bnxt_sriov.c 25 KB
Newer Older
1
2
/* Broadcom NetXtreme-C/E network driver.
 *
3
 * Copyright (c) 2014-2016 Broadcom Corporation
4
 * Copyright (c) 2016-2017 Broadcom Limited
5
6
7
8
9
10
11
12
13
14
15
16
17
18
 *
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation.
 */

#include <linux/module.h>
#include <linux/pci.h>
#include <linux/netdevice.h>
#include <linux/if_vlan.h>
#include <linux/interrupt.h>
#include <linux/etherdevice.h>
#include "bnxt_hsi.h"
#include "bnxt.h"
19
#include "bnxt_ulp.h"
20
#include "bnxt_sriov.h"
21
#include "bnxt_vfr.h"
22
23
24
#include "bnxt_ethtool.h"

#ifdef CONFIG_BNXT_SRIOV
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
static int bnxt_hwrm_fwd_async_event_cmpl(struct bnxt *bp,
					  struct bnxt_vf_info *vf, u16 event_id)
{
	struct hwrm_fwd_async_event_cmpl_output *resp = bp->hwrm_cmd_resp_addr;
	struct hwrm_fwd_async_event_cmpl_input req = {0};
	struct hwrm_async_event_cmpl *async_cmpl;
	int rc = 0;

	bnxt_hwrm_cmd_hdr_init(bp, &req, HWRM_FWD_ASYNC_EVENT_CMPL, -1, -1);
	if (vf)
		req.encap_async_event_target_id = cpu_to_le16(vf->fw_fid);
	else
		/* broadcast this async event to all VFs */
		req.encap_async_event_target_id = cpu_to_le16(0xffff);
	async_cmpl = (struct hwrm_async_event_cmpl *)req.encap_async_event_cmpl;
40
	async_cmpl->type = cpu_to_le16(ASYNC_EVENT_CMPL_TYPE_HWRM_ASYNC_EVENT);
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
	async_cmpl->event_id = cpu_to_le16(event_id);

	mutex_lock(&bp->hwrm_cmd_lock);
	rc = _hwrm_send_message(bp, &req, sizeof(req), HWRM_CMD_TIMEOUT);

	if (rc) {
		netdev_err(bp->dev, "hwrm_fwd_async_event_cmpl failed. rc:%d\n",
			   rc);
		goto fwd_async_event_cmpl_exit;
	}

	if (resp->error_code) {
		netdev_err(bp->dev, "hwrm_fwd_async_event_cmpl error %d\n",
			   resp->error_code);
		rc = -1;
	}

fwd_async_event_cmpl_exit:
	mutex_unlock(&bp->hwrm_cmd_lock);
	return rc;
}

63
64
static int bnxt_vf_ndo_prep(struct bnxt *bp, int vf_id)
{
65
	if (!test_bit(BNXT_STATE_OPEN, &bp->state)) {
66
67
68
69
70
71
72
		netdev_err(bp->dev, "vf ndo called though PF is down\n");
		return -EINVAL;
	}
	if (!bp->pf.active_vfs) {
		netdev_err(bp->dev, "vf ndo called though sriov is disabled\n");
		return -EINVAL;
	}
73
	if (vf_id >= bp->pf.active_vfs) {
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
		netdev_err(bp->dev, "Invalid VF id %d\n", vf_id);
		return -EINVAL;
	}
	return 0;
}

int bnxt_set_vf_spoofchk(struct net_device *dev, int vf_id, bool setting)
{
	struct hwrm_func_cfg_input req = {0};
	struct bnxt *bp = netdev_priv(dev);
	struct bnxt_vf_info *vf;
	bool old_setting = false;
	u32 func_flags;
	int rc;

89
90
91
	if (bp->hwrm_spec_code < 0x10701)
		return -ENOTSUPP;

92
93
94
95
96
97
98
99
100
101
102
103
	rc = bnxt_vf_ndo_prep(bp, vf_id);
	if (rc)
		return rc;

	vf = &bp->pf.vf[vf_id];
	if (vf->flags & BNXT_VF_SPOOFCHK)
		old_setting = true;
	if (old_setting == setting)
		return 0;

	func_flags = vf->func_flags;
	if (setting)
104
		func_flags |= FUNC_CFG_REQ_FLAGS_SRC_MAC_ADDR_CHECK_ENABLE;
105
	else
106
		func_flags |= FUNC_CFG_REQ_FLAGS_SRC_MAC_ADDR_CHECK_DISABLE;
107
108
109
110
	/*TODO: if the driver supports VLAN filter on guest VLAN,
	 * the spoof check should also include vlan anti-spoofing
	 */
	bnxt_hwrm_cmd_hdr_init(bp, &req, HWRM_FUNC_CFG, -1, -1);
111
	req.fid = cpu_to_le16(vf->fw_fid);
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
	req.flags = cpu_to_le32(func_flags);
	rc = hwrm_send_message(bp, &req, sizeof(req), HWRM_CMD_TIMEOUT);
	if (!rc) {
		vf->func_flags = func_flags;
		if (setting)
			vf->flags |= BNXT_VF_SPOOFCHK;
		else
			vf->flags &= ~BNXT_VF_SPOOFCHK;
	}
	return rc;
}

int bnxt_get_vf_config(struct net_device *dev, int vf_id,
		       struct ifla_vf_info *ivi)
{
	struct bnxt *bp = netdev_priv(dev);
	struct bnxt_vf_info *vf;
	int rc;

	rc = bnxt_vf_ndo_prep(bp, vf_id);
	if (rc)
		return rc;

	ivi->vf = vf_id;
	vf = &bp->pf.vf[vf_id];

	memcpy(&ivi->mac, vf->mac_addr, ETH_ALEN);
	ivi->max_tx_rate = vf->max_tx_rate;
	ivi->min_tx_rate = vf->min_tx_rate;
	ivi->vlan = vf->vlan;
142
143
144
145
146
	if (vf->flags & BNXT_VF_QOS)
		ivi->qos = vf->vlan >> VLAN_PRIO_SHIFT;
	else
		ivi->qos = 0;
	ivi->spoofchk = !!(vf->flags & BNXT_VF_SPOOFCHK);
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
	if (!(vf->flags & BNXT_VF_LINK_FORCED))
		ivi->linkstate = IFLA_VF_LINK_STATE_AUTO;
	else if (vf->flags & BNXT_VF_LINK_UP)
		ivi->linkstate = IFLA_VF_LINK_STATE_ENABLE;
	else
		ivi->linkstate = IFLA_VF_LINK_STATE_DISABLE;

	return 0;
}

int bnxt_set_vf_mac(struct net_device *dev, int vf_id, u8 *mac)
{
	struct hwrm_func_cfg_input req = {0};
	struct bnxt *bp = netdev_priv(dev);
	struct bnxt_vf_info *vf;
	int rc;

	rc = bnxt_vf_ndo_prep(bp, vf_id);
	if (rc)
		return rc;
	/* reject bc or mc mac addr, zero mac addr means allow
	 * VF to use its own mac addr
	 */
	if (is_multicast_ether_addr(mac)) {
		netdev_err(dev, "Invalid VF ethernet address\n");
		return -EINVAL;
	}
	vf = &bp->pf.vf[vf_id];

	memcpy(vf->mac_addr, mac, ETH_ALEN);
	bnxt_hwrm_cmd_hdr_init(bp, &req, HWRM_FUNC_CFG, -1, -1);
178
	req.fid = cpu_to_le16(vf->fw_fid);
179
180
181
182
183
184
	req.flags = cpu_to_le32(vf->func_flags);
	req.enables = cpu_to_le32(FUNC_CFG_REQ_ENABLES_DFLT_MAC_ADDR);
	memcpy(req.dflt_mac_addr, mac, ETH_ALEN);
	return hwrm_send_message(bp, &req, sizeof(req), HWRM_CMD_TIMEOUT);
}

185
186
int bnxt_set_vf_vlan(struct net_device *dev, int vf_id, u16 vlan_id, u8 qos,
		     __be16 vlan_proto)
187
188
189
190
191
192
193
{
	struct hwrm_func_cfg_input req = {0};
	struct bnxt *bp = netdev_priv(dev);
	struct bnxt_vf_info *vf;
	u16 vlan_tag;
	int rc;

194
195
196
	if (bp->hwrm_spec_code < 0x10201)
		return -ENOTSUPP;

197
198
199
	if (vlan_proto != htons(ETH_P_8021Q))
		return -EPROTONOSUPPORT;

200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
	rc = bnxt_vf_ndo_prep(bp, vf_id);
	if (rc)
		return rc;

	/* TODO: needed to implement proper handling of user priority,
	 * currently fail the command if there is valid priority
	 */
	if (vlan_id > 4095 || qos)
		return -EINVAL;

	vf = &bp->pf.vf[vf_id];
	vlan_tag = vlan_id;
	if (vlan_tag == vf->vlan)
		return 0;

	bnxt_hwrm_cmd_hdr_init(bp, &req, HWRM_FUNC_CFG, -1, -1);
216
	req.fid = cpu_to_le16(vf->fw_fid);
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
	req.flags = cpu_to_le32(vf->func_flags);
	req.dflt_vlan = cpu_to_le16(vlan_tag);
	req.enables = cpu_to_le32(FUNC_CFG_REQ_ENABLES_DFLT_VLAN);
	rc = hwrm_send_message(bp, &req, sizeof(req), HWRM_CMD_TIMEOUT);
	if (!rc)
		vf->vlan = vlan_tag;
	return rc;
}

int bnxt_set_vf_bw(struct net_device *dev, int vf_id, int min_tx_rate,
		   int max_tx_rate)
{
	struct hwrm_func_cfg_input req = {0};
	struct bnxt *bp = netdev_priv(dev);
	struct bnxt_vf_info *vf;
	u32 pf_link_speed;
	int rc;

	rc = bnxt_vf_ndo_prep(bp, vf_id);
	if (rc)
		return rc;

	vf = &bp->pf.vf[vf_id];
	pf_link_speed = bnxt_fw_to_ethtool_speed(bp->link_info.link_speed);
	if (max_tx_rate > pf_link_speed) {
		netdev_info(bp->dev, "max tx rate %d exceed PF link speed for VF %d\n",
			    max_tx_rate, vf_id);
		return -EINVAL;
	}

	if (min_tx_rate > pf_link_speed || min_tx_rate > max_tx_rate) {
		netdev_info(bp->dev, "min tx rate %d is invalid for VF %d\n",
			    min_tx_rate, vf_id);
		return -EINVAL;
	}
	if (min_tx_rate == vf->min_tx_rate && max_tx_rate == vf->max_tx_rate)
		return 0;
	bnxt_hwrm_cmd_hdr_init(bp, &req, HWRM_FUNC_CFG, -1, -1);
255
	req.fid = cpu_to_le16(vf->fw_fid);
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
	req.flags = cpu_to_le32(vf->func_flags);
	req.enables = cpu_to_le32(FUNC_CFG_REQ_ENABLES_MAX_BW);
	req.max_bw = cpu_to_le32(max_tx_rate);
	req.enables |= cpu_to_le32(FUNC_CFG_REQ_ENABLES_MIN_BW);
	req.min_bw = cpu_to_le32(min_tx_rate);
	rc = hwrm_send_message(bp, &req, sizeof(req), HWRM_CMD_TIMEOUT);
	if (!rc) {
		vf->min_tx_rate = min_tx_rate;
		vf->max_tx_rate = max_tx_rate;
	}
	return rc;
}

int bnxt_set_vf_link_state(struct net_device *dev, int vf_id, int link)
{
	struct bnxt *bp = netdev_priv(dev);
	struct bnxt_vf_info *vf;
	int rc;

	rc = bnxt_vf_ndo_prep(bp, vf_id);
	if (rc)
		return rc;

	vf = &bp->pf.vf[vf_id];

	vf->flags &= ~(BNXT_VF_LINK_UP | BNXT_VF_LINK_FORCED);
	switch (link) {
	case IFLA_VF_LINK_STATE_AUTO:
		vf->flags |= BNXT_VF_LINK_UP;
		break;
	case IFLA_VF_LINK_STATE_DISABLE:
		vf->flags |= BNXT_VF_LINK_FORCED;
		break;
	case IFLA_VF_LINK_STATE_ENABLE:
		vf->flags |= BNXT_VF_LINK_UP | BNXT_VF_LINK_FORCED;
		break;
	default:
		netdev_err(bp->dev, "Invalid link option\n");
		rc = -EINVAL;
		break;
	}
297
298
	if (vf->flags & (BNXT_VF_LINK_UP | BNXT_VF_LINK_FORCED))
		rc = bnxt_hwrm_fwd_async_event_cmpl(bp, vf,
299
			ASYNC_EVENT_CMPL_EVENT_ID_LINK_STATUS_CHANGE);
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
	return rc;
}

static int bnxt_set_vf_attr(struct bnxt *bp, int num_vfs)
{
	int i;
	struct bnxt_vf_info *vf;

	for (i = 0; i < num_vfs; i++) {
		vf = &bp->pf.vf[i];
		memset(vf, 0, sizeof(*vf));
	}
	return 0;
}

315
static int bnxt_hwrm_func_vf_resource_free(struct bnxt *bp, int num_vfs)
316
317
318
319
320
321
322
323
{
	int i, rc = 0;
	struct bnxt_pf_info *pf = &bp->pf;
	struct hwrm_func_vf_resc_free_input req = {0};

	bnxt_hwrm_cmd_hdr_init(bp, &req, HWRM_FUNC_VF_RESC_FREE, -1, -1);

	mutex_lock(&bp->hwrm_cmd_lock);
324
	for (i = pf->first_vf_id; i < pf->first_vf_id + num_vfs; i++) {
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
		req.vf_id = cpu_to_le16(i);
		rc = _hwrm_send_message(bp, &req, sizeof(req),
					HWRM_CMD_TIMEOUT);
		if (rc)
			break;
	}
	mutex_unlock(&bp->hwrm_cmd_lock);
	return rc;
}

static void bnxt_free_vf_resources(struct bnxt *bp)
{
	struct pci_dev *pdev = bp->pdev;
	int i;

	kfree(bp->pf.vf_event_bmap);
	bp->pf.vf_event_bmap = NULL;

	for (i = 0; i < 4; i++) {
		if (bp->pf.hwrm_cmd_req_addr[i]) {
			dma_free_coherent(&pdev->dev, BNXT_PAGE_SIZE,
					  bp->pf.hwrm_cmd_req_addr[i],
					  bp->pf.hwrm_cmd_req_dma_addr[i]);
			bp->pf.hwrm_cmd_req_addr[i] = NULL;
		}
	}

	kfree(bp->pf.vf);
	bp->pf.vf = NULL;
}

static int bnxt_alloc_vf_resources(struct bnxt *bp, int num_vfs)
{
	struct pci_dev *pdev = bp->pdev;
	u32 nr_pages, size, i, j, k = 0;

	bp->pf.vf = kcalloc(num_vfs, sizeof(struct bnxt_vf_info), GFP_KERNEL);
	if (!bp->pf.vf)
		return -ENOMEM;

	bnxt_set_vf_attr(bp, num_vfs);

	size = num_vfs * BNXT_HWRM_REQ_MAX_SIZE;
	nr_pages = size / BNXT_PAGE_SIZE;
	if (size & (BNXT_PAGE_SIZE - 1))
		nr_pages++;

	for (i = 0; i < nr_pages; i++) {
		bp->pf.hwrm_cmd_req_addr[i] =
			dma_alloc_coherent(&pdev->dev, BNXT_PAGE_SIZE,
					   &bp->pf.hwrm_cmd_req_dma_addr[i],
					   GFP_KERNEL);

		if (!bp->pf.hwrm_cmd_req_addr[i])
			return -ENOMEM;

		for (j = 0; j < BNXT_HWRM_REQS_PER_PAGE && k < num_vfs; j++) {
			struct bnxt_vf_info *vf = &bp->pf.vf[k];

			vf->hwrm_cmd_req_addr = bp->pf.hwrm_cmd_req_addr[i] +
						j * BNXT_HWRM_REQ_MAX_SIZE;
			vf->hwrm_cmd_req_dma_addr =
				bp->pf.hwrm_cmd_req_dma_addr[i] + j *
				BNXT_HWRM_REQ_MAX_SIZE;
			k++;
		}
	}

	/* Max 128 VF's */
	bp->pf.vf_event_bmap = kzalloc(16, GFP_KERNEL);
	if (!bp->pf.vf_event_bmap)
		return -ENOMEM;

	bp->pf.hwrm_cmd_req_pages = nr_pages;
	return 0;
}

static int bnxt_hwrm_func_buf_rgtr(struct bnxt *bp)
{
	struct hwrm_func_buf_rgtr_input req = {0};

	bnxt_hwrm_cmd_hdr_init(bp, &req, HWRM_FUNC_BUF_RGTR, -1, -1);

	req.req_buf_num_pages = cpu_to_le16(bp->pf.hwrm_cmd_req_pages);
	req.req_buf_page_size = cpu_to_le16(BNXT_PAGE_SHIFT);
	req.req_buf_len = cpu_to_le16(BNXT_HWRM_REQ_MAX_SIZE);
	req.req_buf_page_addr0 = cpu_to_le64(bp->pf.hwrm_cmd_req_dma_addr[0]);
	req.req_buf_page_addr1 = cpu_to_le64(bp->pf.hwrm_cmd_req_dma_addr[1]);
	req.req_buf_page_addr2 = cpu_to_le64(bp->pf.hwrm_cmd_req_dma_addr[2]);
	req.req_buf_page_addr3 = cpu_to_le64(bp->pf.hwrm_cmd_req_dma_addr[3]);

	return hwrm_send_message(bp, &req, sizeof(req), HWRM_CMD_TIMEOUT);
}

/* only call by PF to reserve resources for VF */
420
static int bnxt_hwrm_func_cfg(struct bnxt *bp, int num_vfs)
421
422
423
{
	u32 rc = 0, mtu, i;
	u16 vf_tx_rings, vf_rx_rings, vf_cp_rings, vf_stat_ctx, vf_vnics;
424
	u16 vf_ring_grps;
425
426
	struct hwrm_func_cfg_input req = {0};
	struct bnxt_pf_info *pf = &bp->pf;
427
	int total_vf_tx_rings = 0;
428
429
430
431

	bnxt_hwrm_cmd_hdr_init(bp, &req, HWRM_FUNC_CFG, -1, -1);

	/* Remaining rings are distributed equally amongs VF's for now */
432
	vf_cp_rings = (pf->max_cp_rings - bp->cp_nr_rings) / num_vfs;
433
	vf_stat_ctx = (pf->max_stat_ctxs - bp->num_stat_ctxs) / num_vfs;
434
	if (bp->flags & BNXT_FLAG_AGG_RINGS)
435
436
		vf_rx_rings = (pf->max_rx_rings - bp->rx_nr_rings * 2) /
			      num_vfs;
437
	else
438
		vf_rx_rings = (pf->max_rx_rings - bp->rx_nr_rings) / num_vfs;
439
	vf_ring_grps = (bp->pf.max_hw_ring_grps - bp->rx_nr_rings) / num_vfs;
440
	vf_tx_rings = (pf->max_tx_rings - bp->tx_nr_rings) / num_vfs;
441
442
	vf_vnics = (pf->max_vnics - bp->nr_vnics) / num_vfs;
	vf_vnics = min_t(u16, vf_vnics, vf_rx_rings);
443
444
445
446
447
448
449
450
451

	req.enables = cpu_to_le32(FUNC_CFG_REQ_ENABLES_MTU |
				  FUNC_CFG_REQ_ENABLES_MRU |
				  FUNC_CFG_REQ_ENABLES_NUM_RSSCOS_CTXS |
				  FUNC_CFG_REQ_ENABLES_NUM_STAT_CTXS |
				  FUNC_CFG_REQ_ENABLES_NUM_CMPL_RINGS |
				  FUNC_CFG_REQ_ENABLES_NUM_TX_RINGS |
				  FUNC_CFG_REQ_ENABLES_NUM_RX_RINGS |
				  FUNC_CFG_REQ_ENABLES_NUM_L2_CTXS |
452
453
				  FUNC_CFG_REQ_ENABLES_NUM_VNICS |
				  FUNC_CFG_REQ_ENABLES_NUM_HW_RING_GRPS);
454
455
456
457
458
459
460
461
462

	mtu = bp->dev->mtu + ETH_HLEN + ETH_FCS_LEN + VLAN_HLEN;
	req.mru = cpu_to_le16(mtu);
	req.mtu = cpu_to_le16(mtu);

	req.num_rsscos_ctxs = cpu_to_le16(1);
	req.num_cmpl_rings = cpu_to_le16(vf_cp_rings);
	req.num_tx_rings = cpu_to_le16(vf_tx_rings);
	req.num_rx_rings = cpu_to_le16(vf_rx_rings);
463
	req.num_hw_ring_grps = cpu_to_le16(vf_ring_grps);
464
465
466
467
468
469
470
	req.num_l2_ctxs = cpu_to_le16(4);

	req.num_vnics = cpu_to_le16(vf_vnics);
	/* FIXME spec currently uses 1 bit for stats ctx */
	req.num_stat_ctxs = cpu_to_le16(vf_stat_ctx);

	mutex_lock(&bp->hwrm_cmd_lock);
471
	for (i = 0; i < num_vfs; i++) {
472
473
		int vf_tx_rsvd = vf_tx_rings;

474
		req.fid = cpu_to_le16(pf->first_vf_id + i);
475
476
477
478
		rc = _hwrm_send_message(bp, &req, sizeof(req),
					HWRM_CMD_TIMEOUT);
		if (rc)
			break;
479
		pf->active_vfs = i + 1;
480
		pf->vf[i].fw_fid = le16_to_cpu(req.fid);
481
482
483
484
485
		rc = __bnxt_hwrm_get_tx_rings(bp, pf->vf[i].fw_fid,
					      &vf_tx_rsvd);
		if (rc)
			break;
		total_vf_tx_rings += vf_tx_rsvd;
486
487
488
	}
	mutex_unlock(&bp->hwrm_cmd_lock);
	if (!rc) {
489
		pf->max_tx_rings -= total_vf_tx_rings;
490
		pf->max_rx_rings -= vf_rx_rings * num_vfs;
491
		pf->max_hw_ring_grps -= vf_ring_grps * num_vfs;
492
493
494
495
		pf->max_cp_rings -= vf_cp_rings * num_vfs;
		pf->max_rsscos_ctxs -= num_vfs;
		pf->max_stat_ctxs -= vf_stat_ctx * num_vfs;
		pf->max_vnics -= vf_vnics * num_vfs;
496
497
498
499
500
501
502
503
504
	}
	return rc;
}

static int bnxt_sriov_enable(struct bnxt *bp, int *num_vfs)
{
	int rc = 0, vfs_supported;
	int min_rx_rings, min_tx_rings, min_rss_ctxs;
	int tx_ok = 0, rx_ok = 0, rss_ok = 0;
505
	int avail_cp, avail_stat;
506
507
508
509
510
511
512

	/* Check if we can enable requested num of vf's. At a mininum
	 * we require 1 RX 1 TX rings for each VF. In this minimum conf
	 * features like TPA will not be available.
	 */
	vfs_supported = *num_vfs;

513
514
515
516
	avail_cp = bp->pf.max_cp_rings - bp->cp_nr_rings;
	avail_stat = bp->pf.max_stat_ctxs - bp->num_stat_ctxs;
	avail_cp = min_t(int, avail_cp, avail_stat);

517
518
519
520
521
522
523
524
525
526
527
528
529
530
	while (vfs_supported) {
		min_rx_rings = vfs_supported;
		min_tx_rings = vfs_supported;
		min_rss_ctxs = vfs_supported;

		if (bp->flags & BNXT_FLAG_AGG_RINGS) {
			if (bp->pf.max_rx_rings - bp->rx_nr_rings * 2 >=
			    min_rx_rings)
				rx_ok = 1;
		} else {
			if (bp->pf.max_rx_rings - bp->rx_nr_rings >=
			    min_rx_rings)
				rx_ok = 1;
		}
531
532
		if (bp->pf.max_vnics - bp->nr_vnics < min_rx_rings ||
		    avail_cp < min_rx_rings)
533
			rx_ok = 0;
534

535
536
		if (bp->pf.max_tx_rings - bp->tx_nr_rings >= min_tx_rings &&
		    avail_cp >= min_tx_rings)
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
			tx_ok = 1;

		if (bp->pf.max_rsscos_ctxs - bp->rsscos_nr_ctxs >= min_rss_ctxs)
			rss_ok = 1;

		if (tx_ok && rx_ok && rss_ok)
			break;

		vfs_supported--;
	}

	if (!vfs_supported) {
		netdev_err(bp->dev, "Cannot enable VF's as all resources are used by PF\n");
		return -EINVAL;
	}

	if (vfs_supported != *num_vfs) {
		netdev_info(bp->dev, "Requested VFs %d, can enable %d\n",
			    *num_vfs, vfs_supported);
		*num_vfs = vfs_supported;
	}

	rc = bnxt_alloc_vf_resources(bp, *num_vfs);
	if (rc)
		goto err_out1;

	/* Reserve resources for VFs */
564
	rc = bnxt_hwrm_func_cfg(bp, *num_vfs);
565
566
567
568
569
570
571
572
	if (rc)
		goto err_out2;

	/* Register buffers for VFs */
	rc = bnxt_hwrm_func_buf_rgtr(bp);
	if (rc)
		goto err_out2;

573
574
	bnxt_ulp_sriov_cfg(bp, *num_vfs);

575
576
577
578
579
580
581
582
	rc = pci_enable_sriov(bp->pdev, *num_vfs);
	if (rc)
		goto err_out2;

	return 0;

err_out2:
	/* Free the resources reserved for various VF's */
583
	bnxt_hwrm_func_vf_resource_free(bp, *num_vfs);
584
585
586
587
588
589
590
591
592

err_out1:
	bnxt_free_vf_resources(bp);

	return rc;
}

void bnxt_sriov_disable(struct bnxt *bp)
{
593
	u16 num_vfs = pci_num_vf(bp->pdev);
594

595
596
	if (!num_vfs)
		return;
597

598
599
600
601
	/* synchronize VF and VF-rep create and destroy */
	mutex_lock(&bp->sriov_lock);
	bnxt_vf_reps_destroy(bp);

602
	if (pci_vfs_assigned(bp->pdev)) {
603
		bnxt_hwrm_fwd_async_event_cmpl(
604
			bp, NULL, ASYNC_EVENT_CMPL_EVENT_ID_PF_DRVR_UNLOAD);
605
606
607
608
609
610
611
		netdev_warn(bp->dev, "Unable to free %d VFs because some are assigned to VMs.\n",
			    num_vfs);
	} else {
		pci_disable_sriov(bp->pdev);
		/* Free the HW resources reserved for various VF's */
		bnxt_hwrm_func_vf_resource_free(bp, num_vfs);
	}
612
	mutex_unlock(&bp->sriov_lock);
613
614
615
616

	bnxt_free_vf_resources(bp);

	bp->pf.active_vfs = 0;
617
	/* Reclaim all resources for the PF. */
618
619
620
	rtnl_lock();
	bnxt_restore_pf_fw_resources(bp);
	rtnl_unlock();
621
622

	bnxt_ulp_sriov_cfg(bp, 0);
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
}

int bnxt_sriov_configure(struct pci_dev *pdev, int num_vfs)
{
	struct net_device *dev = pci_get_drvdata(pdev);
	struct bnxt *bp = netdev_priv(dev);

	if (!(bp->flags & BNXT_FLAG_USING_MSIX)) {
		netdev_warn(dev, "Not allow SRIOV if the irq mode is not MSIX\n");
		return 0;
	}

	rtnl_lock();
	if (!netif_running(dev)) {
		netdev_warn(dev, "Reject SRIOV config request since if is down!\n");
		rtnl_unlock();
		return 0;
	}
	bp->sriov_cfg = true;
	rtnl_unlock();
643
644
645
646
647

	if (pci_vfs_assigned(bp->pdev)) {
		netdev_warn(dev, "Unable to configure SRIOV since some VFs are assigned to VMs.\n");
		num_vfs = 0;
		goto sriov_cfg_exit;
648
649
650
	}

	/* Check if enabled VFs is same as requested */
651
652
653
654
655
656
657
	if (num_vfs && num_vfs == bp->pf.active_vfs)
		goto sriov_cfg_exit;

	/* if there are previous existing VFs, clean them up */
	bnxt_sriov_disable(bp);
	if (!num_vfs)
		goto sriov_cfg_exit;
658
659
660

	bnxt_sriov_enable(bp, &num_vfs);

661
sriov_cfg_exit:
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
	bp->sriov_cfg = false;
	wake_up(&bp->sriov_cfg_wait);

	return num_vfs;
}

static int bnxt_hwrm_fwd_resp(struct bnxt *bp, struct bnxt_vf_info *vf,
			      void *encap_resp, __le64 encap_resp_addr,
			      __le16 encap_resp_cpr, u32 msg_size)
{
	int rc = 0;
	struct hwrm_fwd_resp_input req = {0};
	struct hwrm_fwd_resp_output *resp = bp->hwrm_cmd_resp_addr;

	bnxt_hwrm_cmd_hdr_init(bp, &req, HWRM_FWD_RESP, -1, -1);

	/* Set the new target id */
	req.target_id = cpu_to_le16(vf->fw_fid);
680
	req.encap_resp_target_id = cpu_to_le16(vf->fw_fid);
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
	req.encap_resp_len = cpu_to_le16(msg_size);
	req.encap_resp_addr = encap_resp_addr;
	req.encap_resp_cmpl_ring = encap_resp_cpr;
	memcpy(req.encap_resp, encap_resp, msg_size);

	mutex_lock(&bp->hwrm_cmd_lock);
	rc = _hwrm_send_message(bp, &req, sizeof(req), HWRM_CMD_TIMEOUT);

	if (rc) {
		netdev_err(bp->dev, "hwrm_fwd_resp failed. rc:%d\n", rc);
		goto fwd_resp_exit;
	}

	if (resp->error_code) {
		netdev_err(bp->dev, "hwrm_fwd_resp error %d\n",
			   resp->error_code);
		rc = -1;
	}

fwd_resp_exit:
	mutex_unlock(&bp->hwrm_cmd_lock);
	return rc;
}

static int bnxt_hwrm_fwd_err_resp(struct bnxt *bp, struct bnxt_vf_info *vf,
				  u32 msg_size)
{
	int rc = 0;
	struct hwrm_reject_fwd_resp_input req = {0};
	struct hwrm_reject_fwd_resp_output *resp = bp->hwrm_cmd_resp_addr;

	bnxt_hwrm_cmd_hdr_init(bp, &req, HWRM_REJECT_FWD_RESP, -1, -1);
	/* Set the new target id */
	req.target_id = cpu_to_le16(vf->fw_fid);
715
	req.encap_resp_target_id = cpu_to_le16(vf->fw_fid);
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
	memcpy(req.encap_request, vf->hwrm_cmd_req_addr, msg_size);

	mutex_lock(&bp->hwrm_cmd_lock);
	rc = _hwrm_send_message(bp, &req, sizeof(req), HWRM_CMD_TIMEOUT);

	if (rc) {
		netdev_err(bp->dev, "hwrm_fwd_err_resp failed. rc:%d\n", rc);
		goto fwd_err_resp_exit;
	}

	if (resp->error_code) {
		netdev_err(bp->dev, "hwrm_fwd_err_resp error %d\n",
			   resp->error_code);
		rc = -1;
	}

fwd_err_resp_exit:
	mutex_unlock(&bp->hwrm_cmd_lock);
	return rc;
}

static int bnxt_hwrm_exec_fwd_resp(struct bnxt *bp, struct bnxt_vf_info *vf,
				   u32 msg_size)
{
	int rc = 0;
	struct hwrm_exec_fwd_resp_input req = {0};
	struct hwrm_exec_fwd_resp_output *resp = bp->hwrm_cmd_resp_addr;

	bnxt_hwrm_cmd_hdr_init(bp, &req, HWRM_EXEC_FWD_RESP, -1, -1);
	/* Set the new target id */
	req.target_id = cpu_to_le16(vf->fw_fid);
747
	req.encap_resp_target_id = cpu_to_le16(vf->fw_fid);
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
	memcpy(req.encap_request, vf->hwrm_cmd_req_addr, msg_size);

	mutex_lock(&bp->hwrm_cmd_lock);
	rc = _hwrm_send_message(bp, &req, sizeof(req), HWRM_CMD_TIMEOUT);

	if (rc) {
		netdev_err(bp->dev, "hwrm_exec_fw_resp failed. rc:%d\n", rc);
		goto exec_fwd_resp_exit;
	}

	if (resp->error_code) {
		netdev_err(bp->dev, "hwrm_exec_fw_resp error %d\n",
			   resp->error_code);
		rc = -1;
	}

exec_fwd_resp_exit:
	mutex_unlock(&bp->hwrm_cmd_lock);
	return rc;
}

static int bnxt_vf_validate_set_mac(struct bnxt *bp, struct bnxt_vf_info *vf)
{
	u32 msg_size = sizeof(struct hwrm_cfa_l2_filter_alloc_input);
	struct hwrm_cfa_l2_filter_alloc_input *req =
		(struct hwrm_cfa_l2_filter_alloc_input *)vf->hwrm_cmd_req_addr;

	if (!is_valid_ether_addr(vf->mac_addr) ||
	    ether_addr_equal((const u8 *)req->l2_addr, vf->mac_addr))
		return bnxt_hwrm_exec_fwd_resp(bp, vf, msg_size);
	else
		return bnxt_hwrm_fwd_err_resp(bp, vf, msg_size);
}

static int bnxt_vf_set_link(struct bnxt *bp, struct bnxt_vf_info *vf)
{
	int rc = 0;

	if (!(vf->flags & BNXT_VF_LINK_FORCED)) {
		/* real link */
		rc = bnxt_hwrm_exec_fwd_resp(
			bp, vf, sizeof(struct hwrm_port_phy_qcfg_input));
	} else {
		struct hwrm_port_phy_qcfg_output phy_qcfg_resp;
		struct hwrm_port_phy_qcfg_input *phy_qcfg_req;

		phy_qcfg_req =
		(struct hwrm_port_phy_qcfg_input *)vf->hwrm_cmd_req_addr;
		mutex_lock(&bp->hwrm_cmd_lock);
		memcpy(&phy_qcfg_resp, &bp->link_info.phy_qcfg_resp,
		       sizeof(phy_qcfg_resp));
		mutex_unlock(&bp->hwrm_cmd_lock);
		phy_qcfg_resp.seq_id = phy_qcfg_req->seq_id;

		if (vf->flags & BNXT_VF_LINK_UP) {
			/* if physical link is down, force link up on VF */
804
805
			if (phy_qcfg_resp.link !=
			    PORT_PHY_QCFG_RESP_LINK_LINK) {
806
807
				phy_qcfg_resp.link =
					PORT_PHY_QCFG_RESP_LINK_LINK;
808
809
				phy_qcfg_resp.link_speed = cpu_to_le16(
					PORT_PHY_QCFG_RESP_LINK_SPEED_10GB);
810
811
812
813
				phy_qcfg_resp.duplex_cfg =
					PORT_PHY_QCFG_RESP_DUPLEX_CFG_FULL;
				phy_qcfg_resp.duplex_state =
					PORT_PHY_QCFG_RESP_DUPLEX_STATE_FULL;
814
815
816
817
818
819
820
821
				phy_qcfg_resp.pause =
					(PORT_PHY_QCFG_RESP_PAUSE_TX |
					 PORT_PHY_QCFG_RESP_PAUSE_RX);
			}
		} else {
			/* force link down */
			phy_qcfg_resp.link = PORT_PHY_QCFG_RESP_LINK_NO_LINK;
			phy_qcfg_resp.link_speed = 0;
822
823
			phy_qcfg_resp.duplex_state =
				PORT_PHY_QCFG_RESP_DUPLEX_STATE_HALF;
824
825
826
827
828
829
830
831
832
833
834
835
836
			phy_qcfg_resp.pause = 0;
		}
		rc = bnxt_hwrm_fwd_resp(bp, vf, &phy_qcfg_resp,
					phy_qcfg_req->resp_addr,
					phy_qcfg_req->cmpl_ring,
					sizeof(phy_qcfg_resp));
	}
	return rc;
}

static int bnxt_vf_req_validate_snd(struct bnxt *bp, struct bnxt_vf_info *vf)
{
	int rc = 0;
837
838
	struct input *encap_req = vf->hwrm_cmd_req_addr;
	u32 req_type = le16_to_cpu(encap_req->req_type);
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874

	switch (req_type) {
	case HWRM_CFA_L2_FILTER_ALLOC:
		rc = bnxt_vf_validate_set_mac(bp, vf);
		break;
	case HWRM_FUNC_CFG:
		/* TODO Validate if VF is allowed to change mac address,
		 * mtu, num of rings etc
		 */
		rc = bnxt_hwrm_exec_fwd_resp(
			bp, vf, sizeof(struct hwrm_func_cfg_input));
		break;
	case HWRM_PORT_PHY_QCFG:
		rc = bnxt_vf_set_link(bp, vf);
		break;
	default:
		break;
	}
	return rc;
}

void bnxt_hwrm_exec_fwd_req(struct bnxt *bp)
{
	u32 i = 0, active_vfs = bp->pf.active_vfs, vf_id;

	/* Scan through VF's and process commands */
	while (1) {
		vf_id = find_next_bit(bp->pf.vf_event_bmap, active_vfs, i);
		if (vf_id >= active_vfs)
			break;

		clear_bit(vf_id, bp->pf.vf_event_bmap);
		bnxt_vf_req_validate_snd(bp, &bp->pf.vf[vf_id]);
		i = vf_id + 1;
	}
}
875
876
877
878
879
880
881
882
883
884
885
886
887

void bnxt_update_vf_mac(struct bnxt *bp)
{
	struct hwrm_func_qcaps_input req = {0};
	struct hwrm_func_qcaps_output *resp = bp->hwrm_cmd_resp_addr;

	bnxt_hwrm_cmd_hdr_init(bp, &req, HWRM_FUNC_QCAPS, -1, -1);
	req.fid = cpu_to_le16(0xffff);

	mutex_lock(&bp->hwrm_cmd_lock);
	if (_hwrm_send_message(bp, &req, sizeof(req), HWRM_CMD_TIMEOUT))
		goto update_vf_mac_exit;

888
889
890
891
892
893
894
	/* Store MAC address from the firmware.  There are 2 cases:
	 * 1. MAC address is valid.  It is assigned from the PF and we
	 *    need to override the current VF MAC address with it.
	 * 2. MAC address is zero.  The VF will use a random MAC address by
	 *    default but the stored zero MAC will allow the VF user to change
	 *    the random MAC address using ndo_set_mac_address() if he wants.
	 */
895
896
	if (!ether_addr_equal(resp->mac_address, bp->vf.mac_addr))
		memcpy(bp->vf.mac_addr, resp->mac_address, ETH_ALEN);
897
898
899
900

	/* overwrite netdev dev_addr with admin VF MAC */
	if (is_valid_ether_addr(bp->vf.mac_addr))
		memcpy(bp->dev->dev_addr, bp->vf.mac_addr, ETH_ALEN);
901
902
903
904
update_vf_mac_exit:
	mutex_unlock(&bp->hwrm_cmd_lock);
}

905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
int bnxt_approve_mac(struct bnxt *bp, u8 *mac)
{
	struct hwrm_func_vf_cfg_input req = {0};
	int rc = 0;

	if (!BNXT_VF(bp))
		return 0;

	if (bp->hwrm_spec_code < 0x10202) {
		if (is_valid_ether_addr(bp->vf.mac_addr))
			rc = -EADDRNOTAVAIL;
		goto mac_done;
	}
	bnxt_hwrm_cmd_hdr_init(bp, &req, HWRM_FUNC_VF_CFG, -1, -1);
	req.enables = cpu_to_le32(FUNC_VF_CFG_REQ_ENABLES_DFLT_MAC_ADDR);
	memcpy(req.dflt_mac_addr, mac, ETH_ALEN);
	rc = hwrm_send_message(bp, &req, sizeof(req), HWRM_CMD_TIMEOUT);
mac_done:
	if (rc) {
		rc = -EADDRNOTAVAIL;
		netdev_warn(bp->dev, "VF MAC address %pM not approved by the PF\n",
			    mac);
	}
	return rc;
}
930
931
932
933
934
935
936
937
#else

void bnxt_sriov_disable(struct bnxt *bp)
{
}

void bnxt_hwrm_exec_fwd_req(struct bnxt *bp)
{
938
939
940
941
942
	netdev_err(bp->dev, "Invalid VF message received when SRIOV is not enable\n");
}

void bnxt_update_vf_mac(struct bnxt *bp)
{
943
}
944
945
946
947
948

int bnxt_approve_mac(struct bnxt *bp, u8 *mac)
{
	return 0;
}
949
#endif