-
Notifications
You must be signed in to change notification settings - Fork 171
/
vr_dpdk_host.c
1539 lines (1259 loc) · 38.3 KB
/
vr_dpdk_host.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.
*
* vr_dpdk_host.c -- DPDK vrouter module
*
*/
#include "vr_dpdk.h"
#include "vr_fragment.h"
#include "vr_hash.h"
#include "vr_proto.h"
#include "vr_sandesh.h"
#include <linux/if_ether.h>
#include <netinet/ip.h>
#include <netinet/tcp.h>
#include <sys/user.h>
#include <sys/resource.h>
#include <rte_cycles.h>
#include <rte_errno.h>
#include <rte_ethdev.h>
#include <rte_jhash.h>
#include <rte_malloc.h>
#include <rte_timer.h>
/* Max number of CPUs. We adjust it later in vr_dpdk_host_init() */
unsigned int vr_num_cpus = VR_MAX_CPUS;
/* Global init flag */
static bool vr_host_inited = false;
extern void vr_malloc_stats(unsigned int, unsigned int);
extern void vr_free_stats(unsigned int);
/* RCU callback */
extern void vr_flow_defer_cb(struct vrouter *router, void *arg);
static void *
dpdk_page_alloc(unsigned int size)
{
return rte_malloc(0, size, PAGE_SIZE);
}
static void
dpdk_page_free(void *address, unsigned int size)
{
rte_free(address);
}
static int
dpdk_printf(const char *format, ...)
{
va_list args;
if (RTE_LOGTYPE_DPCORE & rte_logs.type) {
char buf[VR_DPDK_STR_BUF_SZ] = "DPCORE: ";
strncat(buf, format, sizeof(buf) - strlen(buf) - 1);
buf[sizeof(buf) - 1] = '\0';
va_start(args, format);
rte_vlog(RTE_LOG_INFO, RTE_LOGTYPE_DPCORE, buf, args);
va_end(args);
}
return 0;
}
static void *
dpdk_malloc(unsigned int size, unsigned int object)
{
void *mem = rte_malloc(NULL, size, 0);
if (likely(mem != NULL)) {
vr_malloc_stats(size, object);
}
return mem;
}
static void *
dpdk_zalloc(unsigned int size, unsigned int object)
{
void *mem = rte_zmalloc(NULL, size, 0);
if (likely(mem != NULL)) {
vr_malloc_stats(size, object);
}
return mem;
}
static void
dpdk_free(void *mem, unsigned int object)
{
if (mem) {
vr_free_stats(object);
rte_free(mem);
}
return;
}
static uint64_t
dpdk_vtop(void *address)
{
/* TODO: not used */
rte_panic("%s: not used in DPDK mode\n", __func__);
return (uint64_t)0;
}
static struct vr_packet *
dpdk_palloc(unsigned int size)
{
struct rte_mbuf *m;
/* in DPDK we have fixed-sized mbufs only */
RTE_VERIFY(size <= VR_DPDK_MAX_PACKET_SZ);
m = rte_pktmbuf_alloc(vr_dpdk.rss_mempool);
if (!m)
return (NULL);
return vr_dpdk_packet_get(m, NULL);
}
static struct vr_packet *
dpdk_palloc_head(struct vr_packet *pkt, unsigned int size)
{
/* TODO: not implemented */
RTE_LOG(ERR, VROUTER, "%s: not implemented\n", __func__);
return NULL;
}
static struct vr_packet *
dpdk_pexpand_head(struct vr_packet *pkt, unsigned int hspace)
{
/* TODO: not implemented */
return pkt;
}
static void
dpdk_pfree(struct vr_packet *pkt, unsigned short reason)
{
struct vrouter *router = vrouter_get(0);
router->vr_pdrop_stats[rte_lcore_id()][reason]++;
if (pkt)
rte_pktmbuf_free(vr_dpdk_pkt_to_mbuf(pkt));
return;
}
void
vr_dpdk_pfree(struct rte_mbuf *mbuf, unsigned short reason)
{
dpdk_pfree(vr_dpdk_mbuf_to_pkt(mbuf), reason);
}
static void
dpdk_preset(struct vr_packet *pkt)
{
struct rte_mbuf *m;
if (!pkt)
rte_panic("%s: NULL pkt", __func__);
m = vr_dpdk_pkt_to_mbuf(pkt);
/* Reset packet data */
pkt->vp_data = rte_pktmbuf_headroom(m);
pkt->vp_tail = rte_pktmbuf_headroom(m) + rte_pktmbuf_data_len(m);
pkt->vp_len = rte_pktmbuf_data_len(m);
return;
}
/**
* Copy packet mbuf data to another packet mbuf.
*
* @param dst
* The destination packet mbuf.
* @param src
* The source packet mbuf.
*/
static inline void
dpdk_pktmbuf_data_copy(struct rte_mbuf *dst, struct rte_mbuf *src)
{
dst->data_off = src->data_off;
dst->port = src->port;
dst->ol_flags = src->ol_flags;
dst->packet_type = src->packet_type;
dst->data_len = src->data_len;
dst->pkt_len = src->pkt_len;
dst->vlan_tci = src->vlan_tci;
dst->hash = src->hash;
dst->seqn = src->seqn;
dst->userdata = src->userdata;
dst->tx_offload = src->tx_offload;
__rte_mbuf_sanity_check(dst, 1);
__rte_mbuf_sanity_check(src, 0);
/* copy data */
rte_memcpy(rte_pktmbuf_mtod(dst, void *),
rte_pktmbuf_mtod(src, void *), src->data_len);
}
/**
* Creates a copy of the given packet mbuf.
* TODO: remove once rte_pktmbuf_copy() is in DPDK
*
* Walks through all segments of the given packet mbuf, and for each of them:
* - Creates a new packet mbuf from the given pool.
* - Copies data to the newly created mbuf.
* Then updates pkt_len and nb_segs of the "copy" packet mbuf to match values
* from the original packet mbuf.
*
* @param md
* The packet mbuf to be copied.
* @param mp
* The mempool from which the "copy" mbufs are allocated.
* @return
* - The pointer to the new "copy" mbuf on success.
* - NULL if allocation fails.
*/
inline struct rte_mbuf *
vr_dpdk_pktmbuf_copy(struct rte_mbuf *md, struct rte_mempool *mp)
{
struct rte_mbuf *mc, *mi, **prev;
uint32_t pktlen;
uint8_t nseg;
if (unlikely ((mc = rte_pktmbuf_alloc(mp)) == NULL))
return (NULL);
mi = mc;
prev = &mi->next;
pktlen = md->pkt_len;
nseg = 0;
do {
nseg++;
dpdk_pktmbuf_data_copy(mi, md);
*prev = mi;
prev = &mi->next;
} while ((md = md->next) != NULL &&
(mi = rte_pktmbuf_alloc(mp)) != NULL);
*prev = NULL;
mc->nb_segs = nseg;
mc->pkt_len = pktlen;
/* Allocation of new indirect segment failed */
if (unlikely (mi == NULL)) {
rte_pktmbuf_free(mc);
return (NULL);
}
__rte_mbuf_sanity_check(mc, 1);
return (mc);
}
/* VRouter callback */
static struct vr_packet *
dpdk_pclone(struct vr_packet *pkt)
{
/*
* TODO: We have not tested pclone option on DPDK 2.0. - (mbuf leak).
*/
/* Macro RTE_VERSION is workaround, we have mbuf leak in DPDK 2.0 */
#if (RTE_VERSION >= RTE_VERSION_NUM(2, 1, 0, 0))
struct rte_mbuf *m, *m_clone;
struct vr_packet *pkt_clone;
m = vr_dpdk_pkt_to_mbuf(pkt);
m_clone = rte_pktmbuf_clone(m, vr_dpdk.rss_mempool);
if (!m_clone)
return NULL;
/* clone vr_packet data */
pkt_clone = vr_dpdk_mbuf_to_pkt(m_clone);
*pkt_clone = *pkt;
pkt_clone->vp_cpu = vr_get_cpu();
return pkt_clone;
#else
/* if no scatter/gather enabled -> just copy the mbuf */
struct rte_mbuf *m, *m_copy;
struct vr_packet *pkt_copy;
m = vr_dpdk_pkt_to_mbuf(pkt);
m_copy = vr_dpdk_pktmbuf_copy(m, vr_dpdk.rss_mempool);
if (!m_copy)
return NULL;
/* copy vr_packet data */
pkt_copy = vr_dpdk_mbuf_to_pkt(m_copy);
*pkt_copy = *pkt;
/* set head pointer to a copy */
pkt_copy->vp_head = m_copy->buf_addr;
return pkt_copy;
#endif
}
/* Copy the specified number of bytes from the source mbuf to the
* destination buffer.
*/
static int
dpdk_pktmbuf_copy_bits(const struct rte_mbuf *mbuf, int offset,
void *to, int len)
{
/* how many bytes to copy in loop */
int copy = 0;
/* loop pointer to a source data */
void *from;
/* check total packet length */
if (unlikely(offset > (int)rte_pktmbuf_pkt_len(mbuf) - len))
goto fault;
do {
if (offset < rte_pktmbuf_data_len(mbuf)) {
/* copy a piece of data */
from = (void *)(rte_pktmbuf_mtod_offset(mbuf, uintptr_t, offset));
copy = rte_pktmbuf_data_len(mbuf) - offset;
if (copy > len)
copy = len;
rte_memcpy(to, from, copy);
offset = 0;
} else {
offset -= rte_pktmbuf_data_len(mbuf);
}
/* get next mbuf */
to += copy;
len -= copy;
mbuf = mbuf->next;
} while (unlikely(len > 0 && NULL != mbuf));
if (likely(0 == len))
return 0;
fault:
return -EFAULT;
}
/* VRouter callback */
static int
dpdk_pcopy(unsigned char *dst, struct vr_packet *p_src,
unsigned int offset, unsigned int len)
{
int ret;
struct rte_mbuf *src;
src = vr_dpdk_pkt_to_mbuf(p_src);
ret = dpdk_pktmbuf_copy_bits(src, offset, dst, len);
if (ret)
return ret;
return len;
}
static unsigned short
dpdk_pfrag_len(struct vr_packet *pkt)
{
struct rte_mbuf *m;
m = vr_dpdk_pkt_to_mbuf(pkt);
return rte_pktmbuf_pkt_len(m) - rte_pktmbuf_data_len(m);
}
static unsigned short
dpdk_phead_len(struct vr_packet *pkt)
{
struct rte_mbuf *m;
m = vr_dpdk_pkt_to_mbuf(pkt);
return rte_pktmbuf_data_len(m);
}
static void
dpdk_pset_data(struct vr_packet *pkt, unsigned short offset)
{
struct rte_mbuf *m;
m = vr_dpdk_pkt_to_mbuf(pkt);
m->buf_addr = pkt->vp_head;
m->data_off = offset;
return;
}
static unsigned int
dpdk_get_cpu(void)
{
unsigned lcore_id = rte_lcore_id();
/* For the RCU thread we get LCORE_ID_ANY, so memory stats and
* other functions get crashed trying to index per-cpu data.
*/
if (lcore_id < vr_num_cpus)
return lcore_id;
else
return 0;
}
/* DPDK timer callback */
static void
dpdk_timer(struct rte_timer *tim, void *arg)
{
struct vr_timer *vtimer = (struct vr_timer*)arg;
vtimer->vt_timer(vtimer->vt_vr_arg);
}
static int
dpdk_create_timer(struct vr_timer *vtimer)
{
struct rte_timer *timer;
uint64_t hz, ticks;
timer = rte_zmalloc("vr_dpdk_timer", sizeof(struct rte_timer), 0);
if (!timer) {
RTE_LOG(ERR, VROUTER, "Error allocating RTE timer\n");
return -1;
}
/* init timer */
rte_timer_init(timer);
vtimer->vt_os_arg = (void *)timer;
/* reset timer */
hz = rte_get_timer_hz();
ticks = hz * vtimer->vt_msecs / 1000;
if (rte_timer_reset(timer, ticks, PERIODICAL, VR_DPDK_TIMER_LCORE_ID,
dpdk_timer, vtimer) == -1) {
RTE_LOG(ERR, VROUTER, "Error resetting timer\n");
rte_free(timer);
return -1;
}
return 0;
}
static void
dpdk_delete_timer(struct vr_timer *vtimer)
{
struct rte_timer *timer = (struct rte_timer*)vtimer->vt_os_arg;
if (timer) {
rte_timer_stop_sync(timer);
rte_free(timer);
} else {
RTE_LOG(ERR, VROUTER, "No timer to delete\n");
}
}
static void
dpdk_get_time(unsigned long *sec, unsigned long *usec)
{
struct timespec ts;
*sec = *usec = 0;
if (-1 == clock_gettime(CLOCK_REALTIME, &ts))
return;
*sec = ts.tv_sec;
*usec = ts.tv_nsec / 1000;
return;
}
static void
dpdk_get_mono_time(unsigned int *sec, unsigned int *nsec)
{
struct timespec ts;
*sec = *nsec = 0;
if (-1 == clock_gettime(CLOCK_MONOTONIC, &ts))
return;
*sec = ts.tv_sec;
*nsec = ts.tv_nsec;
return;
}
/* Work callback called on NetLink lcore */
static int
dpdk_schedule_work(unsigned int cpu, void (*fn)(void *), void *arg)
{
/* no RCU reader lock needed, just do the work */
fn(arg);
return 0;
}
static void
dpdk_delay_op(void)
{
synchronize_rcu();
return;
}
/* RCU callback called on RCU thread */
static void
dpdk_rcu_cb(struct rcu_head *rh)
{
int i;
struct vr_dpdk_rcu_cb_data *cb_data;
struct vr_defer_data *defer;
struct vr_flow_queue *vfq;
struct vr_packet_node *pnode;
cb_data = CONTAINER_OF(rcd_rcu, struct vr_dpdk_rcu_cb_data, rh);
/* check if we need to pass the callback to packet lcore */
if ((cb_data->rcd_user_cb == vr_flow_defer_cb) &&
cb_data->rcd_user_data) {
defer = (struct vr_defer_data *)cb_data->rcd_user_data;
vfq = ((struct vr_flow_defer_data *)defer->vdd_data)->vfdd_flow_queue;
if (vfq) {
for (i = 0; i < VR_MAX_FLOW_QUEUE_ENTRIES; i++) {
pnode = &vfq->vfq_pnodes[i];
if (pnode->pl_packet) {
RTE_LOG(DEBUG, VROUTER, "%s: lcore %u passing RCU callback "
"to lcore %u\n", __func__, rte_lcore_id(),
VR_DPDK_PACKET_LCORE_ID);
vr_dpdk_lcore_cmd_post(VR_DPDK_PACKET_LCORE_ID,
VR_DPDK_LCORE_RCU_CMD, (uintptr_t)rh);
return;
}
}
RTE_LOG(DEBUG, VROUTER, "%s: lcore %u passing RCU callback to lcore %u\n",
__func__, rte_lcore_id(), VR_DPDK_PACKET_LCORE_ID);
}
}
/* no need to send any packets, so just call the callback */
cb_data->rcd_user_cb(cb_data->rcd_router, cb_data->rcd_user_data);
vr_free(cb_data, VR_DEFER_OBJECT);
}
static void
dpdk_defer(struct vrouter *router, vr_defer_cb user_cb, void *data)
{
struct vr_dpdk_rcu_cb_data *cb_data;
cb_data = CONTAINER_OF(rcd_user_data, struct vr_dpdk_rcu_cb_data, data);
cb_data->rcd_user_cb = user_cb;
cb_data->rcd_router = router;
call_rcu(&cb_data->rcd_rcu, dpdk_rcu_cb);
}
static void *
dpdk_get_defer_data(unsigned int len)
{
struct vr_dpdk_rcu_cb_data *cb_data;
if (!len)
return NULL;
cb_data = dpdk_malloc(sizeof(*cb_data) + len, VR_DEFER_OBJECT);
if (!cb_data) {
return NULL;
}
return cb_data->rcd_user_data;
}
static void
dpdk_put_defer_data(void *data)
{
struct vr_dpdk_rcu_cb_data *cb_data;
if (!data)
return;
cb_data = CONTAINER_OF(rcd_user_data, struct vr_dpdk_rcu_cb_data, data);
dpdk_free(cb_data, VR_DEFER_OBJECT);
return;
}
static void *
dpdk_network_header(struct vr_packet *pkt)
{
if (pkt->vp_network_h < pkt->vp_end)
return pkt->vp_head + pkt->vp_network_h;
/* TODO: for buffer chain? */
rte_panic("%s: buffer chain not supported\n", __func__);
return NULL;
}
static void *
dpdk_inner_network_header(struct vr_packet *pkt)
{
/* TODO: not used? */
rte_panic("%s: not implemented\n", __func__);
return NULL;
}
static void *
dpdk_data_at_offset(struct vr_packet *pkt, unsigned short off)
{
if (off < pkt->vp_end)
return pkt->vp_head + off;
/* TODO: for buffer chain? */
rte_panic("%s: buffer chain not supported\n", __func__);
return NULL;
}
/*
* dpdk_pheader_pointer
* return pointer to data at pkt->vp_data offset if hdr_len bytes
* in continuous memory, otherwise copy data to buf
*/
static void *
dpdk_pheader_pointer(struct vr_packet *pkt, unsigned short hdr_len, void *buf)
{
struct rte_mbuf *m;
int offset;
m = vr_dpdk_pkt_to_mbuf(pkt);
/*
* vp_data is offset from start of buffer,
* so first calculate offset from start of mbuf payload
*/
offset = pkt->vp_data - rte_pktmbuf_headroom(m);
if ((offset + hdr_len) < rte_pktmbuf_data_len(m))
return (void *)((uintptr_t)m->buf_addr + pkt->vp_data);
else {
int len = rte_pktmbuf_data_len(m) - offset;
void *tmp_buf = buf;
rte_memcpy(tmp_buf, rte_pktmbuf_mtod_offset(m, char *, offset), len);
hdr_len -= len;
tmp_buf = (void *)((uintptr_t)tmp_buf + len);
/* iterate thru buffers chain */
while (hdr_len) {
m = m->next;
if (!m)
return (NULL);
if (hdr_len > rte_pktmbuf_data_len(m))
len = rte_pktmbuf_data_len(m);
else
len = hdr_len;
rte_memcpy(tmp_buf, rte_pktmbuf_mtod(m, void *), len);
tmp_buf = (void *)((uintptr_t)tmp_buf + len);
hdr_len -= len;
}
return (buf);
}
}
/* VRouter callback */
static int
dpdk_pcow(struct vr_packet *pkt, unsigned short head_room)
{
struct rte_mbuf *mbuf = vr_dpdk_pkt_to_mbuf(pkt);
if (head_room > rte_pktmbuf_headroom(mbuf)) {
return -ENOMEM;
}
return 0;
}
/*
* dpdk_get_udp_src_port - return a source port for the outer UDP header.
* The source port is based on a hash of the inner IP source/dest addresses,
* vrf (and inner TCP/UDP ports in the future). The label from fmd
* will be used in the future to detect whether it is a L2/L3 packet.
* Returns 0 on error, valid source port otherwise.
*
* Based on linux/vrouter_mod.c:lh_get_udp_src_port
* Copyright (c) 2013, 2014 Juniper Networks, Inc.
*/
static uint16_t
dpdk_get_udp_src_port(struct vr_packet *pkt, struct vr_forwarding_md *fmd,
unsigned short vrf)
{
struct rte_mbuf *mbuf = vr_dpdk_pkt_to_mbuf(pkt);
unsigned int pull_len;
uint32_t ip_src, ip_dst, hashval, port_range;
struct vr_ip *iph;
uint16_t port;
uint16_t sport = 0, dport = 0;
struct vr_fragment *frag;
struct vrouter *router = vrouter_get(0);
uint32_t hash_key[5];
uint16_t *l4_hdr;
struct vr_flow_entry *fentry;
if (likely(mbuf->ol_flags & PKT_RX_RSS_HASH)) {
hashval = mbuf->hash.rss;
} else {
if (unlikely(hashrnd_inited == 0)) {
vr_hashrnd = random();
hashrnd_inited = 1;
}
if (pkt->vp_type == VP_TYPE_IP) {
/* Ideally the below code is only for VP_TYPE_IP and not
* for IP6. But having explicit check for IP only break IP6
*/
pull_len = sizeof(struct iphdr);
pull_len += pkt_get_network_header_off(pkt);
pull_len -= rte_pktmbuf_headroom(mbuf);
/* It's safe to assume the ip hdr is within this mbuf, so we skip
* all the header checks.
*/
iph = (struct vr_ip *)(mbuf->buf_addr + pkt_get_network_header_off(pkt));
if (vr_ip_transport_header_valid(iph)) {
if ((iph->ip_proto == VR_IP_PROTO_TCP) ||
(iph->ip_proto == VR_IP_PROTO_UDP)) {
l4_hdr = (__u16 *) (((char *) iph) + (iph->ip_hl * 4));
sport = *l4_hdr;
dport = *(l4_hdr+1);
}
} else {
/*
* If this fragment required flow lookup, get the source and
* dst port from the frag entry. Otherwise, use 0 as the source
* dst port (which could result in fragments getting a different
* outer UDP source port than non-fragments in the same flow).
*/
frag = vr_fragment_get(router, vrf, iph);
if (frag) {
sport = frag->f_sport;
dport = frag->f_dport;
}
}
if (fmd && fmd->fmd_flow_index >= 0) {
fentry = vr_get_flow_entry(router, fmd->fmd_flow_index);
if (fentry) {
vr_dpdk_mbuf_reset(pkt);
return fentry->fe_udp_src_port;
}
}
ip_src = iph->ip_saddr;
ip_dst = iph->ip_daddr;
hash_key[0] = ip_src;
hash_key[1] = ip_dst;
hash_key[2] = vrf;
hash_key[3] = sport;
hash_key[4] = dport;
hashval = rte_jhash(hash_key, 20, vr_hashrnd);
vr_dpdk_mbuf_reset(pkt);
} else {
/* We treat all non-ip packets as L2 here. For V6 we can extract
* the required fieleds explicity and manipulate the src port
*/
if (pkt_head_len(pkt) < ETH_HLEN)
goto error;
hashval = vr_hash(pkt_data(pkt), ETH_HLEN, vr_hashrnd);
/* Include the VRF to calculate the hash */
hashval = vr_hash_2words(hashval, vrf, vr_hashrnd);
}
} /* !PKT_RX_RSS_HASH */
/*
* Convert the hash value to a value in the port range that we want
* for dynamic UDP ports
*/
port_range = VR_MUDP_PORT_RANGE_END - VR_MUDP_PORT_RANGE_START;
port = (uint16_t) (((uint64_t) hashval * port_range) >> 32);
if (unlikely(port > port_range)) {
/*
* Shouldn't happen...
*/
port = 0;
}
return (port + VR_MUDP_PORT_RANGE_START);
error:
vr_dpdk_mbuf_reset(pkt);
return 0;
}
/**
* dpdk_adjust_tcp_mss - helper adjusting TCP Maximum Segment Size, used in
* dpdk_pkt_from_vm_tcp_mss_adj vRouter callback for packets from the VM and in
* vr_ip_transport_parse to perform MSS adjust for packets sent to the VM.
*/
void
dpdk_adjust_tcp_mss(struct tcphdr *tcph, unsigned short overlay_len,
unsigned char iph_len)
{
int opt_off = sizeof(struct tcphdr);
u_int8_t *opt_ptr = (u_int8_t *) tcph;
u_int16_t pkt_mss, max_mss, mtu;
unsigned int csum;
uint8_t port_id;
struct vrouter *router = vrouter_get(0);
if ((tcph == NULL) || !(tcph->syn) || (router == NULL))
return;
if (router->vr_eth_if == NULL)
return;
while (opt_off < (tcph->doff * 4)) {
switch (opt_ptr[opt_off]) {
case TCPOPT_EOL:
return;
case TCPOPT_NOP:
opt_off++;
continue;
case TCPOPT_MAXSEG:
if ((opt_off + TCPOLEN_MAXSEG) > (tcph->doff*4))
return;
if (opt_ptr[opt_off+1] != TCPOLEN_MAXSEG)
return;
pkt_mss = (opt_ptr[opt_off+2] << 8) | opt_ptr[opt_off+3];
if (router->vr_eth_if == NULL)
return;
port_id = (((struct vr_dpdk_ethdev *)(router->vr_eth_if->vif_os))->
ethdev_port_id);
rte_eth_dev_get_mtu(port_id, &mtu);
max_mss = mtu - (overlay_len + iph_len + sizeof(struct tcphdr));
if (pkt_mss > max_mss) {
opt_ptr[opt_off+2] = (max_mss & 0xff00) >> 8;
opt_ptr[opt_off+3] = max_mss & 0xff;
/* Recalculate checksum */
csum = (unsigned short)(~rte_cpu_to_be_16(tcph->check));
csum = csum + (unsigned short)~pkt_mss;
csum = (csum & 0xffff) + (csum >> 16);
csum += max_mss;
csum = (csum & 0xffff) + (csum >> 16);
tcph->check = rte_cpu_to_be_16(~((unsigned short)csum));
}
return;
default:
if ((opt_off + 1) == (tcph->doff*4))
return;
if (opt_ptr[opt_off+1])
opt_off += opt_ptr[opt_off+1];
else
opt_off++;
continue;
} /* switch */
} /* while */
return;
}
/*
* dpdk_pkt_from_vm_tcp_mss_adj - perform TCP MSS adjust, if required, for packets
* that are sent by a VM. Returns 0 on success, non-zero otherwise.
*/
static int
dpdk_pkt_from_vm_tcp_mss_adj(struct vr_packet *pkt, unsigned short overlay_len)
{
struct rte_mbuf *m = vr_dpdk_pkt_to_mbuf(pkt);
struct vr_ip *ip4h = NULL;
struct vr_ip6 *ip6h = NULL;
struct tcphdr *tcph;
int offset;
unsigned char iph_len = 0, iph_proto = 0;
/* check if whole ip header is in the packet */
if (pkt->vp_type == VP_TYPE_IP) {
offset = sizeof(struct vr_ip);
if (pkt->vp_data + offset < pkt->vp_end)
ip4h = (struct vr_ip *) ((uintptr_t)m->buf_addr + pkt->vp_data);
else
rte_panic("%s: ip header not in first buffer\n", __func__);
iph_proto = ip4h->ip_proto;
iph_len = ip4h->ip_hl * 4;
/*
* If this is a fragment and not the first one, it can be ignored
*/
if (ip4h->ip_frag_off & rte_cpu_to_be_16(IP_OFFMASK))
goto out;
} else if (pkt->vp_type == VP_TYPE_IP6) {
iph_len = offset = sizeof(struct vr_ip6);
if (pkt->vp_data + offset < pkt->vp_end)
ip6h = (struct vr_ip6 *) ((uintptr_t)m->buf_addr + pkt->vp_data);
else
rte_panic("%s: ip header not in first buffer\n", __func__);
iph_proto = ip6h->ip6_nxt;
}
if (iph_proto != VR_IP_PROTO_TCP)
goto out;
/*
* Now we know exact ip header length,
* check if whole tcp header is also in the packet
*/
offset = iph_len + sizeof(struct tcphdr);
if (pkt->vp_data + offset < pkt->vp_end)
tcph = (struct tcphdr *)pkt_data_at_offset(pkt, pkt->vp_data + iph_len);
else
rte_panic("%s: tcp header not in first buffer\n", __func__);
if ((tcph->doff << 2) <= (sizeof(struct tcphdr))) {
/*Nothing to do if there are no TCP options */
goto out;
}
offset += (tcph->doff << 2) - sizeof(struct tcphdr);
if (pkt->vp_data + offset > pkt->vp_end)
rte_panic("%s: tcp header outside first buffer\n", __func__);
dpdk_adjust_tcp_mss(tcph, overlay_len, iph_len);
out:
return 0;
}
static unsigned int
dpdk_pgso_size(struct vr_packet *pkt)
{
/* TODO: not implemented */
return 0;
}
static void
dpdk_add_mpls(struct vrouter *router, unsigned mpls_label)
{
int ret, i;
struct vr_interface *eth_vif;
for (i = 0; i < router->vr_max_interfaces; i++) {
eth_vif = __vrouter_get_interface(router, i);
if (eth_vif && (eth_vif->vif_type == VIF_TYPE_PHYSICAL)
&& (eth_vif->vif_flags & VIF_FLAG_FILTERING_OFFLOAD)) {
RTE_LOG(INFO, VROUTER, "Enabling hardware acceleration on vif %u for MPLS %u\n",
(unsigned)eth_vif->vif_idx, mpls_label);
if (!eth_vif->vif_ip) {
RTE_LOG(ERR, VROUTER, " error accelerating MPLS %u: no IP address set\n",
mpls_label);
continue;
}
ret = vr_dpdk_lcore_mpls_schedule(eth_vif, eth_vif->vif_ip, mpls_label);
if (ret != 0)
RTE_LOG(INFO, VROUTER, " error accelerating MPLS %u: %s (%d)\n",
mpls_label, rte_strerror(-ret), -ret);
}
}