summaryrefslogtreecommitdiffstats
path: root/test/test_classifier.py
blob: 068561230ea0db9f7afbf1011a46e8a5d2af4b0e (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
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
#!/usr/bin/env python3

import binascii
import socket
import unittest

from framework import VppTestCase, VppTestRunner
from scapy.packet import Raw, Packet

from scapy.layers.l2 import Ether
from scapy.layers.inet import IP, UDP, TCP
from util import ppp
from template_classifier import TestClassifier, VarMask, VarMatch
from vpp_ip_route import VppIpRoute, VppRoutePath
from vpp_ip import INVALID_INDEX
from vpp_papi import VppEnum


# Tests split to different test case classes because of issue reported in
# ticket VPP-1336
class TestClassifierIP(TestClassifier):
    """Classifier IP Test Case"""

    @classmethod
    def setUpClass(cls):
        super(TestClassifierIP, cls).setUpClass()

    @classmethod
    def tearDownClass(cls):
        super(TestClassifierIP, cls).tearDownClass()

    def test_iacl_src_ip(self):
        """Source IP iACL test

        Test scenario for basic IP ACL with source IP
            - Create IPv4 stream for pg0 -> pg1 interface.
            - Create iACL with source IP address.
            - Send and verify received packets on pg1 interface.
        """

        # Basic iACL testing with source IP
        pkts = self.create_stream(self.pg0, self.pg1, self.pg_if_packet_sizes)
        self.pg0.add_stream(pkts)

        key = "ip_src"
        self.create_classify_table(key, self.build_ip_mask(src_ip="ffffffff"))
        self.create_classify_session(
            self.acl_tbl_idx.get(key), self.build_ip_match(src_ip=self.pg0.remote_ip4)
        )
        self.input_acl_set_interface(self.pg0, self.acl_tbl_idx.get(key))
        self.acl_active_table = key

        self.pg_enable_capture(self.pg_interfaces)
        self.pg_start()

        pkts = self.pg1.get_capture(len(pkts))
        self.verify_capture(self.pg1, pkts)
        self.pg0.assert_nothing_captured(remark="packets forwarded")
        self.pg2.assert_nothing_captured(remark="packets forwarded")
        self.pg3.assert_nothing_captured(remark="packets forwarded")

    def test_iacl_dst_ip(self):
        """Destination IP iACL test

        Test scenario for basic IP ACL with destination IP
            - Create IPv4 stream for pg0 -> pg1 interface.
            - Create iACL with destination IP address.
            - Send and verify received packets on pg1 interface.
        """

        # Basic iACL testing with destination IP
        pkts = self.create_stream(self.pg0, self.pg1, self.pg_if_packet_sizes)
        self.pg0.add_stream(pkts)

        key = "ip_dst"
        self.create_classify_table(key, self.build_ip_mask(dst_ip="ffffffff"))
        self.create_classify_session(
            self.acl_tbl_idx.get(key), self.build_ip_match(dst_ip=self.pg1.remote_ip4)
        )
        self.input_acl_set_interface(self.pg0, self.acl_tbl_idx.get(key))
        self.acl_active_table = key

        self.pg_enable_capture(self.pg_interfaces)
        self.pg_start()

        pkts = self.pg1.get_capture(len(pkts))
        self.verify_capture(self.pg1, pkts)
        self.pg0.assert_nothing_captured(remark="packets forwarded")
        self.pg2.assert_nothing_captured(remark="packets forwarded")
        self.pg3.assert_nothing_captured(remark="packets forwarded")

    def test_iacl_src_dst_ip(self):
        """Source and destination IP iACL test

        Test scenario for basic IP ACL with source and destination IP
            - Create IPv4 stream for pg0 -> pg1 interface.
            - Create iACL with source and destination IP addresses.
            - Send and verify received packets on pg1 interface.
        """

        # Basic iACL testing with source and destination IP
        pkts = self.create_stream(self.pg0, self.pg1, self.pg_if_packet_sizes)
        self.pg0.add_stream(pkts)

        key = "ip"
        self.create_classify_table(
            key, self.build_ip_mask(src_ip="ffffffff", dst_ip="ffffffff")
        )
        self.create_classify_session(
            self.acl_tbl_idx.get(key),
            self.build_ip_match(src_ip=self.pg0.remote_ip4, dst_ip=self.pg1.remote_ip4),
        )
        self.input_acl_set_interface(self.pg0, self.acl_tbl_idx.get(key))
        self.acl_active_table = key

        self.pg_enable_capture(self.pg_interfaces)
        self.pg_start()

        pkts = self.pg1.get_capture(len(pkts))
        self.verify_capture(self.pg1, pkts)
        self.pg0.assert_nothing_captured(remark="packets forwarded")
        self.pg2.assert_nothing_captured(remark="packets forwarded")
        self.pg3.assert_nothing_captured(remark="packets forwarded")


class TestClassifierUDP(TestClassifier):
    """Classifier UDP proto Test Case"""

    @classmethod
    def setUpClass(cls):
        super(TestClassifierUDP, cls).setUpClass()

    @classmethod
    def tearDownClass(cls):
        super(TestClassifierUDP, cls).tearDownClass()

    def test_iacl_proto_udp(self):
        """UDP protocol iACL test

        Test scenario for basic protocol ACL with UDP protocol
            - Create IPv4 stream for pg0 -> pg1 interface.
            - Create iACL with UDP IP protocol.
            - Send and verify received packets on pg1 interface.
        """

        # Basic iACL testing with UDP protocol
        pkts = self.create_stream(self.pg0, self.pg1, self.pg_if_packet_sizes)
        self.pg0.add_stream(pkts)

        key = "proto_udp"
        self.create_classify_table(key, self.build_ip_mask(proto="ff"))
        self.create_classify_session(
            self.acl_tbl_idx.get(key), self.build_ip_match(proto=socket.IPPROTO_UDP)
        )
        self.input_acl_set_interface(self.pg0, self.acl_tbl_idx.get(key))
        self.acl_active_table = key

        self.pg_enable_capture(self.pg_interfaces)
        self.pg_start()

        pkts = self.pg1.get_capture(len(pkts))
        self.verify_capture(self.pg1, pkts)
        self.pg0.assert_nothing_captured(remark="packets forwarded")
        self.pg2.assert_nothing_captured(remark="packets forwarded")
        self.pg3.assert_nothing_captured(remark="packets forwarded")

    def test_iacl_proto_udp_sport(self):
        """UDP source port iACL test

        Test scenario for basic protocol ACL with UDP and sport
            - Create IPv4 stream for pg0 -> pg1 interface.
            - Create iACL with UDP IP protocol and defined sport.
            - Send and verify received packets on pg1 interface.
        """

        # Basic iACL testing with UDP and sport
        sport = 38
        pkts = self.create_stream(
            self.pg0, self.pg1, self.pg_if_packet_sizes, UDP(sport=sport, dport=5678)
        )
        self.pg0.add_stream(pkts)

        key = "proto_udp_sport"
        self.create_classify_table(key, self.build_ip_mask(proto="ff", src_port="ffff"))
        self.create_classify_session(
            self.acl_tbl_idx.get(key),
            self.build_ip_match(proto=socket.IPPROTO_UDP, src_port=sport),
        )
        self.input_acl_set_interface(self.pg0, self.acl_tbl_idx.get(key))
        self.acl_active_table = key

        self.pg_enable_capture(self.pg_interfaces)
        self.pg_start()

        pkts = self.pg1.get_capture(len(pkts))
        self.verify_capture(self.pg1, pkts)
        self.pg0.assert_nothing_captured(remark="packets forwarded")
        self.pg2.assert_nothing_captured(remark="packets forwarded")
        self.pg3.assert_nothing_captured(remark="packets forwarded")

    def test_iacl_proto_udp_dport(self):
        """UDP destination port iACL test

        Test scenario for basic protocol ACL with UDP and dport
            - Create IPv4 stream for pg0 -> pg1 interface.
            - Create iACL with UDP IP protocol and defined dport.
            - Send and verify received packets on pg1 interface.
        """

        # Basic iACL testing with UDP and dport
        dport = 427
        pkts = self.create_stream(
            self.pg0, self.pg1, self.pg_if_packet_sizes, UDP(sport=1234, dport=dport)
        )
        self.pg0.add_stream(pkts)

        key = "proto_udp_dport"
        self.create_classify_table(key, self.build_ip_mask(proto="ff", dst_port="ffff"))
        self.create_classify_session(
            self.acl_tbl_idx.get(key),
            self.build_ip_match(proto=socket.IPPROTO_UDP, dst_port=dport),
        )
        self.input_acl_set_interface(self.pg0, self.acl_tbl_idx.get(key))
        self.acl_active_table = key

        self.pg_enable_capture(self.pg_interfaces)
        self.pg_start()

        pkts = self.pg1.get_capture(len(pkts))
        self.verify_capture(self.pg1, pkts)
        self.pg0.assert_nothing_captured(remark="packets forwarded")
        self.pg2.assert_nothing_captured(remark="packets forwarded")
        self.pg3.assert_nothing_captured(remark="packets forwarded")

    def test_iacl_proto_udp_sport_dport(self):
        """UDP source and destination ports iACL test

        Test scenario for basic protocol ACL with UDP and sport and dport
            - Create IPv4 stream for pg0 -> pg1 interface.
            - Create iACL with UDP IP protocol and defined sport and dport.
            - Send and verify received packets on pg1 interface.
        """

        # Basic iACL testing with UDP and sport and dport
        sport = 13720
        dport = 9080
        pkts = self.create_stream(
            self.pg0, self.pg1, self.pg_if_packet_sizes, UDP(sport=sport, dport=dport)
        )
        self.pg0.add_stream(pkts)

        key = "proto_udp_ports"
        self.create_classify_table(
            key, self.build_ip_mask(proto="ff", src_port="ffff", dst_port="ffff")
        )
        self.create_classify_session(
            self.acl_tbl_idx.get(key),
            self.build_ip_match(
                proto=socket.IPPROTO_UDP, src_port=sport, dst_port=dport
            ),
        )
        self.input_acl_set_interface(self.pg0, self.acl_tbl_idx.get(key))
        self.acl_active_table = key

        self.pg_enable_capture(self.pg_interfaces)
        self.pg_start()

        pkts = self.pg1.get_capture(len(pkts))
        self.verify_capture(self.pg1, pkts)
        self.pg0.assert_nothing_captured(remark="packets forwarded")
        self.pg2.assert_nothing_captured(remark="packets forwarded")
        self.pg3.assert_nothing_captured(remark="packets forwarded")


class TestClassifierTCP(TestClassifier):
    """Classifier TCP proto Test Case"""

    @classmethod
    def setUpClass(cls):
        super(TestClassifierTCP, cls).setUpClass()

    @classmethod
    def tearDownClass(cls):
        super(TestClassifierTCP, cls).tearDownClass()

    def test_iacl_proto_tcp(self):
        """TCP protocol iACL test

        Test scenario for basic protocol ACL with TCP protocol
            - Create IPv4 stream for pg0 -> pg1 interface.
            - Create iACL with TCP IP protocol.
            - Send and verify received packets on pg1 interface.
        """

        # Basic iACL testing with TCP protocol
        pkts = self.create_stream(
            self.pg0, self.pg1, self.pg_if_packet_sizes, TCP(sport=1234, dport=5678)
        )
        self.pg0.add_stream(pkts)

        key = "proto_tcp"
        self.create_classify_table(key, self.build_ip_mask(proto="ff"))
        self.create_classify_session(
            self.acl_tbl_idx.get(key), self.build_ip_match(proto=socket.IPPROTO_TCP)
        )
        self.input_acl_set_interface(self.pg0, self.acl_tbl_idx.get(key))
        self.acl_active_table = key

        self.pg_enable_capture(self.pg_interfaces)
        self.pg_start()

        pkts = self.pg1.get_capture(len(pkts))
        self.verify_capture(self.pg1, pkts, TCP)
        self.pg0.assert_nothing_captured(remark="packets forwarded")
        self.pg2.assert_nothing_captured(remark="packets forwarded")
        self.pg3.assert_nothing_captured(remark="packets forwarded")

    def test_iacl_proto_tcp_sport(self):
        """TCP source port iACL test

        Test scenario for basic protocol ACL with TCP and sport
            - Create IPv4 stream for pg0 -> pg1 interface.
            - Create iACL with TCP IP protocol and defined sport.
            - Send and verify received packets on pg1 interface.
        """

        # Basic iACL testing with TCP and sport
        sport = 38
        pkts = self.create_stream(
            self.pg0, self.pg1, self.pg_if_packet_sizes, TCP(sport=sport, dport=5678)
        )
        self.pg0.add_stream(pkts)

        key = "proto_tcp_sport"
        self.create_classify_table(key, self.build_ip_mask(proto="ff", src_port="ffff"))
        self.create_classify_session(
            self.acl_tbl_idx.get(key),
            self.build_ip_match(proto=socket.IPPROTO_TCP, src_port=sport),
        )
        self.input_acl_set_interface(self.pg0, self.acl_tbl_idx.get(key))
        self.acl_active_table = key

        self.pg_enable_capture(self.pg_interfaces)
        self.pg_start()

        pkts = self.pg1.get_capture(len(pkts))
        self.verify_capture(self.pg1, pkts, TCP)
        self.pg0.assert_nothing_captured(remark="packets forwarded")
        self.pg2.assert_nothing_captured(remark="packets forwarded")
        self.pg3.assert_nothing_captured(remark="packets forwarded")

    def test_iacl_proto_tcp_dport(self):
        """TCP destination port iACL test

        Test scenario for basic protocol ACL with TCP and dport
            - Create IPv4 stream for pg0 -> pg1 interface.
            - Create iACL with TCP IP protocol and defined dport.
            - Send and verify received packets on pg1 interface.
        """

        # Basic iACL testing with TCP and dport
        dport = 427
        pkts = self.create_stream(
            self.pg0, self.pg1, self.pg_if_packet_sizes, TCP(sport=1234, dport=dport)
        )
        self.pg0.add_stream(pkts)

        key = "proto_tcp_sport"
        self.create_classify_table(key, self.build_ip_mask(proto="ff", dst_port="ffff"))
        self.create_classify_session(
            self.acl_tbl_idx.get(key),
            self.build_ip_match(proto=socket.IPPROTO_TCP, dst_port=dport),
        )
        self.input_acl_set_interface(self.pg0, self.acl_tbl_idx.get(key))
        self.acl_active_table = key

        self.pg_enable_capture(self.pg_interfaces)
        self.pg_start()

        pkts = self.pg1.get_capture(len(pkts))
        self.verify_capture(self.pg1, pkts, TCP)
        self.pg0.assert_nothing_captured(remark="packets forwarded")
        self.pg2.assert_nothing_captured(remark="packets forwarded")
        self.pg3.assert_nothing_captured(remark="packets forwarded")

    def test_iacl_proto_tcp_sport_dport(self):
        """TCP source and destination ports iACL test

        Test scenario for basic protocol ACL with TCP and sport and dport
            - Create IPv4 stream for pg0 -> pg1 interface.
            - Create iACL with TCP IP protocol and defined sport and dport.
            - Send and verify received packets on pg1 interface.
        """

        # Basic iACL testing with TCP and sport and dport
        sport = 13720
        dport = 9080
        pkts = self.create_stream(
            self.pg0, self.pg1, self.pg_if_packet_sizes, TCP(sport=sport, dport=dport)
        )
        self.pg0.add_stream(pkts)

        key = "proto_tcp_ports"
        self.create_classify_table(
            key, self.build_ip_mask(proto="ff", src_port="ffff", dst_port="ffff")
        )
        self.create_classify_session(
            self.acl_tbl_idx.get(key),
            self.build_ip_match(
                proto=socket.IPPROTO_TCP, src_port=sport, dst_port=dport
            ),
        )
        self.input_acl_set_interface(self.pg0, self.acl_tbl_idx.get(key))
        self.acl_active_table = key

        self.pg_enable_capture(self.pg_interfaces)
        self.pg_start()

        pkts = self.pg1.get_capture(len(pkts))
        self.verify_capture(self.pg1, pkts, TCP)
        self.pg0.assert_nothing_captured(remark="packets forwarded")
        self.pg2.assert_nothing_captured(remark="packets forwarded")
        self.pg3.assert_nothing_captured(remark="packets forwarded")


class TestClassifierIPOut(TestClassifier):
    """Classifier output IP Test Case"""

    @classmethod
    def setUpClass(cls):
        super(TestClassifierIPOut, cls).setUpClass()

    @classmethod
    def tearDownClass(cls):
        super(TestClassifierIPOut, cls).tearDownClass()

    def test_acl_ip_out(self):
        """Output IP ACL test

        Test scenario for basic IP ACL with source IP
            - Create IPv4 stream for pg1 -> pg0 interface.
            - Create ACL with source IP address.
            - Send and verify received packets on pg0 interface.
        """

        # Basic oACL testing with source IP
        pkts = self.create_stream(self.pg1, self.pg0, self.pg_if_packet_sizes)
        self.pg1.add_stream(pkts)

        key = "ip_out"
        self.create_classify_table(
            key, self.build_ip_mask(src_ip="ffffffff"), data_offset=0
        )
        self.create_classify_session(
            self.acl_tbl_idx.get(key), self.build_ip_match(src_ip=self.pg1.remote_ip4)
        )
        self.output_acl_set_interface(self.pg0, self.acl_tbl_idx.get(key))
        self.acl_active_table = key

        self.pg_enable_capture(self.pg_interfaces)
        self.pg_start()

        pkts = self.pg0.get_capture(len(pkts))
        self.verify_capture(self.pg0, pkts)
        self.pg1.assert_nothing_captured(remark="packets forwarded")
        self.pg2.assert_nothing_captured(remark="packets forwarded")
        self.pg3.assert_nothing_captured(remark="packets forwarded")


class TestClassifierMAC(TestClassifier):
    """Classifier MAC Test Case"""

    @classmethod
    def setUpClass(cls):
        super(TestClassifierMAC, cls).setUpClass()

    @classmethod
    def tearDownClass(cls):
        super(TestClassifierMAC, cls).tearDownClass()

    def test_acl_mac(self):
        """MAC ACL test

        Test scenario for basic MAC ACL with source MAC
            - Create IPv4 stream for pg0 -> pg2 interface.
            - Create ACL with source MAC address.
            - Send and verify received packets on pg2 interface.
        """

        # Basic iACL testing with source MAC
        pkts = self.create_stream(self.pg0, self.pg2, self.pg_if_packet_sizes)
        self.pg0.add_stream(pkts)

        key = "mac"
        self.create_classify_table(
            key, self.build_mac_mask(src_mac="ffffffffffff"), data_offset=-14
        )
        self.create_classify_session(
            self.acl_tbl_idx.get(key), self.build_mac_match(src_mac=self.pg0.remote_mac)
        )
        self.input_acl_set_interface(self.pg0, self.acl_tbl_idx.get(key))
        self.acl_active_table = key

        self.pg_enable_capture(self.pg_interfaces)
        self.pg_start()

        pkts = self.pg2.get_capture(len(pkts))
        self.verify_capture(self.pg2, pkts)
        self.pg0.assert_nothing_captured(remark="packets forwarded")
        self.pg1.assert_nothing_captured(remark="packets forwarded")
        self.pg3.assert_nothing_captured(remark="packets forwarded")


class TestClassifierComplex(TestClassifier):
    """Large & Nested Classifiers Test Cases"""

    @classmethod
    def setUpClass(cls):
        super(TestClassifierComplex, cls).setUpClass()

    @classmethod
    def tearDownClass(cls):
        super(TestClassifierComplex, cls).tearDownClass()

    def test_iacl_large(self):
        """Large input ACL test

        Test scenario for Large ACL matching on ethernet+ip+udp headers
            - Create IPv4 stream for pg0 -> pg1 interface.
            - Create large acl matching on ethernet+ip+udp header fields
            - Send and verify received packets on pg1 interface.
        """

        # 40b offset = 80bytes - (sizeof(UDP/IP/ETH) + 4b)
        # + 4b as build_ip_ma*() func, do not match against UDP Len & Chksum
        msk = VarMask(offset=40, spec="ffff")
        mth = VarMatch(offset=40, value=0x1234, length=2)

        payload_msk = self.build_payload_mask([msk])
        payload_match = self.build_payload_match([mth])

        sport = 13720
        dport = 9080

        # 36b offset = 80bytes - (sizeof(UDP/IP/ETH))
        packet_ex = bytes.fromhex(("0" * 36) + "1234")
        pkts = self.create_stream(
            self.pg0,
            self.pg1,
            self.pg_if_packet_sizes,
            UDP(sport=sport, dport=dport),
            packet_ex,
        )
        self.pg0.add_stream(pkts)

        key = "large_in"
        self.create_classify_table(
            key,
            self.build_mac_mask(
                src_mac="ffffffffffff", dst_mac="ffffffffffff", ether_type="ffff"
            )
            + self.build_ip_mask(
                proto="ff",
                src_ip="ffffffff",
                dst_ip="ffffffff",
                src_port="ffff",
                dst_port="ffff",
            )
            + payload_msk,
            data_offset=-14,
        )

        self.create_classify_session(
            self.acl_tbl_idx.get(key),
            self.build_mac_match(
                src_mac=self.pg0.remote_mac,
                dst_mac=self.pg0.local_mac,
                # ipv4 next header
                ether_type="0800",
            )
            + self.build_ip_match(
                proto=socket.IPPROTO_UDP,
                src_ip=self.pg0.remote_ip4,
                dst_ip=self.pg1.remote_ip4,
                src_port=sport,
                dst_port=dport,
            )
            + payload_match,
        )

        self.input_acl_set_interface(self.pg0, self.acl_tbl_idx.get(key))
        self.acl_active_table = key

        self.pg_enable_capture(self.pg_interfaces)
        self.pg_start()

        pkts = self.pg1.get_capture(len(pkts))
        self.verify_capture(self.pg1, pkts)
        self.pg0.assert_nothing_captured(remark="packets forwarded")
        self.pg2.assert_nothing_captured(remark="packets forwarded")
        self.pg3.assert_nothing_captured(remark="packets forwarded")

    def test_oacl_large(self):
        """Large output ACL test
        Test scenario for Large ACL matching on ethernet+ip+udp headers
            - Create IPv4 stream for pg1 -> pg0 interface.
            - Create large acl matching on ethernet+ip+udp header fields
            - Send and verify received packets on pg0 interface.
        """

        # 40b offset = 80bytes - (sizeof(UDP/IP/ETH) + 4b)
        # + 4b as build_ip_ma*() func, do not match against UDP Len & Chksum
        msk = VarMask(offset=40, spec="ffff")
        mth = VarMatch(offset=40, value=0x1234, length=2)

        payload_msk = self.build_payload_mask([msk])
        payload_match = self.build_payload_match([mth])

        sport = 13720
        dport = 9080

        # 36b offset = 80bytes - (sizeof(UDP/IP/ETH))
        packet_ex = bytes.fromhex(("0" * 36) + "1234")
        pkts = self.create_stream(
            self.pg1,
            self.pg0,
            self.pg_if_packet_sizes,
            UDP(sport=sport, dport=dport),
            packet_ex,
        )
        self.pg1.add_stream(pkts)

        key = "large_out"
        self.create_classify_table(
            key,
            self.build_mac_mask(
                src_mac="ffffffffffff", dst_mac="ffffffffffff", ether_type="ffff"
            )
            + self.build_ip_mask(
                proto="ff",
                src_ip="ffffffff",
                dst_ip="ffffffff",
                src_port="ffff",
                dst_port="ffff",
            )
            + payload_msk,
            data_offset=-14,
        )

        self.create_classify_session(
            self.acl_tbl_idx.get(key),
            self.build_mac_match(
                src_mac=self.pg0.local_mac,
                dst_mac=self.pg0.remote_mac,
                # ipv4 next header
                ether_type="0800",
            )
            + self.build_ip_match(
                proto=socket.IPPROTO_UDP,
                src_ip=self.pg1.remote_ip4,
                dst_ip=self.pg0.remote_ip4,
                src_port=sport,
                dst_port=dport,
            )
            + payload_match,
        )

        self.output_acl_set_interface(self.pg0, self.acl_tbl_idx.get(key))
        self.acl_active_table = key

        self.pg_enable_capture(self.pg_interfaces)
        self.pg_start()

        pkts = self.pg0.get_capture(len(pkts))
        self.verify_capture(self.pg0, pkts)
        self.pg1.assert_nothing_captured(remark="packets forwarded")
        self.pg2.assert_nothing_captured(remark="packets forwarded")
        self.pg3.assert_nothing_captured(remark="packets forwarded")

    def test_iacl_nested(self):
        """Nested input ACL test

        Test scenario for Large ACL matching on ethernet+ip+udp headers
            - Create IPv4 stream for pg0 -> pg1 interface.
            - Create 1st classifier table, without any entries
            - Create nested acl matching on ethernet+ip+udp header fields
            - Send and verify received packets on pg1 interface.
        """

        sport = 13720
        dport = 9080
        pkts = self.create_stream(
            self.pg0, self.pg1, self.pg_if_packet_sizes, UDP(sport=sport, dport=dport)
        )

        self.pg0.add_stream(pkts)

        subtable_key = "subtable_in"
        self.create_classify_table(
            subtable_key,
            self.build_mac_mask(
                src_mac="ffffffffffff", dst_mac="ffffffffffff", ether_type="ffff"
            )
            + self.build_ip_mask(
                proto="ff",
                src_ip="ffffffff",
                dst_ip="ffffffff",
                src_port="ffff",
                dst_port="ffff",
            ),
            data_offset=-14,
        )

        key = "nested_in"
        self.create_classify_table(
            key,
            self.build_mac_mask(
                src_mac="ffffffffffff", dst_mac="ffffffffffff", ether_type="ffff"
            )
            + self.build_ip_mask(
                proto="ff",
                src_ip="ffffffff",
                dst_ip="ffffffff",
                src_port="ffff",
                dst_port="ffff",
            ),
            next_table_index=self.acl_tbl_idx.get(subtable_key),
        )

        self.create_classify_session(
            self.acl_tbl_idx.get(subtable_key),
            self.build_mac_match(
                src_mac=self.pg0.remote_mac,
                dst_mac=self.pg0.local_mac,
                # ipv4 next header
                ether_type="0800",
            )
            + self.build_ip_match(
                proto=socket.IPPROTO_UDP,
                src_ip=self.pg0.remote_ip4,
                dst_ip=self.pg1.remote_ip4,
                src_port=sport,
                dst_port=dport,
            ),
        )

        self.input_acl_set_interface(self.pg0, self.acl_tbl_idx.get(key))
        self.acl_active_table = key

        self.pg_enable_capture(self.pg_interfaces)
        self.pg_start()

        pkts = self.pg1.get_capture(len(pkts))
        self.verify_capture(self.pg1, pkts)
        self.pg0.assert_nothing_captured(remark="packets forwarded")
        self.pg2.assert_nothing_captured(remark="packets forwarded")
        self.pg3.assert_nothing_captured(remark="packets forwarded")

    def test_oacl_nested(self):
        """Nested output ACL test

        Test scenario for Large ACL matching on ethernet+ip+udp headers
            - Create IPv4 stream for pg1 -> pg0 interface.
            - Create 1st classifier table, without any entries
            - Create nested acl matching on ethernet+ip+udp header fields
            - Send and verify received packets on pg0 interface.
        """

        sport = 13720
        dport = 9080
        pkts = self.create_stream(
            self.pg1, self.pg0, self.pg_if_packet_sizes, UDP(sport=sport, dport=dport)
        )
        self.pg1.add_stream(pkts)

        subtable_key = "subtable_out"
        self.create_classify_table(
            subtable_key,
            self.build_mac_mask(
                src_mac="ffffffffffff", dst_mac="ffffffffffff", ether_type="ffff"
            )
            + self.build_ip_mask(
                proto="ff",
                src_ip="ffffffff",
                dst_ip="ffffffff",
                src_port="ffff",
                dst_port="ffff",
            ),
            data_offset=-14,
        )

        key = "nested_out"
        self.create_classify_table(
            key,
            self.build_mac_mask(
                src_mac="ffffffffffff", dst_mac="ffffffffffff", ether_type="ffff"
            )
            + self.build_ip_mask(
                proto="ff",
                src_ip="ffffffff",
                dst_ip="ffffffff",
                src_port="ffff",
                dst_port="ffff",
            ),
            next_table_index=self.acl_tbl_idx.get(subtable_key),
            data_offset=-14,
        )

        self.create_classify_session(
            self.acl_tbl_idx.get(subtable_key),
            self.build_mac_match(
                src_mac=self.pg0.local_mac,
                dst_mac=self.pg0.remote_mac,
                # ipv4 next header
                ether_type="0800",
            )
            + self.build_ip_match(
                proto=socket.IPPROTO_UDP,
                src_ip=self.pg1.remote_ip4,
                dst_ip=self.pg0.remote_ip4,
                src_port=sport,
                dst_port=dport,
            ),
        )

        self.output_acl_set_interface(self.pg0, self.acl_tbl_idx.get(key))
        self.acl_active_table = key

        self.pg_enable_capture(self.pg_interfaces)
        self.pg_start()

        pkts = self.pg0.get_capture(len(pkts))
        self.verify_capture(self.pg0, pkts)
        self.pg1.assert_nothing_captured(remark="packets forwarded")
        self.pg2.assert_nothing_captured(remark="packets forwarded")
        self.pg3.assert_nothing_captured(remark="packets forwarded")


class TestClassifierPBR(TestClassifier):
    """Classifier PBR Test Case"""

    @classmethod
    def setUpClass(cls):
        super(TestClassifierPBR, cls).setUpClass()

    @classmethod
    def tearDownClass(cls):
        super(TestClassifierPBR, cls).tearDownClass()

    def test_acl_pbr(self):
        """IP PBR test

        Test scenario for PBR with source IP
            - Create IPv4 stream for pg0 -> pg3 interface.
            - Configure PBR fib entry for packet forwarding.
            - Send and verify received packets on pg3 interface.
        """

        # PBR testing with source IP
        pkts = self.create_stream(self.pg0, self.pg3, self.pg_if_packet_sizes)
        self.pg0.add_stream(pkts)

        key = "pbr"
        self.create_classify_table(key, self.build_ip_mask(src_ip="ffffffff"))
        pbr_option = 1
        # this will create the VRF/table in which we will insert the route
        self.create_classify_session(
            self.acl_tbl_idx.get(key),
            self.build_ip_match(src_ip=self.pg0.remote_ip4),
            pbr_option,
            self.pbr_vrfid,
        )
        self.assertTrue(self.verify_vrf(self.pbr_vrfid))
        r = VppIpRoute(
            self,
            self.pg3.local_ip4,
            24,
            [VppRoutePath(self.pg3.remote_ip4, INVALID_INDEX)],
            table_id=self.pbr_vrfid,
        )
        r.add_vpp_config()

        self.input_acl_set_interface(self.pg0, self.acl_tbl_idx.get(key))

        self.pg_enable_capture(self.pg_interfaces)
        self.pg_start()

        pkts = self.pg3.get_capture(len(pkts))
        self.verify_capture(self.pg3, pkts)
        self.input_acl_set_interface(self.pg0, self.acl_tbl_idx.get(key), 0)
        self.pg0.assert_nothing_captured(remark="packets forwarded")
        self.pg1.assert_nothing_captured(remark="packets forwarded")
        self.pg2.assert_nothing_captured(remark="packets forwarded")

        # remove the classify session and the route
        r.remove_vpp_config()
        self.create_classify_session(
            self.acl_tbl_idx.get(key),
            self.build_ip_match(src_ip=self.pg0.remote_ip4),
            pbr_option,
            self.pbr_vrfid,
            is_add=0,
        )

        # and the table should be gone.
        self.assertFalse(self.verify_vrf(self.pbr_vrfid))


class TestClassifierPunt(TestClassifier):
    """Classifier punt Test Case"""

    @classmethod
    def setUpClass(cls):
        super(TestClassifierPunt, cls).setUpClass()

    @classmethod
    def tearDownClass(cls):
        super(TestClassifierPunt, cls).tearDownClass()

    def test_punt_udp(self):
        """IPv4/UDP protocol punt ACL test

        Test scenario for basic punt ACL with UDP protocol
            - Create IPv4 stream for pg0 -> pg1 interface.
            - Create punt ACL with UDP IP protocol.
            - Send and verify received packets on pg1 interface.
        """

        sport = 6754
        dport = 17923

        key = "ip4_udp_punt"
        self.create_classify_table(
            key, self.build_ip_mask(src_ip="ffffffff", proto="ff", src_port="ffff")
        )
        table_index = self.acl_tbl_idx.get(key)
        self.vapi.punt_acl_add_del(ip4_table_index=table_index)
        self.acl_active_table = key

        # punt udp packets to dport received on pg0 through pg1
        self.vapi.set_punt(
            is_add=1,
            punt={
                "type": VppEnum.vl_api_punt_type_t.PUNT_API_TYPE_L4,
                "punt": {
                    "l4": {
                        "af": VppEnum.vl_api_address_family_t.ADDRESS_IP4,
                        "protocol": VppEnum.vl_api_ip_proto_t.IP_API_PROTO_UDP,
                        "port": dport,
                    }
                },
            },
        )
        self.vapi.ip_punt_redirect(
            punt={
                "rx_sw_if_index": self.pg0.sw_if_index,
                "tx_sw_if_index": self.pg1.sw_if_index,
                "nh": self.pg1.remote_ip4,
            }
        )

        pkts = [
            (
                Ether(dst=self.pg0.local_mac, src=self.pg0.remote_mac)
                / IP(src=self.pg0.remote_ip4, dst=self.pg0.local_ip4)
                / UDP(sport=sport, dport=dport)
                / Raw("\x17" * 100)
            )
        ] * 2

        # allow a session but not matching the stream: expect to drop
        self.create_classify_session(
            table_index,
            self.build_ip_match(
                src_ip=self.pg0.remote_ip4,
                proto=socket.IPPROTO_UDP,
                src_port=sport + 10,
            ),
        )
        self.send_and_assert_no_replies(self.pg0, pkts)

        # allow a session matching the stream: expect to pass
        self.create_classify_session(
            table_index,
            self.build_ip_match(
                src_ip=self.pg0.remote_ip4, proto=socket.IPPROTO_UDP, src_port=sport
            ),
        )
        self.send_and_expect_only(self.pg0, pkts, self.pg1)

        # test dump api: ip4 is set, ip6 is not
        r = self.vapi.punt_acl_get()
        self.assertEqual(r.ip4_table_index, table_index)
        self.assertEqual(r.ip6_table_index, 0xFFFFFFFF)

        # cleanup
        self.acl_active_table = ""
        self.vapi.punt_acl_add_del(ip4_table_index=table_index, is_add=0)

        # test dump api: nothing set
        r = self.vapi.punt_acl_get()
        self.assertEqual(r.ip4_table_index, 0xFFFFFFFF)
        self.assertEqual(r.ip6_table_index, 0xFFFFFFFF)


if __name__ == "__main__":
    unittest.main(testRunner=VppTestRunner)
"n">u8 *format_ip4_lookup_trace (u8 * s, va_list * args); VLIB_REGISTER_NODE (ip4_lookup_node) = { .function = ip4_lookup,.name = "ip4-lookup",.vector_size = sizeof (u32),.format_trace = format_ip4_lookup_trace,.n_next_nodes = IP_LOOKUP_N_NEXT,.next_nodes = IP4_LOOKUP_NEXT_NODES,}; VLIB_NODE_FUNCTION_MULTIARCH (ip4_lookup_node, ip4_lookup); always_inline uword ip4_load_balance (vlib_main_t * vm, vlib_node_runtime_t * node, vlib_frame_t * frame) { vlib_combined_counter_main_t *cm = &load_balance_main.lbm_via_counters; u32 n_left_from, n_left_to_next, *from, *to_next; ip_lookup_next_t next; u32 thread_index = vlib_get_thread_index (); from = vlib_frame_vector_args (frame); n_left_from = frame->n_vectors; next = node->cached_next_index; if (node->flags & VLIB_NODE_FLAG_TRACE) ip4_forward_next_trace (vm, node, frame, VLIB_TX); while (n_left_from > 0) { vlib_get_next_frame (vm, node, next, to_next, n_left_to_next); while (n_left_from >= 4 && n_left_to_next >= 2) { ip_lookup_next_t next0, next1; const load_balance_t *lb0, *lb1; vlib_buffer_t *p0, *p1; u32 pi0, lbi0, hc0, pi1, lbi1, hc1; const ip4_header_t *ip0, *ip1; const dpo_id_t *dpo0, *dpo1; /* 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, STORE); vlib_prefetch_buffer_header (p3, STORE); CLIB_PREFETCH (p2->data, sizeof (ip0[0]), STORE); CLIB_PREFETCH (p3->data, sizeof (ip0[0]), STORE); } pi0 = to_next[0] = from[0]; pi1 = to_next[1] = from[1]; from += 2; n_left_from -= 2; to_next += 2; n_left_to_next -= 2; p0 = vlib_get_buffer (vm, pi0); p1 = vlib_get_buffer (vm, pi1); ip0 = vlib_buffer_get_current (p0); ip1 = vlib_buffer_get_current (p1); lbi0 = vnet_buffer (p0)->ip.adj_index[VLIB_TX]; lbi1 = vnet_buffer (p1)->ip.adj_index[VLIB_TX]; lb0 = load_balance_get (lbi0); lb1 = load_balance_get (lbi1); /* * this node is for via FIBs we can re-use the hash value from the * to node if present. * We don't want to use the same hash value at each level in the recursion * graph as that would lead to polarisation */ hc0 = hc1 = 0; if (PREDICT_FALSE (lb0->lb_n_buckets > 1)) { if (PREDICT_TRUE (vnet_buffer (p0)->ip.flow_hash)) { hc0 = vnet_buffer (p0)->ip.flow_hash = vnet_buffer (p0)->ip.flow_hash >> 1; } else { hc0 = vnet_buffer (p0)->ip.flow_hash = ip4_compute_flow_hash (ip0, lb0->lb_hash_config); } dpo0 = load_balance_get_fwd_bucket (lb0, (hc0 & (lb0->lb_n_buckets_minus_1))); } else { dpo0 = load_balance_get_bucket_i (lb0, 0); } if (PREDICT_FALSE (lb1->lb_n_buckets > 1)) { if (PREDICT_TRUE (vnet_buffer (p1)->ip.flow_hash)) { hc1 = vnet_buffer (p1)->ip.flow_hash = vnet_buffer (p1)->ip.flow_hash >> 1; } else { hc1 = vnet_buffer (p1)->ip.flow_hash = ip4_compute_flow_hash (ip1, lb1->lb_hash_config); } dpo1 = load_balance_get_fwd_bucket (lb1, (hc1 & (lb1->lb_n_buckets_minus_1))); } else { dpo1 = load_balance_get_bucket_i (lb1, 0); } next0 = dpo0->dpoi_next_node; next1 = dpo1->dpoi_next_node; vnet_buffer (p0)->ip.adj_index[VLIB_TX] = dpo0->dpoi_index; vnet_buffer (p1)->ip.adj_index[VLIB_TX] = dpo1->dpoi_index; vlib_increment_combined_counter (cm, thread_index, lbi0, 1, vlib_buffer_length_in_chain (vm, p0)); vlib_increment_combined_counter (cm, thread_index, lbi1, 1, vlib_buffer_length_in_chain (vm, p1)); vlib_validate_buffer_enqueue_x2 (vm, node, next, to_next, n_left_to_next, pi0, pi1, next0, next1); } while (n_left_from > 0 && n_left_to_next > 0) { ip_lookup_next_t next0; const load_balance_t *lb0; vlib_buffer_t *p0; u32 pi0, lbi0, hc0; const ip4_header_t *ip0; const dpo_id_t *dpo0; pi0 = from[0]; to_next[0] = pi0; from += 1; to_next += 1; n_left_to_next -= 1; n_left_from -= 1; p0 = vlib_get_buffer (vm, pi0); ip0 = vlib_buffer_get_current (p0); lbi0 = vnet_buffer (p0)->ip.adj_index[VLIB_TX]; lb0 = load_balance_get (lbi0); hc0 = 0; if (PREDICT_FALSE (lb0->lb_n_buckets > 1)) { if (PREDICT_TRUE (vnet_buffer (p0)->ip.flow_hash)) { hc0 = vnet_buffer (p0)->ip.flow_hash = vnet_buffer (p0)->ip.flow_hash >> 1; } else { hc0 = vnet_buffer (p0)->ip.flow_hash = ip4_compute_flow_hash (ip0, lb0->lb_hash_config); } dpo0 = load_balance_get_fwd_bucket (lb0, (hc0 & (lb0->lb_n_buckets_minus_1))); } else { dpo0 = load_balance_get_bucket_i (lb0, 0); } next0 = dpo0->dpoi_next_node; vnet_buffer (p0)->ip.adj_index[VLIB_TX] = dpo0->dpoi_index; vlib_increment_combined_counter (cm, thread_index, lbi0, 1, vlib_buffer_length_in_chain (vm, p0)); vlib_validate_buffer_enqueue_x1 (vm, node, next, to_next, n_left_to_next, pi0, next0); } vlib_put_next_frame (vm, node, next, n_left_to_next); } return frame->n_vectors; } VLIB_REGISTER_NODE (ip4_load_balance_node) = { .function = ip4_load_balance,.name = "ip4-load-balance",.vector_size = sizeof (u32),.sibling_of = "ip4-lookup",.format_trace = format_ip4_lookup_trace,}; VLIB_NODE_FUNCTION_MULTIARCH (ip4_load_balance_node, ip4_load_balance); /* get first interface address */ ip4_address_t * ip4_interface_first_address (ip4_main_t * im, u32 sw_if_index, ip_interface_address_t ** result_ia) { ip_lookup_main_t *lm = &im->lookup_main; ip_interface_address_t *ia = 0; ip4_address_t *result = 0; /* *INDENT-OFF* */ foreach_ip_interface_address (lm, ia, sw_if_index, 1 /* honor unnumbered */ , ({ ip4_address_t * a = ip_interface_address_get_address (lm, ia); result = a; break; })); /* *INDENT-OFF* */ if (result_ia) *result_ia = result ? ia : 0; return result; } static void ip4_add_interface_routes (u32 sw_if_index, ip4_main_t * im, u32 fib_index, ip_interface_address_t * a) { ip_lookup_main_t *lm = &im->lookup_main; ip4_address_t *address = ip_interface_address_get_address (lm, a); fib_prefix_t pfx = { .fp_len = a->address_length, .fp_proto = FIB_PROTOCOL_IP4, .fp_addr.ip4 = *address, }; if (pfx.fp_len <= 30) { /* a /30 or shorter - add a glean for the network address */ fib_table_entry_update_one_path (fib_index, &pfx, FIB_SOURCE_INTERFACE, (FIB_ENTRY_FLAG_CONNECTED | FIB_ENTRY_FLAG_ATTACHED), DPO_PROTO_IP4, /* No next-hop address */ NULL, sw_if_index, // invalid FIB index ~0, 1, // no out-label stack NULL, FIB_ROUTE_PATH_FLAG_NONE); /* Add the two broadcast addresses as drop */ fib_prefix_t net_pfx = { .fp_len = 32, .fp_proto = FIB_PROTOCOL_IP4, .fp_addr.ip4.as_u32 = address->as_u32 & im->fib_masks[pfx.fp_len], }; if (net_pfx.fp_addr.ip4.as_u32 != pfx.fp_addr.ip4.as_u32) fib_table_entry_special_add(fib_index, &net_pfx, FIB_SOURCE_INTERFACE, (FIB_ENTRY_FLAG_DROP | FIB_ENTRY_FLAG_LOOSE_URPF_EXEMPT)); net_pfx.fp_addr.ip4.as_u32 |= ~im->fib_masks[pfx.fp_len]; if (net_pfx.fp_addr.ip4.as_u32 != pfx.fp_addr.ip4.as_u32) fib_table_entry_special_add(fib_index, &net_pfx, FIB_SOURCE_INTERFACE, (FIB_ENTRY_FLAG_DROP | FIB_ENTRY_FLAG_LOOSE_URPF_EXEMPT)); } else if (pfx.fp_len == 31) { u32 mask = clib_host_to_net_u32(1); fib_prefix_t net_pfx = pfx; net_pfx.fp_len = 32; net_pfx.fp_addr.ip4.as_u32 ^= mask; /* a /31 - add the other end as an attached host */ fib_table_entry_update_one_path (fib_index, &net_pfx, FIB_SOURCE_INTERFACE, (FIB_ENTRY_FLAG_ATTACHED), DPO_PROTO_IP4, &net_pfx.fp_addr, sw_if_index, // invalid FIB index ~0, 1, NULL, FIB_ROUTE_PATH_FLAG_NONE); } pfx.fp_len = 32; if (sw_if_index < vec_len (lm->classify_table_index_by_sw_if_index)) { u32 classify_table_index = lm->classify_table_index_by_sw_if_index[sw_if_index]; if (classify_table_index != (u32) ~ 0) { dpo_id_t dpo = DPO_INVALID; dpo_set (&dpo, DPO_CLASSIFY, DPO_PROTO_IP4, classify_dpo_create (DPO_PROTO_IP4, classify_table_index)); fib_table_entry_special_dpo_add (fib_index, &pfx, FIB_SOURCE_CLASSIFY, FIB_ENTRY_FLAG_NONE, &dpo); dpo_reset (&dpo); } } fib_table_entry_update_one_path (fib_index, &pfx, FIB_SOURCE_INTERFACE, (FIB_ENTRY_FLAG_CONNECTED | FIB_ENTRY_FLAG_LOCAL), DPO_PROTO_IP4, &pfx.fp_addr, sw_if_index, // invalid FIB index ~0, 1, NULL, FIB_ROUTE_PATH_FLAG_NONE); } static void ip4_del_interface_routes (ip4_main_t * im, u32 fib_index, ip4_address_t * address, u32 address_length) { fib_prefix_t pfx = { .fp_len = address_length, .fp_proto = FIB_PROTOCOL_IP4, .fp_addr.ip4 = *address, }; if (pfx.fp_len <= 30) { fib_prefix_t net_pfx = { .fp_len = 32, .fp_proto = FIB_PROTOCOL_IP4, .fp_addr.ip4.as_u32 = address->as_u32 & im->fib_masks[pfx.fp_len], }; if (net_pfx.fp_addr.ip4.as_u32 != pfx.fp_addr.ip4.as_u32) fib_table_entry_special_remove(fib_index, &net_pfx, FIB_SOURCE_INTERFACE); net_pfx.fp_addr.ip4.as_u32 |= ~im->fib_masks[pfx.fp_len]; if (net_pfx.fp_addr.ip4.as_u32 != pfx.fp_addr.ip4.as_u32) fib_table_entry_special_remove(fib_index, &net_pfx, FIB_SOURCE_INTERFACE); fib_table_entry_delete (fib_index, &pfx, FIB_SOURCE_INTERFACE); } else if (pfx.fp_len == 31) { u32 mask = clib_host_to_net_u32(1); fib_prefix_t net_pfx = pfx; net_pfx.fp_len = 32; net_pfx.fp_addr.ip4.as_u32 ^= mask; fib_table_entry_delete (fib_index, &net_pfx, FIB_SOURCE_INTERFACE); } pfx.fp_len = 32; fib_table_entry_delete (fib_index, &pfx, FIB_SOURCE_INTERFACE); } void ip4_sw_interface_enable_disable (u32 sw_if_index, u32 is_enable) { ip4_main_t *im = &ip4_main; vec_validate_init_empty (im->ip_enabled_by_sw_if_index, sw_if_index, 0); /* * enable/disable only on the 1<->0 transition */ if (is_enable) { if (1 != ++im->ip_enabled_by_sw_if_index[sw_if_index]) return; } else { ASSERT (im->ip_enabled_by_sw_if_index[sw_if_index] > 0); if (0 != --im->ip_enabled_by_sw_if_index[sw_if_index]) return; } vnet_feature_enable_disable ("ip4-unicast", "ip4-drop", sw_if_index, !is_enable, 0, 0); vnet_feature_enable_disable ("ip4-multicast", "ip4-drop", sw_if_index, !is_enable, 0, 0); } static clib_error_t * ip4_add_del_interface_address_internal (vlib_main_t * vm, u32 sw_if_index, ip4_address_t * address, u32 address_length, u32 is_del) { vnet_main_t *vnm = vnet_get_main (); ip4_main_t *im = &ip4_main; ip_lookup_main_t *lm = &im->lookup_main; clib_error_t *error = 0; u32 if_address_index, elts_before; ip4_address_fib_t ip4_af, *addr_fib = 0; /* local0 interface doesn't support IP addressing */ if (sw_if_index == 0) { return clib_error_create ("local0 interface doesn't support IP addressing"); } vec_validate (im->fib_index_by_sw_if_index, sw_if_index); ip4_addr_fib_init (&ip4_af, address, vec_elt (im->fib_index_by_sw_if_index, sw_if_index)); vec_add1 (addr_fib, ip4_af); /* FIXME-LATER * there is no support for adj-fib handling in the presence of overlapping * subnets on interfaces. Easy fix - disallow overlapping subnets, like * most routers do. */ /* *INDENT-OFF* */ if (!is_del) { /* When adding an address check that it does not conflict with an existing address. */ ip_interface_address_t *ia; foreach_ip_interface_address (&im->lookup_main, ia, sw_if_index, 0 /* honor unnumbered */ , ({ ip4_address_t * x = ip_interface_address_get_address (&im->lookup_main, ia); if (ip4_destination_matches_route (im, address, x, ia->address_length) || ip4_destination_matches_route (im, x, address, address_length)) return clib_error_create ("failed to add %U which conflicts with %U for interface %U", format_ip4_address_and_length, address, address_length, format_ip4_address_and_length, x, ia->address_length, format_vnet_sw_if_index_name, vnm, sw_if_index); })); } /* *INDENT-ON* */ elts_before = pool_elts (lm->if_address_pool); error = ip_interface_address_add_del (lm, sw_if_index, addr_fib, address_length, is_del, &if_address_index); if (error) goto done; ip4_sw_interface_enable_disable (sw_if_index, !is_del); if (is_del) ip4_del_interface_routes (im, ip4_af.fib_index, address, address_length); else ip4_add_interface_routes (sw_if_index, im, ip4_af.fib_index, pool_elt_at_index (lm->if_address_pool, if_address_index)); /* If pool did not grow/shrink: add duplicate address. */ if (elts_before != pool_elts (lm->if_address_pool)) { ip4_add_del_interface_address_callback_t *cb; vec_foreach (cb, im->add_del_interface_address_callbacks) cb->function (im, cb->function_opaque, sw_if_index, address, address_length, if_address_index, is_del); } done: vec_free (addr_fib); return error; } clib_error_t * ip4_add_del_interface_address (vlib_main_t * vm, u32 sw_if_index, ip4_address_t * address, u32 address_length, u32 is_del) { return ip4_add_del_interface_address_internal (vm, sw_if_index, address, address_length, is_del); } /* Built-in ip4 unicast rx feature path definition */ /* *INDENT-OFF* */ VNET_FEATURE_ARC_INIT (ip4_unicast, static) = { .arc_name = "ip4-unicast", .start_nodes = VNET_FEATURES ("ip4-input", "ip4-input-no-checksum"), .arc_index_ptr = &ip4_main.lookup_main.ucast_feature_arc_index, }; VNET_FEATURE_INIT (ip4_flow_classify, static) = { .arc_name = "ip4-unicast", .node_name = "ip4-flow-classify", .runs_before = VNET_FEATURES ("ip4-inacl"), }; VNET_FEATURE_INIT (ip4_inacl, static) = { .arc_name = "ip4-unicast", .node_name = "ip4-inacl", .runs_before = VNET_FEATURES ("ip4-source-check-via-rx"), }; VNET_FEATURE_INIT (ip4_source_check_1, static) = { .arc_name = "ip4-unicast", .node_name = "ip4-source-check-via-rx", .runs_before = VNET_FEATURES ("ip4-source-check-via-any"), }; VNET_FEATURE_INIT (ip4_source_check_2, static) = { .arc_name = "ip4-unicast", .node_name = "ip4-source-check-via-any", .runs_before = VNET_FEATURES ("ip4-policer-classify"), }; VNET_FEATURE_INIT (ip4_source_and_port_range_check_rx, static) = { .arc_name = "ip4-unicast", .node_name = "ip4-source-and-port-range-check-rx", .runs_before = VNET_FEATURES ("ip4-policer-classify"), }; VNET_FEATURE_INIT (ip4_policer_classify, static) = { .arc_name = "ip4-unicast", .node_name = "ip4-policer-classify", .runs_before = VNET_FEATURES ("ipsec-input-ip4"), }; VNET_FEATURE_INIT (ip4_ipsec, static) = { .arc_name = "ip4-unicast", .node_name = "ipsec-input-ip4", .runs_before = VNET_FEATURES ("vpath-input-ip4"), }; VNET_FEATURE_INIT (ip4_vpath, static) = { .arc_name = "ip4-unicast", .node_name = "vpath-input-ip4", .runs_before = VNET_FEATURES ("ip4-vxlan-bypass"), }; VNET_FEATURE_INIT (ip4_vxlan_bypass, static) = { .arc_name = "ip4-unicast", .node_name = "ip4-vxlan-bypass", .runs_before = VNET_FEATURES ("ip4-lookup"), }; VNET_FEATURE_INIT (ip4_drop, static) = { .arc_name = "ip4-unicast", .node_name = "ip4-drop", .runs_before = VNET_FEATURES ("ip4-lookup"), }; VNET_FEATURE_INIT (ip4_lookup, static) = { .arc_name = "ip4-unicast", .node_name = "ip4-lookup", .runs_before = 0, /* not before any other features */ }; /* Built-in ip4 multicast rx feature path definition */ VNET_FEATURE_ARC_INIT (ip4_multicast, static) = { .arc_name = "ip4-multicast", .start_nodes = VNET_FEATURES ("ip4-input", "ip4-input-no-checksum"), .arc_index_ptr = &ip4_main.lookup_main.mcast_feature_arc_index, }; VNET_FEATURE_INIT (ip4_vpath_mc, static) = { .arc_name = "ip4-multicast", .node_name = "vpath-input-ip4", .runs_before = VNET_FEATURES ("ip4-mfib-forward-lookup"), }; VNET_FEATURE_INIT (ip4_mc_drop, static) = { .arc_name = "ip4-multicast", .node_name = "ip4-drop", .runs_before = VNET_FEATURES ("ip4-mfib-forward-lookup"), }; VNET_FEATURE_INIT (ip4_lookup_mc, static) = { .arc_name = "ip4-multicast", .node_name = "ip4-mfib-forward-lookup", .runs_before = 0, /* last feature */ }; /* Source and port-range check ip4 tx feature path definition */ VNET_FEATURE_ARC_INIT (ip4_output, static) = { .arc_name = "ip4-output", .start_nodes = VNET_FEATURES ("ip4-rewrite", "ip4-midchain"), .arc_index_ptr = &ip4_main.lookup_main.output_feature_arc_index, }; VNET_FEATURE_INIT (ip4_source_and_port_range_check_tx, static) = { .arc_name = "ip4-output", .node_name = "ip4-source-and-port-range-check-tx", .runs_before = VNET_FEATURES ("ipsec-output-ip4"), }; VNET_FEATURE_INIT (ip4_ipsec_output, static) = { .arc_name = "ip4-output", .node_name = "ipsec-output-ip4", .runs_before = VNET_FEATURES ("interface-output"), }; /* Built-in ip4 tx feature path definition */ VNET_FEATURE_INIT (ip4_interface_output, static) = { .arc_name = "ip4-output", .node_name = "interface-output", .runs_before = 0, /* not before any other features */ }; /* *INDENT-ON* */ static clib_error_t * ip4_sw_interface_add_del (vnet_main_t * vnm, u32 sw_if_index, u32 is_add) { ip4_main_t *im = &ip4_main; /* Fill in lookup tables with default table (0). */ vec_validate (im->fib_index_by_sw_if_index, sw_if_index); vec_validate (im->mfib_index_by_sw_if_index, sw_if_index); if (!is_add) { ip4_main_t *im4 = &ip4_main; ip_lookup_main_t *lm4 = &im4->lookup_main; ip_interface_address_t *ia = 0; ip4_address_t *address; vlib_main_t *vm = vlib_get_main (); /* *INDENT-OFF* */ foreach_ip_interface_address (lm4, ia, sw_if_index, 1 /* honor unnumbered */, ({ address = ip_interface_address_get_address (lm4, ia); ip4_add_del_interface_address(vm, sw_if_index, address, ia->address_length, 1); })); /* *INDENT-ON* */ } vnet_feature_enable_disable ("ip4-unicast", "ip4-drop", sw_if_index, is_add, 0, 0); vnet_feature_enable_disable ("ip4-multicast", "ip4-drop", sw_if_index, is_add, 0, 0); return /* no error */ 0; } VNET_SW_INTERFACE_ADD_DEL_FUNCTION (ip4_sw_interface_add_del); /* Global IP4 main. */ ip4_main_t ip4_main; clib_error_t * ip4_lookup_init (vlib_main_t * vm) { ip4_main_t *im = &ip4_main; clib_error_t *error; uword i; if ((error = vlib_call_init_function (vm, vnet_feature_init))) return error; for (i = 0; i < ARRAY_LEN (im->fib_masks); i++) { u32 m; if (i < 32) m = pow2_mask (i) << (32 - i); else m = ~0; im->fib_masks[i] = clib_host_to_net_u32 (m); } ip_lookup_init (&im->lookup_main, /* is_ip6 */ 0); /* Create FIB with index 0 and table id of 0. */ fib_table_find_or_create_and_lock (FIB_PROTOCOL_IP4, 0, FIB_SOURCE_DEFAULT_ROUTE); mfib_table_find_or_create_and_lock (FIB_PROTOCOL_IP4, 0, MFIB_SOURCE_DEFAULT_ROUTE); { pg_node_t *pn; pn = pg_get_node (ip4_lookup_node.index); pn->unformat_edit = unformat_pg_ip4_header; } { ethernet_arp_header_t h; memset (&h, 0, sizeof (h)); /* Set target ethernet address to all zeros. */ memset (h.ip4_over_ethernet[1].ethernet, 0, sizeof (h.ip4_over_ethernet[1].ethernet)); #define _16(f,v) h.f = clib_host_to_net_u16 (v); #define _8(f,v) h.f = v; _16 (l2_type, ETHERNET_ARP_HARDWARE_TYPE_ethernet); _16 (l3_type, ETHERNET_TYPE_IP4); _8 (n_l2_address_bytes, 6); _8 (n_l3_address_bytes, 4); _16 (opcode, ETHERNET_ARP_OPCODE_request); #undef _16 #undef _8 vlib_packet_template_init (vm, &im->ip4_arp_request_packet_template, /* data */ &h, sizeof (h), /* alloc chunk size */ 8, "ip4 arp"); } return error; } VLIB_INIT_FUNCTION (ip4_lookup_init); typedef struct { /* Adjacency taken. */ u32 dpo_index; u32 flow_hash; u32 fib_index; /* Packet data, possibly *after* rewrite. */ u8 packet_data[64 - 1 * sizeof (u32)]; } ip4_forward_next_trace_t; u8 * format_ip4_forward_next_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 *); ip4_forward_next_trace_t *t = va_arg (*args, ip4_forward_next_trace_t *); uword indent = format_get_indent (s); s = format (s, "%U%U", format_white_space, indent, format_ip4_header, t->packet_data, sizeof (t->packet_data)); return s; } static u8 * format_ip4_lookup_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 *); ip4_forward_next_trace_t *t = va_arg (*args, ip4_forward_next_trace_t *); uword indent = format_get_indent (s); s = format (s, "fib %d dpo-idx %d flow hash: 0x%08x", t->fib_index, t->dpo_index, t->flow_hash); s = format (s, "\n%U%U", format_white_space, indent, format_ip4_header, t->packet_data, sizeof (t->packet_data)); return s; } static u8 * format_ip4_rewrite_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 *); ip4_forward_next_trace_t *t = va_arg (*args, ip4_forward_next_trace_t *); uword indent = format_get_indent (s); s = format (s, "tx_sw_if_index %d dpo-idx %d : %U flow hash: 0x%08x", t->fib_index, t->dpo_index, format_ip_adjacency, t->dpo_index, FORMAT_IP_ADJACENCY_NONE, t->flow_hash); s = format (s, "\n%U%U", format_white_space, indent, format_ip_adjacency_packet_data, t->dpo_index, t->packet_data, sizeof (t->packet_data)); return s; } /* Common trace function for all ip4-forward next nodes. */ void ip4_forward_next_trace (vlib_main_t * vm, vlib_node_runtime_t * node, vlib_frame_t * frame, vlib_rx_or_tx_t which_adj_index) { u32 *from, n_left; ip4_main_t *im = &ip4_main; n_left = frame->n_vectors; from = vlib_frame_vector_args (frame); while (n_left >= 4) { u32 bi0, bi1; vlib_buffer_t *b0, *b1; ip4_forward_next_trace_t *t0, *t1; /* Prefetch next iteration. */ vlib_prefetch_buffer_with_index (vm, from[2], LOAD); vlib_prefetch_buffer_with_index (vm, from[3], LOAD); bi0 = from[0]; bi1 = from[1]; b0 = vlib_get_buffer (vm, bi0); b1 = vlib_get_buffer (vm, bi1); if (b0->flags & VLIB_BUFFER_IS_TRACED) { t0 = vlib_add_trace (vm, node, b0, sizeof (t0[0])); t0->dpo_index = vnet_buffer (b0)->ip.adj_index[which_adj_index]; t0->flow_hash = vnet_buffer (b0)->ip.flow_hash; t0->fib_index = (vnet_buffer (b0)->sw_if_index[VLIB_TX] != (u32) ~ 0) ? vnet_buffer (b0)->sw_if_index[VLIB_TX] : vec_elt (im->fib_index_by_sw_if_index, vnet_buffer (b0)->sw_if_index[VLIB_RX]); clib_memcpy (t0->packet_data, vlib_buffer_get_current (b0), sizeof (t0->packet_data)); } if (b1->flags & VLIB_BUFFER_IS_TRACED) { t1 = vlib_add_trace (vm, node, b1, sizeof (t1[0])); t1->dpo_index = vnet_buffer (b1)->ip.adj_index[which_adj_index]; t1->flow_hash = vnet_buffer (b1)->ip.flow_hash; t1->fib_index = (vnet_buffer (b1)->sw_if_index[VLIB_TX] != (u32) ~ 0) ? vnet_buffer (b1)->sw_if_index[VLIB_TX] : vec_elt (im->fib_index_by_sw_if_index, vnet_buffer (b1)->sw_if_index[VLIB_RX]); clib_memcpy (t1->packet_data, vlib_buffer_get_current (b1), sizeof (t1->packet_data)); } from += 2; n_left -= 2; } while (n_left >= 1) { u32 bi0; vlib_buffer_t *b0; ip4_forward_next_trace_t *t0; bi0 = from[0]; b0 = vlib_get_buffer (vm, bi0); if (b0->flags & VLIB_BUFFER_IS_TRACED) { t0 = vlib_add_trace (vm, node, b0, sizeof (t0[0])); t0->dpo_index = vnet_buffer (b0)->ip.adj_index[which_adj_index]; t0->flow_hash = vnet_buffer (b0)->ip.flow_hash; t0->fib_index = (vnet_buffer (b0)->sw_if_index[VLIB_TX] != (u32) ~ 0) ? vnet_buffer (b0)->sw_if_index[VLIB_TX] : vec_elt (im->fib_index_by_sw_if_index, vnet_buffer (b0)->sw_if_index[VLIB_RX]); clib_memcpy (t0->packet_data, vlib_buffer_get_current (b0), sizeof (t0->packet_data)); } from += 1; n_left -= 1; } } static uword ip4_drop_or_punt (vlib_main_t * vm, vlib_node_runtime_t * node, vlib_frame_t * frame, ip4_error_t error_code) { u32 *buffers = vlib_frame_vector_args (frame); uword n_packets = frame->n_vectors; vlib_error_drop_buffers (vm, node, buffers, /* stride */ 1, n_packets, /* next */ 0, ip4_input_node.index, error_code); if (node->flags & VLIB_NODE_FLAG_TRACE) ip4_forward_next_trace (vm, node, frame, VLIB_TX); return n_packets; } static uword ip4_drop (vlib_main_t * vm, vlib_node_runtime_t * node, vlib_frame_t * frame) { return ip4_drop_or_punt (vm, node, frame, IP4_ERROR_ADJACENCY_DROP); } static uword ip4_punt (vlib_main_t * vm, vlib_node_runtime_t * node, vlib_frame_t * frame) { return ip4_drop_or_punt (vm, node, frame, IP4_ERROR_ADJACENCY_PUNT); } /* *INDENT-OFF* */ VLIB_REGISTER_NODE (ip4_drop_node, static) = { .function = ip4_drop, .name = "ip4-drop", .vector_size = sizeof (u32), .format_trace = format_ip4_forward_next_trace, .n_next_nodes = 1, .next_nodes = { [0] = "error-drop", }, }; VLIB_NODE_FUNCTION_MULTIARCH (ip4_drop_node, ip4_drop); VLIB_REGISTER_NODE (ip4_punt_node, static) = { .function = ip4_punt, .name = "ip4-punt", .vector_size = sizeof (u32), .format_trace = format_ip4_forward_next_trace, .n_next_nodes = 1, .next_nodes = { [0] = "error-punt", }, }; VLIB_NODE_FUNCTION_MULTIARCH (ip4_punt_node, ip4_punt); /* *INDENT-ON */ /* Compute TCP/UDP/ICMP4 checksum in software. */ u16 ip4_tcp_udp_compute_checksum (vlib_main_t * vm, vlib_buffer_t * p0, ip4_header_t * ip0) { ip_csum_t sum0; u32 ip_header_length, payload_length_host_byte_order; u32 n_this_buffer, n_bytes_left, n_ip_bytes_this_buffer; u16 sum16; void *data_this_buffer; /* Initialize checksum with ip header. */ ip_header_length = ip4_header_bytes (ip0); payload_length_host_byte_order = clib_net_to_host_u16 (ip0->length) - ip_header_length; sum0 = clib_host_to_net_u32 (payload_length_host_byte_order + (ip0->protocol << 16)); if (BITS (uword) == 32) { sum0 = ip_csum_with_carry (sum0, clib_mem_unaligned (&ip0->src_address, u32)); sum0 = ip_csum_with_carry (sum0, clib_mem_unaligned (&ip0->dst_address, u32)); } else sum0 = ip_csum_with_carry (sum0, clib_mem_unaligned (&ip0->src_address, u64)); n_bytes_left = n_this_buffer = payload_length_host_byte_order; data_this_buffer = (void *) ip0 + ip_header_length; n_ip_bytes_this_buffer = p0->current_length - (((u8 *) ip0 - p0->data) - p0->current_data); if (n_this_buffer + ip_header_length > n_ip_bytes_this_buffer) { n_this_buffer = n_ip_bytes_this_buffer > ip_header_length ? n_ip_bytes_this_buffer - ip_header_length : 0; } while (1) { sum0 = ip_incremental_checksum (sum0, data_this_buffer, n_this_buffer); n_bytes_left -= n_this_buffer; if (n_bytes_left == 0) break; ASSERT (p0->flags & VLIB_BUFFER_NEXT_PRESENT); p0 = vlib_get_buffer (vm, p0->next_buffer); data_this_buffer = vlib_buffer_get_current (p0); n_this_buffer = p0->current_length; } sum16 = ~ip_csum_fold (sum0); return sum16; } u32 ip4_tcp_udp_validate_checksum (vlib_main_t * vm, vlib_buffer_t * p0) { ip4_header_t *ip0 = vlib_buffer_get_current (p0); udp_header_t *udp0; u16 sum16; ASSERT (ip0->protocol == IP_PROTOCOL_TCP || ip0->protocol == IP_PROTOCOL_UDP); udp0 = (void *) (ip0 + 1); if (ip0->protocol == IP_PROTOCOL_UDP && udp0->checksum == 0) { p0->flags |= (VNET_BUFFER_F_L4_CHECKSUM_COMPUTED | VNET_BUFFER_F_L4_CHECKSUM_CORRECT); return p0->flags; } sum16 = ip4_tcp_udp_compute_checksum (vm, p0, ip0); p0->flags |= (VNET_BUFFER_F_L4_CHECKSUM_COMPUTED | ((sum16 == 0) << VNET_BUFFER_F_LOG2_L4_CHECKSUM_CORRECT)); return p0->flags; } /* *INDENT-OFF* */ VNET_FEATURE_ARC_INIT (ip4_local) = { .arc_name = "ip4-local", .start_nodes = VNET_FEATURES ("ip4-local"), }; /* *INDENT-ON* */ static inline void ip4_local_validate_l4 (vlib_main_t * vm, vlib_buffer_t * p, ip4_header_t * ip, u8 is_udp, u8 * error, u8 * good_tcp_udp) { u32 flags0; flags0 = ip4_tcp_udp_validate_checksum (vm, p); *good_tcp_udp = (flags0 & VNET_BUFFER_F_L4_CHECKSUM_CORRECT) != 0; if (is_udp) { udp_header_t *udp; u32 ip_len, udp_len; i32 len_diff; udp = ip4_next_header (ip); /* Verify UDP length. */ ip_len = clib_net_to_host_u16 (ip->length); udp_len = clib_net_to_host_u16 (udp->length); len_diff = ip_len - udp_len; *good_tcp_udp &= len_diff >= 0; *error = len_diff < 0 ? IP4_ERROR_UDP_LENGTH : *error; } } #define ip4_local_do_l4_check(is_tcp_udp, flags) \ (is_tcp_udp && !(flags & VNET_BUFFER_F_L4_CHECKSUM_COMPUTED)) static inline uword ip4_local_inline (vlib_main_t * vm, vlib_node_runtime_t * node, vlib_frame_t * frame, int head_of_feature_arc) { ip4_main_t *im = &ip4_main; ip_lookup_main_t *lm = &im->lookup_main; ip_local_next_t next_index; u32 *from, *to_next, n_left_from, n_left_to_next; vlib_node_runtime_t *error_node = vlib_node_get_runtime (vm, ip4_input_node.index); u8 arc_index = vnet_feat_arc_ip4_local.feature_arc_index; from = vlib_frame_vector_args (frame); n_left_from = frame->n_vectors; next_index = node->cached_next_index; if (node->flags & VLIB_NODE_FLAG_TRACE) ip4_forward_next_trace (vm, node, frame, VLIB_TX); while (n_left_from > 0) { vlib_get_next_frame (vm, node, next_index, to_next, n_left_to_next); while (n_left_from >= 4 && n_left_to_next >= 2) { vlib_buffer_t *p0, *p1; ip4_header_t *ip0, *ip1; ip4_fib_mtrie_t *mtrie0, *mtrie1; ip4_fib_mtrie_leaf_t leaf0, leaf1; const dpo_id_t *dpo0, *dpo1; const load_balance_t *lb0, *lb1; u32 pi0, next0, fib_index0, lbi0; u32 pi1, next1, fib_index1, lbi1; u8 error0, is_udp0, is_tcp_udp0, good_tcp_udp0, proto0; u8 error1, is_udp1, is_tcp_udp1, good_tcp_udp1, proto1; u32 sw_if_index0, sw_if_index1; pi0 = to_next[0] = from[0]; pi1 = to_next[1] = from[1]; from += 2; n_left_from -= 2; to_next += 2; n_left_to_next -= 2; next0 = next1 = IP_LOCAL_NEXT_DROP; error0 = error1 = IP4_ERROR_UNKNOWN_PROTOCOL; p0 = vlib_get_buffer (vm, pi0); p1 = vlib_get_buffer (vm, pi1); ip0 = vlib_buffer_get_current (p0); ip1 = vlib_buffer_get_current (p1); vnet_buffer (p0)->l3_hdr_offset = p0->current_data; vnet_buffer (p1)->l3_hdr_offset = p1->current_data; sw_if_index0 = vnet_buffer (p0)->sw_if_index[VLIB_RX]; sw_if_index1 = vnet_buffer (p1)->sw_if_index[VLIB_RX]; /* Treat IP frag packets as "experimental" protocol for now until support of IP frag reassembly is implemented */ proto0 = ip4_is_fragment (ip0) ? 0xfe : ip0->protocol; proto1 = ip4_is_fragment (ip1) ? 0xfe : ip1->protocol; if (head_of_feature_arc == 0) goto skip_checks; is_udp0 = proto0 == IP_PROTOCOL_UDP; is_udp1 = proto1 == IP_PROTOCOL_UDP; is_tcp_udp0 = is_udp0 || proto0 == IP_PROTOCOL_TCP; is_tcp_udp1 = is_udp1 || proto1 == IP_PROTOCOL_TCP; good_tcp_udp0 = (p0->flags & VNET_BUFFER_F_L4_CHECKSUM_CORRECT) != 0; good_tcp_udp1 = (p1->flags & VNET_BUFFER_F_L4_CHECKSUM_CORRECT) != 0; if (PREDICT_FALSE (ip4_local_do_l4_check (is_tcp_udp0, p0->flags) || ip4_local_do_l4_check (is_tcp_udp1, p1->flags))) { if (is_tcp_udp0) ip4_local_validate_l4 (vm, p0, ip0, is_udp0, &error0, &good_tcp_udp0); if (is_tcp_udp1) ip4_local_validate_l4 (vm, p1, ip1, is_udp1, &error1, &good_tcp_udp1); } ASSERT (IP4_ERROR_TCP_CHECKSUM + 1 == IP4_ERROR_UDP_CHECKSUM); error0 = (is_tcp_udp0 && !good_tcp_udp0 ? IP4_ERROR_TCP_CHECKSUM + is_udp0 : error0); error1 = (is_tcp_udp1 && !good_tcp_udp1 ? IP4_ERROR_TCP_CHECKSUM + is_udp1 : error1); fib_index0 = vec_elt (im->fib_index_by_sw_if_index, sw_if_index0); fib_index0 = (vnet_buffer (p0)->sw_if_index[VLIB_TX] == (u32) ~ 0) ? fib_index0 : vnet_buffer (p0)->sw_if_index[VLIB_TX]; fib_index1 = vec_elt (im->fib_index_by_sw_if_index, sw_if_index1); fib_index1 = (vnet_buffer (p1)->sw_if_index[VLIB_TX] == (u32) ~ 0) ? fib_index1 : vnet_buffer (p1)->sw_if_index[VLIB_TX]; mtrie0 = &ip4_fib_get (fib_index0)->mtrie; mtrie1 = &ip4_fib_get (fib_index1)->mtrie; leaf0 = ip4_fib_mtrie_lookup_step_one (mtrie0, &ip0->src_address); leaf1 = ip4_fib_mtrie_lookup_step_one (mtrie1, &ip1->src_address); leaf0 = ip4_fib_mtrie_lookup_step (mtrie0, leaf0, &ip0->src_address, 2); leaf1 = ip4_fib_mtrie_lookup_step (mtrie1, leaf1, &ip1->src_address, 2); leaf0 = ip4_fib_mtrie_lookup_step (mtrie0, leaf0, &ip0->src_address, 3); leaf1 = ip4_fib_mtrie_lookup_step (mtrie1, leaf1, &ip1->src_address, 3); vnet_buffer (p0)->ip.adj_index[VLIB_RX] = lbi0 = ip4_fib_mtrie_leaf_get_adj_index (leaf0); vnet_buffer (p0)->ip.adj_index[VLIB_TX] = lbi0; vnet_buffer (p1)->ip.adj_index[VLIB_RX] = lbi1 = ip4_fib_mtrie_leaf_get_adj_index (leaf1); vnet_buffer (p1)->ip.adj_index[VLIB_TX] = lbi1; lb0 = load_balance_get (lbi0); lb1 = load_balance_get (lbi1); dpo0 = load_balance_get_bucket_i (lb0, 0); dpo1 = load_balance_get_bucket_i (lb1, 0); /* * Must have a route to source otherwise we drop the packet. * ip4 broadcasts are accepted, e.g. to make dhcp client work * * The checks are: * - the source is a recieve => it's from us => bogus, do this * first since it sets a different error code. * - uRPF check for any route to source - accept if passes. * - allow packets destined to the broadcast address from unknown sources */ error0 = ((error0 == IP4_ERROR_UNKNOWN_PROTOCOL && dpo0->dpoi_type == DPO_RECEIVE) ? IP4_ERROR_SPOOFED_LOCAL_PACKETS : error0); error0 = ((error0 == IP4_ERROR_UNKNOWN_PROTOCOL && !fib_urpf_check_size (lb0->lb_urpf) && ip0->dst_address.as_u32 != 0xFFFFFFFF) ? IP4_ERROR_SRC_LOOKUP_MISS : error0); error1 = ((error1 == IP4_ERROR_UNKNOWN_PROTOCOL && dpo1->dpoi_type == DPO_RECEIVE) ? IP4_ERROR_SPOOFED_LOCAL_PACKETS : error1); error1 = ((error1 == IP4_ERROR_UNKNOWN_PROTOCOL && !fib_urpf_check_size (lb1->lb_urpf) && ip1->dst_address.as_u32 != 0xFFFFFFFF) ? IP4_ERROR_SRC_LOOKUP_MISS : error1); skip_checks: next0 = lm->local_next_by_ip_protocol[proto0]; next1 = lm->local_next_by_ip_protocol[proto1]; next0 = error0 != IP4_ERROR_UNKNOWN_PROTOCOL ? IP_LOCAL_NEXT_DROP : next0; next1 = error1 != IP4_ERROR_UNKNOWN_PROTOCOL ? IP_LOCAL_NEXT_DROP : next1; p0->error = error0 ? error_node->errors[error0] : 0; p1->error = error1 ? error_node->errors[error1] : 0; if (head_of_feature_arc) { if (PREDICT_TRUE (error0 == (u8) IP4_ERROR_UNKNOWN_PROTOCOL)) vnet_feature_arc_start (arc_index, sw_if_index0, &next0, p0); if (PREDICT_TRUE (error1 == (u8) IP4_ERROR_UNKNOWN_PROTOCOL)) vnet_feature_arc_start (arc_index, sw_if_index1, &next1, p1); } vlib_validate_buffer_enqueue_x2 (vm, node, next_index, to_next, n_left_to_next, pi0, pi1, next0, next1); } while (n_left_from > 0 && n_left_to_next > 0) { vlib_buffer_t *p0; ip4_header_t *ip0; ip4_fib_mtrie_t *mtrie0; ip4_fib_mtrie_leaf_t leaf0; u32 pi0, next0, fib_index0, lbi0; u8 error0, is_udp0, is_tcp_udp0, good_tcp_udp0, proto0; load_balance_t *lb0; const dpo_id_t *dpo0; u32 sw_if_index0; pi0 = to_next[0] = from[0]; from += 1; n_left_from -= 1; to_next += 1; n_left_to_next -= 1; next0 = IP_LOCAL_NEXT_DROP; error0 = IP4_ERROR_UNKNOWN_PROTOCOL; p0 = vlib_get_buffer (vm, pi0); ip0 = vlib_buffer_get_current (p0); vnet_buffer (p0)->l3_hdr_offset = p0->current_data; sw_if_index0 = vnet_buffer (p0)->sw_if_index[VLIB_RX]; /* Treat IP frag packets as "experimental" protocol for now until support of IP frag reassembly is implemented */ proto0 = ip4_is_fragment (ip0) ? 0xfe : ip0->protocol; if (head_of_feature_arc == 0) goto skip_check; is_udp0 = proto0 == IP_PROTOCOL_UDP; is_tcp_udp0 = is_udp0 || proto0 == IP_PROTOCOL_TCP; good_tcp_udp0 = (p0->flags & VNET_BUFFER_F_L4_CHECKSUM_CORRECT) != 0; if (PREDICT_FALSE (ip4_local_do_l4_check (is_tcp_udp0, p0->flags))) { ip4_local_validate_l4 (vm, p0, ip0, is_udp0, &error0, &good_tcp_udp0); } ASSERT (IP4_ERROR_TCP_CHECKSUM + 1 == IP4_ERROR_UDP_CHECKSUM); error0 = (is_tcp_udp0 && !good_tcp_udp0 ? IP4_ERROR_TCP_CHECKSUM + is_udp0 : error0); fib_index0 = vec_elt (im->fib_index_by_sw_if_index, sw_if_index0); fib_index0 = (vnet_buffer (p0)->sw_if_index[VLIB_TX] == (u32) ~ 0) ? fib_index0 : vnet_buffer (p0)->sw_if_index[VLIB_TX]; mtrie0 = &ip4_fib_get (fib_index0)->mtrie; leaf0 = ip4_fib_mtrie_lookup_step_one (mtrie0, &ip0->src_address); leaf0 = ip4_fib_mtrie_lookup_step (mtrie0, leaf0, &ip0->src_address, 2); leaf0 = ip4_fib_mtrie_lookup_step (mtrie0, leaf0, &ip0->src_address, 3); lbi0 = ip4_fib_mtrie_leaf_get_adj_index (leaf0); vnet_buffer (p0)->ip.adj_index[VLIB_TX] = lbi0; vnet_buffer (p0)->ip.adj_index[VLIB_RX] = lbi0; lb0 = load_balance_get (lbi0); dpo0 = load_balance_get_bucket_i (lb0, 0); error0 = ((error0 == IP4_ERROR_UNKNOWN_PROTOCOL && dpo0->dpoi_type == DPO_RECEIVE) ? IP4_ERROR_SPOOFED_LOCAL_PACKETS : error0); error0 = ((error0 == IP4_ERROR_UNKNOWN_PROTOCOL && !fib_urpf_check_size (lb0->lb_urpf) && ip0->dst_address.as_u32 != 0xFFFFFFFF) ? IP4_ERROR_SRC_LOOKUP_MISS : error0); skip_check: next0 = lm->local_next_by_ip_protocol[proto0]; next0 = error0 != IP4_ERROR_UNKNOWN_PROTOCOL ? IP_LOCAL_NEXT_DROP : next0; p0->error = error0 ? error_node->errors[error0] : 0; if (head_of_feature_arc) { if (PREDICT_TRUE (error0 == (u8) IP4_ERROR_UNKNOWN_PROTOCOL)) vnet_feature_arc_start (arc_index, sw_if_index0, &next0, p0); } vlib_validate_buffer_enqueue_x1 (vm, node, next_index, to_next, n_left_to_next, pi0, next0); } vlib_put_next_frame (vm, node, next_index, n_left_to_next); } return frame->n_vectors; } static uword ip4_local (vlib_main_t * vm, vlib_node_runtime_t * node, vlib_frame_t * frame) { return ip4_local_inline (vm, node, frame, 1 /* head of feature arc */ ); } /* *INDENT-OFF* */ VLIB_REGISTER_NODE (ip4_local_node) = { .function = ip4_local, .name = "ip4-local", .vector_size = sizeof (u32), .format_trace = format_ip4_forward_next_trace, .n_next_nodes = IP_LOCAL_N_NEXT, .next_nodes = { [IP_LOCAL_NEXT_DROP] = "error-drop", [IP_LOCAL_NEXT_PUNT] = "error-punt", [IP_LOCAL_NEXT_UDP_LOOKUP] = "ip4-udp-lookup", [IP_LOCAL_NEXT_ICMP] = "ip4-icmp-input", }, }; /* *INDENT-ON* */ VLIB_NODE_FUNCTION_MULTIARCH (ip4_local_node, ip4_local); static uword ip4_local_end_of_arc (vlib_main_t * vm, vlib_node_runtime_t * node, vlib_frame_t * frame) { return ip4_local_inline (vm, node, frame, 0 /* head of feature arc */ ); } /* *INDENT-OFF* */ VLIB_REGISTER_NODE (ip4_local_end_of_arc_node,static) = { .function = ip4_local_end_of_arc, .name = "ip4-local-end-of-arc", .vector_size = sizeof (u32), .format_trace = format_ip4_forward_next_trace, .sibling_of = "ip4-local", }; VLIB_NODE_FUNCTION_MULTIARCH (ip4_local_end_of_arc_node, ip4_local_end_of_arc) VNET_FEATURE_INIT (ip4_local_end_of_arc, static) = { .arc_name = "ip4-local", .node_name = "ip4-local-end-of-arc", .runs_before = 0, /* not before any other features */ }; /* *INDENT-ON* */ void ip4_register_protocol (u32 protocol, u32 node_index) { vlib_main_t *vm = vlib_get_main (); ip4_main_t *im = &ip4_main; ip_lookup_main_t *lm = &im->lookup_main; ASSERT (protocol < ARRAY_LEN (lm->local_next_by_ip_protocol)); lm->local_next_by_ip_protocol[protocol] = vlib_node_add_next (vm, ip4_local_node.index, node_index); } static clib_error_t * show_ip_local_command_fn (vlib_main_t * vm, unformat_input_t * input, vlib_cli_command_t * cmd) { ip4_main_t *im = &ip4_main; ip_lookup_main_t *lm = &im->lookup_main; int i; vlib_cli_output (vm, "Protocols handled by ip4_local"); for (i = 0; i < ARRAY_LEN (lm->local_next_by_ip_protocol); i++) { if (lm->local_next_by_ip_protocol[i] != IP_LOCAL_NEXT_PUNT) { u32 node_index = vlib_get_node (vm, ip4_local_node.index)-> next_nodes[lm->local_next_by_ip_protocol[i]]; vlib_cli_output (vm, "%d: %U", i, format_vlib_node_name, vm, node_index); } } return 0; } /*? * Display the set of protocols handled by the local IPv4 stack. * * @cliexpar * Example of how to display local protocol table: * @cliexstart{show ip local} * Protocols handled by ip4_local * 1 * 17 * 47 * @cliexend ?*/ /* *INDENT-OFF* */ VLIB_CLI_COMMAND (show_ip_local, static) = { .path = "show ip local", .function = show_ip_local_command_fn, .short_help = "show ip local", }; /* *INDENT-ON* */ always_inline uword ip4_arp_inline (vlib_main_t * vm, vlib_node_runtime_t * node, vlib_frame_t * frame, int is_glean) { vnet_main_t *vnm = vnet_get_main (); ip4_main_t *im = &ip4_main; ip_lookup_main_t *lm = &im->lookup_main; u32 *from, *to_next_drop; uword n_left_from, n_left_to_next_drop, next_index; static f64 time_last_seed_change = -1e100; static u32 hash_seeds[3]; static uword hash_bitmap[256 / BITS (uword)]; f64 time_now; if (node->flags & VLIB_NODE_FLAG_TRACE) ip4_forward_next_trace (vm, node, frame, VLIB_TX); time_now = vlib_time_now (vm); if (time_now - time_last_seed_change > 1e-3) { uword i; u32 *r = clib_random_buffer_get_data (&vm->random_buffer, sizeof (hash_seeds)); for (i = 0; i < ARRAY_LEN (hash_seeds); i++) hash_seeds[i] = r[i]; /* Mark all hash keys as been no-seen before. */ for (i = 0; i < ARRAY_LEN (hash_bitmap); i++) hash_bitmap[i] = 0; time_last_seed_change = time_now; } from = vlib_frame_vector_args (frame); n_left_from = frame->n_vectors; next_index = node->cached_next_index; if (next_index == IP4_ARP_NEXT_DROP) next_index = IP4_ARP_N_NEXT; /* point to first interface */ while (n_left_from > 0) { vlib_get_next_frame (vm, node, IP4_ARP_NEXT_DROP, to_next_drop, n_left_to_next_drop); while (n_left_from > 0 && n_left_to_next_drop > 0) { u32 pi0, adj_index0, a0, b0, c0, m0, sw_if_index0, drop0; ip_adjacency_t *adj0; vlib_buffer_t *p0; ip4_header_t *ip0; uword bm0; pi0 = from[0]; p0 = vlib_get_buffer (vm, pi0); adj_index0 = vnet_buffer (p0)->ip.adj_index[VLIB_TX]; adj0 = adj_get (adj_index0); ip0 = vlib_buffer_get_current (p0); a0 = hash_seeds[0]; b0 = hash_seeds[1]; c0 = hash_seeds[2]; sw_if_index0 = adj0->rewrite_header.sw_if_index; vnet_buffer (p0)->sw_if_index[VLIB_TX] = sw_if_index0; if (is_glean) { /* * this is the Glean case, so we are ARPing for the * packet's destination */ a0 ^= ip0->dst_address.data_u32; } else { a0 ^= adj0->sub_type.nbr.next_hop.ip4.data_u32; } b0 ^= sw_if_index0; hash_v3_mix32 (a0, b0, c0); hash_v3_finalize32 (a0, b0, c0); c0 &= BITS (hash_bitmap) - 1; m0 = (uword) 1 << (c0 % BITS (uword)); c0 = c0 / BITS (uword); bm0 = hash_bitmap[c0]; drop0 = (bm0 & m0) != 0; /* Mark it as seen. */ hash_bitmap[c0] = bm0 | m0; from += 1; n_left_from -= 1; to_next_drop[0] = pi0; to_next_drop += 1; n_left_to_next_drop -= 1; p0->error = node->errors[drop0 ? IP4_ARP_ERROR_DROP : IP4_ARP_ERROR_REQUEST_SENT]; /* * the adj has been updated to a rewrite but the node the DPO that got * us here hasn't - yet. no big deal. we'll drop while we wait. */ if (IP_LOOKUP_NEXT_REWRITE == adj0->lookup_next_index) continue; if (drop0) continue; /* * Can happen if the control-plane is programming tables * with traffic flowing; at least that's today's lame excuse. */ if ((is_glean && adj0->lookup_next_index != IP_LOOKUP_NEXT_GLEAN) || (!is_glean && adj0->lookup_next_index != IP_LOOKUP_NEXT_ARP)) { p0->error = node->errors[IP4_ARP_ERROR_NON_ARP_ADJ]; } else /* Send ARP request. */ { u32 bi0 = 0; vlib_buffer_t *b0; ethernet_arp_header_t *h0; vnet_hw_interface_t *hw_if0; h0 = vlib_packet_template_get_packet (vm, &im->ip4_arp_request_packet_template, &bi0); /* Add rewrite/encap string for ARP packet. */ vnet_rewrite_one_header (adj0[0], h0, sizeof (ethernet_header_t)); hw_if0 = vnet_get_sup_hw_interface (vnm, sw_if_index0); /* Src ethernet address in ARP header. */ clib_memcpy (h0->ip4_over_ethernet[0].ethernet, hw_if0->hw_address, sizeof (h0->ip4_over_ethernet[0].ethernet)); if (is_glean) { /* The interface's source address is stashed in the Glean Adj */ h0->ip4_over_ethernet[0].ip4 = adj0->sub_type.glean.receive_addr.ip4; /* Copy in destination address we are requesting. This is the * glean case, so it's the packet's destination.*/ h0->ip4_over_ethernet[1].ip4.data_u32 = ip0->dst_address.data_u32; } else { /* Src IP address in ARP header. */ if (ip4_src_address_for_packet (lm, sw_if_index0, &h0-> ip4_over_ethernet[0].ip4)) { /* No source address available */ p0->error = node->errors[IP4_ARP_ERROR_NO_SOURCE_ADDRESS]; vlib_buffer_free (vm, &bi0, 1); continue; } /* Copy in destination address we are requesting from the incomplete adj */ h0->ip4_over_ethernet[1].ip4.data_u32 = adj0->sub_type.nbr.next_hop.ip4.as_u32; } vlib_buffer_copy_trace_flag (vm, p0, bi0); b0 = vlib_get_buffer (vm, bi0); VLIB_BUFFER_TRACE_TRAJECTORY_INIT (b0); vnet_buffer (b0)->sw_if_index[VLIB_TX] = sw_if_index0; vlib_buffer_advance (b0, -adj0->rewrite_header.data_bytes); vlib_set_next_frame_buffer (vm, node, adj0->rewrite_header.next_index, bi0); } } vlib_put_next_frame (vm, node, IP4_ARP_NEXT_DROP, n_left_to_next_drop); } return frame->n_vectors; } static uword ip4_arp (vlib_main_t * vm, vlib_node_runtime_t * node, vlib_frame_t * frame) { return (ip4_arp_inline (vm, node, frame, 0)); } static uword ip4_glean (vlib_main_t * vm, vlib_node_runtime_t * node, vlib_frame_t * frame) { return (ip4_arp_inline (vm, node, frame, 1)); } static char *ip4_arp_error_strings[] = { [IP4_ARP_ERROR_DROP] = "address overflow drops", [IP4_ARP_ERROR_REQUEST_SENT] = "ARP requests sent", [IP4_ARP_ERROR_NON_ARP_ADJ] = "ARPs to non-ARP adjacencies", [IP4_ARP_ERROR_REPLICATE_DROP] = "ARP replication completed", [IP4_ARP_ERROR_REPLICATE_FAIL] = "ARP replication failed", [IP4_ARP_ERROR_NO_SOURCE_ADDRESS] = "no source address for ARP request", }; VLIB_REGISTER_NODE (ip4_arp_node) = { .function = ip4_arp,.name = "ip4-arp",.vector_size = sizeof (u32),.format_trace = format_ip4_forward_next_trace,.n_errors = ARRAY_LEN (ip4_arp_error_strings),.error_strings = ip4_arp_error_strings,.n_next_nodes = IP4_ARP_N_NEXT,.next_nodes = { [IP4_ARP_NEXT_DROP] = "error-drop",} ,}; VLIB_REGISTER_NODE (ip4_glean_node) = { .function = ip4_glean,.name = "ip4-glean",.vector_size = sizeof (u32),.format_trace = format_ip4_forward_next_trace,.n_errors = ARRAY_LEN (ip4_arp_error_strings),.error_strings = ip4_arp_error_strings,.n_next_nodes = IP4_ARP_N_NEXT,.next_nodes = { [IP4_ARP_NEXT_DROP] = "error-drop",} ,}; #define foreach_notrace_ip4_arp_error \ _(DROP) \ _(REQUEST_SENT) \ _(REPLICATE_DROP) \ _(REPLICATE_FAIL) clib_error_t * arp_notrace_init (vlib_main_t * vm) { vlib_node_runtime_t *rt = vlib_node_get_runtime (vm, ip4_arp_node.index); /* don't trace ARP request packets */ #define _(a) \ vnet_pcap_drop_trace_filter_add_del \ (rt->errors[IP4_ARP_ERROR_##a], \ 1 /* is_add */); foreach_notrace_ip4_arp_error; #undef _ return 0; } VLIB_INIT_FUNCTION (arp_notrace_init); /* Send an ARP request to see if given destination is reachable on given interface. */ clib_error_t * ip4_probe_neighbor (vlib_main_t * vm, ip4_address_t * dst, u32 sw_if_index) { vnet_main_t *vnm = vnet_get_main (); ip4_main_t *im = &ip4_main; ethernet_arp_header_t *h; ip4_address_t *src; ip_interface_address_t *ia; ip_adjacency_t *adj; vnet_hw_interface_t *hi; vnet_sw_interface_t *si; vlib_buffer_t *b; adj_index_t ai; u32 bi = 0; si = vnet_get_sw_interface (vnm, sw_if_index); if (!(si->flags & VNET_SW_INTERFACE_FLAG_ADMIN_UP)) { return clib_error_return (0, "%U: interface %U down", format_ip4_address, dst, format_vnet_sw_if_index_name, vnm, sw_if_index); } src = ip4_interface_address_matching_destination (im, dst, sw_if_index, &ia); if (!src) { vnm->api_errno = VNET_API_ERROR_NO_MATCHING_INTERFACE; return clib_error_return (0, "no matching interface address for destination %U (interface %U)", format_ip4_address, dst, format_vnet_sw_if_index_name, vnm, sw_if_index); } ip46_address_t nh = { .ip4 = *dst, }; ai = adj_nbr_add_or_lock (FIB_PROTOCOL_IP4, VNET_LINK_IP4, &nh, sw_if_index); adj = adj_get (ai); h = vlib_packet_template_get_packet (vm, &im->ip4_arp_request_packet_template, &bi); hi = vnet_get_sup_hw_interface (vnm, sw_if_index); if (PREDICT_FALSE (!hi->hw_address)) { return clib_error_return (0, "%U: interface %U do not support ip probe", format_ip4_address, dst, format_vnet_sw_if_index_name, vnm, sw_if_index); } clib_memcpy (h->ip4_over_ethernet[0].ethernet, hi->hw_address, sizeof (h->ip4_over_ethernet[0].ethernet)); h->ip4_over_ethernet[0].ip4 = src[0]; h->ip4_over_ethernet[1].ip4 = dst[0]; b = vlib_get_buffer (vm, bi); vnet_buffer (b)->sw_if_index[VLIB_RX] = vnet_buffer (b)->sw_if_index[VLIB_TX] = sw_if_index; /* Add encapsulation string for software interface (e.g. ethernet header). */ vnet_rewrite_one_header (adj[0], h, sizeof (ethernet_header_t)); vlib_buffer_advance (b, -adj->rewrite_header.data_bytes); { vlib_frame_t *f = vlib_get_frame_to_node (vm, hi->output_node_index); u32 *to_next = vlib_frame_vector_args (f); to_next[0] = bi; f->n_vectors = 1; vlib_put_frame_to_node (vm, hi->output_node_index, f); } adj_unlock (ai); return /* no error */ 0; } typedef enum { IP4_REWRITE_NEXT_DROP, IP4_REWRITE_NEXT_ICMP_ERROR, } ip4_rewrite_next_t; always_inline uword ip4_rewrite_inline (vlib_main_t * vm, vlib_node_runtime_t * node, vlib_frame_t * frame, int do_counters, int is_midchain, int is_mcast) { ip_lookup_main_t *lm = &ip4_main.lookup_main; u32 *from = vlib_frame_vector_args (frame); u32 n_left_from, n_left_to_next, *to_next, next_index; vlib_node_runtime_t *error_node = vlib_node_get_runtime (vm, ip4_input_node.index); n_left_from = frame->n_vectors; next_index = node->cached_next_index; u32 thread_index = vlib_get_thread_index (); while (n_left_from > 0) { vlib_get_next_frame (vm, node, next_index, to_next, n_left_to_next); while (n_left_from >= 4 && n_left_to_next >= 2) { ip_adjacency_t *adj0, *adj1; vlib_buffer_t *p0, *p1; ip4_header_t *ip0, *ip1; u32 pi0, rw_len0, next0, error0, checksum0, adj_index0; u32 pi1, rw_len1, next1, error1, checksum1, adj_index1; u32 tx_sw_if_index0, tx_sw_if_index1; /* 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, STORE); vlib_prefetch_buffer_header (p3, STORE); CLIB_PREFETCH (p2->data, sizeof (ip0[0]), STORE); CLIB_PREFETCH (p3->data, sizeof (ip0[0]), STORE); } pi0 = to_next[0] = from[0]; pi1 = to_next[1] = from[1]; from += 2; n_left_from -= 2; to_next += 2; n_left_to_next -= 2; p0 = vlib_get_buffer (vm, pi0); p1 = vlib_get_buffer (vm, pi1); adj_index0 = vnet_buffer (p0)->ip.adj_index[VLIB_TX]; adj_index1 = vnet_buffer (p1)->ip.adj_index[VLIB_TX]; /* * pre-fetch the per-adjacency counters */ if (do_counters) { vlib_prefetch_combined_counter (&adjacency_counters, thread_index, adj_index0); vlib_prefetch_combined_counter (&adjacency_counters, thread_index, adj_index1); } ip0 = vlib_buffer_get_current (p0); ip1 = vlib_buffer_get_current (p1); error0 = error1 = IP4_ERROR_NONE; next0 = next1 = IP4_REWRITE_NEXT_DROP; /* Decrement TTL & update checksum. Works either endian, so no need for byte swap. */ if (PREDICT_TRUE (!(p0->flags & VNET_BUFFER_F_LOCALLY_ORIGINATED))) { i32 ttl0 = ip0->ttl; /* Input node should have reject packets with ttl 0. */ ASSERT (ip0->ttl > 0); checksum0 = ip0->checksum + clib_host_to_net_u16 (0x0100); checksum0 += checksum0 >= 0xffff; ip0->checksum = checksum0; ttl0 -= 1; ip0->ttl = ttl0; /* * If the ttl drops below 1 when forwarding, generate * an ICMP response. */ if (PREDICT_FALSE (ttl0 <= 0)) { error0 = IP4_ERROR_TIME_EXPIRED; vnet_buffer (p0)->sw_if_index[VLIB_TX] = (u32) ~ 0; icmp4_error_set_vnet_buffer (p0, ICMP4_time_exceeded, ICMP4_time_exceeded_ttl_exceeded_in_transit, 0); next0 = IP4_REWRITE_NEXT_ICMP_ERROR; } /* Verify checksum. */ ASSERT ((ip0->checksum == ip4_header_checksum (ip0)) || (p0->flags & VNET_BUFFER_F_OFFLOAD_IP_CKSUM)); } else { p0->flags &= ~VNET_BUFFER_F_LOCALLY_ORIGINATED; } if (PREDICT_TRUE (!(p1->flags & VNET_BUFFER_F_LOCALLY_ORIGINATED))) { i32 ttl1 = ip1->ttl; /* Input node should have reject packets with ttl 0. */ ASSERT (ip1->ttl > 0); checksum1 = ip1->checksum + clib_host_to_net_u16 (0x0100); checksum1 += checksum1 >= 0xffff; ip1->checksum = checksum1; ttl1 -= 1; ip1->ttl = ttl1; /* * If the ttl drops below 1 when forwarding, generate * an ICMP response. */ if (PREDICT_FALSE (ttl1 <= 0)) { error1 = IP4_ERROR_TIME_EXPIRED; vnet_buffer (p1)->sw_if_index[VLIB_TX] = (u32) ~ 0; icmp4_error_set_vnet_buffer (p1, ICMP4_time_exceeded, ICMP4_time_exceeded_ttl_exceeded_in_transit, 0); next1 = IP4_REWRITE_NEXT_ICMP_ERROR; } /* Verify checksum. */ ASSERT ((ip1->checksum == ip4_header_checksum (ip1)) || (p1->flags & VNET_BUFFER_F_OFFLOAD_IP_CKSUM)); } else { p1->flags &= ~VNET_BUFFER_F_LOCALLY_ORIGINATED; } /* Rewrite packet header and updates lengths. */ adj0 = adj_get (adj_index0); adj1 = adj_get (adj_index1); /* Worth pipelining. No guarantee that adj0,1 are hot... */ rw_len0 = adj0[0].rewrite_header.data_bytes; rw_len1 = adj1[0].rewrite_header.data_bytes; vnet_buffer (p0)->ip.save_rewrite_length = rw_len0; vnet_buffer (p1)->ip.save_rewrite_length = rw_len1; /* Check MTU of outgoing interface. */ error0 = (vlib_buffer_length_in_chain (vm, p0) > adj0[0]. rewrite_header.max_l3_packet_bytes ? IP4_ERROR_MTU_EXCEEDED : error0); error1 = (vlib_buffer_length_in_chain (vm, p1) > adj1[0]. rewrite_header.max_l3_packet_bytes ? IP4_ERROR_MTU_EXCEEDED : error1); /* Don't adjust the buffer for ttl issue; icmp-error node wants * to see the IP headerr */ if (PREDICT_TRUE (error0 == IP4_ERROR_NONE)) { next0 = adj0[0].rewrite_header.next_index; p0->current_data -= rw_len0; p0->current_length += rw_len0; tx_sw_if_index0 = adj0[0].rewrite_header.sw_if_index; vnet_buffer (p0)->sw_if_index[VLIB_TX] = tx_sw_if_index0; if (PREDICT_FALSE (adj0[0].rewrite_header.flags & VNET_REWRITE_HAS_FEATURES)) vnet_feature_arc_start (lm->output_feature_arc_index, tx_sw_if_index0, &next0, p0); } if (PREDICT_TRUE (error1 == IP4_ERROR_NONE)) { next1 = adj1[0].rewrite_header.next_index; p1->current_data -= rw_len1; p1->current_length += rw_len1; tx_sw_if_index1 = adj1[0].rewrite_header.sw_if_index; vnet_buffer (p1)->sw_if_index[VLIB_TX] = tx_sw_if_index1; if (PREDICT_FALSE (adj1[0].rewrite_header.flags & VNET_REWRITE_HAS_FEATURES)) vnet_feature_arc_start (lm->output_feature_arc_index, tx_sw_if_index1, &next1, p1); } /* Guess we are only writing on simple Ethernet header. */ vnet_rewrite_two_headers (adj0[0], adj1[0], ip0, ip1, sizeof (ethernet_header_t)); /* * Bump the per-adjacency counters */ if (do_counters) { vlib_increment_combined_counter (&adjacency_counters, thread_index, adj_index0, 1, vlib_buffer_length_in_chain (vm, p0) + rw_len0); vlib_increment_combined_counter (&adjacency_counters, thread_index, adj_index1, 1, vlib_buffer_length_in_chain (vm, p1) + rw_len1); } if (is_midchain) { adj0->sub_type.midchain.fixup_func (vm, adj0, p0); adj1->sub_type.midchain.fixup_func (vm, adj1, p1); } if (is_mcast) { /* * copy bytes from the IP address into the MAC rewrite */ vnet_fixup_one_header (adj0[0], &ip0->dst_address, ip0); vnet_fixup_one_header (adj1[0], &ip1->dst_address, ip1); } vlib_validate_buffer_enqueue_x2 (vm, node, next_index, to_next, n_left_to_next, pi0, pi1, next0, next1); } while (n_left_from > 0 && n_left_to_next > 0) { ip_adjacency_t *adj0; vlib_buffer_t *p0; ip4_header_t *ip0; u32 pi0, rw_len0, adj_index0, next0, error0, checksum0; u32 tx_sw_if_index0; pi0 = to_next[0] = from[0]; p0 = vlib_get_buffer (vm, pi0); adj_index0 = vnet_buffer (p0)->ip.adj_index[VLIB_TX]; adj0 = adj_get (adj_index0); ip0 = vlib_buffer_get_current (p0); error0 = IP4_ERROR_NONE; next0 = IP4_REWRITE_NEXT_DROP; /* drop on error */ /* Decrement TTL & update checksum. */ if (PREDICT_TRUE (!(p0->flags & VNET_BUFFER_F_LOCALLY_ORIGINATED))) { i32 ttl0 = ip0->ttl; checksum0 = ip0->checksum + clib_host_to_net_u16 (0x0100); checksum0 += checksum0 >= 0xffff; ip0->checksum = checksum0; ASSERT (ip0->ttl > 0); ttl0 -= 1; ip0->ttl = ttl0; ASSERT ((ip0->checksum == ip4_header_checksum (ip0)) || (p0->flags & VNET_BUFFER_F_OFFLOAD_IP_CKSUM)); if (PREDICT_FALSE (ttl0 <= 0)) { /* * If the ttl drops below 1 when forwarding, generate * an ICMP response. */ error0 = IP4_ERROR_TIME_EXPIRED; next0 = IP4_REWRITE_NEXT_ICMP_ERROR; vnet_buffer (p0)->sw_if_index[VLIB_TX] = (u32) ~ 0; icmp4_error_set_vnet_buffer (p0, ICMP4_time_exceeded, ICMP4_time_exceeded_ttl_exceeded_in_transit, 0); } } else { p0->flags &= ~VNET_BUFFER_F_LOCALLY_ORIGINATED; } if (do_counters) vlib_prefetch_combined_counter (&adjacency_counters, thread_index, adj_index0); /* Guess we are only writing on simple Ethernet header. */ vnet_rewrite_one_header (adj0[0], ip0, sizeof (ethernet_header_t)); if (is_mcast) { /* * copy bytes from the IP address into the MAC rewrite */ vnet_fixup_one_header (adj0[0], &ip0->dst_address, ip0); } /* Update packet buffer attributes/set output interface. */ rw_len0 = adj0[0].rewrite_header.data_bytes; vnet_buffer (p0)->ip.save_rewrite_length = rw_len0; if (do_counters) vlib_increment_combined_counter (&adjacency_counters, thread_index, adj_index0, 1, vlib_buffer_length_in_chain (vm, p0) + rw_len0); /* Check MTU of outgoing interface. */ error0 = (vlib_buffer_length_in_chain (vm, p0) > adj0[0].rewrite_header.max_l3_packet_bytes ? IP4_ERROR_MTU_EXCEEDED : error0); p0->error = error_node->errors[error0]; /* Don't adjust the buffer for ttl issue; icmp-error node wants * to see the IP headerr */ if (PREDICT_TRUE (error0 == IP4_ERROR_NONE)) { p0->current_data -= rw_len0; p0->current_length += rw_len0; tx_sw_if_index0 = adj0[0].rewrite_header.sw_if_index; vnet_buffer (p0)->sw_if_index[VLIB_TX] = tx_sw_if_index0; next0 = adj0[0].rewrite_header.next_index; if (is_midchain) { adj0->sub_type.midchain.fixup_func (vm, adj0, p0); } if (PREDICT_FALSE (adj0[0].rewrite_header.flags & VNET_REWRITE_HAS_FEATURES)) vnet_feature_arc_start (lm->output_feature_arc_index, tx_sw_if_index0, &next0, p0); } from += 1; n_left_from -= 1; to_next += 1; n_left_to_next -= 1; vlib_validate_buffer_enqueue_x1 (vm, node, next_index, to_next, n_left_to_next, pi0, next0); } vlib_put_next_frame (vm, node, next_index, n_left_to_next); } /* Need to do trace after rewrites to pick up new packet data. */ if (node->flags & VLIB_NODE_FLAG_TRACE) ip4_forward_next_trace (vm, node, frame, VLIB_TX); return frame->n_vectors; } /** @brief IPv4 rewrite node. @node ip4-rewrite This is the IPv4 transit-rewrite node: decrement TTL, fix the ipv4 header checksum, fetch the ip adjacency, check the outbound mtu, apply the adjacency rewrite, and send pkts to the adjacency rewrite header's rewrite_next_index. @param vm vlib_main_t corresponding to the current thread @param node vlib_node_runtime_t @param frame vlib_frame_t whose contents should be dispatched @par Graph mechanics: buffer metadata, next index usage @em Uses: - <code>vnet_buffer(b)->ip.adj_index[VLIB_TX]</code> - the rewrite adjacency index - <code>adj->lookup_next_index</code> - Must be IP_LOOKUP_NEXT_REWRITE or IP_LOOKUP_NEXT_ARP, otherwise the packet will be dropped. - <code>adj->rewrite_header</code> - Rewrite string length, rewrite string, next_index @em Sets: - <code>b->current_data, b->current_length</code> - Updated net of applying the rewrite string <em>Next Indices:</em> - <code> adj->rewrite_header.next_index </code> or @c error-drop */ static uword ip4_rewrite (vlib_main_t * vm, vlib_node_runtime_t * node, vlib_frame_t * frame) { if (adj_are_counters_enabled ()) return ip4_rewrite_inline (vm, node, frame, 1, 0, 0); else return ip4_rewrite_inline (vm, node, frame, 0, 0, 0); } static uword ip4_midchain (vlib_main_t * vm, vlib_node_runtime_t * node, vlib_frame_t * frame) { if (adj_are_counters_enabled ()) return ip4_rewrite_inline (vm, node, frame, 1, 1, 0); else return ip4_rewrite_inline (vm, node, frame, 0, 1, 0); } static uword ip4_rewrite_mcast (vlib_main_t * vm, vlib_node_runtime_t * node, vlib_frame_t * frame) { if (adj_are_counters_enabled ()) return ip4_rewrite_inline (vm, node, frame, 1, 0, 1); else return ip4_rewrite_inline (vm, node, frame, 0, 0, 1); } static uword ip4_mcast_midchain (vlib_main_t * vm, vlib_node_runtime_t * node, vlib_frame_t * frame) { if (adj_are_counters_enabled ()) return ip4_rewrite_inline (vm, node, frame, 1, 1, 1); else return ip4_rewrite_inline (vm, node, frame, 0, 1, 1); } /* *INDENT-OFF* */ VLIB_REGISTER_NODE (ip4_rewrite_node) = { .function = ip4_rewrite, .name = "ip4-rewrite", .vector_size = sizeof (u32), .format_trace = format_ip4_rewrite_trace, .n_next_nodes = 2, .next_nodes = { [IP4_REWRITE_NEXT_DROP] = "error-drop", [IP4_REWRITE_NEXT_ICMP_ERROR] = "ip4-icmp-error", }, }; VLIB_NODE_FUNCTION_MULTIARCH (ip4_rewrite_node, ip4_rewrite) VLIB_REGISTER_NODE (ip4_rewrite_mcast_node) = { .function = ip4_rewrite_mcast, .name = "ip4-rewrite-mcast", .vector_size = sizeof (u32), .format_trace = format_ip4_rewrite_trace, .sibling_of = "ip4-rewrite", }; VLIB_NODE_FUNCTION_MULTIARCH (ip4_rewrite_mcast_node, ip4_rewrite_mcast) VLIB_REGISTER_NODE (ip4_mcast_midchain_node, static) = { .function = ip4_mcast_midchain, .name = "ip4-mcast-midchain", .vector_size = sizeof (u32), .format_trace = format_ip4_rewrite_trace, .sibling_of = "ip4-rewrite", }; VLIB_NODE_FUNCTION_MULTIARCH (ip4_mcast_midchain_node, ip4_mcast_midchain) VLIB_REGISTER_NODE (ip4_midchain_node) = { .function = ip4_midchain, .name = "ip4-midchain", .vector_size = sizeof (u32), .format_trace = format_ip4_forward_next_trace, .sibling_of = "ip4-rewrite", }; VLIB_NODE_FUNCTION_MULTIARCH (ip4_midchain_node, ip4_midchain); /* *INDENT-ON */ int ip4_lookup_validate (ip4_address_t * a, u32 fib_index0) { ip4_fib_mtrie_t *mtrie0; ip4_fib_mtrie_leaf_t leaf0; u32 lbi0; mtrie0 = &ip4_fib_get (fib_index0)->mtrie; leaf0 = ip4_fib_mtrie_lookup_step_one (mtrie0, a); leaf0 = ip4_fib_mtrie_lookup_step (mtrie0, leaf0, a, 2); leaf0 = ip4_fib_mtrie_lookup_step (mtrie0, leaf0, a, 3); lbi0 = ip4_fib_mtrie_leaf_get_adj_index (leaf0); return lbi0 == ip4_fib_table_lookup_lb (ip4_fib_get (fib_index0), a); } static clib_error_t * test_lookup_command_fn (vlib_main_t * vm, unformat_input_t * input, vlib_cli_command_t * cmd) { ip4_fib_t *fib; u32 table_id = 0; f64 count = 1; u32 n; int i; ip4_address_t ip4_base_address; u64 errors = 0; while (unformat_check_input (input) != UNFORMAT_END_OF_INPUT) { if (unformat (input, "table %d", &table_id)) { /* Make sure the entry exists. */ fib = ip4_fib_get (table_id); if ((fib) && (fib->index != table_id)) return clib_error_return (0, "<fib-index> %d does not exist", table_id); } else if (unformat (input, "count %f", &count)) ; else if (unformat (input, "%U", unformat_ip4_address, &ip4_base_address)) ; else return clib_error_return (0, "unknown input `%U'", format_unformat_error, input); } n = count; for (i = 0; i < n; i++) { if (!ip4_lookup_validate (&ip4_base_address, table_id)) errors++; ip4_base_address.as_u32 = clib_host_to_net_u32 (1 + clib_net_to_host_u32 (ip4_base_address.as_u32)); } if (errors) vlib_cli_output (vm, "%llu errors out of %d lookups\n", errors, n); else vlib_cli_output (vm, "No errors in %d lookups\n", n); return 0; } /*? * Perform a lookup of an IPv4 Address (or range of addresses) in the * given FIB table to determine if there is a conflict with the * adjacency table. The fib-id can be determined by using the * '<em>show ip fib</em>' command. If fib-id is not entered, default value * of 0 is used. * * @todo This command uses fib-id, other commands use table-id (not * just a name, they are different indexes). Would like to change this * to table-id for consistency. * * @cliexpar * Example of how to run the test lookup command: * @cliexstart{test lookup 172.16.1.1 table 1 count 2} * No errors in 2 lookups * @cliexend ?*/ /* *INDENT-OFF* */ VLIB_CLI_COMMAND (lookup_test_command, static) = { .path = "test lookup", .short_help = "test lookup <ipv4-addr> [table <fib-id>] [count <nn>]", .function = test_lookup_command_fn, }; /* *INDENT-ON* */ int vnet_set_ip4_flow_hash (u32 table_id, u32 flow_hash_config) { u32 fib_index; fib_index = fib_table_find (FIB_PROTOCOL_IP4, table_id); if (~0 == fib_index) return VNET_API_ERROR_NO_SUCH_FIB; fib_table_set_flow_hash_config (fib_index, FIB_PROTOCOL_IP4, flow_hash_config); return 0; } static clib_error_t * set_ip_flow_hash_command_fn (vlib_main_t * vm, unformat_input_t * input, vlib_cli_command_t * cmd) { int matched = 0; u32 table_id = 0; u32 flow_hash_config = 0; int rv; while (unformat_check_input (input) != UNFORMAT_END_OF_INPUT) { if (unformat (input, "table %d", &table_id)) matched = 1; #define _(a,v) \ else if (unformat (input, #a)) { flow_hash_config |= v; matched=1;} foreach_flow_hash_bit #undef _ else break; } if (matched == 0) return clib_error_return (0, "unknown input `%U'", format_unformat_error, input); rv = vnet_set_ip4_flow_hash (table_id, flow_hash_config); switch (rv) { case 0: break; case VNET_API_ERROR_NO_SUCH_FIB: return clib_error_return (0, "no such FIB table %d", table_id); default: clib_warning ("BUG: illegal flow hash config 0x%x", flow_hash_config); break; } return 0; } /*? * Configure the set of IPv4 fields used by the flow hash. * * @cliexpar * Example of how to set the flow hash on a given table: * @cliexcmd{set ip flow-hash table 7 dst sport dport proto} * Example of display the configured flow hash: * @cliexstart{show ip fib} * ipv4-VRF:0, fib_index 0, flow hash: src dst sport dport proto * 0.0.0.0/0 * unicast-ip4-chain * [@0]: dpo-load-balance: [index:0 buckets:1 uRPF:0 to:[0:0]] * [0] [@0]: dpo-drop ip6 * 0.0.0.0/32 * unicast-ip4-chain * [@0]: dpo-load-balance: [index:1 buckets:1 uRPF:1 to:[0:0]] * [0] [@0]: dpo-drop ip6 * 224.0.0.0/8 * unicast-ip4-chain * [@0]: dpo-load-balance: [index:3 buckets:1 uRPF:3 to:[0:0]] * [0] [@0]: dpo-drop ip6 * 6.0.1.2/32 * unicast-ip4-chain * [@0]: dpo-load-balance: [index:30 buckets:1 uRPF:29 to:[0:0]] * [0] [@3]: arp-ipv4: via 6.0.0.1 af_packet0 * 7.0.0.1/32 * unicast-ip4-chain * [@0]: dpo-load-balance: [index:31 buckets:4 uRPF:30 to:[0:0]] * [0] [@3]: arp-ipv4: via 6.0.0.2 af_packet0 * [1] [@3]: arp-ipv4: via 6.0.0.2 af_packet0 * [2] [@3]: arp-ipv4: via 6.0.0.2 af_packet0 * [3] [@3]: arp-ipv4: via 6.0.0.1 af_packet0 * 240.0.0.0/8 * unicast-ip4-chain * [@0]: dpo-load-balance: [index:2 buckets:1 uRPF:2 to:[0:0]] * [0] [@0]: dpo-drop ip6 * 255.255.255.255/32 * unicast-ip4-chain * [@0]: dpo-load-balance: [index:4 buckets:1 uRPF:4 to:[0:0]] * [0] [@0]: dpo-drop ip6 * ipv4-VRF:7, fib_index 1, flow hash: dst sport dport proto * 0.0.0.0/0 * unicast-ip4-chain * [@0]: dpo-load-balance: [index:12 buckets:1 uRPF:11 to:[0:0]] * [0] [@0]: dpo-drop ip6 * 0.0.0.0/32 * unicast-ip4-chain * [@0]: dpo-load-balance: [index:13 buckets:1 uRPF:12 to:[0:0]] * [0] [@0]: dpo-drop ip6 * 172.16.1.0/24 * unicast-ip4-chain * [@0]: dpo-load-balance: [index:17 buckets:1 uRPF:16 to:[0:0]] * [0] [@4]: ipv4-glean: af_packet0 * 172.16.1.1/32 * unicast-ip4-chain * [@0]: dpo-load-balance: [index:18 buckets:1 uRPF:17 to:[1:84]] * [0] [@2]: dpo-receive: 172.16.1.1 on af_packet0 * 172.16.1.2/32 * unicast-ip4-chain * [@0]: dpo-load-balance: [index:21 buckets:1 uRPF:20 to:[0:0]] * [0] [@5]: ipv4 via 172.16.1.2 af_packet0: IP4: 02:fe:9e:70:7a:2b -> 26:a5:f6:9c:3a:36 * 172.16.2.0/24 * unicast-ip4-chain * [@0]: dpo-load-balance: [index:19 buckets:1 uRPF:18 to:[0:0]] * [0] [@4]: ipv4-glean: af_packet1 * 172.16.2.1/32 * unicast-ip4-chain * [@0]: dpo-load-balance: [index:20 buckets:1 uRPF:19 to:[0:0]] * [0] [@2]: dpo-receive: 172.16.2.1 on af_packet1 * 224.0.0.0/8 * unicast-ip4-chain * [@0]: dpo-load-balance: [index:15 buckets:1 uRPF:14 to:[0:0]] * [0] [@0]: dpo-drop ip6 * 240.0.0.0/8 * unicast-ip4-chain * [@0]: dpo-load-balance: [index:14 buckets:1 uRPF:13 to:[0:0]] * [0] [@0]: dpo-drop ip6 * 255.255.255.255/32 * unicast-ip4-chain * [@0]: dpo-load-balance: [index:16 buckets:1 uRPF:15 to:[0:0]] * [0] [@0]: dpo-drop ip6 * @cliexend ?*/ /* *INDENT-OFF* */ VLIB_CLI_COMMAND (set_ip_flow_hash_command, static) = { .path = "set ip flow-hash", .short_help = "set ip flow-hash table <table-id> [src] [dst] [sport] [dport] [proto] [reverse]", .function = set_ip_flow_hash_command_fn, }; /* *INDENT-ON* */ int vnet_set_ip4_classify_intfc (vlib_main_t * vm, u32 sw_if_index, u32 table_index) { vnet_main_t *vnm = vnet_get_main (); vnet_interface_main_t *im = &vnm->interface_main; ip4_main_t *ipm = &ip4_main; ip_lookup_main_t *lm = &ipm->lookup_main; vnet_classify_main_t *cm = &vnet_classify_main; ip4_address_t *if_addr; if (pool_is_free_index (im->sw_interfaces, sw_if_index)) return VNET_API_ERROR_NO_MATCHING_INTERFACE; if (table_index != ~0 && pool_is_free_index (cm->tables, table_index)) return VNET_API_ERROR_NO_SUCH_ENTRY; vec_validate (lm->classify_table_index_by_sw_if_index, sw_if_index); lm->classify_table_index_by_sw_if_index[sw_if_index] = table_index; if_addr = ip4_interface_first_address (ipm, sw_if_index, NULL); if (NULL != if_addr) { fib_prefix_t pfx = { .fp_len = 32, .fp_proto = FIB_PROTOCOL_IP4, .fp_addr.ip4 = *if_addr, }; u32 fib_index; fib_index = fib_table_get_index_for_sw_if_index (FIB_PROTOCOL_IP4, sw_if_index); if (table_index != (u32) ~ 0) { dpo_id_t dpo = DPO_INVALID; dpo_set (&dpo, DPO_CLASSIFY, DPO_PROTO_IP4, classify_dpo_create (DPO_PROTO_IP4, table_index)); fib_table_entry_special_dpo_add (fib_index, &pfx, FIB_SOURCE_CLASSIFY, FIB_ENTRY_FLAG_NONE, &dpo); dpo_reset (&dpo); } else { fib_table_entry_special_remove (fib_index, &pfx, FIB_SOURCE_CLASSIFY); } } return 0; } static clib_error_t * set_ip_classify_command_fn (vlib_main_t * vm, unformat_input_t * input, vlib_cli_command_t * cmd) { u32 table_index = ~0; int table_index_set = 0; u32 sw_if_index = ~0; int rv; while (unformat_check_input (input) != UNFORMAT_END_OF_INPUT) { if (unformat (input, "table-index %d", &table_index)) table_index_set = 1; else if (unformat (input, "intfc %U", unformat_vnet_sw_interface, vnet_get_main (), &sw_if_index)) ; else break; } if (table_index_set == 0) return clib_error_return (0, "classify table-index must be specified"); if (sw_if_index == ~0) return clib_error_return (0, "interface / subif must be specified"); rv = vnet_set_ip4_classify_intfc (vm, sw_if_index, table_index); switch (rv) { case 0: break; case VNET_API_ERROR_NO_MATCHING_INTERFACE: return clib_error_return (0, "No such interface"); case VNET_API_ERROR_NO_SUCH_ENTRY: return clib_error_return (0, "No such classifier table"); } return 0; } /*? * Assign a classification table to an interface. The classification * table is created using the '<em>classify table</em>' and '<em>classify session</em>' * commands. Once the table is create, use this command to filter packets * on an interface. * * @cliexpar * Example of how to assign a classification table to an interface: * @cliexcmd{set ip classify intfc GigabitEthernet2/0/0 table-index 1} ?*/ /* *INDENT-OFF* */ VLIB_CLI_COMMAND (set_ip_classify_command, static) = { .path = "set ip classify", .short_help = "set ip classify intfc <interface> table-index <classify-idx>", .function = set_ip_classify_command_fn, }; /* *INDENT-ON* */ /* * fd.io coding-style-patch-verification: ON * * Local Variables: * eval: (c-set-style "gnu") * End: */