-
Notifications
You must be signed in to change notification settings - Fork 171
/
dpdk_vrouter.c
1179 lines (1043 loc) · 38.6 KB
/
dpdk_vrouter.c
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
/*
* Copyright (C) 2014 Semihalf.
*
* This program is free software; you can redistribute it and/or
* modify it under the terms of the GNU General Public License as
* published by the Free Software Foundation version 2.
*
* This program is distributed "as is" WITHOUT ANY WARRANTY of any
* kind, whether express or implied; without even the implied warranty
* of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* dpdk_vrouter.c -- vRouter/DPDK application
*
*/
/* For sched_getaffinity() */
#define _GNU_SOURCE
#include <stdint.h>
#include <getopt.h>
#include <signal.h>
#include <sys/time.h>
#include "vr_dpdk.h"
#include "vr_dpdk_usocket.h"
#include "vr_dpdk_virtio.h"
#include "vr_uvhost.h"
#include "vr_bridge.h"
#include "vr_mem.h"
#include "nl_util.h"
#include <rte_errno.h>
#include <rte_ethdev.h>
#include <rte_kni.h>
#include <rte_timer.h>
/* vRouter/DPDK command-line options. */
enum vr_opt_index {
#define NO_DAEMON_OPT "no-daemon"
NO_DAEMON_OPT_INDEX,
#define NO_HUGE_OPT "no-huge"
NO_HUGE_OPT_INDEX,
#define HELP_OPT "help"
HELP_OPT_INDEX,
#define VERSION_OPT "version"
VERSION_OPT_INDEX,
#define MEMPOOL_SIZE_OPT "vr_mempool_sz"
MEMPOOL_SIZE_OPT_INDEX,
#define PACKET_SIZE_OPT "vr_packet_sz"
PACKET_SIZE_OPT_INDEX,
#define VLAN_TCI_OPT "vlan_tci"
VLAN_TCI_OPT_INDEX,
#define VLAN_NAME_OPT "vlan_fwd_intf_name"
VLAN_NAME_OPT_INDEX,
#define VTEST_VLAN_OPT "vtest_vlan"
VTEST_VLAN_OPT_INDEX,
#define VDEV_OPT "vdev"
VDEV_OPT_INDEX,
#define BRIDGE_ENTRIES_OPT "vr_bridge_entries"
BRIDGE_ENTRIES_OPT_INDEX,
#define BRIDGE_OENTRIES_OPT "vr_bridge_oentries"
BRIDGE_OENTRIES_OPT_INDEX,
#define FLOW_ENTRIES_OPT "vr_flow_entries"
FLOW_ENTRIES_OPT_INDEX,
#define OFLOW_ENTRIES_OPT "vr_oflow_entries"
OFLOW_ENTRIES_OPT_INDEX,
#define MPLS_LABELS_OPT "vr_mpls_labels"
MPLS_LABELS_OPT_INDEX,
#define NEXTHOPS_OPT "vr_nexthops"
NEXTHOPS_OPT_INDEX,
#define VRFS_OPT "vr_vrfs"
VRFS_OPT_INDEX,
#define SOCKET_DIR_OPT "vr_socket_dir"
SOCKET_DIR_OPT_INDEX,
#define NETLINK_PORT_OPT "vr_netlink_port"
NETLINK_PORT_OPT_INDEX,
#define SOCKET_MEM_OPT "socket-mem"
SOCKET_MEM_OPT_INDEX,
#define LCORES_OPT "lcores"
LCORES_OPT_INDEX,
MAX_OPT_INDEX
};
/* dp-core parameters */
extern unsigned int vr_bridge_entries;
extern unsigned int vr_bridge_oentries;
extern unsigned int vr_mpls_labels;
extern unsigned int vr_nexthops;
extern unsigned int vr_vrfs;
static int no_daemon_set;
int no_huge_set;
unsigned int vr_mempool_sz = VR_DEF_MEMPOOL_SZ;
unsigned int vr_packet_sz = VR_DEF_MAX_PACKET_SZ;
extern char *ContrailBuildInfo;
/* Global vRouter/DPDK structure */
struct vr_dpdk_global vr_dpdk;
/* EAL command line options for rte_eal_init() */
static char *dpdk_argv[] = {
"dpdk",
"-n", VR_DPDK_MAX_MEMCHANNELS,
"-m", VR_DPDK_DEF_MEM,
/* Up to 5 pairs of argument-value (for vdev, lcores, socket-mem options). */
NULL, NULL,
NULL, NULL,
NULL, NULL,
NULL, NULL,
NULL, NULL
};
/* Timestamp logger */
static FILE *timestamp_log_stream;
/* A packet mbuf pool constructor with vr_packet support */
void vr_dpdk_pktmbuf_pool_init(struct rte_mempool *mp, void *opaque_arg)
{
#if (RTE_VERSION >= RTE_VERSION_NUM(2, 1, 0, 0))
struct rte_pktmbuf_pool_private priv;
/* Set private mbuf size for vr_packet. */
priv.mbuf_data_room_size = mp->elt_size - sizeof(struct rte_mbuf)
- sizeof(struct vr_packet);
priv.mbuf_priv_size = sizeof(struct vr_packet);
rte_pktmbuf_pool_init(mp, &priv);
#else
rte_pktmbuf_pool_init(mp, (void *)(mp->elt_size - sizeof(struct rte_mbuf)
- sizeof(struct vr_packet)));
#endif
}
/* The packet mbuf constructor with vr_packet support */
void
vr_dpdk_pktmbuf_init(struct rte_mempool *mp, void *opaque_arg, void *_m, unsigned i)
{
struct rte_mbuf *m = _m;
struct vr_packet *pkt;
rte_pktmbuf_init(mp, opaque_arg, _m, i);
#if (RTE_VERSION < RTE_VERSION_NUM(2, 1, 0, 0))
/* decrease rte packet size to fit vr_packet struct */
m->buf_len -= sizeof(struct vr_packet);
RTE_VERIFY(0 < m->buf_len);
/* start of buffer is just after vr_packet structure */
m->buf_addr += sizeof(struct vr_packet);
m->buf_physaddr += sizeof(struct vr_packet);
#endif
/* basic vr_packet initialization */
pkt = vr_dpdk_mbuf_to_pkt(m);
pkt->vp_head = (unsigned char *)m->buf_addr;
pkt->vp_end = m->buf_len;
}
/* Create memory pools */
static int
dpdk_mempools_create(void)
{
/* Create the mbuf pool used for RSS */
vr_dpdk.rss_mempool = rte_mempool_create("rss_mempool",
vr_mempool_sz,
VR_DPDK_MBUF_HDR_SZ + vr_packet_sz, VR_DPDK_RSS_MEMPOOL_CACHE_SZ,
sizeof(struct rte_pktmbuf_pool_private),
vr_dpdk_pktmbuf_pool_init, NULL, vr_dpdk_pktmbuf_init, NULL,
rte_socket_id(), 0);
if (vr_dpdk.rss_mempool == NULL) {
RTE_LOG(CRIT, VROUTER, "Error creating RSS mempool: %s (%d)\n",
rte_strerror(rte_errno), rte_errno);
return -rte_errno;
}
/* Create the mbuf pool used for IP fragmentation (direct mbufs) */
vr_dpdk.frag_direct_mempool = rte_mempool_create("frag_direct_mempool",
VR_DPDK_FRAG_DIRECT_MEMPOOL_SZ, VR_DPDK_FRAG_DIRECT_MBUF_SZ,
VR_DPDK_FRAG_DIRECT_MEMPOOL_CACHE_SZ,
sizeof(struct rte_pktmbuf_pool_private), rte_pktmbuf_pool_init,
NULL, rte_pktmbuf_init, NULL, rte_socket_id(), 0);
if (vr_dpdk.frag_direct_mempool == NULL) {
RTE_LOG(CRIT, VROUTER, "Error creating FRAG_DIRECT mempool: %s (%d)\n",
rte_strerror(rte_errno), rte_errno);
return -rte_errno;
}
/* Create the mbuf pool used for IP fragmentation (indirect mbufs) */
vr_dpdk.frag_indirect_mempool = rte_mempool_create("frag_indirect_mempool",
VR_DPDK_FRAG_INDIRECT_MEMPOOL_SZ, VR_DPDK_FRAG_INDIRECT_MBUF_SZ,
VR_DPDK_FRAG_INDIRECT_MEMPOOL_CACHE_SZ, 0, NULL, NULL,
rte_pktmbuf_init, NULL, rte_socket_id(), 0);
if (vr_dpdk.frag_indirect_mempool == NULL) {
RTE_LOG(CRIT, VROUTER, "Error creating FRAG_INDIRECT mempool: %s (%d)\n",
rte_strerror(rte_errno), rte_errno);
return -rte_errno;
}
#if VR_DPDK_USE_HW_FILTERING
int ret, i;
char mempool_name[RTE_MEMPOOL_NAMESIZE];
/* Create a list of free mempools */
vr_dpdk.nb_free_mempools = 0;
for (i = 0; i < VR_DPDK_MAX_VM_MEMPOOLS; i++) {
ret = snprintf(mempool_name, sizeof(mempool_name), "vr_mempool_%d", i);
if (ret >= sizeof(mempool_name)) {
RTE_LOG(INFO, VROUTER, "Error creating VM mempool %d name\n", i);
return -ENOMEM;
}
vr_dpdk.free_mempools[i] = rte_mempool_create(mempool_name,
VR_DPDK_VM_MEMPOOL_SZ, VR_DPDK_MBUF_HDR_SZ + vr_packet_sz, VR_DPDK_VM_MEMPOOL_CACHE_SZ,
sizeof(struct rte_pktmbuf_pool_private),
vr_dpdk_pktmbuf_pool_init, NULL, vr_dpdk_pktmbuf_init, NULL,
rte_socket_id(), 0);
if (vr_dpdk.free_mempools[i] == NULL) {
RTE_LOG(CRIT, VROUTER, "Error creating VM mempool %d: %s (%d)\n",
i, rte_strerror(rte_errno), rte_errno);
return -rte_errno;
}
vr_dpdk.nb_free_mempools++;
}
RTE_LOG(INFO, VROUTER, "Allocated %" PRIu16 " VM mempool(s)\n",
vr_dpdk.nb_free_mempools);
#endif
return 0;
}
/*
* Figure out a number of CPU cores/threads and compute an affinity mask
* which will be passed to EAL initialization in dpdk_init().
*
* Returns:
* new core mask on success
* VR_DPDK_DEF_LCORE_MASK on failure
* 0 if the system does not have enough cores
*/
static uint64_t
dpdk_core_mask_get(long system_cpus_count)
{
cpu_set_t cs;
uint64_t cpu_core_mask = 0;
int i;
long core_mask_count;
if (sched_getaffinity(0, sizeof(cs), &cs) < 0) {
RTE_LOG(ERR, VROUTER, "Error getting affinity."
" Falling back do the default core mask 0x%" PRIx64 "\n",
(uint64_t)(VR_DPDK_DEF_LCORE_MASK));
return VR_DPDK_DEF_LCORE_MASK;
}
/*
* Go through all the CPUs in the cpu_set_t structure to check
* if they are available or not. Build an affinity mask based on that.
* There is no official way to obtain the mask directly, as there is
* no macro for this.
*
* Due to size of uint64_t, maximum number of supported CPUs is 64.
*/
for (i = 0; i < RTE_MIN(CPU_SETSIZE, 64); i++) {
if (CPU_ISSET(i, &cs))
cpu_core_mask |= (uint64_t)1 << i;
}
if (!cpu_core_mask) {
RTE_LOG(ERR, VROUTER, "Error: core mask is zero."
" Falling back do the default core mask 0x%" PRIx64 "\n",
(uint64_t)(VR_DPDK_DEF_LCORE_MASK));
return VR_DPDK_DEF_LCORE_MASK;
}
/*
* Do not allow to run vRouter on all the cores available, as some have
* to be spared for virtual machines.
*/
core_mask_count = __builtin_popcountll((unsigned long long)cpu_core_mask);
if (core_mask_count == system_cpus_count) {
RTE_LOG(NOTICE, VROUTER, "Use taskset(1) to set the core mask."
" Falling back do the default core mask 0x%" PRIx64 "\n",
(uint64_t)(VR_DPDK_DEF_LCORE_MASK));
return VR_DPDK_DEF_LCORE_MASK;
}
return cpu_core_mask;
}
/*
* dpdk_shared_io_core_mask_stringify - stringify shared IO core mask
* Example: if core mask is 0xf and FWD_LCORES_PER_IO is 2,
* there are will be 4 forwarding lcores: 7@0,8@1,9@2,10@3
* and 2 shared IO lcores: 3@(0,1),4@(2,3)
*/
static char *
dpdk_shared_io_core_mask_stringify(uint64_t core_mask)
{
int cpu_id = 0;
int io_lcore_id = VR_DPDK_IO_LCORE_ID;
static char core_mask_string[VR_DPDK_STR_BUF_SZ];
static char io_cpus_string[VR_DPDK_STR_BUF_SZ];
char *p = core_mask_string;
char *iop = io_cpus_string;
int nb_fwd_cores = 0;
if (!VR_DPDK_USE_IO_LCORES || !VR_DPDK_SHARED_IO_LCORES)
return "";
while (core_mask) {
if (core_mask & 1) {
/* add CPU ID to IO lcore string */
if (iop != io_cpus_string)
*iop++ = ',';
iop += snprintf(iop,
sizeof(io_cpus_string) - (iop - io_cpus_string),
"%d", cpu_id);
if (iop - io_cpus_string >= sizeof(io_cpus_string)) {
RTE_LOG(ERR, VROUTER, "Error stringifying IO CPU ID: buffer overflow\n");
return NULL;
}
nb_fwd_cores++;
if (nb_fwd_cores >= VR_DPDK_FWD_LCORES_PER_IO
|| (core_mask >> 1) == 0) {
if (io_lcore_id > VR_DPDK_LAST_IO_LCORE_ID) {
RTE_LOG(WARNING, VROUTER,
"Warning: IO lcores limit exceeded (%d > %d)\n",
io_lcore_id, VR_DPDK_LAST_IO_LCORE_ID);
break;
}
p += snprintf(p,
sizeof(core_mask_string) - (p - core_mask_string),
"%d@(%s),", io_lcore_id, io_cpus_string);
if (p - core_mask_string >= sizeof(core_mask_string)) {
RTE_LOG(ERR, VROUTER, "Error stringifying IO core mask: buffer overflow\n");
return NULL;
}
io_lcore_id++;
iop = io_cpus_string;
nb_fwd_cores = 0;
}
}
core_mask >>= 1;
cpu_id++;
}
*p = '\0';
return core_mask_string;
}
/*
* dpdk_fwd_core_mask_stringify - stringify forwarding and dedicated IO
* core mask
* Example: if core mask is 0x3f and FWD_LCORES_PER_IO is 2,
* there are will be 4 forwarding lcores: 7@1,8@2,9@4,10@5
* and 2 dedicated IO lcores: 3@0,4@3
*/
static char *
dpdk_fwd_core_mask_stringify(uint64_t core_mask)
{
int cpu_id = 0;
int fwd_lcore_id = VR_DPDK_FWD_LCORE_ID;
static char core_mask_string[VR_DPDK_STR_BUF_SZ];
char *p = core_mask_string;
int io_lcore_id = VR_DPDK_IO_LCORE_ID;
int nb_fwd_cores = 0;
while (core_mask) {
if (core_mask & 1) {
if (p != core_mask_string)
*p++ = ',';
if (nb_fwd_cores == 0
&& io_lcore_id == VR_DPDK_LAST_IO_LCORE_ID + 1) {
RTE_LOG(WARNING, VROUTER,
"Warning: IO lcores limit exceeded (%d > %d)\n",
io_lcore_id, VR_DPDK_LAST_IO_LCORE_ID);
io_lcore_id++;
}
if (VR_DPDK_USE_IO_LCORES && !VR_DPDK_SHARED_IO_LCORES
&& nb_fwd_cores == 0
&& io_lcore_id <= VR_DPDK_LAST_IO_LCORE_ID) {
/* first dedicated CPU is an IO lcore */
p += snprintf(p,
sizeof(core_mask_string) - (p - core_mask_string),
"%d@%d", io_lcore_id, cpu_id);
if (p - core_mask_string >= sizeof(core_mask_string)) {
RTE_LOG(ERR, VROUTER, "Error stringifying IO core mask: buffer overflow\n");
return NULL;
}
io_lcore_id++;
} else {
/* forwarding lcore */
p += snprintf(p,
sizeof(core_mask_string) - (p - core_mask_string),
"%d@%d", fwd_lcore_id, cpu_id);
if (p - core_mask_string >= sizeof(core_mask_string)) {
RTE_LOG(ERR, VROUTER, "Error stringifying forwarding core mask: buffer overflow\n");
return NULL;
}
fwd_lcore_id++;
}
nb_fwd_cores++;
if (nb_fwd_cores > VR_DPDK_FWD_LCORES_PER_IO)
nb_fwd_cores = 0;
}
core_mask >>= 1;
cpu_id++;
}
*p = '\0';
return core_mask_string;
}
/*
* dpdk_argv_remove - remove specific argument and value from dpdk_argv[]
* Returns 0 on success, < 0 otherwise.
*/
static int
dpdk_argv_remove(char *arg)
{
int i;
int len = strlen(arg);
for (i = 0; i < RTE_DIM(dpdk_argv) - 1; i++) {
if (strncmp(dpdk_argv[i], arg, len) == 0) {
dpdk_argv[i] = NULL;
dpdk_argv[i + 1] = NULL;
return 0;
}
}
return -1;
}
/*
* dpdk_argv_append - append argument and value to dpdk_argv[]
* Returns 0 on success, < 0 otherwise.
*/
static int
dpdk_argv_append(char *arg, char *val)
{
int i;
for (i = 0; i < RTE_DIM(dpdk_argv) - 1; i++) {
if (dpdk_argv[i] == NULL && dpdk_argv[i + 1] == NULL) {
dpdk_argv[i] = arg;
dpdk_argv[i + 1] = val;
return 0;
}
}
return -1;
}
/*
* dpdk_argv_update - update EAL command line options for rte_eal_init()
* Returns number of arguments in dpdk_argv on success, < 0 otherwise.
*/
static int
dpdk_argv_update(void)
{
long int system_cpus_count;
int i;
uint64_t core_mask, mask_lm_bit;
char *io_core_mask_str;
char *fwd_core_mask_str;
static char lcores_string[VR_DPDK_STR_BUF_SZ];
/* get number of available CPUs */
system_cpus_count = sysconf(_SC_NPROCESSORS_CONF);
if (system_cpus_count == -1) {
system_cpus_count = __builtin_popcountll(
(unsigned long long)VR_DPDK_DEF_LCORE_MASK);
}
if (system_cpus_count == 0)
return -1;
core_mask = dpdk_core_mask_get(system_cpus_count);
/* calculate number of forwarding and IO lcores */
vr_dpdk.nb_fwd_lcores = __builtin_popcountll(core_mask);
vr_dpdk.nb_io_lcores = 0;
if (VR_DPDK_USE_IO_LCORES) {
if (VR_DPDK_SHARED_IO_LCORES) {
vr_dpdk.nb_io_lcores = (vr_dpdk.nb_fwd_lcores + VR_DPDK_FWD_LCORES_PER_IO - 1)
/VR_DPDK_FWD_LCORES_PER_IO;
if (vr_dpdk.nb_io_lcores > VR_DPDK_MAX_IO_LORES)
vr_dpdk.nb_io_lcores = VR_DPDK_MAX_IO_LORES;
} else {
vr_dpdk.nb_io_lcores = (vr_dpdk.nb_fwd_lcores + VR_DPDK_FWD_LCORES_PER_IO)
/(VR_DPDK_FWD_LCORES_PER_IO + 1);
if (vr_dpdk.nb_io_lcores > VR_DPDK_MAX_IO_LORES)
vr_dpdk.nb_io_lcores = VR_DPDK_MAX_IO_LORES;
vr_dpdk.nb_fwd_lcores -= vr_dpdk.nb_io_lcores;
}
}
/* sanity checks */
if (vr_dpdk.nb_fwd_lcores == 0) {
RTE_LOG(ERR, VROUTER, "Error configuring lcores: no forwarding lcores defined\n");
return -1;
}
if (vr_dpdk.nb_io_lcores > 1
&& vr_dpdk.nb_fwd_lcores == VR_DPDK_FWD_LCORES_PER_IO*(vr_dpdk.nb_io_lcores - 1)) {
/*
* The last IO lcore has no forwarding lcores.
* Decrease the number of IO lcores and continue.
*/
RTE_LOG(INFO, VROUTER, "Adjusting number of IO lcores: %u -> %u\n",
vr_dpdk.nb_io_lcores, vr_dpdk.nb_io_lcores - 1);
vr_dpdk.nb_io_lcores--;
/* remove the leftmost bit from the core mask */
mask_lm_bit = core_mask;
mask_lm_bit |= mask_lm_bit >> 32;
mask_lm_bit |= mask_lm_bit >> 16;
mask_lm_bit |= mask_lm_bit >> 8;
mask_lm_bit |= mask_lm_bit >> 4;
mask_lm_bit |= mask_lm_bit >> 2;
mask_lm_bit |= mask_lm_bit >> 1;
mask_lm_bit ^= mask_lm_bit >> 1;
RTE_LOG(INFO, VROUTER, "Adjusting core mask: 0x%"PRIx64" -> 0x%"PRIx64"\n",
core_mask, core_mask ^ mask_lm_bit);
core_mask ^= mask_lm_bit;
}
io_core_mask_str = dpdk_shared_io_core_mask_stringify(core_mask);
if (io_core_mask_str == NULL)
return -1;
fwd_core_mask_str = dpdk_fwd_core_mask_stringify(core_mask);
if (fwd_core_mask_str == NULL)
return -1;
/* lcores order: service, IO, lcores with TX queues, forwaridng lcores */
if (snprintf(lcores_string, sizeof(lcores_string),
"(0-%d)@(0-%ld),%s(%d-%d)@(0-%ld),%s",
VR_DPDK_IO_LCORE_ID - 1, system_cpus_count - 1,
io_core_mask_str,
VR_DPDK_PACKET_LCORE_ID, VR_DPDK_FWD_LCORE_ID - 1, system_cpus_count - 1,
fwd_core_mask_str)
>= sizeof(lcores_string)) {
return -1;
}
/* Append lcores option. */
if (dpdk_argv_append("--"LCORES_OPT, lcores_string) != 0)
return -1;
/* Append no huge option. */
if (no_huge_set && dpdk_argv_append("--"NO_HUGE_OPT, NULL) != 0)
return -1;
/* print out configuration */
if (vr_dpdk.vlan_tag != VLAN_ID_INVALID) {
RTE_LOG(INFO, VROUTER, "Using VLAN TCI: %" PRIu16 "\n", vr_dpdk.vlan_tag);
}
RTE_LOG(INFO, VROUTER, "Bridge Table limit: %" PRIu32 "\n",
vr_bridge_entries);
RTE_LOG(INFO, VROUTER, "Bridge Table overflow limit: %" PRIu32 "\n",
vr_bridge_oentries);
RTE_LOG(INFO, VROUTER, "Flow Table limit: %" PRIu32 "\n",
vr_flow_entries);
RTE_LOG(INFO, VROUTER, "Flow Table overflow limit: %" PRIu32 "\n",
vr_oflow_entries);
RTE_LOG(INFO, VROUTER, "MPLS labels limit: %" PRIu32 "\n",
vr_mpls_labels);
RTE_LOG(INFO, VROUTER, "Nexthops limit: %" PRIu32 "\n",
vr_nexthops);
RTE_LOG(INFO, VROUTER, "VRF tables limit: %" PRIu32 "\n",
vr_vrfs);
RTE_LOG(INFO, VROUTER, "Packet pool size: %" PRIu32 "\n",
vr_mempool_sz);
RTE_LOG(INFO, VROUTER, "Maximum packet size: %" PRIu32 "\n",
vr_packet_sz);
RTE_LOG(INFO, VROUTER, "EAL arguments:\n");
for (i = 1; i < RTE_DIM(dpdk_argv) - 1; i += 2) {
if (dpdk_argv[i] == NULL)
break;
if (dpdk_argv[i + 1] == NULL) {
RTE_LOG(INFO, VROUTER, " %12s\n", dpdk_argv[i]);
i--;
} else {
RTE_LOG(INFO, VROUTER, " %12s \"%s\"\n",
dpdk_argv[i], dpdk_argv[i + 1]);
}
}
return i;
}
static void
version_print(void)
{
RTE_LOG(INFO, VROUTER, "vRouter/DPDK version: %s\n", ContrailBuildInfo);
}
/*
* dpdk_check_sriov_vf - check if any of eth devices is a virtual function.
*/
static void
dpdk_check_sriov_vf(void)
{
int i;
struct rte_eth_dev_info dev_info;
size_t soff;
for (i = 0; i < rte_eth_dev_count(); i++)
{
rte_eth_dev_info_get(i, &dev_info);
/* Check PMD name suffix to detect SR-IOV virtual function. */
soff = strlen(dev_info.driver_name) - sizeof(VR_DPDK_VF_PMD_SFX) + 1;
if (soff > 0 &&
strncmp(dev_info.driver_name + soff, VR_DPDK_VF_PMD_SFX,
sizeof(VR_DPDK_VF_PMD_SFX)) == 0) {
/* Dedicate the first forwarding lcore to VF RX/TX. */
if (dev_info.max_tx_queues < vr_dpdk.nb_fwd_lcores
/* We also need 2 TX queues for Netlink and Packet lcores. */
+ VR_DPDK_FWD_LCORE_ID - VR_DPDK_PACKET_LCORE_ID) {
vr_dpdk.vf_lcore_id = VR_DPDK_FWD_LCORE_ID;
RTE_LOG(INFO, VROUTER,
"%s: Lcore %d: SR-IOV virtual function IO for eth device %d (%s)\n",
__func__, VR_DPDK_FWD_LCORE_ID, i, dev_info.driver_name);
break;
}
}
}
}
/* Init DPDK EAL */
static int
dpdk_init(void)
{
int ret, nb_sys_ports;
version_print();
ret = vr_dpdk_table_mem_init(VR_MEM_FLOW_TABLE_OBJECT, vr_flow_entries,
VR_FLOW_TABLE_SIZE, vr_oflow_entries, VR_OFLOW_TABLE_SIZE);
if (ret < 0) {
RTE_LOG(ERR, VROUTER, "Error initializing flow table: %s (%d)\n",
rte_strerror(-ret), -ret);
return ret;
}
ret = vr_dpdk_table_mem_init(VR_MEM_BRIDGE_TABLE_OBJECT, vr_bridge_entries,
VR_BRIDGE_TABLE_SIZE, vr_bridge_oentries,
VR_BRIDGE_OFLOW_TABLE_SIZE);
if (ret < 0) {
RTE_LOG(ERR, VROUTER, "Error initializing bridge table: %s (%d)\n",
rte_strerror(-ret), -ret);
return ret;
}
ret = dpdk_argv_update();
if (ret == -1) {
RTE_LOG(ERR, VROUTER, "Error updating EAL arguments\n");
return -1;
}
ret = rte_eal_init(ret, dpdk_argv);
if (ret < 0) {
RTE_LOG(ERR, VROUTER, "Error initializing EAL\n");
return ret;
}
/* EAL resets the log stream */
rte_openlog_stream(timestamp_log_stream);
/* disable unwanted logtypes for debug purposes */
rte_set_log_type(VR_DPDK_LOGTYPE_DISABLE, 0);
/* set default log level to INFO */
rte_set_log_level(RTE_LOG_INFO);
ret = dpdk_mempools_create();
if (ret < 0)
return ret;
/* get number of ports found in scan */
nb_sys_ports = rte_eth_dev_count();
RTE_LOG(INFO, VROUTER, "Found %d eth device(s)\n", nb_sys_ports);
/* get number of cores */
RTE_LOG(INFO, VROUTER, "Using %d forwarding lcore(s)\n",
vr_dpdk.nb_fwd_lcores);
RTE_LOG(INFO, VROUTER, "Using %d IO lcore(s)\n",
vr_dpdk.nb_io_lcores);
RTE_LOG(INFO, VROUTER, "Using %d service lcores\n",
rte_lcore_count() - vr_dpdk.nb_fwd_lcores
- vr_dpdk.nb_io_lcores);
/* init timer subsystem */
rte_timer_subsystem_init();
/* Check if any of eth devices is a SR-IOV virtual function. */
dpdk_check_sriov_vf();
/* Init the interface configuration mutex
* ATM we use it just to synchronize access between the NetLink interface
* and kernel KNI events. The datapath is not affected. */
return pthread_mutex_init(&vr_dpdk.if_lock, NULL);
}
/* Shutdown DPDK EAL */
static void
dpdk_exit(void)
{
int i;
vr_dpdk_if_lock();
RTE_LOG(INFO, VROUTER, "Releasing KNI devices...\n");
for (i = 0; i < VR_DPDK_MAX_KNI_INTERFACES; i++) {
if (vr_dpdk.knis[i] != NULL) {
rte_kni_release(vr_dpdk.knis[i]);
vr_dpdk.knis[i] = NULL;
}
}
RTE_LOG(INFO, VROUTER, "Releasing TAP devices...\n");
for (i = 0; i < VR_DPDK_MAX_TAP_INTERFACES; i++) {
if (vr_dpdk.tapdevs[i].tapdev_fd > 0) {
close(vr_dpdk.tapdevs[i].tapdev_fd);
vr_dpdk.tapdevs[i].tapdev_fd = -1;
}
}
RTE_LOG(INFO, VROUTER, "Closing eth devices...\n");
for (i = 0; i < RTE_MAX_ETHPORTS; i++) {
if (vr_dpdk.ethdevs[i].ethdev_ptr != NULL) {
rte_eth_dev_stop(i);
rte_eth_dev_close(i);
vr_dpdk.ethdevs[i].ethdev_ptr = NULL;
}
}
vr_dpdk_if_unlock();
/* destroy interface lock */
if (pthread_mutex_destroy(&vr_dpdk.if_lock)) {
RTE_LOG(ERR, VROUTER, "Error destroying interface lock\n");
}
}
/* Set stop flag for all lcores */
static void
dpdk_stop_flag_set(void)
{
/* check if the flag is already set */
if (unlikely(vr_dpdk_is_stop_flag_set()))
return;
rte_atomic16_inc(&vr_dpdk.stop_flag);
vr_dpdk_lcore_cmd_post_all(VR_DPDK_LCORE_STOP_CMD, 0);
/* wakeup UVHost server to shutdown */
vr_uvhost_wakeup();
}
/* Check if the stop flag is set */
int
vr_dpdk_is_stop_flag_set(void)
{
if (unlikely(rte_atomic16_read(&vr_dpdk.stop_flag)))
return 1;
return 0;
}
/* Custom handling of signals */
static void
dpdk_signal_handler_stop(int signum)
{
RTE_LOG(INFO, VROUTER, "Got signal %d on lcore %u, stopping...\n",
signum, rte_lcore_id());
dpdk_stop_flag_set();
}
static void
dpdk_signal_handler_ignore(int signum)
{
RTE_LOG(INFO, VROUTER, "Got signal %d on lcore %u, ignoring...\n",
signum, rte_lcore_id());
}
/* Setup signal handlers */
static int
dpdk_signals_init(void)
{
struct sigaction act;
sigset_t set;
memset(&act, 0 , sizeof(act));
act.sa_handler = dpdk_signal_handler_stop;
if (sigaction(SIGTERM, &act, NULL) != 0) {
RTE_LOG(CRIT, VROUTER, "Error registering SIGTERM handler\n");
return -1;
}
if (sigaction(SIGINT, &act, NULL) != 0) {
RTE_LOG(CRIT, VROUTER, "Error registering SIGINT handler\n");
return -1;
}
act.sa_handler = dpdk_signal_handler_ignore;
if (sigaction(SIGPIPE, &act, NULL) != 0) {
RTE_LOG(CRIT, VROUTER, "Error registering SIGPIPE handler\n");
return -1;
}
/* Block (ignore) all the signals for this and all the child threads.
* The signals will be unblocked for the master lcore later during
* the lcore intialization.
*/
sigfillset(&set);
if (pthread_sigmask(SIG_BLOCK, &set, NULL) != 0) {
RTE_LOG(CRIT, VROUTER, "Error setting signal mask\n");
return -1;
}
return 0;
}
/*
* vr_dpdk_exit_trigger - function that is called by user space vhost server
* to cause all DPDK threads to exit.
*
* Returns nothing.
*/
void
vr_dpdk_exit_trigger(void)
{
dpdk_stop_flag_set();
return;
}
static struct option long_options[] = {
[NO_DAEMON_OPT_INDEX] = {NO_DAEMON_OPT, no_argument,
&no_daemon_set, 1},
[NO_HUGE_OPT_INDEX] = {NO_HUGE_OPT, no_argument,
&no_huge_set, 1},
[HELP_OPT_INDEX] = {HELP_OPT, no_argument,
NULL, 0},
[VERSION_OPT_INDEX] = {VERSION_OPT, no_argument,
NULL, 0},
[MEMPOOL_SIZE_OPT_INDEX] = {MEMPOOL_SIZE_OPT, required_argument,
NULL, 0},
[PACKET_SIZE_OPT_INDEX] = {PACKET_SIZE_OPT, required_argument,
NULL, 0},
[VLAN_TCI_OPT_INDEX] = {VLAN_TCI_OPT, required_argument,
NULL, 0},
[VLAN_NAME_OPT_INDEX] = {VLAN_NAME_OPT, required_argument,
NULL, 0},
[VTEST_VLAN_OPT_INDEX] = {VTEST_VLAN_OPT, no_argument,
NULL, 0},
[VDEV_OPT_INDEX] = {VDEV_OPT, required_argument,
NULL, 0},
[BRIDGE_ENTRIES_OPT_INDEX] = {BRIDGE_ENTRIES_OPT, required_argument,
NULL, 0},
[BRIDGE_OENTRIES_OPT_INDEX] = {BRIDGE_OENTRIES_OPT, required_argument,
NULL, 0},
[FLOW_ENTRIES_OPT_INDEX] = {FLOW_ENTRIES_OPT, required_argument,
NULL, 0},
[OFLOW_ENTRIES_OPT_INDEX] = {OFLOW_ENTRIES_OPT, required_argument,
NULL, 0},
[MPLS_LABELS_OPT_INDEX] = {MPLS_LABELS_OPT, required_argument,
NULL, 0},
[NEXTHOPS_OPT_INDEX] = {NEXTHOPS_OPT, required_argument,
NULL, 0},
[VRFS_OPT_INDEX] = {VRFS_OPT, required_argument,
NULL, 0},
[SOCKET_DIR_OPT_INDEX] = {SOCKET_DIR_OPT, required_argument,
NULL, 0},
[NETLINK_PORT_OPT_INDEX] = {NETLINK_PORT_OPT, required_argument,
NULL, 0},
[SOCKET_MEM_OPT_INDEX] = {SOCKET_MEM_OPT, required_argument,
NULL, 0},
[MAX_OPT_INDEX] = {NULL, 0,
NULL, 0},
};
static void
Usage(void)
{
printf(
"Usage: contrail-vrouter-dpdk [options]\n"
" --"NO_DAEMON_OPT" Do not demonize the vRouter\n"
" --"NO_HUGE_OPT" Use malloc instead of hugetlbfs\n"
" --"HELP_OPT" This help\n"
" --"VERSION_OPT" Display build information\n"
"\n"
" --"VDEV_OPT" CONF Add a virtual device.\n"
" The argument format is <driver><id>[,key=val,...]\n"
" (ex: --"VDEV_OPT" eth_bond0,mode=4,slave=0000:04:00.0)\n"
" --"SOCKET_MEM_OPT" MB,... Memory to allocate on sockets.\n"
" (ex: --"SOCKET_MEM_OPT" 256,256)\n"
"\n"
" --"VLAN_TCI_OPT" TCI VLAN tag control information to use\n"
" It may be a value between 0 and 4095\n"
" --"VLAN_NAME_OPT" NAME VLAN forwarding interface name\n"
"\n"
" --"BRIDGE_ENTRIES_OPT" NUM Bridge table limit\n"
" --"BRIDGE_OENTRIES_OPT" NUM Bridge table overflow limit\n"
" --"FLOW_ENTRIES_OPT" NUM Flow table limit\n"
" --"OFLOW_ENTRIES_OPT" NUM Flow overflow table limit\n"
" --"MPLS_LABELS_OPT" NUM MPLS table limit\n"
" --"NEXTHOPS_OPT" NUM Nexthop table limit\n"
" --"VRFS_OPT" NUM VRF tables limit\n"
" --"MEMPOOL_SIZE_OPT" NUM Main packet pool size\n"
" --"PACKET_SIZE_OPT" NUM Maximum packet size\n"
);
exit(1);
}
static void
parse_long_opts(int opt_flow_index, char *optarg)
{
errno = 0;
switch (opt_flow_index) {
case NO_DAEMON_OPT_INDEX:
case NO_HUGE_OPT_INDEX:
break;
case VERSION_OPT_INDEX:
version_print();
exit(0);
break;
case MEMPOOL_SIZE_OPT_INDEX:
vr_mempool_sz = (unsigned int)strtoul(optarg, NULL, 0);
if (errno != 0) {
vr_mempool_sz = VR_DEF_MEMPOOL_SZ;
}
break;
case PACKET_SIZE_OPT_INDEX:
vr_packet_sz = (unsigned int)strtoul(optarg, NULL, 0);
if (errno != 0) {
vr_packet_sz = VR_DEF_MAX_PACKET_SZ;
}
break;
/*
* If VLAN tag is set, vRouter will expect tagged packets. The tag
* will be stripped by NIC or in vr_dpdk_ethdev_rx_emulate() and
* injected in dpdk_if_tx().
*
* Received packets with unmatching tag will be forwarded to the VLAN
* forwarding interface, that is created in main(). Packets sent on that
* interface will be immediately forwarded to the physical interface.
*
* See following funtions: dpdk_lcore_fwd_io(), vr_dpdk_ethdev_rx_emulate(),
* vr_dpdk_lcore_vroute(), dpdk_vlan_forwarding_if_add().
*/
case VLAN_TCI_OPT_INDEX:
vr_dpdk.vlan_tag = (uint16_t)strtoul(optarg, NULL, 0);
if (errno != 0) {
vr_dpdk.vlan_tag = VLAN_ID_INVALID;
}
if (vr_dpdk.vlan_tag > 4095)
Usage();
break;
case VTEST_VLAN_OPT_INDEX:
vr_dpdk.vtest_vlan = 1;
break;
/*
* VLAN packets with unmatching tag will be forwarded to the kernel using
* an interface with name defined here.
*/
case VLAN_NAME_OPT_INDEX:
strncpy(vr_dpdk.vlan_name, optarg, sizeof(vr_dpdk.vlan_name) - 1);
break;
case VDEV_OPT_INDEX:
dpdk_argv_append("--"VDEV_OPT, optarg);
break;
case BRIDGE_ENTRIES_OPT_INDEX:
vr_bridge_entries = (unsigned int)strtoul(optarg, NULL, 0);
if (errno != 0) {
vr_bridge_entries = VR_DEF_BRIDGE_ENTRIES;
}
break;
case BRIDGE_OENTRIES_OPT_INDEX:
vr_bridge_oentries = (unsigned int)strtoul(optarg, NULL, 0);
if (errno != 0) {
vr_bridge_oentries = ((vr_bridge_entries / 5) + 1023) & ~1023;
}
break;
case FLOW_ENTRIES_OPT_INDEX:
vr_flow_entries = (unsigned int)strtoul(optarg, NULL, 0);
if (errno != 0) {
vr_flow_entries = VR_DEF_FLOW_ENTRIES;
}