & (VIRTIO_IF_TYPE_TAP | VIRTIO_IF_TYPE_PCI));

  clib_memset (hdr, 0, hdr_sz);

  if (b->flags & VNET_BUFFER_F_GSO)
    {
      if (do_gso)
	set_gso_offsets (b, hdr, is_l2);
      else
	{
	  virtio_interface_drop_inline (vm, node->node_index, &bi, 1,
					VIRTIO_TX_ERROR_GSO_PACKET_DROP);
	  return n_added;
	}
    }
  else if (b->flags & (VNET_BUFFER_F_OFFLOAD_TCP_CKSUM |
		       VNET_BUFFER_F_OFFLOAD_UDP_CKSUM))
    {
      if (csum_offload)
	set_checksum_offsets (b, hdr, is_l2);
      else
	{
	  virtio_interface_drop_inline (vm, node->node_index, &bi, 1,
					VIRTIO_TX_ERROR_CSUM_OFFLOAD_PACKET_DROP);
	  return n_added;
	}
    }

  if (PREDICT_FALSE (b->flags & VLIB_BUFFER_IS_TRACED))
    {
      virtio_tx_trace (vm, node, type, b, bi);
    }

  if (PREDICT_TRUE ((b->flags & VLIB_BUFFER_NEXT_PRESENT) == 0))
    {
      d->addr =
	((type == VIRTIO_IF_TYPE_PCI) ? vlib_buffer_get_current_pa (vm,
								    b) :
	 pointer_to_uword (vlib_buffer_get_current (b))) - hdr_sz;
      d->len = b->current_length + hdr_sz;
      d->flags = 0;
    }
  else if (vif->features & VIRTIO_FEATURE (VIRTIO_RING_F_INDIRECT_DESC))
    {
      /*
       * We are using single vlib_buffer_t for indirect descriptor(s)
       * chain. Single descriptor is 16 bytes and vlib_buffer_t
       * has 2048 bytes space. So maximum long chain can have 128
       * (=2048/16) indirect descriptors.
       * It can easily support 65535 bytes of Jumbo frames with
       * each data buffer size of 512 bytes minimum.
       */
      u32 indirect_buffer = 0;
      if (PREDICT_FALSE (vlib_buffer_alloc (vm, &indirect_buffer, 1) == 0))
	{
	  virtio_interface_drop_inline (vm, node->node_index, &bi, 1,
					VIRTIO_TX_ERROR_INDIRECT_DESC_ALLOC_FAILED);
	  return n_added;
	}

      vlib_buffer_t *indirect_desc = vlib_get_buffer (vm, indirect_buffer);
      indirect_desc->current_data = 0;
      indirect_desc->flags |= VLIB_BUFFER_NEXT_PRESENT;
      indirect_desc->next_buffer = bi;
      bi = indirect_buffer;

      vring_desc_t *id =
	(vring_desc_t *) vlib_buffer_get_current (indirect_desc);
      u32 count = 1;
      if (type == VIRTIO_IF_TYPE_PCI)
	{
	  d->addr = vlib_physmem_get_pa (vm, id);
	  id->addr = vlib_buffer_get_current_pa (vm, b) - hdr_sz;

	  /*
	   * If VIRTIO_F_ANY_LAYOUT is not negotiated, then virtio_net_hdr
	   * should be presented in separate descriptor and data will start
	   * from next descriptor.
	   */
	  if (PREDICT_TRUE
	      (vif->features & VIRTIO_FEATURE (VIRTIO_F_ANY_LAYOUT)))
	    id->len = b->current_length + hdr_sz;
	  else
	    {
	      id->len = hdr_sz;
	      id->flags = VRING_DESC_F_NEXT;
	      id->next = count;
	      count++;
	      id++;
	      id->addr = vlib_buffer_get_current_pa (vm, b);
	      id->len = b->current_length;
	    }
	  while (b->flags & VLIB_BUFFER_NEXT_PRESENT)
	    {
	      id->flags = VRING_DESC_F_NEXT;
	      id->next = count;
	      count++;
	      id++;
	      b = vlib_get_buffer (vm, b->next_buffer);
	      id->addr = vlib_buffer_get_current_pa (vm, b);
	      id->len = b->current_length;
	    }
	}
      else			/* VIRTIO_IF_TYPE_[TAP | TUN] */
	{
	  d->addr = pointer_to_uword (id);
	  /* first buffer in chain */
	  id->addr = pointer_to_uword (vlib_buffer_get_current (b)) - hdr_sz;
	  id->len = b->current_length + hdr_sz;

	  while (b->flags & VLIB_BUFFER_NEXT_PRESENT)
	    {
	      id->flags = VRING_DESC_F_NEXT;
	      id->next = count;
	      count++;
	      id++;
	      b = vlib_get_buffer (vm, b->next_buffer);
	      id->addr = pointer_to_uword (vlib_buffer_get_current (b));
	      id->len = b->current_length;
	    }
	}
      id->flags = 0;
      id->next = 0;
      d->len = count * sizeof (vring_desc_t);
      d->flags = VRING_DESC_F_INDIRECT;
    }
  else if (type == VIRTIO_IF_TYPE_PCI)
    {
      u16 count = next;
      vlib_buffer_t *b_temp = b;
      u16 n_buffers_in_chain = 1;

      /*
       * Check the length of the chain for the required number of
       * descriptors. Return from here, retry to get more descriptors,
       * if chain length is greater than available descriptors.
       */
      while (b_temp->flags & VLIB_BUFFER_NEXT_PRESENT)
	{
	  n_buffers_in_chain++;
	  b_temp = vlib_get_buffer (vm, b_temp->next_buffer);
	}

      if (n_buffers_in_chain > free_desc_count)
	return n_buffers_in_chain;

      d->addr = vlib_buffer_get_current_pa (vm, b) - hdr_sz;
      d->len = b->current_length + hdr_sz;

      while (b->flags & VLIB_BUFFER_NEXT_PRESENT)
	{
	  d->flags = VRING_DESC_F_NEXT;
	  vring->buffers[count] = bi;
	  b->flags &=
	    ~(VLIB_BUFFER_NEXT_PRESENT | VLIB_BUFFER_TOTAL_LENGTH_VALID);
	  bi = b->next_buffer;
	  b->next_buffer = 0;
	  n_added++;
	  count = (count + 1) & mask;
	  d->next = count;
	  d = &vring->desc[count];
	  b = vlib_get_buffer (vm, bi);
	  d->addr = vlib_buffer_get_current_pa (vm, b);
	  d->len = b->current_length;
	}
      d->flags = 0;
      vring->buffers[count] = bi;
      vring->avail->ring[avail & mask] = next;
      n_added++;
      return n_added;
    }
  else
    {
      ASSERT (0);
    }
  vring->buffers[next] = bi;
  vring->avail->ring[avail & mask] = next;
  n_added++;
  return n_added;
}

static_always_inline void
virtio_find_free_desc (virtio_vring_t * vring, u16 size, u16 mask,
		       u16 req, u16 next, u32 * first_free_desc_index,
		       u16 * free_desc_count)
{
  u16 start = 0;
  /* next is used as hint: from where to start looking */
  for (u16 i = 0; i < size; i++, next++)
    {
      if (vring->buffers[next & mask] == ~0)
	{
	  if (*first_free_desc_index == ~0)
	    {
	      *first_free_desc_index = (next & mask);
	      start = i;
	      (*free_desc_count)++;
	      req--;
	      if (req == 0)
		break;
	    }
	  else
	    {
	      if (start + *free_desc_count == i)
		{
		  (*free_desc_count)++;
		  req--;
		  if (req == 0)
		    break;
		}
	      else
		break;
	    }
	}
    }
}

static_always_inline u16
virtio_interface_tx_gso_inline (vlib_main_t * vm, vlib_node_runtime_t * node,
				virtio_if_t * vif,
				virtio_if_type_t type, virtio_vring_t * vring,
				u32 * buffers, u16 n_left, int do_gso,
				int csum_offload)
{
  u16 used, next, avail, n_buffers = 0, n_buffers_left = 0;
  u16 sz = vring->size;
  u16 mask = sz - 1;
  u16 n_vectors = n_left;
  u16 retry_count = 2;

retry:
  /* free consumed buffers */
  virtio_free_used_device_desc (vm, vring, node->node_index);

  used = vring->desc_in_use;
  next = vring->desc_next;
  avail = vring->avail->idx;

  u16 free_desc_count = 0;

  if (PREDICT_FALSE (vring->flags & VRING_TX_OUT_OF_ORDER))
    {
      u32 first_free_desc_index = ~0;

      virtio_find_free_desc (vring, sz, mask, n_left, next,
			     &first_free_desc_index, &free_desc_count);

      if (free_desc_count)
	next = first_free_desc_index;
    }
  else
    free_desc_count = sz - used;

  if (vif->packet_buffering)
    {
      n_buffers = n_buffers_left = virtio_vring_n_buffers (vring->buffering);

      while (n_buffers_left && free_desc_count)
	{
	  u16 n_added = 0;

	  u32 bi = virtio_vring_buffering_read_from_front (vring->buffering);
	  if (bi == ~0)
	    break;

	  n_added =
	    add_buffer_to_slot (vm, node, vif, type, vring, bi,
				free_desc_count, avail, next, mask, do_gso,
				csum_offload);
	  if (PREDICT_FALSE (n_added == 0))
	    {
	      n_buffers_left--;
	      continue;
	    }
	  else if (PREDICT_FALSE (n_added > free_desc_count))
	    break;

	  avail++;
	  next = (next + n_added) & mask;
	  used += n_added;
	  n_buffers_left--;
	  free_desc_count -= n_added;
	}
    }

  while (n_left && free_desc_count)
    {
      u16 n_added = 0;

      n_added =
	add_buffer_to_slot (vm, node, vif, type, vring, buffers[0],
			    free_desc_count, avail, next, mask, do_gso,
			    csum_offload);

      if (PREDICT_FALSE (n_added == 0))
	{
	  buffers++;
	  n_left--;
	  continue;
	}
      else if (PREDICT_FALSE (n_added > free_desc_count))
	break;

      avail++;
      next = (next + n_added) & mask;
      used += n_added;
      buffers++;
      n_left--;
      free_desc_count -= n_added;
    }

  if (n_left != n_vectors || n_buffers != n_buffers_left)
    {
      CLIB_MEMORY_STORE_BARRIER ();
      vring->avail->idx = avail;
      vring->desc_next = next;
      vring->desc_in_use = used;
      if ((vring->used->flags & VRING_USED_F_NO_NOTIFY) == 0)
	virtio_kick (vm, vring, vif);
    }

  if (n_left && retry_count--)
    goto retry;

  return n_left;
}

static_always_inline u16
virtio_interface_tx_inline (vlib_main_t * vm, vlib_node_runtime_t * node,
			    virtio_if_t * vif,
			    virtio_vring_t * vring, virtio_if_type_t type,
			    u32 * buffers, u16 n_left)
{
  vnet_main_t *vnm = vnet_get_main ();
  vnet_hw_interface_t *hw = vnet_get_hw_interface (vnm, vif->hw_if_index);

  if (hw->flags & VNET_HW_INTERFACE_FLAG_SUPPORTS_GSO)
    return virtio_interface_tx_gso_inline (vm, node, vif, type, vring,
					   buffers, n_left, 1 /* do_gso */ ,
					   1 /* checksum offload */ );
  else if (hw->flags & VNET_HW_INTERFACE_FLAG_SUPPORTS_TX_L4_CKSUM_OFFLOAD)
    return virtio_interface_tx_gso_inline (vm, node, vif, type, vring,
					   buffers, n_left,
					   0 /* no do_gso */ ,
					   1 /* checksum offload */ );
  else
    return virtio_interface_tx_gso_inline (vm, node, vif, type, vring,
					   buffers, n_left,
					   0 /* no do_gso */ ,
					   0 /* no checksum offload */ );
}

VNET_DEVICE_CLASS_TX_FN (virtio_device_class) (vlib_main_t * vm,
					       vlib_node_runtime_t * node,
					       vlib_frame_t * frame)
{
  virtio_main_t *nm = &virtio_main;
  vnet_interface_output_runtime_t *rund = (void *) node->runtime_data;
  virtio_if_t *vif = pool_elt_at_index (nm->interfaces, rund->dev_instance);
  u16 qid = vm->thread_index % vif->num_txqs;
  virtio_vring_t *vring = vec_elt_at_index (vif->txq_vrings, qid);
  u16 n_left = frame->n_vectors;
  u32 *buffers = vlib_frame_vector_args (frame);
  u32 to[GRO_TO_VECTOR_SIZE (n_left)];

  clib_spinlock_lock_if_init (&vring->lockp);

  if ((vring->used->flags & VRING_USED_F_NO_NOTIFY) == 0 &&
      (vring->last_kick_avail_idx != vring->avail->idx))
    virtio_kick (vm, vring, vif);

  if (vif->packet_coalesce)
    {
      n_left = vnet_gro_inline (vm, vring->flow_table, buffers, n_left, to);
      buffers = to;
    }

  if (vif->type == VIRTIO_IF_TYPE_TAP)
    n_left = virtio_interface_tx_inline (vm, node, vif, vring,
					 VIRTIO_IF_TYPE_TAP, buffers, n_left);
  else if (vif->type == VIRTIO_IF_TYPE_PCI)
    n_left = virtio_interface_tx_inline (vm, node, vif, vring,
					 VIRTIO_IF_TYPE_PCI, buffers, n_left);
  else if (vif->type == VIRTIO_IF_TYPE_TUN)
    n_left = virtio_interface_tx_inline (vm, node, vif, vring,
					 VIRTIO_IF_TYPE_TUN, buffers, n_left);
  else
    ASSERT (0);

  if (vif->packet_buffering && n_left)
    {
      u16 n_buffered =
	virtio_vring_buffering_store_packets (vring->buffering, buffers,
					      n_left);
      buffers += n_buffered;
      n_left -= n_buffered;
    }
  if (n_left)
    virtio_interface_drop_inline (vm, node->node_index,
				  buffers + frame->n_vectors - n_left, n_left,
				  VIRTIO_TX_ERROR_NO_FREE_SLOTS);

  clib_spinlock_unlock_if_init (&vring->lockp);

  return frame->n_vectors - n_left;
}

static void
virtio_set_interface_next_node (vnet_main_t * vnm, u32 hw_if_index,
				u32 node_index)
{
  virtio_main_t *apm = &virtio_main;
  vnet_hw_interface_t *hw = vnet_get_hw_interface (vnm, hw_if_index);
  virtio_if_t *vif = pool_elt_at_index (apm->interfaces, hw->dev_instance);

  /* Shut off redirection */
  if (node_index == ~0)
    {
      vif->per_interface_next_index = node_index;
      return;
    }

  vif->per_interface_next_index =
    vlib_node_add_next (vlib_get_main (), virtio_input_node.index,
			node_index);
}

static void
virtio_clear_hw_interface_counters (u32 instance)
{
  /* Nothing for now */
}

static clib_error_t *
virtio_interface_rx_mode_change (vnet_main_t * vnm, u32 hw_if_index, u32 qid,
				 vnet_hw_if_rx_mode mode)
{
  vlib_main_t *vm = vnm->vlib_main;
  virtio_main_t *mm = &virtio_main;
  vnet_hw_interface_t *hw = vnet_get_hw_interface (vnm, hw_if_index);
  virtio_if_t *vif = pool_elt_at_index (mm->interfaces, hw->dev_instance);
  virtio_vring_t *rx_vring = vec_elt_at_index (vif->rxq_vrings, qid);

  if (vif->type == VIRTIO_IF_TYPE_PCI && !(vif->support_int_mode))
    {
      rx_vring->avail->flags |= VRING_AVAIL_F_NO_INTERRUPT;
      return clib_error_return (0, "interrupt mode is not supported");
    }

  if (mode == VNET_HW_IF_RX_MODE_POLLING)
    {
      if (vif->packet_coalesce || vif->packet_buffering)
	{
	  if (mm->interrupt_queues_count > 0)
	    mm->interrupt_queues_count--;
	  if (mm->interrupt_queues_count == 0)
	    vlib_process_signal_event (vm,
				       virtio_send_interrupt_node.index,
				       VIRTIO_EVENT_STOP_TIMER, 0);
	}
      rx_vring->avail->flags |= VRING_AVAIL_F_NO_INTERRUPT;
    }
  else
    {
      if (vif->packet_coalesce || vif->packet_buffering)
	{
	  mm->interrupt_queues_count++;
	  if (mm->interrupt_queues_count == 1)
	    vlib_process_signal_event (vm,
				       virtio_send_interrupt_node.index,
				       VIRTIO_EVENT_START_TIMER, 0);
	}
      rx_vring->avail->flags &= ~VRING_AVAIL_F_NO_INTERRUPT;
    }

  rx_vring->mode = mode;

  return 0;
}

static clib_error_t *
virtio_interface_admin_up_down (vnet_main_t * vnm, u32 hw_if_index, u32 flags)
{
  virtio_main_t *mm = &virtio_main;
  vnet_hw_interface_t *hw = vnet_get_hw_interface (vnm, hw_if_index);
  virtio_if_t *vif = pool_elt_at_index (mm->interfaces, hw->dev_instance);

  if (flags & VNET_SW_INTERFACE_FLAG_ADMIN_UP)
    {
      vif->flags |= VIRTIO_IF_FLAG_ADMIN_UP;
      vnet_hw_interface_set_flags (vnm, vif->hw_if_index,
				   VNET_HW_INTERFACE_FLAG_LINK_UP);
    }
  else
    {
      vif->flags &= ~VIRTIO_IF_FLAG_ADMIN_UP;
      vnet_hw_interface_set_flags (vnm, vif->hw_if_index, 0);
    }
  return 0;
}

static clib_error_t *
virtio_subif_add_del_function (vnet_main_t * vnm,
			       u32 hw_if_index,
			       struct vnet_sw_interface_t *st, int is_add)
{
  /* Nothing for now */
  return 0;
}

/* *INDENT-OFF* */
VNET_DEVICE_CLASS (virtio_device_class) = {
  .name = "virtio",
  .format_device_name = format_virtio_device_name,
  .format_device = format_virtio_device,
  .format_tx_trace = format_virtio_tx_trace,
  .tx_function_n_errors = VIRTIO_TX_N_ERROR,
  .tx_function_error_strings = virtio_tx_func_error_strings,
  .rx_redirect_to_node = virtio_set_interface_next_node,
  .clear_counters = virtio_clear_hw_interface_counters,
  .admin_up_down_function = virtio_interface_admin_up_down,
  .subif_add_del_function = virtio_subif_add_del_function,
  .rx_mode_change_function = virtio_interface_rx_mode_change,
};

/* *INDENT-ON* */

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