summaryrefslogtreecommitdiffstats
path: root/RELEASE.md
blob: 142a1dd2011b483a77740042369d481dc2ec3d1f (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
# Release Notes    {#release_notes}

* @subpage release_notes_1802
* @subpage release_notes_1710
* @subpage release_notes_1707
* @subpage release_notes_1704
* @subpage release_notes_17011
* @subpage release_notes_1701
* @subpage release_notes_1609
* @subpage release_notes_1606

@page release_notes_1802 Release notes for VPP 18.02

TBD

@page release_notes_1710 Release notes for VPP 17.10

More than 400 commits since the 1707 release.

## Features
- Infrastructure
  - DPDK 17.08
  - IP reassembly
  - Bounded-index extensible hash bucket-level LRU cache
  - Templated timer wheel improvements

- API
  - C/C++ language binding
  - API stats

- Host stack
  - VPP TCP stack scale/congestion improvements
  - VPP Comms Library (VCL)
  - Overall performance, scale and hardening

- Network features
  - IPSec rework - utilize new FIB
  - VPLS and VPWS implementation

  - NAT
    - Renamed SNAT to NAT
    - Performance / Scale
    - Destination NAT44 with load-balancing
    - In2out translation as an output feature on the outside interface
    - Fullback to 3-tuple key for non TCP/UDP/ICMP sessions

  - Security Groups/ACLs
    - "Replace" semantics for adding a new MacIP acl
    - Test suite tests for MacIP ACLs

  - ONE-LISP
    - Map-server fallback support
    - Preemptive re-fetch of active mappings that are about to expire
    - ND termination

  - PPPoE
    - PPPoE Control Plane packet dispatch
    - PPPoE decapsulation
    - PPPoE encapsulation

## Known issues

For the full list of issues please refer to fd.io [JIRA](https://jira.fd.io).

## Issues fixed

For the full list of fixed issues please refer to:
- fd.io [JIRA](https://jira.fd.io)
- git [commit log](https://git.fd.io/vpp/log/?h=stable/1710)

## API changes

Message Name                         			     Result
bridge_domain_add_del                                        definition changed
bridge_domain_details                                        definition changed
connect_session                                              definition changed
connect_sock                                                 definition changed
connect_sock_reply                                           definition changed
connect_uri_reply                                            definition changed
create_vhost_user_if                                         definition changed
dhcp_client_config                                           definition changed
ip4_arp_event                                                definition changed
ip6_fib_details                                              definition changed
ip6_nd_event                                                 definition changed
ip_add_del_route                                             definition changed
ip_fib_details                                               definition changed
ip_table_add_del                                             definition changed
l2_macs_event                                                only in image
macip_acl_add_replace                                        definition changed
macip_acl_interface_list_details                             only in image
macip_acl_interface_list_dump                                only in image
modify_vhost_user_if                                         definition changed
mpls_fib_details                                             definition changed
mpls_route_add_del                                           definition changed
mpls_table_add_del                                           definition changed
mpls_tunnel_add_del                                          definition changed
nat44_add_del_address_range                                  definition changed
nat44_add_del_interface_addr                                 definition changed
nat44_add_del_lb_static_mapping                              definition changed
nat44_add_del_static_mapping                                 definition changed
nat44_address_details                                        only in image
nat44_address_dump                                           only in image
nat44_interface_add_del_feature                              definition changed
nat44_interface_add_del_output_feature                       definition changed
nat44_interface_addr_details                                 only in image
nat44_interface_addr_dump                                    only in image
nat44_interface_details                                      only in image
nat44_interface_dump                                         only in image
nat44_interface_output_feature_details                       only in image
nat44_interface_output_feature_dump                          only in image
nat44_lb_static_mapping_details                              only in image
nat44_lb_static_mapping_dump                                 only in image
nat44_static_mapping_details                                 only in image
nat44_static_mapping_dump                                    only in image
nat44_user_details                                           only in image
nat44_user_dump                                              only in image
nat44_user_session_details                                   only in image
nat44_user_session_dump                                      only in image
nat_control_ping                                             definition changed
nat_det_add_del_map                                          definition changed
nat_det_close_session_in                                     definition changed
nat_det_close_session_out                                    definition changed
nat_det_forward                                              definition changed
nat_det_get_timeouts                                         definition changed
nat_det_map_details                                          only in image
nat_det_map_dump                                             only in image
nat_det_reverse                                              definition changed
nat_det_session_details                                      only in image
nat_det_session_dump                                         only in image
nat_det_set_timeouts                                         definition changed
nat_ipfix_enable_disable                                     definition changed
nat_set_workers                                              definition changed
nat_show_config                                              definition changed
nat_worker_details                                           only in image
nat_worker_dump                                              only in image
one_add_del_ndp_entry                                        definition changed
one_enable_disable_petr_mode                                 definition changed
one_enable_disable_pitr_mode                                 definition changed
one_enable_disable_xtr_mode                                  definition changed
one_get_transport_protocol                                   definition changed
one_map_register_fallback_threshold                          definition changed
one_map_register_set_ttl                                     definition changed
one_ndp_bd_get                                               definition changed
one_ndp_entries_get                                          definition changed
one_set_transport_protocol                                   definition changed
one_show_petr_mode                                           definition changed
one_show_pitr_mode                                           definition changed
one_show_xtr_mode                                            definition changed
p2p_ethernet_add                                             definition changed
pppoe_add_del_session                                        definition changed
pppoe_session_details                                        only in image
pppoe_session_dump                                           only in image
punt_socket_deregister                                       definition changed
punt_socket_register                                         definition changed
show_one_map_register_fallback_threshold                     definition changed
show_one_map_register_ttl                                    definition changed
snat_interface_add_del_output_feature                        definition changed
snat_interface_output_feature_details                        only in image
snat_interface_output_feature_dump                           only in image
sw_interface_event                                           only in image
sw_interface_set_flags                                       definition changed
sw_interface_span_dump                                       definition changed
sw_interface_span_enable_disable                             definition changed
sw_interface_vhost_user_details                              definition changed
tcp_configure_src_addresses                                  definition changed
vnet_per_interface_combined_counters                         only in image
vnet_per_interface_simple_counters                           only in image
want_interface_combined_stats                                definition changed
want_interface_simple_stats                                  definition changed
want_ip4_fib_stats                                           definition changed
want_ip4_nbr_stats                                           definition changed
want_ip6_fib_stats                                           definition changed
want_ip6_nbr_stats                                           definition changed
want_l2_macs_events                                          definition changed
want_per_interface_combined_stats                            definition changed
want_per_interface_simple_stats                              definition changed
Found 103 api message signature differences

Patches that updated the API files:

./src/plugins/pppoe/pppoe.api
62f9cdd8 Add PPPoE Plugin

./src/plugins/acl/acl.api
c29940c5 ACL-plugin add "replace" semantics for adding a new MacIP acl
de9fbf43 MAC IP ACL interface list dump (as an alternative to the get/reply)

./src/plugins/nat/nat.api
704018cf NAT: Destination NAT44 with load-balancing (VPP-954)
2ba92e32 NAT: Rename snat plugin to nat (VPP-955)

./src/vnet/interface.api
831fb59f Stats refactor
d292ab1e No context in SW interface event
a07bd708 Dedicated SW Interface Event

./src/vnet/dhcp/dhcp.api
51822bf0 DHCP client option 61 "client_id"
4729b1ec DHCP complete event sends mask length

./src/vnet/lldp/lldp.api
99a0e60e Add API support for LLDP config/interface set

./src/vnet/lisp-cp/one.api
d630713d LISP: add neighbor discovery and CP protocol separation APIs
111a5cea LISP: Add APIs for enable/disable xTR/P-ITR/P-ETR modes
7048ff1e LISP: Map-server fallback feature
1e553a00 LISP: make TTL for map register messages configurable

./src/vnet/ethernet/p2p_ethernet.api
15ac81c1 P2P Ethernet

./src/vnet/mpls/mpls.api
2297af01 Add a name to the creation of an IP and MPLS table
28ab9cc1 FIB table add/delete API only
da78f957 L2 over MPLS
a0a908f1 FIB path weight incorrect in dump (VPP-922)
57b5860f FIB path preference

./src/vnet/session/session.api
33e002b1 Fix session connect api message handling.

./src/vnet/span/span.api
5b311202 SPAN/API:enable L2 dump
001fd406 SPAN:add l2 mirror

./src/vnet/devices/virtio/vhost_user.api
4ba75f54 vhost: Remove operation mode in the API

./src/vnet/vxlan-gpe/vxlan_gpe.api
04ffd0ad VPP crash on creating vxlan gpe interface. VPP-875

./src/vnet/tcp/tcp.api
3bbcfab1 TCP source address automation

./src/vnet/ip/ip.api
2297af01 Add a name to the creation of an IP and MPLS table
28ab9cc1 FIB table add/delete API only
57b5860f FIB path preference

./src/vnet/lisp-gpe/lisp_gpe.api
af3d9771 Remove unused retval from gpe_native_fwd_rpath type definition

./src/vnet/l2/l2.api
50570ece Update of free text tag patch for BD
48304141 Support for bridge domain free text tag
e531f4cb Increase default MAC learn limit and check it in learn-update path
8d00fff8 Add support for API client to receive L2 MAC events

./src/vpp/api/vpe.api
8a19f12a Allow individual stats API and introduce stats.api
4802632d Punt socket: Fix coverity error for pathname length mismatch between API and sun_path.
f7a55ad7 PUNT socket: External control plane processes connected via UNIX domain sockets.
75e2f2ac API:fix arp/ND event messages - remove context
99a0e60e Add API support for LLDP config/interface set

./src/vpp/stats/stats.api
831fb59f Stats refactor
8a19f12a Allow individual stats API and introduce stats.api

@page release_notes_1707 Release notes for VPP 17.07

More than 400 commits since the 1704 release.

## Features
- Infrastructure
  - make test; improved debuggability.
  - TAB auto-completion on the CLI
  - DPDK 17.05
  - python 3 support in test infra

- Host stack
  - Improved Linux TCP stack compatibility using IWL test suite (https://jira.fd.io/browse/VPP-720)
  - Improved loss recovery (RFC5681, RFC6582, RF6675)
  - Basic implementation of Eifel detection algorithm (RFC3522)
  - Basic support for buffer chains
  - Refactored session layer API
  - Overall performance, scale and hardening

- Interfaces
  - memif: IP mode, jumbo frames, multi queue
  - virtio-user support
  - vhost-usr; adaptive (poll/interupt) support.

- Network features
  - MPLS Multicast FIB

  - BFD FIB integration

  - NAT64 support

  - GRE over IPv6

  - Segement routing MPLS

  - IOAM configuration for SRv6 localsid

  - LISP
    - NSH support
    - native forward static routes
    - L2 ARP

  - ACL multi-core suuport

  - Flowprobe:
    - Add flowstartns, flowendns and tcpcontrolbits
    - Stateful flows and IPv6, L4 recording

  - GTP-U support

  - VXLAN GPE support for FIB2.0 and bypass.


## Known issues

For the full list of issues please reffer to fd.io [JIRA](https://jira.fd.io).

## Issues fixed

For the full list of fixed issues please reffer to:
- fd.io [JIRA](https://jira.fd.io)
- git [commit log](https://git.fd.io/vpp/log/?h=stable/1707)

@page release_notes_1704 Release notes for VPP 17.04

More than 500 commits since the 1701 release.

## Features
- Infrastructure
  - make test improvements
  - vnet: add device-input threadplacement infra
  - 64 bit per-thread counters
  - process restart cli
  - High performance timer wheels
  - Plugin infrastructure improvements
    - Support for .default_disabled, .version_required
  - Added MAINTAINERS file

- Host stack
  - TCP stack (experimental)
  - DHCPv4 / DHCPv6 relay multi-destination
  - DHCPv4 option 82
  - ND proxy
  - Attached hosts
  - Consolidated DHCPv4 and DHCPv6 implementation

- Interfaces
  - DPDK 17.02 (retire support for DPDK 16.07)
  - Add memif - packet memory interface for intra-host communication
  - vhost: support interrupt mode
  - DPDK as plugin (retired vpp_lite)
  - DPDPK input optimizations
  - Loopback interface allocation scheme

- Network features
  - IP Multicast FIB

  - Bridging
    - Learning on local interfaces
    - Flushing of MACs from the L2 FIB

  - SNAT
    - CGN (Deterministic and dynamic)
    - CGN configurable port allocation algorithm
    - ICMP support
    - Tentant VRF id for SNAT outside addresses
    - Session dump / User dump
    - Port allocation per protocol

  - Security groups
    - Routed interface support
    - L2+L3 unified processing node
    - Improve fragment handling

  - Segement routing v6
    - SR policies with weighted SID lists
    - Binding SID
    - SR steering policies
    - SR Local SIDs
    - Framework to expand local SIDs w/plugins
    - Documentation

  - IOAM
    - UDP Pinger w/path fault isolation
    - IOAM as type 2 metadata in NSH
    - IAOM raw IPFIX collector and analyzer
    - Anycast active server selection
    - Documentation
    - SRv6 Local SID
    - IP6 HBH header and SR header co-existence
    - Active probe

  - LISP
    - Statistics collection
    - Generalize encap for overlay transport (vxlan-gpe support)
    - Improve data plane speed

  - GPE
    - CLI
    - NSH added to encap/decap path
    - Renamed LISP GPE API to GPE

  - MPLS
    - Performance improvements (quad loop)

  - BFD
    - Command line interface
    - Echo function
    - Remote demand mode
    - SHA1 authentication

  - IPsec
    - IKEv2 initiator features

  - VXLAN
    - unify IP4/IP6 control plane handling

## API changes

- Python API: To avoid conflicts between VPP API messages names and
  the Python API binding function names, VPP API methods are put in a
  separate proxy object.
  https://gerrit.fd.io/r/#/c/5570/
  The api methods are now referenced as:
    vpp_handle = VPP(jsonfiles)
    vpp_handle.connect(...)
    vpp = vpp_handle.api
    vpp.show_version()
    vpp_handle.disconnect()

  For backwards compatibility VPP API methods are left in the main
  name space (VPP), but will be removed from 17.07.

  - Python API: Change from cPython to CFFI.

- create_loopback message to be replaced with create_loopback_instance
  create_loopback will be removed from 17.07.
  https://gerrit.fd.io/r/#/c/5572/

## Known issues

For the full list of issues please reffer to fd.io [JIRA](https://jira.fd.io).

## Issues fixed

For the full list of fixed issues please reffer to:
- fd.io [JIRA](https://jira.fd.io)
- git [commit log](https://git.fd.io/vpp/log/?h=stable/1704)

@page release_notes_17011 Release notes for VPP 17.01.1

This is bug fix release.

For the full list of fixed issues please reffer to:
- fd.io [JIRA](https://jira.fd.io)
- git [commit log](https://git.fd.io/vpp/log/?h=stable/1701)


@page release_notes_17011 Release notes for VPP 17.01.1

This is bug fix release.

For the full list of fixed issues please reffer to:
- fd.io [JIRA](https://jira.fd.io)
- git [commit log](https://git.fd.io/vpp/log/?h=stable/1701)

@page release_notes_1701 Release notes for VPP 17.01

@note This release was for a while known as 16.12.

## Features

- [Integrated November 2016 DPDK release](http://www.dpdk.org/doc/guides/rel_notes/release_16_11.html)

- Complete rework of Forwarding Information Base (FIB)

- Performance Improvements
  - Improvements in DPDK input and output nodes
  - Improvements in L2 path
  - Improvmeents in IPv4 lookup node

- Feature Arcs Improvements
  - Consolidation of the code
  - New feature arcs
    - device-input
    - interface-output

- DPDK Cryptodev Support
  - Software and Hardware Crypto Support

- DPDK HQoS support

- Simple Port Analyzer (SPAN)

- Bidirectional Forwarding Detection
  - Basic implementation

- IPFIX Improvements

- L2 GRE over IPSec tunnels

- Link Layer Discovery Protocol (LLDP)

- Vhost-user Improvements
  - Performance Improvements
  - Multiqueue
  - Reconnect

- LISP Enhancements
  - Source/Dest control plane support
  - L2 over LISP and GRE
  - Map-Register/Map-Notify/RLOC-probing support
  - L2 API improvements, overall code hardening

- Plugins:
  - New: ACL
  - New: Flow per Packet
  - Improved: SNAT
    - Mutlithreading
    - Flow export

- Doxygen Enhancements

- Luajit API bindings

- API Refactoring
  - file split
  - message signatures

- Python and Scapy based unit testing infrastructure
  - Infrastructure
  - Various tests

- Packet Generator improvements

- TUN/TAP jumbo frames support

- Other various bug fixes and improvements

## Known issues

For the full list of issues please reffer to fd.io [JIRA](https://jira.fd.io).

## Issues fixed

For the full list of fixed issues please reffer to:
- fd.io [JIRA](https://jira.fd.io)
- git [commit log](https://git.fd.io/vpp/log/?h=stable/1701)

@page release_notes_1609 Release notes for VPP 16.09

## Features

- [Integrated July 2016 DPDK release](http://www.dpdk.org/doc/guides/rel_notes/release_16_07.html)
  - DPDK-vhost is depreciated pending a complete rework of the original integration and
    addressing of rx performance deltas.
  - Patches required for DPDK 16.07:
    - Correctly setting the Packet Type in the IGB, IXGBE and i40e drivers.
    - Correctly setting checksum in the i40e driver.
    - NXP DPAA2 PMD Driver.
    - rte_delay (yield) functionality.

- Add “in tree” plugins:
  - IPv6 ILA.
  - iOAM.
  - Load Balancer.
  - SNAT.

- High-performance (line-rate) “neutron like” L4 port-filtering.

- API refactoring - addressing some of the issues around JVPP bindings.
  - Accommodating plugins [(e.g. NSH_SFC)](https://wiki.fd.io/view/NSH_SFC)
  - Binding for [python](https://wiki.fd.io/view/VPP/Python_API)

- LISP
  - L2 LISP overlays
  -  Multitenancy
  - Multihoming
  - RTR mode
  - Map-resolver failover algorithm

- Support 64-bit vector lengths, huge shared-memory segments.

- Dynamic IP Feature ordering
  - IP Features can now specify features they appear before and after

- 16.09 Builds
  - Ubuntu 14.04 LTS - Trusty Tahr
  - Ubuntu 16.04 LTS - Xenial Xerus
  - CentOS 7
  - More information on [VPP wiki](https://wiki.fd.io/view/VPP/Installing_VPP_binaries_from_packages)

- Performance, characterize and document performance for this release
  [(more information on CSIT page)](https://wiki.fd.io/view/CSIT)

   - IPv4 and IPv6 Scale - performance tests.
     - Bidirectional 10k/100k/1M flows.
     - 64B,570B, 1518B,9000B packet sizes.
   - IPv6 iACL - performance
     - DUT1 and DUT2 are configured with IPv6 routing, two static IPv6 /64 routes and IPv6 iAcl
       security whitelist ingress /64 filter entries applied on links.
     - TG traffic profile contains two L3 flow-groups (flow-group per direction, 253 flows per
       flow-group) with all packets containing Ethernet header, IPv6 header and generated payload.
       MAC addresses are matching MAC addresses of the TG node interfaces.

   - L2XC VXLANoIPv4 - performance
     - DUT1 and DUT2 are configured with L2 cross-connect. VXLAN tunnels are configured between
       L2XCs on DUT1 and DUT2.
     - TG traffic profile contains two L3 flow-groups (flow-group per direction, 253 flows per
       flow-group) with all packets containing Ethernet header, IPv4 header with IP protocol=61
       and generated payload. MAC addresses are matching MAC addresses of the TG node interfaces.

- Documentation
  - Autogenerated CLI documentation.
  - Using doxygen to automate API/Node documentation.
  - [(available online)](https://docs.fd.io/vpp/16.09/)

- Resolved all static analysis issues found by Coverity
  - Beginning of 16.09 cycle: 505 issues.
  - Release: 0 outstanding issues.


## Known issues

Issues in fd.io are tracked in [JIRA](https://jira.fd.io).

Issue | Description
--- | ---
VPP-391 |   vpp debug version assert appeared in the process of start
VPP-380 |   Mapping algorithm compute wrong ea-bits when IPv4 prefix 0.0.0.0/0
VPP-371 |   load_one_plugin:63: Loaded plugin: message from vppctl
VPP-367 |   vpp packages need to depend on specific versions of each other
VPP-312 |   IP6 FIB gets in indeterminate state by duplicating commands
VPP-224 |   Lookup-in-vrf can not be set correctly
VPP-206 |   Fix classify table delete
VPP-203 |   Fix binary API for reading vpp node graph
VPP-147 |   Inconsistent behaviour when adding L2 FIB filter entry
VPP-99  |  VPP doesn't discard DHCPOFFER message with wrong XID


## Issues fixed

Issues in fd.io are tracked in [JIRA](https://jira.fd.io).

Issue | Description
--- | ---
VPP-396 |   Ubuntu systems Graphviz bug
VPP-390 |   vpp-lib rpm fails to include *.so symlinks, causing linking problems with out of tree builds
VPP-388 |   IPSec output feature assumes packets have been ethernet rewritten
VPP-385 |   ARP for indirect adjacencies not working correctly
VPP-361 |   Memory leak on delete of VXLAN over IPv6 tunnel
VPP-357 |   VNI not set correctly when removing LISP fwd entries
VPP-349 |   sw_interface_vhost_user_dump not working
VPP-345 |   net/enic: bad L4 checksum ptype set on ICMP packets
VPP-340 |   MAP-T wrong destination address
VPP-330 |   Use fifo to store LISP pending map-requests
VPP-326 |   map_add_domain VAT command: unable to configure domain with mtu parameter
VPP-318 |   The map_add_domain VAT command accepts invalid arguments
VPP-315 |   Fix "show vxlan-gpe" issue
VPP-310 |   Mapping algorithm compute wrong ea-bits
VPP-239 |   LISP IP forwarding does not tag packets that hit negative mapping entries
VPP-235 |   Invalid help in VAT for sw_interface_set_l2_bridge
VPP-228 |   Mapping algorithm sends packet to wrong IPv6 address
VPP-214 |   vpp-api-test: api_ipsec_sad_add_del_entry: vector "ck" not initialized
VPP-200 |   VPP - TAP port create problem
VPP-189 |   Coverity Issues for 16.09
VPP-184 |   u16 translating to char ,not short
VPP-179 |   Adjacency share-count botch
VPP-163 |   "show ip6 interface" ignores non-global addresses
VPP-155 |   Netmap: Inconsistency in interface state between "show hardware" and "show interface"
VPP-145 |   Dynamically compute IP feature ordering based on constraints
VPP-137 |   VPP sends ARP with wrong requested IP
VPP-118 |   JVpp: 0 length arrays not handled properly in VPP responses
VPP-112 |   linux kernel info missing from build log
VPP-110 |   vxlan encap node should never touch a deleted tunnel
VPP-107 |   RPM build broken in master
VPP-92  |   segment routing is not properly filling out the segment list
VPP-91  |   segment routing add/del tunnel lookup doesn't work
VPP-84  |   af_packet throws a fatal error on EAGAIN
VPP-74  |   Clang compile fails due to warning in vlib/unix/cli.c
VPP-64  |   Top level "make pkg-deb" fails if CDPATH is set in user env.
VPP-48  |   Traceroute does not terminate when VPP is the target
VPP-23  |   CLI pager does not gracefully handle lines longer than the terminal width


@page release_notes_1606 Release notes for VPP 16.06


The FD.io Project, relentlessly focused on data IO speed and efficiency
supporting the creation of high performance, flexible, and scalable software
defined infrastructures, announces the availability of the community’s first
software release (16.06).

In the four months since launching, FD.io has brought together more than 75
developers from 11 different companies including network operators, solution
providers chip vendors, and network equipment vendors who are collaborating to
enhance and innovate around the Vector Packet Processing (VPP) technology. The
FD.io community has quickly formed to grow the number of projects from the
initial VPP project to an additional 6 projects addressing a diverse set of
requirements and usability across a variety of deployment environments.

The 16.06 release brings unprecedented performance: 480Gbps/200mpps with 8
million routes and 2k whitelist entries on standard high volume x86 servers.


## Features

In addition to the existing full suite of vswitch/vrouter features, the new
16.06 release adds:

* Enhanced Switching and Routing:
  * IPv6 Segment Routing multicast support.
  * LISP xTR support.
  * VXLAN over IPv6 underlay.
  * Per interface whitelists.
  * Shared adjacencies in FIB.

* New and improved interface support:
  * Jumbo frame support for vhost-user.
  * Netmap interface support.
  * AF_Packet interface support.

* Expanded and improved programmability:
  * Python API bindings.
  * Enhanced JVPP Java API bindings.
  * Debugging CLI.

* Expanded Hardware and Software Support:
  * Support for ARM 32 targets including Rasberry Pi single-board computer.
  * Support for DPDK 16.04.
lass="o">-1; } } if (PREDICT_FALSE (svm_msg_q_ring_is_full (mq, SESSION_MQ_IO_EVT_RING))) { svm_msg_q_unlock (mq); return -2; } msg = svm_msg_q_alloc_msg_w_ring (mq, SESSION_MQ_IO_EVT_RING); if (PREDICT_FALSE (svm_msg_q_msg_is_invalid (&msg))) { svm_msg_q_unlock (mq); return -2; } evt = (session_event_t *) svm_msg_q_msg_data (mq, &msg); evt->event_type = evt_type; switch (evt_type) { case FIFO_EVENT_RPC: evt->rpc_args.fp = data; evt->rpc_args.arg = args; break; case FIFO_EVENT_APP_TX: case SESSION_IO_EVT_TX_FLUSH: case FIFO_EVENT_BUILTIN_RX: evt->fifo = data; break; case FIFO_EVENT_BUILTIN_TX: case FIFO_EVENT_DISCONNECT: evt->session_handle = session_handle ((stream_session_t *) data); break; default: clib_warning ("evt unhandled!"); svm_msg_q_unlock (mq); return -1; } svm_msg_q_add_and_unlock (mq, &msg); return 0; } int session_send_io_evt_to_thread (svm_fifo_t * f, session_evt_type_t evt_type) { return session_send_evt_to_thread (f, 0, f->master_thread_index, evt_type); } int session_send_io_evt_to_thread_custom (void *data, u32 thread_index, session_evt_type_t evt_type) { return session_send_evt_to_thread (data, 0, thread_index, evt_type); } int session_send_ctrl_evt_to_thread (stream_session_t * s, session_evt_type_t evt_type) { /* only event supported for now is disconnect */ ASSERT (evt_type == FIFO_EVENT_DISCONNECT); return session_send_evt_to_thread (s, 0, s->thread_index, FIFO_EVENT_DISCONNECT); } void session_send_rpc_evt_to_thread (u32 thread_index, void *fp, void *rpc_args) { if (thread_index != vlib_get_thread_index ()) session_send_evt_to_thread (fp, rpc_args, thread_index, FIFO_EVENT_RPC); else { void (*fnp) (void *) = fp; fnp (rpc_args); } } stream_session_t * session_alloc (u32 thread_index) { session_manager_worker_t *wrk = &session_manager_main.wrk[thread_index]; stream_session_t *s; u8 will_expand = 0; pool_get_aligned_will_expand (wrk->sessions, will_expand, CLIB_CACHE_LINE_BYTES); /* If we have peekers, let them finish */ if (PREDICT_FALSE (will_expand && vlib_num_workers ())) { clib_rwlock_writer_lock (&wrk->peekers_rw_locks); pool_get_aligned (wrk->sessions, s, CLIB_CACHE_LINE_BYTES); clib_rwlock_writer_unlock (&wrk->peekers_rw_locks); } else { pool_get_aligned (wrk->sessions, s, CLIB_CACHE_LINE_BYTES); } clib_memset (s, 0, sizeof (*s)); s->session_index = s - wrk->sessions; s->thread_index = thread_index; return s; } void session_free (stream_session_t * s) { pool_put (session_manager_main.wrk[s->thread_index].sessions, s); if (CLIB_DEBUG) clib_memset (s, 0xFA, sizeof (*s)); } void session_free_w_fifos (stream_session_t * s) { segment_manager_dealloc_fifos (s->svm_segment_index, s->server_rx_fifo, s->server_tx_fifo); session_free (s); } int session_alloc_fifos (segment_manager_t * sm, stream_session_t * s) { svm_fifo_t *server_rx_fifo = 0, *server_tx_fifo = 0; u32 fifo_segment_index; int rv; if ((rv = segment_manager_alloc_session_fifos (sm, &server_rx_fifo, &server_tx_fifo, &fifo_segment_index))) return rv; /* Initialize backpointers */ server_rx_fifo->master_session_index = s->session_index; server_rx_fifo->master_thread_index = s->thread_index; server_tx_fifo->master_session_index = s->session_index; server_tx_fifo->master_thread_index = s->thread_index; s->server_rx_fifo = server_rx_fifo; s->server_tx_fifo = server_tx_fifo; s->svm_segment_index = fifo_segment_index; return 0; } static stream_session_t * session_alloc_for_connection (transport_connection_t * tc) { stream_session_t *s; u32 thread_index = tc->thread_index; ASSERT (thread_index == vlib_get_thread_index () || transport_protocol_is_cl (tc->proto)); s = session_alloc (thread_index); s->session_type = session_type_from_proto_and_ip (tc->proto, tc->is_ip4); s->enqueue_epoch = (u64) ~ 0; /* Attach transport to session and vice versa */ s->connection_index = tc->c_index; tc->s_index = s->session_index; return s; } static int session_alloc_and_init (segment_manager_t * sm, transport_connection_t * tc, u8 alloc_fifos, stream_session_t ** ret_s) { stream_session_t *s; int rv; s = session_alloc_for_connection (tc); if (alloc_fifos && (rv = session_alloc_fifos (sm, s))) { session_free (s); *ret_s = 0; return rv; } /* Add to the main lookup table */ session_lookup_add_connection (tc, session_handle (s)); *ret_s = s; return 0; } /** * Discards bytes from buffer chain * * It discards n_bytes_to_drop starting at first buffer after chain_b */ always_inline void session_enqueue_discard_chain_bytes (vlib_main_t * vm, vlib_buffer_t * b, vlib_buffer_t ** chain_b, u32 n_bytes_to_drop) { vlib_buffer_t *next = *chain_b; u32 to_drop = n_bytes_to_drop; ASSERT (b->flags & VLIB_BUFFER_NEXT_PRESENT); while (to_drop && (next->flags & VLIB_BUFFER_NEXT_PRESENT)) { next = vlib_get_buffer (vm, next->next_buffer); if (next->current_length > to_drop) { vlib_buffer_advance (next, to_drop); to_drop = 0; } else { to_drop -= next->current_length; next->current_length = 0; } } *chain_b = next; if (to_drop == 0) b->total_length_not_including_first_buffer -= n_bytes_to_drop; } /** * Enqueue buffer chain tail */ always_inline int session_enqueue_chain_tail (stream_session_t * s, vlib_buffer_t * b, u32 offset, u8 is_in_order) { vlib_buffer_t *chain_b; u32 chain_bi, len, diff; vlib_main_t *vm = vlib_get_main (); u8 *data; u32 written = 0; int rv = 0; if (is_in_order && offset) { diff = offset - b->current_length; if (diff > b->total_length_not_including_first_buffer) return 0; chain_b = b; session_enqueue_discard_chain_bytes (vm, b, &chain_b, diff); chain_bi = vlib_get_buffer_index (vm, chain_b); } else chain_bi = b->next_buffer; do { chain_b = vlib_get_buffer (vm, chain_bi); data = vlib_buffer_get_current (chain_b); len = chain_b->current_length; if (!len) continue; if (is_in_order) { rv = svm_fifo_enqueue_nowait (s->server_rx_fifo, len, data); if (rv == len) { written += rv; } else if (rv < len) { return (rv > 0) ? (written + rv) : written; } else if (rv > len) { written += rv; /* written more than what was left in chain */ if (written > b->total_length_not_including_first_buffer) return written; /* drop the bytes that have already been delivered */ session_enqueue_discard_chain_bytes (vm, b, &chain_b, rv - len); } } else { rv = svm_fifo_enqueue_with_offset (s->server_rx_fifo, offset, len, data); if (rv) { clib_warning ("failed to enqueue multi-buffer seg"); return -1; } offset += len; } } while ((chain_bi = (chain_b->flags & VLIB_BUFFER_NEXT_PRESENT) ? chain_b->next_buffer : 0)); if (is_in_order) return written; return 0; } /* * Enqueue data for delivery to session peer. Does not notify peer of enqueue * event but on request can queue notification events for later delivery by * calling stream_server_flush_enqueue_events(). * * @param tc Transport connection which is to be enqueued data * @param b Buffer to be enqueued * @param offset Offset at which to start enqueueing if out-of-order * @param queue_event Flag to indicate if peer is to be notified or if event * is to be queued. The former is useful when more data is * enqueued and only one event is to be generated. * @param is_in_order Flag to indicate if data is in order * @return Number of bytes enqueued or a negative value if enqueueing failed. */ int session_enqueue_stream_connection (transport_connection_t * tc, vlib_buffer_t * b, u32 offset, u8 queue_event, u8 is_in_order) { stream_session_t *s; int enqueued = 0, rv, in_order_off; s = session_get (tc->s_index, tc->thread_index); if (is_in_order) { enqueued = svm_fifo_enqueue_nowait (s->server_rx_fifo, b->current_length, vlib_buffer_get_current (b)); if (PREDICT_FALSE ((b->flags & VLIB_BUFFER_NEXT_PRESENT) && enqueued >= 0)) { in_order_off = enqueued > b->current_length ? enqueued : 0; rv = session_enqueue_chain_tail (s, b, in_order_off, 1); if (rv > 0) enqueued += rv; } } else { rv = svm_fifo_enqueue_with_offset (s->server_rx_fifo, offset, b->current_length, vlib_buffer_get_current (b)); if (PREDICT_FALSE ((b->flags & VLIB_BUFFER_NEXT_PRESENT) && !rv)) session_enqueue_chain_tail (s, b, offset + b->current_length, 0); /* if something was enqueued, report even this as success for ooo * segment handling */ return rv; } if (queue_event) { /* Queue RX event on this fifo. Eventually these will need to be flushed * by calling stream_server_flush_enqueue_events () */ session_manager_worker_t *wrk; wrk = session_manager_get_worker (s->thread_index); if (s->enqueue_epoch != wrk->current_enqueue_epoch[tc->proto]) { s->enqueue_epoch = wrk->current_enqueue_epoch[tc->proto]; vec_add1 (wrk->session_to_enqueue[tc->proto], s->session_index); } } return enqueued; } int session_enqueue_dgram_connection (stream_session_t * s, session_dgram_hdr_t * hdr, vlib_buffer_t * b, u8 proto, u8 queue_event) { int enqueued = 0, rv, in_order_off; ASSERT (svm_fifo_max_enqueue (s->server_rx_fifo) >= b->current_length + sizeof (*hdr)); svm_fifo_enqueue_nowait (s->server_rx_fifo, sizeof (session_dgram_hdr_t), (u8 *) hdr); enqueued = svm_fifo_enqueue_nowait (s->server_rx_fifo, b->current_length, vlib_buffer_get_current (b)); if (PREDICT_FALSE ((b->flags & VLIB_BUFFER_NEXT_PRESENT) && enqueued >= 0)) { in_order_off = enqueued > b->current_length ? enqueued : 0; rv = session_enqueue_chain_tail (s, b, in_order_off, 1); if (rv > 0) enqueued += rv; } if (queue_event) { /* Queue RX event on this fifo. Eventually these will need to be flushed * by calling stream_server_flush_enqueue_events () */ session_manager_worker_t *wrk; wrk = session_manager_get_worker (s->thread_index); if (s->enqueue_epoch != wrk->current_enqueue_epoch[proto]) { s->enqueue_epoch = wrk->current_enqueue_epoch[proto]; vec_add1 (wrk->session_to_enqueue[proto], s->session_index); } } return enqueued; } /** Check if we have space in rx fifo to push more bytes */ u8 stream_session_no_space (transport_connection_t * tc, u32 thread_index, u16 data_len) { stream_session_t *s = session_get (tc->s_index, thread_index); if (PREDICT_FALSE (s->session_state != SESSION_STATE_READY)) return 1; if (data_len > svm_fifo_max_enqueue (s->server_rx_fifo)) return 1; return 0; } u32 session_tx_fifo_max_dequeue (transport_connection_t * tc) { stream_session_t *s = session_get (tc->s_index, tc->thread_index); if (!s->server_tx_fifo) return 0; return svm_fifo_max_dequeue (s->server_tx_fifo); } int stream_session_peek_bytes (transport_connection_t * tc, u8 * buffer, u32 offset, u32 max_bytes) { stream_session_t *s = session_get (tc->s_index, tc->thread_index); return svm_fifo_peek (s->server_tx_fifo, offset, max_bytes, buffer); } u32 stream_session_dequeue_drop (transport_connection_t * tc, u32 max_bytes) { stream_session_t *s = session_get (tc->s_index, tc->thread_index); return svm_fifo_dequeue_drop (s->server_tx_fifo, max_bytes); } /** * Notify session peer that new data has been enqueued. * * @param s Stream session for which the event is to be generated. * @param lock Flag to indicate if call should lock message queue. * * @return 0 on success or negative number if failed to send notification. */ static inline int session_enqueue_notify (stream_session_t * s) { app_worker_t *app; app = app_worker_get_if_valid (s->app_wrk_index); if (PREDICT_FALSE (!app)) { SESSION_DBG ("invalid s->app_index = %d", s->app_wrk_index); return 0; } /* *INDENT-OFF* */ SESSION_EVT_DBG(SESSION_EVT_ENQ, s, ({ ed->data[0] = FIFO_EVENT_APP_RX; ed->data[1] = svm_fifo_max_dequeue (s->server_rx_fifo); })); /* *INDENT-ON* */ return app_worker_lock_and_send_event (app, s, FIFO_EVENT_APP_RX); } int session_dequeue_notify (stream_session_t * s) { app_worker_t *app; app = app_worker_get_if_valid (s->app_wrk_index); if (PREDICT_FALSE (!app)) return -1; return app_worker_lock_and_send_event (app, s, FIFO_EVENT_APP_TX); } /** * Flushes queue of sessions that are to be notified of new data * enqueued events. * * @param thread_index Thread index for which the flush is to be performed. * @return 0 on success or a positive number indicating the number of * failures due to API queue being full. */ int session_manager_flush_enqueue_events (u8 transport_proto, u32 thread_index) { session_manager_worker_t *wrk = session_manager_get_worker (thread_index); stream_session_t *s; int i, errors = 0; u32 *indices; indices = wrk->session_to_enqueue[transport_proto]; for (i = 0; i < vec_len (indices); i++) { s = session_get_if_valid (indices[i], thread_index); if (PREDICT_FALSE (!s)) { errors++; continue; } if (PREDICT_FALSE (session_enqueue_notify (s))) errors++; } vec_reset_length (indices); wrk->session_to_enqueue[transport_proto] = indices; wrk->current_enqueue_epoch[transport_proto]++; return errors; } int session_manager_flush_all_enqueue_events (u8 transport_proto) { vlib_thread_main_t *vtm = vlib_get_thread_main (); int i, errors = 0; for (i = 0; i < 1 + vtm->n_threads; i++) errors += session_manager_flush_enqueue_events (transport_proto, i); return errors; } /** * Init fifo tail and head pointers * * Useful if transport uses absolute offsets for tracking ooo segments. */ void stream_session_init_fifos_pointers (transport_connection_t * tc, u32 rx_pointer, u32 tx_pointer) { stream_session_t *s; s = session_get (tc->s_index, tc->thread_index); svm_fifo_init_pointers (s->server_rx_fifo, rx_pointer); svm_fifo_init_pointers (s->server_tx_fifo, tx_pointer); } int session_stream_connect_notify (transport_connection_t * tc, u8 is_fail) { u32 opaque = 0, new_ti, new_si; stream_session_t *new_s = 0; segment_manager_t *sm; app_worker_t *app_wrk; application_t *app; u8 alloc_fifos; int error = 0; u64 handle; /* * Find connection handle and cleanup half-open table */ handle = session_lookup_half_open_handle (tc); if (handle == HALF_OPEN_LOOKUP_INVALID_VALUE) { SESSION_DBG ("half-open was removed!"); return -1; } session_lookup_del_half_open (tc); /* Get the app's index from the handle we stored when opening connection * and the opaque (api_context for external apps) from transport session * index */ app_wrk = app_worker_get_if_valid (handle >> 32); if (!app_wrk) return -1; opaque = tc->s_index; app = application_get (app_wrk->app_index); /* * Allocate new session with fifos (svm segments are allocated if needed) */ if (!is_fail) { sm = app_worker_get_connect_segment_manager (app_wrk); alloc_fifos = !application_is_builtin_proxy (app); if (session_alloc_and_init (sm, tc, alloc_fifos, &new_s)) { is_fail = 1; error = -1; } else { new_s->session_state = SESSION_STATE_CONNECTING; new_s->app_wrk_index = app_wrk->wrk_index; new_si = new_s->session_index; new_ti = new_s->thread_index; } } /* * Notify client application */ if (app->cb_fns.session_connected_callback (app_wrk->wrk_index, opaque, new_s, is_fail)) { SESSION_DBG ("failed to notify app"); if (!is_fail) { new_s = session_get (new_si, new_ti); stream_session_disconnect_transport (new_s); } } else { if (!is_fail) { new_s = session_get (new_si, new_ti); new_s->session_state = SESSION_STATE_READY; } } return error; } typedef struct _session_switch_pool_args { u32 session_index; u32 thread_index; u32 new_thread_index; u32 new_session_index; } session_switch_pool_args_t; static void session_switch_pool (void *cb_args) { session_switch_pool_args_t *args = (session_switch_pool_args_t *) cb_args; transport_proto_t tp; stream_session_t *s; ASSERT (args->thread_index == vlib_get_thread_index ()); s = session_get (args->session_index, args->thread_index); s->server_tx_fifo->master_session_index = args->new_session_index; s->server_tx_fifo->master_thread_index = args->new_thread_index; tp = session_get_transport_proto (s); tp_vfts[tp].cleanup (s->connection_index, s->thread_index); session_free (s); clib_mem_free (cb_args); } /** * Move dgram session to the right thread */ int session_dgram_connect_notify (transport_connection_t * tc, u32 old_thread_index, stream_session_t ** new_session) { stream_session_t *new_s; session_switch_pool_args_t *rpc_args; /* * Clone half-open session to the right thread. */ new_s = session_clone_safe (tc->s_index, old_thread_index); new_s->connection_index = tc->c_index; new_s->server_rx_fifo->master_session_index = new_s->session_index; new_s->server_rx_fifo->master_thread_index = new_s->thread_index; new_s->session_state = SESSION_STATE_READY; session_lookup_add_connection (tc, session_handle (new_s)); /* * Ask thread owning the old session to clean it up and make us the tx * fifo owner */ rpc_args = clib_mem_alloc (sizeof (*rpc_args)); rpc_args->new_session_index = new_s->session_index; rpc_args->new_thread_index = new_s->thread_index; rpc_args->session_index = tc->s_index; rpc_args->thread_index = old_thread_index; session_send_rpc_evt_to_thread (rpc_args->thread_index, session_switch_pool, rpc_args); tc->s_index = new_s->session_index; new_s->connection_index = tc->c_index; *new_session = new_s; return 0; } int stream_session_accept_notify (transport_connection_t * tc) { app_worker_t *app_wrk; application_t *app; stream_session_t *s; s = session_get (tc->s_index, tc->thread_index); app_wrk = app_worker_get_if_valid (s->app_wrk_index); if (!app_wrk) return -1; app = application_get (app_wrk->app_index); return app->cb_fns.session_accept_callback (s); } /** * Notification from transport that connection is being closed. * * A disconnect is sent to application but state is not removed. Once * disconnect is acknowledged by application, session disconnect is called. * Ultimately this leads to close being called on transport (passive close). */ void stream_session_disconnect_notify (transport_connection_t * tc) { app_worker_t *app_wrk; application_t *app; stream_session_t *s; s = session_get (tc->s_index, tc->thread_index); if (s->session_state >= SESSION_STATE_TRANSPORT_CLOSING) return; s->session_state = SESSION_STATE_TRANSPORT_CLOSING; app_wrk = app_worker_get_if_valid (s->app_wrk_index); if (!app_wrk) return; app = application_get (app_wrk->app_index); app->cb_fns.session_disconnect_callback (s); } /** * Cleans up session and lookup table. * * Transport connection must still be valid. */ void stream_session_delete (stream_session_t * s) { int rv; /* Delete from the main lookup table. */ if ((rv = session_lookup_del_session (s))) clib_warning ("hash delete error, rv %d", rv); session_free_w_fifos (s); } /** * Notification from transport that connection is being deleted * * This removes the session if it is still valid. It should be called only on * previously fully established sessions. For instance failed connects should * call stream_session_connect_notify and indicate that the connect has * failed. */ void stream_session_delete_notify (transport_connection_t * tc) { stream_session_t *s; /* App might've been removed already */ if (!(s = session_get_if_valid (tc->s_index, tc->thread_index))) return; /* Make sure we don't try to send anything more */ svm_fifo_dequeue_drop_all (s->server_tx_fifo); switch (s->session_state) { case SESSION_STATE_TRANSPORT_CLOSING: /* If transport finishes or times out before we get a reply * from the app, do the whole disconnect since we might still * have lingering events. Cleanup session table in advance * because transport will soon be closed and closed sessions * are assumed to have been removed from the lookup table */ session_lookup_del_session (s); stream_session_disconnect (s); s->session_state = SESSION_STATE_CLOSED; break; case SESSION_STATE_CLOSING: /* Cleanup lookup table. Transport needs to still be valid */ session_lookup_del_session (s); s->session_state = SESSION_STATE_CLOSED; break; case SESSION_STATE_CLOSED: case SESSION_STATE_ACCEPTING: case SESSION_STATE_CLOSED_WAITING: stream_session_delete (s); break; default: stream_session_delete (s); break; } } /** * Notification from transport that session can be closed * * Should be called by transport only if it was closed with non-empty * tx fifo and once it decides to begin the closing procedure prior to * issuing a delete notify. This gives the chance to the session layer * to cleanup any outstanding events. */ void session_stream_close_notify (transport_connection_t * tc) { stream_session_t *s; if (!(s = session_get_if_valid (tc->s_index, tc->thread_index))) return; s->session_state = SESSION_STATE_CLOSED; } /** * Notify application that connection has been reset. */ void stream_session_reset_notify (transport_connection_t * tc) { stream_session_t *s; app_worker_t *app_wrk; application_t *app; s = session_get (tc->s_index, tc->thread_index); svm_fifo_dequeue_drop_all (s->server_tx_fifo); if (s->session_state >= SESSION_STATE_TRANSPORT_CLOSING) return; s->session_state = SESSION_STATE_TRANSPORT_CLOSING; app_wrk = app_worker_get (s->app_wrk_index); app = application_get (app_wrk->app_index); app->cb_fns.session_reset_callback (s); } /** * Accept a stream session. Optionally ping the server by callback. */ int stream_session_accept (transport_connection_t * tc, u32 listener_index, u8 notify) { stream_session_t *s, *listener; app_worker_t *app_wrk; segment_manager_t *sm; int rv; /* Find the server */ listener = listen_session_get (listener_index); app_wrk = application_listener_select_worker (listener, 0); sm = app_worker_get_listen_segment_manager (app_wrk, listener); if ((rv = session_alloc_and_init (sm, tc, 1, &s))) return rv; s->app_wrk_index = app_wrk->wrk_index; s->listener_index = listener_index; s->session_state = SESSION_STATE_ACCEPTING; /* Shoulder-tap the server */ if (notify) { application_t *app = application_get (app_wrk->app_index); return app->cb_fns.session_accept_callback (s); } return 0; } int session_open_cl (u32 app_wrk_index, session_endpoint_t * rmt, u32 opaque) { transport_connection_t *tc; transport_endpoint_cfg_t *tep; segment_manager_t *sm; app_worker_t *app_wrk; stream_session_t *s; application_t *app; int rv; tep = session_endpoint_to_transport_cfg (rmt); rv = tp_vfts[rmt->transport_proto].open (tep); if (rv < 0) { SESSION_DBG ("Transport failed to open connection."); return VNET_API_ERROR_SESSION_CONNECT; } tc = tp_vfts[rmt->transport_proto].get_half_open ((u32) rv); /* For dgram type of service, allocate session and fifos now. */ app_wrk = app_worker_get (app_wrk_index); sm = app_worker_get_connect_segment_manager (app_wrk); if (session_alloc_and_init (sm, tc, 1, &s)) return -1; s->app_wrk_index = app_wrk->wrk_index; s->session_state = SESSION_STATE_OPENED; /* Tell the app about the new event fifo for this session */ app = application_get (app_wrk->app_index); app->cb_fns.session_connected_callback (app_wrk->wrk_index, opaque, s, 0); return 0; } int session_open_vc (u32 app_wrk_index, session_endpoint_t * rmt, u32 opaque) { transport_connection_t *tc; transport_endpoint_cfg_t *tep; u64 handle; int rv; tep = session_endpoint_to_transport_cfg (rmt); rv = tp_vfts[rmt->transport_proto].open (tep); if (rv < 0) { SESSION_DBG ("Transport failed to open connection."); return VNET_API_ERROR_SESSION_CONNECT; } tc = tp_vfts[rmt->transport_proto].get_half_open ((u32) rv); /* If transport offers a stream service, only allocate session once the * connection has been established. * Add connection to half-open table and save app and tc index. The * latter is needed to help establish the connection while the former * is needed when the connect notify comes and we have to notify the * external app */ handle = (((u64) app_wrk_index) << 32) | (u64) tc->c_index; session_lookup_add_half_open (tc, handle); /* Store api_context (opaque) for when the reply comes. Not the nicest * thing but better than allocating a separate half-open pool. */ tc->s_index = opaque; return 0; } int session_open_app (u32 app_wrk_index, session_endpoint_t * rmt, u32 opaque) { session_endpoint_cfg_t *sep = (session_endpoint_cfg_t *) rmt; transport_endpoint_cfg_t *tep_cfg = session_endpoint_to_transport_cfg (sep); sep->app_wrk_index = app_wrk_index; sep->opaque = opaque; return tp_vfts[rmt->transport_proto].open (tep_cfg); } typedef int (*session_open_service_fn) (u32, session_endpoint_t *, u32); /* *INDENT-OFF* */ static session_open_service_fn session_open_srv_fns[TRANSPORT_N_SERVICES] = { session_open_vc, session_open_cl, session_open_app, }; /* *INDENT-ON* */ /** * Ask transport to open connection to remote transport endpoint. * * Stores handle for matching request with reply since the call can be * asynchronous. For instance, for TCP the 3-way handshake must complete * before reply comes. Session is only created once connection is established. * * @param app_index Index of the application requesting the connect * @param st Session type requested. * @param tep Remote transport endpoint * @param opaque Opaque data (typically, api_context) the application expects * on open completion. */ int session_open (u32 app_wrk_index, session_endpoint_t * rmt, u32 opaque) { transport_service_type_t tst = tp_vfts[rmt->transport_proto].service_type; return session_open_srv_fns[tst] (app_wrk_index, rmt, opaque); } /** * Ask transport to listen on session endpoint. * * @param s Session for which listen will be called. Note that unlike * established sessions, listen sessions are not associated to a * thread. * @param sep Local endpoint to be listened on. */ int session_listen (stream_session_t * ls, session_endpoint_cfg_t * sep) { transport_connection_t *tc; transport_endpoint_t *tep; u32 tc_index, s_index; /* Transport bind/listen */ tep = session_endpoint_to_transport (sep); s_index = ls->session_index; tc_index = tp_vfts[sep->transport_proto].bind (s_index, tep); if (tc_index == (u32) ~ 0) return -1; /* Attach transport to session */ ls = listen_session_get (s_index); ls->connection_index = tc_index; /* Add to the main lookup table after transport was initialized */ tc = tp_vfts[sep->transport_proto].get_listener (tc_index); session_lookup_add_connection (tc, s_index); return 0; } /** * Ask transport to stop listening on local transport endpoint. * * @param s Session to stop listening on. It must be in state LISTENING. */ int session_stop_listen (stream_session_t * s) { transport_proto_t tp = session_get_transport_proto (s); transport_connection_t *tc; if (s->session_state != SESSION_STATE_LISTENING) { clib_warning ("not a listening session"); return -1; } tc = tp_vfts[tp].get_listener (s->connection_index); if (!tc) { clib_warning ("no transport"); return VNET_API_ERROR_ADDRESS_NOT_IN_USE; } session_lookup_del_connection (tc); tp_vfts[tp].unbind (s->connection_index); return 0; } /** * Initialize session disconnect. * * Request is always sent to session node to ensure that all outstanding * requests are served before transport is notified. */ void stream_session_disconnect (stream_session_t * s) { u32 thread_index = vlib_get_thread_index (); session_manager_worker_t *wrk; session_event_t *evt; if (!s) return; if (s->session_state >= SESSION_STATE_CLOSING) { /* Session already closed. Clear the tx fifo */ if (s->session_state == SESSION_STATE_CLOSED) svm_fifo_dequeue_drop_all (s->server_tx_fifo); return; } s->session_state = SESSION_STATE_CLOSING; /* If we are in the handler thread, or being called with the worker barrier * held, just append a new event to pending disconnects vector. */ if (vlib_thread_is_main_w_barrier () || thread_index == s->thread_index) { wrk = session_manager_get_worker (s->thread_index); vec_add2 (wrk->pending_disconnects, evt, 1); clib_memset (evt, 0, sizeof (*evt)); evt->session_handle = session_handle (s); evt->event_type = FIFO_EVENT_DISCONNECT; } else session_send_ctrl_evt_to_thread (s, FIFO_EVENT_DISCONNECT); } /** * Notify transport the session can be disconnected. This should eventually * result in a delete notification that allows us to cleanup session state. * Called for both active/passive disconnects. * * Must be called from the session's thread. */ void stream_session_disconnect_transport (stream_session_t * s) { /* If transport is already closed, just free the session */ if (s->session_state == SESSION_STATE_CLOSED) { session_free_w_fifos (s); return; } /* If tx queue wasn't drained, change state to closed waiting for transport. * This way, the transport, if it so wishes, can continue to try sending the * outstanding data (in closed state it cannot). It MUST however at one * point, either after sending everything or after a timeout, call delete * notify. This will finally lead to the complete cleanup of the session. */ if (svm_fifo_max_dequeue (s->server_tx_fifo)) s->session_state = SESSION_STATE_CLOSED_WAITING; else s->session_state = SESSION_STATE_CLOSED; tp_vfts[session_get_transport_proto (s)].close (s->connection_index, s->thread_index); } /** * Cleanup transport and session state. * * Notify transport of the cleanup and free the session. This should * be called only if transport reported some error and is already * closed. */ void stream_session_cleanup (stream_session_t * s) { s->session_state = SESSION_STATE_CLOSED; /* Delete from main lookup table before we axe the the transport */ session_lookup_del_session (s); tp_vfts[session_get_transport_proto (s)].cleanup (s->connection_index, s->thread_index); /* Since we called cleanup, no delete notification will come. So, make * sure the session is properly freed. */ session_free_w_fifos (s); } transport_service_type_t session_transport_service_type (stream_session_t * s) { transport_proto_t tp; tp = session_get_transport_proto (s); return transport_protocol_service_type (tp); } transport_tx_fn_type_t session_transport_tx_fn_type (stream_session_t * s) { transport_proto_t tp; tp = session_get_transport_proto (s); return transport_protocol_tx_fn_type (tp); } u8 session_tx_is_dgram (stream_session_t * s) { return (session_transport_tx_fn_type (s) == TRANSPORT_TX_DGRAM); } /** * Allocate event queues in the shared-memory segment * * That can either be a newly created memfd segment, that will need to be * mapped by all stack users, or the binary api's svm region. The latter is * assumed to be already mapped. NOTE that this assumption DOES NOT hold if * api clients bootstrap shm api over sockets (i.e. use memfd segments) and * vpp uses api svm region for event queues. */ void session_vpp_event_queues_allocate (session_manager_main_t * smm) { u32 evt_q_length = 2048, evt_size = sizeof (session_event_t); ssvm_private_t *eqs = &smm->evt_qs_segment; api_main_t *am = &api_main; u64 eqs_size = 64 << 20; pid_t vpp_pid = getpid (); void *oldheap; int i; if (smm->configured_event_queue_length) evt_q_length = smm->configured_event_queue_length; if (smm->evt_qs_use_memfd_seg) { if (smm->evt_qs_segment_size) eqs_size = smm->evt_qs_segment_size; eqs->ssvm_size = eqs_size; eqs->i_am_master = 1; eqs->my_pid = vpp_pid; eqs->name = format (0, "%s%c", "evt-qs-segment", 0); eqs->requested_va = smm->session_baseva; if (ssvm_master_init (eqs, SSVM_SEGMENT_MEMFD)) { clib_warning ("failed to initialize queue segment"); return; } } if (smm->evt_qs_use_memfd_seg) oldheap = ssvm_push_heap (eqs->sh); else oldheap = svm_push_data_heap (am->vlib_rp); for (i = 0; i < vec_len (smm->wrk); i++) { svm_msg_q_cfg_t _cfg, *cfg = &_cfg; svm_msg_q_ring_cfg_t rc[SESSION_MQ_N_RINGS] = { {evt_q_length, evt_size, 0} , {evt_q_length << 1, 256, 0} }; cfg->consumer_pid = 0; cfg->n_rings = 2; cfg->q_nitems = evt_q_length; cfg->ring_cfgs = rc; smm->wrk[i].vpp_event_queue = svm_msg_q_alloc (cfg); if (smm->evt_qs_use_memfd_seg) { if (svm_msg_q_alloc_consumer_eventfd (smm->wrk[i].vpp_event_queue)) clib_warning ("eventfd returned"); } } if (smm->evt_qs_use_memfd_seg) ssvm_pop_heap (oldheap); else svm_pop_heap (oldheap); } ssvm_private_t * session_manager_get_evt_q_segment (void) { session_manager_main_t *smm = &session_manager_main; if (smm->evt_qs_use_memfd_seg) return &smm->evt_qs_segment; return 0; } /* *INDENT-OFF* */ static session_fifo_rx_fn *session_tx_fns[TRANSPORT_TX_N_FNS] = { session_tx_fifo_peek_and_snd, session_tx_fifo_dequeue_and_snd, session_tx_fifo_dequeue_internal, session_tx_fifo_dequeue_and_snd }; /* *INDENT-ON* */ /** * Initialize session layer for given transport proto and ip version * * Allocates per session type (transport proto + ip version) data structures * and adds arc from session queue node to session type output node. */ void session_register_transport (transport_proto_t transport_proto, const transport_proto_vft_t * vft, u8 is_ip4, u32 output_node) { session_manager_main_t *smm = &session_manager_main; session_type_t session_type; u32 next_index = ~0; session_type = session_type_from_proto_and_ip (transport_proto, is_ip4); vec_validate (smm->session_type_to_next, session_type); vec_validate (smm->session_tx_fns, session_type); /* *INDENT-OFF* */ if (output_node != ~0) { foreach_vlib_main (({ next_index = vlib_node_add_next (this_vlib_main, session_queue_node.index, output_node); })); } /* *INDENT-ON* */ smm->session_type_to_next[session_type] = next_index; smm->session_tx_fns[session_type] = session_tx_fns[vft->tx_type]; } transport_connection_t * session_get_transport (stream_session_t * s) { transport_proto_t tp; if (s->session_state != SESSION_STATE_LISTENING) { tp = session_get_transport_proto (s); return tp_vfts[tp].get_connection (s->connection_index, s->thread_index); } return 0; } transport_connection_t * listen_session_get_transport (stream_session_t * s) { transport_proto_t tp = session_get_transport_proto (s); return tp_vfts[tp].get_listener (s->connection_index); } int listen_session_get_local_session_endpoint (stream_session_t * listener, session_endpoint_t * sep) { transport_proto_t tp = session_get_transport_proto (listener); transport_connection_t *tc; tc = tp_vfts[tp].get_listener (listener->connection_index); if (!tc) { clib_warning ("no transport"); return -1; } /* N.B. The ip should not be copied because this is the local endpoint */ sep->port = tc->lcl_port; sep->transport_proto = tc->proto; sep->is_ip4 = tc->is_ip4; return 0; } void session_flush_frames_main_thread (vlib_main_t * vm) { ASSERT (vlib_get_thread_index () == 0); vlib_process_signal_event_mt (vm, session_queue_process_node.index, SESSION_Q_PROCESS_FLUSH_FRAMES, 0); } static clib_error_t * session_manager_main_enable (vlib_main_t * vm) { segment_manager_main_init_args_t _sm_args = { 0 }, *sm_args = &_sm_args; session_manager_main_t *smm = &session_manager_main; vlib_thread_main_t *vtm = vlib_get_thread_main (); u32 num_threads, preallocated_sessions_per_worker; session_manager_worker_t *wrk; int i, j; num_threads = 1 /* main thread */ + vtm->n_threads; if (num_threads < 1) return clib_error_return (0, "n_thread_stacks not set"); /* Allocate cache line aligned worker contexts */ vec_validate_aligned (smm->wrk, num_threads - 1, CLIB_CACHE_LINE_BYTES); for (i = 0; i < TRANSPORT_N_PROTO; i++) { for (j = 0; j < num_threads; j++) smm->wrk[j].current_enqueue_epoch[i] = 1; } for (i = 0; i < num_threads; i++) { wrk = &smm->wrk[i]; vec_validate (wrk->free_event_vector, 128); _vec_len (wrk->free_event_vector) = 0; vec_validate (wrk->pending_event_vector, 128); _vec_len (wrk->pending_event_vector) = 0; vec_validate (wrk->pending_disconnects, 128); _vec_len (wrk->pending_disconnects) = 0; vec_validate (wrk->postponed_event_vector, 128); _vec_len (wrk->postponed_event_vector) = 0; wrk->last_vlib_time = vlib_time_now (vlib_mains[i]); wrk->dispatch_period = 500e-6; if (num_threads > 1) clib_rwlock_init (&smm->wrk[i].peekers_rw_locks); } #if SESSION_DEBUG vec_validate (smm->last_event_poll_by_thread, num_threads - 1); #endif /* Allocate vpp event queues segment and queue */ session_vpp_event_queues_allocate (smm); /* Initialize fifo segment main baseva and timeout */ sm_args->baseva = smm->session_baseva + smm->evt_qs_segment_size; sm_args->size = smm->session_va_space_size; segment_manager_main_init (sm_args); /* Preallocate sessions */ if (smm->preallocated_sessions) { if (num_threads == 1) { pool_init_fixed (smm->wrk[0].sessions, smm->preallocated_sessions); } else { int j; preallocated_sessions_per_worker = (1.1 * (f64) smm->preallocated_sessions / (f64) (num_threads - 1)); for (j = 1; j < num_threads; j++) { pool_init_fixed (smm->wrk[j].sessions, preallocated_sessions_per_worker); } } } session_lookup_init (); app_namespaces_init (); transport_init (); smm->is_enabled = 1; /* Enable transports */ transport_enable_disable (vm, 1); transport_init_tx_pacers_period (); return 0; } void session_node_enable_disable (u8 is_en) { u8 state = is_en ? VLIB_NODE_STATE_POLLING : VLIB_NODE_STATE_DISABLED; vlib_thread_main_t *vtm = vlib_get_thread_main (); u8 have_workers = vtm->n_threads != 0; /* *INDENT-OFF* */ foreach_vlib_main (({ if (have_workers && ii == 0) { vlib_node_set_state (this_vlib_main, session_queue_process_node.index, state); if (is_en) { vlib_node_t *n = vlib_get_node (this_vlib_main, session_queue_process_node.index); vlib_start_process (this_vlib_main, n->runtime_index); } else { vlib_process_signal_event_mt (this_vlib_main, session_queue_process_node.index, SESSION_Q_PROCESS_STOP, 0); } continue; } vlib_node_set_state (this_vlib_main, session_queue_node.index, state); })); /* *INDENT-ON* */ } clib_error_t * vnet_session_enable_disable (vlib_main_t * vm, u8 is_en) { clib_error_t *error = 0; if (is_en) { if (session_manager_main.is_enabled) return 0; session_node_enable_disable (is_en); error = session_manager_main_enable (vm); } else { session_manager_main.is_enabled = 0; session_node_enable_disable (is_en); } return error; } clib_error_t * session_manager_main_init (vlib_main_t * vm) { session_manager_main_t *smm = &session_manager_main; smm->session_baseva = 0x200000000ULL; smm->session_va_space_size = (u64) 128 << 30; smm->evt_qs_segment_size = 64 << 20; smm->is_enabled = 0; return 0; } VLIB_INIT_FUNCTION (session_manager_main_init); static clib_error_t * session_config_fn (vlib_main_t * vm, unformat_input_t * input) { session_manager_main_t *smm = &session_manager_main; u32 nitems; uword tmp; while (unformat_check_input (input) != UNFORMAT_END_OF_INPUT) { if (unformat (input, "event-queue-length %d", &nitems)) { if (nitems >= 2048) smm->configured_event_queue_length = nitems; else clib_warning ("event queue length %d too small, ignored", nitems); } else if (unformat (input, "preallocated-sessions %d", &smm->preallocated_sessions)) ; else if (unformat (input, "v4-session-table-buckets %d", &smm->configured_v4_session_table_buckets)) ; else if (unformat (input, "v4-halfopen-table-buckets %d", &smm->configured_v4_halfopen_table_buckets)) ; else if (unformat (input, "v6-session-table-buckets %d", &smm->configured_v6_session_table_buckets)) ; else if (unformat (input, "v6-halfopen-table-buckets %d", &smm->configured_v6_halfopen_table_buckets)) ; else if (unformat (input, "v4-session-table-memory %U", unformat_memory_size, &tmp)) { if (tmp >= 0x100000000) return clib_error_return (0, "memory size %llx (%lld) too large", tmp, tmp); smm->configured_v4_session_table_memory = tmp; } else if (unformat (input, "v4-halfopen-table-memory %U", unformat_memory_size, &tmp)) { if (tmp >= 0x100000000) return clib_error_return (0, "memory size %llx (%lld) too large", tmp, tmp); smm->configured_v4_halfopen_table_memory = tmp; } else if (unformat (input, "v6-session-table-memory %U", unformat_memory_size, &tmp)) { if (tmp >= 0x100000000) return clib_error_return (0, "memory size %llx (%lld) too large", tmp, tmp); smm->configured_v6_session_table_memory = tmp; } else if (unformat (input, "v6-halfopen-table-memory %U", unformat_memory_size, &tmp)) { if (tmp >= 0x100000000) return clib_error_return (0, "memory size %llx (%lld) too large", tmp, tmp); smm->configured_v6_halfopen_table_memory = tmp; } else if (unformat (input, "local-endpoints-table-memory %U", unformat_memory_size, &tmp)) { if (tmp >= 0x100000000) return clib_error_return (0, "memory size %llx (%lld) too large", tmp, tmp); smm->local_endpoints_table_memory = tmp; } else if (unformat (input, "local-endpoints-table-buckets %d", &smm->local_endpoints_table_buckets)) ; else if (unformat (input, "evt_qs_memfd_seg")) smm->evt_qs_use_memfd_seg = 1; else return clib_error_return (0, "unknown input `%U'", format_unformat_error, input); } return 0; } VLIB_CONFIG_FUNCTION (session_config_fn, "session"); /* * fd.io coding-style-patch-verification: ON * * Local Variables: * eval: (c-set-style "gnu") * End: */