summaryrefslogtreecommitdiff
path: root/lib/netdev-offload.c
blob: 32eab5910760f1123d8809c0ac3d370f4499c436 (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
/*
 * Copyright (c) 2008 - 2014, 2016, 2017 Nicira, Inc.
 * Copyright (c) 2019 Samsung Electronics Co.,Ltd.
 *
 * Licensed under the Apache License, Version 2.0 (the "License");
 * you may not use this file except in compliance with the License.
 * You may obtain a copy of the License at:
 *
 *     http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */

#include <config.h>
#include "netdev-offload.h"

#include <errno.h>
#include <inttypes.h>
#include <sys/types.h>
#include <netinet/in.h>
#include <stdlib.h>
#include <string.h>
#include <unistd.h>

#include "cmap.h"
#include "coverage.h"
#include "dpif.h"
#include "dp-packet.h"
#include "openvswitch/dynamic-string.h"
#include "fatal-signal.h"
#include "hash.h"
#include "openvswitch/list.h"
#include "netdev-offload-provider.h"
#include "netdev-provider.h"
#include "netdev-vport.h"
#include "odp-netlink.h"
#include "openflow/openflow.h"
#include "packets.h"
#include "openvswitch/ofp-print.h"
#include "openvswitch/poll-loop.h"
#include "seq.h"
#include "openvswitch/shash.h"
#include "smap.h"
#include "socket-util.h"
#include "sset.h"
#include "svec.h"
#include "openvswitch/vlog.h"
#include "flow.h"
#include "util.h"
#ifdef __linux__
#include "tc.h"
#endif

VLOG_DEFINE_THIS_MODULE(netdev_offload);


static bool netdev_flow_api_enabled = false;

/* Protects 'netdev_flow_apis'.  */
static struct ovs_mutex netdev_flow_api_provider_mutex = OVS_MUTEX_INITIALIZER;

/* Contains 'struct netdev_registered_flow_api's. */
static struct cmap netdev_flow_apis = CMAP_INITIALIZER;

struct netdev_registered_flow_api {
    struct cmap_node cmap_node; /* In 'netdev_flow_apis', by flow_api->type. */
    const struct netdev_flow_api *flow_api;

    /* Number of references: one for the flow_api itself and one for every
     * instance of the netdev that uses it. */
    struct ovs_refcount refcnt;
};

static struct netdev_registered_flow_api *
netdev_lookup_flow_api(const char *type)
{
    struct netdev_registered_flow_api *rfa;
    CMAP_FOR_EACH_WITH_HASH (rfa, cmap_node, hash_string(type, 0),
                             &netdev_flow_apis) {
        if (!strcmp(type, rfa->flow_api->type)) {
            return rfa;
        }
    }
    return NULL;
}

/* Registers a new netdev flow api provider. */
int
netdev_register_flow_api_provider(const struct netdev_flow_api *new_flow_api)
    OVS_EXCLUDED(netdev_flow_api_provider_mutex)
{
    int error = 0;

    if (!new_flow_api->init_flow_api) {
        VLOG_WARN("attempted to register invalid flow api provider: %s",
                   new_flow_api->type);
        error = EINVAL;
    }

    ovs_mutex_lock(&netdev_flow_api_provider_mutex);
    if (netdev_lookup_flow_api(new_flow_api->type)) {
        VLOG_WARN("attempted to register duplicate flow api provider: %s",
                   new_flow_api->type);
        error = EEXIST;
    } else {
        struct netdev_registered_flow_api *rfa;

        rfa = xmalloc(sizeof *rfa);
        cmap_insert(&netdev_flow_apis, &rfa->cmap_node,
                    hash_string(new_flow_api->type, 0));
        rfa->flow_api = new_flow_api;
        ovs_refcount_init(&rfa->refcnt);
        VLOG_DBG("netdev: flow API '%s' registered.", new_flow_api->type);
    }
    ovs_mutex_unlock(&netdev_flow_api_provider_mutex);

    return error;
}

/* Unregisters a netdev flow api provider.  'type' must have been previously
 * registered and not currently be in use by any netdevs.  After unregistration
 * netdev flow api of that type cannot be used for netdevs.  (However, the
 * provider may still be accessible from other threads until the next RCU grace
 * period, so the caller must not free or re-register the same netdev_flow_api
 * until that has passed.) */
int
netdev_unregister_flow_api_provider(const char *type)
    OVS_EXCLUDED(netdev_flow_api_provider_mutex)
{
    struct netdev_registered_flow_api *rfa;
    int error;

    ovs_mutex_lock(&netdev_flow_api_provider_mutex);
    rfa = netdev_lookup_flow_api(type);
    if (!rfa) {
        VLOG_WARN("attempted to unregister a flow api provider that is not "
                  "registered: %s", type);
        error = EAFNOSUPPORT;
    } else if (ovs_refcount_unref(&rfa->refcnt) != 1) {
        ovs_refcount_ref(&rfa->refcnt);
        VLOG_WARN("attempted to unregister in use flow api provider: %s",
                  type);
        error = EBUSY;
    } else  {
        cmap_remove(&netdev_flow_apis, &rfa->cmap_node,
                    hash_string(rfa->flow_api->type, 0));
        ovsrcu_postpone(free, rfa);
        error = 0;
    }
    ovs_mutex_unlock(&netdev_flow_api_provider_mutex);

    return error;
}

bool
netdev_flow_api_equals(const struct netdev *netdev1,
                       const struct netdev *netdev2)
{
    const struct netdev_flow_api *netdev_flow_api1 =
        ovsrcu_get(const struct netdev_flow_api *, &netdev1->flow_api);
    const struct netdev_flow_api *netdev_flow_api2 =
        ovsrcu_get(const struct netdev_flow_api *, &netdev2->flow_api);

    return netdev_flow_api1 == netdev_flow_api2;
}

static int
netdev_assign_flow_api(struct netdev *netdev)
{
    struct netdev_registered_flow_api *rfa;

    CMAP_FOR_EACH (rfa, cmap_node, &netdev_flow_apis) {
        if (!rfa->flow_api->init_flow_api(netdev)) {
            ovs_refcount_ref(&rfa->refcnt);
            ovsrcu_set(&netdev->flow_api, rfa->flow_api);
            VLOG_INFO("%s: Assigned flow API '%s'.",
                      netdev_get_name(netdev), rfa->flow_api->type);
            return 0;
        }
        VLOG_DBG("%s: flow API '%s' is not suitable.",
                 netdev_get_name(netdev), rfa->flow_api->type);
    }
    VLOG_INFO("%s: No suitable flow API found.", netdev_get_name(netdev));

    return -1;
}

int
netdev_flow_flush(struct netdev *netdev)
{
    const struct netdev_flow_api *flow_api =
        ovsrcu_get(const struct netdev_flow_api *, &netdev->flow_api);

    return (flow_api && flow_api->flow_flush)
           ? flow_api->flow_flush(netdev)
           : EOPNOTSUPP;
}

int
netdev_flow_dump_create(struct netdev *netdev, struct netdev_flow_dump **dump)
{
    const struct netdev_flow_api *flow_api =
        ovsrcu_get(const struct netdev_flow_api *, &netdev->flow_api);

    return (flow_api && flow_api->flow_dump_create)
           ? flow_api->flow_dump_create(netdev, dump)
           : EOPNOTSUPP;
}

int
netdev_flow_dump_destroy(struct netdev_flow_dump *dump)
{
    const struct netdev_flow_api *flow_api =
        ovsrcu_get(const struct netdev_flow_api *, &dump->netdev->flow_api);

    return (flow_api && flow_api->flow_dump_destroy)
           ? flow_api->flow_dump_destroy(dump)
           : EOPNOTSUPP;
}

bool
netdev_flow_dump_next(struct netdev_flow_dump *dump, struct match *match,
                      struct nlattr **actions, struct dpif_flow_stats *stats,
                      struct dpif_flow_attrs *attrs, ovs_u128 *ufid,
                      struct ofpbuf *rbuffer, struct ofpbuf *wbuffer)
{
    const struct netdev_flow_api *flow_api =
        ovsrcu_get(const struct netdev_flow_api *, &dump->netdev->flow_api);

    return (flow_api && flow_api->flow_dump_next)
           ? flow_api->flow_dump_next(dump, match, actions, stats, attrs,
                                      ufid, rbuffer, wbuffer)
           : false;
}

int
netdev_flow_put(struct netdev *netdev, struct match *match,
                struct nlattr *actions, size_t act_len,
                const ovs_u128 *ufid, struct offload_info *info,
                struct dpif_flow_stats *stats)
{
    const struct netdev_flow_api *flow_api =
        ovsrcu_get(const struct netdev_flow_api *, &netdev->flow_api);

    return (flow_api && flow_api->flow_put)
           ? flow_api->flow_put(netdev, match, actions, act_len, ufid,
                                info, stats)
           : EOPNOTSUPP;
}

int
netdev_flow_get(struct netdev *netdev, struct match *match,
                struct nlattr **actions, const ovs_u128 *ufid,
                struct dpif_flow_stats *stats,
                struct dpif_flow_attrs *attrs, struct ofpbuf *buf)
{
    const struct netdev_flow_api *flow_api =
        ovsrcu_get(const struct netdev_flow_api *, &netdev->flow_api);

    return (flow_api && flow_api->flow_get)
           ? flow_api->flow_get(netdev, match, actions, ufid,
                                stats, attrs, buf)
           : EOPNOTSUPP;
}

int
netdev_flow_del(struct netdev *netdev, const ovs_u128 *ufid,
                struct dpif_flow_stats *stats)
{
    const struct netdev_flow_api *flow_api =
        ovsrcu_get(const struct netdev_flow_api *, &netdev->flow_api);

    return (flow_api && flow_api->flow_del)
           ? flow_api->flow_del(netdev, ufid, stats)
           : EOPNOTSUPP;
}

int
netdev_init_flow_api(struct netdev *netdev)
{
    if (!netdev_is_flow_api_enabled()) {
        return EOPNOTSUPP;
    }

    if (ovsrcu_get(const struct netdev_flow_api *, &netdev->flow_api)) {
        return 0;
    }

    if (netdev_assign_flow_api(netdev)) {
        return EOPNOTSUPP;
    }

    return 0;
}

void
netdev_uninit_flow_api(struct netdev *netdev)
{
    struct netdev_registered_flow_api *rfa;
    const struct netdev_flow_api *flow_api =
            ovsrcu_get(const struct netdev_flow_api *, &netdev->flow_api);

    if (!flow_api) {
        return;
    }

    ovsrcu_set(&netdev->flow_api, NULL);
    rfa = netdev_lookup_flow_api(flow_api->type);
    ovs_refcount_unref(&rfa->refcnt);
}

uint32_t
netdev_get_block_id(struct netdev *netdev)
{
    const struct netdev_class *class = netdev->netdev_class;

    return (class->get_block_id
            ? class->get_block_id(netdev)
            : 0);
}

/*
 * Get the value of the hw info parameter specified by type.
 * Returns the value on success (>= 0). Returns -1 on failure.
 */
int
netdev_get_hw_info(struct netdev *netdev, int type)
{
    int val = -1;

    switch (type) {
    case HW_INFO_TYPE_OOR:
        val = netdev->hw_info.oor;
        break;
    case HW_INFO_TYPE_PEND_COUNT:
        val = netdev->hw_info.pending_count;
        break;
    case HW_INFO_TYPE_OFFL_COUNT:
        val = netdev->hw_info.offload_count;
        break;
    default:
        break;
    }

    return val;
}

/*
 * Set the value of the hw info parameter specified by type.
 */
void
netdev_set_hw_info(struct netdev *netdev, int type, int val)
{
    switch (type) {
    case HW_INFO_TYPE_OOR:
        if (val == 0) {
            VLOG_DBG("Offload rebalance: netdev: %s is not OOR", netdev->name);
        }
        netdev->hw_info.oor = val;
        break;
    case HW_INFO_TYPE_PEND_COUNT:
        netdev->hw_info.pending_count = val;
        break;
    case HW_INFO_TYPE_OFFL_COUNT:
        netdev->hw_info.offload_count = val;
        break;
    default:
        break;
    }
}

/* Protects below port hashmaps. */
static struct ovs_rwlock netdev_hmap_rwlock = OVS_RWLOCK_INITIALIZER;

static struct hmap port_to_netdev OVS_GUARDED_BY(netdev_hmap_rwlock)
    = HMAP_INITIALIZER(&port_to_netdev);
static struct hmap ifindex_to_port OVS_GUARDED_BY(netdev_hmap_rwlock)
    = HMAP_INITIALIZER(&ifindex_to_port);

struct port_to_netdev_data {
    struct hmap_node portno_node; /* By (dpif_class, dpif_port.port_no). */
    struct hmap_node ifindex_node; /* By (dpif_class, ifindex). */
    struct netdev *netdev;
    struct dpif_port dpif_port;
    const struct dpif_class *dpif_class;
    int ifindex;
};

/*
 * Find if any netdev is in OOR state. Return true if there's at least
 * one netdev that's in OOR state; otherwise return false.
 */
bool
netdev_any_oor(void)
    OVS_EXCLUDED(netdev_hmap_rwlock)
{
    struct port_to_netdev_data *data;
    bool oor = false;

    ovs_rwlock_rdlock(&netdev_hmap_rwlock);
    HMAP_FOR_EACH (data, portno_node, &port_to_netdev) {
        struct netdev *dev = data->netdev;

        if (dev->hw_info.oor) {
            oor = true;
            break;
        }
    }
    ovs_rwlock_unlock(&netdev_hmap_rwlock);

    return oor;
}

bool
netdev_is_flow_api_enabled(void)
{
    return netdev_flow_api_enabled;
}

void
netdev_ports_flow_flush(const struct dpif_class *dpif_class)
{
    struct port_to_netdev_data *data;

    ovs_rwlock_rdlock(&netdev_hmap_rwlock);
    HMAP_FOR_EACH (data, portno_node, &port_to_netdev) {
        if (data->dpif_class == dpif_class) {
            netdev_flow_flush(data->netdev);
        }
    }
    ovs_rwlock_unlock(&netdev_hmap_rwlock);
}

struct netdev_flow_dump **
netdev_ports_flow_dump_create(const struct dpif_class *dpif_class, int *ports)
{
    struct port_to_netdev_data *data;
    struct netdev_flow_dump **dumps;
    int count = 0;
    int i = 0;

    ovs_rwlock_rdlock(&netdev_hmap_rwlock);
    HMAP_FOR_EACH (data, portno_node, &port_to_netdev) {
        if (data->dpif_class == dpif_class) {
            count++;
        }
    }

    dumps = count ? xzalloc(sizeof *dumps * count) : NULL;

    HMAP_FOR_EACH (data, portno_node, &port_to_netdev) {
        if (data->dpif_class == dpif_class) {
            if (netdev_flow_dump_create(data->netdev, &dumps[i])) {
                continue;
            }

            dumps[i]->port = data->dpif_port.port_no;
            i++;
        }
    }
    ovs_rwlock_unlock(&netdev_hmap_rwlock);

    *ports = i;
    return dumps;
}

int
netdev_ports_flow_del(const struct dpif_class *dpif_class,
                      const ovs_u128 *ufid,
                      struct dpif_flow_stats *stats)
{
    struct port_to_netdev_data *data;

    ovs_rwlock_rdlock(&netdev_hmap_rwlock);
    HMAP_FOR_EACH (data, portno_node, &port_to_netdev) {
        if (data->dpif_class == dpif_class
            && !netdev_flow_del(data->netdev, ufid, stats)) {
            ovs_rwlock_unlock(&netdev_hmap_rwlock);
            return 0;
        }
    }
    ovs_rwlock_unlock(&netdev_hmap_rwlock);

    return ENOENT;
}

int
netdev_ports_flow_get(const struct dpif_class *dpif_class, struct match *match,
                      struct nlattr **actions, const ovs_u128 *ufid,
                      struct dpif_flow_stats *stats,
                      struct dpif_flow_attrs *attrs, struct ofpbuf *buf)
{
    struct port_to_netdev_data *data;

    ovs_rwlock_rdlock(&netdev_hmap_rwlock);
    HMAP_FOR_EACH (data, portno_node, &port_to_netdev) {
        if (data->dpif_class == dpif_class
            && !netdev_flow_get(data->netdev, match, actions,
                                ufid, stats, attrs, buf)) {
            ovs_rwlock_unlock(&netdev_hmap_rwlock);
            return 0;
        }
    }
    ovs_rwlock_unlock(&netdev_hmap_rwlock);
    return ENOENT;
}

static uint32_t
netdev_ports_hash(odp_port_t port, const struct dpif_class *dpif_class)
{
    return hash_int(odp_to_u32(port), hash_pointer(dpif_class, 0));
}

static struct port_to_netdev_data *
netdev_ports_lookup(odp_port_t port_no, const struct dpif_class *dpif_class)
    OVS_REQ_RDLOCK(netdev_hmap_rwlock)
{
    struct port_to_netdev_data *data;

    HMAP_FOR_EACH_WITH_HASH (data, portno_node,
                             netdev_ports_hash(port_no, dpif_class),
                             &port_to_netdev) {
        if (data->dpif_class == dpif_class
            && data->dpif_port.port_no == port_no) {
            return data;
        }
    }
    return NULL;
}

int
netdev_ports_insert(struct netdev *netdev, const struct dpif_class *dpif_class,
                    struct dpif_port *dpif_port)
{
    struct port_to_netdev_data *data;
    int ifindex = netdev_get_ifindex(netdev);

    if (ifindex < 0) {
        return ENODEV;
    }

    ovs_rwlock_wrlock(&netdev_hmap_rwlock);
    if (netdev_ports_lookup(dpif_port->port_no, dpif_class)) {
        ovs_rwlock_unlock(&netdev_hmap_rwlock);
        return EEXIST;
    }

    data = xzalloc(sizeof *data);
    data->netdev = netdev_ref(netdev);
    data->dpif_class = dpif_class;
    dpif_port_clone(&data->dpif_port, dpif_port);
    data->ifindex = ifindex;

    hmap_insert(&port_to_netdev, &data->portno_node,
                netdev_ports_hash(dpif_port->port_no, dpif_class));
    hmap_insert(&ifindex_to_port, &data->ifindex_node, ifindex);
    ovs_rwlock_unlock(&netdev_hmap_rwlock);

    netdev_init_flow_api(netdev);

    return 0;
}

struct netdev *
netdev_ports_get(odp_port_t port_no, const struct dpif_class *dpif_class)
{
    struct port_to_netdev_data *data;
    struct netdev *ret = NULL;

    ovs_rwlock_rdlock(&netdev_hmap_rwlock);
    data = netdev_ports_lookup(port_no, dpif_class);
    if (data) {
        ret = netdev_ref(data->netdev);
    }
    ovs_rwlock_unlock(&netdev_hmap_rwlock);

    return ret;
}

int
netdev_ports_remove(odp_port_t port_no, const struct dpif_class *dpif_class)
{
    struct port_to_netdev_data *data;
    int ret = ENOENT;

    ovs_rwlock_wrlock(&netdev_hmap_rwlock);
    data = netdev_ports_lookup(port_no, dpif_class);
    if (data) {
        dpif_port_destroy(&data->dpif_port);
        netdev_close(data->netdev); /* unref and possibly close */
        hmap_remove(&port_to_netdev, &data->portno_node);
        hmap_remove(&ifindex_to_port, &data->ifindex_node);
        free(data);
        ret = 0;
    }
    ovs_rwlock_unlock(&netdev_hmap_rwlock);

    return ret;
}

odp_port_t
netdev_ifindex_to_odp_port(int ifindex)
{
    struct port_to_netdev_data *data;
    odp_port_t ret = 0;

    ovs_rwlock_rdlock(&netdev_hmap_rwlock);
    HMAP_FOR_EACH_WITH_HASH (data, ifindex_node, ifindex, &ifindex_to_port) {
        if (data->ifindex == ifindex) {
            ret = data->dpif_port.port_no;
            break;
        }
    }
    ovs_rwlock_unlock(&netdev_hmap_rwlock);

    return ret;
}

static bool netdev_offload_rebalance_policy = false;

bool
netdev_is_offload_rebalance_policy_enabled(void)
{
    return netdev_offload_rebalance_policy;
}

static void
netdev_ports_flow_init(void)
{
    struct port_to_netdev_data *data;

    ovs_rwlock_rdlock(&netdev_hmap_rwlock);
    HMAP_FOR_EACH (data, portno_node, &port_to_netdev) {
       netdev_init_flow_api(data->netdev);
    }
    ovs_rwlock_unlock(&netdev_hmap_rwlock);
}

void
netdev_set_flow_api_enabled(const struct smap *ovs_other_config)
{
    if (smap_get_bool(ovs_other_config, "hw-offload", false)) {
        static struct ovsthread_once once = OVSTHREAD_ONCE_INITIALIZER;

        if (ovsthread_once_start(&once)) {
            netdev_flow_api_enabled = true;

            VLOG_INFO("netdev: Flow API Enabled");

#ifdef __linux__
            tc_set_policy(smap_get_def(ovs_other_config, "tc-policy",
                                       TC_POLICY_DEFAULT));
#endif

            if (smap_get_bool(ovs_other_config, "offload-rebalance", false)) {
                netdev_offload_rebalance_policy = true;
            }

            netdev_ports_flow_init();

            ovsthread_once_done(&once);
        }
    }
}