aboutsummaryrefslogtreecommitdiff
path: root/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib.c
blob: b7c21eb21a218de91a9b053addbfb0813fa4f74b (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
/*
 * Copyright (c) 2017, Mellanox Technologies. All rights reserved.
 *
 * This software is available to you under a choice of one of two
 * licenses.  You may choose to be licensed under the terms of the GNU
 * General Public License (GPL) Version 2, available from the file
 * COPYING in the main directory of this source tree, or the
 * OpenIB.org BSD license below:
 *
 *     Redistribution and use in source and binary forms, with or
 *     without modification, are permitted provided that the following
 *     conditions are met:
 *
 *      - Redistributions of source code must retain the above
 *        copyright notice, this list of conditions and the following
 *        disclaimer.
 *
 *      - Redistributions in binary form must reproduce the above
 *        copyright notice, this list of conditions and the following
 *        disclaimer in the documentation and/or other materials
 *        provided with the distribution.
 *
 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
 * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
 * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
 * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
 * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
 * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
 * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
 * SOFTWARE.
 */

#include <rdma/ib_verbs.h>
#include <linux/mlx5/fs.h>
#include "en.h"
#include "ipoib.h"

#define IB_DEFAULT_Q_KEY   0xb1b
#define MLX5I_PARAMS_DEFAULT_LOG_RQ_SIZE 9

static int mlx5i_open(struct net_device *netdev);
static int mlx5i_close(struct net_device *netdev);
static int mlx5i_change_mtu(struct net_device *netdev, int new_mtu);

static const struct net_device_ops mlx5i_netdev_ops = {
	.ndo_open                = mlx5i_open,
	.ndo_stop                = mlx5i_close,
	.ndo_init                = mlx5i_dev_init,
	.ndo_uninit              = mlx5i_dev_cleanup,
	.ndo_change_mtu          = mlx5i_change_mtu,
	.ndo_do_ioctl            = mlx5i_ioctl,
};

/* IPoIB mlx5 netdev profile */
static void mlx5i_build_nic_params(struct mlx5_core_dev *mdev,
				   struct mlx5e_params *params)
{
	/* Override RQ params as IPoIB supports only LINKED LIST RQ for now */
	MLX5E_SET_PFLAG(params, MLX5E_PFLAG_RX_STRIDING_RQ, false);
	mlx5e_set_rq_type(mdev, params);
	mlx5e_init_rq_type_params(mdev, params);

	/* RQ size in ipoib by default is 512 */
	params->log_rq_mtu_frames = is_kdump_kernel() ?
		MLX5E_PARAMS_MINIMUM_LOG_RQ_SIZE :
		MLX5I_PARAMS_DEFAULT_LOG_RQ_SIZE;

	params->lro_en = false;
	params->hard_mtu = MLX5_IB_GRH_BYTES + MLX5_IPOIB_HARD_LEN;
}

/* Called directly after IPoIB netdevice was created to initialize SW structs */
void mlx5i_init(struct mlx5_core_dev *mdev,
		struct net_device *netdev,
		const struct mlx5e_profile *profile,
		void *ppriv)
{
	struct mlx5e_priv *priv  = mlx5i_epriv(netdev);
	u16 max_mtu;

	/* priv init */
	priv->mdev        = mdev;
	priv->netdev      = netdev;
	priv->profile     = profile;
	priv->ppriv       = ppriv;
	mutex_init(&priv->state_lock);

	mlx5_query_port_max_mtu(mdev, &max_mtu, 1);
	netdev->mtu = max_mtu;

	mlx5e_build_nic_params(mdev, &priv->channels.params,
			       profile->max_nch(mdev), netdev->mtu);
	mlx5i_build_nic_params(mdev, &priv->channels.params);

	mlx5e_timestamp_init(priv);

	/* netdev init */
	netdev->hw_features    |= NETIF_F_SG;
	netdev->hw_features    |= NETIF_F_IP_CSUM;
	netdev->hw_features    |= NETIF_F_IPV6_CSUM;
	netdev->hw_features    |= NETIF_F_GRO;
	netdev->hw_features    |= NETIF_F_TSO;
	netdev->hw_features    |= NETIF_F_TSO6;
	netdev->hw_features    |= NETIF_F_RXCSUM;
	netdev->hw_features    |= NETIF_F_RXHASH;

	netdev->netdev_ops = &mlx5i_netdev_ops;
	netdev->ethtool_ops = &mlx5i_ethtool_ops;
}

/* Called directly before IPoIB netdevice is destroyed to cleanup SW structs */
static void mlx5i_cleanup(struct mlx5e_priv *priv)
{
	/* Do nothing .. */
}

int mlx5i_init_underlay_qp(struct mlx5e_priv *priv)
{
	struct mlx5_core_dev *mdev = priv->mdev;
	struct mlx5i_priv *ipriv = priv->ppriv;
	struct mlx5_core_qp *qp = &ipriv->qp;
	struct mlx5_qp_context *context;
	int ret;

	/* QP states */
	context = kzalloc(sizeof(*context), GFP_KERNEL);
	if (!context)
		return -ENOMEM;

	context->flags = cpu_to_be32(MLX5_QP_PM_MIGRATED << 11);
	context->pri_path.port = 1;
	context->pri_path.pkey_index = cpu_to_be16(ipriv->pkey_index);
	context->qkey = cpu_to_be32(IB_DEFAULT_Q_KEY);

	ret = mlx5_core_qp_modify(mdev, MLX5_CMD_OP_RST2INIT_QP, 0, context, qp);
	if (ret) {
		mlx5_core_err(mdev, "Failed to modify qp RST2INIT, err: %d\n", ret);
		goto err_qp_modify_to_err;
	}
	memset(context, 0, sizeof(*context));

	ret = mlx5_core_qp_modify(mdev, MLX5_CMD_OP_INIT2RTR_QP, 0, context, qp);
	if (ret) {
		mlx5_core_err(mdev, "Failed to modify qp INIT2RTR, err: %d\n", ret);
		goto err_qp_modify_to_err;
	}

	ret = mlx5_core_qp_modify(mdev, MLX5_CMD_OP_RTR2RTS_QP, 0, context, qp);
	if (ret) {
		mlx5_core_err(mdev, "Failed to modify qp RTR2RTS, err: %d\n", ret);
		goto err_qp_modify_to_err;
	}

	kfree(context);
	return 0;

err_qp_modify_to_err:
	mlx5_core_qp_modify(mdev, MLX5_CMD_OP_2ERR_QP, 0, &context, qp);
	kfree(context);
	return ret;
}

void mlx5i_uninit_underlay_qp(struct mlx5e_priv *priv)
{
	struct mlx5i_priv *ipriv = priv->ppriv;
	struct mlx5_core_dev *mdev = priv->mdev;
	struct mlx5_qp_context context;
	int err;

	err = mlx5_core_qp_modify(mdev, MLX5_CMD_OP_2RST_QP, 0, &context,
				  &ipriv->qp);
	if (err)
		mlx5_core_err(mdev, "Failed to modify qp 2RST, err: %d\n", err);
}

#define MLX5_QP_ENHANCED_ULP_STATELESS_MODE 2

int mlx5i_create_underlay_qp(struct mlx5_core_dev *mdev, struct mlx5_core_qp *qp)
{
	u32 *in = NULL;
	void *addr_path;
	int ret = 0;
	int inlen;
	void *qpc;

	inlen = MLX5_ST_SZ_BYTES(create_qp_in);
	in = kvzalloc(inlen, GFP_KERNEL);
	if (!in)
		return -ENOMEM;

	qpc = MLX5_ADDR_OF(create_qp_in, in, qpc);
	MLX5_SET(qpc, qpc, st, MLX5_QP_ST_UD);
	MLX5_SET(qpc, qpc, pm_state, MLX5_QP_PM_MIGRATED);
	MLX5_SET(qpc, qpc, ulp_stateless_offload_mode,
		 MLX5_QP_ENHANCED_ULP_STATELESS_MODE);

	addr_path = MLX5_ADDR_OF(qpc, qpc, primary_address_path);
	MLX5_SET(ads, addr_path, vhca_port_num, 1);
	MLX5_SET(ads, addr_path, grh, 1);

	ret = mlx5_core_create_qp(mdev, qp, in, inlen);
	if (ret) {
		mlx5_core_err(mdev, "Failed creating IPoIB QP err : %d\n", ret);
		goto out;
	}

out:
	kvfree(in);
	return ret;
}

void mlx5i_destroy_underlay_qp(struct mlx5_core_dev *mdev, struct mlx5_core_qp *qp)
{
	mlx5_core_destroy_qp(mdev, qp);
}

static int mlx5i_init_tx(struct mlx5e_priv *priv)
{
	struct mlx5i_priv *ipriv = priv->ppriv;
	int err;

	err = mlx5i_create_underlay_qp(priv->mdev, &ipriv->qp);
	if (err) {
		mlx5_core_warn(priv->mdev, "create underlay QP failed, %d\n", err);
		return err;
	}

	err = mlx5e_create_tis(priv->mdev, 0 /* tc */, ipriv->qp.qpn, &priv->tisn[0]);
	if (err) {
		mlx5_core_warn(priv->mdev, "create tis failed, %d\n", err);
		goto err_destroy_underlay_qp;
	}

	return 0;

err_destroy_underlay_qp:
	mlx5i_destroy_underlay_qp(priv->mdev, &ipriv->qp);
	return err;
}

static void mlx5i_cleanup_tx(struct mlx5e_priv *priv)
{
	struct mlx5i_priv *ipriv = priv->ppriv;

	mlx5e_destroy_tis(priv->mdev, priv->tisn[0]);
	mlx5i_destroy_underlay_qp(priv->mdev, &ipriv->qp);
}

static int mlx5i_create_flow_steering(struct mlx5e_priv *priv)
{
	struct ttc_params ttc_params = {};
	int tt, err;

	priv->fs.ns = mlx5_get_flow_namespace(priv->mdev,
					       MLX5_FLOW_NAMESPACE_KERNEL);

	if (!priv->fs.ns)
		return -EINVAL;

	err = mlx5e_arfs_create_tables(priv);
	if (err) {
		netdev_err(priv->netdev, "Failed to create arfs tables, err=%d\n",
			   err);
		priv->netdev->hw_features &= ~NETIF_F_NTUPLE;
	}

	mlx5e_set_ttc_basic_params(priv, &ttc_params);
	mlx5e_set_inner_ttc_ft_params(&ttc_params);
	for (tt = 0; tt < MLX5E_NUM_INDIR_TIRS; tt++)
		ttc_params.indir_tirn[tt] = priv->inner_indir_tir[tt].tirn;

	err = mlx5e_create_inner_ttc_table(priv, &ttc_params, &priv->fs.inner_ttc);
	if (err) {
		netdev_err(priv->netdev, "Failed to create inner ttc table, err=%d\n",
			   err);
		goto err_destroy_arfs_tables;
	}

	mlx5e_set_ttc_ft_params(&ttc_params);
	for (tt = 0; tt < MLX5E_NUM_INDIR_TIRS; tt++)
		ttc_params.indir_tirn[tt] = priv->indir_tir[tt].tirn;

	err = mlx5e_create_ttc_table(priv, &ttc_params, &priv->fs.ttc);
	if (err) {
		netdev_err(priv->netdev, "Failed to create ttc table, err=%d\n",
			   err);
		goto err_destroy_inner_ttc_table;
	}

	return 0;

err_destroy_inner_ttc_table:
	mlx5e_destroy_inner_ttc_table(priv, &priv->fs.inner_ttc);
err_destroy_arfs_tables:
	mlx5e_arfs_destroy_tables(priv);

	return err;
}

static void mlx5i_destroy_flow_steering(struct mlx5e_priv *priv)
{
	mlx5e_destroy_ttc_table(priv, &priv->fs.ttc);
	mlx5e_destroy_inner_ttc_table(priv, &priv->fs.inner_ttc);
	mlx5e_arfs_destroy_tables(priv);
}

static int mlx5i_init_rx(struct mlx5e_priv *priv)
{
	int err;

	err = mlx5e_create_indirect_rqt(priv);
	if (err)
		return err;

	err = mlx5e_create_direct_rqts(priv);
	if (err)
		goto err_destroy_indirect_rqts;

	err = mlx5e_create_indirect_tirs(priv);
	if (err)
		goto err_destroy_direct_rqts;

	err = mlx5e_create_direct_tirs(priv);
	if (err)
		goto err_destroy_indirect_tirs;

	err = mlx5i_create_flow_steering(priv);
	if (err)
		goto err_destroy_direct_tirs;

	return 0;

err_destroy_direct_tirs:
	mlx5e_destroy_direct_tirs(priv);
err_destroy_indirect_tirs:
	mlx5e_destroy_indirect_tirs(priv);
err_destroy_direct_rqts:
	mlx5e_destroy_direct_rqts(priv);
err_destroy_indirect_rqts:
	mlx5e_destroy_rqt(priv, &priv->indir_rqt);
	return err;
}

static void mlx5i_cleanup_rx(struct mlx5e_priv *priv)
{
	mlx5i_destroy_flow_steering(priv);
	mlx5e_destroy_direct_tirs(priv);
	mlx5e_destroy_indirect_tirs(priv);
	mlx5e_destroy_direct_rqts(priv);
	mlx5e_destroy_rqt(priv, &priv->indir_rqt);
}

static const struct mlx5e_profile mlx5i_nic_profile = {
	.init		   = mlx5i_init,
	.cleanup	   = mlx5i_cleanup,
	.init_tx	   = mlx5i_init_tx,
	.cleanup_tx	   = mlx5i_cleanup_tx,
	.init_rx	   = mlx5i_init_rx,
	.cleanup_rx	   = mlx5i_cleanup_rx,
	.enable		   = NULL, /* mlx5i_enable */
	.disable	   = NULL, /* mlx5i_disable */
	.update_stats	   = NULL, /* mlx5i_update_stats */
	.max_nch	   = mlx5e_get_max_num_channels,
	.update_carrier    = NULL, /* no HW update in IB link */
	.rx_handlers.handle_rx_cqe       = mlx5i_handle_rx_cqe,
	.rx_handlers.handle_rx_cqe_mpwqe = NULL, /* Not supported */
	.max_tc		   = MLX5I_MAX_NUM_TC,
};

/* mlx5i netdev NDos */

static int mlx5i_change_mtu(struct net_device *netdev, int new_mtu)
{
	struct mlx5e_priv *priv = mlx5i_epriv(netdev);
	struct mlx5e_channels new_channels = {};
	struct mlx5e_params *params;
	int err = 0;

	mutex_lock(&priv->state_lock);

	params = &priv->channels.params;

	if (!test_bit(MLX5E_STATE_OPENED, &priv->state)) {
		params->sw_mtu = new_mtu;
		netdev->mtu = params->sw_mtu;
		goto out;
	}

	new_channels.params = *params;
	new_channels.params.sw_mtu = new_mtu;
	err = mlx5e_open_channels(priv, &new_channels);
	if (err)
		goto out;

	mlx5e_switch_priv_channels(priv, &new_channels, NULL);
	netdev->mtu = new_channels.params.sw_mtu;

out:
	mutex_unlock(&priv->state_lock);
	return err;
}

int mlx5i_dev_init(struct net_device *dev)
{
	struct mlx5e_priv    *priv   = mlx5i_epriv(dev);
	struct mlx5i_priv    *ipriv  = priv->ppriv;

	/* Set dev address using underlay QP */
	dev->dev_addr[1] = (ipriv->qp.qpn >> 16) & 0xff;
	dev->dev_addr[2] = (ipriv->qp.qpn >>  8) & 0xff;
	dev->dev_addr[3] = (ipriv->qp.qpn) & 0xff;

	/* Add QPN to net-device mapping to HT */
	mlx5i_pkey_add_qpn(dev ,ipriv->qp.qpn);

	return 0;
}

int mlx5i_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd)
{
	struct mlx5e_priv *priv = mlx5i_epriv(dev);

	switch (cmd) {
	case SIOCSHWTSTAMP:
		return mlx5e_hwstamp_set(priv, ifr);
	case SIOCGHWTSTAMP:
		return mlx5e_hwstamp_get(priv, ifr);
	default:
		return -EOPNOTSUPP;
	}
}

void mlx5i_dev_cleanup(struct net_device *dev)
{
	struct mlx5e_priv    *priv   = mlx5i_epriv(dev);
	struct mlx5i_priv    *ipriv = priv->ppriv;

	mlx5i_uninit_underlay_qp(priv);

	/* Delete QPN to net-device mapping from HT */
	mlx5i_pkey_del_qpn(dev, ipriv->qp.qpn);
}

static int mlx5i_open(struct net_device *netdev)
{
	struct mlx5e_priv *epriv = mlx5i_epriv(netdev);
	struct mlx5i_priv *ipriv = epriv->ppriv;
	struct mlx5_core_dev *mdev = epriv->mdev;
	int err;

	mutex_lock(&epriv->state_lock);

	set_bit(MLX5E_STATE_OPENED, &epriv->state);

	err = mlx5i_init_underlay_qp(epriv);
	if (err) {
		mlx5_core_warn(mdev, "prepare underlay qp state failed, %d\n", err);
		goto err_clear_state_opened_flag;
	}

	err = mlx5_fs_add_rx_underlay_qpn(mdev, ipriv->qp.qpn);
	if (err) {
		mlx5_core_warn(mdev, "attach underlay qp to ft failed, %d\n", err);
		goto err_reset_qp;
	}

	err = mlx5e_open_channels(epriv, &epriv->channels);
	if (err)
		goto err_remove_fs_underlay_qp;

	mlx5e_refresh_tirs(epriv, false);
	mlx5e_activate_priv_channels(epriv);

	mutex_unlock(&epriv->state_lock);
	return 0;

err_remove_fs_underlay_qp:
	mlx5_fs_remove_rx_underlay_qpn(mdev, ipriv->qp.qpn);
err_reset_qp:
	mlx5i_uninit_underlay_qp(epriv);
err_clear_state_opened_flag:
	clear_bit(MLX5E_STATE_OPENED, &epriv->state);
	mutex_unlock(&epriv->state_lock);
	return err;
}

static int mlx5i_close(struct net_device *netdev)
{
	struct mlx5e_priv *epriv = mlx5i_epriv(netdev);
	struct mlx5i_priv *ipriv = epriv->ppriv;
	struct mlx5_core_dev *mdev = epriv->mdev;

	/* May already be CLOSED in case a previous configuration operation
	 * (e.g RX/TX queue size change) that involves close&open failed.
	 */
	mutex_lock(&epriv->state_lock);

	if (!test_bit(MLX5E_STATE_OPENED, &epriv->state))
		goto unlock;

	clear_bit(MLX5E_STATE_OPENED, &epriv->state);

	netif_carrier_off(epriv->netdev);
	mlx5_fs_remove_rx_underlay_qpn(mdev, ipriv->qp.qpn);
	mlx5i_uninit_underlay_qp(epriv);
	mlx5e_deactivate_priv_channels(epriv);
	mlx5e_close_channels(&epriv->channels);
unlock:
	mutex_unlock(&epriv->state_lock);
	return 0;
}

/* IPoIB RDMA netdev callbacks */
static int mlx5i_attach_mcast(struct net_device *netdev, struct ib_device *hca,
			      union ib_gid *gid, u16 lid, int set_qkey,
			      u32 qkey)
{
	struct mlx5e_priv    *epriv = mlx5i_epriv(netdev);
	struct mlx5_core_dev *mdev  = epriv->mdev;
	struct mlx5i_priv    *ipriv = epriv->ppriv;
	int err;

	mlx5_core_dbg(mdev, "attaching QPN 0x%x, MGID %pI6\n", ipriv->qp.qpn, gid->raw);
	err = mlx5_core_attach_mcg(mdev, gid, ipriv->qp.qpn);
	if (err)
		mlx5_core_warn(mdev, "failed attaching QPN 0x%x, MGID %pI6\n",
			       ipriv->qp.qpn, gid->raw);

	if (set_qkey) {
		mlx5_core_dbg(mdev, "%s setting qkey 0x%x\n",
			      netdev->name, qkey);
		ipriv->qkey = qkey;
	}

	return err;
}

static int mlx5i_detach_mcast(struct net_device *netdev, struct ib_device *hca,
			      union ib_gid *gid, u16 lid)
{
	struct mlx5e_priv    *epriv = mlx5i_epriv(netdev);
	struct mlx5_core_dev *mdev  = epriv->mdev;
	struct mlx5i_priv    *ipriv = epriv->ppriv;
	int err;

	mlx5_core_dbg(mdev, "detaching QPN 0x%x, MGID %pI6\n", ipriv->qp.qpn, gid->raw);

	err = mlx5_core_detach_mcg(mdev, gid, ipriv->qp.qpn);
	if (err)
		mlx5_core_dbg(mdev, "failed detaching QPN 0x%x, MGID %pI6\n",
			      ipriv->qp.qpn, gid->raw);

	return err;
}

static int mlx5i_xmit(struct net_device *dev, struct sk_buff *skb,
		      struct ib_ah *address, u32 dqpn)
{
	struct mlx5e_priv *epriv = mlx5i_epriv(dev);
	struct mlx5e_txqsq *sq   = epriv->txq2sq[skb_get_queue_mapping(skb)];
	struct mlx5_ib_ah *mah   = to_mah(address);
	struct mlx5i_priv *ipriv = epriv->ppriv;

	return mlx5i_sq_xmit(sq, skb, &mah->av, dqpn, ipriv->qkey);
}

static void mlx5i_set_pkey_index(struct net_device *netdev, int id)
{
	struct mlx5i_priv *ipriv = netdev_priv(netdev);

	ipriv->pkey_index = (u16)id;
}

static int mlx5i_check_required_hca_cap(struct mlx5_core_dev *mdev)
{
	if (MLX5_CAP_GEN(mdev, port_type) != MLX5_CAP_PORT_TYPE_IB)
		return -EOPNOTSUPP;

	if (!MLX5_CAP_GEN(mdev, ipoib_enhanced_offloads)) {
		mlx5_core_warn(mdev, "IPoIB enhanced offloads are not supported\n");
		return -EOPNOTSUPP;
	}

	return 0;
}

struct net_device *mlx5_rdma_netdev_alloc(struct mlx5_core_dev *mdev,
					  struct ib_device *ibdev,
					  const char *name,
					  void (*setup)(struct net_device *))
{
	const struct mlx5e_profile *profile;
	struct net_device *netdev;
	struct mlx5i_priv *ipriv;
	struct mlx5e_priv *epriv;
	struct rdma_netdev *rn;
	bool sub_interface;
	int nch;
	int err;

	if (mlx5i_check_required_hca_cap(mdev)) {
		mlx5_core_warn(mdev, "Accelerated mode is not supported\n");
		return ERR_PTR(-EOPNOTSUPP);
	}

	/* TODO: Need to find a better way to check if child device*/
	sub_interface = (mdev->mlx5e_res.pdn != 0);

	if (sub_interface)
		profile = mlx5i_pkey_get_profile();
	else
		profile = &mlx5i_nic_profile;

	nch = profile->max_nch(mdev);

	netdev = alloc_netdev_mqs(sizeof(struct mlx5i_priv) + sizeof(struct mlx5e_priv),
				  name, NET_NAME_UNKNOWN,
				  setup,
				  nch * MLX5E_MAX_NUM_TC,
				  nch);
	if (!netdev) {
		mlx5_core_warn(mdev, "alloc_netdev_mqs failed\n");
		return NULL;
	}

	ipriv = netdev_priv(netdev);
	epriv = mlx5i_epriv(netdev);

	epriv->wq = create_singlethread_workqueue("mlx5i");
	if (!epriv->wq)
		goto err_free_netdev;

	ipriv->sub_interface = sub_interface;
	if (!ipriv->sub_interface) {
		err = mlx5i_pkey_qpn_ht_init(netdev);
		if (err) {
			mlx5_core_warn(mdev, "allocate qpn_to_netdev ht failed\n");
			goto destroy_wq;
		}

		/* This should only be called once per mdev */
		err = mlx5e_create_mdev_resources(mdev);
		if (err)
			goto destroy_ht;
	}

	profile->init(mdev, netdev, profile, ipriv);

	mlx5e_attach_netdev(epriv);
	netif_carrier_off(netdev);

	/* set rdma_netdev func pointers */
	rn = &ipriv->rn;
	rn->hca  = ibdev;
	rn->send = mlx5i_xmit;
	rn->attach_mcast = mlx5i_attach_mcast;
	rn->detach_mcast = mlx5i_detach_mcast;
	rn->set_id = mlx5i_set_pkey_index;

	return netdev;

destroy_ht:
	mlx5i_pkey_qpn_ht_cleanup(netdev);
destroy_wq:
	destroy_workqueue(epriv->wq);
err_free_netdev:
	free_netdev(netdev);

	return NULL;
}
EXPORT_SYMBOL(mlx5_rdma_netdev_alloc);

void mlx5_rdma_netdev_free(struct net_device *netdev)
{
	struct mlx5e_priv *priv = mlx5i_epriv(netdev);
	struct mlx5i_priv *ipriv = priv->ppriv;
	const struct mlx5e_profile *profile = priv->profile;

	mlx5e_detach_netdev(priv);
	profile->cleanup(priv);
	destroy_workqueue(priv->wq);

	if (!ipriv->sub_interface) {
		mlx5i_pkey_qpn_ht_cleanup(netdev);
		mlx5e_destroy_mdev_resources(priv->mdev);
	}
	free_netdev(netdev);
}
EXPORT_SYMBOL(mlx5_rdma_netdev_free);