i40e_main.c 321.0 KB
Newer Older
J
Jesse Brandeburg 已提交
1 2 3
/*******************************************************************************
 *
 * Intel Ethernet Controller XL710 Family Linux Driver
4
 * Copyright(c) 2013 - 2016 Intel Corporation.
J
Jesse Brandeburg 已提交
5 6 7 8 9 10 11 12 13 14
 *
 * This program is free software; you can redistribute it and/or modify it
 * under the terms and conditions of the GNU General Public License,
 * version 2, as published by the Free Software Foundation.
 *
 * This program is distributed in the hope it will be useful, but WITHOUT
 * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
 * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
 * more details.
 *
G
Greg Rose 已提交
15 16
 * You should have received a copy of the GNU General Public License along
 * with this program.  If not, see <http://www.gnu.org/licenses/>.
J
Jesse Brandeburg 已提交
17 18 19 20 21 22 23 24 25 26
 *
 * The full GNU General Public License is included in this distribution in
 * the file called "COPYING".
 *
 * Contact Information:
 * e1000-devel Mailing List <e1000-devel@lists.sourceforge.net>
 * Intel Corporation, 5200 N.E. Elam Young Parkway, Hillsboro, OR 97124-6497
 *
 ******************************************************************************/

27 28 29 30
#include <linux/etherdevice.h>
#include <linux/of_net.h>
#include <linux/pci.h>

J
Jesse Brandeburg 已提交
31 32
/* Local includes */
#include "i40e.h"
33
#include "i40e_diag.h"
34
#include <net/udp_tunnel.h>
J
Jesse Brandeburg 已提交
35 36 37 38 39 40 41

const char i40e_driver_name[] = "i40e";
static const char i40e_driver_string[] =
			"Intel(R) Ethernet Connection XL710 Network Driver";

#define DRV_KERN "-k"

42
#define DRV_VERSION_MAJOR 1
43
#define DRV_VERSION_MINOR 6
44
#define DRV_VERSION_BUILD 21
J
Jesse Brandeburg 已提交
45 46 47 48
#define DRV_VERSION __stringify(DRV_VERSION_MAJOR) "." \
	     __stringify(DRV_VERSION_MINOR) "." \
	     __stringify(DRV_VERSION_BUILD)    DRV_KERN
const char i40e_driver_version_str[] = DRV_VERSION;
J
Jesse Brandeburg 已提交
49
static const char i40e_copyright[] = "Copyright (c) 2013 - 2014 Intel Corporation.";
J
Jesse Brandeburg 已提交
50 51 52 53 54 55

/* a bit of forward declarations */
static void i40e_vsi_reinit_locked(struct i40e_vsi *vsi);
static void i40e_handle_reset_warning(struct i40e_pf *pf);
static int i40e_add_vsi(struct i40e_vsi *vsi);
static int i40e_add_veb(struct i40e_veb *veb, struct i40e_vsi *vsi);
56
static int i40e_setup_pf_switch(struct i40e_pf *pf, bool reinit);
J
Jesse Brandeburg 已提交
57 58 59
static int i40e_setup_misc_vector(struct i40e_pf *pf);
static void i40e_determine_queue_usage(struct i40e_pf *pf);
static int i40e_setup_pf_filter_control(struct i40e_pf *pf);
60
static void i40e_fdir_sb_setup(struct i40e_pf *pf);
N
Neerav Parikh 已提交
61
static int i40e_veb_get_bw_info(struct i40e_veb *veb);
J
Jesse Brandeburg 已提交
62 63 64 65 66 67 68 69

/* i40e_pci_tbl - PCI Device ID Table
 *
 * Last entry must be all 0s
 *
 * { Vendor ID, Device ID, SubVendor ID, SubDevice ID,
 *   Class, Class Mask, private data (not used) }
 */
70
static const struct pci_device_id i40e_pci_tbl[] = {
71 72 73 74 75 76 77
	{PCI_VDEVICE(INTEL, I40E_DEV_ID_SFP_XL710), 0},
	{PCI_VDEVICE(INTEL, I40E_DEV_ID_QEMU), 0},
	{PCI_VDEVICE(INTEL, I40E_DEV_ID_KX_B), 0},
	{PCI_VDEVICE(INTEL, I40E_DEV_ID_KX_C), 0},
	{PCI_VDEVICE(INTEL, I40E_DEV_ID_QSFP_A), 0},
	{PCI_VDEVICE(INTEL, I40E_DEV_ID_QSFP_B), 0},
	{PCI_VDEVICE(INTEL, I40E_DEV_ID_QSFP_C), 0},
M
Mitch Williams 已提交
78
	{PCI_VDEVICE(INTEL, I40E_DEV_ID_10G_BASE_T), 0},
79
	{PCI_VDEVICE(INTEL, I40E_DEV_ID_10G_BASE_T4), 0},
80
	{PCI_VDEVICE(INTEL, I40E_DEV_ID_20G_KR2), 0},
81 82
	{PCI_VDEVICE(INTEL, I40E_DEV_ID_KX_X722), 0},
	{PCI_VDEVICE(INTEL, I40E_DEV_ID_QSFP_X722), 0},
83 84 85
	{PCI_VDEVICE(INTEL, I40E_DEV_ID_SFP_X722), 0},
	{PCI_VDEVICE(INTEL, I40E_DEV_ID_1G_BASE_T_X722), 0},
	{PCI_VDEVICE(INTEL, I40E_DEV_ID_10G_BASE_T_X722), 0},
86
	{PCI_VDEVICE(INTEL, I40E_DEV_ID_SFP_I_X722), 0},
87 88
	{PCI_VDEVICE(INTEL, I40E_DEV_ID_20G_KR2), 0},
	{PCI_VDEVICE(INTEL, I40E_DEV_ID_20G_KR2_A), 0},
J
Jesse Brandeburg 已提交
89 90 91 92 93 94 95
	/* required last entry */
	{0, }
};
MODULE_DEVICE_TABLE(pci, i40e_pci_tbl);

#define I40E_MAX_VF_COUNT 128
static int debug = -1;
96 97
module_param(debug, uint, 0);
MODULE_PARM_DESC(debug, "Debug level (0=none,...,16=all), Debug mask (0x8XXXXXXX)");
J
Jesse Brandeburg 已提交
98 99 100 101 102 103

MODULE_AUTHOR("Intel Corporation, <e1000-devel@lists.sourceforge.net>");
MODULE_DESCRIPTION("Intel(R) Ethernet Connection XL710 Network Driver");
MODULE_LICENSE("GPL");
MODULE_VERSION(DRV_VERSION);

104 105
static struct workqueue_struct *i40e_wq;

J
Jesse Brandeburg 已提交
106 107 108 109 110 111 112 113 114 115 116 117 118 119 120
/**
 * i40e_allocate_dma_mem_d - OS specific memory alloc for shared code
 * @hw:   pointer to the HW structure
 * @mem:  ptr to mem struct to fill out
 * @size: size of memory requested
 * @alignment: what to align the allocation to
 **/
int i40e_allocate_dma_mem_d(struct i40e_hw *hw, struct i40e_dma_mem *mem,
			    u64 size, u32 alignment)
{
	struct i40e_pf *pf = (struct i40e_pf *)hw->back;

	mem->size = ALIGN(size, alignment);
	mem->va = dma_zalloc_coherent(&pf->pdev->dev, mem->size,
				      &mem->pa, GFP_KERNEL);
J
Jesse Brandeburg 已提交
121 122
	if (!mem->va)
		return -ENOMEM;
J
Jesse Brandeburg 已提交
123

J
Jesse Brandeburg 已提交
124
	return 0;
J
Jesse Brandeburg 已提交
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
}

/**
 * i40e_free_dma_mem_d - OS specific memory free for shared code
 * @hw:   pointer to the HW structure
 * @mem:  ptr to mem struct to free
 **/
int i40e_free_dma_mem_d(struct i40e_hw *hw, struct i40e_dma_mem *mem)
{
	struct i40e_pf *pf = (struct i40e_pf *)hw->back;

	dma_free_coherent(&pf->pdev->dev, mem->size, mem->va, mem->pa);
	mem->va = NULL;
	mem->pa = 0;
	mem->size = 0;

	return 0;
}

/**
 * i40e_allocate_virt_mem_d - OS specific memory alloc for shared code
 * @hw:   pointer to the HW structure
 * @mem:  ptr to mem struct to fill out
 * @size: size of memory requested
 **/
int i40e_allocate_virt_mem_d(struct i40e_hw *hw, struct i40e_virt_mem *mem,
			     u32 size)
{
	mem->size = size;
	mem->va = kzalloc(size, GFP_KERNEL);

J
Jesse Brandeburg 已提交
156 157
	if (!mem->va)
		return -ENOMEM;
J
Jesse Brandeburg 已提交
158

J
Jesse Brandeburg 已提交
159
	return 0;
J
Jesse Brandeburg 已提交
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
}

/**
 * i40e_free_virt_mem_d - OS specific memory free for shared code
 * @hw:   pointer to the HW structure
 * @mem:  ptr to mem struct to free
 **/
int i40e_free_virt_mem_d(struct i40e_hw *hw, struct i40e_virt_mem *mem)
{
	/* it's ok to kfree a NULL pointer */
	kfree(mem->va);
	mem->va = NULL;
	mem->size = 0;

	return 0;
}

/**
 * i40e_get_lump - find a lump of free generic resource
 * @pf: board private structure
 * @pile: the pile of resource to search
 * @needed: the number of items needed
 * @id: an owner id to stick on the items assigned
 *
 * Returns the base item index of the lump, or negative for error
 *
 * The search_hint trick and lack of advanced fit-finding only work
 * because we're highly likely to have all the same size lump requests.
 * Linear search time and any fragmentation should be minimal.
 **/
static int i40e_get_lump(struct i40e_pf *pf, struct i40e_lump_tracking *pile,
			 u16 needed, u16 id)
{
	int ret = -ENOMEM;
194
	int i, j;
J
Jesse Brandeburg 已提交
195 196 197 198 199 200 201 202 203 204

	if (!pile || needed == 0 || id >= I40E_PILE_VALID_BIT) {
		dev_info(&pf->pdev->dev,
			 "param err: pile=%p needed=%d id=0x%04x\n",
			 pile, needed, id);
		return -EINVAL;
	}

	/* start the linear search with an imperfect hint */
	i = pile->search_hint;
205
	while (i < pile->num_entries) {
J
Jesse Brandeburg 已提交
206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223
		/* skip already allocated entries */
		if (pile->list[i] & I40E_PILE_VALID_BIT) {
			i++;
			continue;
		}

		/* do we have enough in this lump? */
		for (j = 0; (j < needed) && ((i+j) < pile->num_entries); j++) {
			if (pile->list[i+j] & I40E_PILE_VALID_BIT)
				break;
		}

		if (j == needed) {
			/* there was enough, so assign it to the requestor */
			for (j = 0; j < needed; j++)
				pile->list[i+j] = id | I40E_PILE_VALID_BIT;
			ret = i;
			pile->search_hint = i + j;
224
			break;
J
Jesse Brandeburg 已提交
225
		}
J
Jesse Brandeburg 已提交
226 227 228

		/* not enough, so skip over it and continue looking */
		i += j;
J
Jesse Brandeburg 已提交
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
	}

	return ret;
}

/**
 * i40e_put_lump - return a lump of generic resource
 * @pile: the pile of resource to search
 * @index: the base item index
 * @id: the owner id of the items assigned
 *
 * Returns the count of items in the lump
 **/
static int i40e_put_lump(struct i40e_lump_tracking *pile, u16 index, u16 id)
{
	int valid_id = (id | I40E_PILE_VALID_BIT);
	int count = 0;
	int i;

	if (!pile || index >= pile->num_entries)
		return -EINVAL;

	for (i = index;
	     i < pile->num_entries && pile->list[i] == valid_id;
	     i++) {
		pile->list[i] = 0;
		count++;
	}

	if (count && index < pile->search_hint)
		pile->search_hint = index;

	return count;
}

264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279
/**
 * i40e_find_vsi_from_id - searches for the vsi with the given id
 * @pf - the pf structure to search for the vsi
 * @id - id of the vsi it is searching for
 **/
struct i40e_vsi *i40e_find_vsi_from_id(struct i40e_pf *pf, u16 id)
{
	int i;

	for (i = 0; i < pf->num_alloc_vsi; i++)
		if (pf->vsi[i] && (pf->vsi[i]->id == id))
			return pf->vsi[i];

	return NULL;
}

J
Jesse Brandeburg 已提交
280 281 282 283 284 285
/**
 * i40e_service_event_schedule - Schedule the service task to wake up
 * @pf: board private structure
 *
 * If not already scheduled, this puts the task into the work queue
 **/
286
void i40e_service_event_schedule(struct i40e_pf *pf)
J
Jesse Brandeburg 已提交
287 288 289 290
{
	if (!test_bit(__I40E_DOWN, &pf->state) &&
	    !test_bit(__I40E_RESET_RECOVERY_PENDING, &pf->state) &&
	    !test_and_set_bit(__I40E_SERVICE_SCHED, &pf->state))
291
		queue_work(i40e_wq, &pf->service_task);
J
Jesse Brandeburg 已提交
292 293 294 295 296 297 298 299 300 301
}

/**
 * i40e_tx_timeout - Respond to a Tx Hang
 * @netdev: network interface device structure
 *
 * If any port has noticed a Tx timeout, it is likely that the whole
 * device is munged, not just the one netdev port, so go for the full
 * reset.
 **/
302 303 304
#ifdef I40E_FCOE
void i40e_tx_timeout(struct net_device *netdev)
#else
J
Jesse Brandeburg 已提交
305
static void i40e_tx_timeout(struct net_device *netdev)
306
#endif
J
Jesse Brandeburg 已提交
307 308 309 310
{
	struct i40e_netdev_priv *np = netdev_priv(netdev);
	struct i40e_vsi *vsi = np->vsi;
	struct i40e_pf *pf = vsi->back;
311 312 313
	struct i40e_ring *tx_ring = NULL;
	unsigned int i, hung_queue = 0;
	u32 head, val;
J
Jesse Brandeburg 已提交
314 315 316

	pf->tx_timeout_count++;

317 318 319 320 321 322
	/* find the stopped queue the same way the stack does */
	for (i = 0; i < netdev->num_tx_queues; i++) {
		struct netdev_queue *q;
		unsigned long trans_start;

		q = netdev_get_tx_queue(netdev, i);
F
Florian Westphal 已提交
323
		trans_start = q->trans_start;
324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346
		if (netif_xmit_stopped(q) &&
		    time_after(jiffies,
			       (trans_start + netdev->watchdog_timeo))) {
			hung_queue = i;
			break;
		}
	}

	if (i == netdev->num_tx_queues) {
		netdev_info(netdev, "tx_timeout: no netdev hung queue found\n");
	} else {
		/* now that we have an index, find the tx_ring struct */
		for (i = 0; i < vsi->num_queue_pairs; i++) {
			if (vsi->tx_rings[i] && vsi->tx_rings[i]->desc) {
				if (hung_queue ==
				    vsi->tx_rings[i]->queue_index) {
					tx_ring = vsi->tx_rings[i];
					break;
				}
			}
		}
	}

J
Jesse Brandeburg 已提交
347
	if (time_after(jiffies, (pf->tx_timeout_last_recovery + HZ*20)))
348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368
		pf->tx_timeout_recovery_level = 1;  /* reset after some time */
	else if (time_before(jiffies,
		      (pf->tx_timeout_last_recovery + netdev->watchdog_timeo)))
		return;   /* don't do any new action before the next timeout */

	if (tx_ring) {
		head = i40e_get_head(tx_ring);
		/* Read interrupt register */
		if (pf->flags & I40E_FLAG_MSIX_ENABLED)
			val = rd32(&pf->hw,
			     I40E_PFINT_DYN_CTLN(tx_ring->q_vector->v_idx +
						tx_ring->vsi->base_vector - 1));
		else
			val = rd32(&pf->hw, I40E_PFINT_DYN_CTL0);

		netdev_info(netdev, "tx_timeout: VSI_seid: %d, Q %d, NTC: 0x%x, HWB: 0x%x, NTU: 0x%x, TAIL: 0x%x, INT: 0x%x\n",
			    vsi->seid, hung_queue, tx_ring->next_to_clean,
			    head, tx_ring->next_to_use,
			    readl(tx_ring->tail), val);
	}

J
Jesse Brandeburg 已提交
369
	pf->tx_timeout_last_recovery = jiffies;
370 371
	netdev_info(netdev, "tx_timeout recovery level %d, hung_queue %d\n",
		    pf->tx_timeout_recovery_level, hung_queue);
J
Jesse Brandeburg 已提交
372 373 374 375 376 377 378 379 380 381 382 383 384 385 386

	switch (pf->tx_timeout_recovery_level) {
	case 1:
		set_bit(__I40E_PF_RESET_REQUESTED, &pf->state);
		break;
	case 2:
		set_bit(__I40E_CORE_RESET_REQUESTED, &pf->state);
		break;
	case 3:
		set_bit(__I40E_GLOBAL_RESET_REQUESTED, &pf->state);
		break;
	default:
		netdev_err(netdev, "tx_timeout recovery unsuccessful\n");
		break;
	}
387

J
Jesse Brandeburg 已提交
388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410
	i40e_service_event_schedule(pf);
	pf->tx_timeout_recovery_level++;
}

/**
 * i40e_get_vsi_stats_struct - Get System Network Statistics
 * @vsi: the VSI we care about
 *
 * Returns the address of the device statistics structure.
 * The statistics are actually updated from the service task.
 **/
struct rtnl_link_stats64 *i40e_get_vsi_stats_struct(struct i40e_vsi *vsi)
{
	return &vsi->net_stats;
}

/**
 * i40e_get_netdev_stats_struct - Get statistics for netdev interface
 * @netdev: network interface device structure
 *
 * Returns the address of the device statistics structure.
 * The statistics are actually updated from the service task.
 **/
411 412 413 414 415
#ifdef I40E_FCOE
struct rtnl_link_stats64 *i40e_get_netdev_stats_struct(
					     struct net_device *netdev,
					     struct rtnl_link_stats64 *stats)
#else
J
Jesse Brandeburg 已提交
416 417
static struct rtnl_link_stats64 *i40e_get_netdev_stats_struct(
					     struct net_device *netdev,
418
					     struct rtnl_link_stats64 *stats)
419
#endif
J
Jesse Brandeburg 已提交
420 421
{
	struct i40e_netdev_priv *np = netdev_priv(netdev);
422
	struct i40e_ring *tx_ring, *rx_ring;
J
Jesse Brandeburg 已提交
423
	struct i40e_vsi *vsi = np->vsi;
424 425 426
	struct rtnl_link_stats64 *vsi_stats = i40e_get_vsi_stats_struct(vsi);
	int i;

427 428 429
	if (test_bit(__I40E_DOWN, &vsi->state))
		return stats;

J
Jesse Brandeburg 已提交
430 431 432
	if (!vsi->tx_rings)
		return stats;

433 434 435 436 437 438 439 440 441 442
	rcu_read_lock();
	for (i = 0; i < vsi->num_queue_pairs; i++) {
		u64 bytes, packets;
		unsigned int start;

		tx_ring = ACCESS_ONCE(vsi->tx_rings[i]);
		if (!tx_ring)
			continue;

		do {
443
			start = u64_stats_fetch_begin_irq(&tx_ring->syncp);
444 445
			packets = tx_ring->stats.packets;
			bytes   = tx_ring->stats.bytes;
446
		} while (u64_stats_fetch_retry_irq(&tx_ring->syncp, start));
447 448 449 450 451 452

		stats->tx_packets += packets;
		stats->tx_bytes   += bytes;
		rx_ring = &tx_ring[1];

		do {
453
			start = u64_stats_fetch_begin_irq(&rx_ring->syncp);
454 455
			packets = rx_ring->stats.packets;
			bytes   = rx_ring->stats.bytes;
456
		} while (u64_stats_fetch_retry_irq(&rx_ring->syncp, start));
J
Jesse Brandeburg 已提交
457

458 459 460 461 462
		stats->rx_packets += packets;
		stats->rx_bytes   += bytes;
	}
	rcu_read_unlock();

463
	/* following stats updated by i40e_watchdog_subtask() */
464 465 466 467
	stats->multicast	= vsi_stats->multicast;
	stats->tx_errors	= vsi_stats->tx_errors;
	stats->tx_dropped	= vsi_stats->tx_dropped;
	stats->rx_errors	= vsi_stats->rx_errors;
468
	stats->rx_dropped	= vsi_stats->rx_dropped;
469 470
	stats->rx_crc_errors	= vsi_stats->rx_crc_errors;
	stats->rx_length_errors	= vsi_stats->rx_length_errors;
J
Jesse Brandeburg 已提交
471

472
	return stats;
J
Jesse Brandeburg 已提交
473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491
}

/**
 * i40e_vsi_reset_stats - Resets all stats of the given vsi
 * @vsi: the VSI to have its stats reset
 **/
void i40e_vsi_reset_stats(struct i40e_vsi *vsi)
{
	struct rtnl_link_stats64 *ns;
	int i;

	if (!vsi)
		return;

	ns = i40e_get_vsi_stats_struct(vsi);
	memset(ns, 0, sizeof(*ns));
	memset(&vsi->net_stats_offsets, 0, sizeof(vsi->net_stats_offsets));
	memset(&vsi->eth_stats, 0, sizeof(vsi->eth_stats));
	memset(&vsi->eth_stats_offsets, 0, sizeof(vsi->eth_stats_offsets));
492
	if (vsi->rx_rings && vsi->rx_rings[0]) {
J
Jesse Brandeburg 已提交
493
		for (i = 0; i < vsi->num_queue_pairs; i++) {
J
Jesse Brandeburg 已提交
494
			memset(&vsi->rx_rings[i]->stats, 0,
495
			       sizeof(vsi->rx_rings[i]->stats));
J
Jesse Brandeburg 已提交
496
			memset(&vsi->rx_rings[i]->rx_stats, 0,
497
			       sizeof(vsi->rx_rings[i]->rx_stats));
J
Jesse Brandeburg 已提交
498
			memset(&vsi->tx_rings[i]->stats, 0,
499 500 501
			       sizeof(vsi->tx_rings[i]->stats));
			memset(&vsi->tx_rings[i]->tx_stats, 0,
			       sizeof(vsi->tx_rings[i]->tx_stats));
J
Jesse Brandeburg 已提交
502
		}
503
	}
J
Jesse Brandeburg 已提交
504 505 506 507
	vsi->stat_offsets_loaded = false;
}

/**
508
 * i40e_pf_reset_stats - Reset all of the stats for the given PF
J
Jesse Brandeburg 已提交
509 510 511 512
 * @pf: the PF to be reset
 **/
void i40e_pf_reset_stats(struct i40e_pf *pf)
{
513 514
	int i;

J
Jesse Brandeburg 已提交
515 516 517
	memset(&pf->stats, 0, sizeof(pf->stats));
	memset(&pf->stats_offsets, 0, sizeof(pf->stats_offsets));
	pf->stat_offsets_loaded = false;
518 519 520 521 522 523 524 525 526 527

	for (i = 0; i < I40E_MAX_VEB; i++) {
		if (pf->veb[i]) {
			memset(&pf->veb[i]->stats, 0,
			       sizeof(pf->veb[i]->stats));
			memset(&pf->veb[i]->stats_offsets, 0,
			       sizeof(pf->veb[i]->stats_offsets));
			pf->veb[i]->stat_offsets_loaded = false;
		}
	}
528
	pf->hw_csum_rx_error = 0;
J
Jesse Brandeburg 已提交
529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550
}

/**
 * i40e_stat_update48 - read and update a 48 bit stat from the chip
 * @hw: ptr to the hardware info
 * @hireg: the high 32 bit reg to read
 * @loreg: the low 32 bit reg to read
 * @offset_loaded: has the initial offset been loaded yet
 * @offset: ptr to current offset value
 * @stat: ptr to the stat
 *
 * Since the device stats are not reset at PFReset, they likely will not
 * be zeroed when the driver starts.  We'll save the first values read
 * and use them as offsets to be subtracted from the raw values in order
 * to report stats that count from zero.  In the process, we also manage
 * the potential roll-over.
 **/
static void i40e_stat_update48(struct i40e_hw *hw, u32 hireg, u32 loreg,
			       bool offset_loaded, u64 *offset, u64 *stat)
{
	u64 new_data;

551
	if (hw->device_id == I40E_DEV_ID_QEMU) {
J
Jesse Brandeburg 已提交
552 553 554 555 556 557 558 559 560 561
		new_data = rd32(hw, loreg);
		new_data |= ((u64)(rd32(hw, hireg) & 0xFFFF)) << 32;
	} else {
		new_data = rd64(hw, loreg);
	}
	if (!offset_loaded)
		*offset = new_data;
	if (likely(new_data >= *offset))
		*stat = new_data - *offset;
	else
562
		*stat = (new_data + BIT_ULL(48)) - *offset;
J
Jesse Brandeburg 已提交
563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584
	*stat &= 0xFFFFFFFFFFFFULL;
}

/**
 * i40e_stat_update32 - read and update a 32 bit stat from the chip
 * @hw: ptr to the hardware info
 * @reg: the hw reg to read
 * @offset_loaded: has the initial offset been loaded yet
 * @offset: ptr to current offset value
 * @stat: ptr to the stat
 **/
static void i40e_stat_update32(struct i40e_hw *hw, u32 reg,
			       bool offset_loaded, u64 *offset, u64 *stat)
{
	u32 new_data;

	new_data = rd32(hw, reg);
	if (!offset_loaded)
		*offset = new_data;
	if (likely(new_data >= *offset))
		*stat = (u32)(new_data - *offset);
	else
585
		*stat = (u32)((new_data + BIT_ULL(32)) - *offset);
J
Jesse Brandeburg 已提交
586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609
}

/**
 * i40e_update_eth_stats - Update VSI-specific ethernet statistics counters.
 * @vsi: the VSI to be updated
 **/
void i40e_update_eth_stats(struct i40e_vsi *vsi)
{
	int stat_idx = le16_to_cpu(vsi->info.stat_counter_idx);
	struct i40e_pf *pf = vsi->back;
	struct i40e_hw *hw = &pf->hw;
	struct i40e_eth_stats *oes;
	struct i40e_eth_stats *es;     /* device's eth stats */

	es = &vsi->eth_stats;
	oes = &vsi->eth_stats_offsets;

	/* Gather up the stats that the hw collects */
	i40e_stat_update32(hw, I40E_GLV_TEPC(stat_idx),
			   vsi->stat_offsets_loaded,
			   &oes->tx_errors, &es->tx_errors);
	i40e_stat_update32(hw, I40E_GLV_RDPC(stat_idx),
			   vsi->stat_offsets_loaded,
			   &oes->rx_discards, &es->rx_discards);
610 611 612 613 614 615
	i40e_stat_update32(hw, I40E_GLV_RUPP(stat_idx),
			   vsi->stat_offsets_loaded,
			   &oes->rx_unknown_protocol, &es->rx_unknown_protocol);
	i40e_stat_update32(hw, I40E_GLV_TEPC(stat_idx),
			   vsi->stat_offsets_loaded,
			   &oes->tx_errors, &es->tx_errors);
J
Jesse Brandeburg 已提交
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

	i40e_stat_update48(hw, I40E_GLV_GORCH(stat_idx),
			   I40E_GLV_GORCL(stat_idx),
			   vsi->stat_offsets_loaded,
			   &oes->rx_bytes, &es->rx_bytes);
	i40e_stat_update48(hw, I40E_GLV_UPRCH(stat_idx),
			   I40E_GLV_UPRCL(stat_idx),
			   vsi->stat_offsets_loaded,
			   &oes->rx_unicast, &es->rx_unicast);
	i40e_stat_update48(hw, I40E_GLV_MPRCH(stat_idx),
			   I40E_GLV_MPRCL(stat_idx),
			   vsi->stat_offsets_loaded,
			   &oes->rx_multicast, &es->rx_multicast);
	i40e_stat_update48(hw, I40E_GLV_BPRCH(stat_idx),
			   I40E_GLV_BPRCL(stat_idx),
			   vsi->stat_offsets_loaded,
			   &oes->rx_broadcast, &es->rx_broadcast);

	i40e_stat_update48(hw, I40E_GLV_GOTCH(stat_idx),
			   I40E_GLV_GOTCL(stat_idx),
			   vsi->stat_offsets_loaded,
			   &oes->tx_bytes, &es->tx_bytes);
	i40e_stat_update48(hw, I40E_GLV_UPTCH(stat_idx),
			   I40E_GLV_UPTCL(stat_idx),
			   vsi->stat_offsets_loaded,
			   &oes->tx_unicast, &es->tx_unicast);
	i40e_stat_update48(hw, I40E_GLV_MPTCH(stat_idx),
			   I40E_GLV_MPTCL(stat_idx),
			   vsi->stat_offsets_loaded,
			   &oes->tx_multicast, &es->tx_multicast);
	i40e_stat_update48(hw, I40E_GLV_BPTCH(stat_idx),
			   I40E_GLV_BPTCL(stat_idx),
			   vsi->stat_offsets_loaded,
			   &oes->tx_broadcast, &es->tx_broadcast);
	vsi->stat_offsets_loaded = true;
}

/**
 * i40e_update_veb_stats - Update Switch component statistics
 * @veb: the VEB being updated
 **/
static void i40e_update_veb_stats(struct i40e_veb *veb)
{
	struct i40e_pf *pf = veb->pf;
	struct i40e_hw *hw = &pf->hw;
	struct i40e_eth_stats *oes;
	struct i40e_eth_stats *es;     /* device's eth stats */
663 664 665
	struct i40e_veb_tc_stats *veb_oes;
	struct i40e_veb_tc_stats *veb_es;
	int i, idx = 0;
J
Jesse Brandeburg 已提交
666 667 668 669

	idx = veb->stats_idx;
	es = &veb->stats;
	oes = &veb->stats_offsets;
670 671
	veb_es = &veb->tc_stats;
	veb_oes = &veb->tc_stats_offsets;
J
Jesse Brandeburg 已提交
672 673 674 675 676

	/* Gather up the stats that the hw collects */
	i40e_stat_update32(hw, I40E_GLSW_TDPC(idx),
			   veb->stat_offsets_loaded,
			   &oes->tx_discards, &es->tx_discards);
677 678 679 680 681
	if (hw->revision_id > 0)
		i40e_stat_update32(hw, I40E_GLSW_RUPP(idx),
				   veb->stat_offsets_loaded,
				   &oes->rx_unknown_protocol,
				   &es->rx_unknown_protocol);
J
Jesse Brandeburg 已提交
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
	i40e_stat_update48(hw, I40E_GLSW_GORCH(idx), I40E_GLSW_GORCL(idx),
			   veb->stat_offsets_loaded,
			   &oes->rx_bytes, &es->rx_bytes);
	i40e_stat_update48(hw, I40E_GLSW_UPRCH(idx), I40E_GLSW_UPRCL(idx),
			   veb->stat_offsets_loaded,
			   &oes->rx_unicast, &es->rx_unicast);
	i40e_stat_update48(hw, I40E_GLSW_MPRCH(idx), I40E_GLSW_MPRCL(idx),
			   veb->stat_offsets_loaded,
			   &oes->rx_multicast, &es->rx_multicast);
	i40e_stat_update48(hw, I40E_GLSW_BPRCH(idx), I40E_GLSW_BPRCL(idx),
			   veb->stat_offsets_loaded,
			   &oes->rx_broadcast, &es->rx_broadcast);

	i40e_stat_update48(hw, I40E_GLSW_GOTCH(idx), I40E_GLSW_GOTCL(idx),
			   veb->stat_offsets_loaded,
			   &oes->tx_bytes, &es->tx_bytes);
	i40e_stat_update48(hw, I40E_GLSW_UPTCH(idx), I40E_GLSW_UPTCL(idx),
			   veb->stat_offsets_loaded,
			   &oes->tx_unicast, &es->tx_unicast);
	i40e_stat_update48(hw, I40E_GLSW_MPTCH(idx), I40E_GLSW_MPTCL(idx),
			   veb->stat_offsets_loaded,
			   &oes->tx_multicast, &es->tx_multicast);
	i40e_stat_update48(hw, I40E_GLSW_BPTCH(idx), I40E_GLSW_BPTCL(idx),
			   veb->stat_offsets_loaded,
			   &oes->tx_broadcast, &es->tx_broadcast);
707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728
	for (i = 0; i < I40E_MAX_TRAFFIC_CLASS; i++) {
		i40e_stat_update48(hw, I40E_GLVEBTC_RPCH(i, idx),
				   I40E_GLVEBTC_RPCL(i, idx),
				   veb->stat_offsets_loaded,
				   &veb_oes->tc_rx_packets[i],
				   &veb_es->tc_rx_packets[i]);
		i40e_stat_update48(hw, I40E_GLVEBTC_RBCH(i, idx),
				   I40E_GLVEBTC_RBCL(i, idx),
				   veb->stat_offsets_loaded,
				   &veb_oes->tc_rx_bytes[i],
				   &veb_es->tc_rx_bytes[i]);
		i40e_stat_update48(hw, I40E_GLVEBTC_TPCH(i, idx),
				   I40E_GLVEBTC_TPCL(i, idx),
				   veb->stat_offsets_loaded,
				   &veb_oes->tc_tx_packets[i],
				   &veb_es->tc_tx_packets[i]);
		i40e_stat_update48(hw, I40E_GLVEBTC_TBCH(i, idx),
				   I40E_GLVEBTC_TBCL(i, idx),
				   veb->stat_offsets_loaded,
				   &veb_oes->tc_tx_bytes[i],
				   &veb_es->tc_tx_bytes[i]);
	}
J
Jesse Brandeburg 已提交
729 730 731
	veb->stat_offsets_loaded = true;
}

732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747
#ifdef I40E_FCOE
/**
 * i40e_update_fcoe_stats - Update FCoE-specific ethernet statistics counters.
 * @vsi: the VSI that is capable of doing FCoE
 **/
static void i40e_update_fcoe_stats(struct i40e_vsi *vsi)
{
	struct i40e_pf *pf = vsi->back;
	struct i40e_hw *hw = &pf->hw;
	struct i40e_fcoe_stats *ofs;
	struct i40e_fcoe_stats *fs;     /* device's eth stats */
	int idx;

	if (vsi->type != I40E_VSI_FCOE)
		return;

748
	idx = hw->pf_id + I40E_FCOE_PF_STAT_OFFSET;
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
	fs = &vsi->fcoe_stats;
	ofs = &vsi->fcoe_stats_offsets;

	i40e_stat_update32(hw, I40E_GL_FCOEPRC(idx),
			   vsi->fcoe_stat_offsets_loaded,
			   &ofs->rx_fcoe_packets, &fs->rx_fcoe_packets);
	i40e_stat_update48(hw, I40E_GL_FCOEDWRCH(idx), I40E_GL_FCOEDWRCL(idx),
			   vsi->fcoe_stat_offsets_loaded,
			   &ofs->rx_fcoe_dwords, &fs->rx_fcoe_dwords);
	i40e_stat_update32(hw, I40E_GL_FCOERPDC(idx),
			   vsi->fcoe_stat_offsets_loaded,
			   &ofs->rx_fcoe_dropped, &fs->rx_fcoe_dropped);
	i40e_stat_update32(hw, I40E_GL_FCOEPTC(idx),
			   vsi->fcoe_stat_offsets_loaded,
			   &ofs->tx_fcoe_packets, &fs->tx_fcoe_packets);
	i40e_stat_update48(hw, I40E_GL_FCOEDWTCH(idx), I40E_GL_FCOEDWTCL(idx),
			   vsi->fcoe_stat_offsets_loaded,
			   &ofs->tx_fcoe_dwords, &fs->tx_fcoe_dwords);
	i40e_stat_update32(hw, I40E_GL_FCOECRC(idx),
			   vsi->fcoe_stat_offsets_loaded,
			   &ofs->fcoe_bad_fccrc, &fs->fcoe_bad_fccrc);
	i40e_stat_update32(hw, I40E_GL_FCOELAST(idx),
			   vsi->fcoe_stat_offsets_loaded,
			   &ofs->fcoe_last_error, &fs->fcoe_last_error);
	i40e_stat_update32(hw, I40E_GL_FCOEDDPC(idx),
			   vsi->fcoe_stat_offsets_loaded,
			   &ofs->fcoe_ddp_count, &fs->fcoe_ddp_count);

	vsi->fcoe_stat_offsets_loaded = true;
}

#endif
J
Jesse Brandeburg 已提交
781
/**
S
Shannon Nelson 已提交
782
 * i40e_update_vsi_stats - Update the vsi statistics counters.
J
Jesse Brandeburg 已提交
783 784 785 786 787 788
 * @vsi: the VSI to be updated
 *
 * There are a few instances where we store the same stat in a
 * couple of different structs.  This is partly because we have
 * the netdev stats that need to be filled out, which is slightly
 * different from the "eth_stats" defined by the chip and used in
S
Shannon Nelson 已提交
789
 * VF communications.  We sort it out here.
J
Jesse Brandeburg 已提交
790
 **/
S
Shannon Nelson 已提交
791
static void i40e_update_vsi_stats(struct i40e_vsi *vsi)
J
Jesse Brandeburg 已提交
792 793 794 795 796 797 798
{
	struct i40e_pf *pf = vsi->back;
	struct rtnl_link_stats64 *ons;
	struct rtnl_link_stats64 *ns;   /* netdev stats */
	struct i40e_eth_stats *oes;
	struct i40e_eth_stats *es;     /* device's eth stats */
	u32 tx_restart, tx_busy;
799
	u64 tx_lost_interrupt;
800
	struct i40e_ring *p;
J
Jesse Brandeburg 已提交
801
	u32 rx_page, rx_buf;
802 803
	u64 bytes, packets;
	unsigned int start;
804
	u64 tx_linearize;
805
	u64 tx_force_wb;
J
Jesse Brandeburg 已提交
806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823
	u64 rx_p, rx_b;
	u64 tx_p, tx_b;
	u16 q;

	if (test_bit(__I40E_DOWN, &vsi->state) ||
	    test_bit(__I40E_CONFIG_BUSY, &pf->state))
		return;

	ns = i40e_get_vsi_stats_struct(vsi);
	ons = &vsi->net_stats_offsets;
	es = &vsi->eth_stats;
	oes = &vsi->eth_stats_offsets;

	/* Gather up the netdev and vsi stats that the driver collects
	 * on the fly during packet processing
	 */
	rx_b = rx_p = 0;
	tx_b = tx_p = 0;
824
	tx_restart = tx_busy = tx_linearize = tx_force_wb = 0;
825
	tx_lost_interrupt = 0;
J
Jesse Brandeburg 已提交
826 827
	rx_page = 0;
	rx_buf = 0;
828
	rcu_read_lock();
J
Jesse Brandeburg 已提交
829
	for (q = 0; q < vsi->num_queue_pairs; q++) {
830 831 832 833
		/* locate Tx ring */
		p = ACCESS_ONCE(vsi->tx_rings[q]);

		do {
834
			start = u64_stats_fetch_begin_irq(&p->syncp);
835 836
			packets = p->stats.packets;
			bytes = p->stats.bytes;
837
		} while (u64_stats_fetch_retry_irq(&p->syncp, start));
838 839 840 841
		tx_b += bytes;
		tx_p += packets;
		tx_restart += p->tx_stats.restart_queue;
		tx_busy += p->tx_stats.tx_busy;
842
		tx_linearize += p->tx_stats.tx_linearize;
843
		tx_force_wb += p->tx_stats.tx_force_wb;
844
		tx_lost_interrupt += p->tx_stats.tx_lost_interrupt;
J
Jesse Brandeburg 已提交
845

846 847 848
		/* Rx queue is part of the same block as Tx queue */
		p = &p[1];
		do {
849
			start = u64_stats_fetch_begin_irq(&p->syncp);
850 851
			packets = p->stats.packets;
			bytes = p->stats.bytes;
852
		} while (u64_stats_fetch_retry_irq(&p->syncp, start));
853 854
		rx_b += bytes;
		rx_p += packets;
M
Mitch Williams 已提交
855 856
		rx_buf += p->rx_stats.alloc_buff_failed;
		rx_page += p->rx_stats.alloc_page_failed;
J
Jesse Brandeburg 已提交
857
	}
858
	rcu_read_unlock();
J
Jesse Brandeburg 已提交
859 860
	vsi->tx_restart = tx_restart;
	vsi->tx_busy = tx_busy;
861
	vsi->tx_linearize = tx_linearize;
862
	vsi->tx_force_wb = tx_force_wb;
863
	vsi->tx_lost_interrupt = tx_lost_interrupt;
J
Jesse Brandeburg 已提交
864 865 866 867 868 869 870 871 872
	vsi->rx_page_failed = rx_page;
	vsi->rx_buf_failed = rx_buf;

	ns->rx_packets = rx_p;
	ns->rx_bytes = rx_b;
	ns->tx_packets = tx_p;
	ns->tx_bytes = tx_b;

	/* update netdev stats from eth stats */
S
Shannon Nelson 已提交
873
	i40e_update_eth_stats(vsi);
J
Jesse Brandeburg 已提交
874 875 876 877
	ons->tx_errors = oes->tx_errors;
	ns->tx_errors = es->tx_errors;
	ons->multicast = oes->rx_multicast;
	ns->multicast = es->rx_multicast;
878 879
	ons->rx_dropped = oes->rx_discards;
	ns->rx_dropped = es->rx_discards;
J
Jesse Brandeburg 已提交
880 881 882
	ons->tx_dropped = oes->tx_discards;
	ns->tx_dropped = es->tx_discards;

S
Shannon Nelson 已提交
883
	/* pull in a couple PF stats if this is the main vsi */
J
Jesse Brandeburg 已提交
884
	if (vsi == pf->vsi[pf->lan_vsi]) {
S
Shannon Nelson 已提交
885 886 887 888 889
		ns->rx_crc_errors = pf->stats.crc_errors;
		ns->rx_errors = pf->stats.crc_errors + pf->stats.illegal_bytes;
		ns->rx_length_errors = pf->stats.rx_length_errors;
	}
}
J
Jesse Brandeburg 已提交
890

S
Shannon Nelson 已提交
891
/**
892
 * i40e_update_pf_stats - Update the PF statistics counters.
S
Shannon Nelson 已提交
893 894 895 896 897 898 899 900 901
 * @pf: the PF to be updated
 **/
static void i40e_update_pf_stats(struct i40e_pf *pf)
{
	struct i40e_hw_port_stats *osd = &pf->stats_offsets;
	struct i40e_hw_port_stats *nsd = &pf->stats;
	struct i40e_hw *hw = &pf->hw;
	u32 val;
	int i;
J
Jesse Brandeburg 已提交
902

S
Shannon Nelson 已提交
903 904 905 906 907 908 909 910 911 912 913 914
	i40e_stat_update48(hw, I40E_GLPRT_GORCH(hw->port),
			   I40E_GLPRT_GORCL(hw->port),
			   pf->stat_offsets_loaded,
			   &osd->eth.rx_bytes, &nsd->eth.rx_bytes);
	i40e_stat_update48(hw, I40E_GLPRT_GOTCH(hw->port),
			   I40E_GLPRT_GOTCL(hw->port),
			   pf->stat_offsets_loaded,
			   &osd->eth.tx_bytes, &nsd->eth.tx_bytes);
	i40e_stat_update32(hw, I40E_GLPRT_RDPC(hw->port),
			   pf->stat_offsets_loaded,
			   &osd->eth.rx_discards,
			   &nsd->eth.rx_discards);
S
Shannon Nelson 已提交
915 916 917 918 919
	i40e_stat_update48(hw, I40E_GLPRT_UPRCH(hw->port),
			   I40E_GLPRT_UPRCL(hw->port),
			   pf->stat_offsets_loaded,
			   &osd->eth.rx_unicast,
			   &nsd->eth.rx_unicast);
S
Shannon Nelson 已提交
920 921 922 923 924
	i40e_stat_update48(hw, I40E_GLPRT_MPRCH(hw->port),
			   I40E_GLPRT_MPRCL(hw->port),
			   pf->stat_offsets_loaded,
			   &osd->eth.rx_multicast,
			   &nsd->eth.rx_multicast);
S
Shannon Nelson 已提交
925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944
	i40e_stat_update48(hw, I40E_GLPRT_BPRCH(hw->port),
			   I40E_GLPRT_BPRCL(hw->port),
			   pf->stat_offsets_loaded,
			   &osd->eth.rx_broadcast,
			   &nsd->eth.rx_broadcast);
	i40e_stat_update48(hw, I40E_GLPRT_UPTCH(hw->port),
			   I40E_GLPRT_UPTCL(hw->port),
			   pf->stat_offsets_loaded,
			   &osd->eth.tx_unicast,
			   &nsd->eth.tx_unicast);
	i40e_stat_update48(hw, I40E_GLPRT_MPTCH(hw->port),
			   I40E_GLPRT_MPTCL(hw->port),
			   pf->stat_offsets_loaded,
			   &osd->eth.tx_multicast,
			   &nsd->eth.tx_multicast);
	i40e_stat_update48(hw, I40E_GLPRT_BPTCH(hw->port),
			   I40E_GLPRT_BPTCL(hw->port),
			   pf->stat_offsets_loaded,
			   &osd->eth.tx_broadcast,
			   &nsd->eth.tx_broadcast);
J
Jesse Brandeburg 已提交
945

S
Shannon Nelson 已提交
946 947 948 949
	i40e_stat_update32(hw, I40E_GLPRT_TDOLD(hw->port),
			   pf->stat_offsets_loaded,
			   &osd->tx_dropped_link_down,
			   &nsd->tx_dropped_link_down);
J
Jesse Brandeburg 已提交
950

S
Shannon Nelson 已提交
951 952 953
	i40e_stat_update32(hw, I40E_GLPRT_CRCERRS(hw->port),
			   pf->stat_offsets_loaded,
			   &osd->crc_errors, &nsd->crc_errors);
J
Jesse Brandeburg 已提交
954

S
Shannon Nelson 已提交
955 956 957
	i40e_stat_update32(hw, I40E_GLPRT_ILLERRC(hw->port),
			   pf->stat_offsets_loaded,
			   &osd->illegal_bytes, &nsd->illegal_bytes);
J
Jesse Brandeburg 已提交
958

S
Shannon Nelson 已提交
959 960 961 962 963 964 965 966
	i40e_stat_update32(hw, I40E_GLPRT_MLFC(hw->port),
			   pf->stat_offsets_loaded,
			   &osd->mac_local_faults,
			   &nsd->mac_local_faults);
	i40e_stat_update32(hw, I40E_GLPRT_MRFC(hw->port),
			   pf->stat_offsets_loaded,
			   &osd->mac_remote_faults,
			   &nsd->mac_remote_faults);
J
Jesse Brandeburg 已提交
967

S
Shannon Nelson 已提交
968 969 970 971
	i40e_stat_update32(hw, I40E_GLPRT_RLEC(hw->port),
			   pf->stat_offsets_loaded,
			   &osd->rx_length_errors,
			   &nsd->rx_length_errors);
J
Jesse Brandeburg 已提交
972

S
Shannon Nelson 已提交
973 974 975 976 977 978
	i40e_stat_update32(hw, I40E_GLPRT_LXONRXC(hw->port),
			   pf->stat_offsets_loaded,
			   &osd->link_xon_rx, &nsd->link_xon_rx);
	i40e_stat_update32(hw, I40E_GLPRT_LXONTXC(hw->port),
			   pf->stat_offsets_loaded,
			   &osd->link_xon_tx, &nsd->link_xon_tx);
979 980 981
	i40e_stat_update32(hw, I40E_GLPRT_LXOFFRXC(hw->port),
			   pf->stat_offsets_loaded,
			   &osd->link_xoff_rx, &nsd->link_xoff_rx);
S
Shannon Nelson 已提交
982 983 984
	i40e_stat_update32(hw, I40E_GLPRT_LXOFFTXC(hw->port),
			   pf->stat_offsets_loaded,
			   &osd->link_xoff_tx, &nsd->link_xoff_tx);
J
Jesse Brandeburg 已提交
985

S
Shannon Nelson 已提交
986
	for (i = 0; i < 8; i++) {
987 988 989 990
		i40e_stat_update32(hw, I40E_GLPRT_PXOFFRXC(hw->port, i),
				   pf->stat_offsets_loaded,
				   &osd->priority_xoff_rx[i],
				   &nsd->priority_xoff_rx[i]);
S
Shannon Nelson 已提交
991
		i40e_stat_update32(hw, I40E_GLPRT_PXONRXC(hw->port, i),
J
Jesse Brandeburg 已提交
992
				   pf->stat_offsets_loaded,
S
Shannon Nelson 已提交
993 994 995
				   &osd->priority_xon_rx[i],
				   &nsd->priority_xon_rx[i]);
		i40e_stat_update32(hw, I40E_GLPRT_PXONTXC(hw->port, i),
J
Jesse Brandeburg 已提交
996
				   pf->stat_offsets_loaded,
S
Shannon Nelson 已提交
997 998 999
				   &osd->priority_xon_tx[i],
				   &nsd->priority_xon_tx[i]);
		i40e_stat_update32(hw, I40E_GLPRT_PXOFFTXC(hw->port, i),
J
Jesse Brandeburg 已提交
1000
				   pf->stat_offsets_loaded,
S
Shannon Nelson 已提交
1001 1002 1003 1004
				   &osd->priority_xoff_tx[i],
				   &nsd->priority_xoff_tx[i]);
		i40e_stat_update32(hw,
				   I40E_GLPRT_RXON2OFFCNT(hw->port, i),
1005
				   pf->stat_offsets_loaded,
S
Shannon Nelson 已提交
1006 1007
				   &osd->priority_xon_2_xoff[i],
				   &nsd->priority_xon_2_xoff[i]);
J
Jesse Brandeburg 已提交
1008 1009
	}

S
Shannon Nelson 已提交
1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080
	i40e_stat_update48(hw, I40E_GLPRT_PRC64H(hw->port),
			   I40E_GLPRT_PRC64L(hw->port),
			   pf->stat_offsets_loaded,
			   &osd->rx_size_64, &nsd->rx_size_64);
	i40e_stat_update48(hw, I40E_GLPRT_PRC127H(hw->port),
			   I40E_GLPRT_PRC127L(hw->port),
			   pf->stat_offsets_loaded,
			   &osd->rx_size_127, &nsd->rx_size_127);
	i40e_stat_update48(hw, I40E_GLPRT_PRC255H(hw->port),
			   I40E_GLPRT_PRC255L(hw->port),
			   pf->stat_offsets_loaded,
			   &osd->rx_size_255, &nsd->rx_size_255);
	i40e_stat_update48(hw, I40E_GLPRT_PRC511H(hw->port),
			   I40E_GLPRT_PRC511L(hw->port),
			   pf->stat_offsets_loaded,
			   &osd->rx_size_511, &nsd->rx_size_511);
	i40e_stat_update48(hw, I40E_GLPRT_PRC1023H(hw->port),
			   I40E_GLPRT_PRC1023L(hw->port),
			   pf->stat_offsets_loaded,
			   &osd->rx_size_1023, &nsd->rx_size_1023);
	i40e_stat_update48(hw, I40E_GLPRT_PRC1522H(hw->port),
			   I40E_GLPRT_PRC1522L(hw->port),
			   pf->stat_offsets_loaded,
			   &osd->rx_size_1522, &nsd->rx_size_1522);
	i40e_stat_update48(hw, I40E_GLPRT_PRC9522H(hw->port),
			   I40E_GLPRT_PRC9522L(hw->port),
			   pf->stat_offsets_loaded,
			   &osd->rx_size_big, &nsd->rx_size_big);

	i40e_stat_update48(hw, I40E_GLPRT_PTC64H(hw->port),
			   I40E_GLPRT_PTC64L(hw->port),
			   pf->stat_offsets_loaded,
			   &osd->tx_size_64, &nsd->tx_size_64);
	i40e_stat_update48(hw, I40E_GLPRT_PTC127H(hw->port),
			   I40E_GLPRT_PTC127L(hw->port),
			   pf->stat_offsets_loaded,
			   &osd->tx_size_127, &nsd->tx_size_127);
	i40e_stat_update48(hw, I40E_GLPRT_PTC255H(hw->port),
			   I40E_GLPRT_PTC255L(hw->port),
			   pf->stat_offsets_loaded,
			   &osd->tx_size_255, &nsd->tx_size_255);
	i40e_stat_update48(hw, I40E_GLPRT_PTC511H(hw->port),
			   I40E_GLPRT_PTC511L(hw->port),
			   pf->stat_offsets_loaded,
			   &osd->tx_size_511, &nsd->tx_size_511);
	i40e_stat_update48(hw, I40E_GLPRT_PTC1023H(hw->port),
			   I40E_GLPRT_PTC1023L(hw->port),
			   pf->stat_offsets_loaded,
			   &osd->tx_size_1023, &nsd->tx_size_1023);
	i40e_stat_update48(hw, I40E_GLPRT_PTC1522H(hw->port),
			   I40E_GLPRT_PTC1522L(hw->port),
			   pf->stat_offsets_loaded,
			   &osd->tx_size_1522, &nsd->tx_size_1522);
	i40e_stat_update48(hw, I40E_GLPRT_PTC9522H(hw->port),
			   I40E_GLPRT_PTC9522L(hw->port),
			   pf->stat_offsets_loaded,
			   &osd->tx_size_big, &nsd->tx_size_big);

	i40e_stat_update32(hw, I40E_GLPRT_RUC(hw->port),
			   pf->stat_offsets_loaded,
			   &osd->rx_undersize, &nsd->rx_undersize);
	i40e_stat_update32(hw, I40E_GLPRT_RFC(hw->port),
			   pf->stat_offsets_loaded,
			   &osd->rx_fragments, &nsd->rx_fragments);
	i40e_stat_update32(hw, I40E_GLPRT_ROC(hw->port),
			   pf->stat_offsets_loaded,
			   &osd->rx_oversize, &nsd->rx_oversize);
	i40e_stat_update32(hw, I40E_GLPRT_RJC(hw->port),
			   pf->stat_offsets_loaded,
			   &osd->rx_jabber, &nsd->rx_jabber);

1081
	/* FDIR stats */
1082 1083
	i40e_stat_update32(hw,
			   I40E_GLQF_PCNT(I40E_FD_ATR_STAT_IDX(pf->hw.pf_id)),
1084 1085
			   pf->stat_offsets_loaded,
			   &osd->fd_atr_match, &nsd->fd_atr_match);
1086 1087
	i40e_stat_update32(hw,
			   I40E_GLQF_PCNT(I40E_FD_SB_STAT_IDX(pf->hw.pf_id)),
1088 1089
			   pf->stat_offsets_loaded,
			   &osd->fd_sb_match, &nsd->fd_sb_match);
1090 1091 1092 1093
	i40e_stat_update32(hw,
		      I40E_GLQF_PCNT(I40E_FD_ATR_TUNNEL_STAT_IDX(pf->hw.pf_id)),
		      pf->stat_offsets_loaded,
		      &osd->fd_atr_tunnel_match, &nsd->fd_atr_tunnel_match);
1094

S
Shannon Nelson 已提交
1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108
	val = rd32(hw, I40E_PRTPM_EEE_STAT);
	nsd->tx_lpi_status =
		       (val & I40E_PRTPM_EEE_STAT_TX_LPI_STATUS_MASK) >>
			I40E_PRTPM_EEE_STAT_TX_LPI_STATUS_SHIFT;
	nsd->rx_lpi_status =
		       (val & I40E_PRTPM_EEE_STAT_RX_LPI_STATUS_MASK) >>
			I40E_PRTPM_EEE_STAT_RX_LPI_STATUS_SHIFT;
	i40e_stat_update32(hw, I40E_PRTPM_TLPIC,
			   pf->stat_offsets_loaded,
			   &osd->tx_lpi_count, &nsd->tx_lpi_count);
	i40e_stat_update32(hw, I40E_PRTPM_RLPIC,
			   pf->stat_offsets_loaded,
			   &osd->rx_lpi_count, &nsd->rx_lpi_count);

1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120
	if (pf->flags & I40E_FLAG_FD_SB_ENABLED &&
	    !(pf->auto_disable_flags & I40E_FLAG_FD_SB_ENABLED))
		nsd->fd_sb_status = true;
	else
		nsd->fd_sb_status = false;

	if (pf->flags & I40E_FLAG_FD_ATR_ENABLED &&
	    !(pf->auto_disable_flags & I40E_FLAG_FD_ATR_ENABLED))
		nsd->fd_atr_status = true;
	else
		nsd->fd_atr_status = false;

J
Jesse Brandeburg 已提交
1121 1122 1123
	pf->stat_offsets_loaded = true;
}

S
Shannon Nelson 已提交
1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137
/**
 * i40e_update_stats - Update the various statistics counters.
 * @vsi: the VSI to be updated
 *
 * Update the various stats for this VSI and its related entities.
 **/
void i40e_update_stats(struct i40e_vsi *vsi)
{
	struct i40e_pf *pf = vsi->back;

	if (vsi == pf->vsi[pf->lan_vsi])
		i40e_update_pf_stats(pf);

	i40e_update_vsi_stats(vsi);
1138 1139 1140
#ifdef I40E_FCOE
	i40e_update_fcoe_stats(vsi);
#endif
S
Shannon Nelson 已提交
1141 1142
}

J
Jesse Brandeburg 已提交
1143 1144 1145 1146 1147 1148 1149 1150 1151
/**
 * i40e_find_filter - Search VSI filter list for specific mac/vlan filter
 * @vsi: the VSI to be searched
 * @macaddr: the MAC address
 * @vlan: the vlan
 *
 * Returns ptr to the filter object or NULL
 **/
static struct i40e_mac_filter *i40e_find_filter(struct i40e_vsi *vsi,
1152
						const u8 *macaddr, s16 vlan)
J
Jesse Brandeburg 已提交
1153 1154
{
	struct i40e_mac_filter *f;
1155
	u64 key;
J
Jesse Brandeburg 已提交
1156 1157 1158 1159

	if (!vsi || !macaddr)
		return NULL;

1160 1161
	key = i40e_addr_to_hkey(macaddr);
	hash_for_each_possible(vsi->mac_filter_hash, f, hlist, key) {
J
Jesse Brandeburg 已提交
1162
		if ((ether_addr_equal(macaddr, f->macaddr)) &&
1163
		    (vlan == f->vlan))
J
Jesse Brandeburg 已提交
1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176
			return f;
	}
	return NULL;
}

/**
 * i40e_find_mac - Find a mac addr in the macvlan filters list
 * @vsi: the VSI to be searched
 * @macaddr: the MAC address we are searching for
 *
 * Returns the first filter with the provided MAC address or NULL if
 * MAC address was not found
 **/
1177
struct i40e_mac_filter *i40e_find_mac(struct i40e_vsi *vsi, const u8 *macaddr)
J
Jesse Brandeburg 已提交
1178 1179
{
	struct i40e_mac_filter *f;
1180
	u64 key;
J
Jesse Brandeburg 已提交
1181 1182 1183 1184

	if (!vsi || !macaddr)
		return NULL;

1185 1186
	key = i40e_addr_to_hkey(macaddr);
	hash_for_each_possible(vsi->mac_filter_hash, f, hlist, key) {
1187
		if ((ether_addr_equal(macaddr, f->macaddr)))
J
Jesse Brandeburg 已提交
1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200
			return f;
	}
	return NULL;
}

/**
 * i40e_is_vsi_in_vlan - Check if VSI is in vlan mode
 * @vsi: the VSI to be searched
 *
 * Returns true if VSI is in vlan mode or false otherwise
 **/
bool i40e_is_vsi_in_vlan(struct i40e_vsi *vsi)
{
1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223
	/* If we have a PVID, always operate in VLAN mode */
	if (vsi->info.pvid)
		return true;

	/* We need to operate in VLAN mode whenever we have any filters with
	 * a VLAN other than I40E_VLAN_ALL. We could check the table each
	 * time, incurring search cost repeatedly. However, we can notice two
	 * things:
	 *
	 * 1) the only place where we can gain a VLAN filter is in
	 *    i40e_add_filter.
	 *
	 * 2) the only place where filters are actually removed is in
	 *    i40e_vsi_sync_filters_subtask.
	 *
	 * Thus, we can simply use a boolean value, has_vlan_filters which we
	 * will set to true when we add a VLAN filter in i40e_add_filter. Then
	 * we have to perform the full search after deleting filters in
	 * i40e_vsi_sync_filters_subtask, but we already have to search
	 * filters here and can perform the check at the same time. This
	 * results in avoiding embedding a loop for VLAN mode inside another
	 * loop over all the filters, and should maintain correctness as noted
	 * above.
J
Jesse Brandeburg 已提交
1224
	 */
1225
	return vsi->has_vlan_filter;
J
Jesse Brandeburg 已提交
1226 1227 1228 1229 1230 1231 1232 1233 1234
}

/**
 * i40e_add_filter - Add a mac/vlan filter to the VSI
 * @vsi: the VSI to be searched
 * @macaddr: the MAC address
 * @vlan: the vlan
 *
 * Returns ptr to the filter object or NULL when no memory available.
K
Kiran Patil 已提交
1235
 *
1236
 * NOTE: This function is expected to be called with mac_filter_hash_lock
K
Kiran Patil 已提交
1237
 * being held.
J
Jesse Brandeburg 已提交
1238 1239
 **/
struct i40e_mac_filter *i40e_add_filter(struct i40e_vsi *vsi,
1240
					const u8 *macaddr, s16 vlan)
J
Jesse Brandeburg 已提交
1241 1242
{
	struct i40e_mac_filter *f;
1243
	u64 key;
J
Jesse Brandeburg 已提交
1244 1245 1246 1247

	if (!vsi || !macaddr)
		return NULL;

1248 1249 1250 1251 1252 1253 1254
	/* Do not allow broadcast filter to be added since broadcast filter
	 * is added as part of add VSI for any newly created VSI except
	 * FDIR VSI
	 */
	if (is_broadcast_ether_addr(macaddr))
		return NULL;

1255
	f = i40e_find_filter(vsi, macaddr, vlan);
J
Jesse Brandeburg 已提交
1256 1257 1258
	if (!f) {
		f = kzalloc(sizeof(*f), GFP_ATOMIC);
		if (!f)
1259
			return NULL;
J
Jesse Brandeburg 已提交
1260

1261 1262 1263 1264 1265 1266
		/* Update the boolean indicating if we need to function in
		 * VLAN mode.
		 */
		if (vlan >= 0)
			vsi->has_vlan_filter = true;

1267
		ether_addr_copy(f->macaddr, macaddr);
J
Jesse Brandeburg 已提交
1268
		f->vlan = vlan;
1269 1270 1271 1272 1273 1274 1275 1276
		/* If we're in overflow promisc mode, set the state directly
		 * to failed, so we don't bother to try sending the filter
		 * to the hardware.
		 */
		if (test_bit(__I40E_FILTER_OVERFLOW_PROMISC, &vsi->state))
			f->state = I40E_FILTER_FAILED;
		else
			f->state = I40E_FILTER_NEW;
1277 1278 1279 1280
		INIT_HLIST_NODE(&f->hlist);

		key = i40e_addr_to_hkey(macaddr);
		hash_add(vsi->mac_filter_hash, &f->hlist, key);
J
Jesse Brandeburg 已提交
1281 1282 1283 1284 1285

		vsi->flags |= I40E_VSI_FLAG_FILTER_CHANGED;
		vsi->back->flags |= I40E_FLAG_FILTER_SYNC;
	}

1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296
	/* If we're asked to add a filter that has been marked for removal, it
	 * is safe to simply restore it to active state. __i40e_del_filter
	 * will have simply deleted any filters which were previously marked
	 * NEW or FAILED, so if it is currently marked REMOVE it must have
	 * previously been ACTIVE. Since we haven't yet run the sync filters
	 * task, just restore this filter to the ACTIVE state so that the
	 * sync task leaves it in place
	 */
	if (f->state == I40E_FILTER_REMOVE)
		f->state = I40E_FILTER_ACTIVE;

J
Jesse Brandeburg 已提交
1297 1298 1299 1300
	return f;
}

/**
1301 1302 1303 1304 1305 1306 1307
 * __i40e_del_filter - Remove a specific filter from the VSI
 * @vsi: VSI to remove from
 * @f: the filter to remove from the list
 *
 * This function should be called instead of i40e_del_filter only if you know
 * the exact filter you will remove already, such as via i40e_find_filter or
 * i40e_find_mac.
K
Kiran Patil 已提交
1308
 *
1309
 * NOTE: This function is expected to be called with mac_filter_hash_lock
K
Kiran Patil 已提交
1310
 * being held.
1311 1312 1313
 * ANOTHER NOTE: This function MUST be called from within the context of
 * the "safe" variants of any list iterators, e.g. list_for_each_entry_safe()
 * instead of list_for_each_entry().
J
Jesse Brandeburg 已提交
1314
 **/
1315
static void __i40e_del_filter(struct i40e_vsi *vsi, struct i40e_mac_filter *f)
J
Jesse Brandeburg 已提交
1316
{
1317
	if (!f)
J
Jesse Brandeburg 已提交
1318 1319
		return;

1320 1321 1322 1323 1324
	if ((f->state == I40E_FILTER_FAILED) ||
	    (f->state == I40E_FILTER_NEW)) {
		/* this one never got added by the FW. Just remove it,
		 * no need to sync anything.
		 */
1325
		hash_del(&f->hlist);
1326
		kfree(f);
J
Jesse Brandeburg 已提交
1327
	} else {
1328 1329 1330
		f->state = I40E_FILTER_REMOVE;
		vsi->flags |= I40E_VSI_FLAG_FILTER_CHANGED;
		vsi->back->flags |= I40E_FLAG_FILTER_SYNC;
J
Jesse Brandeburg 已提交
1331 1332 1333
	}
}

1334 1335 1336 1337 1338 1339
/**
 * i40e_del_filter - Remove a MAC/VLAN filter from the VSI
 * @vsi: the VSI to be searched
 * @macaddr: the MAC address
 * @vlan: the VLAN
 *
1340
 * NOTE: This function is expected to be called with mac_filter_hash_lock
1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356
 * being held.
 * ANOTHER NOTE: This function MUST be called from within the context of
 * the "safe" variants of any list iterators, e.g. list_for_each_entry_safe()
 * instead of list_for_each_entry().
 **/
void i40e_del_filter(struct i40e_vsi *vsi, const u8 *macaddr, s16 vlan)
{
	struct i40e_mac_filter *f;

	if (!vsi || !macaddr)
		return;

	f = i40e_find_filter(vsi, macaddr, vlan);
	__i40e_del_filter(vsi, f);
}

1357 1358 1359 1360 1361
/**
 * i40e_put_mac_in_vlan - Make macvlan filters from macaddrs and vlans
 * @vsi: the VSI to be searched
 * @macaddr: the mac address to be filtered
 *
1362 1363 1364
 * Goes through all the macvlan filters and adds a macvlan filter for each
 * unique vlan that already exists. If a PVID has been assigned, instead only
 * add the macaddr to that VLAN.
1365
 *
1366
 * Returns last filter added on success, else NULL
1367 1368 1369 1370
 **/
struct i40e_mac_filter *i40e_put_mac_in_vlan(struct i40e_vsi *vsi,
					     const u8 *macaddr)
{
1371
	struct i40e_mac_filter *f, *add = NULL;
1372 1373
	struct hlist_node *h;
	int bkt;
1374 1375 1376 1377

	if (vsi->info.pvid)
		return i40e_add_filter(vsi, macaddr,
				       le16_to_cpu(vsi->info.pvid));
1378

1379
	hash_for_each_safe(vsi->mac_filter_hash, bkt, h, f, hlist) {
1380 1381
		if (f->state == I40E_FILTER_REMOVE)
			continue;
1382 1383 1384
		add = i40e_add_filter(vsi, macaddr, f->vlan);
		if (!add)
			return NULL;
1385 1386
	}

1387
	return add;
1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400
}

/**
 * i40e_del_mac_all_vlan - Remove a MAC filter from all VLANS
 * @vsi: the VSI to be searched
 * @macaddr: the mac address to be removed
 *
 * Removes a given MAC address from a VSI, regardless of VLAN
 *
 * Returns 0 for success, or error
 **/
int i40e_del_mac_all_vlan(struct i40e_vsi *vsi, const u8 *macaddr)
{
1401 1402
	struct i40e_mac_filter *f;
	struct hlist_node *h;
1403
	bool found = false;
1404
	int bkt;
1405

1406 1407 1408
	WARN(!spin_is_locked(&vsi->mac_filter_hash_lock),
	     "Missing mac_filter_hash_lock\n");
	hash_for_each_safe(vsi->mac_filter_hash, bkt, h, f, hlist) {
1409 1410 1411 1412
		if (ether_addr_equal(macaddr, f->macaddr)) {
			__i40e_del_filter(vsi, f);
			found = true;
		}
1413
	}
1414 1415

	if (found)
1416
		return 0;
1417 1418
	else
		return -ENOENT;
1419 1420
}

J
Jesse Brandeburg 已提交
1421 1422 1423 1424 1425 1426 1427
/**
 * i40e_set_mac - NDO callback to set mac address
 * @netdev: network interface device structure
 * @p: pointer to an address structure
 *
 * Returns 0 on success, negative on failure
 **/
1428 1429 1430
#ifdef I40E_FCOE
int i40e_set_mac(struct net_device *netdev, void *p)
#else
J
Jesse Brandeburg 已提交
1431
static int i40e_set_mac(struct net_device *netdev, void *p)
1432
#endif
J
Jesse Brandeburg 已提交
1433 1434 1435
{
	struct i40e_netdev_priv *np = netdev_priv(netdev);
	struct i40e_vsi *vsi = np->vsi;
1436 1437
	struct i40e_pf *pf = vsi->back;
	struct i40e_hw *hw = &pf->hw;
J
Jesse Brandeburg 已提交
1438 1439 1440 1441 1442
	struct sockaddr *addr = p;

	if (!is_valid_ether_addr(addr->sa_data))
		return -EADDRNOTAVAIL;

1443 1444 1445 1446 1447
	if (ether_addr_equal(netdev->dev_addr, addr->sa_data)) {
		netdev_info(netdev, "already using mac address %pM\n",
			    addr->sa_data);
		return 0;
	}
J
Jesse Brandeburg 已提交
1448

1449 1450 1451 1452
	if (test_bit(__I40E_DOWN, &vsi->back->state) ||
	    test_bit(__I40E_RESET_RECOVERY_PENDING, &vsi->back->state))
		return -EADDRNOTAVAIL;

1453 1454 1455 1456 1457 1458
	if (ether_addr_equal(hw->mac.addr, addr->sa_data))
		netdev_info(netdev, "returning to hw mac address %pM\n",
			    hw->mac.addr);
	else
		netdev_info(netdev, "set new mac address %pM\n", addr->sa_data);

1459
	spin_lock_bh(&vsi->mac_filter_hash_lock);
1460 1461
	i40e_del_mac_all_vlan(vsi, netdev->dev_addr);
	i40e_put_mac_in_vlan(vsi, addr->sa_data);
1462
	spin_unlock_bh(&vsi->mac_filter_hash_lock);
1463
	ether_addr_copy(netdev->dev_addr, addr->sa_data);
J
Jesse Brandeburg 已提交
1464 1465
	if (vsi->type == I40E_VSI_MAIN) {
		i40e_status ret;
J
Jesse Brandeburg 已提交
1466

J
Jesse Brandeburg 已提交
1467
		ret = i40e_aq_mac_address_write(&vsi->back->hw,
1468
						I40E_AQC_WRITE_TYPE_LAA_WOL,
J
Jesse Brandeburg 已提交
1469
						addr->sa_data, NULL);
1470 1471 1472 1473
		if (ret)
			netdev_info(netdev, "Ignoring error from firmware on LAA update, status %s, AQ ret %s\n",
				    i40e_stat_str(hw, ret),
				    i40e_aq_str(hw, hw->aq.asq_last_status));
1474 1475
	}

1476 1477 1478 1479 1480
	/* schedule our worker thread which will take care of
	 * applying the new filter changes
	 */
	i40e_service_event_schedule(vsi->back);
	return 0;
J
Jesse Brandeburg 已提交
1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491
}

/**
 * i40e_vsi_setup_queue_map - Setup a VSI queue map based on enabled_tc
 * @vsi: the VSI being setup
 * @ctxt: VSI context structure
 * @enabled_tc: Enabled TCs bitmap
 * @is_add: True if called before Add VSI
 *
 * Setup VSI queue mapping for enabled traffic classes.
 **/
1492 1493 1494 1495 1496 1497
#ifdef I40E_FCOE
void i40e_vsi_setup_queue_map(struct i40e_vsi *vsi,
			      struct i40e_vsi_context *ctxt,
			      u8 enabled_tc,
			      bool is_add)
#else
J
Jesse Brandeburg 已提交
1498 1499 1500 1501
static void i40e_vsi_setup_queue_map(struct i40e_vsi *vsi,
				     struct i40e_vsi_context *ctxt,
				     u8 enabled_tc,
				     bool is_add)
1502
#endif
J
Jesse Brandeburg 已提交
1503 1504 1505 1506 1507 1508 1509 1510 1511
{
	struct i40e_pf *pf = vsi->back;
	u16 sections = 0;
	u8 netdev_tc = 0;
	u16 numtc = 0;
	u16 qcount;
	u8 offset;
	u16 qmap;
	int i;
N
Neerav Parikh 已提交
1512
	u16 num_tc_qps = 0;
J
Jesse Brandeburg 已提交
1513 1514 1515 1516 1517 1518 1519

	sections = I40E_AQ_VSI_PROP_QUEUE_MAP_VALID;
	offset = 0;

	if (enabled_tc && (vsi->back->flags & I40E_FLAG_DCB_ENABLED)) {
		/* Find numtc from enabled TC bitmap */
		for (i = 0; i < I40E_MAX_TRAFFIC_CLASS; i++) {
J
Jesse Brandeburg 已提交
1520
			if (enabled_tc & BIT(i)) /* TC is enabled */
J
Jesse Brandeburg 已提交
1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533
				numtc++;
		}
		if (!numtc) {
			dev_warn(&pf->pdev->dev, "DCB is enabled but no TC enabled, forcing TC0\n");
			numtc = 1;
		}
	} else {
		/* At least TC0 is enabled in case of non-DCB case */
		numtc = 1;
	}

	vsi->tc_config.numtc = numtc;
	vsi->tc_config.enabled_tc = enabled_tc ? enabled_tc : 1;
N
Neerav Parikh 已提交
1534
	/* Number of queues per enabled TC */
1535 1536
	qcount = vsi->alloc_queue_pairs;

1537
	num_tc_qps = qcount / numtc;
1538
	num_tc_qps = min_t(int, num_tc_qps, i40e_pf_get_max_q_per_tc(pf));
J
Jesse Brandeburg 已提交
1539 1540 1541 1542

	/* Setup queue offset/count for all TCs for given VSI */
	for (i = 0; i < I40E_MAX_TRAFFIC_CLASS; i++) {
		/* See if the given TC is enabled for the given VSI */
J
Jesse Brandeburg 已提交
1543
		if (vsi->tc_config.enabled_tc & BIT(i)) {
1544
			/* TC is enabled */
J
Jesse Brandeburg 已提交
1545 1546 1547 1548
			int pow, num_qps;

			switch (vsi->type) {
			case I40E_VSI_MAIN:
1549 1550
				qcount = min_t(int, pf->alloc_rss_size,
					       num_tc_qps);
J
Jesse Brandeburg 已提交
1551
				break;
1552 1553 1554 1555 1556
#ifdef I40E_FCOE
			case I40E_VSI_FCOE:
				qcount = num_tc_qps;
				break;
#endif
J
Jesse Brandeburg 已提交
1557 1558 1559 1560
			case I40E_VSI_FDIR:
			case I40E_VSI_SRIOV:
			case I40E_VSI_VMDQ2:
			default:
N
Neerav Parikh 已提交
1561
				qcount = num_tc_qps;
J
Jesse Brandeburg 已提交
1562 1563 1564
				WARN_ON(i != 0);
				break;
			}
N
Neerav Parikh 已提交
1565 1566
			vsi->tc_config.tc_info[i].qoffset = offset;
			vsi->tc_config.tc_info[i].qcount = qcount;
J
Jesse Brandeburg 已提交
1567

S
Shannon Nelson 已提交
1568
			/* find the next higher power-of-2 of num queue pairs */
N
Neerav Parikh 已提交
1569
			num_qps = qcount;
J
Jesse Brandeburg 已提交
1570
			pow = 0;
1571
			while (num_qps && (BIT_ULL(pow) < qcount)) {
J
Jesse Brandeburg 已提交
1572 1573 1574 1575 1576 1577 1578 1579 1580
				pow++;
				num_qps >>= 1;
			}

			vsi->tc_config.tc_info[i].netdev_tc = netdev_tc++;
			qmap =
			    (offset << I40E_AQ_VSI_TC_QUE_OFFSET_SHIFT) |
			    (pow << I40E_AQ_VSI_TC_QUE_NUMBER_SHIFT);

N
Neerav Parikh 已提交
1581
			offset += qcount;
J
Jesse Brandeburg 已提交
1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597
		} else {
			/* TC is not enabled so set the offset to
			 * default queue and allocate one queue
			 * for the given TC.
			 */
			vsi->tc_config.tc_info[i].qoffset = 0;
			vsi->tc_config.tc_info[i].qcount = 1;
			vsi->tc_config.tc_info[i].netdev_tc = 0;

			qmap = 0;
		}
		ctxt->info.tc_mapping[i] = cpu_to_le16(qmap);
	}

	/* Set actual Tx/Rx queue pairs */
	vsi->num_queue_pairs = offset;
1598 1599 1600
	if ((vsi->type == I40E_VSI_MAIN) && (numtc == 1)) {
		if (vsi->req_queue_pairs > 0)
			vsi->num_queue_pairs = vsi->req_queue_pairs;
1601
		else if (pf->flags & I40E_FLAG_MSIX_ENABLED)
1602 1603
			vsi->num_queue_pairs = pf->num_lan_msix;
	}
J
Jesse Brandeburg 已提交
1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624

	/* Scheduler section valid can only be set for ADD VSI */
	if (is_add) {
		sections |= I40E_AQ_VSI_PROP_SCHED_VALID;

		ctxt->info.up_enable_bits = enabled_tc;
	}
	if (vsi->type == I40E_VSI_SRIOV) {
		ctxt->info.mapping_flags |=
				     cpu_to_le16(I40E_AQ_VSI_QUE_MAP_NONCONTIG);
		for (i = 0; i < vsi->num_queue_pairs; i++)
			ctxt->info.queue_mapping[i] =
					       cpu_to_le16(vsi->base_queue + i);
	} else {
		ctxt->info.mapping_flags |=
					cpu_to_le16(I40E_AQ_VSI_QUE_MAP_CONTIG);
		ctxt->info.queue_mapping[0] = cpu_to_le16(vsi->base_queue);
	}
	ctxt->info.valid_sections |= cpu_to_le16(sections);
}

1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670
/**
 * i40e_addr_sync - Callback for dev_(mc|uc)_sync to add address
 * @netdev: the netdevice
 * @addr: address to add
 *
 * Called by __dev_(mc|uc)_sync when an address needs to be added. We call
 * __dev_(uc|mc)_sync from .set_rx_mode and guarantee to hold the hash lock.
 */
static int i40e_addr_sync(struct net_device *netdev, const u8 *addr)
{
	struct i40e_netdev_priv *np = netdev_priv(netdev);
	struct i40e_vsi *vsi = np->vsi;
	struct i40e_mac_filter *f;

	if (i40e_is_vsi_in_vlan(vsi))
		f = i40e_put_mac_in_vlan(vsi, addr);
	else
		f = i40e_add_filter(vsi, addr, I40E_VLAN_ANY);

	if (f)
		return 0;
	else
		return -ENOMEM;
}

/**
 * i40e_addr_unsync - Callback for dev_(mc|uc)_sync to remove address
 * @netdev: the netdevice
 * @addr: address to add
 *
 * Called by __dev_(mc|uc)_sync when an address needs to be removed. We call
 * __dev_(uc|mc)_sync from .set_rx_mode and guarantee to hold the hash lock.
 */
static int i40e_addr_unsync(struct net_device *netdev, const u8 *addr)
{
	struct i40e_netdev_priv *np = netdev_priv(netdev);
	struct i40e_vsi *vsi = np->vsi;

	if (i40e_is_vsi_in_vlan(vsi))
		i40e_del_mac_all_vlan(vsi, addr);
	else
		i40e_del_filter(vsi, addr, I40E_VLAN_ANY);

	return 0;
}

J
Jesse Brandeburg 已提交
1671 1672 1673 1674
/**
 * i40e_set_rx_mode - NDO callback to set the netdev filters
 * @netdev: network interface device structure
 **/
1675 1676 1677
#ifdef I40E_FCOE
void i40e_set_rx_mode(struct net_device *netdev)
#else
J
Jesse Brandeburg 已提交
1678
static void i40e_set_rx_mode(struct net_device *netdev)
1679
#endif
J
Jesse Brandeburg 已提交
1680 1681 1682 1683
{
	struct i40e_netdev_priv *np = netdev_priv(netdev);
	struct i40e_vsi *vsi = np->vsi;

1684
	spin_lock_bh(&vsi->mac_filter_hash_lock);
K
Kiran Patil 已提交
1685

1686 1687
	__dev_uc_sync(netdev, i40e_addr_sync, i40e_addr_unsync);
	__dev_mc_sync(netdev, i40e_addr_sync, i40e_addr_unsync);
J
Jesse Brandeburg 已提交
1688

1689
	spin_unlock_bh(&vsi->mac_filter_hash_lock);
J
Jesse Brandeburg 已提交
1690 1691 1692 1693 1694 1695

	/* check for other flag changes */
	if (vsi->current_netdev_flags != vsi->netdev->flags) {
		vsi->flags |= I40E_VSI_FLAG_FILTER_CHANGED;
		vsi->back->flags |= I40E_FLAG_FILTER_SYNC;
	}
1696 1697 1698 1699 1700

	/* schedule our worker thread which will take care of
	 * applying the new filter changes
	 */
	i40e_service_event_schedule(vsi->back);
J
Jesse Brandeburg 已提交
1701 1702
}

K
Kiran Patil 已提交
1703
/**
1704 1705
 * i40e_undo_filter_entries - Undo the changes made to MAC filter entries
 * @vsi: Pointer to VSI struct
K
Kiran Patil 已提交
1706 1707 1708
 * @from: Pointer to list which contains MAC filter entries - changes to
 *        those entries needs to be undone.
 *
1709 1710
 * MAC filter entries from list were slated to be sent to firmware, either for
 * addition or deletion.
K
Kiran Patil 已提交
1711
 **/
1712 1713
static void i40e_undo_filter_entries(struct i40e_vsi *vsi,
				     struct hlist_head *from)
K
Kiran Patil 已提交
1714
{
1715 1716 1717 1718 1719
	struct i40e_mac_filter *f;
	struct hlist_node *h;

	hlist_for_each_entry_safe(f, h, from, hlist) {
		u64 key = i40e_addr_to_hkey(f->macaddr);
K
Kiran Patil 已提交
1720 1721

		/* Move the element back into MAC filter list*/
1722 1723
		hlist_del(&f->hlist);
		hash_add(vsi->mac_filter_hash, &f->hlist, key);
K
Kiran Patil 已提交
1724 1725 1726 1727
	}
}

/**
1728 1729 1730 1731 1732 1733
 * i40e_update_filter_state - Update filter state based on return data
 * from firmware
 * @count: Number of filters added
 * @add_list: return data from fw
 * @head: pointer to first filter in current batch
 * @aq_err: status from fw
K
Kiran Patil 已提交
1734
 *
1735 1736
 * MAC filter entries from list were slated to be added to device. Returns
 * number of successful filters. Note that 0 does NOT mean success!
K
Kiran Patil 已提交
1737
 **/
1738 1739 1740 1741
static int
i40e_update_filter_state(int count,
			 struct i40e_aqc_add_macvlan_element_data *add_list,
			 struct i40e_mac_filter *add_head, int aq_err)
K
Kiran Patil 已提交
1742
{
1743 1744
	int retval = 0;
	int i;
K
Kiran Patil 已提交
1745 1746


1747 1748 1749 1750 1751
	if (!aq_err) {
		retval = count;
		/* Everything's good, mark all filters active. */
		for (i = 0; i < count ; i++) {
			add_head->state = I40E_FILTER_ACTIVE;
1752 1753 1754
			add_head = hlist_entry(add_head->hlist.next,
					       typeof(struct i40e_mac_filter),
					       hlist);
1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767
		}
	} else if (aq_err == I40E_AQ_RC_ENOSPC) {
		/* Device ran out of filter space. Check the return value
		 * for each filter to see which ones are active.
		 */
		for (i = 0; i < count ; i++) {
			if (add_list[i].match_method ==
			    I40E_AQC_MM_ERR_NO_RES) {
				add_head->state = I40E_FILTER_FAILED;
			} else {
				add_head->state = I40E_FILTER_ACTIVE;
				retval++;
			}
1768 1769 1770
			add_head = hlist_entry(add_head->hlist.next,
					       typeof(struct i40e_mac_filter),
					       hlist);
1771 1772 1773 1774 1775 1776
		}
	} else {
		/* Some other horrible thing happened, fail all filters */
		retval = 0;
		for (i = 0; i < count ; i++) {
			add_head->state = I40E_FILTER_FAILED;
1777 1778 1779
			add_head = hlist_entry(add_head->hlist.next,
					       typeof(struct i40e_mac_filter),
					       hlist);
1780
		}
K
Kiran Patil 已提交
1781
	}
1782
	return retval;
K
Kiran Patil 已提交
1783 1784
}

1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858
/**
 * i40e_aqc_del_filters - Request firmware to delete a set of filters
 * @vsi: ptr to the VSI
 * @vsi_name: name to display in messages
 * @list: the list of filters to send to firmware
 * @num_del: the number of filters to delete
 * @retval: Set to -EIO on failure to delete
 *
 * Send a request to firmware via AdminQ to delete a set of filters. Uses
 * *retval instead of a return value so that success does not force ret_val to
 * be set to 0. This ensures that a sequence of calls to this function
 * preserve the previous value of *retval on successful delete.
 */
static
void i40e_aqc_del_filters(struct i40e_vsi *vsi, const char *vsi_name,
			  struct i40e_aqc_remove_macvlan_element_data *list,
			  int num_del, int *retval)
{
	struct i40e_hw *hw = &vsi->back->hw;
	i40e_status aq_ret;
	int aq_err;

	aq_ret = i40e_aq_remove_macvlan(hw, vsi->seid, list, num_del, NULL);
	aq_err = hw->aq.asq_last_status;

	/* Explicitly ignore and do not report when firmware returns ENOENT */
	if (aq_ret && !(aq_err == I40E_AQ_RC_ENOENT)) {
		*retval = -EIO;
		dev_info(&vsi->back->pdev->dev,
			 "ignoring delete macvlan error on %s, err %s, aq_err %s\n",
			 vsi_name, i40e_stat_str(hw, aq_ret),
			 i40e_aq_str(hw, aq_err));
	}
}

/**
 * i40e_aqc_add_filters - Request firmware to add a set of filters
 * @vsi: ptr to the VSI
 * @vsi_name: name to display in messages
 * @list: the list of filters to send to firmware
 * @add_head: Position in the add hlist
 * @num_add: the number of filters to add
 * @promisc_change: set to true on exit if promiscuous mode was forced on
 *
 * Send a request to firmware via AdminQ to add a chunk of filters. Will set
 * promisc_changed to true if the firmware has run out of space for more
 * filters.
 */
static
void i40e_aqc_add_filters(struct i40e_vsi *vsi, const char *vsi_name,
			  struct i40e_aqc_add_macvlan_element_data *list,
			  struct i40e_mac_filter *add_head,
			  int num_add, bool *promisc_changed)
{
	struct i40e_hw *hw = &vsi->back->hw;
	i40e_status aq_ret;
	int aq_err, fcnt;

	aq_ret = i40e_aq_add_macvlan(hw, vsi->seid, list, num_add, NULL);
	aq_err = hw->aq.asq_last_status;
	fcnt = i40e_update_filter_state(num_add, list, add_head, aq_ret);
	vsi->active_filters += fcnt;

	if (fcnt != num_add) {
		*promisc_changed = true;
		set_bit(__I40E_FILTER_OVERFLOW_PROMISC, &vsi->state);
		vsi->promisc_threshold = (vsi->active_filters * 3) / 4;
		dev_warn(&vsi->back->pdev->dev,
			 "Error %s adding RX filters on %s, promiscuous mode forced on\n",
			 i40e_aq_str(hw, aq_err),
			 vsi_name);
	}
}

J
Jesse Brandeburg 已提交
1859 1860 1861 1862 1863 1864 1865 1866
/**
 * i40e_sync_vsi_filters - Update the VSI filter list to the HW
 * @vsi: ptr to the VSI
 *
 * Push any outstanding VSI filter changes through the AdminQ.
 *
 * Returns 0 or error value
 **/
J
Jesse Brandeburg 已提交
1867
int i40e_sync_vsi_filters(struct i40e_vsi *vsi)
J
Jesse Brandeburg 已提交
1868
{
1869
	struct hlist_head tmp_add_list, tmp_del_list;
1870
	struct i40e_mac_filter *f, *add_head = NULL;
1871
	struct i40e_hw *hw = &vsi->back->hw;
1872 1873 1874
	unsigned int vlan_any_filters = 0;
	unsigned int non_vlan_filters = 0;
	unsigned int vlan_filters = 0;
1875
	bool promisc_changed = false;
1876
	char vsi_name[16] = "PF";
J
Jesse Brandeburg 已提交
1877
	int filter_list_len = 0;
M
Mitch Williams 已提交
1878
	i40e_status aq_ret = 0;
1879
	u32 changed_flags = 0;
1880
	struct hlist_node *h;
J
Jesse Brandeburg 已提交
1881 1882 1883
	struct i40e_pf *pf;
	int num_add = 0;
	int num_del = 0;
1884
	int retval = 0;
J
Jesse Brandeburg 已提交
1885
	u16 cmd_flags;
1886
	int list_size;
1887
	int bkt;
J
Jesse Brandeburg 已提交
1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901

	/* empty array typed pointers, kcalloc later */
	struct i40e_aqc_add_macvlan_element_data *add_list;
	struct i40e_aqc_remove_macvlan_element_data *del_list;

	while (test_and_set_bit(__I40E_CONFIG_BUSY, &vsi->state))
		usleep_range(1000, 2000);
	pf = vsi->back;

	if (vsi->netdev) {
		changed_flags = vsi->current_netdev_flags ^ vsi->netdev->flags;
		vsi->current_netdev_flags = vsi->netdev->flags;
	}

1902 1903
	INIT_HLIST_HEAD(&tmp_add_list);
	INIT_HLIST_HEAD(&tmp_del_list);
K
Kiran Patil 已提交
1904

1905 1906 1907 1908 1909
	if (vsi->type == I40E_VSI_SRIOV)
		snprintf(vsi_name, sizeof(vsi_name) - 1, "VF %d", vsi->vf_id);
	else if (vsi->type != I40E_VSI_MAIN)
		snprintf(vsi_name, sizeof(vsi_name) - 1, "vsi %d", vsi->seid);

J
Jesse Brandeburg 已提交
1910 1911 1912
	if (vsi->flags & I40E_VSI_FLAG_FILTER_CHANGED) {
		vsi->flags &= ~I40E_VSI_FLAG_FILTER_CHANGED;

1913
		spin_lock_bh(&vsi->mac_filter_hash_lock);
1914
		/* Create a list of filters to delete. */
1915
		hash_for_each_safe(vsi->mac_filter_hash, bkt, h, f, hlist) {
1916 1917
			if (f->state == I40E_FILTER_REMOVE) {
				/* Move the element into temporary del_list */
1918 1919
				hash_del(&f->hlist);
				hlist_add_head(&f->hlist, &tmp_del_list);
1920
				vsi->active_filters--;
1921 1922 1923

				/* Avoid counting removed filters */
				continue;
1924 1925
			}
			if (f->state == I40E_FILTER_NEW) {
1926 1927
				hash_del(&f->hlist);
				hlist_add_head(&f->hlist, &tmp_add_list);
K
Kiran Patil 已提交
1928
			}
1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989

			/* Count the number of each type of filter we have
			 * remaining, ignoring any filters we're about to
			 * delete.
			 */
			if (f->vlan > 0)
				vlan_filters++;
			else if (!f->vlan)
				non_vlan_filters++;
			else
				vlan_any_filters++;
		}

		/* We should never have VLAN=-1 filters at the same time as we
		 * have either VLAN=0 or VLAN>0 filters, so warn about this
		 * case here to help catch any issues.
		 */
		WARN_ON(vlan_any_filters && (vlan_filters + non_vlan_filters));

		/* If we only have VLAN=0 filters remaining, and don't have
		 * any other VLAN filters, we need to convert these VLAN=0
		 * filters into VLAN=-1 (I40E_VLAN_ANY) so that we operate
		 * correctly in non-VLAN mode and receive all traffic tagged
		 * or untagged.
		 */
		if (non_vlan_filters && !vlan_filters) {
			hash_for_each_safe(vsi->mac_filter_hash, bkt, h, f,
					   hlist) {
				/* Only replace VLAN=0 filters */
				if (f->vlan)
					continue;

				/* Allocate a replacement element */
				add_head = kzalloc(sizeof(*add_head),
						   GFP_KERNEL);
				if (!add_head)
					goto err_no_memory_locked;

				/* Copy the filter, with new state and VLAN */
				*add_head = *f;
				add_head->state = I40E_FILTER_NEW;
				add_head->vlan = I40E_VLAN_ANY;

				/* Move the replacement to the add list */
				INIT_HLIST_NODE(&add_head->hlist);
				hlist_add_head(&add_head->hlist,
					       &tmp_add_list);

				/* Move the original to the delete list */
				f->state = I40E_FILTER_REMOVE;
				hash_del(&f->hlist);
				hlist_add_head(&f->hlist, &tmp_del_list);
				vsi->active_filters--;
			}

			/* Also update any filters on the tmp_add list */
			hlist_for_each_entry(f, &tmp_add_list, hlist) {
				if (!f->vlan)
					f->vlan = I40E_VLAN_ANY;
			}
			add_head = NULL;
K
Kiran Patil 已提交
1990
		}
1991
		spin_unlock_bh(&vsi->mac_filter_hash_lock);
K
Kiran Patil 已提交
1992 1993 1994
	}

	/* Now process 'del_list' outside the lock */
1995
	if (!hlist_empty(&tmp_del_list)) {
1996
		filter_list_len = hw->aq.asq_buf_size /
K
Kiran Patil 已提交
1997
			    sizeof(struct i40e_aqc_remove_macvlan_element_data);
1998
		list_size = filter_list_len *
S
Shannon Nelson 已提交
1999
			    sizeof(struct i40e_aqc_remove_macvlan_element_data);
2000
		del_list = kzalloc(list_size, GFP_ATOMIC);
2001 2002
		if (!del_list)
			goto err_no_memory;
K
Kiran Patil 已提交
2003

2004
		hlist_for_each_entry_safe(f, h, &tmp_del_list, hlist) {
J
Jesse Brandeburg 已提交
2005 2006 2007
			cmd_flags = 0;

			/* add to delete list */
2008
			ether_addr_copy(del_list[num_del].mac_addr, f->macaddr);
2009 2010
			if (f->vlan == I40E_VLAN_ANY) {
				del_list[num_del].vlan_tag = 0;
A
Alan Brady 已提交
2011
				cmd_flags |= I40E_AQC_MACVLAN_DEL_IGNORE_VLAN;
2012 2013 2014 2015
			} else {
				del_list[num_del].vlan_tag =
					cpu_to_le16((u16)(f->vlan));
			}
J
Jesse Brandeburg 已提交
2016 2017 2018 2019 2020 2021 2022

			cmd_flags |= I40E_AQC_MACVLAN_DEL_PERFECT_MATCH;
			del_list[num_del].flags = cmd_flags;
			num_del++;

			/* flush a full buffer */
			if (num_del == filter_list_len) {
2023 2024
				i40e_aqc_del_filters(vsi, vsi_name, del_list,
						     num_del, &retval);
2025
				memset(del_list, 0, list_size);
2026
				num_del = 0;
J
Jesse Brandeburg 已提交
2027
			}
K
Kiran Patil 已提交
2028 2029 2030
			/* Release memory for MAC filter entries which were
			 * synced up with HW.
			 */
2031
			hlist_del(&f->hlist);
K
Kiran Patil 已提交
2032
			kfree(f);
J
Jesse Brandeburg 已提交
2033
		}
K
Kiran Patil 已提交
2034

J
Jesse Brandeburg 已提交
2035
		if (num_del) {
2036 2037
			i40e_aqc_del_filters(vsi, vsi_name, del_list,
					     num_del, &retval);
J
Jesse Brandeburg 已提交
2038 2039 2040 2041
		}

		kfree(del_list);
		del_list = NULL;
K
Kiran Patil 已提交
2042 2043
	}

2044 2045 2046 2047 2048 2049 2050 2051
	/* After finishing notifying firmware of the deleted filters, update
	 * the cached value of vsi->has_vlan_filter. Note that we are safe to
	 * use just !!vlan_filters here because if we only have VLAN=0 (that
	 * is, non_vlan_filters) these will all be converted to VLAN=-1 in the
	 * logic above already so this value would still be correct.
	 */
	vsi->has_vlan_filter = !!vlan_filters;

2052
	if (!hlist_empty(&tmp_add_list)) {
2053
		/* Do all the adds now. */
2054
		filter_list_len = hw->aq.asq_buf_size /
S
Shannon Nelson 已提交
2055
			       sizeof(struct i40e_aqc_add_macvlan_element_data);
2056 2057 2058
		list_size = filter_list_len *
			       sizeof(struct i40e_aqc_add_macvlan_element_data);
		add_list = kzalloc(list_size, GFP_ATOMIC);
2059 2060 2061
		if (!add_list)
			goto err_no_memory;

2062
		num_add = 0;
2063
		hlist_for_each_entry(f, &tmp_add_list, hlist) {
2064 2065 2066 2067 2068
			if (test_bit(__I40E_FILTER_OVERFLOW_PROMISC,
				     &vsi->state)) {
				f->state = I40E_FILTER_FAILED;
				continue;
			}
J
Jesse Brandeburg 已提交
2069
			/* add to add array */
2070 2071 2072
			if (num_add == 0)
				add_head = f;
			cmd_flags = 0;
2073
			ether_addr_copy(add_list[num_add].mac_addr, f->macaddr);
2074 2075 2076 2077 2078 2079 2080
			if (f->vlan == I40E_VLAN_ANY) {
				add_list[num_add].vlan_tag = 0;
				cmd_flags |= I40E_AQC_MACVLAN_ADD_IGNORE_VLAN;
			} else {
				add_list[num_add].vlan_tag =
					cpu_to_le16((u16)(f->vlan));
			}
J
Jesse Brandeburg 已提交
2081 2082 2083 2084 2085 2086 2087
			add_list[num_add].queue_number = 0;
			cmd_flags |= I40E_AQC_MACVLAN_ADD_PERFECT_MATCH;
			add_list[num_add].flags = cpu_to_le16(cmd_flags);
			num_add++;

			/* flush a full buffer */
			if (num_add == filter_list_len) {
2088 2089 2090
				i40e_aqc_add_filters(vsi, vsi_name, add_list,
						     add_head, num_add,
						     &promisc_changed);
2091
				memset(add_list, 0, list_size);
J
Jesse Brandeburg 已提交
2092 2093 2094 2095
				num_add = 0;
			}
		}
		if (num_add) {
2096 2097
			i40e_aqc_add_filters(vsi, vsi_name, add_list, add_head,
					     num_add, &promisc_changed);
J
Jesse Brandeburg 已提交
2098
		}
2099 2100 2101
		/* Now move all of the filters from the temp add list back to
		 * the VSI's list.
		 */
2102 2103 2104 2105 2106 2107
		spin_lock_bh(&vsi->mac_filter_hash_lock);
		hlist_for_each_entry_safe(f, h, &tmp_add_list, hlist) {
			u64 key = i40e_addr_to_hkey(f->macaddr);

			hlist_del(&f->hlist);
			hash_add(vsi->mac_filter_hash, &f->hlist, key);
2108
		}
2109
		spin_unlock_bh(&vsi->mac_filter_hash_lock);
J
Jesse Brandeburg 已提交
2110 2111
		kfree(add_list);
		add_list = NULL;
2112
	}
J
Jesse Brandeburg 已提交
2113

2114 2115 2116 2117 2118 2119 2120
	/* Check to see if we can drop out of overflow promiscuous mode. */
	if (test_bit(__I40E_FILTER_OVERFLOW_PROMISC, &vsi->state) &&
	    (vsi->active_filters < vsi->promisc_threshold)) {
		int failed_count = 0;
		/* See if we have any failed filters. We can't drop out of
		 * promiscuous until these have all been deleted.
		 */
2121 2122
		spin_lock_bh(&vsi->mac_filter_hash_lock);
		hash_for_each(vsi->mac_filter_hash, bkt, f, hlist) {
2123 2124 2125
			if (f->state == I40E_FILTER_FAILED)
				failed_count++;
		}
2126
		spin_unlock_bh(&vsi->mac_filter_hash_lock);
2127
		if (!failed_count) {
J
Jesse Brandeburg 已提交
2128
			dev_info(&pf->pdev->dev,
2129 2130 2131 2132 2133
				 "filter logjam cleared on %s, leaving overflow promiscuous mode\n",
				 vsi_name);
			clear_bit(__I40E_FILTER_OVERFLOW_PROMISC, &vsi->state);
			promisc_changed = true;
			vsi->promisc_threshold = 0;
J
Jesse Brandeburg 已提交
2134 2135 2136
		}
	}

2137 2138 2139 2140 2141 2142
	/* if the VF is not trusted do not do promisc */
	if ((vsi->type == I40E_VSI_SRIOV) && !pf->vf[vsi->vf_id].trusted) {
		clear_bit(__I40E_FILTER_OVERFLOW_PROMISC, &vsi->state);
		goto out;
	}

J
Jesse Brandeburg 已提交
2143 2144 2145
	/* check for changes in promiscuous modes */
	if (changed_flags & IFF_ALLMULTI) {
		bool cur_multipromisc;
J
Jesse Brandeburg 已提交
2146

J
Jesse Brandeburg 已提交
2147
		cur_multipromisc = !!(vsi->current_netdev_flags & IFF_ALLMULTI);
M
Mitch Williams 已提交
2148 2149 2150 2151 2152 2153
		aq_ret = i40e_aq_set_vsi_multicast_promiscuous(&vsi->back->hw,
							       vsi->seid,
							       cur_multipromisc,
							       NULL);
		if (aq_ret) {
			retval = i40e_aq_rc_to_posix(aq_ret,
2154
						     hw->aq.asq_last_status);
J
Jesse Brandeburg 已提交
2155
			dev_info(&pf->pdev->dev,
2156 2157
				 "set multi promisc failed on %s, err %s aq_err %s\n",
				 vsi_name,
2158 2159
				 i40e_stat_str(hw, aq_ret),
				 i40e_aq_str(hw, hw->aq.asq_last_status));
M
Mitch Williams 已提交
2160
		}
J
Jesse Brandeburg 已提交
2161
	}
2162 2163 2164
	if ((changed_flags & IFF_PROMISC) ||
	    (promisc_changed &&
	     test_bit(__I40E_FILTER_OVERFLOW_PROMISC, &vsi->state))) {
J
Jesse Brandeburg 已提交
2165
		bool cur_promisc;
J
Jesse Brandeburg 已提交
2166

J
Jesse Brandeburg 已提交
2167 2168 2169
		cur_promisc = (!!(vsi->current_netdev_flags & IFF_PROMISC) ||
			       test_bit(__I40E_FILTER_OVERFLOW_PROMISC,
					&vsi->state));
2170 2171 2172
		if ((vsi->type == I40E_VSI_MAIN) &&
		    (pf->lan_veb != I40E_NO_VEB) &&
		    !(pf->flags & I40E_FLAG_MFP_ENABLED)) {
2173 2174 2175 2176 2177 2178 2179
			/* set defport ON for Main VSI instead of true promisc
			 * this way we will get all unicast/multicast and VLAN
			 * promisc behavior but will not get VF or VMDq traffic
			 * replicated on the Main VSI.
			 */
			if (pf->cur_promisc != cur_promisc) {
				pf->cur_promisc = cur_promisc;
2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193
				if (cur_promisc)
					aq_ret =
					      i40e_aq_set_default_vsi(hw,
								      vsi->seid,
								      NULL);
				else
					aq_ret =
					    i40e_aq_clear_default_vsi(hw,
								      vsi->seid,
								      NULL);
				if (aq_ret) {
					retval = i40e_aq_rc_to_posix(aq_ret,
							hw->aq.asq_last_status);
					dev_info(&pf->pdev->dev,
2194 2195
						 "Set default VSI failed on %s, err %s, aq_err %s\n",
						 vsi_name,
2196 2197 2198 2199
						 i40e_stat_str(hw, aq_ret),
						 i40e_aq_str(hw,
						     hw->aq.asq_last_status));
				}
2200 2201
			}
		} else {
M
Mitch Williams 已提交
2202
			aq_ret = i40e_aq_set_vsi_unicast_promiscuous(
2203
							  hw,
2204
							  vsi->seid,
2205 2206
							  cur_promisc, NULL,
							  true);
M
Mitch Williams 已提交
2207 2208 2209
			if (aq_ret) {
				retval =
				i40e_aq_rc_to_posix(aq_ret,
2210
						    hw->aq.asq_last_status);
2211
				dev_info(&pf->pdev->dev,
2212 2213
					 "set unicast promisc failed on %s, err %s, aq_err %s\n",
					 vsi_name,
2214 2215 2216
					 i40e_stat_str(hw, aq_ret),
					 i40e_aq_str(hw,
						     hw->aq.asq_last_status));
M
Mitch Williams 已提交
2217 2218
			}
			aq_ret = i40e_aq_set_vsi_multicast_promiscuous(
2219
							  hw,
2220 2221
							  vsi->seid,
							  cur_promisc, NULL);
M
Mitch Williams 已提交
2222 2223 2224
			if (aq_ret) {
				retval =
				i40e_aq_rc_to_posix(aq_ret,
2225
						    hw->aq.asq_last_status);
2226
				dev_info(&pf->pdev->dev,
2227 2228
					 "set multicast promisc failed on %s, err %s, aq_err %s\n",
					 vsi_name,
2229 2230 2231
					 i40e_stat_str(hw, aq_ret),
					 i40e_aq_str(hw,
						     hw->aq.asq_last_status));
M
Mitch Williams 已提交
2232
			}
2233
		}
M
Mitch Williams 已提交
2234 2235 2236 2237 2238 2239
		aq_ret = i40e_aq_set_vsi_broadcast(&vsi->back->hw,
						   vsi->seid,
						   cur_promisc, NULL);
		if (aq_ret) {
			retval = i40e_aq_rc_to_posix(aq_ret,
						     pf->hw.aq.asq_last_status);
2240
			dev_info(&pf->pdev->dev,
2241
				 "set brdcast promisc failed, err %s, aq_err %s\n",
2242 2243 2244
					 i40e_stat_str(hw, aq_ret),
					 i40e_aq_str(hw,
						     hw->aq.asq_last_status));
M
Mitch Williams 已提交
2245
		}
J
Jesse Brandeburg 已提交
2246
	}
M
Mitch Williams 已提交
2247
out:
2248 2249 2250 2251
	/* if something went wrong then set the changed flag so we try again */
	if (retval)
		vsi->flags |= I40E_VSI_FLAG_FILTER_CHANGED;

J
Jesse Brandeburg 已提交
2252
	clear_bit(__I40E_CONFIG_BUSY, &vsi->state);
M
Mitch Williams 已提交
2253
	return retval;
2254 2255 2256 2257

err_no_memory:
	/* Restore elements on the temporary add and delete lists */
	spin_lock_bh(&vsi->mac_filter_hash_lock);
2258
err_no_memory_locked:
2259 2260 2261 2262 2263 2264 2265
	i40e_undo_filter_entries(vsi, &tmp_del_list);
	i40e_undo_filter_entries(vsi, &tmp_add_list);
	spin_unlock_bh(&vsi->mac_filter_hash_lock);

	vsi->flags |= I40E_VSI_FLAG_FILTER_CHANGED;
	clear_bit(__I40E_CONFIG_BUSY, &vsi->state);
	return -ENOMEM;
J
Jesse Brandeburg 已提交
2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279
}

/**
 * i40e_sync_filters_subtask - Sync the VSI filter list with HW
 * @pf: board private structure
 **/
static void i40e_sync_filters_subtask(struct i40e_pf *pf)
{
	int v;

	if (!pf || !(pf->flags & I40E_FLAG_FILTER_SYNC))
		return;
	pf->flags &= ~I40E_FLAG_FILTER_SYNC;

M
Mitch Williams 已提交
2280
	for (v = 0; v < pf->num_alloc_vsi; v++) {
J
Jesse Brandeburg 已提交
2281
		if (pf->vsi[v] &&
J
Jesse Brandeburg 已提交
2282 2283 2284 2285 2286 2287 2288 2289 2290
		    (pf->vsi[v]->flags & I40E_VSI_FLAG_FILTER_CHANGED)) {
			int ret = i40e_sync_vsi_filters(pf->vsi[v]);

			if (ret) {
				/* come back and try again later */
				pf->flags |= I40E_FLAG_FILTER_SYNC;
				break;
			}
		}
J
Jesse Brandeburg 已提交
2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310
	}
}

/**
 * i40e_change_mtu - NDO callback to change the Maximum Transfer Unit
 * @netdev: network interface device structure
 * @new_mtu: new value for maximum frame size
 *
 * Returns 0 on success, negative on failure
 **/
static int i40e_change_mtu(struct net_device *netdev, int new_mtu)
{
	struct i40e_netdev_priv *np = netdev_priv(netdev);
	struct i40e_vsi *vsi = np->vsi;

	netdev_info(netdev, "changing MTU from %d to %d\n",
		    netdev->mtu, new_mtu);
	netdev->mtu = new_mtu;
	if (netif_running(netdev))
		i40e_vsi_reinit_locked(vsi);
2311
	i40e_notify_client_of_l2_param_changes(vsi);
J
Jesse Brandeburg 已提交
2312 2313 2314
	return 0;
}

J
Jacob Keller 已提交
2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 2331 2332 2333 2334 2335
/**
 * i40e_ioctl - Access the hwtstamp interface
 * @netdev: network interface device structure
 * @ifr: interface request data
 * @cmd: ioctl command
 **/
int i40e_ioctl(struct net_device *netdev, struct ifreq *ifr, int cmd)
{
	struct i40e_netdev_priv *np = netdev_priv(netdev);
	struct i40e_pf *pf = np->vsi->back;

	switch (cmd) {
	case SIOCGHWTSTAMP:
		return i40e_ptp_get_ts_config(pf, ifr);
	case SIOCSHWTSTAMP:
		return i40e_ptp_set_ts_config(pf, ifr);
	default:
		return -EOPNOTSUPP;
	}
}

J
Jesse Brandeburg 已提交
2336 2337 2338 2339 2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351 2352 2353 2354
/**
 * i40e_vlan_stripping_enable - Turn on vlan stripping for the VSI
 * @vsi: the vsi being adjusted
 **/
void i40e_vlan_stripping_enable(struct i40e_vsi *vsi)
{
	struct i40e_vsi_context ctxt;
	i40e_status ret;

	if ((vsi->info.valid_sections &
	     cpu_to_le16(I40E_AQ_VSI_PROP_VLAN_VALID)) &&
	    ((vsi->info.port_vlan_flags & I40E_AQ_VSI_PVLAN_MODE_MASK) == 0))
		return;  /* already enabled */

	vsi->info.valid_sections = cpu_to_le16(I40E_AQ_VSI_PROP_VLAN_VALID);
	vsi->info.port_vlan_flags = I40E_AQ_VSI_PVLAN_MODE_ALL |
				    I40E_AQ_VSI_PVLAN_EMOD_STR_BOTH;

	ctxt.seid = vsi->seid;
2355
	ctxt.info = vsi->info;
J
Jesse Brandeburg 已提交
2356 2357 2358
	ret = i40e_aq_update_vsi_params(&vsi->back->hw, &ctxt, NULL);
	if (ret) {
		dev_info(&vsi->back->pdev->dev,
2359 2360 2361 2362
			 "update vlan stripping failed, err %s aq_err %s\n",
			 i40e_stat_str(&vsi->back->hw, ret),
			 i40e_aq_str(&vsi->back->hw,
				     vsi->back->hw.aq.asq_last_status));
J
Jesse Brandeburg 已提交
2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 2384 2385
	}
}

/**
 * i40e_vlan_stripping_disable - Turn off vlan stripping for the VSI
 * @vsi: the vsi being adjusted
 **/
void i40e_vlan_stripping_disable(struct i40e_vsi *vsi)
{
	struct i40e_vsi_context ctxt;
	i40e_status ret;

	if ((vsi->info.valid_sections &
	     cpu_to_le16(I40E_AQ_VSI_PROP_VLAN_VALID)) &&
	    ((vsi->info.port_vlan_flags & I40E_AQ_VSI_PVLAN_EMOD_MASK) ==
	     I40E_AQ_VSI_PVLAN_EMOD_MASK))
		return;  /* already disabled */

	vsi->info.valid_sections = cpu_to_le16(I40E_AQ_VSI_PROP_VLAN_VALID);
	vsi->info.port_vlan_flags = I40E_AQ_VSI_PVLAN_MODE_ALL |
				    I40E_AQ_VSI_PVLAN_EMOD_NOTHING;

	ctxt.seid = vsi->seid;
2386
	ctxt.info = vsi->info;
J
Jesse Brandeburg 已提交
2387 2388 2389
	ret = i40e_aq_update_vsi_params(&vsi->back->hw, &ctxt, NULL);
	if (ret) {
		dev_info(&vsi->back->pdev->dev,
2390 2391 2392 2393
			 "update vlan stripping failed, err %s aq_err %s\n",
			 i40e_stat_str(&vsi->back->hw, ret),
			 i40e_aq_str(&vsi->back->hw,
				     vsi->back->hw.aq.asq_last_status));
J
Jesse Brandeburg 已提交
2394 2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405 2406 2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419
	}
}

/**
 * i40e_vlan_rx_register - Setup or shutdown vlan offload
 * @netdev: network interface to be adjusted
 * @features: netdev features to test if VLAN offload is enabled or not
 **/
static void i40e_vlan_rx_register(struct net_device *netdev, u32 features)
{
	struct i40e_netdev_priv *np = netdev_priv(netdev);
	struct i40e_vsi *vsi = np->vsi;

	if (features & NETIF_F_HW_VLAN_CTAG_RX)
		i40e_vlan_stripping_enable(vsi);
	else
		i40e_vlan_stripping_disable(vsi);
}

/**
 * i40e_vsi_add_vlan - Add vsi membership for given vlan
 * @vsi: the vsi being configured
 * @vid: vlan id to be added (0 = untagged only , -1 = any)
 **/
int i40e_vsi_add_vlan(struct i40e_vsi *vsi, s16 vid)
{
2420 2421 2422
	struct i40e_mac_filter *f, *add_f, *del_f;
	struct hlist_node *h;
	int bkt;
J
Jesse Brandeburg 已提交
2423

K
Kiran Patil 已提交
2424
	/* Locked once because all functions invoked below iterates list*/
2425
	spin_lock_bh(&vsi->mac_filter_hash_lock);
K
Kiran Patil 已提交
2426

2427 2428
	if (vsi->netdev) {
		add_f = i40e_add_filter(vsi, vsi->netdev->dev_addr, vid);
J
Jesse Brandeburg 已提交
2429 2430 2431 2432
		if (!add_f) {
			dev_info(&vsi->back->pdev->dev,
				 "Could not add vlan filter %d for %pM\n",
				 vid, vsi->netdev->dev_addr);
2433
			spin_unlock_bh(&vsi->mac_filter_hash_lock);
J
Jesse Brandeburg 已提交
2434 2435 2436 2437
			return -ENOMEM;
		}
	}

2438
	hash_for_each_safe(vsi->mac_filter_hash, bkt, h, f, hlist) {
2439 2440
		if (f->state == I40E_FILTER_REMOVE)
			continue;
2441
		add_f = i40e_add_filter(vsi, f->macaddr, vid);
J
Jesse Brandeburg 已提交
2442 2443 2444 2445
		if (!add_f) {
			dev_info(&vsi->back->pdev->dev,
				 "Could not add vlan filter %d for %pM\n",
				 vid, f->macaddr);
2446
			spin_unlock_bh(&vsi->mac_filter_hash_lock);
J
Jesse Brandeburg 已提交
2447 2448 2449 2450 2451 2452 2453
			return -ENOMEM;
		}
	}

	/* Now if we add a vlan tag, make sure to check if it is the first
	 * tag (i.e. a "tag" -1 does exist) and if so replace the -1 "tag"
	 * with 0, so we now accept untagged and specified tagged traffic
2454
	 * (and not all tags along with untagged)
J
Jesse Brandeburg 已提交
2455
	 */
2456 2457 2458 2459 2460
	if (vid > 0 && vsi->netdev) {
		del_f = i40e_find_filter(vsi, vsi->netdev->dev_addr,
					 I40E_VLAN_ANY);
		if (del_f) {
			__i40e_del_filter(vsi, del_f);
2461
			add_f = i40e_add_filter(vsi, vsi->netdev->dev_addr, 0);
J
Jesse Brandeburg 已提交
2462 2463 2464 2465
			if (!add_f) {
				dev_info(&vsi->back->pdev->dev,
					 "Could not add filter 0 for %pM\n",
					 vsi->netdev->dev_addr);
2466
				spin_unlock_bh(&vsi->mac_filter_hash_lock);
J
Jesse Brandeburg 已提交
2467 2468 2469
				return -ENOMEM;
			}
		}
2470
	}
J
Jesse Brandeburg 已提交
2471

2472 2473
	/* Do not assume that I40E_VLAN_ANY should be reset to VLAN 0 */
	if (vid > 0 && !vsi->info.pvid) {
2474
		hash_for_each_safe(vsi->mac_filter_hash, bkt, h, f, hlist) {
2475 2476
			if (f->state == I40E_FILTER_REMOVE)
				continue;
2477 2478 2479
			del_f = i40e_find_filter(vsi, f->macaddr,
						 I40E_VLAN_ANY);
			if (!del_f)
K
Kiran Patil 已提交
2480
				continue;
2481
			__i40e_del_filter(vsi, del_f);
2482
			add_f = i40e_add_filter(vsi, f->macaddr, 0);
K
Kiran Patil 已提交
2483 2484 2485 2486
			if (!add_f) {
				dev_info(&vsi->back->pdev->dev,
					 "Could not add filter 0 for %pM\n",
					f->macaddr);
2487
				spin_unlock_bh(&vsi->mac_filter_hash_lock);
K
Kiran Patil 已提交
2488
				return -ENOMEM;
J
Jesse Brandeburg 已提交
2489 2490 2491 2492
			}
		}
	}

2493
	spin_unlock_bh(&vsi->mac_filter_hash_lock);
K
Kiran Patil 已提交
2494

2495 2496 2497 2498 2499
	/* schedule our worker thread which will take care of
	 * applying the new filter changes
	 */
	i40e_service_event_schedule(vsi->back);
	return 0;
J
Jesse Brandeburg 已提交
2500 2501 2502 2503 2504 2505 2506
}

/**
 * i40e_vsi_kill_vlan - Remove vsi membership for given vlan
 * @vsi: the vsi being configured
 * @vid: vlan id to be removed (0 = untagged only , -1 = any)
 **/
F
Filip Sadowski 已提交
2507
void i40e_vsi_kill_vlan(struct i40e_vsi *vsi, s16 vid)
J
Jesse Brandeburg 已提交
2508 2509
{
	struct net_device *netdev = vsi->netdev;
2510
	struct i40e_mac_filter *f;
2511 2512
	struct hlist_node *h;
	int bkt;
J
Jesse Brandeburg 已提交
2513

K
Kiran Patil 已提交
2514
	/* Locked once because all functions invoked below iterates list */
2515
	spin_lock_bh(&vsi->mac_filter_hash_lock);
K
Kiran Patil 已提交
2516

2517 2518
	if (vsi->netdev)
		i40e_del_filter(vsi, netdev->dev_addr, vid);
J
Jesse Brandeburg 已提交
2519

2520
	hash_for_each_safe(vsi->mac_filter_hash, bkt, h, f, hlist) {
2521 2522 2523
		if (f->vlan == vid)
			__i40e_del_filter(vsi, f);
	}
J
Jesse Brandeburg 已提交
2524

2525
	spin_unlock_bh(&vsi->mac_filter_hash_lock);
K
Kiran Patil 已提交
2526

2527 2528 2529 2530
	/* schedule our worker thread which will take care of
	 * applying the new filter changes
	 */
	i40e_service_event_schedule(vsi->back);
J
Jesse Brandeburg 已提交
2531 2532 2533 2534 2535 2536
}

/**
 * i40e_vlan_rx_add_vid - Add a vlan id filter to HW offload
 * @netdev: network interface to be adjusted
 * @vid: vlan id to be added
J
Jesse Brandeburg 已提交
2537 2538
 *
 * net_device_ops implementation for adding vlan ids
J
Jesse Brandeburg 已提交
2539
 **/
2540 2541 2542 2543
#ifdef I40E_FCOE
int i40e_vlan_rx_add_vid(struct net_device *netdev,
			 __always_unused __be16 proto, u16 vid)
#else
J
Jesse Brandeburg 已提交
2544 2545
static int i40e_vlan_rx_add_vid(struct net_device *netdev,
				__always_unused __be16 proto, u16 vid)
2546
#endif
J
Jesse Brandeburg 已提交
2547 2548 2549
{
	struct i40e_netdev_priv *np = netdev_priv(netdev);
	struct i40e_vsi *vsi = np->vsi;
J
Jesse Brandeburg 已提交
2550
	int ret = 0;
J
Jesse Brandeburg 已提交
2551

2552
	if (vid >= VLAN_N_VID)
J
Jesse Brandeburg 已提交
2553 2554
		return -EINVAL;

2555 2556 2557 2558 2559
	/* If the network stack called us with vid = 0 then
	 * it is asking to receive priority tagged packets with
	 * vlan id 0.  Our HW receives them by default when configured
	 * to receive untagged packets so there is no need to add an
	 * extra filter for vlan 0 tagged packets.
J
Jesse Brandeburg 已提交
2560
	 */
2561 2562
	if (vid)
		ret = i40e_vsi_add_vlan(vsi, vid);
J
Jesse Brandeburg 已提交
2563

2564
	if (!ret)
J
Jesse Brandeburg 已提交
2565
		set_bit(vid, vsi->active_vlans);
J
Jesse Brandeburg 已提交
2566

J
Jesse Brandeburg 已提交
2567
	return ret;
J
Jesse Brandeburg 已提交
2568 2569 2570 2571 2572 2573
}

/**
 * i40e_vlan_rx_kill_vid - Remove a vlan id filter from HW offload
 * @netdev: network interface to be adjusted
 * @vid: vlan id to be removed
J
Jesse Brandeburg 已提交
2574
 *
A
Akeem G Abodunrin 已提交
2575
 * net_device_ops implementation for removing vlan ids
J
Jesse Brandeburg 已提交
2576
 **/
2577 2578 2579 2580
#ifdef I40E_FCOE
int i40e_vlan_rx_kill_vid(struct net_device *netdev,
			  __always_unused __be16 proto, u16 vid)
#else
J
Jesse Brandeburg 已提交
2581 2582
static int i40e_vlan_rx_kill_vid(struct net_device *netdev,
				 __always_unused __be16 proto, u16 vid)
2583
#endif
J
Jesse Brandeburg 已提交
2584 2585 2586 2587 2588 2589
{
	struct i40e_netdev_priv *np = netdev_priv(netdev);
	struct i40e_vsi *vsi = np->vsi;

	/* return code is ignored as there is nothing a user
	 * can do about failure to remove and a log message was
J
Jesse Brandeburg 已提交
2590
	 * already printed from the other function
J
Jesse Brandeburg 已提交
2591 2592 2593 2594
	 */
	i40e_vsi_kill_vlan(vsi, vid);

	clear_bit(vid, vsi->active_vlans);
J
Jesse Brandeburg 已提交
2595

J
Jesse Brandeburg 已提交
2596 2597 2598
	return 0;
}

2599 2600 2601 2602 2603 2604 2605 2606 2607 2608 2609 2610 2611 2612 2613 2614 2615 2616 2617 2618 2619 2620 2621 2622 2623 2624 2625 2626 2627 2628 2629 2630 2631 2632 2633 2634 2635 2636
/**
 * i40e_macaddr_init - explicitly write the mac address filters
 *
 * @vsi: pointer to the vsi
 * @macaddr: the MAC address
 *
 * This is needed when the macaddr has been obtained by other
 * means than the default, e.g., from Open Firmware or IDPROM.
 * Returns 0 on success, negative on failure
 **/
static int i40e_macaddr_init(struct i40e_vsi *vsi, u8 *macaddr)
{
	int ret;
	struct i40e_aqc_add_macvlan_element_data element;

	ret = i40e_aq_mac_address_write(&vsi->back->hw,
					I40E_AQC_WRITE_TYPE_LAA_WOL,
					macaddr, NULL);
	if (ret) {
		dev_info(&vsi->back->pdev->dev,
			 "Addr change for VSI failed: %d\n", ret);
		return -EADDRNOTAVAIL;
	}

	memset(&element, 0, sizeof(element));
	ether_addr_copy(element.mac_addr, macaddr);
	element.flags = cpu_to_le16(I40E_AQC_MACVLAN_ADD_PERFECT_MATCH);
	ret = i40e_aq_add_macvlan(&vsi->back->hw, vsi->seid, &element, 1, NULL);
	if (ret) {
		dev_info(&vsi->back->pdev->dev,
			 "add filter failed err %s aq_err %s\n",
			 i40e_stat_str(&vsi->back->hw, ret),
			 i40e_aq_str(&vsi->back->hw,
				     vsi->back->hw.aq.asq_last_status));
	}
	return ret;
}

J
Jesse Brandeburg 已提交
2637 2638 2639 2640 2641 2642 2643 2644 2645 2646 2647 2648 2649 2650 2651 2652 2653 2654 2655 2656 2657 2658 2659
/**
 * i40e_restore_vlan - Reinstate vlans when vsi/netdev comes back up
 * @vsi: the vsi being brought back up
 **/
static void i40e_restore_vlan(struct i40e_vsi *vsi)
{
	u16 vid;

	if (!vsi->netdev)
		return;

	i40e_vlan_rx_register(vsi->netdev, vsi->netdev->features);

	for_each_set_bit(vid, vsi->active_vlans, VLAN_N_VID)
		i40e_vlan_rx_add_vid(vsi->netdev, htons(ETH_P_8021Q),
				     vid);
}

/**
 * i40e_vsi_add_pvid - Add pvid for the VSI
 * @vsi: the vsi being adjusted
 * @vid: the vlan id to set as a PVID
 **/
J
Jesse Brandeburg 已提交
2660
int i40e_vsi_add_pvid(struct i40e_vsi *vsi, u16 vid)
J
Jesse Brandeburg 已提交
2661 2662
{
	struct i40e_vsi_context ctxt;
2663
	i40e_status ret;
J
Jesse Brandeburg 已提交
2664 2665 2666

	vsi->info.valid_sections = cpu_to_le16(I40E_AQ_VSI_PROP_VLAN_VALID);
	vsi->info.pvid = cpu_to_le16(vid);
G
Greg Rose 已提交
2667 2668
	vsi->info.port_vlan_flags = I40E_AQ_VSI_PVLAN_MODE_TAGGED |
				    I40E_AQ_VSI_PVLAN_INSERT_PVID |
G
Greg Rose 已提交
2669
				    I40E_AQ_VSI_PVLAN_EMOD_STR;
J
Jesse Brandeburg 已提交
2670 2671

	ctxt.seid = vsi->seid;
2672
	ctxt.info = vsi->info;
2673 2674
	ret = i40e_aq_update_vsi_params(&vsi->back->hw, &ctxt, NULL);
	if (ret) {
J
Jesse Brandeburg 已提交
2675
		dev_info(&vsi->back->pdev->dev,
2676 2677 2678 2679
			 "add pvid failed, err %s aq_err %s\n",
			 i40e_stat_str(&vsi->back->hw, ret),
			 i40e_aq_str(&vsi->back->hw,
				     vsi->back->hw.aq.asq_last_status));
J
Jesse Brandeburg 已提交
2680
		return -ENOENT;
J
Jesse Brandeburg 已提交
2681 2682
	}

J
Jesse Brandeburg 已提交
2683
	return 0;
J
Jesse Brandeburg 已提交
2684 2685 2686 2687 2688 2689 2690 2691 2692 2693
}

/**
 * i40e_vsi_remove_pvid - Remove the pvid from the VSI
 * @vsi: the vsi being adjusted
 *
 * Just use the vlan_rx_register() service to put it back to normal
 **/
void i40e_vsi_remove_pvid(struct i40e_vsi *vsi)
{
G
Greg Rose 已提交
2694 2695
	i40e_vlan_stripping_disable(vsi);

J
Jesse Brandeburg 已提交
2696 2697 2698 2699 2700 2701 2702 2703 2704 2705 2706 2707 2708 2709 2710 2711 2712 2713
	vsi->info.pvid = 0;
}

/**
 * i40e_vsi_setup_tx_resources - Allocate VSI Tx queue resources
 * @vsi: ptr to the VSI
 *
 * If this function returns with an error, then it's possible one or
 * more of the rings is populated (while the rest are not).  It is the
 * callers duty to clean those orphaned rings.
 *
 * Return 0 on success, negative on failure
 **/
static int i40e_vsi_setup_tx_resources(struct i40e_vsi *vsi)
{
	int i, err = 0;

	for (i = 0; i < vsi->num_queue_pairs && !err; i++)
2714
		err = i40e_setup_tx_descriptors(vsi->tx_rings[i]);
J
Jesse Brandeburg 已提交
2715 2716 2717 2718 2719 2720 2721 2722 2723 2724 2725 2726 2727 2728

	return err;
}

/**
 * i40e_vsi_free_tx_resources - Free Tx resources for VSI queues
 * @vsi: ptr to the VSI
 *
 * Free VSI's transmit software resources
 **/
static void i40e_vsi_free_tx_resources(struct i40e_vsi *vsi)
{
	int i;

2729 2730 2731
	if (!vsi->tx_rings)
		return;

J
Jesse Brandeburg 已提交
2732
	for (i = 0; i < vsi->num_queue_pairs; i++)
2733
		if (vsi->tx_rings[i] && vsi->tx_rings[i]->desc)
2734
			i40e_free_tx_resources(vsi->tx_rings[i]);
J
Jesse Brandeburg 已提交
2735 2736 2737 2738 2739 2740 2741 2742 2743 2744 2745 2746 2747 2748 2749 2750 2751
}

/**
 * i40e_vsi_setup_rx_resources - Allocate VSI queues Rx resources
 * @vsi: ptr to the VSI
 *
 * If this function returns with an error, then it's possible one or
 * more of the rings is populated (while the rest are not).  It is the
 * callers duty to clean those orphaned rings.
 *
 * Return 0 on success, negative on failure
 **/
static int i40e_vsi_setup_rx_resources(struct i40e_vsi *vsi)
{
	int i, err = 0;

	for (i = 0; i < vsi->num_queue_pairs && !err; i++)
2752
		err = i40e_setup_rx_descriptors(vsi->rx_rings[i]);
2753 2754 2755
#ifdef I40E_FCOE
	i40e_fcoe_setup_ddp_resources(vsi);
#endif
J
Jesse Brandeburg 已提交
2756 2757 2758 2759 2760 2761 2762 2763 2764 2765 2766 2767 2768
	return err;
}

/**
 * i40e_vsi_free_rx_resources - Free Rx Resources for VSI queues
 * @vsi: ptr to the VSI
 *
 * Free all receive software resources
 **/
static void i40e_vsi_free_rx_resources(struct i40e_vsi *vsi)
{
	int i;

2769 2770 2771
	if (!vsi->rx_rings)
		return;

J
Jesse Brandeburg 已提交
2772
	for (i = 0; i < vsi->num_queue_pairs; i++)
2773
		if (vsi->rx_rings[i] && vsi->rx_rings[i]->desc)
2774
			i40e_free_rx_resources(vsi->rx_rings[i]);
2775 2776 2777
#ifdef I40E_FCOE
	i40e_fcoe_free_ddp_resources(vsi);
#endif
J
Jesse Brandeburg 已提交
2778 2779
}

2780 2781 2782 2783 2784 2785 2786 2787 2788 2789 2790 2791
/**
 * i40e_config_xps_tx_ring - Configure XPS for a Tx ring
 * @ring: The Tx ring to configure
 *
 * This enables/disables XPS for a given Tx descriptor ring
 * based on the TCs enabled for the VSI that ring belongs to.
 **/
static void i40e_config_xps_tx_ring(struct i40e_ring *ring)
{
	struct i40e_vsi *vsi = ring->vsi;
	cpumask_var_t mask;

2792 2793 2794 2795 2796 2797
	if (!ring->q_vector || !ring->netdev)
		return;

	/* Single TC mode enable XPS */
	if (vsi->tc_config.numtc <= 1) {
		if (!test_and_set_bit(__I40E_TX_XPS_INIT_DONE, &ring->state))
2798 2799 2800
			netif_set_xps_queue(ring->netdev,
					    &ring->q_vector->affinity_mask,
					    ring->queue_index);
2801 2802 2803 2804 2805
	} else if (alloc_cpumask_var(&mask, GFP_KERNEL)) {
		/* Disable XPS to allow selection based on TC */
		bitmap_zero(cpumask_bits(mask), nr_cpumask_bits);
		netif_set_xps_queue(ring->netdev, mask, ring->queue_index);
		free_cpumask_var(mask);
2806
	}
2807 2808 2809 2810 2811

	/* schedule our worker thread which will take care of
	 * applying the new filter changes
	 */
	i40e_service_event_schedule(vsi->back);
2812 2813
}

J
Jesse Brandeburg 已提交
2814 2815 2816 2817 2818 2819 2820 2821 2822 2823 2824 2825 2826 2827 2828 2829
/**
 * i40e_configure_tx_ring - Configure a transmit ring context and rest
 * @ring: The Tx ring to configure
 *
 * Configure the Tx descriptor ring in the HMC context.
 **/
static int i40e_configure_tx_ring(struct i40e_ring *ring)
{
	struct i40e_vsi *vsi = ring->vsi;
	u16 pf_q = vsi->base_queue + ring->queue_index;
	struct i40e_hw *hw = &vsi->back->hw;
	struct i40e_hmc_obj_txq tx_ctx;
	i40e_status err = 0;
	u32 qtx_ctl = 0;

	/* some ATR related tx ring init */
J
Jesse Brandeburg 已提交
2830
	if (vsi->back->flags & I40E_FLAG_FD_ATR_ENABLED) {
J
Jesse Brandeburg 已提交
2831 2832 2833 2834 2835 2836
		ring->atr_sample_rate = vsi->back->atr_sample_rate;
		ring->atr_count = 0;
	} else {
		ring->atr_sample_rate = 0;
	}

2837 2838
	/* configure XPS */
	i40e_config_xps_tx_ring(ring);
J
Jesse Brandeburg 已提交
2839 2840 2841 2842 2843 2844 2845

	/* clear the context structure first */
	memset(&tx_ctx, 0, sizeof(tx_ctx));

	tx_ctx.new_context = 1;
	tx_ctx.base = (ring->dma / 128);
	tx_ctx.qlen = ring->count;
J
Jesse Brandeburg 已提交
2846 2847
	tx_ctx.fd_ena = !!(vsi->back->flags & (I40E_FLAG_FD_SB_ENABLED |
					       I40E_FLAG_FD_ATR_ENABLED));
2848 2849 2850
#ifdef I40E_FCOE
	tx_ctx.fc_ena = (vsi->type == I40E_VSI_FCOE);
#endif
J
Jacob Keller 已提交
2851
	tx_ctx.timesync_ena = !!(vsi->back->flags & I40E_FLAG_PTP);
2852 2853 2854 2855 2856
	/* FDIR VSI tx ring can still use RS bit and writebacks */
	if (vsi->type != I40E_VSI_FDIR)
		tx_ctx.head_wb_ena = 1;
	tx_ctx.head_wb_addr = ring->dma +
			      (ring->count * sizeof(struct i40e_tx_desc));
J
Jesse Brandeburg 已提交
2857 2858 2859 2860 2861 2862 2863 2864 2865 2866 2867 2868 2869 2870 2871 2872 2873 2874 2875 2876 2877 2878 2879 2880 2881 2882 2883 2884 2885 2886 2887 2888 2889

	/* As part of VSI creation/update, FW allocates certain
	 * Tx arbitration queue sets for each TC enabled for
	 * the VSI. The FW returns the handles to these queue
	 * sets as part of the response buffer to Add VSI,
	 * Update VSI, etc. AQ commands. It is expected that
	 * these queue set handles be associated with the Tx
	 * queues by the driver as part of the TX queue context
	 * initialization. This has to be done regardless of
	 * DCB as by default everything is mapped to TC0.
	 */
	tx_ctx.rdylist = le16_to_cpu(vsi->info.qs_handle[ring->dcb_tc]);
	tx_ctx.rdylist_act = 0;

	/* clear the context in the HMC */
	err = i40e_clear_lan_tx_queue_context(hw, pf_q);
	if (err) {
		dev_info(&vsi->back->pdev->dev,
			 "Failed to clear LAN Tx queue context on Tx ring %d (pf_q %d), error: %d\n",
			 ring->queue_index, pf_q, err);
		return -ENOMEM;
	}

	/* set the context in the HMC */
	err = i40e_set_lan_tx_queue_context(hw, pf_q, &tx_ctx);
	if (err) {
		dev_info(&vsi->back->pdev->dev,
			 "Failed to set LAN Tx queue context on Tx ring %d (pf_q %d, error: %d\n",
			 ring->queue_index, pf_q, err);
		return -ENOMEM;
	}

	/* Now associate this queue with this PCI function */
M
Mitch Williams 已提交
2890
	if (vsi->type == I40E_VSI_VMDQ2) {
2891
		qtx_ctl = I40E_QTX_CTL_VM_QUEUE;
M
Mitch Williams 已提交
2892 2893 2894
		qtx_ctl |= ((vsi->id) << I40E_QTX_CTL_VFVM_INDX_SHIFT) &
			   I40E_QTX_CTL_VFVM_INDX_MASK;
	} else {
2895
		qtx_ctl = I40E_QTX_CTL_PF_QUEUE;
M
Mitch Williams 已提交
2896 2897
	}

2898 2899
	qtx_ctl |= ((hw->pf_id << I40E_QTX_CTL_PF_INDX_SHIFT) &
		    I40E_QTX_CTL_PF_INDX_MASK);
J
Jesse Brandeburg 已提交
2900 2901 2902 2903 2904 2905 2906 2907 2908 2909 2910 2911 2912 2913 2914 2915 2916 2917 2918 2919 2920 2921 2922 2923 2924 2925 2926 2927 2928 2929 2930 2931 2932 2933 2934 2935
	wr32(hw, I40E_QTX_CTL(pf_q), qtx_ctl);
	i40e_flush(hw);

	/* cache tail off for easier writes later */
	ring->tail = hw->hw_addr + I40E_QTX_TAIL(pf_q);

	return 0;
}

/**
 * i40e_configure_rx_ring - Configure a receive ring context
 * @ring: The Rx ring to configure
 *
 * Configure the Rx descriptor ring in the HMC context.
 **/
static int i40e_configure_rx_ring(struct i40e_ring *ring)
{
	struct i40e_vsi *vsi = ring->vsi;
	u32 chain_len = vsi->back->hw.func_caps.rx_buf_chain_len;
	u16 pf_q = vsi->base_queue + ring->queue_index;
	struct i40e_hw *hw = &vsi->back->hw;
	struct i40e_hmc_obj_rxq rx_ctx;
	i40e_status err = 0;

	ring->state = 0;

	/* clear the context structure first */
	memset(&rx_ctx, 0, sizeof(rx_ctx));

	ring->rx_buf_len = vsi->rx_buf_len;

	rx_ctx.dbuff = ring->rx_buf_len >> I40E_RXQ_CTX_DBUFF_SHIFT;

	rx_ctx.base = (ring->dma / 128);
	rx_ctx.qlen = ring->count;

2936 2937
	/* use 32 byte descriptors */
	rx_ctx.dsize = 1;
J
Jesse Brandeburg 已提交
2938

2939 2940 2941
	/* descriptor type is always zero
	 * rx_ctx.dtype = 0;
	 */
2942
	rx_ctx.hsplit_0 = 0;
J
Jesse Brandeburg 已提交
2943

2944
	rx_ctx.rxmax = min_t(u16, vsi->max_frame, chain_len * ring->rx_buf_len);
2945 2946 2947 2948
	if (hw->revision_id == 0)
		rx_ctx.lrxqthresh = 0;
	else
		rx_ctx.lrxqthresh = 2;
J
Jesse Brandeburg 已提交
2949 2950
	rx_ctx.crcstrip = 1;
	rx_ctx.l2tsel = 1;
J
Jesse Brandeburg 已提交
2951 2952
	/* this controls whether VLAN is stripped from inner headers */
	rx_ctx.showiv = 0;
2953 2954 2955
#ifdef I40E_FCOE
	rx_ctx.fc_ena = (vsi->type == I40E_VSI_FCOE);
#endif
2956 2957
	/* set the prefena field to 1 because the manual says to */
	rx_ctx.prefena = 1;
J
Jesse Brandeburg 已提交
2958 2959 2960 2961 2962 2963 2964 2965 2966 2967 2968 2969 2970 2971 2972 2973 2974 2975 2976 2977 2978 2979 2980

	/* clear the context in the HMC */
	err = i40e_clear_lan_rx_queue_context(hw, pf_q);
	if (err) {
		dev_info(&vsi->back->pdev->dev,
			 "Failed to clear LAN Rx queue context on Rx ring %d (pf_q %d), error: %d\n",
			 ring->queue_index, pf_q, err);
		return -ENOMEM;
	}

	/* set the context in the HMC */
	err = i40e_set_lan_rx_queue_context(hw, pf_q, &rx_ctx);
	if (err) {
		dev_info(&vsi->back->pdev->dev,
			 "Failed to set LAN Rx queue context on Rx ring %d (pf_q %d), error: %d\n",
			 ring->queue_index, pf_q, err);
		return -ENOMEM;
	}

	/* cache tail for quicker writes, and clear the reg before use */
	ring->tail = hw->hw_addr + I40E_QRX_TAIL(pf_q);
	writel(0, ring->tail);

2981
	i40e_alloc_rx_buffers(ring, I40E_DESC_UNUSED(ring));
J
Jesse Brandeburg 已提交
2982 2983 2984 2985 2986 2987 2988 2989 2990 2991 2992 2993 2994 2995 2996

	return 0;
}

/**
 * i40e_vsi_configure_tx - Configure the VSI for Tx
 * @vsi: VSI structure describing this set of rings and resources
 *
 * Configure the Tx VSI for operation.
 **/
static int i40e_vsi_configure_tx(struct i40e_vsi *vsi)
{
	int err = 0;
	u16 i;

2997 2998
	for (i = 0; (i < vsi->num_queue_pairs) && !err; i++)
		err = i40e_configure_tx_ring(vsi->tx_rings[i]);
J
Jesse Brandeburg 已提交
2999 3000 3001 3002 3003 3004 3005 3006 3007 3008 3009 3010 3011 3012 3013 3014 3015 3016 3017 3018 3019

	return err;
}

/**
 * i40e_vsi_configure_rx - Configure the VSI for Rx
 * @vsi: the VSI being configured
 *
 * Configure the Rx VSI for operation.
 **/
static int i40e_vsi_configure_rx(struct i40e_vsi *vsi)
{
	int err = 0;
	u16 i;

	if (vsi->netdev && (vsi->netdev->mtu > ETH_DATA_LEN))
		vsi->max_frame = vsi->netdev->mtu + ETH_HLEN
			       + ETH_FCS_LEN + VLAN_HLEN;
	else
		vsi->max_frame = I40E_RXBUFFER_2048;

3020
	vsi->rx_buf_len = I40E_RXBUFFER_2048;
J
Jesse Brandeburg 已提交
3021

3022 3023 3024 3025 3026 3027 3028 3029 3030
#ifdef I40E_FCOE
	/* setup rx buffer for FCoE */
	if ((vsi->type == I40E_VSI_FCOE) &&
	    (vsi->back->flags & I40E_FLAG_FCOE_ENABLED)) {
		vsi->rx_buf_len = I40E_RXBUFFER_3072;
		vsi->max_frame = I40E_RXBUFFER_3072;
	}

#endif /* I40E_FCOE */
J
Jesse Brandeburg 已提交
3031 3032
	/* round up for the chip's needs */
	vsi->rx_buf_len = ALIGN(vsi->rx_buf_len,
3033
				BIT_ULL(I40E_RXQ_CTX_DBUFF_SHIFT));
J
Jesse Brandeburg 已提交
3034 3035 3036

	/* set up individual rings */
	for (i = 0; i < vsi->num_queue_pairs && !err; i++)
3037
		err = i40e_configure_rx_ring(vsi->rx_rings[i]);
J
Jesse Brandeburg 已提交
3038 3039 3040 3041 3042 3043 3044 3045 3046 3047

	return err;
}

/**
 * i40e_vsi_config_dcb_rings - Update rings to reflect DCB TC
 * @vsi: ptr to the VSI
 **/
static void i40e_vsi_config_dcb_rings(struct i40e_vsi *vsi)
{
3048
	struct i40e_ring *tx_ring, *rx_ring;
J
Jesse Brandeburg 已提交
3049 3050 3051
	u16 qoffset, qcount;
	int i, n;

3052 3053 3054 3055 3056 3057 3058 3059 3060
	if (!(vsi->back->flags & I40E_FLAG_DCB_ENABLED)) {
		/* Reset the TC information */
		for (i = 0; i < vsi->num_queue_pairs; i++) {
			rx_ring = vsi->rx_rings[i];
			tx_ring = vsi->tx_rings[i];
			rx_ring->dcb_tc = 0;
			tx_ring->dcb_tc = 0;
		}
	}
J
Jesse Brandeburg 已提交
3061 3062

	for (n = 0; n < I40E_MAX_TRAFFIC_CLASS; n++) {
3063
		if (!(vsi->tc_config.enabled_tc & BIT_ULL(n)))
J
Jesse Brandeburg 已提交
3064 3065 3066 3067 3068
			continue;

		qoffset = vsi->tc_config.tc_info[n].qoffset;
		qcount = vsi->tc_config.tc_info[n].qcount;
		for (i = qoffset; i < (qoffset + qcount); i++) {
3069 3070
			rx_ring = vsi->rx_rings[i];
			tx_ring = vsi->tx_rings[i];
J
Jesse Brandeburg 已提交
3071 3072 3073 3074 3075 3076 3077 3078 3079 3080 3081 3082
			rx_ring->dcb_tc = n;
			tx_ring->dcb_tc = n;
		}
	}
}

/**
 * i40e_set_vsi_rx_mode - Call set_rx_mode on a VSI
 * @vsi: ptr to the VSI
 **/
static void i40e_set_vsi_rx_mode(struct i40e_vsi *vsi)
{
3083 3084 3085
	struct i40e_pf *pf = vsi->back;
	int err;

J
Jesse Brandeburg 已提交
3086 3087
	if (vsi->netdev)
		i40e_set_rx_mode(vsi->netdev);
3088 3089 3090 3091 3092 3093 3094 3095

	if (!!(pf->flags & I40E_FLAG_PF_MAC)) {
		err = i40e_macaddr_init(vsi, pf->hw.mac.addr);
		if (err) {
			dev_warn(&pf->pdev->dev,
				 "could not set up macaddr; err %d\n", err);
		}
	}
J
Jesse Brandeburg 已提交
3096 3097
}

3098 3099 3100 3101 3102 3103 3104 3105 3106 3107 3108 3109 3110
/**
 * i40e_fdir_filter_restore - Restore the Sideband Flow Director filters
 * @vsi: Pointer to the targeted VSI
 *
 * This function replays the hlist on the hw where all the SB Flow Director
 * filters were saved.
 **/
static void i40e_fdir_filter_restore(struct i40e_vsi *vsi)
{
	struct i40e_fdir_filter *filter;
	struct i40e_pf *pf = vsi->back;
	struct hlist_node *node;

3111 3112 3113
	if (!(pf->flags & I40E_FLAG_FD_SB_ENABLED))
		return;

3114 3115 3116 3117 3118 3119
	hlist_for_each_entry_safe(filter, node,
				  &pf->fdir_filter_list, fdir_node) {
		i40e_add_del_fdir(vsi, filter, true);
	}
}

J
Jesse Brandeburg 已提交
3120 3121 3122 3123 3124 3125 3126 3127 3128 3129 3130 3131 3132 3133 3134 3135 3136 3137 3138 3139 3140 3141 3142 3143 3144 3145 3146 3147 3148 3149 3150 3151 3152 3153 3154 3155
/**
 * i40e_vsi_configure - Set up the VSI for action
 * @vsi: the VSI being configured
 **/
static int i40e_vsi_configure(struct i40e_vsi *vsi)
{
	int err;

	i40e_set_vsi_rx_mode(vsi);
	i40e_restore_vlan(vsi);
	i40e_vsi_config_dcb_rings(vsi);
	err = i40e_vsi_configure_tx(vsi);
	if (!err)
		err = i40e_vsi_configure_rx(vsi);

	return err;
}

/**
 * i40e_vsi_configure_msix - MSIX mode Interrupt Config in the HW
 * @vsi: the VSI being configured
 **/
static void i40e_vsi_configure_msix(struct i40e_vsi *vsi)
{
	struct i40e_pf *pf = vsi->back;
	struct i40e_hw *hw = &pf->hw;
	u16 vector;
	int i, q;
	u32 qp;

	/* The interrupt indexing is offset by 1 in the PFINT_ITRn
	 * and PFINT_LNKLSTn registers, e.g.:
	 *   PFINT_ITRn[0..n-1] gets msix-1..msix-n  (qpair interrupts)
	 */
	qp = vsi->base_queue;
	vector = vsi->base_vector;
3156
	for (i = 0; i < vsi->num_q_vectors; i++, vector++) {
3157 3158
		struct i40e_q_vector *q_vector = vsi->q_vectors[i];

3159
		q_vector->itr_countdown = ITR_COUNTDOWN_START;
3160
		q_vector->rx.itr = ITR_TO_REG(vsi->rx_rings[i]->rx_itr_setting);
J
Jesse Brandeburg 已提交
3161 3162 3163
		q_vector->rx.latency_range = I40E_LOW_LATENCY;
		wr32(hw, I40E_PFINT_ITRN(I40E_RX_ITR, vector - 1),
		     q_vector->rx.itr);
3164
		q_vector->tx.itr = ITR_TO_REG(vsi->tx_rings[i]->tx_itr_setting);
J
Jesse Brandeburg 已提交
3165 3166 3167
		q_vector->tx.latency_range = I40E_LOW_LATENCY;
		wr32(hw, I40E_PFINT_ITRN(I40E_TX_ITR, vector - 1),
		     q_vector->tx.itr);
3168 3169
		wr32(hw, I40E_PFINT_RATEN(vector - 1),
		     INTRL_USEC_TO_REG(vsi->int_rate_limit));
J
Jesse Brandeburg 已提交
3170 3171 3172 3173

		/* Linked list for the queuepairs assigned to this vector */
		wr32(hw, I40E_PFINT_LNKLSTN(vector - 1), qp);
		for (q = 0; q < q_vector->num_ringpairs; q++) {
3174 3175
			u32 val;

J
Jesse Brandeburg 已提交
3176 3177 3178 3179 3180 3181 3182 3183 3184 3185 3186 3187 3188 3189 3190 3191 3192 3193 3194 3195 3196 3197 3198 3199 3200 3201 3202 3203 3204 3205 3206 3207 3208
			val = I40E_QINT_RQCTL_CAUSE_ENA_MASK |
			      (I40E_RX_ITR << I40E_QINT_RQCTL_ITR_INDX_SHIFT)  |
			      (vector      << I40E_QINT_RQCTL_MSIX_INDX_SHIFT) |
			      (qp          << I40E_QINT_RQCTL_NEXTQ_INDX_SHIFT)|
			      (I40E_QUEUE_TYPE_TX
				      << I40E_QINT_RQCTL_NEXTQ_TYPE_SHIFT);

			wr32(hw, I40E_QINT_RQCTL(qp), val);

			val = I40E_QINT_TQCTL_CAUSE_ENA_MASK |
			      (I40E_TX_ITR << I40E_QINT_TQCTL_ITR_INDX_SHIFT)  |
			      (vector      << I40E_QINT_TQCTL_MSIX_INDX_SHIFT) |
			      ((qp+1)      << I40E_QINT_TQCTL_NEXTQ_INDX_SHIFT)|
			      (I40E_QUEUE_TYPE_RX
				      << I40E_QINT_TQCTL_NEXTQ_TYPE_SHIFT);

			/* Terminate the linked list */
			if (q == (q_vector->num_ringpairs - 1))
				val |= (I40E_QUEUE_END_OF_LIST
					   << I40E_QINT_TQCTL_NEXTQ_INDX_SHIFT);

			wr32(hw, I40E_QINT_TQCTL(qp), val);
			qp++;
		}
	}

	i40e_flush(hw);
}

/**
 * i40e_enable_misc_int_causes - enable the non-queue interrupts
 * @hw: ptr to the hardware info
 **/
3209
static void i40e_enable_misc_int_causes(struct i40e_pf *pf)
J
Jesse Brandeburg 已提交
3210
{
3211
	struct i40e_hw *hw = &pf->hw;
J
Jesse Brandeburg 已提交
3212 3213 3214 3215 3216 3217 3218 3219 3220 3221 3222 3223 3224 3225 3226
	u32 val;

	/* clear things first */
	wr32(hw, I40E_PFINT_ICR0_ENA, 0);  /* disable all */
	rd32(hw, I40E_PFINT_ICR0);         /* read to clear */

	val = I40E_PFINT_ICR0_ENA_ECC_ERR_MASK       |
	      I40E_PFINT_ICR0_ENA_MAL_DETECT_MASK    |
	      I40E_PFINT_ICR0_ENA_GRST_MASK          |
	      I40E_PFINT_ICR0_ENA_PCI_EXCEPTION_MASK |
	      I40E_PFINT_ICR0_ENA_GPIO_MASK          |
	      I40E_PFINT_ICR0_ENA_HMC_ERR_MASK       |
	      I40E_PFINT_ICR0_ENA_VFLR_MASK          |
	      I40E_PFINT_ICR0_ENA_ADMINQ_MASK;

3227 3228 3229
	if (pf->flags & I40E_FLAG_IWARP_ENABLED)
		val |= I40E_PFINT_ICR0_ENA_PE_CRITERR_MASK;

3230 3231 3232
	if (pf->flags & I40E_FLAG_PTP)
		val |= I40E_PFINT_ICR0_ENA_TIMESYNC_MASK;

J
Jesse Brandeburg 已提交
3233 3234 3235
	wr32(hw, I40E_PFINT_ICR0_ENA, val);

	/* SW_ITR_IDX = 0, but don't change INTENA */
3236 3237
	wr32(hw, I40E_PFINT_DYN_CTL0, I40E_PFINT_DYN_CTL0_SW_ITR_INDX_MASK |
					I40E_PFINT_DYN_CTL0_INTENA_MSK_MASK);
J
Jesse Brandeburg 已提交
3238 3239 3240 3241 3242 3243 3244 3245 3246 3247 3248

	/* OTHER_ITR_IDX = 0 */
	wr32(hw, I40E_PFINT_STAT_CTL0, 0);
}

/**
 * i40e_configure_msi_and_legacy - Legacy mode interrupt config in the HW
 * @vsi: the VSI being configured
 **/
static void i40e_configure_msi_and_legacy(struct i40e_vsi *vsi)
{
3249
	struct i40e_q_vector *q_vector = vsi->q_vectors[0];
J
Jesse Brandeburg 已提交
3250 3251 3252 3253 3254
	struct i40e_pf *pf = vsi->back;
	struct i40e_hw *hw = &pf->hw;
	u32 val;

	/* set the ITR configuration */
3255
	q_vector->itr_countdown = ITR_COUNTDOWN_START;
3256
	q_vector->rx.itr = ITR_TO_REG(vsi->rx_rings[0]->rx_itr_setting);
J
Jesse Brandeburg 已提交
3257 3258
	q_vector->rx.latency_range = I40E_LOW_LATENCY;
	wr32(hw, I40E_PFINT_ITR0(I40E_RX_ITR), q_vector->rx.itr);
3259
	q_vector->tx.itr = ITR_TO_REG(vsi->tx_rings[0]->tx_itr_setting);
J
Jesse Brandeburg 已提交
3260 3261 3262
	q_vector->tx.latency_range = I40E_LOW_LATENCY;
	wr32(hw, I40E_PFINT_ITR0(I40E_TX_ITR), q_vector->tx.itr);

3263
	i40e_enable_misc_int_causes(pf);
J
Jesse Brandeburg 已提交
3264 3265 3266 3267

	/* FIRSTQ_INDX = 0, FIRSTQ_TYPE = 0 (rx) */
	wr32(hw, I40E_PFINT_LNKLST0, 0);

3268
	/* Associate the queue pair to the vector and enable the queue int */
J
Jesse Brandeburg 已提交
3269 3270 3271 3272 3273 3274 3275 3276 3277 3278 3279 3280 3281 3282
	val = I40E_QINT_RQCTL_CAUSE_ENA_MASK		      |
	      (I40E_RX_ITR << I40E_QINT_RQCTL_ITR_INDX_SHIFT) |
	      (I40E_QUEUE_TYPE_TX << I40E_QINT_TQCTL_NEXTQ_TYPE_SHIFT);

	wr32(hw, I40E_QINT_RQCTL(0), val);

	val = I40E_QINT_TQCTL_CAUSE_ENA_MASK		      |
	      (I40E_TX_ITR << I40E_QINT_TQCTL_ITR_INDX_SHIFT) |
	      (I40E_QUEUE_END_OF_LIST << I40E_QINT_TQCTL_NEXTQ_INDX_SHIFT);

	wr32(hw, I40E_QINT_TQCTL(0), val);
	i40e_flush(hw);
}

3283 3284 3285 3286 3287 3288 3289 3290 3291 3292 3293 3294 3295
/**
 * i40e_irq_dynamic_disable_icr0 - Disable default interrupt generation for icr0
 * @pf: board private structure
 **/
void i40e_irq_dynamic_disable_icr0(struct i40e_pf *pf)
{
	struct i40e_hw *hw = &pf->hw;

	wr32(hw, I40E_PFINT_DYN_CTL0,
	     I40E_ITR_NONE << I40E_PFINT_DYN_CTLN_ITR_INDX_SHIFT);
	i40e_flush(hw);
}

J
Jesse Brandeburg 已提交
3296 3297 3298
/**
 * i40e_irq_dynamic_enable_icr0 - Enable default interrupt generation for icr0
 * @pf: board private structure
3299
 * @clearpba: true when all pending interrupt events should be cleared
J
Jesse Brandeburg 已提交
3300
 **/
3301
void i40e_irq_dynamic_enable_icr0(struct i40e_pf *pf, bool clearpba)
J
Jesse Brandeburg 已提交
3302 3303 3304 3305 3306
{
	struct i40e_hw *hw = &pf->hw;
	u32 val;

	val = I40E_PFINT_DYN_CTL0_INTENA_MASK   |
3307
	      (clearpba ? I40E_PFINT_DYN_CTL0_CLEARPBA_MASK : 0) |
J
Jesse Brandeburg 已提交
3308 3309 3310 3311 3312 3313 3314 3315 3316 3317 3318 3319 3320 3321 3322
	      (I40E_ITR_NONE << I40E_PFINT_DYN_CTL0_ITR_INDX_SHIFT);

	wr32(hw, I40E_PFINT_DYN_CTL0, val);
	i40e_flush(hw);
}

/**
 * i40e_msix_clean_rings - MSIX mode Interrupt Handler
 * @irq: interrupt number
 * @data: pointer to a q_vector
 **/
static irqreturn_t i40e_msix_clean_rings(int irq, void *data)
{
	struct i40e_q_vector *q_vector = data;

3323
	if (!q_vector->tx.ring && !q_vector->rx.ring)
J
Jesse Brandeburg 已提交
3324 3325
		return IRQ_HANDLED;

3326
	napi_schedule_irqoff(&q_vector->napi);
J
Jesse Brandeburg 已提交
3327 3328 3329 3330

	return IRQ_HANDLED;
}

3331 3332 3333 3334 3335 3336 3337 3338 3339 3340 3341 3342 3343 3344 3345 3346 3347 3348 3349 3350 3351 3352 3353 3354 3355 3356 3357
/**
 * i40e_irq_affinity_notify - Callback for affinity changes
 * @notify: context as to what irq was changed
 * @mask: the new affinity mask
 *
 * This is a callback function used by the irq_set_affinity_notifier function
 * so that we may register to receive changes to the irq affinity masks.
 **/
static void i40e_irq_affinity_notify(struct irq_affinity_notify *notify,
				     const cpumask_t *mask)
{
	struct i40e_q_vector *q_vector =
		container_of(notify, struct i40e_q_vector, affinity_notify);

	q_vector->affinity_mask = *mask;
}

/**
 * i40e_irq_affinity_release - Callback for affinity notifier release
 * @ref: internal core kernel usage
 *
 * This is a callback function used by the irq_set_affinity_notifier function
 * to inform the current notification subscriber that they will no longer
 * receive notifications.
 **/
static void i40e_irq_affinity_release(struct kref *ref) {}

J
Jesse Brandeburg 已提交
3358 3359 3360 3361 3362 3363 3364 3365 3366 3367 3368 3369 3370 3371 3372
/**
 * i40e_vsi_request_irq_msix - Initialize MSI-X interrupts
 * @vsi: the VSI being configured
 * @basename: name for the vector
 *
 * Allocates MSI-X vectors and requests interrupts from the kernel.
 **/
static int i40e_vsi_request_irq_msix(struct i40e_vsi *vsi, char *basename)
{
	int q_vectors = vsi->num_q_vectors;
	struct i40e_pf *pf = vsi->back;
	int base = vsi->base_vector;
	int rx_int_idx = 0;
	int tx_int_idx = 0;
	int vector, err;
3373
	int irq_num;
J
Jesse Brandeburg 已提交
3374 3375

	for (vector = 0; vector < q_vectors; vector++) {
3376
		struct i40e_q_vector *q_vector = vsi->q_vectors[vector];
J
Jesse Brandeburg 已提交
3377

3378 3379
		irq_num = pf->msix_entries[base + vector].vector;

3380
		if (q_vector->tx.ring && q_vector->rx.ring) {
J
Jesse Brandeburg 已提交
3381 3382 3383
			snprintf(q_vector->name, sizeof(q_vector->name) - 1,
				 "%s-%s-%d", basename, "TxRx", rx_int_idx++);
			tx_int_idx++;
3384
		} else if (q_vector->rx.ring) {
J
Jesse Brandeburg 已提交
3385 3386
			snprintf(q_vector->name, sizeof(q_vector->name) - 1,
				 "%s-%s-%d", basename, "rx", rx_int_idx++);
3387
		} else if (q_vector->tx.ring) {
J
Jesse Brandeburg 已提交
3388 3389 3390 3391 3392 3393
			snprintf(q_vector->name, sizeof(q_vector->name) - 1,
				 "%s-%s-%d", basename, "tx", tx_int_idx++);
		} else {
			/* skip this unused q_vector */
			continue;
		}
3394
		err = request_irq(irq_num,
J
Jesse Brandeburg 已提交
3395 3396 3397 3398 3399 3400
				  vsi->irq_handler,
				  0,
				  q_vector->name,
				  q_vector);
		if (err) {
			dev_info(&pf->pdev->dev,
S
Shannon Nelson 已提交
3401
				 "MSIX request_irq failed, error: %d\n", err);
J
Jesse Brandeburg 已提交
3402 3403
			goto free_queue_irqs;
		}
3404 3405 3406 3407 3408

		/* register for affinity change notifications */
		q_vector->affinity_notify.notify = i40e_irq_affinity_notify;
		q_vector->affinity_notify.release = i40e_irq_affinity_release;
		irq_set_affinity_notifier(irq_num, &q_vector->affinity_notify);
J
Jesse Brandeburg 已提交
3409
		/* assign the mask for this irq */
3410
		irq_set_affinity_hint(irq_num, &q_vector->affinity_mask);
J
Jesse Brandeburg 已提交
3411 3412
	}

3413
	vsi->irqs_ready = true;
J
Jesse Brandeburg 已提交
3414 3415 3416 3417 3418
	return 0;

free_queue_irqs:
	while (vector) {
		vector--;
3419 3420 3421 3422
		irq_num = pf->msix_entries[base + vector].vector;
		irq_set_affinity_notifier(irq_num, NULL);
		irq_set_affinity_hint(irq_num, NULL);
		free_irq(irq_num, &vsi->q_vectors[vector]);
J
Jesse Brandeburg 已提交
3423 3424 3425 3426 3427 3428 3429 3430 3431 3432 3433 3434 3435 3436 3437 3438
	}
	return err;
}

/**
 * i40e_vsi_disable_irq - Mask off queue interrupt generation on the VSI
 * @vsi: the VSI being un-configured
 **/
static void i40e_vsi_disable_irq(struct i40e_vsi *vsi)
{
	struct i40e_pf *pf = vsi->back;
	struct i40e_hw *hw = &pf->hw;
	int base = vsi->base_vector;
	int i;

	for (i = 0; i < vsi->num_queue_pairs; i++) {
3439 3440
		wr32(hw, I40E_QINT_TQCTL(vsi->tx_rings[i]->reg_idx), 0);
		wr32(hw, I40E_QINT_RQCTL(vsi->rx_rings[i]->reg_idx), 0);
J
Jesse Brandeburg 已提交
3441 3442 3443 3444 3445 3446 3447 3448 3449 3450 3451 3452 3453 3454 3455 3456 3457 3458 3459 3460 3461 3462 3463 3464 3465 3466 3467 3468 3469
	}

	if (pf->flags & I40E_FLAG_MSIX_ENABLED) {
		for (i = vsi->base_vector;
		     i < (vsi->num_q_vectors + vsi->base_vector); i++)
			wr32(hw, I40E_PFINT_DYN_CTLN(i - 1), 0);

		i40e_flush(hw);
		for (i = 0; i < vsi->num_q_vectors; i++)
			synchronize_irq(pf->msix_entries[i + base].vector);
	} else {
		/* Legacy and MSI mode - this stops all interrupt handling */
		wr32(hw, I40E_PFINT_ICR0_ENA, 0);
		wr32(hw, I40E_PFINT_DYN_CTL0, 0);
		i40e_flush(hw);
		synchronize_irq(pf->pdev->irq);
	}
}

/**
 * i40e_vsi_enable_irq - Enable IRQ for the given VSI
 * @vsi: the VSI being configured
 **/
static int i40e_vsi_enable_irq(struct i40e_vsi *vsi)
{
	struct i40e_pf *pf = vsi->back;
	int i;

	if (pf->flags & I40E_FLAG_MSIX_ENABLED) {
3470
		for (i = 0; i < vsi->num_q_vectors; i++)
J
Jesse Brandeburg 已提交
3471 3472
			i40e_irq_dynamic_enable(vsi, i);
	} else {
3473
		i40e_irq_dynamic_enable_icr0(pf, true);
J
Jesse Brandeburg 已提交
3474 3475
	}

3476
	i40e_flush(&pf->hw);
J
Jesse Brandeburg 已提交
3477 3478 3479 3480 3481 3482 3483 3484 3485 3486 3487 3488 3489 3490 3491 3492 3493 3494 3495 3496 3497 3498 3499 3500 3501 3502 3503
	return 0;
}

/**
 * i40e_stop_misc_vector - Stop the vector that handles non-queue events
 * @pf: board private structure
 **/
static void i40e_stop_misc_vector(struct i40e_pf *pf)
{
	/* Disable ICR 0 */
	wr32(&pf->hw, I40E_PFINT_ICR0_ENA, 0);
	i40e_flush(&pf->hw);
}

/**
 * i40e_intr - MSI/Legacy and non-queue interrupt handler
 * @irq: interrupt number
 * @data: pointer to a q_vector
 *
 * This is the handler used for all MSI/Legacy interrupts, and deals
 * with both queue and non-queue interrupts.  This is also used in
 * MSIX mode to handle the non-queue interrupts.
 **/
static irqreturn_t i40e_intr(int irq, void *data)
{
	struct i40e_pf *pf = (struct i40e_pf *)data;
	struct i40e_hw *hw = &pf->hw;
3504
	irqreturn_t ret = IRQ_NONE;
J
Jesse Brandeburg 已提交
3505 3506 3507 3508
	u32 icr0, icr0_remaining;
	u32 val, ena_mask;

	icr0 = rd32(hw, I40E_PFINT_ICR0);
3509
	ena_mask = rd32(hw, I40E_PFINT_ICR0_ENA);
J
Jesse Brandeburg 已提交
3510

3511 3512
	/* if sharing a legacy IRQ, we might get called w/o an intr pending */
	if ((icr0 & I40E_PFINT_ICR0_INTEVENT_MASK) == 0)
3513
		goto enable_intr;
J
Jesse Brandeburg 已提交
3514

S
Shannon Nelson 已提交
3515 3516 3517 3518 3519
	/* if interrupt but no bits showing, must be SWINT */
	if (((icr0 & ~I40E_PFINT_ICR0_INTEVENT_MASK) == 0) ||
	    (icr0 & I40E_PFINT_ICR0_SWINT_MASK))
		pf->sw_int_count++;

3520 3521 3522 3523 3524 3525 3526
	if ((pf->flags & I40E_FLAG_IWARP_ENABLED) &&
	    (ena_mask & I40E_PFINT_ICR0_ENA_PE_CRITERR_MASK)) {
		ena_mask &= ~I40E_PFINT_ICR0_ENA_PE_CRITERR_MASK;
		icr0 &= ~I40E_PFINT_ICR0_ENA_PE_CRITERR_MASK;
		dev_info(&pf->pdev->dev, "cleared PE_CRITERR\n");
	}

J
Jesse Brandeburg 已提交
3527 3528
	/* only q0 is used in MSI/Legacy mode, and none are used in MSIX */
	if (icr0 & I40E_PFINT_ICR0_QUEUE_0_MASK) {
3529 3530
		struct i40e_vsi *vsi = pf->vsi[pf->lan_vsi];
		struct i40e_q_vector *q_vector = vsi->q_vectors[0];
J
Jesse Brandeburg 已提交
3531

3532 3533 3534 3535 3536 3537
		/* We do not have a way to disarm Queue causes while leaving
		 * interrupt enabled for all other causes, ideally
		 * interrupt should be disabled while we are in NAPI but
		 * this is not a performance path and napi_schedule()
		 * can deal with rescheduling.
		 */
J
Jesse Brandeburg 已提交
3538
		if (!test_bit(__I40E_DOWN, &pf->state))
3539
			napi_schedule_irqoff(&q_vector->napi);
J
Jesse Brandeburg 已提交
3540 3541 3542 3543 3544
	}

	if (icr0 & I40E_PFINT_ICR0_ADMINQ_MASK) {
		ena_mask &= ~I40E_PFINT_ICR0_ENA_ADMINQ_MASK;
		set_bit(__I40E_ADMINQ_EVENT_PENDING, &pf->state);
3545
		i40e_debug(&pf->hw, I40E_DEBUG_NVM, "AdminQ event\n");
J
Jesse Brandeburg 已提交
3546 3547 3548 3549 3550 3551 3552 3553 3554 3555 3556 3557 3558 3559 3560 3561 3562 3563 3564
	}

	if (icr0 & I40E_PFINT_ICR0_MAL_DETECT_MASK) {
		ena_mask &= ~I40E_PFINT_ICR0_ENA_MAL_DETECT_MASK;
		set_bit(__I40E_MDD_EVENT_PENDING, &pf->state);
	}

	if (icr0 & I40E_PFINT_ICR0_VFLR_MASK) {
		ena_mask &= ~I40E_PFINT_ICR0_ENA_VFLR_MASK;
		set_bit(__I40E_VFLR_EVENT_PENDING, &pf->state);
	}

	if (icr0 & I40E_PFINT_ICR0_GRST_MASK) {
		if (!test_bit(__I40E_RESET_RECOVERY_PENDING, &pf->state))
			set_bit(__I40E_RESET_INTR_RECEIVED, &pf->state);
		ena_mask &= ~I40E_PFINT_ICR0_ENA_GRST_MASK;
		val = rd32(hw, I40E_GLGEN_RSTAT);
		val = (val & I40E_GLGEN_RSTAT_RESET_TYPE_MASK)
		       >> I40E_GLGEN_RSTAT_RESET_TYPE_SHIFT;
3565
		if (val == I40E_RESET_CORER) {
J
Jesse Brandeburg 已提交
3566
			pf->corer_count++;
3567
		} else if (val == I40E_RESET_GLOBR) {
J
Jesse Brandeburg 已提交
3568
			pf->globr_count++;
3569
		} else if (val == I40E_RESET_EMPR) {
J
Jesse Brandeburg 已提交
3570
			pf->empr_count++;
3571
			set_bit(__I40E_EMP_RESET_INTR_RECEIVED, &pf->state);
3572
		}
J
Jesse Brandeburg 已提交
3573 3574
	}

3575 3576 3577
	if (icr0 & I40E_PFINT_ICR0_HMC_ERR_MASK) {
		icr0 &= ~I40E_PFINT_ICR0_HMC_ERR_MASK;
		dev_info(&pf->pdev->dev, "HMC error interrupt\n");
3578 3579 3580
		dev_info(&pf->pdev->dev, "HMC error info 0x%x, HMC error data 0x%x\n",
			 rd32(hw, I40E_PFHMC_ERRORINFO),
			 rd32(hw, I40E_PFHMC_ERRORDATA));
3581 3582
	}

J
Jacob Keller 已提交
3583 3584 3585 3586
	if (icr0 & I40E_PFINT_ICR0_TIMESYNC_MASK) {
		u32 prttsyn_stat = rd32(hw, I40E_PRTTSYN_STAT_0);

		if (prttsyn_stat & I40E_PRTTSYN_STAT_0_TXTIME_MASK) {
3587
			icr0 &= ~I40E_PFINT_ICR0_ENA_TIMESYNC_MASK;
J
Jacob Keller 已提交
3588 3589 3590 3591
			i40e_ptp_tx_hwtstamp(pf);
		}
	}

J
Jesse Brandeburg 已提交
3592 3593 3594 3595 3596 3597 3598 3599
	/* If a critical error is pending we have no choice but to reset the
	 * device.
	 * Report and mask out any remaining unexpected interrupts.
	 */
	icr0_remaining = icr0 & ena_mask;
	if (icr0_remaining) {
		dev_info(&pf->pdev->dev, "unhandled interrupt icr0=0x%08x\n",
			 icr0_remaining);
3600
		if ((icr0_remaining & I40E_PFINT_ICR0_PE_CRITERR_MASK) ||
J
Jesse Brandeburg 已提交
3601
		    (icr0_remaining & I40E_PFINT_ICR0_PCI_EXCEPTION_MASK) ||
3602
		    (icr0_remaining & I40E_PFINT_ICR0_ECC_ERR_MASK)) {
3603 3604 3605
			dev_info(&pf->pdev->dev, "device will be reset\n");
			set_bit(__I40E_PF_RESET_REQUESTED, &pf->state);
			i40e_service_event_schedule(pf);
J
Jesse Brandeburg 已提交
3606 3607 3608
		}
		ena_mask &= ~icr0_remaining;
	}
3609
	ret = IRQ_HANDLED;
J
Jesse Brandeburg 已提交
3610

3611
enable_intr:
J
Jesse Brandeburg 已提交
3612 3613 3614 3615
	/* re-enable interrupt causes */
	wr32(hw, I40E_PFINT_ICR0_ENA, ena_mask);
	if (!test_bit(__I40E_DOWN, &pf->state)) {
		i40e_service_event_schedule(pf);
3616
		i40e_irq_dynamic_enable_icr0(pf, false);
J
Jesse Brandeburg 已提交
3617 3618
	}

3619
	return ret;
J
Jesse Brandeburg 已提交
3620 3621
}

3622 3623 3624 3625 3626 3627 3628 3629 3630 3631 3632 3633 3634 3635 3636 3637 3638 3639 3640 3641 3642 3643 3644 3645 3646 3647 3648 3649 3650 3651 3652 3653 3654 3655 3656 3657
/**
 * i40e_clean_fdir_tx_irq - Reclaim resources after transmit completes
 * @tx_ring:  tx ring to clean
 * @budget:   how many cleans we're allowed
 *
 * Returns true if there's any budget left (e.g. the clean is finished)
 **/
static bool i40e_clean_fdir_tx_irq(struct i40e_ring *tx_ring, int budget)
{
	struct i40e_vsi *vsi = tx_ring->vsi;
	u16 i = tx_ring->next_to_clean;
	struct i40e_tx_buffer *tx_buf;
	struct i40e_tx_desc *tx_desc;

	tx_buf = &tx_ring->tx_bi[i];
	tx_desc = I40E_TX_DESC(tx_ring, i);
	i -= tx_ring->count;

	do {
		struct i40e_tx_desc *eop_desc = tx_buf->next_to_watch;

		/* if next_to_watch is not set then there is no work pending */
		if (!eop_desc)
			break;

		/* prevent any other reads prior to eop_desc */
		read_barrier_depends();

		/* if the descriptor isn't done, no work yet to do */
		if (!(eop_desc->cmd_type_offset_bsz &
		      cpu_to_le64(I40E_TX_DESC_DTYPE_DESC_DONE)))
			break;

		/* clear next_to_watch to prevent false hangs */
		tx_buf->next_to_watch = NULL;

3658 3659 3660 3661 3662 3663 3664 3665 3666 3667 3668
		tx_desc->buffer_addr = 0;
		tx_desc->cmd_type_offset_bsz = 0;
		/* move past filter desc */
		tx_buf++;
		tx_desc++;
		i++;
		if (unlikely(!i)) {
			i -= tx_ring->count;
			tx_buf = tx_ring->tx_bi;
			tx_desc = I40E_TX_DESC(tx_ring, 0);
		}
3669 3670 3671 3672 3673
		/* unmap skb header data */
		dma_unmap_single(tx_ring->dev,
				 dma_unmap_addr(tx_buf, dma),
				 dma_unmap_len(tx_buf, len),
				 DMA_TO_DEVICE);
3674 3675
		if (tx_buf->tx_flags & I40E_TX_FLAGS_FD_SB)
			kfree(tx_buf->raw_buf);
3676

3677 3678 3679
		tx_buf->raw_buf = NULL;
		tx_buf->tx_flags = 0;
		tx_buf->next_to_watch = NULL;
3680
		dma_unmap_len_set(tx_buf, len, 0);
3681 3682
		tx_desc->buffer_addr = 0;
		tx_desc->cmd_type_offset_bsz = 0;
3683

3684
		/* move us past the eop_desc for start of next FD desc */
3685 3686 3687 3688 3689 3690 3691 3692 3693 3694 3695 3696 3697 3698 3699 3700
		tx_buf++;
		tx_desc++;
		i++;
		if (unlikely(!i)) {
			i -= tx_ring->count;
			tx_buf = tx_ring->tx_bi;
			tx_desc = I40E_TX_DESC(tx_ring, 0);
		}

		/* update budget accounting */
		budget--;
	} while (likely(budget));

	i += tx_ring->count;
	tx_ring->next_to_clean = i;

J
Jesse Brandeburg 已提交
3701
	if (vsi->back->flags & I40E_FLAG_MSIX_ENABLED)
3702
		i40e_irq_dynamic_enable(vsi, tx_ring->q_vector->v_idx);
J
Jesse Brandeburg 已提交
3703

3704 3705 3706 3707 3708 3709 3710 3711 3712 3713 3714 3715 3716 3717 3718 3719 3720 3721 3722 3723 3724 3725
	return budget > 0;
}

/**
 * i40e_fdir_clean_ring - Interrupt Handler for FDIR SB ring
 * @irq: interrupt number
 * @data: pointer to a q_vector
 **/
static irqreturn_t i40e_fdir_clean_ring(int irq, void *data)
{
	struct i40e_q_vector *q_vector = data;
	struct i40e_vsi *vsi;

	if (!q_vector->tx.ring)
		return IRQ_HANDLED;

	vsi = q_vector->tx.ring->vsi;
	i40e_clean_fdir_tx_irq(q_vector->tx.ring, vsi->work_limit);

	return IRQ_HANDLED;
}

J
Jesse Brandeburg 已提交
3726
/**
3727
 * i40e_map_vector_to_qp - Assigns the queue pair to the vector
J
Jesse Brandeburg 已提交
3728 3729
 * @vsi: the VSI being configured
 * @v_idx: vector index
3730
 * @qp_idx: queue pair index
J
Jesse Brandeburg 已提交
3731
 **/
3732
static void i40e_map_vector_to_qp(struct i40e_vsi *vsi, int v_idx, int qp_idx)
J
Jesse Brandeburg 已提交
3733
{
3734
	struct i40e_q_vector *q_vector = vsi->q_vectors[v_idx];
3735 3736
	struct i40e_ring *tx_ring = vsi->tx_rings[qp_idx];
	struct i40e_ring *rx_ring = vsi->rx_rings[qp_idx];
J
Jesse Brandeburg 已提交
3737 3738

	tx_ring->q_vector = q_vector;
3739 3740
	tx_ring->next = q_vector->tx.ring;
	q_vector->tx.ring = tx_ring;
J
Jesse Brandeburg 已提交
3741
	q_vector->tx.count++;
3742 3743 3744 3745 3746

	rx_ring->q_vector = q_vector;
	rx_ring->next = q_vector->rx.ring;
	q_vector->rx.ring = rx_ring;
	q_vector->rx.count++;
J
Jesse Brandeburg 已提交
3747 3748 3749 3750 3751 3752 3753 3754 3755 3756 3757 3758 3759 3760 3761
}

/**
 * i40e_vsi_map_rings_to_vectors - Maps descriptor rings to vectors
 * @vsi: the VSI being configured
 *
 * This function maps descriptor rings to the queue-specific vectors
 * we were allotted through the MSI-X enabling code.  Ideally, we'd have
 * one vector per queue pair, but on a constrained vector budget, we
 * group the queue pairs as "efficiently" as possible.
 **/
static void i40e_vsi_map_rings_to_vectors(struct i40e_vsi *vsi)
{
	int qp_remaining = vsi->num_queue_pairs;
	int q_vectors = vsi->num_q_vectors;
3762
	int num_ringpairs;
J
Jesse Brandeburg 已提交
3763 3764 3765 3766 3767
	int v_start = 0;
	int qp_idx = 0;

	/* If we don't have enough vectors for a 1-to-1 mapping, we'll have to
	 * group them so there are multiple queues per vector.
3768 3769 3770 3771
	 * It is also important to go through all the vectors available to be
	 * sure that if we don't use all the vectors, that the remaining vectors
	 * are cleared. This is especially important when decreasing the
	 * number of queues in use.
J
Jesse Brandeburg 已提交
3772
	 */
3773
	for (; v_start < q_vectors; v_start++) {
3774 3775 3776 3777 3778 3779 3780 3781 3782 3783 3784 3785
		struct i40e_q_vector *q_vector = vsi->q_vectors[v_start];

		num_ringpairs = DIV_ROUND_UP(qp_remaining, q_vectors - v_start);

		q_vector->num_ringpairs = num_ringpairs;

		q_vector->rx.count = 0;
		q_vector->tx.count = 0;
		q_vector->rx.ring = NULL;
		q_vector->tx.ring = NULL;

		while (num_ringpairs--) {
3786
			i40e_map_vector_to_qp(vsi, v_start, qp_idx);
3787 3788
			qp_idx++;
			qp_remaining--;
J
Jesse Brandeburg 已提交
3789 3790 3791 3792 3793 3794 3795 3796 3797 3798 3799 3800 3801 3802 3803 3804 3805 3806
		}
	}
}

/**
 * i40e_vsi_request_irq - Request IRQ from the OS
 * @vsi: the VSI being configured
 * @basename: name for the vector
 **/
static int i40e_vsi_request_irq(struct i40e_vsi *vsi, char *basename)
{
	struct i40e_pf *pf = vsi->back;
	int err;

	if (pf->flags & I40E_FLAG_MSIX_ENABLED)
		err = i40e_vsi_request_irq_msix(vsi, basename);
	else if (pf->flags & I40E_FLAG_MSI_ENABLED)
		err = request_irq(pf->pdev->irq, i40e_intr, 0,
3807
				  pf->int_name, pf);
J
Jesse Brandeburg 已提交
3808 3809
	else
		err = request_irq(pf->pdev->irq, i40e_intr, IRQF_SHARED,
3810
				  pf->int_name, pf);
J
Jesse Brandeburg 已提交
3811 3812 3813 3814 3815 3816 3817 3818 3819

	if (err)
		dev_info(&pf->pdev->dev, "request_irq failed, Error %d\n", err);

	return err;
}

#ifdef CONFIG_NET_POLL_CONTROLLER
/**
3820
 * i40e_netpoll - A Polling 'interrupt' handler
J
Jesse Brandeburg 已提交
3821 3822 3823 3824 3825
 * @netdev: network interface device structure
 *
 * This is used by netconsole to send skbs without having to re-enable
 * interrupts.  It's not called while the normal interrupt routine is executing.
 **/
3826 3827 3828
#ifdef I40E_FCOE
void i40e_netpoll(struct net_device *netdev)
#else
J
Jesse Brandeburg 已提交
3829
static void i40e_netpoll(struct net_device *netdev)
3830
#endif
J
Jesse Brandeburg 已提交
3831 3832 3833 3834 3835 3836 3837 3838 3839 3840 3841 3842
{
	struct i40e_netdev_priv *np = netdev_priv(netdev);
	struct i40e_vsi *vsi = np->vsi;
	struct i40e_pf *pf = vsi->back;
	int i;

	/* if interface is down do nothing */
	if (test_bit(__I40E_DOWN, &vsi->state))
		return;

	if (pf->flags & I40E_FLAG_MSIX_ENABLED) {
		for (i = 0; i < vsi->num_q_vectors; i++)
3843
			i40e_msix_clean_rings(0, vsi->q_vectors[i]);
J
Jesse Brandeburg 已提交
3844 3845 3846 3847 3848 3849
	} else {
		i40e_intr(pf->pdev->irq, netdev);
	}
}
#endif

3850 3851 3852 3853 3854 3855 3856 3857 3858 3859 3860 3861 3862 3863 3864 3865 3866 3867 3868 3869 3870
/**
 * i40e_pf_txq_wait - Wait for a PF's Tx queue to be enabled or disabled
 * @pf: the PF being configured
 * @pf_q: the PF queue
 * @enable: enable or disable state of the queue
 *
 * This routine will wait for the given Tx queue of the PF to reach the
 * enabled or disabled state.
 * Returns -ETIMEDOUT in case of failing to reach the requested state after
 * multiple retries; else will return 0 in case of success.
 **/
static int i40e_pf_txq_wait(struct i40e_pf *pf, int pf_q, bool enable)
{
	int i;
	u32 tx_reg;

	for (i = 0; i < I40E_QUEUE_WAIT_RETRY_LIMIT; i++) {
		tx_reg = rd32(&pf->hw, I40E_QTX_ENA(pf_q));
		if (enable == !!(tx_reg & I40E_QTX_ENA_QENA_STAT_MASK))
			break;

3871
		usleep_range(10, 20);
3872 3873 3874 3875 3876 3877 3878
	}
	if (i >= I40E_QUEUE_WAIT_RETRY_LIMIT)
		return -ETIMEDOUT;

	return 0;
}

J
Jesse Brandeburg 已提交
3879 3880 3881 3882 3883 3884 3885 3886 3887
/**
 * i40e_vsi_control_tx - Start or stop a VSI's rings
 * @vsi: the VSI being configured
 * @enable: start or stop the rings
 **/
static int i40e_vsi_control_tx(struct i40e_vsi *vsi, bool enable)
{
	struct i40e_pf *pf = vsi->back;
	struct i40e_hw *hw = &pf->hw;
3888
	int i, j, pf_q, ret = 0;
J
Jesse Brandeburg 已提交
3889 3890 3891 3892
	u32 tx_reg;

	pf_q = vsi->base_queue;
	for (i = 0; i < vsi->num_queue_pairs; i++, pf_q++) {
3893 3894 3895 3896

		/* warn the TX unit of coming changes */
		i40e_pre_tx_queue_cfg(&pf->hw, pf_q, enable);
		if (!enable)
3897
			usleep_range(10, 20);
3898

3899
		for (j = 0; j < 50; j++) {
J
Jesse Brandeburg 已提交
3900
			tx_reg = rd32(hw, I40E_QTX_ENA(pf_q));
3901 3902 3903 3904 3905
			if (((tx_reg >> I40E_QTX_ENA_QENA_REQ_SHIFT) & 1) ==
			    ((tx_reg >> I40E_QTX_ENA_QENA_STAT_SHIFT) & 1))
				break;
			usleep_range(1000, 2000);
		}
3906
		/* Skip if the queue is already in the requested state */
3907
		if (enable == !!(tx_reg & I40E_QTX_ENA_QENA_STAT_MASK))
3908
			continue;
J
Jesse Brandeburg 已提交
3909 3910

		/* turn on/off the queue */
3911 3912
		if (enable) {
			wr32(hw, I40E_QTX_HEAD(pf_q), 0);
3913
			tx_reg |= I40E_QTX_ENA_QENA_REQ_MASK;
3914
		} else {
J
Jesse Brandeburg 已提交
3915
			tx_reg &= ~I40E_QTX_ENA_QENA_REQ_MASK;
3916
		}
J
Jesse Brandeburg 已提交
3917 3918

		wr32(hw, I40E_QTX_ENA(pf_q), tx_reg);
3919 3920 3921
		/* No waiting for the Tx queue to disable */
		if (!enable && test_bit(__I40E_PORT_TX_SUSPENDED, &pf->state))
			continue;
J
Jesse Brandeburg 已提交
3922 3923

		/* wait for the change to finish */
3924 3925 3926
		ret = i40e_pf_txq_wait(pf, pf_q, enable);
		if (ret) {
			dev_info(&pf->pdev->dev,
S
Shannon Nelson 已提交
3927 3928
				 "VSI seid %d Tx ring %d %sable timeout\n",
				 vsi->seid, pf_q, (enable ? "en" : "dis"));
3929
			break;
J
Jesse Brandeburg 已提交
3930 3931 3932
		}
	}

3933 3934
	if (hw->revision_id == 0)
		mdelay(50);
3935 3936 3937 3938 3939 3940 3941 3942 3943 3944 3945 3946 3947 3948 3949 3950 3951 3952 3953 3954 3955 3956 3957 3958
	return ret;
}

/**
 * i40e_pf_rxq_wait - Wait for a PF's Rx queue to be enabled or disabled
 * @pf: the PF being configured
 * @pf_q: the PF queue
 * @enable: enable or disable state of the queue
 *
 * This routine will wait for the given Rx queue of the PF to reach the
 * enabled or disabled state.
 * Returns -ETIMEDOUT in case of failing to reach the requested state after
 * multiple retries; else will return 0 in case of success.
 **/
static int i40e_pf_rxq_wait(struct i40e_pf *pf, int pf_q, bool enable)
{
	int i;
	u32 rx_reg;

	for (i = 0; i < I40E_QUEUE_WAIT_RETRY_LIMIT; i++) {
		rx_reg = rd32(&pf->hw, I40E_QRX_ENA(pf_q));
		if (enable == !!(rx_reg & I40E_QRX_ENA_QENA_STAT_MASK))
			break;

3959
		usleep_range(10, 20);
3960 3961 3962
	}
	if (i >= I40E_QUEUE_WAIT_RETRY_LIMIT)
		return -ETIMEDOUT;
3963

J
Jesse Brandeburg 已提交
3964 3965 3966 3967 3968 3969 3970 3971 3972 3973 3974 3975
	return 0;
}

/**
 * i40e_vsi_control_rx - Start or stop a VSI's rings
 * @vsi: the VSI being configured
 * @enable: start or stop the rings
 **/
static int i40e_vsi_control_rx(struct i40e_vsi *vsi, bool enable)
{
	struct i40e_pf *pf = vsi->back;
	struct i40e_hw *hw = &pf->hw;
3976
	int i, j, pf_q, ret = 0;
J
Jesse Brandeburg 已提交
3977 3978 3979 3980
	u32 rx_reg;

	pf_q = vsi->base_queue;
	for (i = 0; i < vsi->num_queue_pairs; i++, pf_q++) {
3981
		for (j = 0; j < 50; j++) {
J
Jesse Brandeburg 已提交
3982
			rx_reg = rd32(hw, I40E_QRX_ENA(pf_q));
3983 3984 3985 3986 3987
			if (((rx_reg >> I40E_QRX_ENA_QENA_REQ_SHIFT) & 1) ==
			    ((rx_reg >> I40E_QRX_ENA_QENA_STAT_SHIFT) & 1))
				break;
			usleep_range(1000, 2000);
		}
J
Jesse Brandeburg 已提交
3988

3989 3990 3991
		/* Skip if the queue is already in the requested state */
		if (enable == !!(rx_reg & I40E_QRX_ENA_QENA_STAT_MASK))
			continue;
J
Jesse Brandeburg 已提交
3992 3993 3994

		/* turn on/off the queue */
		if (enable)
3995
			rx_reg |= I40E_QRX_ENA_QENA_REQ_MASK;
J
Jesse Brandeburg 已提交
3996
		else
3997
			rx_reg &= ~I40E_QRX_ENA_QENA_REQ_MASK;
J
Jesse Brandeburg 已提交
3998
		wr32(hw, I40E_QRX_ENA(pf_q), rx_reg);
3999 4000 4001
		/* No waiting for the Tx queue to disable */
		if (!enable && test_bit(__I40E_PORT_TX_SUSPENDED, &pf->state))
			continue;
J
Jesse Brandeburg 已提交
4002 4003

		/* wait for the change to finish */
4004 4005 4006
		ret = i40e_pf_rxq_wait(pf, pf_q, enable);
		if (ret) {
			dev_info(&pf->pdev->dev,
S
Shannon Nelson 已提交
4007 4008
				 "VSI seid %d Rx ring %d %sable timeout\n",
				 vsi->seid, pf_q, (enable ? "en" : "dis"));
4009
			break;
J
Jesse Brandeburg 已提交
4010 4011 4012
		}
	}

4013
	return ret;
J
Jesse Brandeburg 已提交
4014 4015 4016
}

/**
F
Filip Sadowski 已提交
4017
 * i40e_vsi_start_rings - Start a VSI's rings
J
Jesse Brandeburg 已提交
4018 4019
 * @vsi: the VSI being configured
 **/
F
Filip Sadowski 已提交
4020
int i40e_vsi_start_rings(struct i40e_vsi *vsi)
J
Jesse Brandeburg 已提交
4021
{
4022
	int ret = 0;
J
Jesse Brandeburg 已提交
4023 4024

	/* do rx first for enable and last for disable */
F
Filip Sadowski 已提交
4025 4026 4027 4028
	ret = i40e_vsi_control_rx(vsi, true);
	if (ret)
		return ret;
	ret = i40e_vsi_control_tx(vsi, true);
J
Jesse Brandeburg 已提交
4029 4030 4031 4032

	return ret;
}

F
Filip Sadowski 已提交
4033 4034 4035 4036 4037 4038 4039 4040 4041 4042 4043 4044 4045
/**
 * i40e_vsi_stop_rings - Stop a VSI's rings
 * @vsi: the VSI being configured
 **/
void i40e_vsi_stop_rings(struct i40e_vsi *vsi)
{
	/* do rx first for enable and last for disable
	 * Ignore return value, we need to shutdown whatever we can
	 */
	i40e_vsi_control_tx(vsi, false);
	i40e_vsi_control_rx(vsi, false);
}

J
Jesse Brandeburg 已提交
4046 4047 4048 4049 4050 4051 4052 4053 4054 4055 4056 4057 4058 4059 4060 4061
/**
 * i40e_vsi_free_irq - Free the irq association with the OS
 * @vsi: the VSI being configured
 **/
static void i40e_vsi_free_irq(struct i40e_vsi *vsi)
{
	struct i40e_pf *pf = vsi->back;
	struct i40e_hw *hw = &pf->hw;
	int base = vsi->base_vector;
	u32 val, qp;
	int i;

	if (pf->flags & I40E_FLAG_MSIX_ENABLED) {
		if (!vsi->q_vectors)
			return;

4062 4063 4064 4065
		if (!vsi->irqs_ready)
			return;

		vsi->irqs_ready = false;
J
Jesse Brandeburg 已提交
4066
		for (i = 0; i < vsi->num_q_vectors; i++) {
4067 4068 4069 4070 4071
			int irq_num;
			u16 vector;

			vector = i + base;
			irq_num = pf->msix_entries[vector].vector;
J
Jesse Brandeburg 已提交
4072 4073

			/* free only the irqs that were actually requested */
S
Shannon Nelson 已提交
4074 4075
			if (!vsi->q_vectors[i] ||
			    !vsi->q_vectors[i]->num_ringpairs)
J
Jesse Brandeburg 已提交
4076 4077
				continue;

4078 4079
			/* clear the affinity notifier in the IRQ descriptor */
			irq_set_affinity_notifier(irq_num, NULL);
J
Jesse Brandeburg 已提交
4080
			/* clear the affinity_mask in the IRQ descriptor */
4081 4082 4083
			irq_set_affinity_hint(irq_num, NULL);
			synchronize_irq(irq_num);
			free_irq(irq_num, vsi->q_vectors[i]);
J
Jesse Brandeburg 已提交
4084 4085 4086 4087 4088 4089 4090 4091 4092 4093 4094 4095 4096 4097 4098 4099 4100 4101 4102 4103 4104 4105 4106 4107 4108 4109 4110 4111 4112 4113 4114 4115 4116 4117 4118 4119 4120 4121 4122 4123 4124 4125 4126 4127 4128 4129 4130 4131 4132 4133 4134 4135 4136 4137 4138 4139 4140 4141 4142 4143 4144 4145 4146 4147 4148 4149 4150 4151 4152 4153 4154 4155 4156 4157 4158 4159 4160 4161 4162 4163 4164 4165

			/* Tear down the interrupt queue link list
			 *
			 * We know that they come in pairs and always
			 * the Rx first, then the Tx.  To clear the
			 * link list, stick the EOL value into the
			 * next_q field of the registers.
			 */
			val = rd32(hw, I40E_PFINT_LNKLSTN(vector - 1));
			qp = (val & I40E_PFINT_LNKLSTN_FIRSTQ_INDX_MASK)
				>> I40E_PFINT_LNKLSTN_FIRSTQ_INDX_SHIFT;
			val |= I40E_QUEUE_END_OF_LIST
				<< I40E_PFINT_LNKLSTN_FIRSTQ_INDX_SHIFT;
			wr32(hw, I40E_PFINT_LNKLSTN(vector - 1), val);

			while (qp != I40E_QUEUE_END_OF_LIST) {
				u32 next;

				val = rd32(hw, I40E_QINT_RQCTL(qp));

				val &= ~(I40E_QINT_RQCTL_MSIX_INDX_MASK  |
					 I40E_QINT_RQCTL_MSIX0_INDX_MASK |
					 I40E_QINT_RQCTL_CAUSE_ENA_MASK  |
					 I40E_QINT_RQCTL_INTEVENT_MASK);

				val |= (I40E_QINT_RQCTL_ITR_INDX_MASK |
					 I40E_QINT_RQCTL_NEXTQ_INDX_MASK);

				wr32(hw, I40E_QINT_RQCTL(qp), val);

				val = rd32(hw, I40E_QINT_TQCTL(qp));

				next = (val & I40E_QINT_TQCTL_NEXTQ_INDX_MASK)
					>> I40E_QINT_TQCTL_NEXTQ_INDX_SHIFT;

				val &= ~(I40E_QINT_TQCTL_MSIX_INDX_MASK  |
					 I40E_QINT_TQCTL_MSIX0_INDX_MASK |
					 I40E_QINT_TQCTL_CAUSE_ENA_MASK  |
					 I40E_QINT_TQCTL_INTEVENT_MASK);

				val |= (I40E_QINT_TQCTL_ITR_INDX_MASK |
					 I40E_QINT_TQCTL_NEXTQ_INDX_MASK);

				wr32(hw, I40E_QINT_TQCTL(qp), val);
				qp = next;
			}
		}
	} else {
		free_irq(pf->pdev->irq, pf);

		val = rd32(hw, I40E_PFINT_LNKLST0);
		qp = (val & I40E_PFINT_LNKLSTN_FIRSTQ_INDX_MASK)
			>> I40E_PFINT_LNKLSTN_FIRSTQ_INDX_SHIFT;
		val |= I40E_QUEUE_END_OF_LIST
			<< I40E_PFINT_LNKLST0_FIRSTQ_INDX_SHIFT;
		wr32(hw, I40E_PFINT_LNKLST0, val);

		val = rd32(hw, I40E_QINT_RQCTL(qp));
		val &= ~(I40E_QINT_RQCTL_MSIX_INDX_MASK  |
			 I40E_QINT_RQCTL_MSIX0_INDX_MASK |
			 I40E_QINT_RQCTL_CAUSE_ENA_MASK  |
			 I40E_QINT_RQCTL_INTEVENT_MASK);

		val |= (I40E_QINT_RQCTL_ITR_INDX_MASK |
			I40E_QINT_RQCTL_NEXTQ_INDX_MASK);

		wr32(hw, I40E_QINT_RQCTL(qp), val);

		val = rd32(hw, I40E_QINT_TQCTL(qp));

		val &= ~(I40E_QINT_TQCTL_MSIX_INDX_MASK  |
			 I40E_QINT_TQCTL_MSIX0_INDX_MASK |
			 I40E_QINT_TQCTL_CAUSE_ENA_MASK  |
			 I40E_QINT_TQCTL_INTEVENT_MASK);

		val |= (I40E_QINT_TQCTL_ITR_INDX_MASK |
			I40E_QINT_TQCTL_NEXTQ_INDX_MASK);

		wr32(hw, I40E_QINT_TQCTL(qp), val);
	}
}

4166 4167 4168 4169 4170 4171 4172 4173 4174 4175 4176 4177
/**
 * i40e_free_q_vector - Free memory allocated for specific interrupt vector
 * @vsi: the VSI being configured
 * @v_idx: Index of vector to be freed
 *
 * This function frees the memory allocated to the q_vector.  In addition if
 * NAPI is enabled it will delete any references to the NAPI struct prior
 * to freeing the q_vector.
 **/
static void i40e_free_q_vector(struct i40e_vsi *vsi, int v_idx)
{
	struct i40e_q_vector *q_vector = vsi->q_vectors[v_idx];
4178
	struct i40e_ring *ring;
4179 4180 4181 4182 4183

	if (!q_vector)
		return;

	/* disassociate q_vector from rings */
4184 4185 4186 4187 4188
	i40e_for_each_ring(ring, q_vector->tx)
		ring->q_vector = NULL;

	i40e_for_each_ring(ring, q_vector->rx)
		ring->q_vector = NULL;
4189 4190 4191 4192 4193 4194 4195 4196 4197 4198

	/* only VSI w/ an associated netdev is set up w/ NAPI */
	if (vsi->netdev)
		netif_napi_del(&q_vector->napi);

	vsi->q_vectors[v_idx] = NULL;

	kfree_rcu(q_vector, rcu);
}

J
Jesse Brandeburg 已提交
4199 4200 4201 4202 4203 4204 4205 4206 4207 4208 4209
/**
 * i40e_vsi_free_q_vectors - Free memory allocated for interrupt vectors
 * @vsi: the VSI being un-configured
 *
 * This frees the memory allocated to the q_vectors and
 * deletes references to the NAPI struct.
 **/
static void i40e_vsi_free_q_vectors(struct i40e_vsi *vsi)
{
	int v_idx;

4210 4211
	for (v_idx = 0; v_idx < vsi->num_q_vectors; v_idx++)
		i40e_free_q_vector(vsi, v_idx);
J
Jesse Brandeburg 已提交
4212 4213 4214 4215 4216 4217 4218 4219 4220 4221 4222 4223 4224
}

/**
 * i40e_reset_interrupt_capability - Disable interrupt setup in OS
 * @pf: board private structure
 **/
static void i40e_reset_interrupt_capability(struct i40e_pf *pf)
{
	/* If we're in Legacy mode, the interrupt was cleaned in vsi_close */
	if (pf->flags & I40E_FLAG_MSIX_ENABLED) {
		pci_disable_msix(pf->pdev);
		kfree(pf->msix_entries);
		pf->msix_entries = NULL;
4225 4226
		kfree(pf->irq_pile);
		pf->irq_pile = NULL;
J
Jesse Brandeburg 已提交
4227 4228 4229 4230 4231 4232 4233 4234 4235 4236 4237 4238 4239 4240 4241 4242 4243
	} else if (pf->flags & I40E_FLAG_MSI_ENABLED) {
		pci_disable_msi(pf->pdev);
	}
	pf->flags &= ~(I40E_FLAG_MSIX_ENABLED | I40E_FLAG_MSI_ENABLED);
}

/**
 * i40e_clear_interrupt_scheme - Clear the current interrupt scheme settings
 * @pf: board private structure
 *
 * We go through and clear interrupt specific resources and reset the structure
 * to pre-load conditions
 **/
static void i40e_clear_interrupt_scheme(struct i40e_pf *pf)
{
	int i;

4244
	i40e_stop_misc_vector(pf);
4245
	if (pf->flags & I40E_FLAG_MSIX_ENABLED && pf->msix_entries) {
4246 4247 4248 4249
		synchronize_irq(pf->msix_entries[0].vector);
		free_irq(pf->msix_entries[0].vector, pf);
	}

4250 4251 4252
	i40e_put_lump(pf->irq_pile, pf->iwarp_base_vector,
		      I40E_IWARP_IRQ_PILE_ID);

J
Jesse Brandeburg 已提交
4253
	i40e_put_lump(pf->irq_pile, 0, I40E_PILE_VALID_BIT-1);
M
Mitch Williams 已提交
4254
	for (i = 0; i < pf->num_alloc_vsi; i++)
J
Jesse Brandeburg 已提交
4255 4256 4257 4258 4259 4260 4261 4262 4263 4264 4265 4266 4267 4268 4269 4270 4271
		if (pf->vsi[i])
			i40e_vsi_free_q_vectors(pf->vsi[i]);
	i40e_reset_interrupt_capability(pf);
}

/**
 * i40e_napi_enable_all - Enable NAPI for all q_vectors in the VSI
 * @vsi: the VSI being configured
 **/
static void i40e_napi_enable_all(struct i40e_vsi *vsi)
{
	int q_idx;

	if (!vsi->netdev)
		return;

	for (q_idx = 0; q_idx < vsi->num_q_vectors; q_idx++)
4272
		napi_enable(&vsi->q_vectors[q_idx]->napi);
J
Jesse Brandeburg 已提交
4273 4274 4275 4276 4277 4278 4279 4280 4281 4282 4283 4284 4285 4286
}

/**
 * i40e_napi_disable_all - Disable NAPI for all q_vectors in the VSI
 * @vsi: the VSI being configured
 **/
static void i40e_napi_disable_all(struct i40e_vsi *vsi)
{
	int q_idx;

	if (!vsi->netdev)
		return;

	for (q_idx = 0; q_idx < vsi->num_q_vectors; q_idx++)
4287
		napi_disable(&vsi->q_vectors[q_idx]->napi);
J
Jesse Brandeburg 已提交
4288 4289
}

4290 4291 4292 4293 4294 4295
/**
 * i40e_vsi_close - Shut down a VSI
 * @vsi: the vsi to be quelled
 **/
static void i40e_vsi_close(struct i40e_vsi *vsi)
{
4296 4297
	bool reset = false;

4298 4299 4300 4301 4302
	if (!test_and_set_bit(__I40E_DOWN, &vsi->state))
		i40e_down(vsi);
	i40e_vsi_free_irq(vsi);
	i40e_vsi_free_tx_resources(vsi);
	i40e_vsi_free_rx_resources(vsi);
4303
	vsi->current_netdev_flags = 0;
4304 4305 4306
	if (test_bit(__I40E_RESET_RECOVERY_PENDING, &vsi->back->state))
		reset = true;
	i40e_notify_client_of_netdev_close(vsi, reset);
4307 4308
}

J
Jesse Brandeburg 已提交
4309 4310 4311 4312 4313 4314 4315 4316 4317
/**
 * i40e_quiesce_vsi - Pause a given VSI
 * @vsi: the VSI being paused
 **/
static void i40e_quiesce_vsi(struct i40e_vsi *vsi)
{
	if (test_bit(__I40E_DOWN, &vsi->state))
		return;

4318 4319 4320 4321
	/* No need to disable FCoE VSI when Tx suspended */
	if ((test_bit(__I40E_PORT_TX_SUSPENDED, &vsi->back->state)) &&
	    vsi->type == I40E_VSI_FCOE) {
		dev_dbg(&vsi->back->pdev->dev,
S
Shannon Nelson 已提交
4322
			 "VSI seid %d skipping FCoE VSI disable\n", vsi->seid);
4323 4324 4325
		return;
	}

J
Jesse Brandeburg 已提交
4326
	set_bit(__I40E_NEEDS_RESTART, &vsi->state);
J
Jesse Brandeburg 已提交
4327
	if (vsi->netdev && netif_running(vsi->netdev))
J
Jesse Brandeburg 已提交
4328
		vsi->netdev->netdev_ops->ndo_stop(vsi->netdev);
J
Jesse Brandeburg 已提交
4329
	else
4330
		i40e_vsi_close(vsi);
J
Jesse Brandeburg 已提交
4331 4332 4333 4334 4335 4336 4337 4338 4339 4340 4341 4342 4343 4344 4345
}

/**
 * i40e_unquiesce_vsi - Resume a given VSI
 * @vsi: the VSI being resumed
 **/
static void i40e_unquiesce_vsi(struct i40e_vsi *vsi)
{
	if (!test_bit(__I40E_NEEDS_RESTART, &vsi->state))
		return;

	clear_bit(__I40E_NEEDS_RESTART, &vsi->state);
	if (vsi->netdev && netif_running(vsi->netdev))
		vsi->netdev->netdev_ops->ndo_open(vsi->netdev);
	else
4346
		i40e_vsi_open(vsi);   /* this clears the DOWN bit */
J
Jesse Brandeburg 已提交
4347 4348 4349 4350 4351 4352 4353 4354 4355 4356
}

/**
 * i40e_pf_quiesce_all_vsi - Pause all VSIs on a PF
 * @pf: the PF
 **/
static void i40e_pf_quiesce_all_vsi(struct i40e_pf *pf)
{
	int v;

M
Mitch Williams 已提交
4357
	for (v = 0; v < pf->num_alloc_vsi; v++) {
J
Jesse Brandeburg 已提交
4358 4359 4360 4361 4362 4363 4364 4365 4366 4367 4368 4369 4370
		if (pf->vsi[v])
			i40e_quiesce_vsi(pf->vsi[v]);
	}
}

/**
 * i40e_pf_unquiesce_all_vsi - Resume all VSIs on a PF
 * @pf: the PF
 **/
static void i40e_pf_unquiesce_all_vsi(struct i40e_pf *pf)
{
	int v;

M
Mitch Williams 已提交
4371
	for (v = 0; v < pf->num_alloc_vsi; v++) {
J
Jesse Brandeburg 已提交
4372 4373 4374 4375 4376
		if (pf->vsi[v])
			i40e_unquiesce_vsi(pf->vsi[v]);
	}
}

4377 4378
#ifdef CONFIG_I40E_DCB
/**
4379
 * i40e_vsi_wait_queues_disabled - Wait for VSI's queues to be disabled
4380 4381
 * @vsi: the VSI being configured
 *
4382
 * This function waits for the given VSI's queues to be disabled.
4383
 **/
4384
static int i40e_vsi_wait_queues_disabled(struct i40e_vsi *vsi)
4385 4386 4387 4388 4389 4390 4391 4392 4393 4394
{
	struct i40e_pf *pf = vsi->back;
	int i, pf_q, ret;

	pf_q = vsi->base_queue;
	for (i = 0; i < vsi->num_queue_pairs; i++, pf_q++) {
		/* Check and wait for the disable status of the queue */
		ret = i40e_pf_txq_wait(pf, pf_q, false);
		if (ret) {
			dev_info(&pf->pdev->dev,
S
Shannon Nelson 已提交
4395 4396
				 "VSI seid %d Tx ring %d disable timeout\n",
				 vsi->seid, pf_q);
4397 4398 4399 4400
			return ret;
		}
	}

4401 4402 4403 4404 4405 4406 4407 4408 4409 4410 4411 4412
	pf_q = vsi->base_queue;
	for (i = 0; i < vsi->num_queue_pairs; i++, pf_q++) {
		/* Check and wait for the disable status of the queue */
		ret = i40e_pf_rxq_wait(pf, pf_q, false);
		if (ret) {
			dev_info(&pf->pdev->dev,
				 "VSI seid %d Rx ring %d disable timeout\n",
				 vsi->seid, pf_q);
			return ret;
		}
	}

4413 4414 4415 4416
	return 0;
}

/**
4417
 * i40e_pf_wait_queues_disabled - Wait for all queues of PF VSIs to be disabled
4418 4419
 * @pf: the PF
 *
4420
 * This function waits for the queues to be in disabled state for all the
4421 4422
 * VSIs that are managed by this PF.
 **/
4423
static int i40e_pf_wait_queues_disabled(struct i40e_pf *pf)
4424 4425 4426 4427
{
	int v, ret = 0;

	for (v = 0; v < pf->hw.func_caps.num_vsis; v++) {
4428 4429
		/* No need to wait for FCoE VSI queues */
		if (pf->vsi[v] && pf->vsi[v]->type != I40E_VSI_FCOE) {
4430
			ret = i40e_vsi_wait_queues_disabled(pf->vsi[v]);
4431 4432 4433 4434 4435 4436 4437 4438 4439
			if (ret)
				break;
		}
	}

	return ret;
}

#endif
4440 4441 4442 4443 4444 4445 4446 4447 4448 4449 4450 4451 4452 4453 4454 4455

/**
 * i40e_detect_recover_hung_queue - Function to detect and recover hung_queue
 * @q_idx: TX queue number
 * @vsi: Pointer to VSI struct
 *
 * This function checks specified queue for given VSI. Detects hung condition.
 * Sets hung bit since it is two step process. Before next run of service task
 * if napi_poll runs, it reset 'hung' bit for respective q_vector. If not,
 * hung condition remain unchanged and during subsequent run, this function
 * issues SW interrupt to recover from hung condition.
 **/
static void i40e_detect_recover_hung_queue(int q_idx, struct i40e_vsi *vsi)
{
	struct i40e_ring *tx_ring = NULL;
	struct i40e_pf	*pf;
4456
	u32 head, val, tx_pending_hw;
4457 4458 4459 4460 4461 4462 4463 4464 4465 4466 4467 4468 4469 4470 4471 4472 4473 4474 4475 4476 4477 4478 4479 4480 4481 4482 4483
	int i;

	pf = vsi->back;

	/* now that we have an index, find the tx_ring struct */
	for (i = 0; i < vsi->num_queue_pairs; i++) {
		if (vsi->tx_rings[i] && vsi->tx_rings[i]->desc) {
			if (q_idx == vsi->tx_rings[i]->queue_index) {
				tx_ring = vsi->tx_rings[i];
				break;
			}
		}
	}

	if (!tx_ring)
		return;

	/* Read interrupt register */
	if (pf->flags & I40E_FLAG_MSIX_ENABLED)
		val = rd32(&pf->hw,
			   I40E_PFINT_DYN_CTLN(tx_ring->q_vector->v_idx +
					       tx_ring->vsi->base_vector - 1));
	else
		val = rd32(&pf->hw, I40E_PFINT_DYN_CTL0);

	head = i40e_get_head(tx_ring);

4484
	tx_pending_hw = i40e_get_tx_pending(tx_ring, false);
4485

4486 4487 4488 4489 4490
	/* HW is done executing descriptors, updated HEAD write back,
	 * but SW hasn't processed those descriptors. If interrupt is
	 * not generated from this point ON, it could result into
	 * dev_watchdog detecting timeout on those netdev_queue,
	 * hence proactively trigger SW interrupt.
4491
	 */
4492
	if (tx_pending_hw && (!(val & I40E_PFINT_DYN_CTLN_INTENA_MASK))) {
4493 4494 4495
		/* NAPI Poll didn't run and clear since it was set */
		if (test_and_clear_bit(I40E_Q_VECTOR_HUNG_DETECT,
				       &tx_ring->q_vector->hung_detected)) {
4496 4497
			netdev_info(vsi->netdev, "VSI_seid %d, Hung TX queue %d, tx_pending_hw: %d, NTC:0x%x, HWB: 0x%x, NTU: 0x%x, TAIL: 0x%x\n",
				    vsi->seid, q_idx, tx_pending_hw,
4498 4499 4500 4501 4502 4503 4504 4505 4506 4507 4508 4509
				    tx_ring->next_to_clean, head,
				    tx_ring->next_to_use,
				    readl(tx_ring->tail));
			netdev_info(vsi->netdev, "VSI_seid %d, Issuing force_wb for TX queue %d, Interrupt Reg: 0x%x\n",
				    vsi->seid, q_idx, val);
			i40e_force_wb(vsi, tx_ring->q_vector);
		} else {
			/* First Chance - detected possible hung */
			set_bit(I40E_Q_VECTOR_HUNG_DETECT,
				&tx_ring->q_vector->hung_detected);
		}
	}
4510 4511 4512 4513 4514 4515 4516 4517 4518 4519 4520

	/* This is the case where we have interrupts missing,
	 * so the tx_pending in HW will most likely be 0, but we
	 * will have tx_pending in SW since the WB happened but the
	 * interrupt got lost.
	 */
	if ((!tx_pending_hw) && i40e_get_tx_pending(tx_ring, true) &&
	    (!(val & I40E_PFINT_DYN_CTLN_INTENA_MASK))) {
		if (napi_reschedule(&tx_ring->q_vector->napi))
			tx_ring->tx_stats.tx_lost_interrupt++;
	}
4521 4522 4523 4524 4525 4526 4527 4528 4529 4530 4531 4532 4533 4534 4535 4536 4537 4538 4539 4540 4541 4542 4543 4544 4545 4546 4547 4548 4549 4550 4551 4552 4553 4554 4555 4556 4557 4558 4559 4560 4561 4562 4563 4564 4565 4566 4567 4568 4569
}

/**
 * i40e_detect_recover_hung - Function to detect and recover hung_queues
 * @pf:  pointer to PF struct
 *
 * LAN VSI has netdev and netdev has TX queues. This function is to check
 * each of those TX queues if they are hung, trigger recovery by issuing
 * SW interrupt.
 **/
static void i40e_detect_recover_hung(struct i40e_pf *pf)
{
	struct net_device *netdev;
	struct i40e_vsi *vsi;
	int i;

	/* Only for LAN VSI */
	vsi = pf->vsi[pf->lan_vsi];

	if (!vsi)
		return;

	/* Make sure, VSI state is not DOWN/RECOVERY_PENDING */
	if (test_bit(__I40E_DOWN, &vsi->back->state) ||
	    test_bit(__I40E_RESET_RECOVERY_PENDING, &vsi->back->state))
		return;

	/* Make sure type is MAIN VSI */
	if (vsi->type != I40E_VSI_MAIN)
		return;

	netdev = vsi->netdev;
	if (!netdev)
		return;

	/* Bail out if netif_carrier is not OK */
	if (!netif_carrier_ok(netdev))
		return;

	/* Go thru' TX queues for netdev */
	for (i = 0; i < netdev->num_tx_queues; i++) {
		struct netdev_queue *q;

		q = netdev_get_tx_queue(netdev, i);
		if (q)
			i40e_detect_recover_hung_queue(i, vsi);
	}
}

4570 4571
/**
 * i40e_get_iscsi_tc_map - Return TC map for iSCSI APP
4572
 * @pf: pointer to PF
4573 4574 4575 4576 4577 4578 4579 4580 4581 4582 4583 4584 4585 4586 4587 4588 4589 4590
 *
 * Get TC map for ISCSI PF type that will include iSCSI TC
 * and LAN TC.
 **/
static u8 i40e_get_iscsi_tc_map(struct i40e_pf *pf)
{
	struct i40e_dcb_app_priority_table app;
	struct i40e_hw *hw = &pf->hw;
	u8 enabled_tc = 1; /* TC0 is always enabled */
	u8 tc, i;
	/* Get the iSCSI APP TLV */
	struct i40e_dcbx_config *dcbcfg = &hw->local_dcbx_config;

	for (i = 0; i < dcbcfg->numapps; i++) {
		app = dcbcfg->app[i];
		if (app.selector == I40E_APP_SEL_TCPIP &&
		    app.protocolid == I40E_APP_PROTOID_ISCSI) {
			tc = dcbcfg->etscfg.prioritytable[app.priority];
J
Jesse Brandeburg 已提交
4591
			enabled_tc |= BIT(tc);
4592 4593 4594 4595 4596 4597 4598
			break;
		}
	}

	return enabled_tc;
}

J
Jesse Brandeburg 已提交
4599 4600 4601 4602 4603 4604 4605 4606
/**
 * i40e_dcb_get_num_tc -  Get the number of TCs from DCBx config
 * @dcbcfg: the corresponding DCBx configuration structure
 *
 * Return the number of TCs from given DCBx configuration
 **/
static u8 i40e_dcb_get_num_tc(struct i40e_dcbx_config *dcbcfg)
{
4607
	int i, tc_unused = 0;
J
Jesse Brandeburg 已提交
4608
	u8 num_tc = 0;
4609
	u8 ret = 0;
J
Jesse Brandeburg 已提交
4610 4611 4612

	/* Scan the ETS Config Priority Table to find
	 * traffic class enabled for a given priority
4613
	 * and create a bitmask of enabled TCs
J
Jesse Brandeburg 已提交
4614
	 */
4615 4616
	for (i = 0; i < I40E_MAX_USER_PRIORITY; i++)
		num_tc |= BIT(dcbcfg->etscfg.prioritytable[i]);
J
Jesse Brandeburg 已提交
4617

4618 4619
	/* Now scan the bitmask to check for
	 * contiguous TCs starting with TC0
J
Jesse Brandeburg 已提交
4620
	 */
4621 4622 4623 4624 4625 4626 4627 4628 4629 4630 4631 4632 4633 4634 4635 4636 4637 4638
	for (i = 0; i < I40E_MAX_TRAFFIC_CLASS; i++) {
		if (num_tc & BIT(i)) {
			if (!tc_unused) {
				ret++;
			} else {
				pr_err("Non-contiguous TC - Disabling DCB\n");
				return 1;
			}
		} else {
			tc_unused = 1;
		}
	}

	/* There is always at least TC0 */
	if (!ret)
		ret = 1;

	return ret;
J
Jesse Brandeburg 已提交
4639 4640 4641 4642 4643 4644 4645 4646 4647 4648 4649 4650 4651 4652 4653 4654
}

/**
 * i40e_dcb_get_enabled_tc - Get enabled traffic classes
 * @dcbcfg: the corresponding DCBx configuration structure
 *
 * Query the current DCB configuration and return the number of
 * traffic classes enabled from the given DCBX config
 **/
static u8 i40e_dcb_get_enabled_tc(struct i40e_dcbx_config *dcbcfg)
{
	u8 num_tc = i40e_dcb_get_num_tc(dcbcfg);
	u8 enabled_tc = 1;
	u8 i;

	for (i = 0; i < num_tc; i++)
4655
		enabled_tc |= BIT(i);
J
Jesse Brandeburg 已提交
4656 4657 4658 4659 4660 4661 4662 4663 4664 4665 4666 4667 4668

	return enabled_tc;
}

/**
 * i40e_pf_get_num_tc - Get enabled traffic classes for PF
 * @pf: PF being queried
 *
 * Return number of traffic classes enabled for the given PF
 **/
static u8 i40e_pf_get_num_tc(struct i40e_pf *pf)
{
	struct i40e_hw *hw = &pf->hw;
4669
	u8 i, enabled_tc = 1;
J
Jesse Brandeburg 已提交
4670 4671 4672 4673 4674 4675 4676
	u8 num_tc = 0;
	struct i40e_dcbx_config *dcbcfg = &hw->local_dcbx_config;

	/* If DCB is not enabled then always in single TC */
	if (!(pf->flags & I40E_FLAG_DCB_ENABLED))
		return 1;

4677 4678 4679 4680
	/* SFP mode will be enabled for all TCs on port */
	if (!(pf->flags & I40E_FLAG_MFP_ENABLED))
		return i40e_dcb_get_num_tc(dcbcfg);

J
Jesse Brandeburg 已提交
4681
	/* MFP mode return count of enabled TCs for this PF */
4682 4683 4684
	if (pf->hw.func_caps.iscsi)
		enabled_tc =  i40e_get_iscsi_tc_map(pf);
	else
4685
		return 1; /* Only TC0 */
J
Jesse Brandeburg 已提交
4686

4687
	for (i = 0; i < I40E_MAX_TRAFFIC_CLASS; i++) {
J
Jesse Brandeburg 已提交
4688
		if (enabled_tc & BIT(i))
4689 4690 4691
			num_tc++;
	}
	return num_tc;
J
Jesse Brandeburg 已提交
4692 4693 4694 4695 4696 4697 4698 4699 4700 4701 4702 4703
}

/**
 * i40e_pf_get_pf_tc_map - Get bitmap for enabled traffic classes
 * @pf: PF being queried
 *
 * Return a bitmap for enabled traffic classes for this PF.
 **/
static u8 i40e_pf_get_tc_map(struct i40e_pf *pf)
{
	/* If DCB is not enabled for this PF then just return default TC */
	if (!(pf->flags & I40E_FLAG_DCB_ENABLED))
4704
		return I40E_DEFAULT_TRAFFIC_CLASS;
J
Jesse Brandeburg 已提交
4705 4706

	/* SFP mode we want PF to be enabled for all TCs */
4707 4708 4709
	if (!(pf->flags & I40E_FLAG_MFP_ENABLED))
		return i40e_dcb_get_enabled_tc(&pf->hw.local_dcbx_config);

4710
	/* MFP enabled and iSCSI PF type */
4711 4712 4713
	if (pf->hw.func_caps.iscsi)
		return i40e_get_iscsi_tc_map(pf);
	else
4714
		return I40E_DEFAULT_TRAFFIC_CLASS;
J
Jesse Brandeburg 已提交
4715 4716 4717 4718 4719 4720 4721 4722 4723 4724 4725 4726 4727 4728
}

/**
 * i40e_vsi_get_bw_info - Query VSI BW Information
 * @vsi: the VSI being queried
 *
 * Returns 0 on success, negative value on failure
 **/
static int i40e_vsi_get_bw_info(struct i40e_vsi *vsi)
{
	struct i40e_aqc_query_vsi_ets_sla_config_resp bw_ets_config = {0};
	struct i40e_aqc_query_vsi_bw_config_resp bw_config = {0};
	struct i40e_pf *pf = vsi->back;
	struct i40e_hw *hw = &pf->hw;
4729
	i40e_status ret;
J
Jesse Brandeburg 已提交
4730 4731 4732 4733
	u32 tc_bw_max;
	int i;

	/* Get the VSI level BW configuration */
4734 4735
	ret = i40e_aq_query_vsi_bw_config(hw, vsi->seid, &bw_config, NULL);
	if (ret) {
J
Jesse Brandeburg 已提交
4736
		dev_info(&pf->pdev->dev,
4737 4738 4739
			 "couldn't get PF vsi bw config, err %s aq_err %s\n",
			 i40e_stat_str(&pf->hw, ret),
			 i40e_aq_str(&pf->hw, pf->hw.aq.asq_last_status));
J
Jesse Brandeburg 已提交
4740
		return -EINVAL;
J
Jesse Brandeburg 已提交
4741 4742 4743
	}

	/* Get the VSI level BW configuration per TC */
4744 4745 4746
	ret = i40e_aq_query_vsi_ets_sla_config(hw, vsi->seid, &bw_ets_config,
					       NULL);
	if (ret) {
J
Jesse Brandeburg 已提交
4747
		dev_info(&pf->pdev->dev,
4748 4749 4750
			 "couldn't get PF vsi ets bw config, err %s aq_err %s\n",
			 i40e_stat_str(&pf->hw, ret),
			 i40e_aq_str(&pf->hw, pf->hw.aq.asq_last_status));
J
Jesse Brandeburg 已提交
4751
		return -EINVAL;
J
Jesse Brandeburg 已提交
4752 4753 4754 4755 4756 4757 4758 4759 4760 4761 4762 4763 4764 4765 4766 4767 4768 4769 4770 4771 4772
	}

	if (bw_config.tc_valid_bits != bw_ets_config.tc_valid_bits) {
		dev_info(&pf->pdev->dev,
			 "Enabled TCs mismatch from querying VSI BW info 0x%08x 0x%08x\n",
			 bw_config.tc_valid_bits,
			 bw_ets_config.tc_valid_bits);
		/* Still continuing */
	}

	vsi->bw_limit = le16_to_cpu(bw_config.port_bw_limit);
	vsi->bw_max_quanta = bw_config.max_bw;
	tc_bw_max = le16_to_cpu(bw_ets_config.tc_bw_max[0]) |
		    (le16_to_cpu(bw_ets_config.tc_bw_max[1]) << 16);
	for (i = 0; i < I40E_MAX_TRAFFIC_CLASS; i++) {
		vsi->bw_ets_share_credits[i] = bw_ets_config.share_credits[i];
		vsi->bw_ets_limit_credits[i] =
					le16_to_cpu(bw_ets_config.credits[i]);
		/* 3 bits out of 4 for each TC */
		vsi->bw_ets_max_quanta[i] = (u8)((tc_bw_max >> (i*4)) & 0x7);
	}
J
Jesse Brandeburg 已提交
4773

J
Jesse Brandeburg 已提交
4774
	return 0;
J
Jesse Brandeburg 已提交
4775 4776 4777 4778 4779 4780 4781 4782 4783 4784
}

/**
 * i40e_vsi_configure_bw_alloc - Configure VSI BW allocation per TC
 * @vsi: the VSI being configured
 * @enabled_tc: TC bitmap
 * @bw_credits: BW shared credits per TC
 *
 * Returns 0 on success, negative value on failure
 **/
J
Jesse Brandeburg 已提交
4785
static int i40e_vsi_configure_bw_alloc(struct i40e_vsi *vsi, u8 enabled_tc,
J
Jesse Brandeburg 已提交
4786 4787 4788
				       u8 *bw_share)
{
	struct i40e_aqc_configure_vsi_tc_bw_data bw_data;
4789
	i40e_status ret;
J
Jesse Brandeburg 已提交
4790
	int i;
J
Jesse Brandeburg 已提交
4791 4792 4793 4794 4795

	bw_data.tc_valid_bits = enabled_tc;
	for (i = 0; i < I40E_MAX_TRAFFIC_CLASS; i++)
		bw_data.tc_bw_credits[i] = bw_share[i];

4796 4797 4798
	ret = i40e_aq_config_vsi_tc_bw(&vsi->back->hw, vsi->seid, &bw_data,
				       NULL);
	if (ret) {
J
Jesse Brandeburg 已提交
4799
		dev_info(&vsi->back->pdev->dev,
J
Jesse Brandeburg 已提交
4800 4801
			 "AQ command Config VSI BW allocation per TC failed = %d\n",
			 vsi->back->hw.aq.asq_last_status);
J
Jesse Brandeburg 已提交
4802
		return -EINVAL;
J
Jesse Brandeburg 已提交
4803 4804 4805 4806 4807
	}

	for (i = 0; i < I40E_MAX_TRAFFIC_CLASS; i++)
		vsi->info.qs_handle[i] = bw_data.qs_handles[i];

J
Jesse Brandeburg 已提交
4808
	return 0;
J
Jesse Brandeburg 已提交
4809 4810 4811 4812 4813 4814 4815 4816 4817 4818 4819 4820 4821 4822 4823 4824 4825 4826 4827 4828 4829 4830 4831 4832 4833 4834 4835 4836 4837 4838 4839 4840 4841 4842 4843 4844 4845 4846
}

/**
 * i40e_vsi_config_netdev_tc - Setup the netdev TC configuration
 * @vsi: the VSI being configured
 * @enabled_tc: TC map to be enabled
 *
 **/
static void i40e_vsi_config_netdev_tc(struct i40e_vsi *vsi, u8 enabled_tc)
{
	struct net_device *netdev = vsi->netdev;
	struct i40e_pf *pf = vsi->back;
	struct i40e_hw *hw = &pf->hw;
	u8 netdev_tc = 0;
	int i;
	struct i40e_dcbx_config *dcbcfg = &hw->local_dcbx_config;

	if (!netdev)
		return;

	if (!enabled_tc) {
		netdev_reset_tc(netdev);
		return;
	}

	/* Set up actual enabled TCs on the VSI */
	if (netdev_set_num_tc(netdev, vsi->tc_config.numtc))
		return;

	/* set per TC queues for the VSI */
	for (i = 0; i < I40E_MAX_TRAFFIC_CLASS; i++) {
		/* Only set TC queues for enabled tcs
		 *
		 * e.g. For a VSI that has TC0 and TC3 enabled the
		 * enabled_tc bitmap would be 0x00001001; the driver
		 * will set the numtc for netdev as 2 that will be
		 * referenced by the netdev layer as TC 0 and 1.
		 */
J
Jesse Brandeburg 已提交
4847
		if (vsi->tc_config.enabled_tc & BIT(i))
J
Jesse Brandeburg 已提交
4848 4849 4850 4851 4852 4853 4854 4855 4856 4857 4858 4859 4860 4861 4862 4863 4864 4865 4866 4867 4868 4869 4870 4871 4872 4873 4874 4875 4876 4877 4878 4879 4880 4881 4882 4883 4884 4885 4886 4887 4888 4889 4890 4891 4892 4893 4894 4895 4896 4897 4898 4899 4900 4901 4902 4903 4904 4905 4906 4907 4908
			netdev_set_tc_queue(netdev,
					vsi->tc_config.tc_info[i].netdev_tc,
					vsi->tc_config.tc_info[i].qcount,
					vsi->tc_config.tc_info[i].qoffset);
	}

	/* Assign UP2TC map for the VSI */
	for (i = 0; i < I40E_MAX_USER_PRIORITY; i++) {
		/* Get the actual TC# for the UP */
		u8 ets_tc = dcbcfg->etscfg.prioritytable[i];
		/* Get the mapped netdev TC# for the UP */
		netdev_tc =  vsi->tc_config.tc_info[ets_tc].netdev_tc;
		netdev_set_prio_tc_map(netdev, i, netdev_tc);
	}
}

/**
 * i40e_vsi_update_queue_map - Update our copy of VSi info with new queue map
 * @vsi: the VSI being configured
 * @ctxt: the ctxt buffer returned from AQ VSI update param command
 **/
static void i40e_vsi_update_queue_map(struct i40e_vsi *vsi,
				      struct i40e_vsi_context *ctxt)
{
	/* copy just the sections touched not the entire info
	 * since not all sections are valid as returned by
	 * update vsi params
	 */
	vsi->info.mapping_flags = ctxt->info.mapping_flags;
	memcpy(&vsi->info.queue_mapping,
	       &ctxt->info.queue_mapping, sizeof(vsi->info.queue_mapping));
	memcpy(&vsi->info.tc_mapping, ctxt->info.tc_mapping,
	       sizeof(vsi->info.tc_mapping));
}

/**
 * i40e_vsi_config_tc - Configure VSI Tx Scheduler for given TC map
 * @vsi: VSI to be configured
 * @enabled_tc: TC bitmap
 *
 * This configures a particular VSI for TCs that are mapped to the
 * given TC bitmap. It uses default bandwidth share for TCs across
 * VSIs to configure TC for a particular VSI.
 *
 * NOTE:
 * It is expected that the VSI queues have been quisced before calling
 * this function.
 **/
static int i40e_vsi_config_tc(struct i40e_vsi *vsi, u8 enabled_tc)
{
	u8 bw_share[I40E_MAX_TRAFFIC_CLASS] = {0};
	struct i40e_vsi_context ctxt;
	int ret = 0;
	int i;

	/* Check if enabled_tc is same as existing or new TCs */
	if (vsi->tc_config.enabled_tc == enabled_tc)
		return ret;

	/* Enable ETS TCs with equal BW Share for now across all VSIs */
	for (i = 0; i < I40E_MAX_TRAFFIC_CLASS; i++) {
J
Jesse Brandeburg 已提交
4909
		if (enabled_tc & BIT(i))
J
Jesse Brandeburg 已提交
4910 4911 4912 4913 4914 4915 4916 4917 4918 4919 4920 4921 4922 4923 4924 4925
			bw_share[i] = 1;
	}

	ret = i40e_vsi_configure_bw_alloc(vsi, enabled_tc, bw_share);
	if (ret) {
		dev_info(&vsi->back->pdev->dev,
			 "Failed configuring TC map %d for VSI %d\n",
			 enabled_tc, vsi->seid);
		goto out;
	}

	/* Update Queue Pairs Mapping for currently enabled UPs */
	ctxt.seid = vsi->seid;
	ctxt.pf_num = vsi->back->hw.pf_id;
	ctxt.vf_num = 0;
	ctxt.uplink_seid = vsi->uplink_seid;
4926
	ctxt.info = vsi->info;
J
Jesse Brandeburg 已提交
4927 4928
	i40e_vsi_setup_queue_map(vsi, &ctxt, enabled_tc, false);

4929 4930 4931 4932 4933 4934
	if (vsi->back->flags & I40E_FLAG_IWARP_ENABLED) {
		ctxt.info.valid_sections |=
				cpu_to_le16(I40E_AQ_VSI_PROP_QUEUE_OPT_VALID);
		ctxt.info.queueing_opt_flags |= I40E_AQ_VSI_QUE_OPT_TCP_ENA;
	}

J
Jesse Brandeburg 已提交
4935 4936 4937 4938
	/* Update the VSI after updating the VSI queue-mapping information */
	ret = i40e_aq_update_vsi_params(&vsi->back->hw, &ctxt, NULL);
	if (ret) {
		dev_info(&vsi->back->pdev->dev,
4939 4940 4941 4942
			 "Update vsi tc config failed, err %s aq_err %s\n",
			 i40e_stat_str(&vsi->back->hw, ret),
			 i40e_aq_str(&vsi->back->hw,
				     vsi->back->hw.aq.asq_last_status));
J
Jesse Brandeburg 已提交
4943 4944 4945 4946 4947 4948 4949 4950 4951 4952
		goto out;
	}
	/* update the local VSI info with updated queue map */
	i40e_vsi_update_queue_map(vsi, &ctxt);
	vsi->info.valid_sections = 0;

	/* Update current VSI BW information */
	ret = i40e_vsi_get_bw_info(vsi);
	if (ret) {
		dev_info(&vsi->back->pdev->dev,
4953 4954 4955 4956
			 "Failed updating vsi bw info, err %s aq_err %s\n",
			 i40e_stat_str(&vsi->back->hw, ret),
			 i40e_aq_str(&vsi->back->hw,
				     vsi->back->hw.aq.asq_last_status));
J
Jesse Brandeburg 已提交
4957 4958 4959 4960 4961 4962 4963 4964 4965
		goto out;
	}

	/* Update the netdev TC setup */
	i40e_vsi_config_netdev_tc(vsi, enabled_tc);
out:
	return ret;
}

N
Neerav Parikh 已提交
4966 4967 4968 4969 4970 4971 4972 4973 4974 4975 4976 4977 4978 4979 4980 4981 4982 4983 4984 4985 4986 4987 4988
/**
 * i40e_veb_config_tc - Configure TCs for given VEB
 * @veb: given VEB
 * @enabled_tc: TC bitmap
 *
 * Configures given TC bitmap for VEB (switching) element
 **/
int i40e_veb_config_tc(struct i40e_veb *veb, u8 enabled_tc)
{
	struct i40e_aqc_configure_switching_comp_bw_config_data bw_data = {0};
	struct i40e_pf *pf = veb->pf;
	int ret = 0;
	int i;

	/* No TCs or already enabled TCs just return */
	if (!enabled_tc || veb->enabled_tc == enabled_tc)
		return ret;

	bw_data.tc_valid_bits = enabled_tc;
	/* bw_data.absolute_credits is not set (relative) */

	/* Enable ETS TCs with equal BW Share for now */
	for (i = 0; i < I40E_MAX_TRAFFIC_CLASS; i++) {
J
Jesse Brandeburg 已提交
4989
		if (enabled_tc & BIT(i))
N
Neerav Parikh 已提交
4990 4991 4992 4993 4994 4995 4996
			bw_data.tc_bw_share_credits[i] = 1;
	}

	ret = i40e_aq_config_switch_comp_bw_config(&pf->hw, veb->seid,
						   &bw_data, NULL);
	if (ret) {
		dev_info(&pf->pdev->dev,
4997 4998 4999
			 "VEB bw config failed, err %s aq_err %s\n",
			 i40e_stat_str(&pf->hw, ret),
			 i40e_aq_str(&pf->hw, pf->hw.aq.asq_last_status));
N
Neerav Parikh 已提交
5000 5001 5002 5003 5004 5005 5006
		goto out;
	}

	/* Update the BW information */
	ret = i40e_veb_get_bw_info(veb);
	if (ret) {
		dev_info(&pf->pdev->dev,
5007 5008 5009
			 "Failed getting veb bw config, err %s aq_err %s\n",
			 i40e_stat_str(&pf->hw, ret),
			 i40e_aq_str(&pf->hw, pf->hw.aq.asq_last_status));
N
Neerav Parikh 已提交
5010 5011 5012 5013 5014 5015 5016 5017 5018 5019 5020 5021 5022 5023 5024 5025 5026 5027 5028 5029 5030 5031 5032 5033 5034 5035 5036 5037 5038 5039 5040 5041 5042 5043 5044 5045
	}

out:
	return ret;
}

#ifdef CONFIG_I40E_DCB
/**
 * i40e_dcb_reconfigure - Reconfigure all VEBs and VSIs
 * @pf: PF struct
 *
 * Reconfigure VEB/VSIs on a given PF; it is assumed that
 * the caller would've quiesce all the VSIs before calling
 * this function
 **/
static void i40e_dcb_reconfigure(struct i40e_pf *pf)
{
	u8 tc_map = 0;
	int ret;
	u8 v;

	/* Enable the TCs available on PF to all VEBs */
	tc_map = i40e_pf_get_tc_map(pf);
	for (v = 0; v < I40E_MAX_VEB; v++) {
		if (!pf->veb[v])
			continue;
		ret = i40e_veb_config_tc(pf->veb[v], tc_map);
		if (ret) {
			dev_info(&pf->pdev->dev,
				 "Failed configuring TC for VEB seid=%d\n",
				 pf->veb[v]->seid);
			/* Will try to configure as many components */
		}
	}

	/* Update each VSI */
M
Mitch Williams 已提交
5046
	for (v = 0; v < pf->num_alloc_vsi; v++) {
N
Neerav Parikh 已提交
5047 5048 5049 5050
		if (!pf->vsi[v])
			continue;

		/* - Enable all TCs for the LAN VSI
5051 5052 5053 5054
#ifdef I40E_FCOE
		 * - For FCoE VSI only enable the TC configured
		 *   as per the APP TLV
#endif
N
Neerav Parikh 已提交
5055 5056 5057 5058 5059
		 * - For all others keep them at TC0 for now
		 */
		if (v == pf->lan_vsi)
			tc_map = i40e_pf_get_tc_map(pf);
		else
5060
			tc_map = I40E_DEFAULT_TRAFFIC_CLASS;
5061 5062 5063 5064
#ifdef I40E_FCOE
		if (pf->vsi[v]->type == I40E_VSI_FCOE)
			tc_map = i40e_get_fcoe_tc_map(pf);
#endif /* #ifdef I40E_FCOE */
N
Neerav Parikh 已提交
5065 5066 5067 5068 5069 5070 5071 5072

		ret = i40e_vsi_config_tc(pf->vsi[v], tc_map);
		if (ret) {
			dev_info(&pf->pdev->dev,
				 "Failed configuring TC for VSI seid=%d\n",
				 pf->vsi[v]->seid);
			/* Will try to configure as many components */
		} else {
5073 5074
			/* Re-configure VSI vectors based on updated TC map */
			i40e_vsi_map_rings_to_vectors(pf->vsi[v]);
N
Neerav Parikh 已提交
5075 5076 5077 5078 5079 5080
			if (pf->vsi[v]->netdev)
				i40e_dcbnl_set_all(pf->vsi[v]);
		}
	}
}

5081 5082 5083 5084 5085 5086 5087 5088 5089 5090 5091 5092 5093 5094 5095
/**
 * i40e_resume_port_tx - Resume port Tx
 * @pf: PF struct
 *
 * Resume a port's Tx and issue a PF reset in case of failure to
 * resume.
 **/
static int i40e_resume_port_tx(struct i40e_pf *pf)
{
	struct i40e_hw *hw = &pf->hw;
	int ret;

	ret = i40e_aq_resume_port_tx(hw, NULL);
	if (ret) {
		dev_info(&pf->pdev->dev,
5096 5097 5098
			 "Resume Port Tx failed, err %s aq_err %s\n",
			  i40e_stat_str(&pf->hw, ret),
			  i40e_aq_str(&pf->hw, pf->hw.aq.asq_last_status));
5099 5100 5101 5102 5103 5104 5105 5106
		/* Schedule PF reset to recover */
		set_bit(__I40E_PF_RESET_REQUESTED, &pf->state);
		i40e_service_event_schedule(pf);
	}

	return ret;
}

N
Neerav Parikh 已提交
5107 5108 5109 5110 5111 5112 5113 5114 5115 5116 5117 5118
/**
 * i40e_init_pf_dcb - Initialize DCB configuration
 * @pf: PF being configured
 *
 * Query the current DCB configuration and cache it
 * in the hardware structure
 **/
static int i40e_init_pf_dcb(struct i40e_pf *pf)
{
	struct i40e_hw *hw = &pf->hw;
	int err = 0;

5119
	/* Do not enable DCB for SW1 and SW2 images even if the FW is capable */
5120
	if (pf->flags & I40E_FLAG_NO_DCB_SUPPORT)
5121 5122
		goto out;

N
Neerav Parikh 已提交
5123 5124 5125 5126 5127 5128 5129 5130 5131 5132 5133 5134 5135 5136 5137 5138
	/* Get the initial DCB configuration */
	err = i40e_init_dcb(hw);
	if (!err) {
		/* Device/Function is not DCBX capable */
		if ((!hw->func_caps.dcb) ||
		    (hw->dcbx_status == I40E_DCBX_STATUS_DISABLED)) {
			dev_info(&pf->pdev->dev,
				 "DCBX offload is not supported or is disabled for this PF.\n");

			if (pf->flags & I40E_FLAG_MFP_ENABLED)
				goto out;

		} else {
			/* When status is not DISABLED then DCBX in FW */
			pf->dcbx_cap = DCB_CAP_DCBX_LLD_MANAGED |
				       DCB_CAP_DCBX_VER_IEEE;
5139 5140

			pf->flags |= I40E_FLAG_DCB_CAPABLE;
5141 5142 5143
			/* Enable DCB tagging only when more than one TC
			 * or explicitly disable if only one TC
			 */
5144 5145
			if (i40e_dcb_get_num_tc(&hw->local_dcbx_config) > 1)
				pf->flags |= I40E_FLAG_DCB_ENABLED;
5146 5147
			else
				pf->flags &= ~I40E_FLAG_DCB_ENABLED;
5148 5149
			dev_dbg(&pf->pdev->dev,
				"DCBX offload is supported for this PF.\n");
N
Neerav Parikh 已提交
5150
		}
5151
	} else {
5152
		dev_info(&pf->pdev->dev,
5153 5154 5155
			 "Query for DCB configuration failed, err %s aq_err %s\n",
			 i40e_stat_str(&pf->hw, err),
			 i40e_aq_str(&pf->hw, pf->hw.aq.asq_last_status));
N
Neerav Parikh 已提交
5156 5157 5158 5159 5160 5161
	}

out:
	return err;
}
#endif /* CONFIG_I40E_DCB */
J
Jesse Brandeburg 已提交
5162 5163 5164 5165 5166 5167
#define SPEED_SIZE 14
#define FC_SIZE 8
/**
 * i40e_print_link_message - print link up or down
 * @vsi: the VSI for which link needs a message
 */
M
Matt Jared 已提交
5168
void i40e_print_link_message(struct i40e_vsi *vsi, bool isup)
J
Jesse Brandeburg 已提交
5169
{
5170 5171
	char *speed = "Unknown";
	char *fc = "Unknown";
J
Jesse Brandeburg 已提交
5172

M
Matt Jared 已提交
5173 5174 5175
	if (vsi->current_isup == isup)
		return;
	vsi->current_isup = isup;
J
Jesse Brandeburg 已提交
5176 5177 5178 5179 5180
	if (!isup) {
		netdev_info(vsi->netdev, "NIC Link is Down\n");
		return;
	}

5181 5182 5183 5184 5185 5186 5187 5188 5189
	/* Warn user if link speed on NPAR enabled partition is not at
	 * least 10GB
	 */
	if (vsi->back->hw.func_caps.npar_enable &&
	    (vsi->back->hw.phy.link_info.link_speed == I40E_LINK_SPEED_1GB ||
	     vsi->back->hw.phy.link_info.link_speed == I40E_LINK_SPEED_100MB))
		netdev_warn(vsi->netdev,
			    "The partition detected link speed that is less than 10Gbps\n");

J
Jesse Brandeburg 已提交
5190 5191
	switch (vsi->back->hw.phy.link_info.link_speed) {
	case I40E_LINK_SPEED_40GB:
5192
		speed = "40 G";
J
Jesse Brandeburg 已提交
5193
		break;
5194
	case I40E_LINK_SPEED_20GB:
5195
		speed = "20 G";
5196
		break;
J
Jesse Brandeburg 已提交
5197
	case I40E_LINK_SPEED_10GB:
5198
		speed = "10 G";
J
Jesse Brandeburg 已提交
5199 5200
		break;
	case I40E_LINK_SPEED_1GB:
5201
		speed = "1000 M";
J
Jesse Brandeburg 已提交
5202
		break;
M
Mitch Williams 已提交
5203
	case I40E_LINK_SPEED_100MB:
5204
		speed = "100 M";
M
Mitch Williams 已提交
5205
		break;
J
Jesse Brandeburg 已提交
5206 5207 5208 5209 5210 5211
	default:
		break;
	}

	switch (vsi->back->hw.fc.current_mode) {
	case I40E_FC_FULL:
5212
		fc = "RX/TX";
J
Jesse Brandeburg 已提交
5213 5214
		break;
	case I40E_FC_TX_PAUSE:
5215
		fc = "TX";
J
Jesse Brandeburg 已提交
5216 5217
		break;
	case I40E_FC_RX_PAUSE:
5218
		fc = "RX";
J
Jesse Brandeburg 已提交
5219 5220
		break;
	default:
5221
		fc = "None";
J
Jesse Brandeburg 已提交
5222 5223 5224
		break;
	}

5225
	netdev_info(vsi->netdev, "NIC Link is Up %sbps Full Duplex, Flow Control: %s\n",
J
Jesse Brandeburg 已提交
5226 5227
		    speed, fc);
}
N
Neerav Parikh 已提交
5228

J
Jesse Brandeburg 已提交
5229 5230 5231 5232 5233 5234 5235 5236 5237 5238 5239 5240 5241 5242 5243
/**
 * i40e_up_complete - Finish the last steps of bringing up a connection
 * @vsi: the VSI being configured
 **/
static int i40e_up_complete(struct i40e_vsi *vsi)
{
	struct i40e_pf *pf = vsi->back;
	int err;

	if (pf->flags & I40E_FLAG_MSIX_ENABLED)
		i40e_vsi_configure_msix(vsi);
	else
		i40e_configure_msi_and_legacy(vsi);

	/* start rings */
F
Filip Sadowski 已提交
5244
	err = i40e_vsi_start_rings(vsi);
J
Jesse Brandeburg 已提交
5245 5246 5247 5248 5249 5250 5251 5252 5253
	if (err)
		return err;

	clear_bit(__I40E_DOWN, &vsi->state);
	i40e_napi_enable_all(vsi);
	i40e_vsi_enable_irq(vsi);

	if ((pf->hw.phy.link_info.link_info & I40E_AQ_LINK_UP) &&
	    (vsi->netdev)) {
J
Jesse Brandeburg 已提交
5254
		i40e_print_link_message(vsi, true);
J
Jesse Brandeburg 已提交
5255 5256
		netif_tx_start_all_queues(vsi->netdev);
		netif_carrier_on(vsi->netdev);
A
Anjali Singhai 已提交
5257
	} else if (vsi->netdev) {
J
Jesse Brandeburg 已提交
5258
		i40e_print_link_message(vsi, false);
5259 5260 5261 5262 5263 5264 5265
		/* need to check for qualified module here*/
		if ((pf->hw.phy.link_info.link_info &
			I40E_AQ_MEDIA_AVAILABLE) &&
		    (!(pf->hw.phy.link_info.an_info &
			I40E_AQ_QUALIFIED_MODULE)))
			netdev_err(vsi->netdev,
				   "the driver failed to link because an unqualified module was detected.");
J
Jesse Brandeburg 已提交
5266
	}
5267 5268

	/* replay FDIR SB filters */
5269 5270 5271 5272
	if (vsi->type == I40E_VSI_FDIR) {
		/* reset fd counters */
		pf->fd_add_err = pf->fd_atr_cnt = 0;
		if (pf->fd_tcp_rule > 0) {
5273
			pf->auto_disable_flags |= I40E_FLAG_FD_ATR_ENABLED;
5274 5275
			if (I40E_DEBUG_FD & pf->hw.debug_mask)
				dev_info(&pf->pdev->dev, "Forcing ATR off, sideband rules for TCP/IPv4 exist\n");
5276 5277
			pf->fd_tcp_rule = 0;
		}
5278
		i40e_fdir_filter_restore(vsi);
5279
	}
5280 5281 5282 5283 5284

	/* On the next run of the service_task, notify any clients of the new
	 * opened netdev
	 */
	pf->flags |= I40E_FLAG_SERVICE_CLIENT_REQUESTED;
J
Jesse Brandeburg 已提交
5285 5286 5287 5288 5289 5290 5291 5292 5293 5294 5295 5296 5297 5298 5299 5300 5301 5302 5303 5304 5305 5306 5307 5308 5309 5310 5311 5312 5313 5314 5315 5316 5317 5318 5319 5320 5321 5322 5323 5324 5325 5326 5327 5328 5329 5330 5331 5332 5333 5334 5335 5336 5337 5338 5339 5340
	i40e_service_event_schedule(pf);

	return 0;
}

/**
 * i40e_vsi_reinit_locked - Reset the VSI
 * @vsi: the VSI being configured
 *
 * Rebuild the ring structs after some configuration
 * has changed, e.g. MTU size.
 **/
static void i40e_vsi_reinit_locked(struct i40e_vsi *vsi)
{
	struct i40e_pf *pf = vsi->back;

	WARN_ON(in_interrupt());
	while (test_and_set_bit(__I40E_CONFIG_BUSY, &pf->state))
		usleep_range(1000, 2000);
	i40e_down(vsi);

	i40e_up(vsi);
	clear_bit(__I40E_CONFIG_BUSY, &pf->state);
}

/**
 * i40e_up - Bring the connection back up after being down
 * @vsi: the VSI being configured
 **/
int i40e_up(struct i40e_vsi *vsi)
{
	int err;

	err = i40e_vsi_configure(vsi);
	if (!err)
		err = i40e_up_complete(vsi);

	return err;
}

/**
 * i40e_down - Shutdown the connection processing
 * @vsi: the VSI being stopped
 **/
void i40e_down(struct i40e_vsi *vsi)
{
	int i;

	/* It is assumed that the caller of this function
	 * sets the vsi->state __I40E_DOWN bit.
	 */
	if (vsi->netdev) {
		netif_carrier_off(vsi->netdev);
		netif_tx_disable(vsi->netdev);
	}
	i40e_vsi_disable_irq(vsi);
F
Filip Sadowski 已提交
5341
	i40e_vsi_stop_rings(vsi);
J
Jesse Brandeburg 已提交
5342 5343 5344
	i40e_napi_disable_all(vsi);

	for (i = 0; i < vsi->num_queue_pairs; i++) {
5345 5346
		i40e_clean_tx_ring(vsi->tx_rings[i]);
		i40e_clean_rx_ring(vsi->rx_rings[i]);
J
Jesse Brandeburg 已提交
5347
	}
5348 5349 5350

	i40e_notify_client_of_netdev_close(vsi, false);

J
Jesse Brandeburg 已提交
5351 5352 5353 5354 5355 5356 5357 5358 5359 5360 5361 5362 5363 5364 5365 5366 5367 5368 5369 5370 5371 5372 5373 5374 5375 5376 5377 5378 5379 5380 5381 5382 5383 5384 5385 5386
}

/**
 * i40e_setup_tc - configure multiple traffic classes
 * @netdev: net device to configure
 * @tc: number of traffic classes to enable
 **/
static int i40e_setup_tc(struct net_device *netdev, u8 tc)
{
	struct i40e_netdev_priv *np = netdev_priv(netdev);
	struct i40e_vsi *vsi = np->vsi;
	struct i40e_pf *pf = vsi->back;
	u8 enabled_tc = 0;
	int ret = -EINVAL;
	int i;

	/* Check if DCB enabled to continue */
	if (!(pf->flags & I40E_FLAG_DCB_ENABLED)) {
		netdev_info(netdev, "DCB is not enabled for adapter\n");
		goto exit;
	}

	/* Check if MFP enabled */
	if (pf->flags & I40E_FLAG_MFP_ENABLED) {
		netdev_info(netdev, "Configuring TC not supported in MFP mode\n");
		goto exit;
	}

	/* Check whether tc count is within enabled limit */
	if (tc > i40e_pf_get_num_tc(pf)) {
		netdev_info(netdev, "TC count greater than enabled on link for adapter\n");
		goto exit;
	}

	/* Generate TC map for number of tc requested */
	for (i = 0; i < tc; i++)
J
Jesse Brandeburg 已提交
5387
		enabled_tc |= BIT(i);
J
Jesse Brandeburg 已提交
5388 5389 5390 5391 5392 5393 5394 5395 5396 5397 5398 5399 5400 5401 5402 5403 5404 5405 5406 5407 5408 5409 5410

	/* Requesting same TC configuration as already enabled */
	if (enabled_tc == vsi->tc_config.enabled_tc)
		return 0;

	/* Quiesce VSI queues */
	i40e_quiesce_vsi(vsi);

	/* Configure VSI for enabled TCs */
	ret = i40e_vsi_config_tc(vsi, enabled_tc);
	if (ret) {
		netdev_info(netdev, "Failed configuring TC for VSI seid=%d\n",
			    vsi->seid);
		goto exit;
	}

	/* Unquiesce VSI */
	i40e_unquiesce_vsi(vsi);

exit:
	return ret;
}

5411
#ifdef I40E_FCOE
5412 5413
int __i40e_setup_tc(struct net_device *netdev, u32 handle, __be16 proto,
		    struct tc_to_netdev *tc)
5414
#else
5415 5416
static int __i40e_setup_tc(struct net_device *netdev, u32 handle, __be16 proto,
			   struct tc_to_netdev *tc)
5417 5418
#endif
{
5419
	if (handle != TC_H_ROOT || tc->type != TC_SETUP_MQPRIO)
5420
		return -EINVAL;
5421
	return i40e_setup_tc(netdev, tc->tc);
5422 5423
}

J
Jesse Brandeburg 已提交
5424 5425 5426 5427 5428 5429 5430 5431 5432 5433 5434 5435
/**
 * i40e_open - Called when a network interface is made active
 * @netdev: network interface device structure
 *
 * The open entry point is called when a network interface is made
 * active by the system (IFF_UP).  At this point all resources needed
 * for transmit and receive operations are allocated, the interrupt
 * handler is registered with the OS, the netdev watchdog subtask is
 * enabled, and the stack is notified that the interface is ready.
 *
 * Returns 0 on success, negative value on failure
 **/
5436
int i40e_open(struct net_device *netdev)
J
Jesse Brandeburg 已提交
5437 5438 5439 5440 5441 5442
{
	struct i40e_netdev_priv *np = netdev_priv(netdev);
	struct i40e_vsi *vsi = np->vsi;
	struct i40e_pf *pf = vsi->back;
	int err;

5443 5444 5445
	/* disallow open during test or if eeprom is broken */
	if (test_bit(__I40E_TESTING, &pf->state) ||
	    test_bit(__I40E_BAD_EEPROM, &pf->state))
J
Jesse Brandeburg 已提交
5446 5447 5448 5449
		return -EBUSY;

	netif_carrier_off(netdev);

5450 5451 5452 5453
	err = i40e_vsi_open(vsi);
	if (err)
		return err;

5454 5455 5456 5457 5458 5459 5460 5461
	/* configure global TSO hardware offload settings */
	wr32(&pf->hw, I40E_GLLAN_TSOMSK_F, be32_to_cpu(TCP_FLAG_PSH |
						       TCP_FLAG_FIN) >> 16);
	wr32(&pf->hw, I40E_GLLAN_TSOMSK_M, be32_to_cpu(TCP_FLAG_PSH |
						       TCP_FLAG_FIN |
						       TCP_FLAG_CWR) >> 16);
	wr32(&pf->hw, I40E_GLLAN_TSOMSK_L, be32_to_cpu(TCP_FLAG_CWR) >> 16);

5462
	udp_tunnel_get_rx_info(netdev);
5463

5464 5465 5466 5467 5468 5469 5470 5471 5472 5473 5474 5475 5476 5477
	return 0;
}

/**
 * i40e_vsi_open -
 * @vsi: the VSI to open
 *
 * Finish initialization of the VSI.
 *
 * Returns 0 on success, negative value on failure
 **/
int i40e_vsi_open(struct i40e_vsi *vsi)
{
	struct i40e_pf *pf = vsi->back;
5478
	char int_name[I40E_INT_NAME_STR_LEN];
5479 5480
	int err;

J
Jesse Brandeburg 已提交
5481 5482 5483 5484 5485 5486 5487 5488 5489 5490 5491 5492
	/* allocate descriptors */
	err = i40e_vsi_setup_tx_resources(vsi);
	if (err)
		goto err_setup_tx;
	err = i40e_vsi_setup_rx_resources(vsi);
	if (err)
		goto err_setup_rx;

	err = i40e_vsi_configure(vsi);
	if (err)
		goto err_setup_rx;

5493 5494 5495 5496 5497 5498
	if (vsi->netdev) {
		snprintf(int_name, sizeof(int_name) - 1, "%s-%s",
			 dev_driver_string(&pf->pdev->dev), vsi->netdev->name);
		err = i40e_vsi_request_irq(vsi, int_name);
		if (err)
			goto err_setup_rx;
J
Jesse Brandeburg 已提交
5499

5500 5501 5502 5503 5504
		/* Notify the stack of the actual queue counts. */
		err = netif_set_real_num_tx_queues(vsi->netdev,
						   vsi->num_queue_pairs);
		if (err)
			goto err_set_queues;
5505

5506 5507 5508 5509
		err = netif_set_real_num_rx_queues(vsi->netdev,
						   vsi->num_queue_pairs);
		if (err)
			goto err_set_queues;
5510 5511

	} else if (vsi->type == I40E_VSI_FDIR) {
5512
		snprintf(int_name, sizeof(int_name) - 1, "%s-%s:fdir",
5513 5514
			 dev_driver_string(&pf->pdev->dev),
			 dev_name(&pf->pdev->dev));
5515
		err = i40e_vsi_request_irq(vsi, int_name);
5516

5517
	} else {
5518
		err = -EINVAL;
5519 5520
		goto err_setup_rx;
	}
5521

J
Jesse Brandeburg 已提交
5522 5523 5524 5525 5526 5527 5528 5529
	err = i40e_up_complete(vsi);
	if (err)
		goto err_up_complete;

	return 0;

err_up_complete:
	i40e_down(vsi);
5530
err_set_queues:
J
Jesse Brandeburg 已提交
5531 5532 5533 5534 5535 5536
	i40e_vsi_free_irq(vsi);
err_setup_rx:
	i40e_vsi_free_rx_resources(vsi);
err_setup_tx:
	i40e_vsi_free_tx_resources(vsi);
	if (vsi == pf->vsi[pf->lan_vsi])
5537
		i40e_do_reset(pf, BIT_ULL(__I40E_PF_RESET_REQUESTED));
J
Jesse Brandeburg 已提交
5538 5539 5540 5541

	return err;
}

5542 5543
/**
 * i40e_fdir_filter_exit - Cleans up the Flow Director accounting
5544
 * @pf: Pointer to PF
5545 5546 5547 5548 5549 5550 5551 5552 5553 5554 5555 5556 5557 5558 5559 5560 5561
 *
 * This function destroys the hlist where all the Flow Director
 * filters were saved.
 **/
static void i40e_fdir_filter_exit(struct i40e_pf *pf)
{
	struct i40e_fdir_filter *filter;
	struct hlist_node *node2;

	hlist_for_each_entry_safe(filter, node2,
				  &pf->fdir_filter_list, fdir_node) {
		hlist_del(&filter->fdir_node);
		kfree(filter);
	}
	pf->fdir_pf_active_filters = 0;
}

J
Jesse Brandeburg 已提交
5562 5563 5564 5565 5566 5567 5568 5569 5570 5571
/**
 * i40e_close - Disables a network interface
 * @netdev: network interface device structure
 *
 * The close entry point is called when an interface is de-activated
 * by the OS.  The hardware is still under the driver's control, but
 * this netdev interface is disabled.
 *
 * Returns 0, this is not allowed to fail
 **/
5572
int i40e_close(struct net_device *netdev)
J
Jesse Brandeburg 已提交
5573 5574 5575 5576
{
	struct i40e_netdev_priv *np = netdev_priv(netdev);
	struct i40e_vsi *vsi = np->vsi;

5577
	i40e_vsi_close(vsi);
J
Jesse Brandeburg 已提交
5578 5579 5580 5581 5582 5583 5584 5585 5586 5587 5588 5589 5590 5591 5592 5593 5594 5595 5596

	return 0;
}

/**
 * i40e_do_reset - Start a PF or Core Reset sequence
 * @pf: board private structure
 * @reset_flags: which reset is requested
 *
 * The essential difference in resets is that the PF Reset
 * doesn't clear the packet buffers, doesn't reset the PE
 * firmware, and doesn't bother the other PFs on the chip.
 **/
void i40e_do_reset(struct i40e_pf *pf, u32 reset_flags)
{
	u32 val;

	WARN_ON(in_interrupt());

5597

J
Jesse Brandeburg 已提交
5598
	/* do the biggest reset indicated */
5599
	if (reset_flags & BIT_ULL(__I40E_GLOBAL_RESET_REQUESTED)) {
J
Jesse Brandeburg 已提交
5600 5601 5602 5603 5604 5605 5606 5607 5608

		/* Request a Global Reset
		 *
		 * This will start the chip's countdown to the actual full
		 * chip reset event, and a warning interrupt to be sent
		 * to all PFs, including the requestor.  Our handler
		 * for the warning interrupt will deal with the shutdown
		 * and recovery of the switch setup.
		 */
J
Jesse Brandeburg 已提交
5609
		dev_dbg(&pf->pdev->dev, "GlobalR requested\n");
J
Jesse Brandeburg 已提交
5610 5611 5612 5613
		val = rd32(&pf->hw, I40E_GLGEN_RTRIG);
		val |= I40E_GLGEN_RTRIG_GLOBR_MASK;
		wr32(&pf->hw, I40E_GLGEN_RTRIG, val);

5614
	} else if (reset_flags & BIT_ULL(__I40E_CORE_RESET_REQUESTED)) {
J
Jesse Brandeburg 已提交
5615 5616 5617 5618 5619

		/* Request a Core Reset
		 *
		 * Same as Global Reset, except does *not* include the MAC/PHY
		 */
J
Jesse Brandeburg 已提交
5620
		dev_dbg(&pf->pdev->dev, "CoreR requested\n");
J
Jesse Brandeburg 已提交
5621 5622 5623 5624 5625
		val = rd32(&pf->hw, I40E_GLGEN_RTRIG);
		val |= I40E_GLGEN_RTRIG_CORER_MASK;
		wr32(&pf->hw, I40E_GLGEN_RTRIG, val);
		i40e_flush(&pf->hw);

5626
	} else if (reset_flags & BIT_ULL(__I40E_PF_RESET_REQUESTED)) {
J
Jesse Brandeburg 已提交
5627 5628 5629 5630 5631 5632 5633 5634 5635

		/* Request a PF Reset
		 *
		 * Resets only the PF-specific registers
		 *
		 * This goes directly to the tear-down and rebuild of
		 * the switch, since we need to do all the recovery as
		 * for the Core Reset.
		 */
J
Jesse Brandeburg 已提交
5636
		dev_dbg(&pf->pdev->dev, "PFR requested\n");
J
Jesse Brandeburg 已提交
5637 5638
		i40e_handle_reset_warning(pf);

5639
	} else if (reset_flags & BIT_ULL(__I40E_REINIT_REQUESTED)) {
J
Jesse Brandeburg 已提交
5640 5641 5642 5643 5644
		int v;

		/* Find the VSI(s) that requested a re-init */
		dev_info(&pf->pdev->dev,
			 "VSI reinit requested\n");
M
Mitch Williams 已提交
5645
		for (v = 0; v < pf->num_alloc_vsi; v++) {
J
Jesse Brandeburg 已提交
5646
			struct i40e_vsi *vsi = pf->vsi[v];
J
Jesse Brandeburg 已提交
5647

J
Jesse Brandeburg 已提交
5648 5649 5650 5651 5652 5653
			if (vsi != NULL &&
			    test_bit(__I40E_REINIT_REQUESTED, &vsi->state)) {
				i40e_vsi_reinit_locked(pf->vsi[v]);
				clear_bit(__I40E_REINIT_REQUESTED, &vsi->state);
			}
		}
5654
	} else if (reset_flags & BIT_ULL(__I40E_DOWN_REQUESTED)) {
5655 5656 5657 5658 5659 5660
		int v;

		/* Find the VSI(s) that needs to be brought down */
		dev_info(&pf->pdev->dev, "VSI down requested\n");
		for (v = 0; v < pf->num_alloc_vsi; v++) {
			struct i40e_vsi *vsi = pf->vsi[v];
J
Jesse Brandeburg 已提交
5661

5662 5663 5664 5665 5666 5667 5668
			if (vsi != NULL &&
			    test_bit(__I40E_DOWN_REQUESTED, &vsi->state)) {
				set_bit(__I40E_DOWN, &vsi->state);
				i40e_down(vsi);
				clear_bit(__I40E_DOWN_REQUESTED, &vsi->state);
			}
		}
J
Jesse Brandeburg 已提交
5669 5670 5671 5672 5673 5674
	} else {
		dev_info(&pf->pdev->dev,
			 "bad reset request 0x%08x\n", reset_flags);
	}
}

N
Neerav Parikh 已提交
5675 5676 5677 5678 5679 5680 5681 5682 5683 5684 5685 5686 5687 5688 5689 5690 5691 5692 5693 5694 5695 5696
#ifdef CONFIG_I40E_DCB
/**
 * i40e_dcb_need_reconfig - Check if DCB needs reconfig
 * @pf: board private structure
 * @old_cfg: current DCB config
 * @new_cfg: new DCB config
 **/
bool i40e_dcb_need_reconfig(struct i40e_pf *pf,
			    struct i40e_dcbx_config *old_cfg,
			    struct i40e_dcbx_config *new_cfg)
{
	bool need_reconfig = false;

	/* Check if ETS configuration has changed */
	if (memcmp(&new_cfg->etscfg,
		   &old_cfg->etscfg,
		   sizeof(new_cfg->etscfg))) {
		/* If Priority Table has changed reconfig is needed */
		if (memcmp(&new_cfg->etscfg.prioritytable,
			   &old_cfg->etscfg.prioritytable,
			   sizeof(new_cfg->etscfg.prioritytable))) {
			need_reconfig = true;
J
Jesse Brandeburg 已提交
5697
			dev_dbg(&pf->pdev->dev, "ETS UP2TC changed.\n");
N
Neerav Parikh 已提交
5698 5699 5700 5701 5702
		}

		if (memcmp(&new_cfg->etscfg.tcbwtable,
			   &old_cfg->etscfg.tcbwtable,
			   sizeof(new_cfg->etscfg.tcbwtable)))
J
Jesse Brandeburg 已提交
5703
			dev_dbg(&pf->pdev->dev, "ETS TC BW Table changed.\n");
N
Neerav Parikh 已提交
5704 5705 5706 5707

		if (memcmp(&new_cfg->etscfg.tsatable,
			   &old_cfg->etscfg.tsatable,
			   sizeof(new_cfg->etscfg.tsatable)))
J
Jesse Brandeburg 已提交
5708
			dev_dbg(&pf->pdev->dev, "ETS TSA Table changed.\n");
N
Neerav Parikh 已提交
5709 5710 5711 5712 5713 5714 5715
	}

	/* Check if PFC configuration has changed */
	if (memcmp(&new_cfg->pfc,
		   &old_cfg->pfc,
		   sizeof(new_cfg->pfc))) {
		need_reconfig = true;
J
Jesse Brandeburg 已提交
5716
		dev_dbg(&pf->pdev->dev, "PFC config change detected.\n");
N
Neerav Parikh 已提交
5717 5718 5719 5720 5721
	}

	/* Check if APP Table has changed */
	if (memcmp(&new_cfg->app,
		   &old_cfg->app,
5722
		   sizeof(new_cfg->app))) {
N
Neerav Parikh 已提交
5723
		need_reconfig = true;
J
Jesse Brandeburg 已提交
5724
		dev_dbg(&pf->pdev->dev, "APP Table change detected.\n");
5725
	}
N
Neerav Parikh 已提交
5726

S
Shannon Nelson 已提交
5727
	dev_dbg(&pf->pdev->dev, "dcb need_reconfig=%d\n", need_reconfig);
N
Neerav Parikh 已提交
5728 5729 5730 5731 5732 5733 5734 5735 5736 5737 5738 5739 5740 5741 5742 5743 5744 5745 5746
	return need_reconfig;
}

/**
 * i40e_handle_lldp_event - Handle LLDP Change MIB event
 * @pf: board private structure
 * @e: event info posted on ARQ
 **/
static int i40e_handle_lldp_event(struct i40e_pf *pf,
				  struct i40e_arq_event_info *e)
{
	struct i40e_aqc_lldp_get_mib *mib =
		(struct i40e_aqc_lldp_get_mib *)&e->desc.params.raw;
	struct i40e_hw *hw = &pf->hw;
	struct i40e_dcbx_config tmp_dcbx_cfg;
	bool need_reconfig = false;
	int ret = 0;
	u8 type;

5747
	/* Not DCB capable or capability disabled */
5748
	if (!(pf->flags & I40E_FLAG_DCB_CAPABLE))
5749 5750
		return ret;

N
Neerav Parikh 已提交
5751 5752 5753
	/* Ignore if event is not for Nearest Bridge */
	type = ((mib->type >> I40E_AQ_LLDP_BRIDGE_TYPE_SHIFT)
		& I40E_AQ_LLDP_BRIDGE_TYPE_MASK);
S
Shannon Nelson 已提交
5754
	dev_dbg(&pf->pdev->dev, "LLDP event mib bridge type 0x%x\n", type);
N
Neerav Parikh 已提交
5755 5756 5757 5758 5759
	if (type != I40E_AQ_LLDP_BRIDGE_TYPE_NEAREST_BRIDGE)
		return ret;

	/* Check MIB Type and return if event for Remote MIB update */
	type = mib->type & I40E_AQ_LLDP_MIB_TYPE_MASK;
5760
	dev_dbg(&pf->pdev->dev,
S
Shannon Nelson 已提交
5761
		"LLDP event mib type %s\n", type ? "remote" : "local");
N
Neerav Parikh 已提交
5762 5763 5764 5765 5766 5767 5768 5769
	if (type == I40E_AQ_LLDP_MIB_REMOTE) {
		/* Update the remote cached instance and return */
		ret = i40e_aq_get_dcb_config(hw, I40E_AQ_LLDP_MIB_REMOTE,
				I40E_AQ_LLDP_BRIDGE_TYPE_NEAREST_BRIDGE,
				&hw->remote_dcbx_config);
		goto exit;
	}

5770
	/* Store the old configuration */
5771
	tmp_dcbx_cfg = hw->local_dcbx_config;
5772

5773 5774
	/* Reset the old DCBx configuration data */
	memset(&hw->local_dcbx_config, 0, sizeof(hw->local_dcbx_config));
5775 5776
	/* Get updated DCBX data from firmware */
	ret = i40e_get_dcb_config(&pf->hw);
N
Neerav Parikh 已提交
5777
	if (ret) {
5778 5779 5780 5781
		dev_info(&pf->pdev->dev,
			 "Failed querying DCB configuration data from firmware, err %s aq_err %s\n",
			 i40e_stat_str(&pf->hw, ret),
			 i40e_aq_str(&pf->hw, pf->hw.aq.asq_last_status));
N
Neerav Parikh 已提交
5782 5783 5784 5785
		goto exit;
	}

	/* No change detected in DCBX configs */
5786 5787
	if (!memcmp(&tmp_dcbx_cfg, &hw->local_dcbx_config,
		    sizeof(tmp_dcbx_cfg))) {
J
Jesse Brandeburg 已提交
5788
		dev_dbg(&pf->pdev->dev, "No change detected in DCBX configuration.\n");
N
Neerav Parikh 已提交
5789 5790 5791
		goto exit;
	}

5792 5793
	need_reconfig = i40e_dcb_need_reconfig(pf, &tmp_dcbx_cfg,
					       &hw->local_dcbx_config);
N
Neerav Parikh 已提交
5794

5795
	i40e_dcbnl_flush_apps(pf, &tmp_dcbx_cfg, &hw->local_dcbx_config);
N
Neerav Parikh 已提交
5796 5797 5798 5799

	if (!need_reconfig)
		goto exit;

5800
	/* Enable DCB tagging only when more than one TC */
5801
	if (i40e_dcb_get_num_tc(&hw->local_dcbx_config) > 1)
5802 5803 5804 5805
		pf->flags |= I40E_FLAG_DCB_ENABLED;
	else
		pf->flags &= ~I40E_FLAG_DCB_ENABLED;

5806
	set_bit(__I40E_PORT_TX_SUSPENDED, &pf->state);
N
Neerav Parikh 已提交
5807 5808 5809 5810 5811 5812
	/* Reconfiguration needed quiesce all VSIs */
	i40e_pf_quiesce_all_vsi(pf);

	/* Changes in configuration update VEB/VSI */
	i40e_dcb_reconfigure(pf);

5813 5814
	ret = i40e_resume_port_tx(pf);

5815
	clear_bit(__I40E_PORT_TX_SUSPENDED, &pf->state);
5816
	/* In case of error no point in resuming VSIs */
5817 5818 5819
	if (ret)
		goto exit;

5820 5821
	/* Wait for the PF's queues to be disabled */
	ret = i40e_pf_wait_queues_disabled(pf);
5822 5823 5824 5825 5826
	if (ret) {
		/* Schedule PF reset to recover */
		set_bit(__I40E_PF_RESET_REQUESTED, &pf->state);
		i40e_service_event_schedule(pf);
	} else {
5827
		i40e_pf_unquiesce_all_vsi(pf);
5828 5829
		/* Notify the client for the DCB changes */
		i40e_notify_client_of_l2_param_changes(pf->vsi[pf->lan_vsi]);
5830 5831
	}

N
Neerav Parikh 已提交
5832 5833 5834 5835 5836
exit:
	return ret;
}
#endif /* CONFIG_I40E_DCB */

5837 5838 5839 5840 5841 5842 5843 5844 5845 5846 5847 5848 5849
/**
 * i40e_do_reset_safe - Protected reset path for userland calls.
 * @pf: board private structure
 * @reset_flags: which reset is requested
 *
 **/
void i40e_do_reset_safe(struct i40e_pf *pf, u32 reset_flags)
{
	rtnl_lock();
	i40e_do_reset(pf, reset_flags);
	rtnl_unlock();
}

J
Jesse Brandeburg 已提交
5850 5851 5852 5853 5854 5855 5856 5857 5858 5859 5860 5861 5862 5863 5864 5865 5866 5867 5868
/**
 * i40e_handle_lan_overflow_event - Handler for LAN queue overflow event
 * @pf: board private structure
 * @e: event info posted on ARQ
 *
 * Handler for LAN Queue Overflow Event generated by the firmware for PF
 * and VF queues
 **/
static void i40e_handle_lan_overflow_event(struct i40e_pf *pf,
					   struct i40e_arq_event_info *e)
{
	struct i40e_aqc_lan_overflow *data =
		(struct i40e_aqc_lan_overflow *)&e->desc.params.raw;
	u32 queue = le32_to_cpu(data->prtdcb_rupto);
	u32 qtx_ctl = le32_to_cpu(data->otx_ctl);
	struct i40e_hw *hw = &pf->hw;
	struct i40e_vf *vf;
	u16 vf_id;

J
Jesse Brandeburg 已提交
5869 5870
	dev_dbg(&pf->pdev->dev, "overflow Rx Queue Number = %d QTX_CTL=0x%08x\n",
		queue, qtx_ctl);
J
Jesse Brandeburg 已提交
5871 5872 5873 5874 5875 5876 5877 5878 5879 5880 5881 5882 5883 5884 5885 5886 5887 5888 5889 5890 5891

	/* Queue belongs to VF, find the VF and issue VF reset */
	if (((qtx_ctl & I40E_QTX_CTL_PFVF_Q_MASK)
	    >> I40E_QTX_CTL_PFVF_Q_SHIFT) == I40E_QTX_CTL_VF_QUEUE) {
		vf_id = (u16)((qtx_ctl & I40E_QTX_CTL_VFVM_INDX_MASK)
			 >> I40E_QTX_CTL_VFVM_INDX_SHIFT);
		vf_id -= hw->func_caps.vf_base_id;
		vf = &pf->vf[vf_id];
		i40e_vc_notify_vf_reset(vf);
		/* Allow VF to process pending reset notification */
		msleep(20);
		i40e_reset_vf(vf, false);
	}
}

/**
 * i40e_service_event_complete - Finish up the service event
 * @pf: board private structure
 **/
static void i40e_service_event_complete(struct i40e_pf *pf)
{
5892
	WARN_ON(!test_bit(__I40E_SERVICE_SCHED, &pf->state));
J
Jesse Brandeburg 已提交
5893 5894

	/* flush memory to make sure state is correct before next watchog */
5895
	smp_mb__before_atomic();
J
Jesse Brandeburg 已提交
5896 5897 5898
	clear_bit(__I40E_SERVICE_SCHED, &pf->state);
}

5899
/**
5900 5901 5902
 * i40e_get_cur_guaranteed_fd_count - Get the consumed guaranteed FD filters
 * @pf: board private structure
 **/
5903
u32 i40e_get_cur_guaranteed_fd_count(struct i40e_pf *pf)
5904
{
5905
	u32 val, fcnt_prog;
5906 5907 5908 5909 5910 5911 5912

	val = rd32(&pf->hw, I40E_PFQF_FDSTAT);
	fcnt_prog = (val & I40E_PFQF_FDSTAT_GUARANT_CNT_MASK);
	return fcnt_prog;
}

/**
5913
 * i40e_get_current_fd_count - Get total FD filters programmed for this PF
5914 5915
 * @pf: board private structure
 **/
5916
u32 i40e_get_current_fd_count(struct i40e_pf *pf)
5917
{
5918 5919
	u32 val, fcnt_prog;

5920 5921 5922 5923 5924 5925
	val = rd32(&pf->hw, I40E_PFQF_FDSTAT);
	fcnt_prog = (val & I40E_PFQF_FDSTAT_GUARANT_CNT_MASK) +
		    ((val & I40E_PFQF_FDSTAT_BEST_CNT_MASK) >>
		      I40E_PFQF_FDSTAT_BEST_CNT_SHIFT);
	return fcnt_prog;
}
5926

5927 5928 5929 5930 5931 5932 5933 5934 5935 5936 5937 5938 5939 5940 5941
/**
 * i40e_get_global_fd_count - Get total FD filters programmed on device
 * @pf: board private structure
 **/
u32 i40e_get_global_fd_count(struct i40e_pf *pf)
{
	u32 val, fcnt_prog;

	val = rd32(&pf->hw, I40E_GLQF_FDCNT_0);
	fcnt_prog = (val & I40E_GLQF_FDCNT_0_GUARANT_CNT_MASK) +
		    ((val & I40E_GLQF_FDCNT_0_BESTCNT_MASK) >>
		     I40E_GLQF_FDCNT_0_BESTCNT_SHIFT);
	return fcnt_prog;
}

5942 5943 5944 5945 5946 5947
/**
 * i40e_fdir_check_and_reenable - Function to reenabe FD ATR or SB if disabled
 * @pf: board private structure
 **/
void i40e_fdir_check_and_reenable(struct i40e_pf *pf)
{
5948
	struct i40e_fdir_filter *filter;
5949
	u32 fcnt_prog, fcnt_avail;
5950
	struct hlist_node *node;
5951

5952 5953 5954
	if (test_bit(__I40E_FD_FLUSH_REQUESTED, &pf->state))
		return;

5955 5956 5957
	/* Check if, FD SB or ATR was auto disabled and if there is enough room
	 * to re-enable
	 */
5958
	fcnt_prog = i40e_get_global_fd_count(pf);
5959
	fcnt_avail = pf->fdir_pf_filter_count;
5960 5961 5962
	if ((fcnt_prog < (fcnt_avail - I40E_FDIR_BUFFER_HEAD_ROOM)) ||
	    (pf->fd_add_err == 0) ||
	    (i40e_get_current_atr_cnt(pf) < pf->fd_atr_cnt)) {
5963 5964 5965
		if ((pf->flags & I40E_FLAG_FD_SB_ENABLED) &&
		    (pf->auto_disable_flags & I40E_FLAG_FD_SB_ENABLED)) {
			pf->auto_disable_flags &= ~I40E_FLAG_FD_SB_ENABLED;
5966 5967
			if (I40E_DEBUG_FD & pf->hw.debug_mask)
				dev_info(&pf->pdev->dev, "FD Sideband/ntuple is being enabled since we have space in the table now\n");
5968 5969
		}
	}
5970 5971 5972 5973

	/* Wait for some more space to be available to turn on ATR. We also
	 * must check that no existing ntuple rules for TCP are in effect
	 */
5974 5975
	if (fcnt_prog < (fcnt_avail - I40E_FDIR_BUFFER_HEAD_ROOM * 2)) {
		if ((pf->flags & I40E_FLAG_FD_ATR_ENABLED) &&
5976 5977
		    (pf->auto_disable_flags & I40E_FLAG_FD_ATR_ENABLED) &&
		    (pf->fd_tcp_rule == 0)) {
5978
			pf->auto_disable_flags &= ~I40E_FLAG_FD_ATR_ENABLED;
5979
			if (I40E_DEBUG_FD & pf->hw.debug_mask)
5980
				dev_info(&pf->pdev->dev, "ATR is being enabled since we have space in the table and there are no conflicting ntuple rules\n");
5981 5982
		}
	}
5983 5984 5985 5986 5987 5988 5989 5990 5991 5992 5993 5994

	/* if hw had a problem adding a filter, delete it */
	if (pf->fd_inv > 0) {
		hlist_for_each_entry_safe(filter, node,
					  &pf->fdir_filter_list, fdir_node) {
			if (filter->fd_id == pf->fd_inv) {
				hlist_del(&filter->fdir_node);
				kfree(filter);
				pf->fdir_pf_active_filters--;
			}
		}
	}
5995 5996
}

5997
#define I40E_MIN_FD_FLUSH_INTERVAL 10
5998
#define I40E_MIN_FD_FLUSH_SB_ATR_UNSTABLE 30
5999 6000 6001 6002 6003 6004
/**
 * i40e_fdir_flush_and_replay - Function to flush all FD filters and replay SB
 * @pf: board private structure
 **/
static void i40e_fdir_flush_and_replay(struct i40e_pf *pf)
{
6005
	unsigned long min_flush_time;
6006
	int flush_wait_retry = 50;
6007 6008
	bool disable_atr = false;
	int fd_room;
6009 6010
	int reg;

6011 6012 6013
	if (!time_after(jiffies, pf->fd_flush_timestamp +
				 (I40E_MIN_FD_FLUSH_INTERVAL * HZ)))
		return;
6014

6015 6016 6017 6018 6019 6020 6021 6022 6023 6024 6025 6026 6027 6028 6029
	/* If the flush is happening too quick and we have mostly SB rules we
	 * should not re-enable ATR for some time.
	 */
	min_flush_time = pf->fd_flush_timestamp +
			 (I40E_MIN_FD_FLUSH_SB_ATR_UNSTABLE * HZ);
	fd_room = pf->fdir_pf_filter_count - pf->fdir_pf_active_filters;

	if (!(time_after(jiffies, min_flush_time)) &&
	    (fd_room < I40E_FDIR_BUFFER_HEAD_ROOM_FOR_ATR)) {
		if (I40E_DEBUG_FD & pf->hw.debug_mask)
			dev_info(&pf->pdev->dev, "ATR disabled, not enough FD filter space.\n");
		disable_atr = true;
	}

	pf->fd_flush_timestamp = jiffies;
6030
	pf->auto_disable_flags |= I40E_FLAG_FD_ATR_ENABLED;
6031 6032 6033 6034 6035 6036 6037 6038 6039 6040 6041 6042 6043 6044 6045 6046 6047 6048 6049
	/* flush all filters */
	wr32(&pf->hw, I40E_PFQF_CTL_1,
	     I40E_PFQF_CTL_1_CLEARFDTABLE_MASK);
	i40e_flush(&pf->hw);
	pf->fd_flush_cnt++;
	pf->fd_add_err = 0;
	do {
		/* Check FD flush status every 5-6msec */
		usleep_range(5000, 6000);
		reg = rd32(&pf->hw, I40E_PFQF_CTL_1);
		if (!(reg & I40E_PFQF_CTL_1_CLEARFDTABLE_MASK))
			break;
	} while (flush_wait_retry--);
	if (reg & I40E_PFQF_CTL_1_CLEARFDTABLE_MASK) {
		dev_warn(&pf->pdev->dev, "FD table did not flush, needs more time\n");
	} else {
		/* replay sideband filters */
		i40e_fdir_filter_restore(pf->vsi[pf->lan_vsi]);
		if (!disable_atr)
6050
			pf->auto_disable_flags &= ~I40E_FLAG_FD_ATR_ENABLED;
6051 6052 6053
		clear_bit(__I40E_FD_FLUSH_REQUESTED, &pf->state);
		if (I40E_DEBUG_FD & pf->hw.debug_mask)
			dev_info(&pf->pdev->dev, "FD Filter table flushed and FD-SB replayed.\n");
6054 6055 6056 6057 6058 6059 6060
	}
}

/**
 * i40e_get_current_atr_count - Get the count of total FD ATR filters programmed
 * @pf: board private structure
 **/
6061
u32 i40e_get_current_atr_cnt(struct i40e_pf *pf)
6062 6063 6064 6065 6066 6067 6068 6069 6070 6071 6072
{
	return i40e_get_current_fd_count(pf) - pf->fdir_pf_active_filters;
}

/* We can see up to 256 filter programming desc in transit if the filters are
 * being applied really fast; before we see the first
 * filter miss error on Rx queue 0. Accumulating enough error messages before
 * reacting will make sure we don't cause flush too often.
 */
#define I40E_MAX_FD_PROGRAM_ERROR 256

J
Jesse Brandeburg 已提交
6073 6074 6075 6076 6077 6078 6079 6080 6081 6082
/**
 * i40e_fdir_reinit_subtask - Worker thread to reinit FDIR filter table
 * @pf: board private structure
 **/
static void i40e_fdir_reinit_subtask(struct i40e_pf *pf)
{

	/* if interface is down do nothing */
	if (test_bit(__I40E_DOWN, &pf->state))
		return;
6083

6084
	if (test_bit(__I40E_FD_FLUSH_REQUESTED, &pf->state))
6085 6086
		i40e_fdir_flush_and_replay(pf);

6087 6088
	i40e_fdir_check_and_reenable(pf);

J
Jesse Brandeburg 已提交
6089 6090 6091 6092 6093 6094 6095 6096 6097
}

/**
 * i40e_vsi_link_event - notify VSI of a link event
 * @vsi: vsi to be notified
 * @link_up: link up or down
 **/
static void i40e_vsi_link_event(struct i40e_vsi *vsi, bool link_up)
{
6098
	if (!vsi || test_bit(__I40E_DOWN, &vsi->state))
J
Jesse Brandeburg 已提交
6099 6100 6101 6102
		return;

	switch (vsi->type) {
	case I40E_VSI_MAIN:
6103 6104 6105
#ifdef I40E_FCOE
	case I40E_VSI_FCOE:
#endif
J
Jesse Brandeburg 已提交
6106 6107 6108 6109 6110 6111 6112 6113 6114 6115 6116 6117 6118 6119 6120
		if (!vsi->netdev || !vsi->netdev_registered)
			break;

		if (link_up) {
			netif_carrier_on(vsi->netdev);
			netif_tx_wake_all_queues(vsi->netdev);
		} else {
			netif_carrier_off(vsi->netdev);
			netif_tx_stop_all_queues(vsi->netdev);
		}
		break;

	case I40E_VSI_SRIOV:
	case I40E_VSI_VMDQ2:
	case I40E_VSI_CTRL:
6121
	case I40E_VSI_IWARP:
J
Jesse Brandeburg 已提交
6122 6123 6124 6125 6126 6127 6128 6129 6130 6131 6132 6133 6134 6135 6136 6137 6138 6139 6140 6141 6142 6143 6144 6145 6146 6147 6148
	case I40E_VSI_MIRROR:
	default:
		/* there is no notification for other VSIs */
		break;
	}
}

/**
 * i40e_veb_link_event - notify elements on the veb of a link event
 * @veb: veb to be notified
 * @link_up: link up or down
 **/
static void i40e_veb_link_event(struct i40e_veb *veb, bool link_up)
{
	struct i40e_pf *pf;
	int i;

	if (!veb || !veb->pf)
		return;
	pf = veb->pf;

	/* depth first... */
	for (i = 0; i < I40E_MAX_VEB; i++)
		if (pf->veb[i] && (pf->veb[i]->uplink_seid == veb->seid))
			i40e_veb_link_event(pf->veb[i], link_up);

	/* ... now the local VSIs */
M
Mitch Williams 已提交
6149
	for (i = 0; i < pf->num_alloc_vsi; i++)
J
Jesse Brandeburg 已提交
6150 6151 6152 6153 6154 6155 6156 6157 6158 6159
		if (pf->vsi[i] && (pf->vsi[i]->uplink_seid == veb->seid))
			i40e_vsi_link_event(pf->vsi[i], link_up);
}

/**
 * i40e_link_event - Update netif_carrier status
 * @pf: board private structure
 **/
static void i40e_link_event(struct i40e_pf *pf)
{
M
Mitch Williams 已提交
6160
	struct i40e_vsi *vsi = pf->vsi[pf->lan_vsi];
6161
	u8 new_link_speed, old_link_speed;
6162 6163
	i40e_status status;
	bool new_link, old_link;
J
Jesse Brandeburg 已提交
6164

6165 6166 6167
	/* save off old link status information */
	pf->hw.phy.link_info_old = pf->hw.phy.link_info;

6168 6169 6170
	/* set this to force the get_link_status call to refresh state */
	pf->hw.phy.get_link_info = true;

J
Jesse Brandeburg 已提交
6171
	old_link = (pf->hw.phy.link_info_old.link_info & I40E_AQ_LINK_UP);
6172 6173 6174 6175 6176 6177 6178 6179

	status = i40e_get_link_status(&pf->hw, &new_link);
	if (status) {
		dev_dbg(&pf->pdev->dev, "couldn't get link state, status: %d\n",
			status);
		return;
	}

6180 6181
	old_link_speed = pf->hw.phy.link_info_old.link_speed;
	new_link_speed = pf->hw.phy.link_info.link_speed;
J
Jesse Brandeburg 已提交
6182

6183
	if (new_link == old_link &&
6184
	    new_link_speed == old_link_speed &&
M
Mitch Williams 已提交
6185 6186
	    (test_bit(__I40E_DOWN, &vsi->state) ||
	     new_link == netif_carrier_ok(vsi->netdev)))
J
Jesse Brandeburg 已提交
6187
		return;
M
Mitch Williams 已提交
6188 6189 6190

	if (!test_bit(__I40E_DOWN, &vsi->state))
		i40e_print_link_message(vsi, new_link);
J
Jesse Brandeburg 已提交
6191 6192 6193 6194 6195 6196 6197

	/* Notify the base of the switch tree connected to
	 * the link.  Floating VEBs are not notified.
	 */
	if (pf->lan_veb != I40E_NO_VEB && pf->veb[pf->lan_veb])
		i40e_veb_link_event(pf->veb[pf->lan_veb], new_link);
	else
M
Mitch Williams 已提交
6198
		i40e_vsi_link_event(vsi, new_link);
J
Jesse Brandeburg 已提交
6199 6200 6201

	if (pf->vf)
		i40e_vc_notify_link_state(pf);
J
Jacob Keller 已提交
6202 6203 6204

	if (pf->flags & I40E_FLAG_PTP)
		i40e_ptp_set_increment(pf);
J
Jesse Brandeburg 已提交
6205 6206 6207
}

/**
6208
 * i40e_watchdog_subtask - periodic checks not using event driven response
J
Jesse Brandeburg 已提交
6209 6210 6211 6212 6213 6214 6215 6216 6217 6218 6219
 * @pf: board private structure
 **/
static void i40e_watchdog_subtask(struct i40e_pf *pf)
{
	int i;

	/* if interface is down do nothing */
	if (test_bit(__I40E_DOWN, &pf->state) ||
	    test_bit(__I40E_CONFIG_BUSY, &pf->state))
		return;

6220 6221 6222 6223 6224 6225
	/* make sure we don't do these things too often */
	if (time_before(jiffies, (pf->service_timer_previous +
				  pf->service_timer_period)))
		return;
	pf->service_timer_previous = jiffies;

S
Shannon Nelson 已提交
6226 6227
	if (pf->flags & I40E_FLAG_LINK_POLLING_ENABLED)
		i40e_link_event(pf);
6228

J
Jesse Brandeburg 已提交
6229 6230 6231
	/* Update the stats for active netdevs so the network stack
	 * can look at updated numbers whenever it cares to
	 */
M
Mitch Williams 已提交
6232
	for (i = 0; i < pf->num_alloc_vsi; i++)
J
Jesse Brandeburg 已提交
6233 6234 6235
		if (pf->vsi[i] && pf->vsi[i]->netdev)
			i40e_update_stats(pf->vsi[i]);

6236 6237 6238 6239 6240 6241
	if (pf->flags & I40E_FLAG_VEB_STATS_ENABLED) {
		/* Update the stats for the active switching components */
		for (i = 0; i < I40E_MAX_VEB; i++)
			if (pf->veb[i])
				i40e_update_veb_stats(pf->veb[i]);
	}
J
Jacob Keller 已提交
6242 6243

	i40e_ptp_rx_hang(pf->vsi[pf->lan_vsi]);
J
Jesse Brandeburg 已提交
6244 6245 6246 6247 6248 6249 6250 6251 6252 6253
}

/**
 * i40e_reset_subtask - Set up for resetting the device and driver
 * @pf: board private structure
 **/
static void i40e_reset_subtask(struct i40e_pf *pf)
{
	u32 reset_flags = 0;

6254
	rtnl_lock();
J
Jesse Brandeburg 已提交
6255
	if (test_bit(__I40E_REINIT_REQUESTED, &pf->state)) {
J
Jesse Brandeburg 已提交
6256
		reset_flags |= BIT(__I40E_REINIT_REQUESTED);
J
Jesse Brandeburg 已提交
6257 6258 6259
		clear_bit(__I40E_REINIT_REQUESTED, &pf->state);
	}
	if (test_bit(__I40E_PF_RESET_REQUESTED, &pf->state)) {
J
Jesse Brandeburg 已提交
6260
		reset_flags |= BIT(__I40E_PF_RESET_REQUESTED);
J
Jesse Brandeburg 已提交
6261 6262 6263
		clear_bit(__I40E_PF_RESET_REQUESTED, &pf->state);
	}
	if (test_bit(__I40E_CORE_RESET_REQUESTED, &pf->state)) {
J
Jesse Brandeburg 已提交
6264
		reset_flags |= BIT(__I40E_CORE_RESET_REQUESTED);
J
Jesse Brandeburg 已提交
6265 6266 6267
		clear_bit(__I40E_CORE_RESET_REQUESTED, &pf->state);
	}
	if (test_bit(__I40E_GLOBAL_RESET_REQUESTED, &pf->state)) {
J
Jesse Brandeburg 已提交
6268
		reset_flags |= BIT(__I40E_GLOBAL_RESET_REQUESTED);
J
Jesse Brandeburg 已提交
6269 6270
		clear_bit(__I40E_GLOBAL_RESET_REQUESTED, &pf->state);
	}
6271
	if (test_bit(__I40E_DOWN_REQUESTED, &pf->state)) {
J
Jesse Brandeburg 已提交
6272
		reset_flags |= BIT(__I40E_DOWN_REQUESTED);
6273 6274
		clear_bit(__I40E_DOWN_REQUESTED, &pf->state);
	}
J
Jesse Brandeburg 已提交
6275 6276 6277 6278 6279 6280

	/* If there's a recovery already waiting, it takes
	 * precedence before starting a new reset sequence.
	 */
	if (test_bit(__I40E_RESET_INTR_RECEIVED, &pf->state)) {
		i40e_handle_reset_warning(pf);
6281
		goto unlock;
J
Jesse Brandeburg 已提交
6282 6283 6284 6285 6286 6287 6288
	}

	/* If we're already down or resetting, just bail */
	if (reset_flags &&
	    !test_bit(__I40E_DOWN, &pf->state) &&
	    !test_bit(__I40E_CONFIG_BUSY, &pf->state))
		i40e_do_reset(pf, reset_flags);
6289 6290 6291

unlock:
	rtnl_unlock();
J
Jesse Brandeburg 已提交
6292 6293 6294 6295 6296 6297 6298 6299 6300 6301 6302 6303 6304
}

/**
 * i40e_handle_link_event - Handle link event
 * @pf: board private structure
 * @e: event info posted on ARQ
 **/
static void i40e_handle_link_event(struct i40e_pf *pf,
				   struct i40e_arq_event_info *e)
{
	struct i40e_aqc_get_link_status *status =
		(struct i40e_aqc_get_link_status *)&e->desc.params.raw;

6305 6306 6307 6308 6309 6310 6311 6312
	/* Do a new status request to re-enable LSE reporting
	 * and load new status information into the hw struct
	 * This completely ignores any state information
	 * in the ARQ event info, instead choosing to always
	 * issue the AQ update link status command.
	 */
	i40e_link_event(pf);

6313 6314 6315 6316 6317 6318
	/* check for unqualified module, if link is down */
	if ((status->link_info & I40E_AQ_MEDIA_AVAILABLE) &&
	    (!(status->an_info & I40E_AQ_QUALIFIED_MODULE)) &&
	    (!(status->link_info & I40E_AQ_LINK_UP)))
		dev_err(&pf->pdev->dev,
			"The driver failed to link because an unqualified module was detected.\n");
J
Jesse Brandeburg 已提交
6319 6320 6321 6322 6323 6324 6325 6326 6327 6328 6329 6330 6331
}

/**
 * i40e_clean_adminq_subtask - Clean the AdminQ rings
 * @pf: board private structure
 **/
static void i40e_clean_adminq_subtask(struct i40e_pf *pf)
{
	struct i40e_arq_event_info event;
	struct i40e_hw *hw = &pf->hw;
	u16 pending, i = 0;
	i40e_status ret;
	u16 opcode;
6332
	u32 oldval;
J
Jesse Brandeburg 已提交
6333 6334
	u32 val;

6335 6336 6337 6338
	/* Do not run clean AQ when PF reset fails */
	if (test_bit(__I40E_RESET_FAILED, &pf->state))
		return;

6339 6340 6341 6342
	/* check for error indications */
	val = rd32(&pf->hw, pf->hw.aq.arq.len);
	oldval = val;
	if (val & I40E_PF_ARQLEN_ARQVFE_MASK) {
M
Mitch Williams 已提交
6343 6344
		if (hw->debug_mask & I40E_DEBUG_AQ)
			dev_info(&pf->pdev->dev, "ARQ VF Error detected\n");
6345 6346 6347
		val &= ~I40E_PF_ARQLEN_ARQVFE_MASK;
	}
	if (val & I40E_PF_ARQLEN_ARQOVFL_MASK) {
M
Mitch Williams 已提交
6348 6349
		if (hw->debug_mask & I40E_DEBUG_AQ)
			dev_info(&pf->pdev->dev, "ARQ Overflow Error detected\n");
6350
		val &= ~I40E_PF_ARQLEN_ARQOVFL_MASK;
6351
		pf->arq_overflows++;
6352 6353
	}
	if (val & I40E_PF_ARQLEN_ARQCRIT_MASK) {
M
Mitch Williams 已提交
6354 6355
		if (hw->debug_mask & I40E_DEBUG_AQ)
			dev_info(&pf->pdev->dev, "ARQ Critical Error detected\n");
6356 6357 6358 6359 6360 6361 6362 6363
		val &= ~I40E_PF_ARQLEN_ARQCRIT_MASK;
	}
	if (oldval != val)
		wr32(&pf->hw, pf->hw.aq.arq.len, val);

	val = rd32(&pf->hw, pf->hw.aq.asq.len);
	oldval = val;
	if (val & I40E_PF_ATQLEN_ATQVFE_MASK) {
M
Mitch Williams 已提交
6364 6365
		if (pf->hw.debug_mask & I40E_DEBUG_AQ)
			dev_info(&pf->pdev->dev, "ASQ VF Error detected\n");
6366 6367 6368
		val &= ~I40E_PF_ATQLEN_ATQVFE_MASK;
	}
	if (val & I40E_PF_ATQLEN_ATQOVFL_MASK) {
M
Mitch Williams 已提交
6369 6370
		if (pf->hw.debug_mask & I40E_DEBUG_AQ)
			dev_info(&pf->pdev->dev, "ASQ Overflow Error detected\n");
6371 6372 6373
		val &= ~I40E_PF_ATQLEN_ATQOVFL_MASK;
	}
	if (val & I40E_PF_ATQLEN_ATQCRIT_MASK) {
M
Mitch Williams 已提交
6374 6375
		if (pf->hw.debug_mask & I40E_DEBUG_AQ)
			dev_info(&pf->pdev->dev, "ASQ Critical Error detected\n");
6376 6377 6378 6379 6380
		val &= ~I40E_PF_ATQLEN_ATQCRIT_MASK;
	}
	if (oldval != val)
		wr32(&pf->hw, pf->hw.aq.asq.len, val);

M
Mitch Williams 已提交
6381 6382
	event.buf_len = I40E_MAX_AQ_BUF_SIZE;
	event.msg_buf = kzalloc(event.buf_len, GFP_KERNEL);
J
Jesse Brandeburg 已提交
6383 6384 6385 6386 6387
	if (!event.msg_buf)
		return;

	do {
		ret = i40e_clean_arq_element(hw, &event, &pending);
M
Mitch Williams 已提交
6388
		if (ret == I40E_ERR_ADMIN_QUEUE_NO_WORK)
J
Jesse Brandeburg 已提交
6389
			break;
M
Mitch Williams 已提交
6390
		else if (ret) {
J
Jesse Brandeburg 已提交
6391 6392 6393 6394 6395 6396 6397 6398 6399 6400 6401 6402 6403 6404 6405 6406
			dev_info(&pf->pdev->dev, "ARQ event error %d\n", ret);
			break;
		}

		opcode = le16_to_cpu(event.desc.opcode);
		switch (opcode) {

		case i40e_aqc_opc_get_link_status:
			i40e_handle_link_event(pf, &event);
			break;
		case i40e_aqc_opc_send_msg_to_pf:
			ret = i40e_vc_process_vf_msg(pf,
					le16_to_cpu(event.desc.retval),
					le32_to_cpu(event.desc.cookie_high),
					le32_to_cpu(event.desc.cookie_low),
					event.msg_buf,
M
Mitch Williams 已提交
6407
					event.msg_len);
J
Jesse Brandeburg 已提交
6408 6409
			break;
		case i40e_aqc_opc_lldp_update_mib:
J
Jesse Brandeburg 已提交
6410
			dev_dbg(&pf->pdev->dev, "ARQ: Update LLDP MIB event received\n");
N
Neerav Parikh 已提交
6411 6412 6413 6414 6415
#ifdef CONFIG_I40E_DCB
			rtnl_lock();
			ret = i40e_handle_lldp_event(pf, &event);
			rtnl_unlock();
#endif /* CONFIG_I40E_DCB */
J
Jesse Brandeburg 已提交
6416 6417
			break;
		case i40e_aqc_opc_event_lan_overflow:
J
Jesse Brandeburg 已提交
6418
			dev_dbg(&pf->pdev->dev, "ARQ LAN queue overflow event received\n");
J
Jesse Brandeburg 已提交
6419 6420
			i40e_handle_lan_overflow_event(pf, &event);
			break;
6421 6422 6423
		case i40e_aqc_opc_send_msg_to_peer:
			dev_info(&pf->pdev->dev, "ARQ: Msg from other pf\n");
			break;
6424 6425
		case i40e_aqc_opc_nvm_erase:
		case i40e_aqc_opc_nvm_update:
6426
		case i40e_aqc_opc_oem_post_update:
6427 6428 6429
			i40e_debug(&pf->hw, I40E_DEBUG_NVM,
				   "ARQ NVM operation 0x%04x completed\n",
				   opcode);
6430
			break;
J
Jesse Brandeburg 已提交
6431 6432
		default:
			dev_info(&pf->pdev->dev,
6433
				 "ARQ: Unknown event 0x%04x ignored\n",
6434
				 opcode);
J
Jesse Brandeburg 已提交
6435 6436 6437 6438 6439 6440 6441 6442 6443 6444 6445 6446 6447 6448
			break;
		}
	} while (pending && (i++ < pf->adminq_work_limit));

	clear_bit(__I40E_ADMINQ_EVENT_PENDING, &pf->state);
	/* re-enable Admin queue interrupt cause */
	val = rd32(hw, I40E_PFINT_ICR0_ENA);
	val |=  I40E_PFINT_ICR0_ENA_ADMINQ_MASK;
	wr32(hw, I40E_PFINT_ICR0_ENA, val);
	i40e_flush(hw);

	kfree(event.msg_buf);
}

6449 6450 6451 6452 6453 6454 6455 6456 6457 6458 6459 6460 6461 6462 6463 6464 6465 6466 6467 6468 6469 6470 6471 6472 6473
/**
 * i40e_verify_eeprom - make sure eeprom is good to use
 * @pf: board private structure
 **/
static void i40e_verify_eeprom(struct i40e_pf *pf)
{
	int err;

	err = i40e_diag_eeprom_test(&pf->hw);
	if (err) {
		/* retry in case of garbage read */
		err = i40e_diag_eeprom_test(&pf->hw);
		if (err) {
			dev_info(&pf->pdev->dev, "eeprom check failed (%d), Tx/Rx traffic disabled\n",
				 err);
			set_bit(__I40E_BAD_EEPROM, &pf->state);
		}
	}

	if (!err && test_bit(__I40E_BAD_EEPROM, &pf->state)) {
		dev_info(&pf->pdev->dev, "eeprom check passed, Tx/Rx traffic enabled\n");
		clear_bit(__I40E_BAD_EEPROM, &pf->state);
	}
}

6474 6475
/**
 * i40e_enable_pf_switch_lb
6476
 * @pf: pointer to the PF structure
6477 6478 6479 6480 6481 6482 6483
 *
 * enable switch loop back or die - no point in a return value
 **/
static void i40e_enable_pf_switch_lb(struct i40e_pf *pf)
{
	struct i40e_vsi *vsi = pf->vsi[pf->lan_vsi];
	struct i40e_vsi_context ctxt;
6484
	int ret;
6485 6486 6487 6488

	ctxt.seid = pf->main_vsi_seid;
	ctxt.pf_num = pf->hw.pf_id;
	ctxt.vf_num = 0;
6489 6490
	ret = i40e_aq_get_vsi_params(&pf->hw, &ctxt, NULL);
	if (ret) {
6491
		dev_info(&pf->pdev->dev,
6492 6493 6494
			 "couldn't get PF vsi config, err %s aq_err %s\n",
			 i40e_stat_str(&pf->hw, ret),
			 i40e_aq_str(&pf->hw, pf->hw.aq.asq_last_status));
6495 6496 6497 6498 6499 6500
		return;
	}
	ctxt.flags = I40E_AQ_VSI_TYPE_PF;
	ctxt.info.valid_sections = cpu_to_le16(I40E_AQ_VSI_PROP_SWITCH_VALID);
	ctxt.info.switch_id |= cpu_to_le16(I40E_AQ_VSI_SW_ID_FLAG_ALLOW_LB);

6501 6502
	ret = i40e_aq_update_vsi_params(&vsi->back->hw, &ctxt, NULL);
	if (ret) {
6503
		dev_info(&pf->pdev->dev,
6504 6505 6506
			 "update vsi switch failed, err %s aq_err %s\n",
			 i40e_stat_str(&pf->hw, ret),
			 i40e_aq_str(&pf->hw, pf->hw.aq.asq_last_status));
6507 6508 6509 6510 6511
	}
}

/**
 * i40e_disable_pf_switch_lb
6512
 * @pf: pointer to the PF structure
6513 6514 6515 6516 6517 6518 6519
 *
 * disable switch loop back or die - no point in a return value
 **/
static void i40e_disable_pf_switch_lb(struct i40e_pf *pf)
{
	struct i40e_vsi *vsi = pf->vsi[pf->lan_vsi];
	struct i40e_vsi_context ctxt;
6520
	int ret;
6521 6522 6523 6524

	ctxt.seid = pf->main_vsi_seid;
	ctxt.pf_num = pf->hw.pf_id;
	ctxt.vf_num = 0;
6525 6526
	ret = i40e_aq_get_vsi_params(&pf->hw, &ctxt, NULL);
	if (ret) {
6527
		dev_info(&pf->pdev->dev,
6528 6529 6530
			 "couldn't get PF vsi config, err %s aq_err %s\n",
			 i40e_stat_str(&pf->hw, ret),
			 i40e_aq_str(&pf->hw, pf->hw.aq.asq_last_status));
6531 6532 6533 6534 6535 6536
		return;
	}
	ctxt.flags = I40E_AQ_VSI_TYPE_PF;
	ctxt.info.valid_sections = cpu_to_le16(I40E_AQ_VSI_PROP_SWITCH_VALID);
	ctxt.info.switch_id &= ~cpu_to_le16(I40E_AQ_VSI_SW_ID_FLAG_ALLOW_LB);

6537 6538
	ret = i40e_aq_update_vsi_params(&vsi->back->hw, &ctxt, NULL);
	if (ret) {
6539
		dev_info(&pf->pdev->dev,
6540 6541 6542
			 "update vsi switch failed, err %s aq_err %s\n",
			 i40e_stat_str(&pf->hw, ret),
			 i40e_aq_str(&pf->hw, pf->hw.aq.asq_last_status));
6543 6544 6545
	}
}

6546 6547 6548 6549 6550 6551 6552 6553 6554 6555 6556 6557
/**
 * i40e_config_bridge_mode - Configure the HW bridge mode
 * @veb: pointer to the bridge instance
 *
 * Configure the loop back mode for the LAN VSI that is downlink to the
 * specified HW bridge instance. It is expected this function is called
 * when a new HW bridge is instantiated.
 **/
static void i40e_config_bridge_mode(struct i40e_veb *veb)
{
	struct i40e_pf *pf = veb->pf;

6558 6559 6560
	if (pf->hw.debug_mask & I40E_DEBUG_LAN)
		dev_info(&pf->pdev->dev, "enabling bridge mode: %s\n",
			 veb->bridge_mode == BRIDGE_MODE_VEPA ? "VEPA" : "VEB");
6561 6562 6563 6564 6565 6566
	if (veb->bridge_mode & BRIDGE_MODE_VEPA)
		i40e_disable_pf_switch_lb(pf);
	else
		i40e_enable_pf_switch_lb(pf);
}

J
Jesse Brandeburg 已提交
6567 6568 6569 6570 6571 6572 6573 6574 6575 6576 6577 6578 6579 6580 6581 6582 6583
/**
 * i40e_reconstitute_veb - rebuild the VEB and anything connected to it
 * @veb: pointer to the VEB instance
 *
 * This is a recursive function that first builds the attached VSIs then
 * recurses in to build the next layer of VEB.  We track the connections
 * through our own index numbers because the seid's from the HW could
 * change across the reset.
 **/
static int i40e_reconstitute_veb(struct i40e_veb *veb)
{
	struct i40e_vsi *ctl_vsi = NULL;
	struct i40e_pf *pf = veb->pf;
	int v, veb_idx;
	int ret;

	/* build VSI that owns this VEB, temporarily attached to base VEB */
M
Mitch Williams 已提交
6584
	for (v = 0; v < pf->num_alloc_vsi && !ctl_vsi; v++) {
J
Jesse Brandeburg 已提交
6585 6586 6587 6588 6589 6590 6591 6592 6593 6594 6595 6596 6597 6598 6599 6600 6601 6602
		if (pf->vsi[v] &&
		    pf->vsi[v]->veb_idx == veb->idx &&
		    pf->vsi[v]->flags & I40E_VSI_FLAG_VEB_OWNER) {
			ctl_vsi = pf->vsi[v];
			break;
		}
	}
	if (!ctl_vsi) {
		dev_info(&pf->pdev->dev,
			 "missing owner VSI for veb_idx %d\n", veb->idx);
		ret = -ENOENT;
		goto end_reconstitute;
	}
	if (ctl_vsi != pf->vsi[pf->lan_vsi])
		ctl_vsi->uplink_seid = pf->vsi[pf->lan_vsi]->uplink_seid;
	ret = i40e_add_vsi(ctl_vsi);
	if (ret) {
		dev_info(&pf->pdev->dev,
6603 6604
			 "rebuild of veb_idx %d owner VSI failed: %d\n",
			 veb->idx, ret);
J
Jesse Brandeburg 已提交
6605 6606 6607 6608 6609 6610 6611 6612 6613
		goto end_reconstitute;
	}
	i40e_vsi_reset_stats(ctl_vsi);

	/* create the VEB in the switch and move the VSI onto the VEB */
	ret = i40e_add_veb(veb, ctl_vsi);
	if (ret)
		goto end_reconstitute;

6614 6615 6616 6617
	if (pf->flags & I40E_FLAG_VEB_MODE_ENABLED)
		veb->bridge_mode = BRIDGE_MODE_VEB;
	else
		veb->bridge_mode = BRIDGE_MODE_VEPA;
6618
	i40e_config_bridge_mode(veb);
6619

J
Jesse Brandeburg 已提交
6620
	/* create the remaining VSIs attached to this VEB */
M
Mitch Williams 已提交
6621
	for (v = 0; v < pf->num_alloc_vsi; v++) {
J
Jesse Brandeburg 已提交
6622 6623 6624 6625 6626
		if (!pf->vsi[v] || pf->vsi[v] == ctl_vsi)
			continue;

		if (pf->vsi[v]->veb_idx == veb->idx) {
			struct i40e_vsi *vsi = pf->vsi[v];
J
Jesse Brandeburg 已提交
6627

J
Jesse Brandeburg 已提交
6628 6629 6630 6631 6632 6633 6634 6635 6636 6637 6638 6639 6640 6641 6642 6643 6644 6645 6646 6647 6648 6649 6650 6651 6652 6653 6654 6655 6656 6657 6658 6659 6660 6661 6662 6663 6664 6665 6666 6667 6668 6669 6670 6671 6672 6673 6674 6675 6676 6677 6678 6679 6680 6681 6682 6683
			vsi->uplink_seid = veb->seid;
			ret = i40e_add_vsi(vsi);
			if (ret) {
				dev_info(&pf->pdev->dev,
					 "rebuild of vsi_idx %d failed: %d\n",
					 v, ret);
				goto end_reconstitute;
			}
			i40e_vsi_reset_stats(vsi);
		}
	}

	/* create any VEBs attached to this VEB - RECURSION */
	for (veb_idx = 0; veb_idx < I40E_MAX_VEB; veb_idx++) {
		if (pf->veb[veb_idx] && pf->veb[veb_idx]->veb_idx == veb->idx) {
			pf->veb[veb_idx]->uplink_seid = veb->seid;
			ret = i40e_reconstitute_veb(pf->veb[veb_idx]);
			if (ret)
				break;
		}
	}

end_reconstitute:
	return ret;
}

/**
 * i40e_get_capabilities - get info about the HW
 * @pf: the PF struct
 **/
static int i40e_get_capabilities(struct i40e_pf *pf)
{
	struct i40e_aqc_list_capabilities_element_resp *cap_buf;
	u16 data_size;
	int buf_len;
	int err;

	buf_len = 40 * sizeof(struct i40e_aqc_list_capabilities_element_resp);
	do {
		cap_buf = kzalloc(buf_len, GFP_KERNEL);
		if (!cap_buf)
			return -ENOMEM;

		/* this loads the data into the hw struct for us */
		err = i40e_aq_discover_capabilities(&pf->hw, cap_buf, buf_len,
					    &data_size,
					    i40e_aqc_opc_list_func_capabilities,
					    NULL);
		/* data loaded, buffer no longer needed */
		kfree(cap_buf);

		if (pf->hw.aq.asq_last_status == I40E_AQ_RC_ENOMEM) {
			/* retry with a larger buffer */
			buf_len = data_size;
		} else if (pf->hw.aq.asq_last_status != I40E_AQ_RC_OK) {
			dev_info(&pf->pdev->dev,
6684 6685 6686 6687
				 "capability discovery failed, err %s aq_err %s\n",
				 i40e_stat_str(&pf->hw, err),
				 i40e_aq_str(&pf->hw,
					     pf->hw.aq.asq_last_status));
J
Jesse Brandeburg 已提交
6688 6689 6690 6691 6692 6693 6694 6695 6696 6697 6698 6699 6700 6701 6702
			return -ENODEV;
		}
	} while (err);

	if (pf->hw.debug_mask & I40E_DEBUG_USER)
		dev_info(&pf->pdev->dev,
			 "pf=%d, num_vfs=%d, msix_pf=%d, msix_vf=%d, fd_g=%d, fd_b=%d, pf_max_q=%d num_vsi=%d\n",
			 pf->hw.pf_id, pf->hw.func_caps.num_vfs,
			 pf->hw.func_caps.num_msix_vectors,
			 pf->hw.func_caps.num_msix_vectors_vf,
			 pf->hw.func_caps.fd_filters_guaranteed,
			 pf->hw.func_caps.fd_filters_best_effort,
			 pf->hw.func_caps.num_tx_qp,
			 pf->hw.func_caps.num_vsis);

6703 6704 6705 6706 6707 6708 6709 6710 6711
#define DEF_NUM_VSI (1 + (pf->hw.func_caps.fcoe ? 1 : 0) \
		       + pf->hw.func_caps.num_vfs)
	if (pf->hw.revision_id == 0 && (DEF_NUM_VSI > pf->hw.func_caps.num_vsis)) {
		dev_info(&pf->pdev->dev,
			 "got num_vsis %d, setting num_vsis to %d\n",
			 pf->hw.func_caps.num_vsis, DEF_NUM_VSI);
		pf->hw.func_caps.num_vsis = DEF_NUM_VSI;
	}

J
Jesse Brandeburg 已提交
6712 6713 6714
	return 0;
}

6715 6716
static int i40e_vsi_clear(struct i40e_vsi *vsi);

J
Jesse Brandeburg 已提交
6717
/**
6718
 * i40e_fdir_sb_setup - initialize the Flow Director resources for Sideband
J
Jesse Brandeburg 已提交
6719 6720
 * @pf: board private structure
 **/
6721
static void i40e_fdir_sb_setup(struct i40e_pf *pf)
J
Jesse Brandeburg 已提交
6722 6723 6724
{
	struct i40e_vsi *vsi;

6725 6726 6727 6728 6729 6730 6731 6732 6733
	/* quick workaround for an NVM issue that leaves a critical register
	 * uninitialized
	 */
	if (!rd32(&pf->hw, I40E_GLQF_HKEY(0))) {
		static const u32 hkey[] = {
			0xe640d33f, 0xcdfe98ab, 0x73fa7161, 0x0d7a7d36,
			0xeacb7d61, 0xaa4f05b6, 0x9c5c89ed, 0xfc425ddb,
			0xa4654832, 0xfc7461d4, 0x8f827619, 0xf5c63c21,
			0x95b3a76d};
6734
		int i;
6735 6736 6737 6738 6739

		for (i = 0; i <= I40E_GLQF_HKEY_MAX_INDEX; i++)
			wr32(&pf->hw, I40E_GLQF_HKEY(i), hkey[i]);
	}

6740
	if (!(pf->flags & I40E_FLAG_FD_SB_ENABLED))
J
Jesse Brandeburg 已提交
6741 6742
		return;

6743
	/* find existing VSI and see if it needs configuring */
6744
	vsi = i40e_find_vsi_by_type(pf, I40E_VSI_FDIR);
6745 6746

	/* create a new VSI if none exists */
J
Jesse Brandeburg 已提交
6747
	if (!vsi) {
6748 6749
		vsi = i40e_vsi_setup(pf, I40E_VSI_FDIR,
				     pf->vsi[pf->lan_vsi]->seid, 0);
J
Jesse Brandeburg 已提交
6750 6751
		if (!vsi) {
			dev_info(&pf->pdev->dev, "Couldn't create FDir VSI\n");
6752 6753
			pf->flags &= ~I40E_FLAG_FD_SB_ENABLED;
			return;
J
Jesse Brandeburg 已提交
6754
		}
6755
	}
J
Jesse Brandeburg 已提交
6756

6757
	i40e_vsi_setup_irqhandler(vsi, i40e_fdir_clean_ring);
J
Jesse Brandeburg 已提交
6758 6759 6760 6761 6762 6763 6764 6765
}

/**
 * i40e_fdir_teardown - release the Flow Director resources
 * @pf: board private structure
 **/
static void i40e_fdir_teardown(struct i40e_pf *pf)
{
6766
	struct i40e_vsi *vsi;
J
Jesse Brandeburg 已提交
6767

6768
	i40e_fdir_filter_exit(pf);
6769 6770 6771
	vsi = i40e_find_vsi_by_type(pf, I40E_VSI_FDIR);
	if (vsi)
		i40e_vsi_release(vsi);
J
Jesse Brandeburg 已提交
6772 6773 6774
}

/**
A
Anjali Singhai Jain 已提交
6775
 * i40e_prep_for_reset - prep for the core to reset
J
Jesse Brandeburg 已提交
6776 6777
 * @pf: board private structure
 *
6778
 * Close up the VFs and other things in prep for PF Reset.
A
Anjali Singhai Jain 已提交
6779
  **/
S
Shannon Nelson 已提交
6780
static void i40e_prep_for_reset(struct i40e_pf *pf)
J
Jesse Brandeburg 已提交
6781 6782
{
	struct i40e_hw *hw = &pf->hw;
6783
	i40e_status ret = 0;
J
Jesse Brandeburg 已提交
6784 6785 6786 6787
	u32 v;

	clear_bit(__I40E_RESET_INTR_RECEIVED, &pf->state);
	if (test_and_set_bit(__I40E_RESET_RECOVERY_PENDING, &pf->state))
S
Shannon Nelson 已提交
6788
		return;
M
Mitch Williams 已提交
6789 6790
	if (i40e_check_asq_alive(&pf->hw))
		i40e_vc_notify_reset(pf);
J
Jesse Brandeburg 已提交
6791

J
Jesse Brandeburg 已提交
6792
	dev_dbg(&pf->pdev->dev, "Tearing down internal switch for reset\n");
J
Jesse Brandeburg 已提交
6793 6794 6795 6796

	/* quiesce the VSIs and their queues that are not already DOWN */
	i40e_pf_quiesce_all_vsi(pf);

M
Mitch Williams 已提交
6797
	for (v = 0; v < pf->num_alloc_vsi; v++) {
J
Jesse Brandeburg 已提交
6798 6799 6800 6801 6802 6803
		if (pf->vsi[v])
			pf->vsi[v]->seid = 0;
	}

	i40e_shutdown_adminq(&pf->hw);

A
Anjali Singhai Jain 已提交
6804
	/* call shutdown HMC */
6805 6806
	if (hw->hmc.hmc_obj) {
		ret = i40e_shutdown_lan_hmc(hw);
S
Shannon Nelson 已提交
6807
		if (ret)
6808 6809
			dev_warn(&pf->pdev->dev,
				 "shutdown_lan_hmc failed: %d\n", ret);
A
Anjali Singhai Jain 已提交
6810 6811 6812
	}
}

J
Jesse Brandeburg 已提交
6813 6814 6815 6816 6817 6818 6819 6820 6821 6822 6823 6824
/**
 * i40e_send_version - update firmware with driver version
 * @pf: PF struct
 */
static void i40e_send_version(struct i40e_pf *pf)
{
	struct i40e_driver_version dv;

	dv.major_version = DRV_VERSION_MAJOR;
	dv.minor_version = DRV_VERSION_MINOR;
	dv.build_version = DRV_VERSION_BUILD;
	dv.subbuild_version = 0;
6825
	strlcpy(dv.driver_string, DRV_VERSION, sizeof(dv.driver_string));
J
Jesse Brandeburg 已提交
6826 6827 6828
	i40e_aq_send_driver_version(&pf->hw, &dv, NULL);
}

A
Anjali Singhai Jain 已提交
6829
/**
J
Jesse Brandeburg 已提交
6830
 * i40e_reset_and_rebuild - reset and rebuild using a saved config
A
Anjali Singhai Jain 已提交
6831
 * @pf: board private structure
6832
 * @reinit: if the Main VSI needs to re-initialized.
A
Anjali Singhai Jain 已提交
6833
 **/
6834
static void i40e_reset_and_rebuild(struct i40e_pf *pf, bool reinit)
A
Anjali Singhai Jain 已提交
6835 6836
{
	struct i40e_hw *hw = &pf->hw;
6837
	u8 set_fc_aq_fail = 0;
A
Anjali Singhai Jain 已提交
6838
	i40e_status ret;
6839
	u32 val;
A
Anjali Singhai Jain 已提交
6840 6841
	u32 v;

J
Jesse Brandeburg 已提交
6842 6843 6844 6845 6846
	/* Now we wait for GRST to settle out.
	 * We don't have to delete the VEBs or VSIs from the hw switch
	 * because the reset will make them disappear.
	 */
	ret = i40e_pf_reset(hw);
6847
	if (ret) {
J
Jesse Brandeburg 已提交
6848
		dev_info(&pf->pdev->dev, "PF reset failed, %d\n", ret);
6849 6850
		set_bit(__I40E_RESET_FAILED, &pf->state);
		goto clear_recovery;
6851
	}
J
Jesse Brandeburg 已提交
6852 6853 6854
	pf->pfr_count++;

	if (test_bit(__I40E_DOWN, &pf->state))
6855
		goto clear_recovery;
J
Jesse Brandeburg 已提交
6856
	dev_dbg(&pf->pdev->dev, "Rebuilding internal switch\n");
J
Jesse Brandeburg 已提交
6857 6858 6859 6860

	/* rebuild the basics for the AdminQ, HMC, and initial HW switch */
	ret = i40e_init_adminq(&pf->hw);
	if (ret) {
6861 6862 6863
		dev_info(&pf->pdev->dev, "Rebuild AdminQ failed, err %s aq_err %s\n",
			 i40e_stat_str(&pf->hw, ret),
			 i40e_aq_str(&pf->hw, pf->hw.aq.asq_last_status));
6864
		goto clear_recovery;
J
Jesse Brandeburg 已提交
6865 6866
	}

6867
	/* re-verify the eeprom if we just had an EMP reset */
6868
	if (test_and_clear_bit(__I40E_EMP_RESET_INTR_RECEIVED, &pf->state))
6869 6870
		i40e_verify_eeprom(pf);

6871
	i40e_clear_pxe_mode(hw);
J
Jesse Brandeburg 已提交
6872
	ret = i40e_get_capabilities(pf);
6873
	if (ret)
J
Jesse Brandeburg 已提交
6874 6875 6876 6877 6878 6879 6880 6881 6882 6883 6884 6885 6886 6887 6888
		goto end_core_reset;

	ret = i40e_init_lan_hmc(hw, hw->func_caps.num_tx_qp,
				hw->func_caps.num_rx_qp,
				pf->fcoe_hmc_cntx_num, pf->fcoe_hmc_filt_num);
	if (ret) {
		dev_info(&pf->pdev->dev, "init_lan_hmc failed: %d\n", ret);
		goto end_core_reset;
	}
	ret = i40e_configure_lan_hmc(hw, I40E_HMC_MODEL_DIRECT_ONLY);
	if (ret) {
		dev_info(&pf->pdev->dev, "configure_lan_hmc failed: %d\n", ret);
		goto end_core_reset;
	}

N
Neerav Parikh 已提交
6889 6890 6891
#ifdef CONFIG_I40E_DCB
	ret = i40e_init_pf_dcb(pf);
	if (ret) {
6892 6893 6894
		dev_info(&pf->pdev->dev, "DCB init failed %d, disabled\n", ret);
		pf->flags &= ~I40E_FLAG_DCB_CAPABLE;
		/* Continue without DCB enabled */
N
Neerav Parikh 已提交
6895 6896
	}
#endif /* CONFIG_I40E_DCB */
6897
#ifdef I40E_FCOE
6898
	i40e_init_pf_fcoe(pf);
N
Neerav Parikh 已提交
6899

6900
#endif
J
Jesse Brandeburg 已提交
6901
	/* do basic switch setup */
6902
	ret = i40e_setup_pf_switch(pf, reinit);
J
Jesse Brandeburg 已提交
6903 6904 6905
	if (ret)
		goto end_core_reset;

S
Shannon Nelson 已提交
6906 6907
	/* The driver only wants link up/down and module qualification
	 * reports from firmware.  Note the negative logic.
J
Jesse Brandeburg 已提交
6908 6909
	 */
	ret = i40e_aq_set_phy_int_mask(&pf->hw,
S
Shannon Nelson 已提交
6910
				       ~(I40E_AQ_EVENT_LINK_UPDOWN |
6911
					 I40E_AQ_EVENT_MEDIA_NA |
S
Shannon Nelson 已提交
6912
					 I40E_AQ_EVENT_MODULE_QUAL_FAIL), NULL);
J
Jesse Brandeburg 已提交
6913
	if (ret)
6914 6915 6916
		dev_info(&pf->pdev->dev, "set phy mask fail, err %s aq_err %s\n",
			 i40e_stat_str(&pf->hw, ret),
			 i40e_aq_str(&pf->hw, pf->hw.aq.asq_last_status));
J
Jesse Brandeburg 已提交
6917

6918 6919 6920
	/* make sure our flow control settings are restored */
	ret = i40e_set_fc(&pf->hw, &set_fc_aq_fail, true);
	if (ret)
6921 6922 6923
		dev_dbg(&pf->pdev->dev, "setting flow control: ret = %s last_status = %s\n",
			i40e_stat_str(&pf->hw, ret),
			i40e_aq_str(&pf->hw, pf->hw.aq.asq_last_status));
6924

J
Jesse Brandeburg 已提交
6925 6926 6927 6928 6929 6930 6931 6932
	/* Rebuild the VSIs and VEBs that existed before reset.
	 * They are still in our local switch element arrays, so only
	 * need to rebuild the switch model in the HW.
	 *
	 * If there were VEBs but the reconstitution failed, we'll try
	 * try to recover minimal use by getting the basic PF VSI working.
	 */
	if (pf->vsi[pf->lan_vsi]->uplink_seid != pf->mac_seid) {
J
Jesse Brandeburg 已提交
6933
		dev_dbg(&pf->pdev->dev, "attempting to rebuild switch\n");
J
Jesse Brandeburg 已提交
6934 6935 6936 6937 6938 6939 6940 6941 6942 6943 6944 6945 6946 6947 6948 6949 6950 6951 6952 6953 6954 6955 6956 6957 6958 6959 6960 6961 6962 6963 6964 6965 6966 6967 6968
		/* find the one VEB connected to the MAC, and find orphans */
		for (v = 0; v < I40E_MAX_VEB; v++) {
			if (!pf->veb[v])
				continue;

			if (pf->veb[v]->uplink_seid == pf->mac_seid ||
			    pf->veb[v]->uplink_seid == 0) {
				ret = i40e_reconstitute_veb(pf->veb[v]);

				if (!ret)
					continue;

				/* If Main VEB failed, we're in deep doodoo,
				 * so give up rebuilding the switch and set up
				 * for minimal rebuild of PF VSI.
				 * If orphan failed, we'll report the error
				 * but try to keep going.
				 */
				if (pf->veb[v]->uplink_seid == pf->mac_seid) {
					dev_info(&pf->pdev->dev,
						 "rebuild of switch failed: %d, will try to set up simple PF connection\n",
						 ret);
					pf->vsi[pf->lan_vsi]->uplink_seid
								= pf->mac_seid;
					break;
				} else if (pf->veb[v]->uplink_seid == 0) {
					dev_info(&pf->pdev->dev,
						 "rebuild of orphan VEB failed: %d\n",
						 ret);
				}
			}
		}
	}

	if (pf->vsi[pf->lan_vsi]->uplink_seid == pf->mac_seid) {
S
Shannon Nelson 已提交
6969
		dev_dbg(&pf->pdev->dev, "attempting to rebuild PF VSI\n");
J
Jesse Brandeburg 已提交
6970 6971 6972 6973 6974 6975 6976 6977 6978
		/* no VEB, so rebuild only the Main VSI */
		ret = i40e_add_vsi(pf->vsi[pf->lan_vsi]);
		if (ret) {
			dev_info(&pf->pdev->dev,
				 "rebuild of Main VSI failed: %d\n", ret);
			goto end_core_reset;
		}
	}

6979 6980 6981 6982 6983 6984 6985 6986 6987 6988 6989 6990 6991 6992
	/* Reconfigure hardware for allowing smaller MSS in the case
	 * of TSO, so that we avoid the MDD being fired and causing
	 * a reset in the case of small MSS+TSO.
	 */
#define I40E_REG_MSS          0x000E64DC
#define I40E_REG_MSS_MIN_MASK 0x3FF0000
#define I40E_64BYTE_MSS       0x400000
	val = rd32(hw, I40E_REG_MSS);
	if ((val & I40E_REG_MSS_MIN_MASK) > I40E_64BYTE_MSS) {
		val &= ~I40E_REG_MSS_MIN_MASK;
		val |= I40E_64BYTE_MSS;
		wr32(hw, I40E_REG_MSS, val);
	}

6993
	if (pf->flags & I40E_FLAG_RESTART_AUTONEG) {
6994 6995 6996
		msleep(75);
		ret = i40e_aq_set_link_restart_an(&pf->hw, true, NULL);
		if (ret)
6997 6998 6999 7000
			dev_info(&pf->pdev->dev, "link restart failed, err %s aq_err %s\n",
				 i40e_stat_str(&pf->hw, ret),
				 i40e_aq_str(&pf->hw,
					     pf->hw.aq.asq_last_status));
7001
	}
J
Jesse Brandeburg 已提交
7002 7003 7004 7005
	/* reinit the misc interrupt */
	if (pf->flags & I40E_FLAG_MSIX_ENABLED)
		ret = i40e_setup_misc_vector(pf);

7006 7007 7008 7009 7010 7011 7012 7013 7014
	/* Add a filter to drop all Flow control frames from any VSI from being
	 * transmitted. By doing so we stop a malicious VF from sending out
	 * PAUSE or PFC frames and potentially controlling traffic for other
	 * PF/VF VSIs.
	 * The FW can still send Flow control frames if enabled.
	 */
	i40e_add_filter_to_drop_tx_flow_control_frames(&pf->hw,
						       pf->main_vsi_seid);

J
Jesse Brandeburg 已提交
7015 7016 7017
	/* restart the VSIs that were rebuilt and running before the reset */
	i40e_pf_unquiesce_all_vsi(pf);

M
Mitch Williams 已提交
7018 7019 7020 7021 7022
	if (pf->num_alloc_vfs) {
		for (v = 0; v < pf->num_alloc_vfs; v++)
			i40e_reset_vf(&pf->vf[v], true);
	}

J
Jesse Brandeburg 已提交
7023
	/* tell the firmware that we're starting */
J
Jesse Brandeburg 已提交
7024
	i40e_send_version(pf);
J
Jesse Brandeburg 已提交
7025 7026

end_core_reset:
7027 7028
	clear_bit(__I40E_RESET_FAILED, &pf->state);
clear_recovery:
J
Jesse Brandeburg 已提交
7029 7030 7031
	clear_bit(__I40E_RESET_RECOVERY_PENDING, &pf->state);
}

A
Anjali Singhai Jain 已提交
7032
/**
7033
 * i40e_handle_reset_warning - prep for the PF to reset, reset and rebuild
A
Anjali Singhai Jain 已提交
7034 7035 7036 7037 7038 7039 7040
 * @pf: board private structure
 *
 * Close up the VFs and other things in prep for a Core Reset,
 * then get ready to rebuild the world.
 **/
static void i40e_handle_reset_warning(struct i40e_pf *pf)
{
S
Shannon Nelson 已提交
7041 7042
	i40e_prep_for_reset(pf);
	i40e_reset_and_rebuild(pf, false);
A
Anjali Singhai Jain 已提交
7043 7044
}

J
Jesse Brandeburg 已提交
7045 7046
/**
 * i40e_handle_mdd_event
7047
 * @pf: pointer to the PF structure
J
Jesse Brandeburg 已提交
7048 7049 7050 7051 7052 7053 7054
 *
 * Called from the MDD irq handler to identify possibly malicious vfs
 **/
static void i40e_handle_mdd_event(struct i40e_pf *pf)
{
	struct i40e_hw *hw = &pf->hw;
	bool mdd_detected = false;
7055
	bool pf_mdd_detected = false;
J
Jesse Brandeburg 已提交
7056 7057 7058 7059 7060 7061 7062 7063 7064 7065
	struct i40e_vf *vf;
	u32 reg;
	int i;

	if (!test_bit(__I40E_MDD_EVENT_PENDING, &pf->state))
		return;

	/* find what triggered the MDD event */
	reg = rd32(hw, I40E_GL_MDET_TX);
	if (reg & I40E_GL_MDET_TX_VALID_MASK) {
7066 7067
		u8 pf_num = (reg & I40E_GL_MDET_TX_PF_NUM_MASK) >>
				I40E_GL_MDET_TX_PF_NUM_SHIFT;
7068
		u16 vf_num = (reg & I40E_GL_MDET_TX_VF_NUM_MASK) >>
7069
				I40E_GL_MDET_TX_VF_NUM_SHIFT;
7070
		u8 event = (reg & I40E_GL_MDET_TX_EVENT_MASK) >>
7071
				I40E_GL_MDET_TX_EVENT_SHIFT;
7072 7073 7074
		u16 queue = ((reg & I40E_GL_MDET_TX_QUEUE_MASK) >>
				I40E_GL_MDET_TX_QUEUE_SHIFT) -
				pf->hw.func_caps.base_queue;
7075
		if (netif_msg_tx_err(pf))
7076
			dev_info(&pf->pdev->dev, "Malicious Driver Detection event 0x%02x on TX queue %d PF number 0x%02x VF number 0x%02x\n",
7077
				 event, queue, pf_num, vf_num);
J
Jesse Brandeburg 已提交
7078 7079 7080 7081 7082
		wr32(hw, I40E_GL_MDET_TX, 0xffffffff);
		mdd_detected = true;
	}
	reg = rd32(hw, I40E_GL_MDET_RX);
	if (reg & I40E_GL_MDET_RX_VALID_MASK) {
7083 7084
		u8 func = (reg & I40E_GL_MDET_RX_FUNCTION_MASK) >>
				I40E_GL_MDET_RX_FUNCTION_SHIFT;
7085
		u8 event = (reg & I40E_GL_MDET_RX_EVENT_MASK) >>
7086
				I40E_GL_MDET_RX_EVENT_SHIFT;
7087 7088 7089
		u16 queue = ((reg & I40E_GL_MDET_RX_QUEUE_MASK) >>
				I40E_GL_MDET_RX_QUEUE_SHIFT) -
				pf->hw.func_caps.base_queue;
7090 7091 7092
		if (netif_msg_rx_err(pf))
			dev_info(&pf->pdev->dev, "Malicious Driver Detection event 0x%02x on RX queue %d of function 0x%02x\n",
				 event, queue, func);
J
Jesse Brandeburg 已提交
7093 7094 7095 7096
		wr32(hw, I40E_GL_MDET_RX, 0xffffffff);
		mdd_detected = true;
	}

7097 7098 7099 7100
	if (mdd_detected) {
		reg = rd32(hw, I40E_PF_MDET_TX);
		if (reg & I40E_PF_MDET_TX_VALID_MASK) {
			wr32(hw, I40E_PF_MDET_TX, 0xFFFF);
7101
			dev_info(&pf->pdev->dev, "TX driver issue detected, PF reset issued\n");
7102 7103 7104 7105 7106
			pf_mdd_detected = true;
		}
		reg = rd32(hw, I40E_PF_MDET_RX);
		if (reg & I40E_PF_MDET_RX_VALID_MASK) {
			wr32(hw, I40E_PF_MDET_RX, 0xFFFF);
7107
			dev_info(&pf->pdev->dev, "RX driver issue detected, PF reset issued\n");
7108 7109 7110 7111 7112 7113 7114 7115 7116
			pf_mdd_detected = true;
		}
		/* Queue belongs to the PF, initiate a reset */
		if (pf_mdd_detected) {
			set_bit(__I40E_PF_RESET_REQUESTED, &pf->state);
			i40e_service_event_schedule(pf);
		}
	}

J
Jesse Brandeburg 已提交
7117 7118 7119 7120 7121 7122 7123
	/* see if one of the VFs needs its hand slapped */
	for (i = 0; i < pf->num_alloc_vfs && mdd_detected; i++) {
		vf = &(pf->vf[i]);
		reg = rd32(hw, I40E_VP_MDET_TX(i));
		if (reg & I40E_VP_MDET_TX_VALID_MASK) {
			wr32(hw, I40E_VP_MDET_TX(i), 0xFFFF);
			vf->num_mdd_events++;
7124 7125
			dev_info(&pf->pdev->dev, "TX driver issue detected on VF %d\n",
				 i);
J
Jesse Brandeburg 已提交
7126 7127 7128 7129 7130 7131
		}

		reg = rd32(hw, I40E_VP_MDET_RX(i));
		if (reg & I40E_VP_MDET_RX_VALID_MASK) {
			wr32(hw, I40E_VP_MDET_RX(i), 0xFFFF);
			vf->num_mdd_events++;
7132 7133
			dev_info(&pf->pdev->dev, "RX driver issue detected on VF %d\n",
				 i);
J
Jesse Brandeburg 已提交
7134 7135 7136 7137 7138 7139 7140 7141 7142 7143 7144 7145 7146 7147 7148 7149 7150 7151 7152
		}

		if (vf->num_mdd_events > I40E_DEFAULT_NUM_MDD_EVENTS_ALLOWED) {
			dev_info(&pf->pdev->dev,
				 "Too many MDD events on VF %d, disabled\n", i);
			dev_info(&pf->pdev->dev,
				 "Use PF Control I/F to re-enable the VF\n");
			set_bit(I40E_VF_STAT_DISABLED, &vf->vf_states);
		}
	}

	/* re-enable mdd interrupt cause */
	clear_bit(__I40E_MDD_EVENT_PENDING, &pf->state);
	reg = rd32(hw, I40E_PFINT_ICR0_ENA);
	reg |=  I40E_PFINT_ICR0_ENA_MAL_DETECT_MASK;
	wr32(hw, I40E_PFINT_ICR0_ENA, reg);
	i40e_flush(hw);
}

7153
/**
7154
 * i40e_sync_udp_filters_subtask - Sync the VSI filter list with HW
7155 7156
 * @pf: board private structure
 **/
7157
static void i40e_sync_udp_filters_subtask(struct i40e_pf *pf)
7158 7159 7160 7161 7162 7163
{
	struct i40e_hw *hw = &pf->hw;
	i40e_status ret;
	__be16 port;
	int i;

7164
	if (!(pf->flags & I40E_FLAG_UDP_FILTER_SYNC))
7165 7166
		return;

7167
	pf->flags &= ~I40E_FLAG_UDP_FILTER_SYNC;
7168 7169

	for (i = 0; i < I40E_MAX_PF_UDP_OFFLOAD_PORTS; i++) {
7170 7171 7172
		if (pf->pending_udp_bitmap & BIT_ULL(i)) {
			pf->pending_udp_bitmap &= ~BIT_ULL(i);
			port = pf->udp_ports[i].index;
S
Shannon Nelson 已提交
7173
			if (port)
7174 7175 7176
				ret = i40e_aq_add_udp_tunnel(hw, port,
							pf->udp_ports[i].type,
							NULL, NULL);
S
Shannon Nelson 已提交
7177 7178
			else
				ret = i40e_aq_del_udp_tunnel(hw, i, NULL);
7179 7180

			if (ret) {
7181 7182 7183 7184 7185 7186 7187
				dev_dbg(&pf->pdev->dev,
					"%s %s port %d, index %d failed, err %s aq_err %s\n",
					pf->udp_ports[i].type ? "vxlan" : "geneve",
					port ? "add" : "delete",
					ntohs(port), i,
					i40e_stat_str(&pf->hw, ret),
					i40e_aq_str(&pf->hw,
7188
						    pf->hw.aq.asq_last_status));
7189
				pf->udp_ports[i].index = 0;
7190 7191 7192 7193 7194
			}
		}
	}
}

J
Jesse Brandeburg 已提交
7195 7196 7197 7198 7199 7200 7201 7202 7203 7204 7205
/**
 * i40e_service_task - Run the driver's async subtasks
 * @work: pointer to work_struct containing our data
 **/
static void i40e_service_task(struct work_struct *work)
{
	struct i40e_pf *pf = container_of(work,
					  struct i40e_pf,
					  service_task);
	unsigned long start_time = jiffies;

7206 7207 7208 7209 7210 7211
	/* don't bother with service tasks if a reset is in progress */
	if (test_bit(__I40E_RESET_RECOVERY_PENDING, &pf->state)) {
		i40e_service_event_complete(pf);
		return;
	}

7212
	i40e_detect_recover_hung(pf);
7213
	i40e_sync_filters_subtask(pf);
J
Jesse Brandeburg 已提交
7214 7215 7216 7217 7218
	i40e_reset_subtask(pf);
	i40e_handle_mdd_event(pf);
	i40e_vc_process_vflr_event(pf);
	i40e_watchdog_subtask(pf);
	i40e_fdir_reinit_subtask(pf);
7219
	i40e_client_subtask(pf);
J
Jesse Brandeburg 已提交
7220
	i40e_sync_filters_subtask(pf);
7221
	i40e_sync_udp_filters_subtask(pf);
J
Jesse Brandeburg 已提交
7222 7223 7224 7225 7226 7227 7228 7229 7230 7231 7232 7233 7234 7235 7236 7237 7238 7239 7240 7241 7242 7243 7244 7245 7246 7247 7248 7249 7250 7251 7252 7253 7254 7255 7256 7257 7258 7259 7260 7261 7262 7263 7264 7265 7266 7267 7268 7269 7270 7271 7272 7273
	i40e_clean_adminq_subtask(pf);

	i40e_service_event_complete(pf);

	/* If the tasks have taken longer than one timer cycle or there
	 * is more work to be done, reschedule the service task now
	 * rather than wait for the timer to tick again.
	 */
	if (time_after(jiffies, (start_time + pf->service_timer_period)) ||
	    test_bit(__I40E_ADMINQ_EVENT_PENDING, &pf->state)		 ||
	    test_bit(__I40E_MDD_EVENT_PENDING, &pf->state)		 ||
	    test_bit(__I40E_VFLR_EVENT_PENDING, &pf->state))
		i40e_service_event_schedule(pf);
}

/**
 * i40e_service_timer - timer callback
 * @data: pointer to PF struct
 **/
static void i40e_service_timer(unsigned long data)
{
	struct i40e_pf *pf = (struct i40e_pf *)data;

	mod_timer(&pf->service_timer,
		  round_jiffies(jiffies + pf->service_timer_period));
	i40e_service_event_schedule(pf);
}

/**
 * i40e_set_num_rings_in_vsi - Determine number of rings in the VSI
 * @vsi: the VSI being configured
 **/
static int i40e_set_num_rings_in_vsi(struct i40e_vsi *vsi)
{
	struct i40e_pf *pf = vsi->back;

	switch (vsi->type) {
	case I40E_VSI_MAIN:
		vsi->alloc_queue_pairs = pf->num_lan_qps;
		vsi->num_desc = ALIGN(I40E_DEFAULT_NUM_DESCRIPTORS,
				      I40E_REQ_DESCRIPTOR_MULTIPLE);
		if (pf->flags & I40E_FLAG_MSIX_ENABLED)
			vsi->num_q_vectors = pf->num_lan_msix;
		else
			vsi->num_q_vectors = 1;

		break;

	case I40E_VSI_FDIR:
		vsi->alloc_queue_pairs = 1;
		vsi->num_desc = ALIGN(I40E_FDIR_RING_COUNT,
				      I40E_REQ_DESCRIPTOR_MULTIPLE);
7274
		vsi->num_q_vectors = pf->num_fdsb_msix;
J
Jesse Brandeburg 已提交
7275 7276 7277 7278 7279 7280 7281 7282 7283 7284 7285 7286 7287 7288 7289
		break;

	case I40E_VSI_VMDQ2:
		vsi->alloc_queue_pairs = pf->num_vmdq_qps;
		vsi->num_desc = ALIGN(I40E_DEFAULT_NUM_DESCRIPTORS,
				      I40E_REQ_DESCRIPTOR_MULTIPLE);
		vsi->num_q_vectors = pf->num_vmdq_msix;
		break;

	case I40E_VSI_SRIOV:
		vsi->alloc_queue_pairs = pf->num_vf_qps;
		vsi->num_desc = ALIGN(I40E_DEFAULT_NUM_DESCRIPTORS,
				      I40E_REQ_DESCRIPTOR_MULTIPLE);
		break;

7290 7291 7292 7293 7294 7295 7296 7297 7298
#ifdef I40E_FCOE
	case I40E_VSI_FCOE:
		vsi->alloc_queue_pairs = pf->num_fcoe_qps;
		vsi->num_desc = ALIGN(I40E_DEFAULT_NUM_DESCRIPTORS,
				      I40E_REQ_DESCRIPTOR_MULTIPLE);
		vsi->num_q_vectors = pf->num_fcoe_msix;
		break;

#endif /* I40E_FCOE */
J
Jesse Brandeburg 已提交
7299 7300 7301 7302 7303 7304 7305 7306
	default:
		WARN_ON(1);
		return -ENODATA;
	}

	return 0;
}

A
Anjali Singhai Jain 已提交
7307 7308 7309
/**
 * i40e_vsi_alloc_arrays - Allocate queue and vector pointer arrays for the vsi
 * @type: VSI pointer
7310
 * @alloc_qvectors: a bool to specify if q_vectors need to be allocated.
A
Anjali Singhai Jain 已提交
7311 7312 7313 7314
 *
 * On error: returns error code (negative)
 * On success: returns 0
 **/
7315
static int i40e_vsi_alloc_arrays(struct i40e_vsi *vsi, bool alloc_qvectors)
A
Anjali Singhai Jain 已提交
7316 7317 7318 7319
{
	int size;
	int ret = 0;

7320
	/* allocate memory for both Tx and Rx ring pointers */
A
Anjali Singhai Jain 已提交
7321 7322 7323 7324 7325 7326
	size = sizeof(struct i40e_ring *) * vsi->alloc_queue_pairs * 2;
	vsi->tx_rings = kzalloc(size, GFP_KERNEL);
	if (!vsi->tx_rings)
		return -ENOMEM;
	vsi->rx_rings = &vsi->tx_rings[vsi->alloc_queue_pairs];

7327 7328
	if (alloc_qvectors) {
		/* allocate memory for q_vector pointers */
7329
		size = sizeof(struct i40e_q_vector *) * vsi->num_q_vectors;
7330 7331 7332 7333 7334
		vsi->q_vectors = kzalloc(size, GFP_KERNEL);
		if (!vsi->q_vectors) {
			ret = -ENOMEM;
			goto err_vectors;
		}
A
Anjali Singhai Jain 已提交
7335 7336 7337 7338 7339 7340 7341 7342
	}
	return ret;

err_vectors:
	kfree(vsi->tx_rings);
	return ret;
}

J
Jesse Brandeburg 已提交
7343 7344 7345 7346 7347 7348 7349 7350 7351 7352 7353 7354 7355 7356 7357 7358 7359 7360 7361 7362 7363 7364 7365 7366 7367
/**
 * i40e_vsi_mem_alloc - Allocates the next available struct vsi in the PF
 * @pf: board private structure
 * @type: type of VSI
 *
 * On error: returns error code (negative)
 * On success: returns vsi index in PF (positive)
 **/
static int i40e_vsi_mem_alloc(struct i40e_pf *pf, enum i40e_vsi_type type)
{
	int ret = -ENODEV;
	struct i40e_vsi *vsi;
	int vsi_idx;
	int i;

	/* Need to protect the allocation of the VSIs at the PF level */
	mutex_lock(&pf->switch_mutex);

	/* VSI list may be fragmented if VSI creation/destruction has
	 * been happening.  We can afford to do a quick scan to look
	 * for any free VSIs in the list.
	 *
	 * find next empty vsi slot, looping back around if necessary
	 */
	i = pf->next_vsi;
M
Mitch Williams 已提交
7368
	while (i < pf->num_alloc_vsi && pf->vsi[i])
J
Jesse Brandeburg 已提交
7369
		i++;
M
Mitch Williams 已提交
7370
	if (i >= pf->num_alloc_vsi) {
J
Jesse Brandeburg 已提交
7371 7372 7373 7374 7375
		i = 0;
		while (i < pf->next_vsi && pf->vsi[i])
			i++;
	}

M
Mitch Williams 已提交
7376
	if (i < pf->num_alloc_vsi && !pf->vsi[i]) {
J
Jesse Brandeburg 已提交
7377 7378 7379
		vsi_idx = i;             /* Found one! */
	} else {
		ret = -ENODEV;
7380
		goto unlock_pf;  /* out of VSI slots! */
J
Jesse Brandeburg 已提交
7381 7382 7383 7384 7385 7386
	}
	pf->next_vsi = ++i;

	vsi = kzalloc(sizeof(*vsi), GFP_KERNEL);
	if (!vsi) {
		ret = -ENOMEM;
7387
		goto unlock_pf;
J
Jesse Brandeburg 已提交
7388 7389 7390 7391 7392 7393
	}
	vsi->type = type;
	vsi->back = pf;
	set_bit(__I40E_DOWN, &vsi->state);
	vsi->flags = 0;
	vsi->idx = vsi_idx;
7394
	vsi->int_rate_limit = 0;
7395 7396
	vsi->rss_table_size = (vsi->type == I40E_VSI_MAIN) ?
				pf->rss_table_size : 64;
J
Jesse Brandeburg 已提交
7397 7398
	vsi->netdev_registered = false;
	vsi->work_limit = I40E_DEFAULT_IRQ_WORK;
7399
	hash_init(vsi->mac_filter_hash);
7400
	vsi->irqs_ready = false;
J
Jesse Brandeburg 已提交
7401

7402 7403 7404 7405
	ret = i40e_set_num_rings_in_vsi(vsi);
	if (ret)
		goto err_rings;

7406
	ret = i40e_vsi_alloc_arrays(vsi, true);
A
Anjali Singhai Jain 已提交
7407
	if (ret)
7408
		goto err_rings;
7409

J
Jesse Brandeburg 已提交
7410 7411 7412
	/* Setup default MSIX irq handler for VSI */
	i40e_vsi_setup_irqhandler(vsi, i40e_msix_clean_rings);

K
Kiran Patil 已提交
7413
	/* Initialize VSI lock */
7414
	spin_lock_init(&vsi->mac_filter_hash_lock);
J
Jesse Brandeburg 已提交
7415 7416
	pf->vsi[vsi_idx] = vsi;
	ret = vsi_idx;
7417 7418
	goto unlock_pf;

7419
err_rings:
7420 7421 7422
	pf->next_vsi = i - 1;
	kfree(vsi);
unlock_pf:
J
Jesse Brandeburg 已提交
7423 7424 7425 7426
	mutex_unlock(&pf->switch_mutex);
	return ret;
}

A
Anjali Singhai Jain 已提交
7427 7428 7429
/**
 * i40e_vsi_free_arrays - Free queue and vector pointer arrays for the VSI
 * @type: VSI pointer
7430
 * @free_qvectors: a bool to specify if q_vectors need to be freed.
A
Anjali Singhai Jain 已提交
7431 7432 7433 7434
 *
 * On error: returns error code (negative)
 * On success: returns 0
 **/
7435
static void i40e_vsi_free_arrays(struct i40e_vsi *vsi, bool free_qvectors)
A
Anjali Singhai Jain 已提交
7436 7437
{
	/* free the ring and vector containers */
7438 7439 7440 7441
	if (free_qvectors) {
		kfree(vsi->q_vectors);
		vsi->q_vectors = NULL;
	}
A
Anjali Singhai Jain 已提交
7442 7443 7444 7445 7446
	kfree(vsi->tx_rings);
	vsi->tx_rings = NULL;
	vsi->rx_rings = NULL;
}

7447 7448 7449 7450 7451 7452 7453 7454 7455 7456 7457 7458 7459 7460 7461 7462 7463
/**
 * i40e_clear_rss_config_user - clear the user configured RSS hash keys
 * and lookup table
 * @vsi: Pointer to VSI structure
 */
static void i40e_clear_rss_config_user(struct i40e_vsi *vsi)
{
	if (!vsi)
		return;

	kfree(vsi->rss_hkey_user);
	vsi->rss_hkey_user = NULL;

	kfree(vsi->rss_lut_user);
	vsi->rss_lut_user = NULL;
}

J
Jesse Brandeburg 已提交
7464 7465 7466 7467 7468 7469 7470 7471 7472 7473 7474 7475 7476 7477 7478 7479 7480 7481 7482 7483 7484 7485 7486 7487 7488 7489 7490 7491 7492 7493 7494 7495
/**
 * i40e_vsi_clear - Deallocate the VSI provided
 * @vsi: the VSI being un-configured
 **/
static int i40e_vsi_clear(struct i40e_vsi *vsi)
{
	struct i40e_pf *pf;

	if (!vsi)
		return 0;

	if (!vsi->back)
		goto free_vsi;
	pf = vsi->back;

	mutex_lock(&pf->switch_mutex);
	if (!pf->vsi[vsi->idx]) {
		dev_err(&pf->pdev->dev, "pf->vsi[%d] is NULL, just free vsi[%d](%p,type %d)\n",
			vsi->idx, vsi->idx, vsi, vsi->type);
		goto unlock_vsi;
	}

	if (pf->vsi[vsi->idx] != vsi) {
		dev_err(&pf->pdev->dev,
			"pf->vsi[%d](%p, type %d) != vsi[%d](%p,type %d): no free!\n",
			pf->vsi[vsi->idx]->idx,
			pf->vsi[vsi->idx],
			pf->vsi[vsi->idx]->type,
			vsi->idx, vsi, vsi->type);
		goto unlock_vsi;
	}

7496
	/* updates the PF for this cleared vsi */
J
Jesse Brandeburg 已提交
7497 7498 7499
	i40e_put_lump(pf->qp_pile, vsi->base_queue, vsi->idx);
	i40e_put_lump(pf->irq_pile, vsi->base_vector, vsi->idx);

7500
	i40e_vsi_free_arrays(vsi, true);
7501
	i40e_clear_rss_config_user(vsi);
7502

J
Jesse Brandeburg 已提交
7503 7504 7505 7506 7507 7508 7509 7510 7511 7512 7513 7514
	pf->vsi[vsi->idx] = NULL;
	if (vsi->idx < pf->next_vsi)
		pf->next_vsi = vsi->idx;

unlock_vsi:
	mutex_unlock(&pf->switch_mutex);
free_vsi:
	kfree(vsi);

	return 0;
}

7515 7516 7517 7518
/**
 * i40e_vsi_clear_rings - Deallocates the Rx and Tx rings for the provided VSI
 * @vsi: the VSI being cleaned
 **/
7519
static void i40e_vsi_clear_rings(struct i40e_vsi *vsi)
7520 7521 7522
{
	int i;

7523
	if (vsi->tx_rings && vsi->tx_rings[0]) {
N
Neerav Parikh 已提交
7524
		for (i = 0; i < vsi->alloc_queue_pairs; i++) {
7525 7526 7527 7528
			kfree_rcu(vsi->tx_rings[i], rcu);
			vsi->tx_rings[i] = NULL;
			vsi->rx_rings[i] = NULL;
		}
7529
	}
7530 7531
}

J
Jesse Brandeburg 已提交
7532 7533 7534 7535 7536 7537
/**
 * i40e_alloc_rings - Allocates the Rx and Tx rings for the provided VSI
 * @vsi: the VSI being configured
 **/
static int i40e_alloc_rings(struct i40e_vsi *vsi)
{
7538
	struct i40e_ring *tx_ring, *rx_ring;
J
Jesse Brandeburg 已提交
7539 7540 7541 7542
	struct i40e_pf *pf = vsi->back;
	int i;

	/* Set basic values in the rings to be used later during open() */
N
Neerav Parikh 已提交
7543
	for (i = 0; i < vsi->alloc_queue_pairs; i++) {
7544
		/* allocate space for both Tx and Rx in one shot */
7545 7546 7547
		tx_ring = kzalloc(sizeof(struct i40e_ring) * 2, GFP_KERNEL);
		if (!tx_ring)
			goto err_out;
J
Jesse Brandeburg 已提交
7548 7549 7550 7551 7552 7553 7554 7555 7556 7557

		tx_ring->queue_index = i;
		tx_ring->reg_idx = vsi->base_queue + i;
		tx_ring->ring_active = false;
		tx_ring->vsi = vsi;
		tx_ring->netdev = vsi->netdev;
		tx_ring->dev = &pf->pdev->dev;
		tx_ring->count = vsi->num_desc;
		tx_ring->size = 0;
		tx_ring->dcb_tc = 0;
7558 7559
		if (vsi->back->flags & I40E_FLAG_WB_ON_ITR_CAPABLE)
			tx_ring->flags = I40E_TXR_FLAGS_WB_ON_ITR;
7560
		tx_ring->tx_itr_setting = pf->tx_itr_default;
7561
		vsi->tx_rings[i] = tx_ring;
J
Jesse Brandeburg 已提交
7562

7563
		rx_ring = &tx_ring[1];
J
Jesse Brandeburg 已提交
7564 7565 7566 7567 7568 7569 7570 7571 7572
		rx_ring->queue_index = i;
		rx_ring->reg_idx = vsi->base_queue + i;
		rx_ring->ring_active = false;
		rx_ring->vsi = vsi;
		rx_ring->netdev = vsi->netdev;
		rx_ring->dev = &pf->pdev->dev;
		rx_ring->count = vsi->num_desc;
		rx_ring->size = 0;
		rx_ring->dcb_tc = 0;
7573
		rx_ring->rx_itr_setting = pf->rx_itr_default;
7574
		vsi->rx_rings[i] = rx_ring;
J
Jesse Brandeburg 已提交
7575 7576 7577
	}

	return 0;
7578 7579 7580 7581

err_out:
	i40e_vsi_clear_rings(vsi);
	return -ENOMEM;
J
Jesse Brandeburg 已提交
7582 7583 7584 7585 7586 7587 7588 7589 7590 7591 7592
}

/**
 * i40e_reserve_msix_vectors - Reserve MSI-X vectors in the kernel
 * @pf: board private structure
 * @vectors: the number of MSI-X vectors to request
 *
 * Returns the number of vectors reserved, or error
 **/
static int i40e_reserve_msix_vectors(struct i40e_pf *pf, int vectors)
{
7593 7594 7595
	vectors = pci_enable_msix_range(pf->pdev, pf->msix_entries,
					I40E_MIN_MSIX, vectors);
	if (vectors < 0) {
J
Jesse Brandeburg 已提交
7596
		dev_info(&pf->pdev->dev,
7597
			 "MSI-X vector reservation failed: %d\n", vectors);
J
Jesse Brandeburg 已提交
7598 7599 7600 7601 7602 7603 7604 7605 7606 7607 7608 7609
		vectors = 0;
	}

	return vectors;
}

/**
 * i40e_init_msix - Setup the MSIX capability
 * @pf: board private structure
 *
 * Work with the OS to set up the MSIX vectors needed.
 *
7610
 * Returns the number of vectors reserved or negative on failure
J
Jesse Brandeburg 已提交
7611 7612 7613 7614
 **/
static int i40e_init_msix(struct i40e_pf *pf)
{
	struct i40e_hw *hw = &pf->hw;
S
Shannon Nelson 已提交
7615
	int vectors_left;
J
Jesse Brandeburg 已提交
7616
	int v_budget, i;
7617
	int v_actual;
7618
	int iwarp_requested = 0;
J
Jesse Brandeburg 已提交
7619 7620 7621 7622 7623 7624 7625

	if (!(pf->flags & I40E_FLAG_MSIX_ENABLED))
		return -ENODEV;

	/* The number of vectors we'll request will be comprised of:
	 *   - Add 1 for "other" cause for Admin Queue events, etc.
	 *   - The number of LAN queue pairs
7626 7627 7628 7629 7630
	 *	- Queues being used for RSS.
	 *		We don't need as many as max_rss_size vectors.
	 *		use rss_size instead in the calculation since that
	 *		is governed by number of cpus in the system.
	 *	- assumes symmetric Tx/Rx pairing
J
Jesse Brandeburg 已提交
7631
	 *   - The number of VMDq pairs
7632
	 *   - The CPU count within the NUMA node if iWARP is enabled
7633 7634 7635
#ifdef I40E_FCOE
	 *   - The number of FCOE qps.
#endif
J
Jesse Brandeburg 已提交
7636 7637 7638 7639 7640
	 * Once we count this up, try the request.
	 *
	 * If we can't get what we want, we'll simplify to nearly nothing
	 * and try again.  If that still fails, we punt.
	 */
S
Shannon Nelson 已提交
7641 7642 7643 7644 7645 7646 7647 7648 7649 7650 7651 7652 7653 7654 7655 7656 7657
	vectors_left = hw->func_caps.num_msix_vectors;
	v_budget = 0;

	/* reserve one vector for miscellaneous handler */
	if (vectors_left) {
		v_budget++;
		vectors_left--;
	}

	/* reserve vectors for the main PF traffic queues */
	pf->num_lan_msix = min_t(int, num_online_cpus(), vectors_left);
	vectors_left -= pf->num_lan_msix;
	v_budget += pf->num_lan_msix;

	/* reserve one vector for sideband flow director */
	if (pf->flags & I40E_FLAG_FD_SB_ENABLED) {
		if (vectors_left) {
7658
			pf->num_fdsb_msix = 1;
S
Shannon Nelson 已提交
7659 7660 7661
			v_budget++;
			vectors_left--;
		} else {
7662
			pf->num_fdsb_msix = 0;
S
Shannon Nelson 已提交
7663 7664
		}
	}
7665

7666
#ifdef I40E_FCOE
S
Shannon Nelson 已提交
7667
	/* can we reserve enough for FCoE? */
7668
	if (pf->flags & I40E_FLAG_FCOE_ENABLED) {
S
Shannon Nelson 已提交
7669 7670 7671 7672 7673 7674
		if (!vectors_left)
			pf->num_fcoe_msix = 0;
		else if (vectors_left >= pf->num_fcoe_qps)
			pf->num_fcoe_msix = pf->num_fcoe_qps;
		else
			pf->num_fcoe_msix = 1;
7675
		v_budget += pf->num_fcoe_msix;
S
Shannon Nelson 已提交
7676
		vectors_left -= pf->num_fcoe_msix;
7677
	}
S
Shannon Nelson 已提交
7678

7679
#endif
7680 7681
	/* can we reserve enough for iWARP? */
	if (pf->flags & I40E_FLAG_IWARP_ENABLED) {
7682 7683
		iwarp_requested = pf->num_iwarp_msix;

7684 7685 7686 7687 7688 7689 7690 7691
		if (!vectors_left)
			pf->num_iwarp_msix = 0;
		else if (vectors_left < pf->num_iwarp_msix)
			pf->num_iwarp_msix = 1;
		v_budget += pf->num_iwarp_msix;
		vectors_left -= pf->num_iwarp_msix;
	}

S
Shannon Nelson 已提交
7692 7693 7694 7695 7696
	/* any vectors left over go for VMDq support */
	if (pf->flags & I40E_FLAG_VMDQ_ENABLED) {
		int vmdq_vecs_wanted = pf->num_vmdq_vsis * pf->num_vmdq_qps;
		int vmdq_vecs = min_t(int, vectors_left, vmdq_vecs_wanted);

7697 7698 7699 7700 7701 7702 7703 7704 7705 7706 7707 7708 7709
		if (!vectors_left) {
			pf->num_vmdq_msix = 0;
			pf->num_vmdq_qps = 0;
		} else {
			/* if we're short on vectors for what's desired, we limit
			 * the queues per vmdq.  If this is still more than are
			 * available, the user will need to change the number of
			 * queues/vectors used by the PF later with the ethtool
			 * channels command
			 */
			if (vmdq_vecs < vmdq_vecs_wanted)
				pf->num_vmdq_qps = 1;
			pf->num_vmdq_msix = pf->num_vmdq_qps;
S
Shannon Nelson 已提交
7710

7711 7712 7713
			v_budget += vmdq_vecs;
			vectors_left -= vmdq_vecs;
		}
S
Shannon Nelson 已提交
7714
	}
J
Jesse Brandeburg 已提交
7715 7716 7717 7718 7719 7720 7721 7722

	pf->msix_entries = kcalloc(v_budget, sizeof(struct msix_entry),
				   GFP_KERNEL);
	if (!pf->msix_entries)
		return -ENOMEM;

	for (i = 0; i < v_budget; i++)
		pf->msix_entries[i].entry = i;
7723
	v_actual = i40e_reserve_msix_vectors(pf, v_budget);
7724

7725
	if (v_actual < I40E_MIN_MSIX) {
J
Jesse Brandeburg 已提交
7726 7727 7728
		pf->flags &= ~I40E_FLAG_MSIX_ENABLED;
		kfree(pf->msix_entries);
		pf->msix_entries = NULL;
7729
		pci_disable_msix(pf->pdev);
J
Jesse Brandeburg 已提交
7730 7731
		return -ENODEV;

7732
	} else if (v_actual == I40E_MIN_MSIX) {
J
Jesse Brandeburg 已提交
7733 7734 7735 7736 7737 7738
		/* Adjust for minimal MSIX use */
		pf->num_vmdq_vsis = 0;
		pf->num_vmdq_qps = 0;
		pf->num_lan_qps = 1;
		pf->num_lan_msix = 1;

7739 7740 7741 7742 7743 7744
	} else if (!vectors_left) {
		/* If we have limited resources, we will start with no vectors
		 * for the special features and then allocate vectors to some
		 * of these features based on the policy and at the end disable
		 * the features that did not get any vectors.
		 */
7745 7746
		int vec;

7747 7748
		dev_info(&pf->pdev->dev,
			 "MSI-X vector limit reached, attempting to redistribute vectors\n");
7749
		/* reserve the misc vector */
7750
		vec = v_actual - 1;
7751

J
Jesse Brandeburg 已提交
7752 7753
		/* Scale vector usage down */
		pf->num_vmdq_msix = 1;    /* force VMDqs to only one vector */
7754
		pf->num_vmdq_vsis = 1;
S
Shannon Nelson 已提交
7755
		pf->num_vmdq_qps = 1;
7756 7757 7758 7759
#ifdef I40E_FCOE
		pf->num_fcoe_qps = 0;
		pf->num_fcoe_msix = 0;
#endif
J
Jesse Brandeburg 已提交
7760 7761 7762 7763 7764 7765 7766

		/* partition out the remaining vectors */
		switch (vec) {
		case 2:
			pf->num_lan_msix = 1;
			break;
		case 3:
7767 7768 7769 7770 7771 7772
			if (pf->flags & I40E_FLAG_IWARP_ENABLED) {
				pf->num_lan_msix = 1;
				pf->num_iwarp_msix = 1;
			} else {
				pf->num_lan_msix = 2;
			}
7773 7774 7775 7776 7777 7778 7779
#ifdef I40E_FCOE
			/* give one vector to FCoE */
			if (pf->flags & I40E_FLAG_FCOE_ENABLED) {
				pf->num_lan_msix = 1;
				pf->num_fcoe_msix = 1;
			}
#endif
J
Jesse Brandeburg 已提交
7780 7781
			break;
		default:
7782 7783 7784 7785 7786 7787 7788 7789 7790
			if (pf->flags & I40E_FLAG_IWARP_ENABLED) {
				pf->num_iwarp_msix = min_t(int, (vec / 3),
						 iwarp_requested);
				pf->num_vmdq_vsis = min_t(int, (vec / 3),
						  I40E_DEFAULT_NUM_VMDQ_VSI);
			} else {
				pf->num_vmdq_vsis = min_t(int, (vec / 2),
						  I40E_DEFAULT_NUM_VMDQ_VSI);
			}
7791 7792 7793 7794
			if (pf->flags & I40E_FLAG_FD_SB_ENABLED) {
				pf->num_fdsb_msix = 1;
				vec--;
			}
7795 7796 7797
			pf->num_lan_msix = min_t(int,
			       (vec - (pf->num_iwarp_msix + pf->num_vmdq_vsis)),
							      pf->num_lan_msix);
7798
			pf->num_lan_qps = pf->num_lan_msix;
7799 7800 7801 7802 7803 7804 7805
#ifdef I40E_FCOE
			/* give one vector to FCoE */
			if (pf->flags & I40E_FLAG_FCOE_ENABLED) {
				pf->num_fcoe_msix = 1;
				vec--;
			}
#endif
J
Jesse Brandeburg 已提交
7806 7807 7808 7809
			break;
		}
	}

7810 7811 7812 7813 7814
	if ((pf->flags & I40E_FLAG_FD_SB_ENABLED) &&
	    (pf->num_fdsb_msix == 0)) {
		dev_info(&pf->pdev->dev, "Sideband Flowdir disabled, not enough MSI-X vectors\n");
		pf->flags &= ~I40E_FLAG_FD_SB_ENABLED;
	}
7815 7816 7817 7818 7819
	if ((pf->flags & I40E_FLAG_VMDQ_ENABLED) &&
	    (pf->num_vmdq_msix == 0)) {
		dev_info(&pf->pdev->dev, "VMDq disabled, not enough MSI-X vectors\n");
		pf->flags &= ~I40E_FLAG_VMDQ_ENABLED;
	}
7820 7821 7822 7823 7824 7825

	if ((pf->flags & I40E_FLAG_IWARP_ENABLED) &&
	    (pf->num_iwarp_msix == 0)) {
		dev_info(&pf->pdev->dev, "IWARP disabled, not enough MSI-X vectors\n");
		pf->flags &= ~I40E_FLAG_IWARP_ENABLED;
	}
7826 7827 7828 7829 7830 7831 7832
#ifdef I40E_FCOE

	if ((pf->flags & I40E_FLAG_FCOE_ENABLED) && (pf->num_fcoe_msix == 0)) {
		dev_info(&pf->pdev->dev, "FCOE disabled, not enough MSI-X vectors\n");
		pf->flags &= ~I40E_FLAG_FCOE_ENABLED;
	}
#endif
7833 7834 7835 7836 7837 7838 7839
	i40e_debug(&pf->hw, I40E_DEBUG_INIT,
		   "MSI-X vector distribution: PF %d, VMDq %d, FDSB %d, iWARP %d\n",
		   pf->num_lan_msix,
		   pf->num_vmdq_msix * pf->num_vmdq_vsis,
		   pf->num_fdsb_msix,
		   pf->num_iwarp_msix);

7840
	return v_actual;
J
Jesse Brandeburg 已提交
7841 7842
}

7843
/**
7844
 * i40e_vsi_alloc_q_vector - Allocate memory for a single interrupt vector
7845 7846
 * @vsi: the VSI being configured
 * @v_idx: index of the vector in the vsi struct
7847
 * @cpu: cpu to be used on affinity_mask
7848 7849 7850
 *
 * We allocate one q_vector.  If allocation fails we return -ENOMEM.
 **/
7851
static int i40e_vsi_alloc_q_vector(struct i40e_vsi *vsi, int v_idx, int cpu)
7852 7853 7854 7855 7856 7857 7858 7859 7860 7861
{
	struct i40e_q_vector *q_vector;

	/* allocate q_vector */
	q_vector = kzalloc(sizeof(struct i40e_q_vector), GFP_KERNEL);
	if (!q_vector)
		return -ENOMEM;

	q_vector->vsi = vsi;
	q_vector->v_idx = v_idx;
7862 7863
	cpumask_set_cpu(cpu, &q_vector->affinity_mask);

7864 7865
	if (vsi->netdev)
		netif_napi_add(vsi->netdev, &q_vector->napi,
J
Jesse Brandeburg 已提交
7866
			       i40e_napi_poll, NAPI_POLL_WEIGHT);
7867

7868 7869 7870
	q_vector->rx.latency_range = I40E_LOW_LATENCY;
	q_vector->tx.latency_range = I40E_LOW_LATENCY;

7871 7872 7873 7874 7875 7876
	/* tie q_vector and vsi together */
	vsi->q_vectors[v_idx] = q_vector;

	return 0;
}

J
Jesse Brandeburg 已提交
7877
/**
7878
 * i40e_vsi_alloc_q_vectors - Allocate memory for interrupt vectors
J
Jesse Brandeburg 已提交
7879 7880 7881 7882 7883
 * @vsi: the VSI being configured
 *
 * We allocate one q_vector per queue interrupt.  If allocation fails we
 * return -ENOMEM.
 **/
7884
static int i40e_vsi_alloc_q_vectors(struct i40e_vsi *vsi)
J
Jesse Brandeburg 已提交
7885 7886
{
	struct i40e_pf *pf = vsi->back;
7887
	int err, v_idx, num_q_vectors, current_cpu;
J
Jesse Brandeburg 已提交
7888 7889 7890 7891 7892 7893 7894 7895 7896

	/* if not MSIX, give the one vector only to the LAN VSI */
	if (pf->flags & I40E_FLAG_MSIX_ENABLED)
		num_q_vectors = vsi->num_q_vectors;
	else if (vsi == pf->vsi[pf->lan_vsi])
		num_q_vectors = 1;
	else
		return -EINVAL;

7897 7898
	current_cpu = cpumask_first(cpu_online_mask);

J
Jesse Brandeburg 已提交
7899
	for (v_idx = 0; v_idx < num_q_vectors; v_idx++) {
7900
		err = i40e_vsi_alloc_q_vector(vsi, v_idx, current_cpu);
7901 7902
		if (err)
			goto err_out;
7903 7904 7905
		current_cpu = cpumask_next(current_cpu, cpu_online_mask);
		if (unlikely(current_cpu >= nr_cpu_ids))
			current_cpu = cpumask_first(cpu_online_mask);
J
Jesse Brandeburg 已提交
7906 7907 7908
	}

	return 0;
7909 7910 7911 7912 7913 7914

err_out:
	while (v_idx--)
		i40e_free_q_vector(vsi, v_idx);

	return err;
J
Jesse Brandeburg 已提交
7915 7916 7917 7918 7919 7920
}

/**
 * i40e_init_interrupt_scheme - Determine proper interrupt scheme
 * @pf: board private structure to initialize
 **/
7921
static int i40e_init_interrupt_scheme(struct i40e_pf *pf)
J
Jesse Brandeburg 已提交
7922
{
7923 7924
	int vectors = 0;
	ssize_t size;
J
Jesse Brandeburg 已提交
7925 7926

	if (pf->flags & I40E_FLAG_MSIX_ENABLED) {
7927 7928
		vectors = i40e_init_msix(pf);
		if (vectors < 0) {
J
Jesse Brandeburg 已提交
7929
			pf->flags &= ~(I40E_FLAG_MSIX_ENABLED	|
7930
				       I40E_FLAG_IWARP_ENABLED	|
7931 7932 7933
#ifdef I40E_FCOE
				       I40E_FLAG_FCOE_ENABLED	|
#endif
J
Jesse Brandeburg 已提交
7934
				       I40E_FLAG_RSS_ENABLED	|
7935
				       I40E_FLAG_DCB_CAPABLE	|
7936
				       I40E_FLAG_DCB_ENABLED	|
J
Jesse Brandeburg 已提交
7937 7938 7939 7940
				       I40E_FLAG_SRIOV_ENABLED	|
				       I40E_FLAG_FD_SB_ENABLED	|
				       I40E_FLAG_FD_ATR_ENABLED	|
				       I40E_FLAG_VMDQ_ENABLED);
J
Jesse Brandeburg 已提交
7941 7942 7943 7944 7945 7946 7947 7948

			/* rework the queue expectations without MSIX */
			i40e_determine_queue_usage(pf);
		}
	}

	if (!(pf->flags & I40E_FLAG_MSIX_ENABLED) &&
	    (pf->flags & I40E_FLAG_MSI_ENABLED)) {
C
Catherine Sullivan 已提交
7949
		dev_info(&pf->pdev->dev, "MSI-X not available, trying MSI\n");
7950 7951 7952 7953
		vectors = pci_enable_msi(pf->pdev);
		if (vectors < 0) {
			dev_info(&pf->pdev->dev, "MSI init failed - %d\n",
				 vectors);
J
Jesse Brandeburg 已提交
7954 7955
			pf->flags &= ~I40E_FLAG_MSI_ENABLED;
		}
7956
		vectors = 1;  /* one MSI or Legacy vector */
J
Jesse Brandeburg 已提交
7957 7958
	}

7959
	if (!(pf->flags & (I40E_FLAG_MSIX_ENABLED | I40E_FLAG_MSI_ENABLED)))
C
Catherine Sullivan 已提交
7960
		dev_info(&pf->pdev->dev, "MSI-X and MSI not available, falling back to Legacy IRQ\n");
7961

7962 7963 7964
	/* set up vector assignment tracking */
	size = sizeof(struct i40e_lump_tracking) + (sizeof(u16) * vectors);
	pf->irq_pile = kzalloc(size, GFP_KERNEL);
7965 7966 7967 7968
	if (!pf->irq_pile) {
		dev_err(&pf->pdev->dev, "error allocating irq_pile memory\n");
		return -ENOMEM;
	}
7969 7970 7971
	pf->irq_pile->num_entries = vectors;
	pf->irq_pile->search_hint = 0;

7972
	/* track first vector for misc interrupts, ignore return */
7973
	(void)i40e_get_lump(pf, pf->irq_pile, 1, I40E_PILE_VALID_BIT - 1);
7974 7975

	return 0;
J
Jesse Brandeburg 已提交
7976 7977 7978 7979 7980 7981 7982 7983 7984 7985 7986 7987 7988 7989 7990 7991 7992 7993 7994 7995
}

/**
 * i40e_setup_misc_vector - Setup the misc vector to handle non queue events
 * @pf: board private structure
 *
 * This sets up the handler for MSIX 0, which is used to manage the
 * non-queue interrupts, e.g. AdminQ and errors.  This is not used
 * when in MSI or Legacy interrupt mode.
 **/
static int i40e_setup_misc_vector(struct i40e_pf *pf)
{
	struct i40e_hw *hw = &pf->hw;
	int err = 0;

	/* Only request the irq if this is the first time through, and
	 * not when we're rebuilding after a Reset
	 */
	if (!test_bit(__I40E_RESET_RECOVERY_PENDING, &pf->state)) {
		err = request_irq(pf->msix_entries[0].vector,
7996
				  i40e_intr, 0, pf->int_name, pf);
J
Jesse Brandeburg 已提交
7997 7998
		if (err) {
			dev_info(&pf->pdev->dev,
C
Catherine Sullivan 已提交
7999
				 "request_irq for %s failed: %d\n",
8000
				 pf->int_name, err);
J
Jesse Brandeburg 已提交
8001 8002 8003 8004
			return -EFAULT;
		}
	}

8005
	i40e_enable_misc_int_causes(pf);
J
Jesse Brandeburg 已提交
8006 8007 8008 8009 8010 8011 8012

	/* associate no queues to the misc vector */
	wr32(hw, I40E_PFINT_LNKLST0, I40E_QUEUE_END_OF_LIST);
	wr32(hw, I40E_PFINT_ITR0(I40E_RX_ITR), I40E_ITR_8K);

	i40e_flush(hw);

8013
	i40e_irq_dynamic_enable_icr0(pf, true);
J
Jesse Brandeburg 已提交
8014 8015 8016 8017 8018

	return err;
}

/**
8019 8020 8021 8022
 * i40e_config_rss_aq - Prepare for RSS using AQ commands
 * @vsi: vsi structure
 * @seed: RSS hash seed
 **/
8023 8024
static int i40e_config_rss_aq(struct i40e_vsi *vsi, const u8 *seed,
			      u8 *lut, u16 lut_size)
8025 8026 8027
{
	struct i40e_pf *pf = vsi->back;
	struct i40e_hw *hw = &pf->hw;
8028
	int ret = 0;
8029

8030 8031 8032 8033 8034 8035 8036 8037 8038 8039 8040
	if (seed) {
		struct i40e_aqc_get_set_rss_key_data *seed_dw =
			(struct i40e_aqc_get_set_rss_key_data *)seed;
		ret = i40e_aq_set_rss_key(hw, vsi->id, seed_dw);
		if (ret) {
			dev_info(&pf->pdev->dev,
				 "Cannot set RSS key, err %s aq_err %s\n",
				 i40e_stat_str(hw, ret),
				 i40e_aq_str(hw, hw->aq.asq_last_status));
			return ret;
		}
8041
	}
8042 8043
	if (lut) {
		bool pf_lut = vsi->type == I40E_VSI_MAIN ? true : false;
8044

8045 8046 8047 8048 8049 8050 8051 8052 8053
		ret = i40e_aq_set_rss_lut(hw, vsi->id, pf_lut, lut, lut_size);
		if (ret) {
			dev_info(&pf->pdev->dev,
				 "Cannot set RSS lut, err %s aq_err %s\n",
				 i40e_stat_str(hw, ret),
				 i40e_aq_str(hw, hw->aq.asq_last_status));
			return ret;
		}
	}
8054 8055 8056
	return ret;
}

8057 8058 8059 8060 8061 8062 8063 8064 8065 8066 8067 8068 8069 8070 8071 8072 8073 8074 8075 8076 8077 8078 8079 8080 8081 8082 8083 8084 8085 8086 8087 8088 8089 8090 8091 8092 8093 8094 8095 8096 8097 8098 8099 8100 8101 8102
/**
 * i40e_get_rss_aq - Get RSS keys and lut by using AQ commands
 * @vsi: Pointer to vsi structure
 * @seed: Buffter to store the hash keys
 * @lut: Buffer to store the lookup table entries
 * @lut_size: Size of buffer to store the lookup table entries
 *
 * Return 0 on success, negative on failure
 */
static int i40e_get_rss_aq(struct i40e_vsi *vsi, const u8 *seed,
			   u8 *lut, u16 lut_size)
{
	struct i40e_pf *pf = vsi->back;
	struct i40e_hw *hw = &pf->hw;
	int ret = 0;

	if (seed) {
		ret = i40e_aq_get_rss_key(hw, vsi->id,
			(struct i40e_aqc_get_set_rss_key_data *)seed);
		if (ret) {
			dev_info(&pf->pdev->dev,
				 "Cannot get RSS key, err %s aq_err %s\n",
				 i40e_stat_str(&pf->hw, ret),
				 i40e_aq_str(&pf->hw,
					     pf->hw.aq.asq_last_status));
			return ret;
		}
	}

	if (lut) {
		bool pf_lut = vsi->type == I40E_VSI_MAIN ? true : false;

		ret = i40e_aq_get_rss_lut(hw, vsi->id, pf_lut, lut, lut_size);
		if (ret) {
			dev_info(&pf->pdev->dev,
				 "Cannot get RSS lut, err %s aq_err %s\n",
				 i40e_stat_str(&pf->hw, ret),
				 i40e_aq_str(&pf->hw,
					     pf->hw.aq.asq_last_status));
			return ret;
		}
	}

	return ret;
}

8103 8104 8105 8106 8107 8108 8109 8110 8111 8112 8113 8114 8115 8116
/**
 * i40e_vsi_config_rss - Prepare for VSI(VMDq) RSS if used
 * @vsi: VSI structure
 **/
static int i40e_vsi_config_rss(struct i40e_vsi *vsi)
{
	u8 seed[I40E_HKEY_ARRAY_SIZE];
	struct i40e_pf *pf = vsi->back;
	u8 *lut;
	int ret;

	if (!(pf->flags & I40E_FLAG_RSS_AQ_CAPABLE))
		return 0;

8117 8118 8119 8120 8121 8122
	if (!vsi->rss_size)
		vsi->rss_size = min_t(int, pf->alloc_rss_size,
				      vsi->num_queue_pairs);
	if (!vsi->rss_size)
		return -EINVAL;

8123 8124 8125
	lut = kzalloc(vsi->rss_table_size, GFP_KERNEL);
	if (!lut)
		return -ENOMEM;
8126 8127 8128 8129 8130 8131 8132 8133 8134 8135 8136
	/* Use the user configured hash keys and lookup table if there is one,
	 * otherwise use default
	 */
	if (vsi->rss_lut_user)
		memcpy(lut, vsi->rss_lut_user, vsi->rss_table_size);
	else
		i40e_fill_rss_lut(pf, lut, vsi->rss_table_size, vsi->rss_size);
	if (vsi->rss_hkey_user)
		memcpy(seed, vsi->rss_hkey_user, I40E_HKEY_ARRAY_SIZE);
	else
		netdev_rss_key_fill((void *)seed, I40E_HKEY_ARRAY_SIZE);
8137 8138 8139 8140 8141 8142
	ret = i40e_config_rss_aq(vsi, seed, lut, vsi->rss_table_size);
	kfree(lut);

	return ret;
}

8143
/**
8144
 * i40e_config_rss_reg - Configure RSS keys and lut by writing registers
8145
 * @vsi: Pointer to vsi structure
8146
 * @seed: RSS hash seed
8147 8148 8149 8150
 * @lut: Lookup table
 * @lut_size: Lookup table size
 *
 * Returns 0 on success, negative on failure
J
Jesse Brandeburg 已提交
8151
 **/
8152 8153
static int i40e_config_rss_reg(struct i40e_vsi *vsi, const u8 *seed,
			       const u8 *lut, u16 lut_size)
J
Jesse Brandeburg 已提交
8154
{
8155
	struct i40e_pf *pf = vsi->back;
8156
	struct i40e_hw *hw = &pf->hw;
8157
	u16 vf_id = vsi->vf_id;
8158
	u8 i;
J
Jesse Brandeburg 已提交
8159

8160
	/* Fill out hash function seed */
8161 8162 8163
	if (seed) {
		u32 *seed_dw = (u32 *)seed;

8164 8165 8166 8167 8168 8169 8170 8171 8172 8173 8174 8175
		if (vsi->type == I40E_VSI_MAIN) {
			for (i = 0; i <= I40E_PFQF_HKEY_MAX_INDEX; i++)
				i40e_write_rx_ctl(hw, I40E_PFQF_HKEY(i),
						  seed_dw[i]);
		} else if (vsi->type == I40E_VSI_SRIOV) {
			for (i = 0; i <= I40E_VFQF_HKEY1_MAX_INDEX; i++)
				i40e_write_rx_ctl(hw,
						  I40E_VFQF_HKEY1(i, vf_id),
						  seed_dw[i]);
		} else {
			dev_err(&pf->pdev->dev, "Cannot set RSS seed - invalid VSI type\n");
		}
8176 8177 8178 8179 8180
	}

	if (lut) {
		u32 *lut_dw = (u32 *)lut;

8181 8182 8183 8184 8185 8186 8187 8188 8189 8190 8191 8192 8193 8194 8195
		if (vsi->type == I40E_VSI_MAIN) {
			if (lut_size != I40E_HLUT_ARRAY_SIZE)
				return -EINVAL;
			for (i = 0; i <= I40E_PFQF_HLUT_MAX_INDEX; i++)
				wr32(hw, I40E_PFQF_HLUT(i), lut_dw[i]);
		} else if (vsi->type == I40E_VSI_SRIOV) {
			if (lut_size != I40E_VF_HLUT_ARRAY_SIZE)
				return -EINVAL;
			for (i = 0; i <= I40E_VFQF_HLUT_MAX_INDEX; i++)
				i40e_write_rx_ctl(hw,
						  I40E_VFQF_HLUT1(i, vf_id),
						  lut_dw[i]);
		} else {
			dev_err(&pf->pdev->dev, "Cannot set RSS LUT - invalid VSI type\n");
		}
8196 8197 8198 8199 8200 8201
	}
	i40e_flush(hw);

	return 0;
}

8202 8203 8204 8205 8206 8207 8208 8209 8210 8211 8212 8213 8214 8215 8216 8217 8218 8219 8220 8221
/**
 * i40e_get_rss_reg - Get the RSS keys and lut by reading registers
 * @vsi: Pointer to VSI structure
 * @seed: Buffer to store the keys
 * @lut: Buffer to store the lookup table entries
 * @lut_size: Size of buffer to store the lookup table entries
 *
 * Returns 0 on success, negative on failure
 */
static int i40e_get_rss_reg(struct i40e_vsi *vsi, u8 *seed,
			    u8 *lut, u16 lut_size)
{
	struct i40e_pf *pf = vsi->back;
	struct i40e_hw *hw = &pf->hw;
	u16 i;

	if (seed) {
		u32 *seed_dw = (u32 *)seed;

		for (i = 0; i <= I40E_PFQF_HKEY_MAX_INDEX; i++)
8222
			seed_dw[i] = i40e_read_rx_ctl(hw, I40E_PFQF_HKEY(i));
8223 8224 8225 8226 8227 8228 8229 8230 8231 8232 8233 8234 8235 8236 8237 8238 8239 8240 8241 8242 8243 8244 8245 8246 8247 8248 8249 8250 8251 8252 8253 8254 8255 8256 8257 8258 8259 8260 8261 8262 8263 8264 8265
	}
	if (lut) {
		u32 *lut_dw = (u32 *)lut;

		if (lut_size != I40E_HLUT_ARRAY_SIZE)
			return -EINVAL;
		for (i = 0; i <= I40E_PFQF_HLUT_MAX_INDEX; i++)
			lut_dw[i] = rd32(hw, I40E_PFQF_HLUT(i));
	}

	return 0;
}

/**
 * i40e_config_rss - Configure RSS keys and lut
 * @vsi: Pointer to VSI structure
 * @seed: RSS hash seed
 * @lut: Lookup table
 * @lut_size: Lookup table size
 *
 * Returns 0 on success, negative on failure
 */
int i40e_config_rss(struct i40e_vsi *vsi, u8 *seed, u8 *lut, u16 lut_size)
{
	struct i40e_pf *pf = vsi->back;

	if (pf->flags & I40E_FLAG_RSS_AQ_CAPABLE)
		return i40e_config_rss_aq(vsi, seed, lut, lut_size);
	else
		return i40e_config_rss_reg(vsi, seed, lut, lut_size);
}

/**
 * i40e_get_rss - Get RSS keys and lut
 * @vsi: Pointer to VSI structure
 * @seed: Buffer to store the keys
 * @lut: Buffer to store the lookup table entries
 * lut_size: Size of buffer to store the lookup table entries
 *
 * Returns 0 on success, negative on failure
 */
int i40e_get_rss(struct i40e_vsi *vsi, u8 *seed, u8 *lut, u16 lut_size)
{
8266 8267 8268 8269 8270 8271
	struct i40e_pf *pf = vsi->back;

	if (pf->flags & I40E_FLAG_RSS_AQ_CAPABLE)
		return i40e_get_rss_aq(vsi, seed, lut, lut_size);
	else
		return i40e_get_rss_reg(vsi, seed, lut, lut_size);
8272 8273
}

8274 8275 8276 8277 8278 8279 8280
/**
 * i40e_fill_rss_lut - Fill the RSS lookup table with default values
 * @pf: Pointer to board private structure
 * @lut: Lookup table
 * @rss_table_size: Lookup table size
 * @rss_size: Range of queue number for hashing
 */
8281 8282
void i40e_fill_rss_lut(struct i40e_pf *pf, u8 *lut,
		       u16 rss_table_size, u16 rss_size)
8283 8284 8285 8286 8287 8288 8289
{
	u16 i;

	for (i = 0; i < rss_table_size; i++)
		lut[i] = i % rss_size;
}

8290
/**
8291
 * i40e_pf_config_rss - Prepare for RSS if used
8292 8293
 * @pf: board private structure
 **/
8294
static int i40e_pf_config_rss(struct i40e_pf *pf)
8295 8296 8297
{
	struct i40e_vsi *vsi = pf->vsi[pf->lan_vsi];
	u8 seed[I40E_HKEY_ARRAY_SIZE];
8298
	u8 *lut;
8299 8300 8301
	struct i40e_hw *hw = &pf->hw;
	u32 reg_val;
	u64 hena;
8302
	int ret;
8303

J
Jesse Brandeburg 已提交
8304
	/* By default we enable TCP/UDP with IPv4/IPv6 ptypes */
8305 8306
	hena = (u64)i40e_read_rx_ctl(hw, I40E_PFQF_HENA(0)) |
		((u64)i40e_read_rx_ctl(hw, I40E_PFQF_HENA(1)) << 32);
8307 8308
	hena |= i40e_pf_get_default_rss_hena(pf);

8309 8310
	i40e_write_rx_ctl(hw, I40E_PFQF_HENA(0), (u32)hena);
	i40e_write_rx_ctl(hw, I40E_PFQF_HENA(1), (u32)(hena >> 32));
J
Jesse Brandeburg 已提交
8311

8312
	/* Determine the RSS table size based on the hardware capabilities */
8313
	reg_val = i40e_read_rx_ctl(hw, I40E_PFQF_CTL_0);
8314 8315 8316
	reg_val = (pf->rss_table_size == 512) ?
			(reg_val | I40E_PFQF_CTL_0_HASHLUTSIZE_512) :
			(reg_val & ~I40E_PFQF_CTL_0_HASHLUTSIZE_512);
8317
	i40e_write_rx_ctl(hw, I40E_PFQF_CTL_0, reg_val);
8318

8319 8320
	/* Determine the RSS size of the VSI */
	if (!vsi->rss_size)
8321 8322
		vsi->rss_size = min_t(int, pf->alloc_rss_size,
				      vsi->num_queue_pairs);
8323 8324
	if (!vsi->rss_size)
		return -EINVAL;
8325

8326 8327 8328 8329
	lut = kzalloc(vsi->rss_table_size, GFP_KERNEL);
	if (!lut)
		return -ENOMEM;

8330 8331 8332 8333 8334
	/* Use user configured lut if there is one, otherwise use default */
	if (vsi->rss_lut_user)
		memcpy(lut, vsi->rss_lut_user, vsi->rss_table_size);
	else
		i40e_fill_rss_lut(pf, lut, vsi->rss_table_size, vsi->rss_size);
8335

8336 8337 8338 8339 8340 8341 8342
	/* Use user configured hash key if there is one, otherwise
	 * use default.
	 */
	if (vsi->rss_hkey_user)
		memcpy(seed, vsi->rss_hkey_user, I40E_HKEY_ARRAY_SIZE);
	else
		netdev_rss_key_fill((void *)seed, I40E_HKEY_ARRAY_SIZE);
8343
	ret = i40e_config_rss(vsi, seed, lut, vsi->rss_table_size);
8344 8345 8346
	kfree(lut);

	return ret;
J
Jesse Brandeburg 已提交
8347 8348
}

8349 8350 8351 8352 8353 8354 8355 8356 8357 8358
/**
 * i40e_reconfig_rss_queues - change number of queues for rss and rebuild
 * @pf: board private structure
 * @queue_count: the requested queue count for rss.
 *
 * returns 0 if rss is not enabled, if enabled returns the final rss queue
 * count which may be different from the requested queue count.
 **/
int i40e_reconfig_rss_queues(struct i40e_pf *pf, int queue_count)
{
8359 8360 8361
	struct i40e_vsi *vsi = pf->vsi[pf->lan_vsi];
	int new_rss_size;

8362 8363 8364
	if (!(pf->flags & I40E_FLAG_RSS_ENABLED))
		return 0;

8365
	new_rss_size = min_t(int, queue_count, pf->rss_size_max);
8366

8367 8368
	if (queue_count != vsi->num_queue_pairs) {
		vsi->req_queue_pairs = queue_count;
8369 8370
		i40e_prep_for_reset(pf);

8371
		pf->alloc_rss_size = new_rss_size;
8372 8373

		i40e_reset_and_rebuild(pf, true);
8374 8375 8376 8377 8378 8379 8380 8381 8382 8383 8384

		/* Discard the user configured hash keys and lut, if less
		 * queues are enabled.
		 */
		if (queue_count < vsi->rss_size) {
			i40e_clear_rss_config_user(vsi);
			dev_dbg(&pf->pdev->dev,
				"discard user configured hash keys and lut\n");
		}

		/* Reset vsi->rss_size, as number of enabled queues changed */
8385 8386
		vsi->rss_size = min_t(int, pf->alloc_rss_size,
				      vsi->num_queue_pairs);
8387

8388
		i40e_pf_config_rss(pf);
8389
	}
8390 8391
	dev_info(&pf->pdev->dev, "User requested queue count/HW max RSS count:  %d/%d\n",
		 vsi->req_queue_pairs, pf->rss_size_max);
8392
	return pf->alloc_rss_size;
8393 8394
}

8395 8396 8397 8398 8399 8400 8401 8402 8403 8404 8405 8406 8407 8408 8409 8410 8411 8412 8413 8414 8415 8416 8417 8418 8419 8420 8421 8422 8423 8424 8425 8426
/**
 * i40e_get_npar_bw_setting - Retrieve BW settings for this PF partition
 * @pf: board private structure
 **/
i40e_status i40e_get_npar_bw_setting(struct i40e_pf *pf)
{
	i40e_status status;
	bool min_valid, max_valid;
	u32 max_bw, min_bw;

	status = i40e_read_bw_from_alt_ram(&pf->hw, &max_bw, &min_bw,
					   &min_valid, &max_valid);

	if (!status) {
		if (min_valid)
			pf->npar_min_bw = min_bw;
		if (max_valid)
			pf->npar_max_bw = max_bw;
	}

	return status;
}

/**
 * i40e_set_npar_bw_setting - Set BW settings for this PF partition
 * @pf: board private structure
 **/
i40e_status i40e_set_npar_bw_setting(struct i40e_pf *pf)
{
	struct i40e_aqc_configure_partition_bw_data bw_data;
	i40e_status status;

8427
	/* Set the valid bit for this PF */
8428
	bw_data.pf_valid_bits = cpu_to_le16(BIT(pf->hw.pf_id));
8429 8430 8431 8432 8433 8434 8435 8436 8437 8438 8439 8440 8441 8442 8443 8444 8445 8446 8447 8448 8449 8450 8451 8452 8453 8454 8455 8456 8457 8458 8459 8460 8461
	bw_data.max_bw[pf->hw.pf_id] = pf->npar_max_bw & I40E_ALT_BW_VALUE_MASK;
	bw_data.min_bw[pf->hw.pf_id] = pf->npar_min_bw & I40E_ALT_BW_VALUE_MASK;

	/* Set the new bandwidths */
	status = i40e_aq_configure_partition_bw(&pf->hw, &bw_data, NULL);

	return status;
}

/**
 * i40e_commit_npar_bw_setting - Commit BW settings for this PF partition
 * @pf: board private structure
 **/
i40e_status i40e_commit_npar_bw_setting(struct i40e_pf *pf)
{
	/* Commit temporary BW setting to permanent NVM image */
	enum i40e_admin_queue_err last_aq_status;
	i40e_status ret;
	u16 nvm_word;

	if (pf->hw.partition_id != 1) {
		dev_info(&pf->pdev->dev,
			 "Commit BW only works on partition 1! This is partition %d",
			 pf->hw.partition_id);
		ret = I40E_NOT_SUPPORTED;
		goto bw_commit_out;
	}

	/* Acquire NVM for read access */
	ret = i40e_acquire_nvm(&pf->hw, I40E_RESOURCE_READ);
	last_aq_status = pf->hw.aq.asq_last_status;
	if (ret) {
		dev_info(&pf->pdev->dev,
8462 8463 8464
			 "Cannot acquire NVM for read access, err %s aq_err %s\n",
			 i40e_stat_str(&pf->hw, ret),
			 i40e_aq_str(&pf->hw, last_aq_status));
8465 8466 8467 8468 8469 8470 8471 8472 8473 8474 8475 8476 8477 8478
		goto bw_commit_out;
	}

	/* Read word 0x10 of NVM - SW compatibility word 1 */
	ret = i40e_aq_read_nvm(&pf->hw,
			       I40E_SR_NVM_CONTROL_WORD,
			       0x10, sizeof(nvm_word), &nvm_word,
			       false, NULL);
	/* Save off last admin queue command status before releasing
	 * the NVM
	 */
	last_aq_status = pf->hw.aq.asq_last_status;
	i40e_release_nvm(&pf->hw);
	if (ret) {
8479 8480 8481
		dev_info(&pf->pdev->dev, "NVM read error, err %s aq_err %s\n",
			 i40e_stat_str(&pf->hw, ret),
			 i40e_aq_str(&pf->hw, last_aq_status));
8482 8483 8484 8485 8486 8487 8488 8489 8490 8491 8492
		goto bw_commit_out;
	}

	/* Wait a bit for NVM release to complete */
	msleep(50);

	/* Acquire NVM for write access */
	ret = i40e_acquire_nvm(&pf->hw, I40E_RESOURCE_WRITE);
	last_aq_status = pf->hw.aq.asq_last_status;
	if (ret) {
		dev_info(&pf->pdev->dev,
8493 8494 8495
			 "Cannot acquire NVM for write access, err %s aq_err %s\n",
			 i40e_stat_str(&pf->hw, ret),
			 i40e_aq_str(&pf->hw, last_aq_status));
8496 8497 8498 8499 8500 8501 8502 8503 8504 8505 8506 8507 8508 8509 8510 8511 8512
		goto bw_commit_out;
	}
	/* Write it back out unchanged to initiate update NVM,
	 * which will force a write of the shadow (alt) RAM to
	 * the NVM - thus storing the bandwidth values permanently.
	 */
	ret = i40e_aq_update_nvm(&pf->hw,
				 I40E_SR_NVM_CONTROL_WORD,
				 0x10, sizeof(nvm_word),
				 &nvm_word, true, NULL);
	/* Save off last admin queue command status before releasing
	 * the NVM
	 */
	last_aq_status = pf->hw.aq.asq_last_status;
	i40e_release_nvm(&pf->hw);
	if (ret)
		dev_info(&pf->pdev->dev,
8513 8514 8515
			 "BW settings NOT SAVED, err %s aq_err %s\n",
			 i40e_stat_str(&pf->hw, ret),
			 i40e_aq_str(&pf->hw, last_aq_status));
8516 8517 8518 8519 8520
bw_commit_out:

	return ret;
}

J
Jesse Brandeburg 已提交
8521 8522 8523 8524 8525 8526 8527 8528 8529 8530 8531 8532 8533 8534 8535 8536
/**
 * i40e_sw_init - Initialize general software structures (struct i40e_pf)
 * @pf: board private structure to initialize
 *
 * i40e_sw_init initializes the Adapter private data structure.
 * Fields are initialized based on PCI device information and
 * OS network device settings (MTU size).
 **/
static int i40e_sw_init(struct i40e_pf *pf)
{
	int err = 0;
	int size;

	/* Set default capability flags */
	pf->flags = I40E_FLAG_RX_CSUM_ENABLED |
		    I40E_FLAG_MSI_ENABLED     |
8537 8538
		    I40E_FLAG_MSIX_ENABLED;

8539 8540 8541 8542
	/* Set default ITR */
	pf->rx_itr_default = I40E_ITR_DYNAMIC | I40E_ITR_RX_DEF;
	pf->tx_itr_default = I40E_ITR_DYNAMIC | I40E_ITR_TX_DEF;

8543 8544 8545
	/* Depending on PF configurations, it is possible that the RSS
	 * maximum might end up larger than the available queues
	 */
8546
	pf->rss_size_max = BIT(pf->hw.func_caps.rss_table_entry_width);
8547
	pf->alloc_rss_size = 1;
8548
	pf->rss_table_size = pf->hw.func_caps.rss_table_size;
8549 8550
	pf->rss_size_max = min_t(int, pf->rss_size_max,
				 pf->hw.func_caps.num_tx_qp);
J
Jesse Brandeburg 已提交
8551 8552
	if (pf->hw.func_caps.rss) {
		pf->flags |= I40E_FLAG_RSS_ENABLED;
8553 8554
		pf->alloc_rss_size = min_t(int, pf->rss_size_max,
					   num_online_cpus());
J
Jesse Brandeburg 已提交
8555 8556
	}

8557
	/* MFP mode enabled */
8558
	if (pf->hw.func_caps.npar_enable || pf->hw.func_caps.flex10_enable) {
8559 8560
		pf->flags |= I40E_FLAG_MFP_ENABLED;
		dev_info(&pf->pdev->dev, "MFP mode Enabled\n");
8561 8562 8563 8564 8565 8566 8567
		if (i40e_get_npar_bw_setting(pf))
			dev_warn(&pf->pdev->dev,
				 "Could not get NPAR bw settings\n");
		else
			dev_info(&pf->pdev->dev,
				 "Min BW = %8.8x, Max BW = %8.8x\n",
				 pf->npar_min_bw, pf->npar_max_bw);
8568 8569
	}

8570 8571 8572 8573 8574
	/* FW/NVM is not yet fixed in this regard */
	if ((pf->hw.func_caps.fd_filters_guaranteed > 0) ||
	    (pf->hw.func_caps.fd_filters_best_effort > 0)) {
		pf->flags |= I40E_FLAG_FD_ATR_ENABLED;
		pf->atr_sample_rate = I40E_DEFAULT_ATR_SAMPLE_RATE;
8575 8576
		if (pf->flags & I40E_FLAG_MFP_ENABLED &&
		    pf->hw.num_partitions > 1)
8577
			dev_info(&pf->pdev->dev,
A
Anjali Singhai Jain 已提交
8578
				 "Flow Director Sideband mode Disabled in MFP mode\n");
8579 8580
		else
			pf->flags |= I40E_FLAG_FD_SB_ENABLED;
8581 8582 8583 8584
		pf->fdir_pf_filter_count =
				 pf->hw.func_caps.fd_filters_guaranteed;
		pf->hw.fdir_shared_filter_count =
				 pf->hw.func_caps.fd_filters_best_effort;
J
Jesse Brandeburg 已提交
8585 8586
	}

8587
	if (i40e_is_mac_710(&pf->hw) &&
8588
	    (((pf->hw.aq.fw_maj_ver == 4) && (pf->hw.aq.fw_min_ver < 33)) ||
8589
	    (pf->hw.aq.fw_maj_ver < 4))) {
8590
		pf->flags |= I40E_FLAG_RESTART_AUTONEG;
8591 8592 8593 8594 8595 8596 8597 8598 8599 8600 8601 8602 8603 8604 8605
		/* No DCB support  for FW < v4.33 */
		pf->flags |= I40E_FLAG_NO_DCB_SUPPORT;
	}

	/* Disable FW LLDP if FW < v4.3 */
	if (i40e_is_mac_710(&pf->hw) &&
	    (((pf->hw.aq.fw_maj_ver == 4) && (pf->hw.aq.fw_min_ver < 3)) ||
	    (pf->hw.aq.fw_maj_ver < 4)))
		pf->flags |= I40E_FLAG_STOP_FW_LLDP;

	/* Use the FW Set LLDP MIB API if FW > v4.40 */
	if (i40e_is_mac_710(&pf->hw) &&
	    (((pf->hw.aq.fw_maj_ver == 4) && (pf->hw.aq.fw_min_ver >= 40)) ||
	    (pf->hw.aq.fw_maj_ver >= 5)))
		pf->flags |= I40E_FLAG_USE_SET_LLDP_MIB;
8606

J
Jesse Brandeburg 已提交
8607 8608
	if (pf->hw.func_caps.vmdq) {
		pf->num_vmdq_vsis = I40E_DEFAULT_NUM_VMDQ_VSI;
8609
		pf->flags |= I40E_FLAG_VMDQ_ENABLED;
J
Jesse Brandeburg 已提交
8610
		pf->num_vmdq_qps = i40e_default_queues_per_vmdq(pf);
J
Jesse Brandeburg 已提交
8611 8612
	}

8613 8614 8615 8616 8617 8618
	if (pf->hw.func_caps.iwarp) {
		pf->flags |= I40E_FLAG_IWARP_ENABLED;
		/* IWARP needs one extra vector for CQP just like MISC.*/
		pf->num_iwarp_msix = (int)num_online_cpus() + 1;
	}

8619
#ifdef I40E_FCOE
8620
	i40e_init_pf_fcoe(pf);
8621 8622

#endif /* I40E_FCOE */
J
Jesse Brandeburg 已提交
8623
#ifdef CONFIG_PCI_IOV
8624
	if (pf->hw.func_caps.num_vfs && pf->hw.partition_id == 1) {
J
Jesse Brandeburg 已提交
8625 8626 8627 8628 8629 8630 8631
		pf->num_vf_qps = I40E_DEFAULT_QUEUES_PER_VF;
		pf->flags |= I40E_FLAG_SRIOV_ENABLED;
		pf->num_req_vfs = min_t(int,
					pf->hw.func_caps.num_vfs,
					I40E_MAX_VF_COUNT);
	}
#endif /* CONFIG_PCI_IOV */
8632 8633 8634 8635 8636 8637
	if (pf->hw.mac.type == I40E_MAC_X722) {
		pf->flags |= I40E_FLAG_RSS_AQ_CAPABLE |
			     I40E_FLAG_128_QP_RSS_CAPABLE |
			     I40E_FLAG_HW_ATR_EVICT_CAPABLE |
			     I40E_FLAG_OUTER_UDP_CSUM_CAPABLE |
			     I40E_FLAG_WB_ON_ITR_CAPABLE |
8638
			     I40E_FLAG_MULTIPLE_TCP_UDP_RSS_PCTYPE |
8639
			     I40E_FLAG_NO_PCI_LINK_CHECK |
8640
			     I40E_FLAG_USE_SET_LLDP_MIB |
8641
			     I40E_FLAG_GENEVE_OFFLOAD_CAPABLE;
8642 8643 8644 8645 8646
	} else if ((pf->hw.aq.api_maj_ver > 1) ||
		   ((pf->hw.aq.api_maj_ver == 1) &&
		    (pf->hw.aq.api_min_ver > 4))) {
		/* Supported in FW API version higher than 1.4 */
		pf->flags |= I40E_FLAG_GENEVE_OFFLOAD_CAPABLE;
8647 8648 8649
		pf->auto_disable_flags = I40E_FLAG_HW_ATR_EVICT_CAPABLE;
	} else {
		pf->auto_disable_flags = I40E_FLAG_HW_ATR_EVICT_CAPABLE;
8650
	}
8651

J
Jesse Brandeburg 已提交
8652 8653 8654 8655
	pf->eeprom_version = 0xDEAD;
	pf->lan_veb = I40E_NO_VEB;
	pf->lan_vsi = I40E_NO_VSI;

8656 8657 8658
	/* By default FW has this off for performance reasons */
	pf->flags &= ~I40E_FLAG_VEB_STATS_ENABLED;

J
Jesse Brandeburg 已提交
8659 8660 8661 8662 8663 8664 8665 8666 8667 8668 8669
	/* set up queue assignment tracking */
	size = sizeof(struct i40e_lump_tracking)
		+ (sizeof(u16) * pf->hw.func_caps.num_tx_qp);
	pf->qp_pile = kzalloc(size, GFP_KERNEL);
	if (!pf->qp_pile) {
		err = -ENOMEM;
		goto sw_init_done;
	}
	pf->qp_pile->num_entries = pf->hw.func_caps.num_tx_qp;
	pf->qp_pile->search_hint = 0;

8670 8671
	pf->tx_timeout_recovery_level = 1;

J
Jesse Brandeburg 已提交
8672 8673
	mutex_init(&pf->switch_mutex);

G
Greg Rose 已提交
8674 8675 8676 8677
	/* If NPAR is enabled nudge the Tx scheduler */
	if (pf->hw.func_caps.npar_enable && (!i40e_get_npar_bw_setting(pf)))
		i40e_set_npar_bw_setting(pf);

J
Jesse Brandeburg 已提交
8678 8679 8680 8681
sw_init_done:
	return err;
}

8682 8683 8684 8685 8686 8687 8688 8689 8690 8691 8692 8693 8694 8695 8696 8697 8698 8699
/**
 * i40e_set_ntuple - set the ntuple feature flag and take action
 * @pf: board private structure to initialize
 * @features: the feature set that the stack is suggesting
 *
 * returns a bool to indicate if reset needs to happen
 **/
bool i40e_set_ntuple(struct i40e_pf *pf, netdev_features_t features)
{
	bool need_reset = false;

	/* Check if Flow Director n-tuple support was enabled or disabled.  If
	 * the state changed, we need to reset.
	 */
	if (features & NETIF_F_NTUPLE) {
		/* Enable filters and mark for reset */
		if (!(pf->flags & I40E_FLAG_FD_SB_ENABLED))
			need_reset = true;
8700 8701 8702
		/* enable FD_SB only if there is MSI-X vector */
		if (pf->num_fdsb_msix > 0)
			pf->flags |= I40E_FLAG_FD_SB_ENABLED;
8703 8704 8705 8706 8707 8708 8709
	} else {
		/* turn off filters, mark for reset and clear SW filter list */
		if (pf->flags & I40E_FLAG_FD_SB_ENABLED) {
			need_reset = true;
			i40e_fdir_filter_exit(pf);
		}
		pf->flags &= ~I40E_FLAG_FD_SB_ENABLED;
8710
		pf->auto_disable_flags &= ~I40E_FLAG_FD_SB_ENABLED;
8711 8712 8713
		/* reset fd counters */
		pf->fd_add_err = pf->fd_atr_cnt = pf->fd_tcp_rule = 0;
		pf->fdir_pf_active_filters = 0;
8714 8715
		/* if ATR was auto disabled it can be re-enabled. */
		if ((pf->flags & I40E_FLAG_FD_ATR_ENABLED) &&
8716
		    (pf->auto_disable_flags & I40E_FLAG_FD_ATR_ENABLED)) {
8717
			pf->auto_disable_flags &= ~I40E_FLAG_FD_ATR_ENABLED;
8718 8719 8720
			if (I40E_DEBUG_FD & pf->hw.debug_mask)
				dev_info(&pf->pdev->dev, "ATR re-enabled.\n");
		}
8721 8722 8723 8724
	}
	return need_reset;
}

8725 8726 8727 8728 8729 8730 8731 8732 8733 8734 8735 8736 8737 8738 8739 8740 8741 8742 8743 8744 8745 8746
/**
 * i40e_clear_rss_lut - clear the rx hash lookup table
 * @vsi: the VSI being configured
 **/
static void i40e_clear_rss_lut(struct i40e_vsi *vsi)
{
	struct i40e_pf *pf = vsi->back;
	struct i40e_hw *hw = &pf->hw;
	u16 vf_id = vsi->vf_id;
	u8 i;

	if (vsi->type == I40E_VSI_MAIN) {
		for (i = 0; i <= I40E_PFQF_HLUT_MAX_INDEX; i++)
			wr32(hw, I40E_PFQF_HLUT(i), 0);
	} else if (vsi->type == I40E_VSI_SRIOV) {
		for (i = 0; i <= I40E_VFQF_HLUT_MAX_INDEX; i++)
			i40e_write_rx_ctl(hw, I40E_VFQF_HLUT1(i, vf_id), 0);
	} else {
		dev_err(&pf->pdev->dev, "Cannot set RSS LUT - invalid VSI type\n");
	}
}

J
Jesse Brandeburg 已提交
8747 8748 8749 8750 8751 8752 8753 8754 8755 8756
/**
 * i40e_set_features - set the netdev feature flags
 * @netdev: ptr to the netdev being adjusted
 * @features: the feature set that the stack is suggesting
 **/
static int i40e_set_features(struct net_device *netdev,
			     netdev_features_t features)
{
	struct i40e_netdev_priv *np = netdev_priv(netdev);
	struct i40e_vsi *vsi = np->vsi;
8757 8758
	struct i40e_pf *pf = vsi->back;
	bool need_reset;
J
Jesse Brandeburg 已提交
8759

8760 8761 8762 8763 8764 8765
	if (features & NETIF_F_RXHASH && !(netdev->features & NETIF_F_RXHASH))
		i40e_pf_config_rss(pf);
	else if (!(features & NETIF_F_RXHASH) &&
		 netdev->features & NETIF_F_RXHASH)
		i40e_clear_rss_lut(vsi);

J
Jesse Brandeburg 已提交
8766 8767 8768 8769 8770
	if (features & NETIF_F_HW_VLAN_CTAG_RX)
		i40e_vlan_stripping_enable(vsi);
	else
		i40e_vlan_stripping_disable(vsi);

8771 8772 8773
	need_reset = i40e_set_ntuple(pf, features);

	if (need_reset)
8774
		i40e_do_reset(pf, BIT_ULL(__I40E_PF_RESET_REQUESTED));
8775

J
Jesse Brandeburg 已提交
8776 8777 8778
	return 0;
}

8779
/**
8780
 * i40e_get_udp_port_idx - Lookup a possibly offloaded for Rx UDP port
8781 8782 8783 8784 8785
 * @pf: board private structure
 * @port: The UDP port to look up
 *
 * Returns the index number or I40E_MAX_PF_UDP_OFFLOAD_PORTS if port not found
 **/
8786
static u8 i40e_get_udp_port_idx(struct i40e_pf *pf, __be16 port)
8787 8788 8789 8790
{
	u8 i;

	for (i = 0; i < I40E_MAX_PF_UDP_OFFLOAD_PORTS; i++) {
8791
		if (pf->udp_ports[i].index == port)
8792 8793 8794 8795 8796 8797 8798
			return i;
	}

	return i;
}

/**
8799
 * i40e_udp_tunnel_add - Get notifications about UDP tunnel ports that come up
8800
 * @netdev: This physical port's netdev
8801
 * @ti: Tunnel endpoint information
8802
 **/
8803 8804
static void i40e_udp_tunnel_add(struct net_device *netdev,
				struct udp_tunnel_info *ti)
8805 8806 8807 8808
{
	struct i40e_netdev_priv *np = netdev_priv(netdev);
	struct i40e_vsi *vsi = np->vsi;
	struct i40e_pf *pf = vsi->back;
8809
	__be16 port = ti->port;
8810 8811 8812
	u8 next_idx;
	u8 idx;

8813
	idx = i40e_get_udp_port_idx(pf, port);
8814 8815 8816

	/* Check if port already exists */
	if (idx < I40E_MAX_PF_UDP_OFFLOAD_PORTS) {
8817
		netdev_info(netdev, "port %d already offloaded\n",
S
Shannon Nelson 已提交
8818
			    ntohs(port));
8819 8820 8821 8822
		return;
	}

	/* Now check if there is space to add the new port */
8823
	next_idx = i40e_get_udp_port_idx(pf, 0);
8824 8825

	if (next_idx == I40E_MAX_PF_UDP_OFFLOAD_PORTS) {
8826
		netdev_info(netdev, "maximum number of offloaded UDP ports reached, not adding port %d\n",
8827 8828 8829 8830
			    ntohs(port));
		return;
	}

8831 8832 8833 8834 8835 8836 8837 8838 8839 8840
	switch (ti->type) {
	case UDP_TUNNEL_TYPE_VXLAN:
		pf->udp_ports[next_idx].type = I40E_AQC_TUNNEL_TYPE_VXLAN;
		break;
	case UDP_TUNNEL_TYPE_GENEVE:
		if (!(pf->flags & I40E_FLAG_GENEVE_OFFLOAD_CAPABLE))
			return;
		pf->udp_ports[next_idx].type = I40E_AQC_TUNNEL_TYPE_NGE;
		break;
	default:
8841 8842 8843 8844 8845 8846 8847
		return;
	}

	/* New port: add it and mark its index in the bitmap */
	pf->udp_ports[next_idx].index = port;
	pf->pending_udp_bitmap |= BIT_ULL(next_idx);
	pf->flags |= I40E_FLAG_UDP_FILTER_SYNC;
8848 8849
}

8850
/**
8851
 * i40e_udp_tunnel_del - Get notifications about UDP tunnel ports that go away
8852
 * @netdev: This physical port's netdev
8853
 * @ti: Tunnel endpoint information
8854
 **/
8855 8856
static void i40e_udp_tunnel_del(struct net_device *netdev,
				struct udp_tunnel_info *ti)
8857 8858 8859 8860
{
	struct i40e_netdev_priv *np = netdev_priv(netdev);
	struct i40e_vsi *vsi = np->vsi;
	struct i40e_pf *pf = vsi->back;
8861
	__be16 port = ti->port;
8862 8863 8864 8865 8866
	u8 idx;

	idx = i40e_get_udp_port_idx(pf, port);

	/* Check if port already exists */
8867 8868
	if (idx >= I40E_MAX_PF_UDP_OFFLOAD_PORTS)
		goto not_found;
8869

8870 8871 8872 8873 8874 8875 8876 8877 8878 8879 8880
	switch (ti->type) {
	case UDP_TUNNEL_TYPE_VXLAN:
		if (pf->udp_ports[idx].type != I40E_AQC_TUNNEL_TYPE_VXLAN)
			goto not_found;
		break;
	case UDP_TUNNEL_TYPE_GENEVE:
		if (pf->udp_ports[idx].type != I40E_AQC_TUNNEL_TYPE_NGE)
			goto not_found;
		break;
	default:
		goto not_found;
8881
	}
8882 8883 8884 8885 8886 8887 8888 8889 8890 8891 8892 8893

	/* if port exists, set it to 0 (mark for deletion)
	 * and make it pending
	 */
	pf->udp_ports[idx].index = 0;
	pf->pending_udp_bitmap |= BIT_ULL(idx);
	pf->flags |= I40E_FLAG_UDP_FILTER_SYNC;

	return;
not_found:
	netdev_warn(netdev, "UDP port %d was not found, not deleting\n",
		    ntohs(port));
8894 8895
}

8896
static int i40e_get_phys_port_id(struct net_device *netdev,
8897
				 struct netdev_phys_item_id *ppid)
8898 8899 8900 8901 8902 8903 8904 8905 8906 8907 8908 8909 8910 8911
{
	struct i40e_netdev_priv *np = netdev_priv(netdev);
	struct i40e_pf *pf = np->vsi->back;
	struct i40e_hw *hw = &pf->hw;

	if (!(pf->flags & I40E_FLAG_PORT_ID_VALID))
		return -EOPNOTSUPP;

	ppid->id_len = min_t(int, sizeof(hw->mac.port_addr), sizeof(ppid->id));
	memcpy(ppid->id, hw->mac.port_addr, ppid->id_len);

	return 0;
}

8912 8913 8914 8915 8916 8917 8918 8919
/**
 * i40e_ndo_fdb_add - add an entry to the hardware database
 * @ndm: the input from the stack
 * @tb: pointer to array of nladdr (unused)
 * @dev: the net device pointer
 * @addr: the MAC address entry being added
 * @flags: instructions from stack about fdb operation
 */
8920 8921
static int i40e_ndo_fdb_add(struct ndmsg *ndm, struct nlattr *tb[],
			    struct net_device *dev,
8922
			    const unsigned char *addr, u16 vid,
8923 8924 8925 8926 8927 8928 8929 8930 8931
			    u16 flags)
{
	struct i40e_netdev_priv *np = netdev_priv(dev);
	struct i40e_pf *pf = np->vsi->back;
	int err = 0;

	if (!(pf->flags & I40E_FLAG_SRIOV_ENABLED))
		return -EOPNOTSUPP;

8932 8933 8934 8935 8936
	if (vid) {
		pr_info("%s: vlans aren't supported yet for dev_uc|mc_add()\n", dev->name);
		return -EINVAL;
	}

8937 8938 8939 8940 8941 8942 8943 8944 8945 8946 8947 8948 8949 8950 8951 8952 8953 8954 8955 8956 8957 8958
	/* Hardware does not support aging addresses so if a
	 * ndm_state is given only allow permanent addresses
	 */
	if (ndm->ndm_state && !(ndm->ndm_state & NUD_PERMANENT)) {
		netdev_info(dev, "FDB only supports static addresses\n");
		return -EINVAL;
	}

	if (is_unicast_ether_addr(addr) || is_link_local_ether_addr(addr))
		err = dev_uc_add_excl(dev, addr);
	else if (is_multicast_ether_addr(addr))
		err = dev_mc_add_excl(dev, addr);
	else
		err = -EINVAL;

	/* Only return duplicate errors if NLM_F_EXCL is set */
	if (err == -EEXIST && !(flags & NLM_F_EXCL))
		err = 0;

	return err;
}

8959 8960 8961 8962 8963 8964 8965 8966 8967 8968 8969 8970 8971
/**
 * i40e_ndo_bridge_setlink - Set the hardware bridge mode
 * @dev: the netdev being configured
 * @nlh: RTNL message
 *
 * Inserts a new hardware bridge if not already created and
 * enables the bridging mode requested (VEB or VEPA). If the
 * hardware bridge has already been inserted and the request
 * is to change the mode then that requires a PF reset to
 * allow rebuild of the components with required hardware
 * bridge mode enabled.
 **/
static int i40e_ndo_bridge_setlink(struct net_device *dev,
8972 8973
				   struct nlmsghdr *nlh,
				   u16 flags)
8974 8975 8976 8977 8978 8979 8980 8981 8982 8983 8984 8985 8986 8987 8988 8989 8990 8991 8992 8993 8994 8995 8996 8997 8998 8999 9000 9001 9002 9003 9004 9005 9006 9007 9008 9009 9010 9011 9012 9013 9014 9015 9016 9017 9018 9019
{
	struct i40e_netdev_priv *np = netdev_priv(dev);
	struct i40e_vsi *vsi = np->vsi;
	struct i40e_pf *pf = vsi->back;
	struct i40e_veb *veb = NULL;
	struct nlattr *attr, *br_spec;
	int i, rem;

	/* Only for PF VSI for now */
	if (vsi->seid != pf->vsi[pf->lan_vsi]->seid)
		return -EOPNOTSUPP;

	/* Find the HW bridge for PF VSI */
	for (i = 0; i < I40E_MAX_VEB && !veb; i++) {
		if (pf->veb[i] && pf->veb[i]->seid == vsi->uplink_seid)
			veb = pf->veb[i];
	}

	br_spec = nlmsg_find_attr(nlh, sizeof(struct ifinfomsg), IFLA_AF_SPEC);

	nla_for_each_nested(attr, br_spec, rem) {
		__u16 mode;

		if (nla_type(attr) != IFLA_BRIDGE_MODE)
			continue;

		mode = nla_get_u16(attr);
		if ((mode != BRIDGE_MODE_VEPA) &&
		    (mode != BRIDGE_MODE_VEB))
			return -EINVAL;

		/* Insert a new HW bridge */
		if (!veb) {
			veb = i40e_veb_setup(pf, 0, vsi->uplink_seid, vsi->seid,
					     vsi->tc_config.enabled_tc);
			if (veb) {
				veb->bridge_mode = mode;
				i40e_config_bridge_mode(veb);
			} else {
				/* No Bridge HW offload available */
				return -ENOENT;
			}
			break;
		} else if (mode != veb->bridge_mode) {
			/* Existing HW bridge but different mode needs reset */
			veb->bridge_mode = mode;
9020 9021 9022 9023 9024 9025
			/* TODO: If no VFs or VMDq VSIs, disallow VEB mode */
			if (mode == BRIDGE_MODE_VEB)
				pf->flags |= I40E_FLAG_VEB_MODE_ENABLED;
			else
				pf->flags &= ~I40E_FLAG_VEB_MODE_ENABLED;
			i40e_do_reset(pf, BIT_ULL(__I40E_PF_RESET_REQUESTED));
9026 9027 9028 9029 9030 9031 9032 9033 9034 9035 9036 9037 9038 9039
			break;
		}
	}

	return 0;
}

/**
 * i40e_ndo_bridge_getlink - Get the hardware bridge mode
 * @skb: skb buff
 * @pid: process id
 * @seq: RTNL message seq #
 * @dev: the netdev being configured
 * @filter_mask: unused
9040
 * @nlflags: netlink flags passed in
9041 9042 9043 9044 9045 9046
 *
 * Return the mode in which the hardware bridge is operating in
 * i.e VEB or VEPA.
 **/
static int i40e_ndo_bridge_getlink(struct sk_buff *skb, u32 pid, u32 seq,
				   struct net_device *dev,
9047 9048
				   u32 __always_unused filter_mask,
				   int nlflags)
9049 9050 9051 9052 9053 9054 9055 9056 9057 9058 9059 9060 9061 9062 9063 9064 9065 9066 9067 9068
{
	struct i40e_netdev_priv *np = netdev_priv(dev);
	struct i40e_vsi *vsi = np->vsi;
	struct i40e_pf *pf = vsi->back;
	struct i40e_veb *veb = NULL;
	int i;

	/* Only for PF VSI for now */
	if (vsi->seid != pf->vsi[pf->lan_vsi]->seid)
		return -EOPNOTSUPP;

	/* Find the HW bridge for the PF VSI */
	for (i = 0; i < I40E_MAX_VEB && !veb; i++) {
		if (pf->veb[i] && pf->veb[i]->seid == vsi->uplink_seid)
			veb = pf->veb[i];
	}

	if (!veb)
		return 0;

9069
	return ndo_dflt_bridge_getlink(skb, pid, seq, dev, veb->bridge_mode,
9070
				       0, 0, nlflags, filter_mask, NULL);
9071 9072
}

9073 9074 9075
/**
 * i40e_features_check - Validate encapsulated packet conforms to limits
 * @skb: skb buff
J
Jean Sacren 已提交
9076
 * @dev: This physical port's netdev
9077 9078 9079 9080 9081 9082
 * @features: Offload features that the stack believes apply
 **/
static netdev_features_t i40e_features_check(struct sk_buff *skb,
					     struct net_device *dev,
					     netdev_features_t features)
{
9083 9084 9085 9086 9087 9088 9089 9090 9091 9092 9093 9094 9095 9096 9097 9098 9099 9100 9101 9102 9103 9104 9105 9106 9107 9108 9109 9110 9111 9112 9113 9114 9115 9116 9117 9118 9119 9120 9121 9122 9123 9124
	size_t len;

	/* No point in doing any of this if neither checksum nor GSO are
	 * being requested for this frame.  We can rule out both by just
	 * checking for CHECKSUM_PARTIAL
	 */
	if (skb->ip_summed != CHECKSUM_PARTIAL)
		return features;

	/* We cannot support GSO if the MSS is going to be less than
	 * 64 bytes.  If it is then we need to drop support for GSO.
	 */
	if (skb_is_gso(skb) && (skb_shinfo(skb)->gso_size < 64))
		features &= ~NETIF_F_GSO_MASK;

	/* MACLEN can support at most 63 words */
	len = skb_network_header(skb) - skb->data;
	if (len & ~(63 * 2))
		goto out_err;

	/* IPLEN and EIPLEN can support at most 127 dwords */
	len = skb_transport_header(skb) - skb_network_header(skb);
	if (len & ~(127 * 4))
		goto out_err;

	if (skb->encapsulation) {
		/* L4TUNLEN can support 127 words */
		len = skb_inner_network_header(skb) - skb_transport_header(skb);
		if (len & ~(127 * 2))
			goto out_err;

		/* IPLEN can support at most 127 dwords */
		len = skb_inner_transport_header(skb) -
		      skb_inner_network_header(skb);
		if (len & ~(127 * 4))
			goto out_err;
	}

	/* No need to validate L4LEN as TCP is the only protocol with a
	 * a flexible value and we support all possible values supported
	 * by TCP, which is at most 15 dwords
	 */
9125 9126

	return features;
9127 9128
out_err:
	return features & ~(NETIF_F_CSUM_MASK | NETIF_F_GSO_MASK);
9129 9130
}

9131
static const struct net_device_ops i40e_netdev_ops = {
J
Jesse Brandeburg 已提交
9132 9133 9134 9135 9136 9137 9138 9139
	.ndo_open		= i40e_open,
	.ndo_stop		= i40e_close,
	.ndo_start_xmit		= i40e_lan_xmit_frame,
	.ndo_get_stats64	= i40e_get_netdev_stats_struct,
	.ndo_set_rx_mode	= i40e_set_rx_mode,
	.ndo_validate_addr	= eth_validate_addr,
	.ndo_set_mac_address	= i40e_set_mac,
	.ndo_change_mtu		= i40e_change_mtu,
J
Jacob Keller 已提交
9140
	.ndo_do_ioctl		= i40e_ioctl,
J
Jesse Brandeburg 已提交
9141 9142 9143 9144 9145 9146
	.ndo_tx_timeout		= i40e_tx_timeout,
	.ndo_vlan_rx_add_vid	= i40e_vlan_rx_add_vid,
	.ndo_vlan_rx_kill_vid	= i40e_vlan_rx_kill_vid,
#ifdef CONFIG_NET_POLL_CONTROLLER
	.ndo_poll_controller	= i40e_netpoll,
#endif
9147
	.ndo_setup_tc		= __i40e_setup_tc,
9148 9149 9150 9151
#ifdef I40E_FCOE
	.ndo_fcoe_enable	= i40e_fcoe_enable,
	.ndo_fcoe_disable	= i40e_fcoe_disable,
#endif
J
Jesse Brandeburg 已提交
9152 9153 9154
	.ndo_set_features	= i40e_set_features,
	.ndo_set_vf_mac		= i40e_ndo_set_vf_mac,
	.ndo_set_vf_vlan	= i40e_ndo_set_vf_port_vlan,
9155
	.ndo_set_vf_rate	= i40e_ndo_set_vf_bw,
J
Jesse Brandeburg 已提交
9156
	.ndo_get_vf_config	= i40e_ndo_get_vf_config,
M
Mitch Williams 已提交
9157
	.ndo_set_vf_link_state	= i40e_ndo_set_vf_link_state,
9158
	.ndo_set_vf_spoofchk	= i40e_ndo_set_vf_spoofchk,
9159
	.ndo_set_vf_trust	= i40e_ndo_set_vf_trust,
9160 9161
	.ndo_udp_tunnel_add	= i40e_udp_tunnel_add,
	.ndo_udp_tunnel_del	= i40e_udp_tunnel_del,
9162
	.ndo_get_phys_port_id	= i40e_get_phys_port_id,
9163
	.ndo_fdb_add		= i40e_ndo_fdb_add,
9164
	.ndo_features_check	= i40e_features_check,
9165 9166
	.ndo_bridge_getlink	= i40e_ndo_bridge_getlink,
	.ndo_bridge_setlink	= i40e_ndo_bridge_setlink,
J
Jesse Brandeburg 已提交
9167 9168 9169 9170 9171 9172 9173 9174 9175 9176 9177 9178 9179 9180 9181 9182 9183 9184
};

/**
 * i40e_config_netdev - Setup the netdev flags
 * @vsi: the VSI being configured
 *
 * Returns 0 on success, negative value on failure
 **/
static int i40e_config_netdev(struct i40e_vsi *vsi)
{
	struct i40e_pf *pf = vsi->back;
	struct i40e_hw *hw = &pf->hw;
	struct i40e_netdev_priv *np;
	struct net_device *netdev;
	u8 mac_addr[ETH_ALEN];
	int etherdev_size;

	etherdev_size = sizeof(struct i40e_netdev_priv);
9185
	netdev = alloc_etherdev_mq(etherdev_size, vsi->alloc_queue_pairs);
J
Jesse Brandeburg 已提交
9186 9187 9188 9189 9190 9191 9192
	if (!netdev)
		return -ENOMEM;

	vsi->netdev = netdev;
	np = netdev_priv(netdev);
	np->vsi = vsi;

9193 9194 9195 9196 9197 9198 9199 9200 9201
	netdev->hw_enc_features |= NETIF_F_SG			|
				   NETIF_F_IP_CSUM		|
				   NETIF_F_IPV6_CSUM		|
				   NETIF_F_HIGHDMA		|
				   NETIF_F_SOFT_FEATURES	|
				   NETIF_F_TSO			|
				   NETIF_F_TSO_ECN		|
				   NETIF_F_TSO6			|
				   NETIF_F_GSO_GRE		|
9202
				   NETIF_F_GSO_GRE_CSUM		|
9203
				   NETIF_F_GSO_IPXIP4		|
9204
				   NETIF_F_GSO_IPXIP6		|
9205 9206
				   NETIF_F_GSO_UDP_TUNNEL	|
				   NETIF_F_GSO_UDP_TUNNEL_CSUM	|
9207
				   NETIF_F_GSO_PARTIAL		|
9208 9209 9210
				   NETIF_F_SCTP_CRC		|
				   NETIF_F_RXHASH		|
				   NETIF_F_RXCSUM		|
9211
				   0;
J
Jesse Brandeburg 已提交
9212

9213
	if (!(pf->flags & I40E_FLAG_OUTER_UDP_CSUM_CAPABLE))
9214 9215 9216
		netdev->gso_partial_features |= NETIF_F_GSO_UDP_TUNNEL_CSUM;

	netdev->gso_partial_features |= NETIF_F_GSO_GRE_CSUM;
9217 9218

	/* record features VLANs can make use of */
9219 9220
	netdev->vlan_features |= netdev->hw_enc_features |
				 NETIF_F_TSO_MANGLEID;
J
Jesse Brandeburg 已提交
9221

9222
	if (!(pf->flags & I40E_FLAG_MFP_ENABLED))
9223 9224 9225 9226 9227
		netdev->hw_features |= NETIF_F_NTUPLE;

	netdev->hw_features |= netdev->hw_enc_features	|
			       NETIF_F_HW_VLAN_CTAG_TX	|
			       NETIF_F_HW_VLAN_CTAG_RX;
9228

9229
	netdev->features |= netdev->hw_features | NETIF_F_HW_VLAN_CTAG_FILTER;
9230
	netdev->hw_enc_features |= NETIF_F_TSO_MANGLEID;
J
Jesse Brandeburg 已提交
9231 9232 9233

	if (vsi->type == I40E_VSI_MAIN) {
		SET_NETDEV_DEV(netdev, &pf->pdev->dev);
9234
		ether_addr_copy(mac_addr, hw->mac.perm_addr);
9235
		spin_lock_bh(&vsi->mac_filter_hash_lock);
9236
		i40e_add_filter(vsi, mac_addr, I40E_VLAN_ANY);
9237
		spin_unlock_bh(&vsi->mac_filter_hash_lock);
J
Jesse Brandeburg 已提交
9238 9239 9240 9241 9242
	} else {
		/* relate the VSI_VMDQ name to the VSI_MAIN name */
		snprintf(netdev->name, IFNAMSIZ, "%sv%%d",
			 pf->vsi[pf->lan_vsi]->netdev->name);
		random_ether_addr(mac_addr);
K
Kiran Patil 已提交
9243

9244
		spin_lock_bh(&vsi->mac_filter_hash_lock);
9245
		i40e_add_filter(vsi, mac_addr, I40E_VLAN_ANY);
9246
		spin_unlock_bh(&vsi->mac_filter_hash_lock);
J
Jesse Brandeburg 已提交
9247
	}
K
Kiran Patil 已提交
9248

9249 9250
	ether_addr_copy(netdev->dev_addr, mac_addr);
	ether_addr_copy(netdev->perm_addr, mac_addr);
9251

J
Jesse Brandeburg 已提交
9252 9253 9254 9255 9256 9257 9258 9259
	netdev->priv_flags |= IFF_UNICAST_FLT;
	netdev->priv_flags |= IFF_SUPP_NOFCS;
	/* Setup netdev TC information */
	i40e_vsi_config_netdev_tc(vsi, vsi->tc_config.enabled_tc);

	netdev->netdev_ops = &i40e_netdev_ops;
	netdev->watchdog_timeo = 5 * HZ;
	i40e_set_ethtool_ops(netdev);
9260 9261 9262
#ifdef I40E_FCOE
	i40e_fcoe_config_netdev(netdev, vsi);
#endif
J
Jesse Brandeburg 已提交
9263

9264 9265 9266 9267 9268
	/* MTU range: 68 - 9706 */
	netdev->min_mtu = ETH_MIN_MTU;
	netdev->max_mtu = I40E_MAX_RXBUFFER -
			  (ETH_HLEN + ETH_FCS_LEN + VLAN_HLEN);

J
Jesse Brandeburg 已提交
9269 9270 9271 9272 9273 9274 9275 9276 9277 9278 9279 9280 9281 9282 9283 9284 9285 9286
	return 0;
}

/**
 * i40e_vsi_delete - Delete a VSI from the switch
 * @vsi: the VSI being removed
 *
 * Returns 0 on success, negative value on failure
 **/
static void i40e_vsi_delete(struct i40e_vsi *vsi)
{
	/* remove default VSI is not allowed */
	if (vsi == vsi->back->vsi[vsi->back->lan_vsi])
		return;

	i40e_aq_delete_element(&vsi->back->hw, vsi->seid, NULL);
}

9287 9288 9289 9290 9291 9292 9293 9294 9295 9296 9297 9298 9299 9300 9301 9302
/**
 * i40e_is_vsi_uplink_mode_veb - Check if the VSI's uplink bridge mode is VEB
 * @vsi: the VSI being queried
 *
 * Returns 1 if HW bridge mode is VEB and return 0 in case of VEPA mode
 **/
int i40e_is_vsi_uplink_mode_veb(struct i40e_vsi *vsi)
{
	struct i40e_veb *veb;
	struct i40e_pf *pf = vsi->back;

	/* Uplink is not a bridge so default to VEB */
	if (vsi->veb_idx == I40E_NO_VEB)
		return 1;

	veb = pf->veb[vsi->veb_idx];
9303 9304 9305 9306 9307 9308
	if (!veb) {
		dev_info(&pf->pdev->dev,
			 "There is no veb associated with the bridge\n");
		return -ENOENT;
	}

9309
	/* Uplink is a bridge in VEPA mode */
9310
	if (veb->bridge_mode & BRIDGE_MODE_VEPA) {
9311
		return 0;
9312 9313 9314 9315
	} else {
		/* Uplink is a bridge in VEB mode */
		return 1;
	}
9316

9317 9318
	/* VEPA is now default bridge, so return 0 */
	return 0;
9319 9320
}

J
Jesse Brandeburg 已提交
9321 9322 9323 9324 9325 9326 9327 9328 9329 9330
/**
 * i40e_add_vsi - Add a VSI to the switch
 * @vsi: the VSI being configured
 *
 * This initializes a VSI context depending on the VSI type to be added and
 * passes it down to the add_vsi aq command.
 **/
static int i40e_add_vsi(struct i40e_vsi *vsi)
{
	int ret = -ENODEV;
9331
	i40e_status aq_ret = 0;
J
Jesse Brandeburg 已提交
9332 9333 9334
	struct i40e_pf *pf = vsi->back;
	struct i40e_hw *hw = &pf->hw;
	struct i40e_vsi_context ctxt;
9335 9336 9337
	struct i40e_mac_filter *f;
	struct hlist_node *h;
	int bkt;
K
Kiran Patil 已提交
9338

J
Jesse Brandeburg 已提交
9339 9340 9341 9342 9343 9344 9345 9346 9347 9348 9349 9350 9351 9352 9353 9354 9355 9356
	u8 enabled_tc = 0x1; /* TC0 enabled */
	int f_count = 0;

	memset(&ctxt, 0, sizeof(ctxt));
	switch (vsi->type) {
	case I40E_VSI_MAIN:
		/* The PF's main VSI is already setup as part of the
		 * device initialization, so we'll not bother with
		 * the add_vsi call, but we will retrieve the current
		 * VSI context.
		 */
		ctxt.seid = pf->main_vsi_seid;
		ctxt.pf_num = pf->hw.pf_id;
		ctxt.vf_num = 0;
		ret = i40e_aq_get_vsi_params(&pf->hw, &ctxt, NULL);
		ctxt.flags = I40E_AQ_VSI_TYPE_PF;
		if (ret) {
			dev_info(&pf->pdev->dev,
9357 9358 9359 9360
				 "couldn't get PF vsi config, err %s aq_err %s\n",
				 i40e_stat_str(&pf->hw, ret),
				 i40e_aq_str(&pf->hw,
					     pf->hw.aq.asq_last_status));
J
Jesse Brandeburg 已提交
9361 9362
			return -ENOENT;
		}
9363
		vsi->info = ctxt.info;
J
Jesse Brandeburg 已提交
9364 9365 9366 9367 9368 9369 9370 9371
		vsi->info.valid_sections = 0;

		vsi->seid = ctxt.seid;
		vsi->id = ctxt.vsi_number;

		enabled_tc = i40e_pf_get_tc_map(pf);

		/* MFP mode setup queue map and update VSI */
9372 9373
		if ((pf->flags & I40E_FLAG_MFP_ENABLED) &&
		    !(pf->hw.func_caps.iscsi)) { /* NIC type PF */
J
Jesse Brandeburg 已提交
9374 9375 9376 9377 9378 9379 9380 9381
			memset(&ctxt, 0, sizeof(ctxt));
			ctxt.seid = pf->main_vsi_seid;
			ctxt.pf_num = pf->hw.pf_id;
			ctxt.vf_num = 0;
			i40e_vsi_setup_queue_map(vsi, &ctxt, enabled_tc, false);
			ret = i40e_aq_update_vsi_params(hw, &ctxt, NULL);
			if (ret) {
				dev_info(&pf->pdev->dev,
9382 9383 9384 9385
					 "update vsi failed, err %s aq_err %s\n",
					 i40e_stat_str(&pf->hw, ret),
					 i40e_aq_str(&pf->hw,
						    pf->hw.aq.asq_last_status));
J
Jesse Brandeburg 已提交
9386 9387 9388 9389 9390 9391 9392 9393 9394 9395
				ret = -ENOENT;
				goto err;
			}
			/* update the local VSI info queue map */
			i40e_vsi_update_queue_map(vsi, &ctxt);
			vsi->info.valid_sections = 0;
		} else {
			/* Default/Main VSI is only enabled for TC0
			 * reconfigure it to enable all TCs that are
			 * available on the port in SFP mode.
9396 9397
			 * For MFP case the iSCSI PF would use this
			 * flow to enable LAN+iSCSI TC.
J
Jesse Brandeburg 已提交
9398 9399 9400 9401
			 */
			ret = i40e_vsi_config_tc(vsi, enabled_tc);
			if (ret) {
				dev_info(&pf->pdev->dev,
9402 9403 9404 9405 9406
					 "failed to configure TCs for main VSI tc_map 0x%08x, err %s aq_err %s\n",
					 enabled_tc,
					 i40e_stat_str(&pf->hw, ret),
					 i40e_aq_str(&pf->hw,
						    pf->hw.aq.asq_last_status));
J
Jesse Brandeburg 已提交
9407 9408 9409 9410 9411 9412
				ret = -ENOENT;
			}
		}
		break;

	case I40E_VSI_FDIR:
9413 9414 9415
		ctxt.pf_num = hw->pf_id;
		ctxt.vf_num = 0;
		ctxt.uplink_seid = vsi->uplink_seid;
9416
		ctxt.connection_type = I40E_AQ_VSI_CONN_TYPE_NORMAL;
9417
		ctxt.flags = I40E_AQ_VSI_TYPE_PF;
9418 9419
		if ((pf->flags & I40E_FLAG_VEB_MODE_ENABLED) &&
		    (i40e_is_vsi_uplink_mode_veb(vsi))) {
9420
			ctxt.info.valid_sections |=
9421
			     cpu_to_le16(I40E_AQ_VSI_PROP_SWITCH_VALID);
9422
			ctxt.info.switch_id =
9423
			   cpu_to_le16(I40E_AQ_VSI_SW_ID_FLAG_ALLOW_LB);
9424
		}
J
Jesse Brandeburg 已提交
9425 9426 9427 9428 9429 9430 9431
		i40e_vsi_setup_queue_map(vsi, &ctxt, enabled_tc, true);
		break;

	case I40E_VSI_VMDQ2:
		ctxt.pf_num = hw->pf_id;
		ctxt.vf_num = 0;
		ctxt.uplink_seid = vsi->uplink_seid;
9432
		ctxt.connection_type = I40E_AQ_VSI_CONN_TYPE_NORMAL;
J
Jesse Brandeburg 已提交
9433 9434 9435 9436 9437
		ctxt.flags = I40E_AQ_VSI_TYPE_VMDQ2;

		/* This VSI is connected to VEB so the switch_id
		 * should be set to zero by default.
		 */
9438 9439 9440 9441 9442 9443
		if (i40e_is_vsi_uplink_mode_veb(vsi)) {
			ctxt.info.valid_sections |=
				cpu_to_le16(I40E_AQ_VSI_PROP_SWITCH_VALID);
			ctxt.info.switch_id =
				cpu_to_le16(I40E_AQ_VSI_SW_ID_FLAG_ALLOW_LB);
		}
J
Jesse Brandeburg 已提交
9444 9445 9446 9447 9448 9449 9450 9451 9452

		/* Setup the VSI tx/rx queue map for TC0 only for now */
		i40e_vsi_setup_queue_map(vsi, &ctxt, enabled_tc, true);
		break;

	case I40E_VSI_SRIOV:
		ctxt.pf_num = hw->pf_id;
		ctxt.vf_num = vsi->vf_id + hw->func_caps.vf_base_id;
		ctxt.uplink_seid = vsi->uplink_seid;
9453
		ctxt.connection_type = I40E_AQ_VSI_CONN_TYPE_NORMAL;
J
Jesse Brandeburg 已提交
9454 9455 9456 9457 9458
		ctxt.flags = I40E_AQ_VSI_TYPE_VF;

		/* This VSI is connected to VEB so the switch_id
		 * should be set to zero by default.
		 */
9459 9460 9461 9462 9463 9464
		if (i40e_is_vsi_uplink_mode_veb(vsi)) {
			ctxt.info.valid_sections |=
				cpu_to_le16(I40E_AQ_VSI_PROP_SWITCH_VALID);
			ctxt.info.switch_id =
				cpu_to_le16(I40E_AQ_VSI_SW_ID_FLAG_ALLOW_LB);
		}
J
Jesse Brandeburg 已提交
9465

9466 9467 9468 9469
		if (vsi->back->flags & I40E_FLAG_IWARP_ENABLED) {
			ctxt.info.valid_sections |=
				cpu_to_le16(I40E_AQ_VSI_PROP_QUEUE_OPT_VALID);
			ctxt.info.queueing_opt_flags |=
9470 9471
				(I40E_AQ_VSI_QUE_OPT_TCP_ENA |
				 I40E_AQ_VSI_QUE_OPT_RSS_LUT_VSI);
9472 9473
		}

J
Jesse Brandeburg 已提交
9474 9475
		ctxt.info.valid_sections |= cpu_to_le16(I40E_AQ_VSI_PROP_VLAN_VALID);
		ctxt.info.port_vlan_flags |= I40E_AQ_VSI_PVLAN_MODE_ALL;
9476 9477 9478 9479 9480 9481 9482
		if (pf->vf[vsi->vf_id].spoofchk) {
			ctxt.info.valid_sections |=
				cpu_to_le16(I40E_AQ_VSI_PROP_SECURITY_VALID);
			ctxt.info.sec_flags |=
				(I40E_AQ_VSI_SEC_FLAG_ENABLE_VLAN_CHK |
				 I40E_AQ_VSI_SEC_FLAG_ENABLE_MAC_CHK);
		}
J
Jesse Brandeburg 已提交
9483 9484 9485 9486
		/* Setup the VSI tx/rx queue map for TC0 only for now */
		i40e_vsi_setup_queue_map(vsi, &ctxt, enabled_tc, true);
		break;

9487 9488 9489 9490 9491 9492 9493 9494 9495 9496
#ifdef I40E_FCOE
	case I40E_VSI_FCOE:
		ret = i40e_fcoe_vsi_init(vsi, &ctxt);
		if (ret) {
			dev_info(&pf->pdev->dev, "failed to initialize FCoE VSI\n");
			return ret;
		}
		break;

#endif /* I40E_FCOE */
9497 9498 9499 9500
	case I40E_VSI_IWARP:
		/* send down message to iWARP */
		break;

J
Jesse Brandeburg 已提交
9501 9502 9503 9504 9505 9506 9507 9508
	default:
		return -ENODEV;
	}

	if (vsi->type != I40E_VSI_MAIN) {
		ret = i40e_aq_add_vsi(hw, &ctxt, NULL);
		if (ret) {
			dev_info(&vsi->back->pdev->dev,
9509 9510 9511 9512
				 "add vsi failed, err %s aq_err %s\n",
				 i40e_stat_str(&pf->hw, ret),
				 i40e_aq_str(&pf->hw,
					     pf->hw.aq.asq_last_status));
J
Jesse Brandeburg 已提交
9513 9514 9515
			ret = -ENOENT;
			goto err;
		}
9516
		vsi->info = ctxt.info;
J
Jesse Brandeburg 已提交
9517 9518 9519 9520
		vsi->info.valid_sections = 0;
		vsi->seid = ctxt.seid;
		vsi->id = ctxt.vsi_number;
	}
9521 9522 9523 9524 9525 9526 9527 9528 9529 9530 9531 9532
	/* Except FDIR VSI, for all othet VSI set the broadcast filter */
	if (vsi->type != I40E_VSI_FDIR) {
		aq_ret = i40e_aq_set_vsi_broadcast(hw, vsi->seid, true, NULL);
		if (aq_ret) {
			ret = i40e_aq_rc_to_posix(aq_ret,
						  hw->aq.asq_last_status);
			dev_info(&pf->pdev->dev,
				 "set brdcast promisc failed, err %s, aq_err %s\n",
				 i40e_stat_str(hw, aq_ret),
				 i40e_aq_str(hw, hw->aq.asq_last_status));
		}
	}
J
Jesse Brandeburg 已提交
9533

9534 9535
	vsi->active_filters = 0;
	clear_bit(__I40E_FILTER_OVERFLOW_PROMISC, &vsi->state);
9536
	spin_lock_bh(&vsi->mac_filter_hash_lock);
J
Jesse Brandeburg 已提交
9537
	/* If macvlan filters already exist, force them to get loaded */
9538
	hash_for_each_safe(vsi->mac_filter_hash, bkt, h, f, hlist) {
9539
		f->state = I40E_FILTER_NEW;
J
Jesse Brandeburg 已提交
9540
		f_count++;
K
Kiran Patil 已提交
9541
	}
9542
	spin_unlock_bh(&vsi->mac_filter_hash_lock);
9543

J
Jesse Brandeburg 已提交
9544 9545 9546 9547 9548 9549 9550 9551 9552
	if (f_count) {
		vsi->flags |= I40E_VSI_FLAG_FILTER_CHANGED;
		pf->flags |= I40E_FLAG_FILTER_SYNC;
	}

	/* Update VSI BW information */
	ret = i40e_vsi_get_bw_info(vsi);
	if (ret) {
		dev_info(&pf->pdev->dev,
9553 9554 9555
			 "couldn't get vsi bw info, err %s aq_err %s\n",
			 i40e_stat_str(&pf->hw, ret),
			 i40e_aq_str(&pf->hw, pf->hw.aq.asq_last_status));
J
Jesse Brandeburg 已提交
9556 9557 9558 9559 9560 9561 9562 9563 9564 9565 9566 9567 9568 9569 9570 9571
		/* VSI is already added so not tearing that up */
		ret = 0;
	}

err:
	return ret;
}

/**
 * i40e_vsi_release - Delete a VSI and free its resources
 * @vsi: the VSI being removed
 *
 * Returns 0 on success or < 0 on error
 **/
int i40e_vsi_release(struct i40e_vsi *vsi)
{
9572 9573
	struct i40e_mac_filter *f;
	struct hlist_node *h;
J
Jesse Brandeburg 已提交
9574 9575 9576
	struct i40e_veb *veb = NULL;
	struct i40e_pf *pf;
	u16 uplink_seid;
9577
	int i, n, bkt;
J
Jesse Brandeburg 已提交
9578 9579 9580 9581 9582 9583 9584 9585 9586 9587 9588 9589 9590 9591 9592 9593 9594 9595 9596 9597 9598 9599 9600 9601

	pf = vsi->back;

	/* release of a VEB-owner or last VSI is not allowed */
	if (vsi->flags & I40E_VSI_FLAG_VEB_OWNER) {
		dev_info(&pf->pdev->dev, "VSI %d has existing VEB %d\n",
			 vsi->seid, vsi->uplink_seid);
		return -ENODEV;
	}
	if (vsi == pf->vsi[pf->lan_vsi] &&
	    !test_bit(__I40E_DOWN, &pf->state)) {
		dev_info(&pf->pdev->dev, "Can't remove PF VSI\n");
		return -ENODEV;
	}

	uplink_seid = vsi->uplink_seid;
	if (vsi->type != I40E_VSI_SRIOV) {
		if (vsi->netdev_registered) {
			vsi->netdev_registered = false;
			if (vsi->netdev) {
				/* results in a call to i40e_close() */
				unregister_netdev(vsi->netdev);
			}
		} else {
9602
			i40e_vsi_close(vsi);
J
Jesse Brandeburg 已提交
9603 9604 9605 9606
		}
		i40e_vsi_disable_irq(vsi);
	}

9607
	spin_lock_bh(&vsi->mac_filter_hash_lock);
9608 9609 9610 9611 9612 9613 9614 9615

	/* clear the sync flag on all filters */
	if (vsi->netdev) {
		__dev_uc_unsync(vsi->netdev, NULL);
		__dev_mc_unsync(vsi->netdev, NULL);
	}

	/* make sure any remaining filters are marked for deletion */
9616
	hash_for_each_safe(vsi->mac_filter_hash, bkt, h, f, hlist)
9617
		__i40e_del_filter(vsi, f);
9618

9619
	spin_unlock_bh(&vsi->mac_filter_hash_lock);
K
Kiran Patil 已提交
9620

J
Jesse Brandeburg 已提交
9621
	i40e_sync_vsi_filters(vsi);
J
Jesse Brandeburg 已提交
9622 9623 9624

	i40e_vsi_delete(vsi);
	i40e_vsi_free_q_vectors(vsi);
9625 9626 9627 9628
	if (vsi->netdev) {
		free_netdev(vsi->netdev);
		vsi->netdev = NULL;
	}
J
Jesse Brandeburg 已提交
9629 9630 9631 9632 9633 9634 9635 9636 9637 9638 9639
	i40e_vsi_clear_rings(vsi);
	i40e_vsi_clear(vsi);

	/* If this was the last thing on the VEB, except for the
	 * controlling VSI, remove the VEB, which puts the controlling
	 * VSI onto the next level down in the switch.
	 *
	 * Well, okay, there's one more exception here: don't remove
	 * the orphan VEBs yet.  We'll wait for an explicit remove request
	 * from up the network stack.
	 */
M
Mitch Williams 已提交
9640
	for (n = 0, i = 0; i < pf->num_alloc_vsi; i++) {
J
Jesse Brandeburg 已提交
9641 9642 9643 9644 9645 9646 9647 9648 9649 9650 9651 9652 9653 9654 9655 9656 9657 9658 9659 9660 9661 9662 9663 9664 9665 9666 9667 9668 9669 9670 9671 9672 9673 9674 9675
		if (pf->vsi[i] &&
		    pf->vsi[i]->uplink_seid == uplink_seid &&
		    (pf->vsi[i]->flags & I40E_VSI_FLAG_VEB_OWNER) == 0) {
			n++;      /* count the VSIs */
		}
	}
	for (i = 0; i < I40E_MAX_VEB; i++) {
		if (!pf->veb[i])
			continue;
		if (pf->veb[i]->uplink_seid == uplink_seid)
			n++;     /* count the VEBs */
		if (pf->veb[i]->seid == uplink_seid)
			veb = pf->veb[i];
	}
	if (n == 0 && veb && veb->uplink_seid != 0)
		i40e_veb_release(veb);

	return 0;
}

/**
 * i40e_vsi_setup_vectors - Set up the q_vectors for the given VSI
 * @vsi: ptr to the VSI
 *
 * This should only be called after i40e_vsi_mem_alloc() which allocates the
 * corresponding SW VSI structure and initializes num_queue_pairs for the
 * newly allocated VSI.
 *
 * Returns 0 on success or negative on failure
 **/
static int i40e_vsi_setup_vectors(struct i40e_vsi *vsi)
{
	int ret = -ENOENT;
	struct i40e_pf *pf = vsi->back;

9676
	if (vsi->q_vectors[0]) {
J
Jesse Brandeburg 已提交
9677 9678 9679 9680 9681 9682
		dev_info(&pf->pdev->dev, "VSI %d has existing q_vectors\n",
			 vsi->seid);
		return -EEXIST;
	}

	if (vsi->base_vector) {
9683
		dev_info(&pf->pdev->dev, "VSI %d has non-zero base vector %d\n",
J
Jesse Brandeburg 已提交
9684 9685 9686 9687
			 vsi->seid, vsi->base_vector);
		return -EEXIST;
	}

9688
	ret = i40e_vsi_alloc_q_vectors(vsi);
J
Jesse Brandeburg 已提交
9689 9690 9691 9692 9693 9694 9695 9696
	if (ret) {
		dev_info(&pf->pdev->dev,
			 "failed to allocate %d q_vector for VSI %d, ret=%d\n",
			 vsi->num_q_vectors, vsi->seid, ret);
		vsi->num_q_vectors = 0;
		goto vector_setup_out;
	}

9697 9698 9699 9700 9701
	/* In Legacy mode, we do not have to get any other vector since we
	 * piggyback on the misc/ICR0 for queue interrupts.
	*/
	if (!(pf->flags & I40E_FLAG_MSIX_ENABLED))
		return ret;
9702 9703 9704
	if (vsi->num_q_vectors)
		vsi->base_vector = i40e_get_lump(pf, pf->irq_pile,
						 vsi->num_q_vectors, vsi->idx);
J
Jesse Brandeburg 已提交
9705 9706
	if (vsi->base_vector < 0) {
		dev_info(&pf->pdev->dev,
9707 9708
			 "failed to get tracking for %d vectors for VSI %d, err=%d\n",
			 vsi->num_q_vectors, vsi->seid, vsi->base_vector);
J
Jesse Brandeburg 已提交
9709 9710 9711 9712 9713 9714 9715 9716 9717
		i40e_vsi_free_q_vectors(vsi);
		ret = -ENOENT;
		goto vector_setup_out;
	}

vector_setup_out:
	return ret;
}

9718 9719 9720 9721 9722 9723 9724 9725 9726 9727 9728
/**
 * i40e_vsi_reinit_setup - return and reallocate resources for a VSI
 * @vsi: pointer to the vsi.
 *
 * This re-allocates a vsi's queue resources.
 *
 * Returns pointer to the successfully allocated and configured VSI sw struct
 * on success, otherwise returns NULL on failure.
 **/
static struct i40e_vsi *i40e_vsi_reinit_setup(struct i40e_vsi *vsi)
{
J
John Underwood 已提交
9729
	struct i40e_pf *pf;
9730 9731 9732
	u8 enabled_tc;
	int ret;

J
John Underwood 已提交
9733 9734 9735 9736 9737
	if (!vsi)
		return NULL;

	pf = vsi->back;

9738 9739 9740 9741 9742 9743 9744 9745 9746 9747 9748
	i40e_put_lump(pf->qp_pile, vsi->base_queue, vsi->idx);
	i40e_vsi_clear_rings(vsi);

	i40e_vsi_free_arrays(vsi, false);
	i40e_set_num_rings_in_vsi(vsi);
	ret = i40e_vsi_alloc_arrays(vsi, false);
	if (ret)
		goto err_vsi;

	ret = i40e_get_lump(pf, pf->qp_pile, vsi->alloc_queue_pairs, vsi->idx);
	if (ret < 0) {
9749
		dev_info(&pf->pdev->dev,
9750
			 "failed to get tracking for %d queues for VSI %d err %d\n",
9751
			 vsi->alloc_queue_pairs, vsi->seid, ret);
9752 9753 9754 9755 9756 9757 9758 9759 9760 9761 9762 9763 9764 9765 9766 9767 9768 9769 9770 9771 9772 9773 9774 9775 9776 9777 9778 9779 9780 9781 9782 9783 9784 9785 9786
		goto err_vsi;
	}
	vsi->base_queue = ret;

	/* Update the FW view of the VSI. Force a reset of TC and queue
	 * layout configurations.
	 */
	enabled_tc = pf->vsi[pf->lan_vsi]->tc_config.enabled_tc;
	pf->vsi[pf->lan_vsi]->tc_config.enabled_tc = 0;
	pf->vsi[pf->lan_vsi]->seid = pf->main_vsi_seid;
	i40e_vsi_config_tc(pf->vsi[pf->lan_vsi], enabled_tc);

	/* assign it some queues */
	ret = i40e_alloc_rings(vsi);
	if (ret)
		goto err_rings;

	/* map all of the rings to the q_vectors */
	i40e_vsi_map_rings_to_vectors(vsi);
	return vsi;

err_rings:
	i40e_vsi_free_q_vectors(vsi);
	if (vsi->netdev_registered) {
		vsi->netdev_registered = false;
		unregister_netdev(vsi->netdev);
		free_netdev(vsi->netdev);
		vsi->netdev = NULL;
	}
	i40e_aq_delete_element(&pf->hw, vsi->seid, NULL);
err_vsi:
	i40e_vsi_clear(vsi);
	return NULL;
}

J
Jesse Brandeburg 已提交
9787 9788 9789 9790 9791 9792 9793 9794 9795 9796 9797 9798 9799 9800 9801 9802 9803 9804 9805 9806 9807 9808 9809 9810 9811 9812 9813 9814 9815 9816 9817 9818 9819 9820 9821 9822 9823 9824 9825 9826 9827 9828 9829
/**
 * i40e_vsi_setup - Set up a VSI by a given type
 * @pf: board private structure
 * @type: VSI type
 * @uplink_seid: the switch element to link to
 * @param1: usage depends upon VSI type. For VF types, indicates VF id
 *
 * This allocates the sw VSI structure and its queue resources, then add a VSI
 * to the identified VEB.
 *
 * Returns pointer to the successfully allocated and configure VSI sw struct on
 * success, otherwise returns NULL on failure.
 **/
struct i40e_vsi *i40e_vsi_setup(struct i40e_pf *pf, u8 type,
				u16 uplink_seid, u32 param1)
{
	struct i40e_vsi *vsi = NULL;
	struct i40e_veb *veb = NULL;
	int ret, i;
	int v_idx;

	/* The requested uplink_seid must be either
	 *     - the PF's port seid
	 *              no VEB is needed because this is the PF
	 *              or this is a Flow Director special case VSI
	 *     - seid of an existing VEB
	 *     - seid of a VSI that owns an existing VEB
	 *     - seid of a VSI that doesn't own a VEB
	 *              a new VEB is created and the VSI becomes the owner
	 *     - seid of the PF VSI, which is what creates the first VEB
	 *              this is a special case of the previous
	 *
	 * Find which uplink_seid we were given and create a new VEB if needed
	 */
	for (i = 0; i < I40E_MAX_VEB; i++) {
		if (pf->veb[i] && pf->veb[i]->seid == uplink_seid) {
			veb = pf->veb[i];
			break;
		}
	}

	if (!veb && uplink_seid != pf->mac_seid) {

M
Mitch Williams 已提交
9830
		for (i = 0; i < pf->num_alloc_vsi; i++) {
J
Jesse Brandeburg 已提交
9831 9832 9833 9834 9835 9836 9837 9838 9839 9840 9841 9842 9843 9844 9845 9846 9847
			if (pf->vsi[i] && pf->vsi[i]->seid == uplink_seid) {
				vsi = pf->vsi[i];
				break;
			}
		}
		if (!vsi) {
			dev_info(&pf->pdev->dev, "no such uplink_seid %d\n",
				 uplink_seid);
			return NULL;
		}

		if (vsi->uplink_seid == pf->mac_seid)
			veb = i40e_veb_setup(pf, 0, pf->mac_seid, vsi->seid,
					     vsi->tc_config.enabled_tc);
		else if ((vsi->flags & I40E_VSI_FLAG_VEB_OWNER) == 0)
			veb = i40e_veb_setup(pf, 0, vsi->uplink_seid, vsi->seid,
					     vsi->tc_config.enabled_tc);
9848 9849 9850
		if (veb) {
			if (vsi->seid != pf->vsi[pf->lan_vsi]->seid) {
				dev_info(&vsi->back->pdev->dev,
S
Shannon Nelson 已提交
9851
					 "New VSI creation error, uplink seid of LAN VSI expected.\n");
9852 9853
				return NULL;
			}
9854 9855 9856 9857 9858 9859 9860 9861
			/* We come up by default in VEPA mode if SRIOV is not
			 * already enabled, in which case we can't force VEPA
			 * mode.
			 */
			if (!(pf->flags & I40E_FLAG_VEB_MODE_ENABLED)) {
				veb->bridge_mode = BRIDGE_MODE_VEPA;
				pf->flags &= ~I40E_FLAG_VEB_MODE_ENABLED;
			}
9862
			i40e_config_bridge_mode(veb);
9863
		}
J
Jesse Brandeburg 已提交
9864 9865 9866 9867 9868 9869 9870 9871 9872 9873 9874 9875 9876 9877 9878 9879 9880 9881
		for (i = 0; i < I40E_MAX_VEB && !veb; i++) {
			if (pf->veb[i] && pf->veb[i]->seid == vsi->uplink_seid)
				veb = pf->veb[i];
		}
		if (!veb) {
			dev_info(&pf->pdev->dev, "couldn't add VEB\n");
			return NULL;
		}

		vsi->flags |= I40E_VSI_FLAG_VEB_OWNER;
		uplink_seid = veb->seid;
	}

	/* get vsi sw struct */
	v_idx = i40e_vsi_mem_alloc(pf, type);
	if (v_idx < 0)
		goto err_alloc;
	vsi = pf->vsi[v_idx];
9882 9883
	if (!vsi)
		goto err_alloc;
J
Jesse Brandeburg 已提交
9884 9885 9886 9887 9888 9889 9890 9891
	vsi->type = type;
	vsi->veb_idx = (veb ? veb->idx : I40E_NO_VEB);

	if (type == I40E_VSI_MAIN)
		pf->lan_vsi = v_idx;
	else if (type == I40E_VSI_SRIOV)
		vsi->vf_id = param1;
	/* assign it some queues */
9892 9893
	ret = i40e_get_lump(pf, pf->qp_pile, vsi->alloc_queue_pairs,
				vsi->idx);
J
Jesse Brandeburg 已提交
9894
	if (ret < 0) {
9895 9896 9897
		dev_info(&pf->pdev->dev,
			 "failed to get tracking for %d queues for VSI %d err=%d\n",
			 vsi->alloc_queue_pairs, vsi->seid, ret);
J
Jesse Brandeburg 已提交
9898 9899 9900 9901 9902 9903 9904 9905 9906 9907 9908 9909 9910
		goto err_vsi;
	}
	vsi->base_queue = ret;

	/* get a VSI from the hardware */
	vsi->uplink_seid = uplink_seid;
	ret = i40e_add_vsi(vsi);
	if (ret)
		goto err_vsi;

	switch (vsi->type) {
	/* setup the netdev if needed */
	case I40E_VSI_MAIN:
9911 9912 9913 9914 9915 9916 9917 9918 9919 9920 9921
		/* Apply relevant filters if a platform-specific mac
		 * address was selected.
		 */
		if (!!(pf->flags & I40E_FLAG_PF_MAC)) {
			ret = i40e_macaddr_init(vsi, pf->hw.mac.addr);
			if (ret) {
				dev_warn(&pf->pdev->dev,
					 "could not set up macaddr; err %d\n",
					 ret);
			}
		}
J
Jesse Brandeburg 已提交
9922
	case I40E_VSI_VMDQ2:
9923
	case I40E_VSI_FCOE:
J
Jesse Brandeburg 已提交
9924 9925 9926 9927 9928 9929 9930 9931
		ret = i40e_config_netdev(vsi);
		if (ret)
			goto err_netdev;
		ret = register_netdev(vsi->netdev);
		if (ret)
			goto err_netdev;
		vsi->netdev_registered = true;
		netif_carrier_off(vsi->netdev);
N
Neerav Parikh 已提交
9932 9933 9934 9935
#ifdef CONFIG_I40E_DCB
		/* Setup DCB netlink interface */
		i40e_dcbnl_setup(vsi);
#endif /* CONFIG_I40E_DCB */
J
Jesse Brandeburg 已提交
9936 9937 9938 9939 9940 9941 9942 9943 9944 9945 9946 9947 9948 9949 9950 9951 9952 9953 9954 9955 9956 9957 9958
		/* fall through */

	case I40E_VSI_FDIR:
		/* set up vectors and rings if needed */
		ret = i40e_vsi_setup_vectors(vsi);
		if (ret)
			goto err_msix;

		ret = i40e_alloc_rings(vsi);
		if (ret)
			goto err_rings;

		/* map all of the rings to the q_vectors */
		i40e_vsi_map_rings_to_vectors(vsi);

		i40e_vsi_reset_stats(vsi);
		break;

	default:
		/* no netdev or rings for the other VSI types */
		break;
	}

9959 9960 9961 9962
	if ((pf->flags & I40E_FLAG_RSS_AQ_CAPABLE) &&
	    (vsi->type == I40E_VSI_VMDQ2)) {
		ret = i40e_vsi_config_rss(vsi);
	}
J
Jesse Brandeburg 已提交
9963 9964 9965 9966 9967 9968 9969 9970 9971 9972 9973 9974 9975 9976 9977 9978 9979 9980 9981 9982 9983 9984 9985 9986 9987 9988 9989 9990 9991 9992 9993 9994 9995 9996 9997 9998 9999 10000 10001
	return vsi;

err_rings:
	i40e_vsi_free_q_vectors(vsi);
err_msix:
	if (vsi->netdev_registered) {
		vsi->netdev_registered = false;
		unregister_netdev(vsi->netdev);
		free_netdev(vsi->netdev);
		vsi->netdev = NULL;
	}
err_netdev:
	i40e_aq_delete_element(&pf->hw, vsi->seid, NULL);
err_vsi:
	i40e_vsi_clear(vsi);
err_alloc:
	return NULL;
}

/**
 * i40e_veb_get_bw_info - Query VEB BW information
 * @veb: the veb to query
 *
 * Query the Tx scheduler BW configuration data for given VEB
 **/
static int i40e_veb_get_bw_info(struct i40e_veb *veb)
{
	struct i40e_aqc_query_switching_comp_ets_config_resp ets_data;
	struct i40e_aqc_query_switching_comp_bw_config_resp bw_data;
	struct i40e_pf *pf = veb->pf;
	struct i40e_hw *hw = &pf->hw;
	u32 tc_bw_max;
	int ret = 0;
	int i;

	ret = i40e_aq_query_switch_comp_bw_config(hw, veb->seid,
						  &bw_data, NULL);
	if (ret) {
		dev_info(&pf->pdev->dev,
10002 10003 10004
			 "query veb bw config failed, err %s aq_err %s\n",
			 i40e_stat_str(&pf->hw, ret),
			 i40e_aq_str(&pf->hw, hw->aq.asq_last_status));
J
Jesse Brandeburg 已提交
10005 10006 10007 10008 10009 10010 10011
		goto out;
	}

	ret = i40e_aq_query_switch_comp_ets_config(hw, veb->seid,
						   &ets_data, NULL);
	if (ret) {
		dev_info(&pf->pdev->dev,
10012 10013 10014
			 "query veb bw ets config failed, err %s aq_err %s\n",
			 i40e_stat_str(&pf->hw, ret),
			 i40e_aq_str(&pf->hw, hw->aq.asq_last_status));
J
Jesse Brandeburg 已提交
10015 10016 10017 10018 10019 10020
		goto out;
	}

	veb->bw_limit = le16_to_cpu(ets_data.port_bw_limit);
	veb->bw_max_quanta = ets_data.tc_bw_max;
	veb->is_abs_credits = bw_data.absolute_credits_enable;
10021
	veb->enabled_tc = ets_data.tc_valid_bits;
J
Jesse Brandeburg 已提交
10022 10023 10024 10025 10026 10027 10028 10029 10030 10031 10032 10033 10034 10035 10036 10037 10038 10039 10040 10041 10042 10043 10044 10045 10046 10047 10048 10049 10050 10051 10052 10053 10054 10055 10056 10057 10058 10059 10060 10061 10062 10063 10064 10065 10066 10067 10068 10069 10070 10071 10072 10073 10074 10075 10076 10077 10078 10079 10080 10081 10082 10083 10084 10085 10086 10087 10088 10089 10090 10091 10092 10093 10094 10095 10096 10097 10098 10099 10100 10101 10102 10103 10104 10105 10106 10107
	tc_bw_max = le16_to_cpu(bw_data.tc_bw_max[0]) |
		    (le16_to_cpu(bw_data.tc_bw_max[1]) << 16);
	for (i = 0; i < I40E_MAX_TRAFFIC_CLASS; i++) {
		veb->bw_tc_share_credits[i] = bw_data.tc_bw_share_credits[i];
		veb->bw_tc_limit_credits[i] =
					le16_to_cpu(bw_data.tc_bw_limits[i]);
		veb->bw_tc_max_quanta[i] = ((tc_bw_max >> (i*4)) & 0x7);
	}

out:
	return ret;
}

/**
 * i40e_veb_mem_alloc - Allocates the next available struct veb in the PF
 * @pf: board private structure
 *
 * On error: returns error code (negative)
 * On success: returns vsi index in PF (positive)
 **/
static int i40e_veb_mem_alloc(struct i40e_pf *pf)
{
	int ret = -ENOENT;
	struct i40e_veb *veb;
	int i;

	/* Need to protect the allocation of switch elements at the PF level */
	mutex_lock(&pf->switch_mutex);

	/* VEB list may be fragmented if VEB creation/destruction has
	 * been happening.  We can afford to do a quick scan to look
	 * for any free slots in the list.
	 *
	 * find next empty veb slot, looping back around if necessary
	 */
	i = 0;
	while ((i < I40E_MAX_VEB) && (pf->veb[i] != NULL))
		i++;
	if (i >= I40E_MAX_VEB) {
		ret = -ENOMEM;
		goto err_alloc_veb;  /* out of VEB slots! */
	}

	veb = kzalloc(sizeof(*veb), GFP_KERNEL);
	if (!veb) {
		ret = -ENOMEM;
		goto err_alloc_veb;
	}
	veb->pf = pf;
	veb->idx = i;
	veb->enabled_tc = 1;

	pf->veb[i] = veb;
	ret = i;
err_alloc_veb:
	mutex_unlock(&pf->switch_mutex);
	return ret;
}

/**
 * i40e_switch_branch_release - Delete a branch of the switch tree
 * @branch: where to start deleting
 *
 * This uses recursion to find the tips of the branch to be
 * removed, deleting until we get back to and can delete this VEB.
 **/
static void i40e_switch_branch_release(struct i40e_veb *branch)
{
	struct i40e_pf *pf = branch->pf;
	u16 branch_seid = branch->seid;
	u16 veb_idx = branch->idx;
	int i;

	/* release any VEBs on this VEB - RECURSION */
	for (i = 0; i < I40E_MAX_VEB; i++) {
		if (!pf->veb[i])
			continue;
		if (pf->veb[i]->uplink_seid == branch->seid)
			i40e_switch_branch_release(pf->veb[i]);
	}

	/* Release the VSIs on this VEB, but not the owner VSI.
	 *
	 * NOTE: Removing the last VSI on a VEB has the SIDE EFFECT of removing
	 *       the VEB itself, so don't use (*branch) after this loop.
	 */
M
Mitch Williams 已提交
10108
	for (i = 0; i < pf->num_alloc_vsi; i++) {
J
Jesse Brandeburg 已提交
10109 10110 10111 10112 10113 10114 10115 10116 10117 10118 10119 10120 10121 10122 10123 10124 10125 10126 10127 10128 10129 10130 10131 10132 10133 10134 10135 10136 10137 10138 10139 10140 10141 10142 10143 10144 10145 10146 10147 10148 10149 10150 10151 10152 10153 10154 10155 10156 10157 10158 10159
		if (!pf->vsi[i])
			continue;
		if (pf->vsi[i]->uplink_seid == branch_seid &&
		   (pf->vsi[i]->flags & I40E_VSI_FLAG_VEB_OWNER) == 0) {
			i40e_vsi_release(pf->vsi[i]);
		}
	}

	/* There's one corner case where the VEB might not have been
	 * removed, so double check it here and remove it if needed.
	 * This case happens if the veb was created from the debugfs
	 * commands and no VSIs were added to it.
	 */
	if (pf->veb[veb_idx])
		i40e_veb_release(pf->veb[veb_idx]);
}

/**
 * i40e_veb_clear - remove veb struct
 * @veb: the veb to remove
 **/
static void i40e_veb_clear(struct i40e_veb *veb)
{
	if (!veb)
		return;

	if (veb->pf) {
		struct i40e_pf *pf = veb->pf;

		mutex_lock(&pf->switch_mutex);
		if (pf->veb[veb->idx] == veb)
			pf->veb[veb->idx] = NULL;
		mutex_unlock(&pf->switch_mutex);
	}

	kfree(veb);
}

/**
 * i40e_veb_release - Delete a VEB and free its resources
 * @veb: the VEB being removed
 **/
void i40e_veb_release(struct i40e_veb *veb)
{
	struct i40e_vsi *vsi = NULL;
	struct i40e_pf *pf;
	int i, n = 0;

	pf = veb->pf;

	/* find the remaining VSI and check for extras */
M
Mitch Williams 已提交
10160
	for (i = 0; i < pf->num_alloc_vsi; i++) {
J
Jesse Brandeburg 已提交
10161 10162 10163 10164 10165 10166 10167 10168 10169 10170 10171 10172 10173 10174 10175 10176 10177 10178 10179 10180 10181 10182 10183 10184 10185 10186 10187 10188 10189 10190 10191 10192 10193 10194 10195 10196 10197
		if (pf->vsi[i] && pf->vsi[i]->uplink_seid == veb->seid) {
			n++;
			vsi = pf->vsi[i];
		}
	}
	if (n != 1) {
		dev_info(&pf->pdev->dev,
			 "can't remove VEB %d with %d VSIs left\n",
			 veb->seid, n);
		return;
	}

	/* move the remaining VSI to uplink veb */
	vsi->flags &= ~I40E_VSI_FLAG_VEB_OWNER;
	if (veb->uplink_seid) {
		vsi->uplink_seid = veb->uplink_seid;
		if (veb->uplink_seid == pf->mac_seid)
			vsi->veb_idx = I40E_NO_VEB;
		else
			vsi->veb_idx = veb->veb_idx;
	} else {
		/* floating VEB */
		vsi->uplink_seid = pf->vsi[pf->lan_vsi]->uplink_seid;
		vsi->veb_idx = pf->vsi[pf->lan_vsi]->veb_idx;
	}

	i40e_aq_delete_element(&pf->hw, veb->seid, NULL);
	i40e_veb_clear(veb);
}

/**
 * i40e_add_veb - create the VEB in the switch
 * @veb: the VEB to be instantiated
 * @vsi: the controlling VSI
 **/
static int i40e_add_veb(struct i40e_veb *veb, struct i40e_vsi *vsi)
{
10198
	struct i40e_pf *pf = veb->pf;
10199
	bool enable_stats = !!(pf->flags & I40E_FLAG_VEB_STATS_ENABLED);
J
Jesse Brandeburg 已提交
10200 10201
	int ret;

10202
	ret = i40e_aq_add_veb(&pf->hw, veb->uplink_seid, vsi->seid,
10203
			      veb->enabled_tc, false,
10204
			      &veb->seid, enable_stats, NULL);
10205 10206

	/* get a VEB from the hardware */
J
Jesse Brandeburg 已提交
10207
	if (ret) {
10208 10209 10210 10211
		dev_info(&pf->pdev->dev,
			 "couldn't add VEB, err %s aq_err %s\n",
			 i40e_stat_str(&pf->hw, ret),
			 i40e_aq_str(&pf->hw, pf->hw.aq.asq_last_status));
J
Jesse Brandeburg 已提交
10212 10213 10214 10215
		return -EPERM;
	}

	/* get statistics counter */
10216
	ret = i40e_aq_get_veb_parameters(&pf->hw, veb->seid, NULL, NULL,
J
Jesse Brandeburg 已提交
10217 10218
					 &veb->stats_idx, NULL, NULL, NULL);
	if (ret) {
10219 10220 10221 10222
		dev_info(&pf->pdev->dev,
			 "couldn't get VEB statistics idx, err %s aq_err %s\n",
			 i40e_stat_str(&pf->hw, ret),
			 i40e_aq_str(&pf->hw, pf->hw.aq.asq_last_status));
J
Jesse Brandeburg 已提交
10223 10224 10225 10226
		return -EPERM;
	}
	ret = i40e_veb_get_bw_info(veb);
	if (ret) {
10227 10228 10229 10230 10231
		dev_info(&pf->pdev->dev,
			 "couldn't get VEB bw info, err %s aq_err %s\n",
			 i40e_stat_str(&pf->hw, ret),
			 i40e_aq_str(&pf->hw, pf->hw.aq.asq_last_status));
		i40e_aq_delete_element(&pf->hw, veb->seid, NULL);
J
Jesse Brandeburg 已提交
10232 10233 10234 10235 10236 10237 10238 10239 10240 10241 10242 10243 10244 10245 10246 10247 10248 10249 10250 10251 10252 10253 10254 10255 10256 10257 10258 10259 10260 10261 10262 10263 10264 10265 10266 10267 10268 10269 10270 10271 10272 10273 10274 10275
		return -ENOENT;
	}

	vsi->uplink_seid = veb->seid;
	vsi->veb_idx = veb->idx;
	vsi->flags |= I40E_VSI_FLAG_VEB_OWNER;

	return 0;
}

/**
 * i40e_veb_setup - Set up a VEB
 * @pf: board private structure
 * @flags: VEB setup flags
 * @uplink_seid: the switch element to link to
 * @vsi_seid: the initial VSI seid
 * @enabled_tc: Enabled TC bit-map
 *
 * This allocates the sw VEB structure and links it into the switch
 * It is possible and legal for this to be a duplicate of an already
 * existing VEB.  It is also possible for both uplink and vsi seids
 * to be zero, in order to create a floating VEB.
 *
 * Returns pointer to the successfully allocated VEB sw struct on
 * success, otherwise returns NULL on failure.
 **/
struct i40e_veb *i40e_veb_setup(struct i40e_pf *pf, u16 flags,
				u16 uplink_seid, u16 vsi_seid,
				u8 enabled_tc)
{
	struct i40e_veb *veb, *uplink_veb = NULL;
	int vsi_idx, veb_idx;
	int ret;

	/* if one seid is 0, the other must be 0 to create a floating relay */
	if ((uplink_seid == 0 || vsi_seid == 0) &&
	    (uplink_seid + vsi_seid != 0)) {
		dev_info(&pf->pdev->dev,
			 "one, not both seid's are 0: uplink=%d vsi=%d\n",
			 uplink_seid, vsi_seid);
		return NULL;
	}

	/* make sure there is such a vsi and uplink */
M
Mitch Williams 已提交
10276
	for (vsi_idx = 0; vsi_idx < pf->num_alloc_vsi; vsi_idx++)
J
Jesse Brandeburg 已提交
10277 10278
		if (pf->vsi[vsi_idx] && pf->vsi[vsi_idx]->seid == vsi_seid)
			break;
M
Mitch Williams 已提交
10279
	if (vsi_idx >= pf->num_alloc_vsi && vsi_seid != 0) {
J
Jesse Brandeburg 已提交
10280 10281 10282 10283 10284 10285 10286 10287 10288 10289 10290 10291 10292 10293 10294 10295 10296 10297 10298 10299 10300 10301 10302 10303 10304 10305 10306 10307 10308 10309 10310 10311 10312 10313
		dev_info(&pf->pdev->dev, "vsi seid %d not found\n",
			 vsi_seid);
		return NULL;
	}

	if (uplink_seid && uplink_seid != pf->mac_seid) {
		for (veb_idx = 0; veb_idx < I40E_MAX_VEB; veb_idx++) {
			if (pf->veb[veb_idx] &&
			    pf->veb[veb_idx]->seid == uplink_seid) {
				uplink_veb = pf->veb[veb_idx];
				break;
			}
		}
		if (!uplink_veb) {
			dev_info(&pf->pdev->dev,
				 "uplink seid %d not found\n", uplink_seid);
			return NULL;
		}
	}

	/* get veb sw struct */
	veb_idx = i40e_veb_mem_alloc(pf);
	if (veb_idx < 0)
		goto err_alloc;
	veb = pf->veb[veb_idx];
	veb->flags = flags;
	veb->uplink_seid = uplink_seid;
	veb->veb_idx = (uplink_veb ? uplink_veb->idx : I40E_NO_VEB);
	veb->enabled_tc = (enabled_tc ? enabled_tc : 0x1);

	/* create the VEB in the switch */
	ret = i40e_add_veb(veb, pf->vsi[vsi_idx]);
	if (ret)
		goto err_veb;
S
Shannon Nelson 已提交
10314 10315
	if (vsi_idx == pf->lan_vsi)
		pf->lan_veb = veb->idx;
J
Jesse Brandeburg 已提交
10316 10317 10318 10319 10320 10321 10322 10323 10324 10325

	return veb;

err_veb:
	i40e_veb_clear(veb);
err_alloc:
	return NULL;
}

/**
10326
 * i40e_setup_pf_switch_element - set PF vars based on switch type
J
Jesse Brandeburg 已提交
10327 10328 10329 10330 10331 10332 10333 10334 10335 10336 10337 10338 10339 10340 10341 10342 10343 10344 10345 10346 10347 10348 10349 10350 10351 10352 10353 10354 10355 10356 10357 10358 10359 10360 10361 10362 10363 10364 10365 10366 10367 10368 10369 10370 10371 10372 10373 10374 10375 10376 10377 10378 10379 10380 10381 10382 10383 10384 10385 10386 10387 10388 10389 10390 10391 10392 10393 10394 10395 10396 10397 10398 10399 10400 10401 10402 10403 10404 10405 10406 10407 10408 10409 10410 10411 10412 10413 10414 10415 10416 10417 10418 10419 10420 10421 10422 10423 10424 10425 10426 10427 10428 10429 10430 10431 10432 10433 10434 10435 10436
 * @pf: board private structure
 * @ele: element we are building info from
 * @num_reported: total number of elements
 * @printconfig: should we print the contents
 *
 * helper function to assist in extracting a few useful SEID values.
 **/
static void i40e_setup_pf_switch_element(struct i40e_pf *pf,
				struct i40e_aqc_switch_config_element_resp *ele,
				u16 num_reported, bool printconfig)
{
	u16 downlink_seid = le16_to_cpu(ele->downlink_seid);
	u16 uplink_seid = le16_to_cpu(ele->uplink_seid);
	u8 element_type = ele->element_type;
	u16 seid = le16_to_cpu(ele->seid);

	if (printconfig)
		dev_info(&pf->pdev->dev,
			 "type=%d seid=%d uplink=%d downlink=%d\n",
			 element_type, seid, uplink_seid, downlink_seid);

	switch (element_type) {
	case I40E_SWITCH_ELEMENT_TYPE_MAC:
		pf->mac_seid = seid;
		break;
	case I40E_SWITCH_ELEMENT_TYPE_VEB:
		/* Main VEB? */
		if (uplink_seid != pf->mac_seid)
			break;
		if (pf->lan_veb == I40E_NO_VEB) {
			int v;

			/* find existing or else empty VEB */
			for (v = 0; v < I40E_MAX_VEB; v++) {
				if (pf->veb[v] && (pf->veb[v]->seid == seid)) {
					pf->lan_veb = v;
					break;
				}
			}
			if (pf->lan_veb == I40E_NO_VEB) {
				v = i40e_veb_mem_alloc(pf);
				if (v < 0)
					break;
				pf->lan_veb = v;
			}
		}

		pf->veb[pf->lan_veb]->seid = seid;
		pf->veb[pf->lan_veb]->uplink_seid = pf->mac_seid;
		pf->veb[pf->lan_veb]->pf = pf;
		pf->veb[pf->lan_veb]->veb_idx = I40E_NO_VEB;
		break;
	case I40E_SWITCH_ELEMENT_TYPE_VSI:
		if (num_reported != 1)
			break;
		/* This is immediately after a reset so we can assume this is
		 * the PF's VSI
		 */
		pf->mac_seid = uplink_seid;
		pf->pf_seid = downlink_seid;
		pf->main_vsi_seid = seid;
		if (printconfig)
			dev_info(&pf->pdev->dev,
				 "pf_seid=%d main_vsi_seid=%d\n",
				 pf->pf_seid, pf->main_vsi_seid);
		break;
	case I40E_SWITCH_ELEMENT_TYPE_PF:
	case I40E_SWITCH_ELEMENT_TYPE_VF:
	case I40E_SWITCH_ELEMENT_TYPE_EMP:
	case I40E_SWITCH_ELEMENT_TYPE_BMC:
	case I40E_SWITCH_ELEMENT_TYPE_PE:
	case I40E_SWITCH_ELEMENT_TYPE_PA:
		/* ignore these for now */
		break;
	default:
		dev_info(&pf->pdev->dev, "unknown element type=%d seid=%d\n",
			 element_type, seid);
		break;
	}
}

/**
 * i40e_fetch_switch_configuration - Get switch config from firmware
 * @pf: board private structure
 * @printconfig: should we print the contents
 *
 * Get the current switch configuration from the device and
 * extract a few useful SEID values.
 **/
int i40e_fetch_switch_configuration(struct i40e_pf *pf, bool printconfig)
{
	struct i40e_aqc_get_switch_config_resp *sw_config;
	u16 next_seid = 0;
	int ret = 0;
	u8 *aq_buf;
	int i;

	aq_buf = kzalloc(I40E_AQ_LARGE_BUF, GFP_KERNEL);
	if (!aq_buf)
		return -ENOMEM;

	sw_config = (struct i40e_aqc_get_switch_config_resp *)aq_buf;
	do {
		u16 num_reported, num_total;

		ret = i40e_aq_get_switch_config(&pf->hw, sw_config,
						I40E_AQ_LARGE_BUF,
						&next_seid, NULL);
		if (ret) {
			dev_info(&pf->pdev->dev,
10437 10438 10439 10440
				 "get switch config failed err %s aq_err %s\n",
				 i40e_stat_str(&pf->hw, ret),
				 i40e_aq_str(&pf->hw,
					     pf->hw.aq.asq_last_status));
J
Jesse Brandeburg 已提交
10441 10442 10443 10444 10445 10446 10447 10448 10449 10450 10451 10452 10453 10454 10455 10456 10457 10458 10459 10460 10461 10462 10463 10464 10465 10466 10467 10468
			kfree(aq_buf);
			return -ENOENT;
		}

		num_reported = le16_to_cpu(sw_config->header.num_reported);
		num_total = le16_to_cpu(sw_config->header.num_total);

		if (printconfig)
			dev_info(&pf->pdev->dev,
				 "header: %d reported %d total\n",
				 num_reported, num_total);

		for (i = 0; i < num_reported; i++) {
			struct i40e_aqc_switch_config_element_resp *ele =
				&sw_config->element[i];

			i40e_setup_pf_switch_element(pf, ele, num_reported,
						     printconfig);
		}
	} while (next_seid != 0);

	kfree(aq_buf);
	return ret;
}

/**
 * i40e_setup_pf_switch - Setup the HW switch on startup or after reset
 * @pf: board private structure
10469
 * @reinit: if the Main VSI needs to re-initialized.
J
Jesse Brandeburg 已提交
10470 10471 10472
 *
 * Returns 0 on success, negative value on failure
 **/
10473
static int i40e_setup_pf_switch(struct i40e_pf *pf, bool reinit)
J
Jesse Brandeburg 已提交
10474
{
10475
	u16 flags = 0;
J
Jesse Brandeburg 已提交
10476 10477 10478 10479 10480 10481
	int ret;

	/* find out what's out there already */
	ret = i40e_fetch_switch_configuration(pf, false);
	if (ret) {
		dev_info(&pf->pdev->dev,
10482 10483 10484
			 "couldn't fetch switch config, err %s aq_err %s\n",
			 i40e_stat_str(&pf->hw, ret),
			 i40e_aq_str(&pf->hw, pf->hw.aq.asq_last_status));
J
Jesse Brandeburg 已提交
10485 10486 10487 10488
		return ret;
	}
	i40e_pf_reset_stats(pf);

10489 10490 10491 10492 10493 10494 10495 10496 10497 10498 10499 10500 10501 10502 10503 10504 10505 10506 10507 10508 10509 10510 10511 10512 10513 10514
	/* set the switch config bit for the whole device to
	 * support limited promisc or true promisc
	 * when user requests promisc. The default is limited
	 * promisc.
	*/

	if ((pf->hw.pf_id == 0) &&
	    !(pf->flags & I40E_FLAG_TRUE_PROMISC_SUPPORT))
		flags = I40E_AQ_SET_SWITCH_CFG_PROMISC;

	if (pf->hw.pf_id == 0) {
		u16 valid_flags;

		valid_flags = I40E_AQ_SET_SWITCH_CFG_PROMISC;
		ret = i40e_aq_set_switch_config(&pf->hw, flags, valid_flags,
						NULL);
		if (ret && pf->hw.aq.asq_last_status != I40E_AQ_RC_ESRCH) {
			dev_info(&pf->pdev->dev,
				 "couldn't set switch config bits, err %s aq_err %s\n",
				 i40e_stat_str(&pf->hw, ret),
				 i40e_aq_str(&pf->hw,
					     pf->hw.aq.asq_last_status));
			/* not a fatal problem, just keep going */
		}
	}

J
Jesse Brandeburg 已提交
10515
	/* first time setup */
10516
	if (pf->lan_vsi == I40E_NO_VSI || reinit) {
J
Jesse Brandeburg 已提交
10517 10518 10519 10520 10521 10522 10523 10524 10525 10526
		struct i40e_vsi *vsi = NULL;
		u16 uplink_seid;

		/* Set up the PF VSI associated with the PF's main VSI
		 * that is already in the HW switch
		 */
		if (pf->lan_veb != I40E_NO_VEB && pf->veb[pf->lan_veb])
			uplink_seid = pf->veb[pf->lan_veb]->seid;
		else
			uplink_seid = pf->mac_seid;
10527 10528 10529 10530
		if (pf->lan_vsi == I40E_NO_VSI)
			vsi = i40e_vsi_setup(pf, I40E_VSI_MAIN, uplink_seid, 0);
		else if (reinit)
			vsi = i40e_vsi_reinit_setup(pf->vsi[pf->lan_vsi]);
J
Jesse Brandeburg 已提交
10531 10532 10533 10534 10535 10536 10537 10538
		if (!vsi) {
			dev_info(&pf->pdev->dev, "setup of MAIN VSI failed\n");
			i40e_fdir_teardown(pf);
			return -EAGAIN;
		}
	} else {
		/* force a reset of TC and queue layout configurations */
		u8 enabled_tc = pf->vsi[pf->lan_vsi]->tc_config.enabled_tc;
J
Jesse Brandeburg 已提交
10539

J
Jesse Brandeburg 已提交
10540 10541 10542 10543 10544 10545
		pf->vsi[pf->lan_vsi]->tc_config.enabled_tc = 0;
		pf->vsi[pf->lan_vsi]->seid = pf->main_vsi_seid;
		i40e_vsi_config_tc(pf->vsi[pf->lan_vsi], enabled_tc);
	}
	i40e_vlan_stripping_disable(pf->vsi[pf->lan_vsi]);

10546 10547
	i40e_fdir_sb_setup(pf);

J
Jesse Brandeburg 已提交
10548 10549 10550 10551 10552 10553 10554 10555 10556 10557 10558 10559
	/* Setup static PF queue filter control settings */
	ret = i40e_setup_pf_filter_control(pf);
	if (ret) {
		dev_info(&pf->pdev->dev, "setup_pf_filter_control failed: %d\n",
			 ret);
		/* Failure here should not stop continuing other steps */
	}

	/* enable RSS in the HW, even for only one queue, as the stack can use
	 * the hash
	 */
	if ((pf->flags & I40E_FLAG_RSS_ENABLED))
10560
		i40e_pf_config_rss(pf);
J
Jesse Brandeburg 已提交
10561 10562

	/* fill in link information and enable LSE reporting */
10563
	i40e_update_link_info(&pf->hw);
10564 10565
	i40e_link_event(pf);

10566
	/* Initialize user-specific link properties */
J
Jesse Brandeburg 已提交
10567 10568
	pf->fc_autoneg_status = ((pf->hw.phy.link_info.an_info &
				  I40E_AQ_AN_COMPLETED) ? true : false);
10569

J
Jacob Keller 已提交
10570 10571
	i40e_ptp_init(pf);

J
Jesse Brandeburg 已提交
10572 10573 10574 10575 10576 10577 10578 10579 10580 10581 10582 10583
	return ret;
}

/**
 * i40e_determine_queue_usage - Work out queue distribution
 * @pf: board private structure
 **/
static void i40e_determine_queue_usage(struct i40e_pf *pf)
{
	int queues_left;

	pf->num_lan_qps = 0;
10584 10585 10586
#ifdef I40E_FCOE
	pf->num_fcoe_qps = 0;
#endif
J
Jesse Brandeburg 已提交
10587 10588 10589 10590 10591 10592 10593

	/* Find the max queues to be put into basic use.  We'll always be
	 * using TC0, whether or not DCB is running, and TC0 will get the
	 * big RSS set.
	 */
	queues_left = pf->hw.func_caps.num_tx_qp;

10594
	if ((queues_left == 1) ||
10595
	    !(pf->flags & I40E_FLAG_MSIX_ENABLED)) {
J
Jesse Brandeburg 已提交
10596 10597
		/* one qp for PF, no queues for anything else */
		queues_left = 0;
10598
		pf->alloc_rss_size = pf->num_lan_qps = 1;
J
Jesse Brandeburg 已提交
10599 10600

		/* make sure all the fancies are disabled */
J
Jesse Brandeburg 已提交
10601
		pf->flags &= ~(I40E_FLAG_RSS_ENABLED	|
10602
			       I40E_FLAG_IWARP_ENABLED	|
10603 10604 10605
#ifdef I40E_FCOE
			       I40E_FLAG_FCOE_ENABLED	|
#endif
J
Jesse Brandeburg 已提交
10606 10607
			       I40E_FLAG_FD_SB_ENABLED	|
			       I40E_FLAG_FD_ATR_ENABLED	|
10608
			       I40E_FLAG_DCB_CAPABLE	|
10609
			       I40E_FLAG_DCB_ENABLED	|
J
Jesse Brandeburg 已提交
10610 10611
			       I40E_FLAG_SRIOV_ENABLED	|
			       I40E_FLAG_VMDQ_ENABLED);
10612 10613
	} else if (!(pf->flags & (I40E_FLAG_RSS_ENABLED |
				  I40E_FLAG_FD_SB_ENABLED |
10614
				  I40E_FLAG_FD_ATR_ENABLED |
10615
				  I40E_FLAG_DCB_CAPABLE))) {
10616
		/* one qp for PF */
10617
		pf->alloc_rss_size = pf->num_lan_qps = 1;
10618 10619 10620
		queues_left -= pf->num_lan_qps;

		pf->flags &= ~(I40E_FLAG_RSS_ENABLED	|
10621
			       I40E_FLAG_IWARP_ENABLED	|
10622 10623 10624
#ifdef I40E_FCOE
			       I40E_FLAG_FCOE_ENABLED	|
#endif
10625 10626 10627 10628
			       I40E_FLAG_FD_SB_ENABLED	|
			       I40E_FLAG_FD_ATR_ENABLED	|
			       I40E_FLAG_DCB_ENABLED	|
			       I40E_FLAG_VMDQ_ENABLED);
J
Jesse Brandeburg 已提交
10629
	} else {
10630
		/* Not enough queues for all TCs */
10631
		if ((pf->flags & I40E_FLAG_DCB_CAPABLE) &&
10632
		    (queues_left < I40E_MAX_TRAFFIC_CLASS)) {
10633 10634
			pf->flags &= ~(I40E_FLAG_DCB_CAPABLE |
					I40E_FLAG_DCB_ENABLED);
10635 10636
			dev_info(&pf->pdev->dev, "not enough queues for DCB. DCB is disabled.\n");
		}
10637 10638 10639 10640 10641
		pf->num_lan_qps = max_t(int, pf->rss_size_max,
					num_online_cpus());
		pf->num_lan_qps = min_t(int, pf->num_lan_qps,
					pf->hw.func_caps.num_tx_qp);

10642 10643 10644
		queues_left -= pf->num_lan_qps;
	}

10645 10646 10647 10648 10649 10650 10651 10652 10653 10654 10655 10656 10657 10658 10659 10660
#ifdef I40E_FCOE
	if (pf->flags & I40E_FLAG_FCOE_ENABLED) {
		if (I40E_DEFAULT_FCOE <= queues_left) {
			pf->num_fcoe_qps = I40E_DEFAULT_FCOE;
		} else if (I40E_MINIMUM_FCOE <= queues_left) {
			pf->num_fcoe_qps = I40E_MINIMUM_FCOE;
		} else {
			pf->num_fcoe_qps = 0;
			pf->flags &= ~I40E_FLAG_FCOE_ENABLED;
			dev_info(&pf->pdev->dev, "not enough queues for FCoE. FCoE feature will be disabled\n");
		}

		queues_left -= pf->num_fcoe_qps;
	}

#endif
10661 10662 10663 10664 10665 10666 10667
	if (pf->flags & I40E_FLAG_FD_SB_ENABLED) {
		if (queues_left > 1) {
			queues_left -= 1; /* save 1 queue for FD */
		} else {
			pf->flags &= ~I40E_FLAG_FD_SB_ENABLED;
			dev_info(&pf->pdev->dev, "not enough queues for Flow Director. Flow Director feature is disabled\n");
		}
J
Jesse Brandeburg 已提交
10668 10669 10670 10671
	}

	if ((pf->flags & I40E_FLAG_SRIOV_ENABLED) &&
	    pf->num_vf_qps && pf->num_req_vfs && queues_left) {
10672 10673
		pf->num_req_vfs = min_t(int, pf->num_req_vfs,
					(queues_left / pf->num_vf_qps));
J
Jesse Brandeburg 已提交
10674 10675 10676 10677 10678 10679 10680 10681 10682 10683
		queues_left -= (pf->num_req_vfs * pf->num_vf_qps);
	}

	if ((pf->flags & I40E_FLAG_VMDQ_ENABLED) &&
	    pf->num_vmdq_vsis && pf->num_vmdq_qps && queues_left) {
		pf->num_vmdq_vsis = min_t(int, pf->num_vmdq_vsis,
					  (queues_left / pf->num_vmdq_qps));
		queues_left -= (pf->num_vmdq_vsis * pf->num_vmdq_qps);
	}

10684
	pf->queues_left = queues_left;
10685 10686 10687 10688
	dev_dbg(&pf->pdev->dev,
		"qs_avail=%d FD SB=%d lan_qs=%d lan_tc0=%d vf=%d*%d vmdq=%d*%d, remaining=%d\n",
		pf->hw.func_caps.num_tx_qp,
		!!(pf->flags & I40E_FLAG_FD_SB_ENABLED),
10689 10690 10691
		pf->num_lan_qps, pf->alloc_rss_size, pf->num_req_vfs,
		pf->num_vf_qps, pf->num_vmdq_vsis, pf->num_vmdq_qps,
		queues_left);
10692
#ifdef I40E_FCOE
10693
	dev_dbg(&pf->pdev->dev, "fcoe queues = %d\n", pf->num_fcoe_qps);
10694
#endif
J
Jesse Brandeburg 已提交
10695 10696 10697 10698 10699 10700
}

/**
 * i40e_setup_pf_filter_control - Setup PF static filter control
 * @pf: PF to be setup
 *
10701
 * i40e_setup_pf_filter_control sets up a PF's initial filter control
J
Jesse Brandeburg 已提交
10702 10703 10704 10705 10706 10707 10708 10709 10710 10711 10712 10713 10714
 * settings. If PE/FCoE are enabled then it will also set the per PF
 * based filter sizes required for them. It also enables Flow director,
 * ethertype and macvlan type filter settings for the pf.
 *
 * Returns 0 on success, negative on failure
 **/
static int i40e_setup_pf_filter_control(struct i40e_pf *pf)
{
	struct i40e_filter_control_settings *settings = &pf->filter_settings;

	settings->hash_lut_size = I40E_HASH_LUT_SIZE_128;

	/* Flow Director is enabled */
J
Jesse Brandeburg 已提交
10715
	if (pf->flags & (I40E_FLAG_FD_SB_ENABLED | I40E_FLAG_FD_ATR_ENABLED))
J
Jesse Brandeburg 已提交
10716 10717 10718 10719 10720 10721 10722 10723 10724 10725 10726 10727
		settings->enable_fdir = true;

	/* Ethtype and MACVLAN filters enabled for PF */
	settings->enable_ethtype = true;
	settings->enable_macvlan = true;

	if (i40e_set_filter_control(&pf->hw, settings))
		return -ENOENT;

	return 0;
}

J
Jesse Brandeburg 已提交
10728
#define INFO_STRING_LEN 255
10729
#define REMAIN(__x) (INFO_STRING_LEN - (__x))
J
Jesse Brandeburg 已提交
10730 10731 10732
static void i40e_print_features(struct i40e_pf *pf)
{
	struct i40e_hw *hw = &pf->hw;
10733 10734
	char *buf;
	int i;
J
Jesse Brandeburg 已提交
10735

10736 10737
	buf = kmalloc(INFO_STRING_LEN, GFP_KERNEL);
	if (!buf)
J
Jesse Brandeburg 已提交
10738 10739
		return;

10740
	i = snprintf(buf, INFO_STRING_LEN, "Features: PF-id[%d]", hw->pf_id);
J
Jesse Brandeburg 已提交
10741
#ifdef CONFIG_PCI_IOV
10742
	i += snprintf(&buf[i], REMAIN(i), " VFs: %d", pf->num_req_vfs);
J
Jesse Brandeburg 已提交
10743
#endif
10744
	i += snprintf(&buf[i], REMAIN(i), " VSIs: %d QP: %d",
10745
		      pf->hw.func_caps.num_vsis,
10746
		      pf->vsi[pf->lan_vsi]->num_queue_pairs);
J
Jesse Brandeburg 已提交
10747
	if (pf->flags & I40E_FLAG_RSS_ENABLED)
10748
		i += snprintf(&buf[i], REMAIN(i), " RSS");
J
Jesse Brandeburg 已提交
10749
	if (pf->flags & I40E_FLAG_FD_ATR_ENABLED)
10750
		i += snprintf(&buf[i], REMAIN(i), " FD_ATR");
10751
	if (pf->flags & I40E_FLAG_FD_SB_ENABLED) {
10752 10753
		i += snprintf(&buf[i], REMAIN(i), " FD_SB");
		i += snprintf(&buf[i], REMAIN(i), " NTUPLE");
10754
	}
10755
	if (pf->flags & I40E_FLAG_DCB_CAPABLE)
10756 10757
		i += snprintf(&buf[i], REMAIN(i), " DCB");
	i += snprintf(&buf[i], REMAIN(i), " VxLAN");
10758
	i += snprintf(&buf[i], REMAIN(i), " Geneve");
J
Jesse Brandeburg 已提交
10759
	if (pf->flags & I40E_FLAG_PTP)
10760
		i += snprintf(&buf[i], REMAIN(i), " PTP");
10761 10762
#ifdef I40E_FCOE
	if (pf->flags & I40E_FLAG_FCOE_ENABLED)
10763
		i += snprintf(&buf[i], REMAIN(i), " FCOE");
10764
#endif
10765
	if (pf->flags & I40E_FLAG_VEB_MODE_ENABLED)
10766
		i += snprintf(&buf[i], REMAIN(i), " VEB");
10767
	else
10768
		i += snprintf(&buf[i], REMAIN(i), " VEPA");
J
Jesse Brandeburg 已提交
10769

10770 10771
	dev_info(&pf->pdev->dev, "%s\n", buf);
	kfree(buf);
10772
	WARN_ON(i > INFO_STRING_LEN);
J
Jesse Brandeburg 已提交
10773 10774
}

10775 10776 10777 10778 10779 10780 10781 10782 10783 10784 10785 10786 10787 10788
/**
 * i40e_get_platform_mac_addr - get platform-specific MAC address
 *
 * @pdev: PCI device information struct
 * @pf: board private structure
 *
 * Look up the MAC address in Open Firmware  on systems that support it,
 * and use IDPROM on SPARC if no OF address is found. On return, the
 * I40E_FLAG_PF_MAC will be wset in pf->flags if a platform-specific value
 * has been selected.
 **/
static void i40e_get_platform_mac_addr(struct pci_dev *pdev, struct i40e_pf *pf)
{
	pf->flags &= ~I40E_FLAG_PF_MAC;
10789
	if (!eth_platform_get_mac_address(&pdev->dev, pf->hw.mac.addr))
10790 10791 10792
		pf->flags |= I40E_FLAG_PF_MAC;
}

J
Jesse Brandeburg 已提交
10793 10794 10795 10796 10797
/**
 * i40e_probe - Device initialization routine
 * @pdev: PCI device information struct
 * @ent: entry in i40e_pci_tbl
 *
10798 10799
 * i40e_probe initializes a PF identified by a pci_dev structure.
 * The OS initialization, configuring of the PF private structure,
J
Jesse Brandeburg 已提交
10800 10801 10802 10803 10804 10805
 * and a hardware reset occur.
 *
 * Returns 0 on success, negative on failure
 **/
static int i40e_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
{
10806
	struct i40e_aq_get_phy_abilities_resp abilities;
J
Jesse Brandeburg 已提交
10807 10808
	struct i40e_pf *pf;
	struct i40e_hw *hw;
10809
	static u16 pfs_found;
10810
	u16 wol_nvm_bits;
10811
	u16 link_status;
10812
	int err;
10813
	u32 val;
10814
	u32 i;
10815
	u8 set_fc_aq_fail;
J
Jesse Brandeburg 已提交
10816 10817 10818 10819 10820 10821

	err = pci_enable_device_mem(pdev);
	if (err)
		return err;

	/* set up for high or low dma */
10822 10823
	err = dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(64));
	if (err) {
10824 10825 10826 10827 10828 10829
		err = dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(32));
		if (err) {
			dev_err(&pdev->dev,
				"DMA configuration failed: 0x%x\n", err);
			goto err_dma;
		}
J
Jesse Brandeburg 已提交
10830 10831 10832
	}

	/* set up pci connections */
10833
	err = pci_request_mem_regions(pdev, i40e_driver_name);
J
Jesse Brandeburg 已提交
10834 10835 10836 10837 10838 10839 10840 10841 10842 10843 10844 10845 10846 10847 10848 10849 10850 10851 10852 10853 10854 10855 10856 10857 10858
	if (err) {
		dev_info(&pdev->dev,
			 "pci_request_selected_regions failed %d\n", err);
		goto err_pci_reg;
	}

	pci_enable_pcie_error_reporting(pdev);
	pci_set_master(pdev);

	/* Now that we have a PCI connection, we need to do the
	 * low level device setup.  This is primarily setting up
	 * the Admin Queue structures and then querying for the
	 * device's current profile information.
	 */
	pf = kzalloc(sizeof(*pf), GFP_KERNEL);
	if (!pf) {
		err = -ENOMEM;
		goto err_pf_alloc;
	}
	pf->next_vsi = 0;
	pf->pdev = pdev;
	set_bit(__I40E_DOWN, &pf->state);

	hw = &pf->hw;
	hw->back = pf;
A
Anjali Singhai 已提交
10859

S
Shannon Nelson 已提交
10860 10861
	pf->ioremap_len = min_t(int, pci_resource_len(pdev, 0),
				I40E_MAX_CSR_SPACE);
A
Anjali Singhai 已提交
10862

S
Shannon Nelson 已提交
10863
	hw->hw_addr = ioremap(pci_resource_start(pdev, 0), pf->ioremap_len);
J
Jesse Brandeburg 已提交
10864 10865 10866 10867
	if (!hw->hw_addr) {
		err = -EIO;
		dev_info(&pdev->dev, "ioremap(0x%04x, 0x%04x) failed: 0x%x\n",
			 (unsigned int)pci_resource_start(pdev, 0),
S
Shannon Nelson 已提交
10868
			 pf->ioremap_len, err);
J
Jesse Brandeburg 已提交
10869 10870 10871 10872 10873 10874 10875 10876 10877
		goto err_ioremap;
	}
	hw->vendor_id = pdev->vendor;
	hw->device_id = pdev->device;
	pci_read_config_byte(pdev, PCI_REVISION_ID, &hw->revision_id);
	hw->subsystem_vendor_id = pdev->subsystem_vendor;
	hw->subsystem_device_id = pdev->subsystem_device;
	hw->bus.device = PCI_SLOT(pdev->devfn);
	hw->bus.func = PCI_FUNC(pdev->devfn);
10878
	pf->instance = pfs_found;
J
Jesse Brandeburg 已提交
10879

10880 10881 10882 10883 10884 10885
	/* set up the locks for the AQ, do this only once in probe
	 * and destroy them only once in remove
	 */
	mutex_init(&hw->aq.asq_mutex);
	mutex_init(&hw->aq.arq_mutex);

10886 10887 10888 10889 10890 10891
	pf->msg_enable = netif_msg_init(debug,
					NETIF_MSG_DRV |
					NETIF_MSG_PROBE |
					NETIF_MSG_LINK);
	if (debug < -1)
		pf->hw.debug_mask = debug;
S
Shannon Nelson 已提交
10892

10893 10894 10895 10896 10897 10898 10899 10900 10901 10902 10903
	/* do a special CORER for clearing PXE mode once at init */
	if (hw->revision_id == 0 &&
	    (rd32(hw, I40E_GLLAN_RCTL_0) & I40E_GLLAN_RCTL_0_PXE_MODE_MASK)) {
		wr32(hw, I40E_GLGEN_RTRIG, I40E_GLGEN_RTRIG_CORER_MASK);
		i40e_flush(hw);
		msleep(200);
		pf->corer_count++;

		i40e_clear_pxe_mode(hw);
	}

J
Jesse Brandeburg 已提交
10904
	/* Reset here to make sure all is clean and to define PF 'n' */
10905
	i40e_clear_hw(hw);
J
Jesse Brandeburg 已提交
10906 10907 10908 10909 10910 10911 10912 10913 10914 10915 10916 10917
	err = i40e_pf_reset(hw);
	if (err) {
		dev_info(&pdev->dev, "Initial pf_reset failed: %d\n", err);
		goto err_pf_reset;
	}
	pf->pfr_count++;

	hw->aq.num_arq_entries = I40E_AQ_LEN;
	hw->aq.num_asq_entries = I40E_AQ_LEN;
	hw->aq.arq_buf_size = I40E_MAX_AQ_BUF_SIZE;
	hw->aq.asq_buf_size = I40E_MAX_AQ_BUF_SIZE;
	pf->adminq_work_limit = I40E_AQ_WORK_LIMIT;
10918

10919
	snprintf(pf->int_name, sizeof(pf->int_name) - 1,
10920 10921
		 "%s-%s:misc",
		 dev_driver_string(&pf->pdev->dev), dev_name(&pdev->dev));
J
Jesse Brandeburg 已提交
10922 10923 10924

	err = i40e_init_shared_code(hw);
	if (err) {
10925 10926
		dev_warn(&pdev->dev, "unidentified MAC or BLANK NVM: %d\n",
			 err);
J
Jesse Brandeburg 已提交
10927 10928 10929
		goto err_pf_reset;
	}

10930 10931 10932
	/* set up a default setting for link flow control */
	pf->hw.fc.requested_mode = I40E_FC_NONE;

J
Jesse Brandeburg 已提交
10933
	err = i40e_init_adminq(hw);
C
Carolyn Wyborny 已提交
10934 10935 10936 10937 10938 10939 10940 10941 10942 10943
	if (err) {
		if (err == I40E_ERR_FIRMWARE_API_VERSION)
			dev_info(&pdev->dev,
				 "The driver for the device stopped because the NVM image is newer than expected. You must install the most recent version of the network driver.\n");
		else
			dev_info(&pdev->dev,
				 "The driver for the device stopped because the device firmware failed to init. Try updating your NVM image.\n");

		goto err_pf_reset;
	}
10944

10945 10946 10947 10948 10949
	/* provide nvm, fw, api versions */
	dev_info(&pdev->dev, "fw %d.%d.%05d api %d.%d nvm %s\n",
		 hw->aq.fw_maj_ver, hw->aq.fw_min_ver, hw->aq.fw_build,
		 hw->aq.api_maj_ver, hw->aq.api_min_ver,
		 i40e_nvm_version_str(hw));
10950

10951 10952
	if (hw->aq.api_maj_ver == I40E_FW_API_VERSION_MAJOR &&
	    hw->aq.api_min_ver > I40E_FW_API_VERSION_MINOR)
10953
		dev_info(&pdev->dev,
10954 10955 10956
			 "The driver for the device detected a newer version of the NVM image than expected. Please install the most recent version of the network driver.\n");
	else if (hw->aq.api_maj_ver < I40E_FW_API_VERSION_MAJOR ||
		 hw->aq.api_min_ver < (I40E_FW_API_VERSION_MINOR - 1))
10957
		dev_info(&pdev->dev,
10958
			 "The driver for the device detected an older version of the NVM image than expected. Please update the NVM image.\n");
10959

10960 10961
	i40e_verify_eeprom(pf);

10962 10963 10964 10965
	/* Rev 0 hardware was never productized */
	if (hw->revision_id < 1)
		dev_warn(&pdev->dev, "This device is a pre-production adapter/LOM. Please be aware there may be issues with your hardware. If you are experiencing problems please contact your Intel or hardware representative who provided you with this hardware.\n");

10966
	i40e_clear_pxe_mode(hw);
J
Jesse Brandeburg 已提交
10967 10968 10969 10970 10971 10972 10973 10974 10975 10976 10977 10978 10979 10980 10981 10982 10983 10984 10985 10986 10987 10988 10989 10990 10991
	err = i40e_get_capabilities(pf);
	if (err)
		goto err_adminq_setup;

	err = i40e_sw_init(pf);
	if (err) {
		dev_info(&pdev->dev, "sw_init failed: %d\n", err);
		goto err_sw_init;
	}

	err = i40e_init_lan_hmc(hw, hw->func_caps.num_tx_qp,
				hw->func_caps.num_rx_qp,
				pf->fcoe_hmc_cntx_num, pf->fcoe_hmc_filt_num);
	if (err) {
		dev_info(&pdev->dev, "init_lan_hmc failed: %d\n", err);
		goto err_init_lan_hmc;
	}

	err = i40e_configure_lan_hmc(hw, I40E_HMC_MODEL_DIRECT_ONLY);
	if (err) {
		dev_info(&pdev->dev, "configure_lan_hmc failed: %d\n", err);
		err = -ENOENT;
		goto err_configure_lan_hmc;
	}

10992 10993 10994 10995
	/* Disable LLDP for NICs that have firmware versions lower than v4.3.
	 * Ignore error return codes because if it was already disabled via
	 * hardware settings this will fail
	 */
10996
	if (pf->flags & I40E_FLAG_STOP_FW_LLDP) {
10997 10998 10999 11000
		dev_info(&pdev->dev, "Stopping firmware LLDP agent.\n");
		i40e_aq_stop_lldp(hw, true, NULL);
	}

J
Jesse Brandeburg 已提交
11001
	i40e_get_mac_addr(hw, hw->mac.addr);
11002 11003
	/* allow a platform config to override the HW addr */
	i40e_get_platform_mac_addr(pdev, pf);
11004
	if (!is_valid_ether_addr(hw->mac.addr)) {
J
Jesse Brandeburg 已提交
11005 11006 11007 11008 11009
		dev_info(&pdev->dev, "invalid MAC address %pM\n", hw->mac.addr);
		err = -EIO;
		goto err_mac_addr;
	}
	dev_info(&pdev->dev, "MAC address: %pM\n", hw->mac.addr);
11010
	ether_addr_copy(hw->mac.perm_addr, hw->mac.addr);
11011 11012 11013
	i40e_get_port_mac_addr(hw, hw->mac.port_addr);
	if (is_valid_ether_addr(hw->mac.port_addr))
		pf->flags |= I40E_FLAG_PORT_ID_VALID;
11014 11015 11016 11017 11018 11019 11020 11021 11022 11023 11024 11025
#ifdef I40E_FCOE
	err = i40e_get_san_mac_addr(hw, hw->mac.san_addr);
	if (err)
		dev_info(&pdev->dev,
			 "(non-fatal) SAN MAC retrieval failed: %d\n", err);
	if (!is_valid_ether_addr(hw->mac.san_addr)) {
		dev_warn(&pdev->dev, "invalid SAN MAC address %pM, falling back to LAN MAC\n",
			 hw->mac.san_addr);
		ether_addr_copy(hw->mac.san_addr, hw->mac.addr);
	}
	dev_info(&pf->pdev->dev, "SAN MAC: %pM\n", hw->mac.san_addr);
#endif /* I40E_FCOE */
J
Jesse Brandeburg 已提交
11026 11027 11028

	pci_set_drvdata(pdev, pf);
	pci_save_state(pdev);
N
Neerav Parikh 已提交
11029 11030 11031
#ifdef CONFIG_I40E_DCB
	err = i40e_init_pf_dcb(pf);
	if (err) {
11032
		dev_info(&pdev->dev, "DCB init failed %d, disabled\n", err);
11033
		pf->flags &= ~(I40E_FLAG_DCB_CAPABLE | I40E_FLAG_DCB_ENABLED);
11034
		/* Continue without DCB enabled */
N
Neerav Parikh 已提交
11035 11036
	}
#endif /* CONFIG_I40E_DCB */
J
Jesse Brandeburg 已提交
11037 11038 11039 11040 11041 11042 11043 11044 11045

	/* set up periodic task facility */
	setup_timer(&pf->service_timer, i40e_service_timer, (unsigned long)pf);
	pf->service_timer_period = HZ;

	INIT_WORK(&pf->service_task, i40e_service_task);
	clear_bit(__I40E_SERVICE_SCHED, &pf->state);
	pf->flags |= I40E_FLAG_NEED_LINK_UPDATE;

11046 11047
	/* NVM bit on means WoL disabled for the port */
	i40e_read_nvm_word(hw, I40E_SR_NVM_WAKE_ON_LAN, &wol_nvm_bits);
J
Jesse Brandeburg 已提交
11048
	if (BIT (hw->port) & wol_nvm_bits || hw->partition_id != 1)
11049 11050 11051
		pf->wol_en = false;
	else
		pf->wol_en = true;
S
Shannon Nelson 已提交
11052 11053
	device_set_wakeup_enable(&pf->pdev->dev, pf->wol_en);

J
Jesse Brandeburg 已提交
11054 11055
	/* set up the main switch operations */
	i40e_determine_queue_usage(pf);
11056 11057 11058
	err = i40e_init_interrupt_scheme(pf);
	if (err)
		goto err_switch_setup;
J
Jesse Brandeburg 已提交
11059

M
Mitch Williams 已提交
11060 11061 11062 11063
	/* The number of VSIs reported by the FW is the minimum guaranteed
	 * to us; HW supports far more and we share the remaining pool with
	 * the other PFs. We allocate space for more than the guarantee with
	 * the understanding that we might not get them all later.
J
Jesse Brandeburg 已提交
11064
	 */
M
Mitch Williams 已提交
11065 11066 11067 11068 11069 11070
	if (pf->hw.func_caps.num_vsis < I40E_MIN_VSI_ALLOC)
		pf->num_alloc_vsi = I40E_MIN_VSI_ALLOC;
	else
		pf->num_alloc_vsi = pf->hw.func_caps.num_vsis;

	/* Set up the *vsi struct and our local tracking of the MAIN PF vsi. */
11071 11072
	pf->vsi = kcalloc(pf->num_alloc_vsi, sizeof(struct i40e_vsi *),
			  GFP_KERNEL);
11073 11074
	if (!pf->vsi) {
		err = -ENOMEM;
J
Jesse Brandeburg 已提交
11075
		goto err_switch_setup;
11076
	}
J
Jesse Brandeburg 已提交
11077

11078 11079 11080 11081 11082 11083 11084 11085 11086
#ifdef CONFIG_PCI_IOV
	/* prep for VF support */
	if ((pf->flags & I40E_FLAG_SRIOV_ENABLED) &&
	    (pf->flags & I40E_FLAG_MSIX_ENABLED) &&
	    !test_bit(__I40E_BAD_EEPROM, &pf->state)) {
		if (pci_num_vf(pdev))
			pf->flags |= I40E_FLAG_VEB_MODE_ENABLED;
	}
#endif
11087
	err = i40e_setup_pf_switch(pf, false);
J
Jesse Brandeburg 已提交
11088 11089 11090 11091
	if (err) {
		dev_info(&pdev->dev, "setup_pf_switch failed: %d\n", err);
		goto err_vsis;
	}
11092 11093 11094 11095 11096 11097 11098 11099 11100 11101 11102 11103 11104 11105 11106 11107 11108 11109 11110

	/* Make sure flow control is set according to current settings */
	err = i40e_set_fc(hw, &set_fc_aq_fail, true);
	if (set_fc_aq_fail & I40E_SET_FC_AQ_FAIL_GET)
		dev_dbg(&pf->pdev->dev,
			"Set fc with err %s aq_err %s on get_phy_cap\n",
			i40e_stat_str(hw, err),
			i40e_aq_str(hw, hw->aq.asq_last_status));
	if (set_fc_aq_fail & I40E_SET_FC_AQ_FAIL_SET)
		dev_dbg(&pf->pdev->dev,
			"Set fc with err %s aq_err %s on set_phy_config\n",
			i40e_stat_str(hw, err),
			i40e_aq_str(hw, hw->aq.asq_last_status));
	if (set_fc_aq_fail & I40E_SET_FC_AQ_FAIL_UPDATE)
		dev_dbg(&pf->pdev->dev,
			"Set fc with err %s aq_err %s on get_link_info\n",
			i40e_stat_str(hw, err),
			i40e_aq_str(hw, hw->aq.asq_last_status));

11111
	/* if FDIR VSI was set up, start it now */
M
Mitch Williams 已提交
11112
	for (i = 0; i < pf->num_alloc_vsi; i++) {
11113 11114 11115 11116 11117
		if (pf->vsi[i] && pf->vsi[i]->type == I40E_VSI_FDIR) {
			i40e_vsi_open(pf->vsi[i]);
			break;
		}
	}
J
Jesse Brandeburg 已提交
11118

S
Shannon Nelson 已提交
11119 11120
	/* The driver only wants link up/down and module qualification
	 * reports from firmware.  Note the negative logic.
J
Jesse Brandeburg 已提交
11121 11122
	 */
	err = i40e_aq_set_phy_int_mask(&pf->hw,
S
Shannon Nelson 已提交
11123
				       ~(I40E_AQ_EVENT_LINK_UPDOWN |
11124
					 I40E_AQ_EVENT_MEDIA_NA |
S
Shannon Nelson 已提交
11125
					 I40E_AQ_EVENT_MODULE_QUAL_FAIL), NULL);
J
Jesse Brandeburg 已提交
11126
	if (err)
11127 11128 11129
		dev_info(&pf->pdev->dev, "set phy mask fail, err %s aq_err %s\n",
			 i40e_stat_str(&pf->hw, err),
			 i40e_aq_str(&pf->hw, pf->hw.aq.asq_last_status));
J
Jesse Brandeburg 已提交
11130

11131 11132 11133 11134 11135 11136 11137 11138 11139 11140 11141
	/* Reconfigure hardware for allowing smaller MSS in the case
	 * of TSO, so that we avoid the MDD being fired and causing
	 * a reset in the case of small MSS+TSO.
	 */
	val = rd32(hw, I40E_REG_MSS);
	if ((val & I40E_REG_MSS_MIN_MASK) > I40E_64BYTE_MSS) {
		val &= ~I40E_REG_MSS_MIN_MASK;
		val |= I40E_64BYTE_MSS;
		wr32(hw, I40E_REG_MSS, val);
	}

11142
	if (pf->flags & I40E_FLAG_RESTART_AUTONEG) {
11143 11144 11145
		msleep(75);
		err = i40e_aq_set_link_restart_an(&pf->hw, true, NULL);
		if (err)
11146 11147 11148 11149
			dev_info(&pf->pdev->dev, "link restart failed, err %s aq_err %s\n",
				 i40e_stat_str(&pf->hw, err),
				 i40e_aq_str(&pf->hw,
					     pf->hw.aq.asq_last_status));
11150
	}
J
Jesse Brandeburg 已提交
11151 11152 11153 11154 11155 11156 11157 11158 11159 11160 11161 11162 11163 11164 11165 11166 11167 11168 11169 11170
	/* The main driver is (mostly) up and happy. We need to set this state
	 * before setting up the misc vector or we get a race and the vector
	 * ends up disabled forever.
	 */
	clear_bit(__I40E_DOWN, &pf->state);

	/* In case of MSIX we are going to setup the misc vector right here
	 * to handle admin queue events etc. In case of legacy and MSI
	 * the misc functionality and queue processing is combined in
	 * the same vector and that gets setup at open.
	 */
	if (pf->flags & I40E_FLAG_MSIX_ENABLED) {
		err = i40e_setup_misc_vector(pf);
		if (err) {
			dev_info(&pdev->dev,
				 "setup of misc vector failed: %d\n", err);
			goto err_vsis;
		}
	}

11171
#ifdef CONFIG_PCI_IOV
J
Jesse Brandeburg 已提交
11172 11173
	/* prep for VF support */
	if ((pf->flags & I40E_FLAG_SRIOV_ENABLED) &&
11174 11175
	    (pf->flags & I40E_FLAG_MSIX_ENABLED) &&
	    !test_bit(__I40E_BAD_EEPROM, &pf->state)) {
J
Jesse Brandeburg 已提交
11176 11177 11178 11179 11180
		/* disable link interrupts for VFs */
		val = rd32(hw, I40E_PFGEN_PORTMDIO_NUM);
		val &= ~I40E_PFGEN_PORTMDIO_NUM_VFLINK_STAT_ENA_MASK;
		wr32(hw, I40E_PFGEN_PORTMDIO_NUM, val);
		i40e_flush(hw);
M
Mitch Williams 已提交
11181 11182 11183 11184 11185 11186 11187 11188 11189 11190

		if (pci_num_vf(pdev)) {
			dev_info(&pdev->dev,
				 "Active VFs found, allocating resources.\n");
			err = i40e_alloc_vfs(pf, pci_num_vf(pdev));
			if (err)
				dev_info(&pdev->dev,
					 "Error %d allocating resources for existing VFs\n",
					 err);
		}
J
Jesse Brandeburg 已提交
11191
	}
11192
#endif /* CONFIG_PCI_IOV */
J
Jesse Brandeburg 已提交
11193

11194 11195 11196 11197 11198 11199 11200 11201 11202 11203 11204
	if (pf->flags & I40E_FLAG_IWARP_ENABLED) {
		pf->iwarp_base_vector = i40e_get_lump(pf, pf->irq_pile,
						      pf->num_iwarp_msix,
						      I40E_IWARP_IRQ_PILE_ID);
		if (pf->iwarp_base_vector < 0) {
			dev_info(&pdev->dev,
				 "failed to get tracking for %d vectors for IWARP err=%d\n",
				 pf->num_iwarp_msix, pf->iwarp_base_vector);
			pf->flags &= ~I40E_FLAG_IWARP_ENABLED;
		}
	}
11205

J
Jesse Brandeburg 已提交
11206 11207 11208
	i40e_dbg_pf_init(pf);

	/* tell the firmware that we're starting */
J
Jesse Brandeburg 已提交
11209
	i40e_send_version(pf);
J
Jesse Brandeburg 已提交
11210 11211 11212 11213 11214

	/* since everything's happy, start the service_task timer */
	mod_timer(&pf->service_timer,
		  round_jiffies(jiffies + pf->service_timer_period));

11215 11216 11217 11218 11219 11220
	/* add this PF to client device list and launch a client service task */
	err = i40e_lan_add_device(pf);
	if (err)
		dev_info(&pdev->dev, "Failed to add PF to client API service list: %d\n",
			 err);

11221 11222 11223 11224 11225
#ifdef I40E_FCOE
	/* create FCoE interface */
	i40e_fcoe_vsi_setup(pf);

#endif
11226 11227 11228 11229 11230 11231 11232 11233 11234 11235 11236 11237 11238 11239 11240 11241 11242 11243 11244 11245 11246 11247 11248 11249 11250 11251 11252 11253 11254 11255 11256 11257 11258 11259 11260 11261 11262 11263 11264 11265 11266 11267 11268 11269 11270 11271 11272 11273 11274
#define PCI_SPEED_SIZE 8
#define PCI_WIDTH_SIZE 8
	/* Devices on the IOSF bus do not have this information
	 * and will report PCI Gen 1 x 1 by default so don't bother
	 * checking them.
	 */
	if (!(pf->flags & I40E_FLAG_NO_PCI_LINK_CHECK)) {
		char speed[PCI_SPEED_SIZE] = "Unknown";
		char width[PCI_WIDTH_SIZE] = "Unknown";

		/* Get the negotiated link width and speed from PCI config
		 * space
		 */
		pcie_capability_read_word(pf->pdev, PCI_EXP_LNKSTA,
					  &link_status);

		i40e_set_pci_config_data(hw, link_status);

		switch (hw->bus.speed) {
		case i40e_bus_speed_8000:
			strncpy(speed, "8.0", PCI_SPEED_SIZE); break;
		case i40e_bus_speed_5000:
			strncpy(speed, "5.0", PCI_SPEED_SIZE); break;
		case i40e_bus_speed_2500:
			strncpy(speed, "2.5", PCI_SPEED_SIZE); break;
		default:
			break;
		}
		switch (hw->bus.width) {
		case i40e_bus_width_pcie_x8:
			strncpy(width, "8", PCI_WIDTH_SIZE); break;
		case i40e_bus_width_pcie_x4:
			strncpy(width, "4", PCI_WIDTH_SIZE); break;
		case i40e_bus_width_pcie_x2:
			strncpy(width, "2", PCI_WIDTH_SIZE); break;
		case i40e_bus_width_pcie_x1:
			strncpy(width, "1", PCI_WIDTH_SIZE); break;
		default:
			break;
		}

		dev_info(&pdev->dev, "PCI-Express: Speed %sGT/s Width x%s\n",
			 speed, width);

		if (hw->bus.width < i40e_bus_width_pcie_x8 ||
		    hw->bus.speed < i40e_bus_speed_8000) {
			dev_warn(&pdev->dev, "PCI-Express bandwidth available for this device may be insufficient for optimal performance.\n");
			dev_warn(&pdev->dev, "Please move the device to a different PCI-e link with more lanes and/or higher transfer rate.\n");
		}
11275 11276
	}

11277 11278 11279
	/* get the requested speeds from the fw */
	err = i40e_aq_get_phy_capabilities(hw, false, false, &abilities, NULL);
	if (err)
11280 11281 11282
		dev_dbg(&pf->pdev->dev, "get requested speeds ret =  %s last_status =  %s\n",
			i40e_stat_str(&pf->hw, err),
			i40e_aq_str(&pf->hw, pf->hw.aq.asq_last_status));
11283 11284
	pf->hw.phy.link_info.requested_speeds = abilities.link_speed;

11285 11286 11287 11288 11289 11290 11291 11292
	/* get the supported phy types from the fw */
	err = i40e_aq_get_phy_capabilities(hw, false, true, &abilities, NULL);
	if (err)
		dev_dbg(&pf->pdev->dev, "get supported phy types ret =  %s last_status =  %s\n",
			i40e_stat_str(&pf->hw, err),
			i40e_aq_str(&pf->hw, pf->hw.aq.asq_last_status));
	pf->hw.phy.phy_types = le32_to_cpu(abilities.phy_type);

11293 11294 11295 11296 11297 11298 11299 11300 11301
	/* Add a filter to drop all Flow control frames from any VSI from being
	 * transmitted. By doing so we stop a malicious VF from sending out
	 * PAUSE or PFC frames and potentially controlling traffic for other
	 * PF/VF VSIs.
	 * The FW can still send Flow control frames if enabled.
	 */
	i40e_add_filter_to_drop_tx_flow_control_frames(&pf->hw,
						       pf->main_vsi_seid);

11302 11303 11304 11305
	if ((pf->hw.device_id == I40E_DEV_ID_10G_BASE_T) ||
	    (pf->hw.device_id == I40E_DEV_ID_10G_BASE_T4))
		pf->flags |= I40E_FLAG_HAVE_10GBASET_PHY;

J
Jesse Brandeburg 已提交
11306 11307 11308
	/* print a string summarizing features */
	i40e_print_features(pf);

J
Jesse Brandeburg 已提交
11309 11310 11311 11312 11313 11314 11315
	return 0;

	/* Unwind what we've done if something failed in the setup */
err_vsis:
	set_bit(__I40E_DOWN, &pf->state);
	i40e_clear_interrupt_scheme(pf);
	kfree(pf->vsi);
11316 11317
err_switch_setup:
	i40e_reset_interrupt_capability(pf);
J
Jesse Brandeburg 已提交
11318 11319 11320 11321 11322 11323 11324 11325 11326 11327 11328 11329 11330 11331
	del_timer_sync(&pf->service_timer);
err_mac_addr:
err_configure_lan_hmc:
	(void)i40e_shutdown_lan_hmc(hw);
err_init_lan_hmc:
	kfree(pf->qp_pile);
err_sw_init:
err_adminq_setup:
err_pf_reset:
	iounmap(hw->hw_addr);
err_ioremap:
	kfree(pf);
err_pf_alloc:
	pci_disable_pcie_error_reporting(pdev);
11332
	pci_release_mem_regions(pdev);
J
Jesse Brandeburg 已提交
11333 11334 11335 11336 11337 11338 11339 11340 11341 11342 11343 11344 11345 11346 11347 11348 11349 11350
err_pci_reg:
err_dma:
	pci_disable_device(pdev);
	return err;
}

/**
 * i40e_remove - Device removal routine
 * @pdev: PCI device information struct
 *
 * i40e_remove is called by the PCI subsystem to alert the driver
 * that is should release a PCI device.  This could be caused by a
 * Hot-Plug event, or because the driver is going to be removed from
 * memory.
 **/
static void i40e_remove(struct pci_dev *pdev)
{
	struct i40e_pf *pf = pci_get_drvdata(pdev);
11351
	struct i40e_hw *hw = &pf->hw;
J
Jesse Brandeburg 已提交
11352 11353 11354 11355 11356
	i40e_status ret_code;
	int i;

	i40e_dbg_pf_exit(pf);

J
Jacob Keller 已提交
11357 11358
	i40e_ptp_stop(pf);

11359
	/* Disable RSS in hw */
11360 11361
	i40e_write_rx_ctl(hw, I40E_PFQF_HENA(0), 0);
	i40e_write_rx_ctl(hw, I40E_PFQF_HENA(1), 0);
11362

J
Jesse Brandeburg 已提交
11363
	/* no more scheduling of any task */
11364
	set_bit(__I40E_SUSPENDED, &pf->state);
J
Jesse Brandeburg 已提交
11365
	set_bit(__I40E_DOWN, &pf->state);
11366 11367 11368 11369
	if (pf->service_timer.data)
		del_timer_sync(&pf->service_timer);
	if (pf->service_task.func)
		cancel_work_sync(&pf->service_task);
J
Jesse Brandeburg 已提交
11370

11371 11372 11373 11374 11375
	if (pf->flags & I40E_FLAG_SRIOV_ENABLED) {
		i40e_free_vfs(pf);
		pf->flags &= ~I40E_FLAG_SRIOV_ENABLED;
	}

J
Jesse Brandeburg 已提交
11376 11377 11378 11379 11380 11381 11382 11383 11384 11385 11386 11387 11388 11389 11390 11391 11392 11393 11394 11395
	i40e_fdir_teardown(pf);

	/* If there is a switch structure or any orphans, remove them.
	 * This will leave only the PF's VSI remaining.
	 */
	for (i = 0; i < I40E_MAX_VEB; i++) {
		if (!pf->veb[i])
			continue;

		if (pf->veb[i]->uplink_seid == pf->mac_seid ||
		    pf->veb[i]->uplink_seid == 0)
			i40e_switch_branch_release(pf->veb[i]);
	}

	/* Now we can shutdown the PF's VSI, just before we kill
	 * adminq and hmc.
	 */
	if (pf->vsi[pf->lan_vsi])
		i40e_vsi_release(pf->vsi[pf->lan_vsi]);

11396 11397 11398 11399 11400 11401 11402
	/* remove attached clients */
	ret_code = i40e_lan_del_device(pf);
	if (ret_code) {
		dev_warn(&pdev->dev, "Failed to delete client device: %d\n",
			 ret_code);
	}

J
Jesse Brandeburg 已提交
11403
	/* shutdown and destroy the HMC */
11404 11405
	if (hw->hmc.hmc_obj) {
		ret_code = i40e_shutdown_lan_hmc(hw);
11406 11407 11408 11409 11410
		if (ret_code)
			dev_warn(&pdev->dev,
				 "Failed to destroy the HMC resources: %d\n",
				 ret_code);
	}
J
Jesse Brandeburg 已提交
11411 11412

	/* shutdown the adminq */
H
Henry Tieman 已提交
11413
	i40e_shutdown_adminq(hw);
J
Jesse Brandeburg 已提交
11414

11415 11416 11417 11418
	/* destroy the locks only once, here */
	mutex_destroy(&hw->aq.arq_mutex);
	mutex_destroy(&hw->aq.asq_mutex);

J
Jesse Brandeburg 已提交
11419 11420
	/* Clear all dynamic memory lists of rings, q_vectors, and VSIs */
	i40e_clear_interrupt_scheme(pf);
M
Mitch Williams 已提交
11421
	for (i = 0; i < pf->num_alloc_vsi; i++) {
J
Jesse Brandeburg 已提交
11422 11423 11424 11425 11426 11427 11428 11429 11430 11431 11432 11433 11434 11435 11436
		if (pf->vsi[i]) {
			i40e_vsi_clear_rings(pf->vsi[i]);
			i40e_vsi_clear(pf->vsi[i]);
			pf->vsi[i] = NULL;
		}
	}

	for (i = 0; i < I40E_MAX_VEB; i++) {
		kfree(pf->veb[i]);
		pf->veb[i] = NULL;
	}

	kfree(pf->qp_pile);
	kfree(pf->vsi);

11437
	iounmap(hw->hw_addr);
J
Jesse Brandeburg 已提交
11438
	kfree(pf);
11439
	pci_release_mem_regions(pdev);
J
Jesse Brandeburg 已提交
11440 11441 11442 11443 11444 11445 11446 11447 11448 11449 11450 11451 11452 11453 11454 11455 11456 11457 11458 11459

	pci_disable_pcie_error_reporting(pdev);
	pci_disable_device(pdev);
}

/**
 * i40e_pci_error_detected - warning that something funky happened in PCI land
 * @pdev: PCI device information struct
 *
 * Called to warn that something happened and the error handling steps
 * are in progress.  Allows the driver to quiesce things, be ready for
 * remediation.
 **/
static pci_ers_result_t i40e_pci_error_detected(struct pci_dev *pdev,
						enum pci_channel_state error)
{
	struct i40e_pf *pf = pci_get_drvdata(pdev);

	dev_info(&pdev->dev, "%s: error %d\n", __func__, error);

11460 11461 11462 11463 11464 11465
	if (!pf) {
		dev_info(&pdev->dev,
			 "Cannot recover - error happened during device probe\n");
		return PCI_ERS_RESULT_DISCONNECT;
	}

J
Jesse Brandeburg 已提交
11466
	/* shutdown all operations */
11467 11468 11469 11470 11471
	if (!test_bit(__I40E_SUSPENDED, &pf->state)) {
		rtnl_lock();
		i40e_prep_for_reset(pf);
		rtnl_unlock();
	}
J
Jesse Brandeburg 已提交
11472 11473 11474 11475 11476 11477 11478 11479 11480 11481 11482 11483 11484 11485 11486 11487 11488 11489 11490 11491 11492

	/* Request a slot reset */
	return PCI_ERS_RESULT_NEED_RESET;
}

/**
 * i40e_pci_error_slot_reset - a PCI slot reset just happened
 * @pdev: PCI device information struct
 *
 * Called to find if the driver can work with the device now that
 * the pci slot has been reset.  If a basic connection seems good
 * (registers are readable and have sane content) then return a
 * happy little PCI_ERS_RESULT_xxx.
 **/
static pci_ers_result_t i40e_pci_error_slot_reset(struct pci_dev *pdev)
{
	struct i40e_pf *pf = pci_get_drvdata(pdev);
	pci_ers_result_t result;
	int err;
	u32 reg;

S
Shannon Nelson 已提交
11493
	dev_dbg(&pdev->dev, "%s\n", __func__);
J
Jesse Brandeburg 已提交
11494 11495 11496 11497 11498 11499 11500 11501 11502 11503 11504 11505 11506 11507 11508 11509 11510 11511 11512 11513 11514 11515 11516 11517 11518 11519 11520 11521 11522 11523 11524 11525 11526 11527 11528 11529 11530 11531 11532
	if (pci_enable_device_mem(pdev)) {
		dev_info(&pdev->dev,
			 "Cannot re-enable PCI device after reset.\n");
		result = PCI_ERS_RESULT_DISCONNECT;
	} else {
		pci_set_master(pdev);
		pci_restore_state(pdev);
		pci_save_state(pdev);
		pci_wake_from_d3(pdev, false);

		reg = rd32(&pf->hw, I40E_GLGEN_RTRIG);
		if (reg == 0)
			result = PCI_ERS_RESULT_RECOVERED;
		else
			result = PCI_ERS_RESULT_DISCONNECT;
	}

	err = pci_cleanup_aer_uncorrect_error_status(pdev);
	if (err) {
		dev_info(&pdev->dev,
			 "pci_cleanup_aer_uncorrect_error_status failed 0x%0x\n",
			 err);
		/* non-fatal, continue */
	}

	return result;
}

/**
 * i40e_pci_error_resume - restart operations after PCI error recovery
 * @pdev: PCI device information struct
 *
 * Called to allow the driver to bring things back up after PCI error
 * and/or reset recovery has finished.
 **/
static void i40e_pci_error_resume(struct pci_dev *pdev)
{
	struct i40e_pf *pf = pci_get_drvdata(pdev);

S
Shannon Nelson 已提交
11533
	dev_dbg(&pdev->dev, "%s\n", __func__);
11534 11535 11536 11537
	if (test_bit(__I40E_SUSPENDED, &pf->state))
		return;

	rtnl_lock();
J
Jesse Brandeburg 已提交
11538
	i40e_handle_reset_warning(pf);
11539
	rtnl_unlock();
11540 11541 11542 11543 11544 11545 11546 11547 11548
}

/**
 * i40e_shutdown - PCI callback for shutting down
 * @pdev: PCI device information struct
 **/
static void i40e_shutdown(struct pci_dev *pdev)
{
	struct i40e_pf *pf = pci_get_drvdata(pdev);
S
Shannon Nelson 已提交
11549
	struct i40e_hw *hw = &pf->hw;
11550 11551 11552 11553 11554 11555 11556

	set_bit(__I40E_SUSPENDED, &pf->state);
	set_bit(__I40E_DOWN, &pf->state);
	rtnl_lock();
	i40e_prep_for_reset(pf);
	rtnl_unlock();

S
Shannon Nelson 已提交
11557 11558 11559
	wr32(hw, I40E_PFPM_APM, (pf->wol_en ? I40E_PFPM_APM_APME_MASK : 0));
	wr32(hw, I40E_PFPM_WUFC, (pf->wol_en ? I40E_PFPM_WUFC_MAG_MASK : 0));

11560 11561 11562 11563 11564 11565 11566 11567 11568 11569 11570 11571 11572
	del_timer_sync(&pf->service_timer);
	cancel_work_sync(&pf->service_task);
	i40e_fdir_teardown(pf);

	rtnl_lock();
	i40e_prep_for_reset(pf);
	rtnl_unlock();

	wr32(hw, I40E_PFPM_APM,
	     (pf->wol_en ? I40E_PFPM_APM_APME_MASK : 0));
	wr32(hw, I40E_PFPM_WUFC,
	     (pf->wol_en ? I40E_PFPM_WUFC_MAG_MASK : 0));

11573 11574
	i40e_clear_interrupt_scheme(pf);

11575
	if (system_state == SYSTEM_POWER_OFF) {
S
Shannon Nelson 已提交
11576
		pci_wake_from_d3(pdev, pf->wol_en);
11577 11578 11579 11580 11581 11582 11583 11584 11585 11586 11587 11588
		pci_set_power_state(pdev, PCI_D3hot);
	}
}

#ifdef CONFIG_PM
/**
 * i40e_suspend - PCI callback for moving to D3
 * @pdev: PCI device information struct
 **/
static int i40e_suspend(struct pci_dev *pdev, pm_message_t state)
{
	struct i40e_pf *pf = pci_get_drvdata(pdev);
S
Shannon Nelson 已提交
11589
	struct i40e_hw *hw = &pf->hw;
G
Greg Rose 已提交
11590
	int retval = 0;
11591 11592 11593

	set_bit(__I40E_SUSPENDED, &pf->state);
	set_bit(__I40E_DOWN, &pf->state);
11594

11595 11596 11597 11598
	rtnl_lock();
	i40e_prep_for_reset(pf);
	rtnl_unlock();

S
Shannon Nelson 已提交
11599 11600 11601
	wr32(hw, I40E_PFPM_APM, (pf->wol_en ? I40E_PFPM_APM_APME_MASK : 0));
	wr32(hw, I40E_PFPM_WUFC, (pf->wol_en ? I40E_PFPM_WUFC_MAG_MASK : 0));

11602 11603
	i40e_stop_misc_vector(pf);

G
Greg Rose 已提交
11604 11605 11606 11607
	retval = pci_save_state(pdev);
	if (retval)
		return retval;

S
Shannon Nelson 已提交
11608
	pci_wake_from_d3(pdev, pf->wol_en);
11609 11610
	pci_set_power_state(pdev, PCI_D3hot);

G
Greg Rose 已提交
11611
	return retval;
J
Jesse Brandeburg 已提交
11612 11613
}

11614 11615 11616 11617 11618 11619 11620 11621 11622 11623 11624 11625 11626 11627 11628 11629 11630 11631
/**
 * i40e_resume - PCI callback for waking up from D3
 * @pdev: PCI device information struct
 **/
static int i40e_resume(struct pci_dev *pdev)
{
	struct i40e_pf *pf = pci_get_drvdata(pdev);
	u32 err;

	pci_set_power_state(pdev, PCI_D0);
	pci_restore_state(pdev);
	/* pci_restore_state() clears dev->state_saves, so
	 * call pci_save_state() again to restore it.
	 */
	pci_save_state(pdev);

	err = pci_enable_device_mem(pdev);
	if (err) {
S
Shannon Nelson 已提交
11632
		dev_err(&pdev->dev, "Cannot enable PCI device from suspend\n");
11633 11634 11635 11636 11637 11638 11639 11640 11641 11642 11643 11644 11645 11646 11647 11648 11649 11650 11651
		return err;
	}
	pci_set_master(pdev);

	/* no wakeup events while running */
	pci_wake_from_d3(pdev, false);

	/* handling the reset will rebuild the device state */
	if (test_and_clear_bit(__I40E_SUSPENDED, &pf->state)) {
		clear_bit(__I40E_DOWN, &pf->state);
		rtnl_lock();
		i40e_reset_and_rebuild(pf, false);
		rtnl_unlock();
	}

	return 0;
}

#endif
J
Jesse Brandeburg 已提交
11652 11653 11654 11655 11656 11657 11658 11659 11660 11661 11662
static const struct pci_error_handlers i40e_err_handler = {
	.error_detected = i40e_pci_error_detected,
	.slot_reset = i40e_pci_error_slot_reset,
	.resume = i40e_pci_error_resume,
};

static struct pci_driver i40e_driver = {
	.name     = i40e_driver_name,
	.id_table = i40e_pci_tbl,
	.probe    = i40e_probe,
	.remove   = i40e_remove,
11663 11664 11665 11666 11667
#ifdef CONFIG_PM
	.suspend  = i40e_suspend,
	.resume   = i40e_resume,
#endif
	.shutdown = i40e_shutdown,
J
Jesse Brandeburg 已提交
11668 11669 11670 11671 11672 11673 11674 11675 11676 11677 11678 11679 11680 11681 11682
	.err_handler = &i40e_err_handler,
	.sriov_configure = i40e_pci_sriov_configure,
};

/**
 * i40e_init_module - Driver registration routine
 *
 * i40e_init_module is the first routine called when the driver is
 * loaded. All it does is register with the PCI subsystem.
 **/
static int __init i40e_init_module(void)
{
	pr_info("%s: %s - version %s\n", i40e_driver_name,
		i40e_driver_string, i40e_driver_version_str);
	pr_info("%s: %s\n", i40e_driver_name, i40e_copyright);
11683

11684 11685 11686 11687
	/* we will see if single thread per module is enough for now,
	 * it can't be any worse than using the system workqueue which
	 * was already single threaded
	 */
11688 11689
	i40e_wq = alloc_workqueue("%s", WQ_UNBOUND | WQ_MEM_RECLAIM, 1,
				  i40e_driver_name);
11690 11691 11692 11693 11694
	if (!i40e_wq) {
		pr_err("%s: Failed to create workqueue\n", i40e_driver_name);
		return -ENOMEM;
	}

J
Jesse Brandeburg 已提交
11695 11696 11697 11698 11699 11700 11701 11702 11703 11704 11705 11706 11707 11708
	i40e_dbg_init();
	return pci_register_driver(&i40e_driver);
}
module_init(i40e_init_module);

/**
 * i40e_exit_module - Driver exit cleanup routine
 *
 * i40e_exit_module is called just before the driver is removed
 * from memory.
 **/
static void __exit i40e_exit_module(void)
{
	pci_unregister_driver(&i40e_driver);
11709
	destroy_workqueue(i40e_wq);
J
Jesse Brandeburg 已提交
11710 11711 11712
	i40e_dbg_exit();
}
module_exit(i40e_exit_module);