i40e_main.c 326.0 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 43 44
#define DRV_VERSION_MAJOR 2
#define DRV_VERSION_MINOR 1
#define DRV_VERSION_BUILD 7
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 81
	{PCI_VDEVICE(INTEL, I40E_DEV_ID_KX_X722), 0},
	{PCI_VDEVICE(INTEL, I40E_DEV_ID_QSFP_X722), 0},
82 83 84
	{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},
85
	{PCI_VDEVICE(INTEL, I40E_DEV_ID_SFP_I_X722), 0},
86 87
	{PCI_VDEVICE(INTEL, I40E_DEV_ID_20G_KR2), 0},
	{PCI_VDEVICE(INTEL, I40E_DEV_ID_20G_KR2_A), 0},
88 89
	{PCI_VDEVICE(INTEL, I40E_DEV_ID_25G_B), 0},
	{PCI_VDEVICE(INTEL, I40E_DEV_ID_25G_SFP28), 0},
J
Jesse Brandeburg 已提交
90 91 92 93 94 95 96
	/* required last entry */
	{0, }
};
MODULE_DEVICE_TABLE(pci, i40e_pci_tbl);

#define I40E_MAX_VF_COUNT 128
static int debug = -1;
97 98
module_param(debug, uint, 0);
MODULE_PARM_DESC(debug, "Debug level (0=none,...,16=all), Debug mask (0x8XXXXXXX)");
J
Jesse Brandeburg 已提交
99 100 101 102 103 104

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);

105 106
static struct workqueue_struct *i40e_wq;

J
Jesse Brandeburg 已提交
107 108 109 110 111 112 113 114 115 116 117 118 119 120 121
/**
 * 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
122 123
	if (!mem->va)
		return -ENOMEM;
J
Jesse Brandeburg 已提交
124

J
Jesse Brandeburg 已提交
125
	return 0;
J
Jesse Brandeburg 已提交
126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156
}

/**
 * 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
157 158
	if (!mem->va)
		return -ENOMEM;
J
Jesse Brandeburg 已提交
159

J
Jesse Brandeburg 已提交
160
	return 0;
J
Jesse Brandeburg 已提交
161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194
}

/**
 * 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;
195
	int i, j;
J
Jesse Brandeburg 已提交
196 197 198 199 200 201 202 203 204 205

	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;
206
	while (i < pile->num_entries) {
J
Jesse Brandeburg 已提交
207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224
		/* 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;
225
			break;
J
Jesse Brandeburg 已提交
226
		}
J
Jesse Brandeburg 已提交
227 228 229

		/* not enough, so skip over it and continue looking */
		i += j;
J
Jesse Brandeburg 已提交
230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264
	}

	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;
}

265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280
/**
 * 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
281 282 283 284 285 286
/**
 * 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
 **/
287
void i40e_service_event_schedule(struct i40e_pf *pf)
J
Jesse Brandeburg 已提交
288 289
{
	if (!test_bit(__I40E_DOWN, &pf->state) &&
M
Mitch Williams 已提交
290
	    !test_bit(__I40E_RESET_RECOVERY_PENDING, &pf->state))
291
		queue_work(i40e_wq, &pf->service_task);
J
Jesse Brandeburg 已提交
292 293 294 295 296 297 298 299 300 301 302 303 304 305 306
}

/**
 * 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.
 **/
static void i40e_tx_timeout(struct net_device *netdev)
{
	struct i40e_netdev_priv *np = netdev_priv(netdev);
	struct i40e_vsi *vsi = np->vsi;
	struct i40e_pf *pf = vsi->back;
307 308 309
	struct i40e_ring *tx_ring = NULL;
	unsigned int i, hung_queue = 0;
	u32 head, val;
J
Jesse Brandeburg 已提交
310 311 312

	pf->tx_timeout_count++;

313 314 315 316 317 318
	/* 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
319
		trans_start = q->trans_start;
320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342
		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
343
	if (time_after(jiffies, (pf->tx_timeout_last_recovery + HZ*20)))
344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364
		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
365
	pf->tx_timeout_last_recovery = jiffies;
366 367
	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
368 369 370 371 372 373 374 375 376 377 378 379 380 381 382

	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;
	}
383

J
Jesse Brandeburg 已提交
384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406
	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.
 **/
407
static void i40e_get_netdev_stats_struct(struct net_device *netdev,
408
				  struct rtnl_link_stats64 *stats)
J
Jesse Brandeburg 已提交
409 410
{
	struct i40e_netdev_priv *np = netdev_priv(netdev);
411
	struct i40e_ring *tx_ring, *rx_ring;
J
Jesse Brandeburg 已提交
412
	struct i40e_vsi *vsi = np->vsi;
413 414 415
	struct rtnl_link_stats64 *vsi_stats = i40e_get_vsi_stats_struct(vsi);
	int i;

416
	if (test_bit(__I40E_DOWN, &vsi->state))
417
		return;
418

J
Jesse Brandeburg 已提交
419
	if (!vsi->tx_rings)
420
		return;
J
Jesse Brandeburg 已提交
421

422 423 424 425 426 427 428 429 430 431
	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 {
432
			start = u64_stats_fetch_begin_irq(&tx_ring->syncp);
433 434
			packets = tx_ring->stats.packets;
			bytes   = tx_ring->stats.bytes;
435
		} while (u64_stats_fetch_retry_irq(&tx_ring->syncp, start));
436 437 438 439 440 441

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

		do {
442
			start = u64_stats_fetch_begin_irq(&rx_ring->syncp);
443 444
			packets = rx_ring->stats.packets;
			bytes   = rx_ring->stats.bytes;
445
		} while (u64_stats_fetch_retry_irq(&rx_ring->syncp, start));
J
Jesse Brandeburg 已提交
446

447 448 449 450 451
		stats->rx_packets += packets;
		stats->rx_bytes   += bytes;
	}
	rcu_read_unlock();

452
	/* following stats updated by i40e_watchdog_subtask() */
453 454 455 456
	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;
457
	stats->rx_dropped	= vsi_stats->rx_dropped;
458 459
	stats->rx_crc_errors	= vsi_stats->rx_crc_errors;
	stats->rx_length_errors	= vsi_stats->rx_length_errors;
J
Jesse Brandeburg 已提交
460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478
}

/**
 * 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));
479
	if (vsi->rx_rings && vsi->rx_rings[0]) {
J
Jesse Brandeburg 已提交
480
		for (i = 0; i < vsi->num_queue_pairs; i++) {
J
Jesse Brandeburg 已提交
481
			memset(&vsi->rx_rings[i]->stats, 0,
482
			       sizeof(vsi->rx_rings[i]->stats));
J
Jesse Brandeburg 已提交
483
			memset(&vsi->rx_rings[i]->rx_stats, 0,
484
			       sizeof(vsi->rx_rings[i]->rx_stats));
J
Jesse Brandeburg 已提交
485
			memset(&vsi->tx_rings[i]->stats, 0,
486 487 488
			       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
489
		}
490
	}
J
Jesse Brandeburg 已提交
491 492 493 494
	vsi->stat_offsets_loaded = false;
}

/**
495
 * i40e_pf_reset_stats - Reset all of the stats for the given PF
J
Jesse Brandeburg 已提交
496 497 498 499
 * @pf: the PF to be reset
 **/
void i40e_pf_reset_stats(struct i40e_pf *pf)
{
500 501
	int i;

J
Jesse Brandeburg 已提交
502 503 504
	memset(&pf->stats, 0, sizeof(pf->stats));
	memset(&pf->stats_offsets, 0, sizeof(pf->stats_offsets));
	pf->stat_offsets_loaded = false;
505 506 507 508 509 510 511 512 513 514

	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;
		}
	}
515
	pf->hw_csum_rx_error = 0;
J
Jesse Brandeburg 已提交
516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537
}

/**
 * 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;

538
	if (hw->device_id == I40E_DEV_ID_QEMU) {
J
Jesse Brandeburg 已提交
539 540 541 542 543 544 545 546 547 548
		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
549
		*stat = (new_data + BIT_ULL(48)) - *offset;
J
Jesse Brandeburg 已提交
550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571
	*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
572
		*stat = (u32)((new_data + BIT_ULL(32)) - *offset);
J
Jesse Brandeburg 已提交
573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596
}

/**
 * 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);
597 598 599 600 601 602
	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
603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649

	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 */
650 651 652
	struct i40e_veb_tc_stats *veb_oes;
	struct i40e_veb_tc_stats *veb_es;
	int i, idx = 0;
J
Jesse Brandeburg 已提交
653 654 655 656

	idx = veb->stats_idx;
	es = &veb->stats;
	oes = &veb->stats_offsets;
657 658
	veb_es = &veb->tc_stats;
	veb_oes = &veb->tc_stats_offsets;
J
Jesse Brandeburg 已提交
659 660 661 662 663

	/* 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);
664 665 666 667 668
	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
669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693
	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);
694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715
	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
716 717 718 719
	veb->stat_offsets_loaded = true;
}

/**
S
Shannon Nelson 已提交
720
 * i40e_update_vsi_stats - Update the vsi statistics counters.
J
Jesse Brandeburg 已提交
721 722 723 724 725 726
 * @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
727
 * VF communications.  We sort it out here.
J
Jesse Brandeburg 已提交
728
 **/
S
Shannon Nelson 已提交
729
static void i40e_update_vsi_stats(struct i40e_vsi *vsi)
J
Jesse Brandeburg 已提交
730 731 732 733 734 735 736
{
	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;
737
	u64 tx_lost_interrupt;
738
	struct i40e_ring *p;
J
Jesse Brandeburg 已提交
739
	u32 rx_page, rx_buf;
740 741
	u64 bytes, packets;
	unsigned int start;
742
	u64 tx_linearize;
743
	u64 tx_force_wb;
J
Jesse Brandeburg 已提交
744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761
	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;
762
	tx_restart = tx_busy = tx_linearize = tx_force_wb = 0;
763
	tx_lost_interrupt = 0;
J
Jesse Brandeburg 已提交
764 765
	rx_page = 0;
	rx_buf = 0;
766
	rcu_read_lock();
J
Jesse Brandeburg 已提交
767
	for (q = 0; q < vsi->num_queue_pairs; q++) {
768 769 770 771
		/* locate Tx ring */
		p = ACCESS_ONCE(vsi->tx_rings[q]);

		do {
772
			start = u64_stats_fetch_begin_irq(&p->syncp);
773 774
			packets = p->stats.packets;
			bytes = p->stats.bytes;
775
		} while (u64_stats_fetch_retry_irq(&p->syncp, start));
776 777 778 779
		tx_b += bytes;
		tx_p += packets;
		tx_restart += p->tx_stats.restart_queue;
		tx_busy += p->tx_stats.tx_busy;
780
		tx_linearize += p->tx_stats.tx_linearize;
781
		tx_force_wb += p->tx_stats.tx_force_wb;
782
		tx_lost_interrupt += p->tx_stats.tx_lost_interrupt;
J
Jesse Brandeburg 已提交
783

784 785 786
		/* Rx queue is part of the same block as Tx queue */
		p = &p[1];
		do {
787
			start = u64_stats_fetch_begin_irq(&p->syncp);
788 789
			packets = p->stats.packets;
			bytes = p->stats.bytes;
790
		} while (u64_stats_fetch_retry_irq(&p->syncp, start));
791 792
		rx_b += bytes;
		rx_p += packets;
M
Mitch Williams 已提交
793 794
		rx_buf += p->rx_stats.alloc_buff_failed;
		rx_page += p->rx_stats.alloc_page_failed;
J
Jesse Brandeburg 已提交
795
	}
796
	rcu_read_unlock();
J
Jesse Brandeburg 已提交
797 798
	vsi->tx_restart = tx_restart;
	vsi->tx_busy = tx_busy;
799
	vsi->tx_linearize = tx_linearize;
800
	vsi->tx_force_wb = tx_force_wb;
801
	vsi->tx_lost_interrupt = tx_lost_interrupt;
J
Jesse Brandeburg 已提交
802 803 804 805 806 807 808 809 810
	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
811
	i40e_update_eth_stats(vsi);
J
Jesse Brandeburg 已提交
812 813 814 815
	ons->tx_errors = oes->tx_errors;
	ns->tx_errors = es->tx_errors;
	ons->multicast = oes->rx_multicast;
	ns->multicast = es->rx_multicast;
816 817
	ons->rx_dropped = oes->rx_discards;
	ns->rx_dropped = es->rx_discards;
J
Jesse Brandeburg 已提交
818 819 820
	ons->tx_dropped = oes->tx_discards;
	ns->tx_dropped = es->tx_discards;

S
Shannon Nelson 已提交
821
	/* pull in a couple PF stats if this is the main vsi */
J
Jesse Brandeburg 已提交
822
	if (vsi == pf->vsi[pf->lan_vsi]) {
S
Shannon Nelson 已提交
823 824 825 826 827
		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
828

S
Shannon Nelson 已提交
829
/**
830
 * i40e_update_pf_stats - Update the PF statistics counters.
S
Shannon Nelson 已提交
831 832 833 834 835 836 837 838 839
 * @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
840

S
Shannon Nelson 已提交
841 842 843 844 845 846 847 848 849 850 851 852
	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
853 854 855 856 857
	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
858 859 860 861 862
	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
863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882
	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
883

S
Shannon Nelson 已提交
884 885 886 887
	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
888

S
Shannon Nelson 已提交
889 890 891
	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
892

S
Shannon Nelson 已提交
893 894 895
	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
896

S
Shannon Nelson 已提交
897 898 899 900 901 902 903 904
	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
905

S
Shannon Nelson 已提交
906 907 908 909
	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
910

S
Shannon Nelson 已提交
911 912 913 914 915 916
	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);
917 918 919
	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
920 921 922
	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
923

S
Shannon Nelson 已提交
924
	for (i = 0; i < 8; i++) {
925 926 927 928
		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
929
		i40e_stat_update32(hw, I40E_GLPRT_PXONRXC(hw->port, i),
J
Jesse Brandeburg 已提交
930
				   pf->stat_offsets_loaded,
S
Shannon Nelson 已提交
931 932 933
				   &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
934
				   pf->stat_offsets_loaded,
S
Shannon Nelson 已提交
935 936 937
				   &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
938
				   pf->stat_offsets_loaded,
S
Shannon Nelson 已提交
939 940 941 942
				   &osd->priority_xoff_tx[i],
				   &nsd->priority_xoff_tx[i]);
		i40e_stat_update32(hw,
				   I40E_GLPRT_RXON2OFFCNT(hw->port, i),
943
				   pf->stat_offsets_loaded,
S
Shannon Nelson 已提交
944 945
				   &osd->priority_xon_2_xoff[i],
				   &nsd->priority_xon_2_xoff[i]);
J
Jesse Brandeburg 已提交
946 947
	}

S
Shannon Nelson 已提交
948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018
	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);

1019
	/* FDIR stats */
1020 1021
	i40e_stat_update32(hw,
			   I40E_GLQF_PCNT(I40E_FD_ATR_STAT_IDX(pf->hw.pf_id)),
1022 1023
			   pf->stat_offsets_loaded,
			   &osd->fd_atr_match, &nsd->fd_atr_match);
1024 1025
	i40e_stat_update32(hw,
			   I40E_GLQF_PCNT(I40E_FD_SB_STAT_IDX(pf->hw.pf_id)),
1026 1027
			   pf->stat_offsets_loaded,
			   &osd->fd_sb_match, &nsd->fd_sb_match);
1028 1029 1030 1031
	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);
1032

S
Shannon Nelson 已提交
1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046
	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);

1047
	if (pf->flags & I40E_FLAG_FD_SB_ENABLED &&
1048
	    !(pf->hw_disabled_flags & I40E_FLAG_FD_SB_ENABLED))
1049 1050 1051 1052 1053
		nsd->fd_sb_status = true;
	else
		nsd->fd_sb_status = false;

	if (pf->flags & I40E_FLAG_FD_ATR_ENABLED &&
1054
	    !(pf->hw_disabled_flags & I40E_FLAG_FD_ATR_ENABLED))
1055 1056 1057 1058
		nsd->fd_atr_status = true;
	else
		nsd->fd_atr_status = false;

J
Jesse Brandeburg 已提交
1059 1060 1061
	pf->stat_offsets_loaded = true;
}

S
Shannon Nelson 已提交
1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077
/**
 * 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);
}

J
Jesse Brandeburg 已提交
1078 1079 1080 1081 1082 1083 1084 1085 1086
/**
 * 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,
1087
						const u8 *macaddr, s16 vlan)
J
Jesse Brandeburg 已提交
1088 1089
{
	struct i40e_mac_filter *f;
1090
	u64 key;
J
Jesse Brandeburg 已提交
1091 1092 1093 1094

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

1095 1096
	key = i40e_addr_to_hkey(macaddr);
	hash_for_each_possible(vsi->mac_filter_hash, f, hlist, key) {
J
Jesse Brandeburg 已提交
1097
		if ((ether_addr_equal(macaddr, f->macaddr)) &&
1098
		    (vlan == f->vlan))
J
Jesse Brandeburg 已提交
1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111
			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
 **/
1112
struct i40e_mac_filter *i40e_find_mac(struct i40e_vsi *vsi, const u8 *macaddr)
J
Jesse Brandeburg 已提交
1113 1114
{
	struct i40e_mac_filter *f;
1115
	u64 key;
J
Jesse Brandeburg 已提交
1116 1117 1118 1119

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

1120 1121
	key = i40e_addr_to_hkey(macaddr);
	hash_for_each_possible(vsi->mac_filter_hash, f, hlist, key) {
1122
		if ((ether_addr_equal(macaddr, f->macaddr)))
J
Jesse Brandeburg 已提交
1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135
			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)
{
1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148
	/* 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
1149
	 *    i40e_sync_filters_subtask.
1150 1151 1152 1153
	 *
	 * 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
1154
	 * i40e_sync_filters_subtask, but we already have to search
1155 1156 1157 1158
	 * 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
1159
	 */
1160
	return vsi->has_vlan_filter;
J
Jesse Brandeburg 已提交
1161 1162
}

1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196
/**
 * i40e_correct_mac_vlan_filters - Correct non-VLAN filters if necessary
 * @vsi: the VSI to configure
 * @tmp_add_list: list of filters ready to be added
 * @tmp_del_list: list of filters ready to be deleted
 * @vlan_filters: the number of active VLAN filters
 *
 * Update VLAN=0 and VLAN=-1 (I40E_VLAN_ANY) filters properly so that they
 * behave as expected. If we have any active VLAN filters remaining or about
 * to be added then we need to update non-VLAN filters to be marked as VLAN=0
 * so that they only match against untagged traffic. If we no longer have any
 * active VLAN filters, we need to make all non-VLAN filters marked as VLAN=-1
 * so that they match against both tagged and untagged traffic. In this way,
 * we ensure that we correctly receive the desired traffic. This ensures that
 * when we have an active VLAN we will receive only untagged traffic and
 * traffic matching active VLANs. If we have no active VLANs then we will
 * operate in non-VLAN mode and receive all traffic, tagged or untagged.
 *
 * Finally, in a similar fashion, this function also corrects filters when
 * there is an active PVID assigned to this VSI.
 *
 * In case of memory allocation failure return -ENOMEM. Otherwise, return 0.
 *
 * This function is only expected to be called from within
 * i40e_sync_vsi_filters.
 *
 * NOTE: This function expects to be called while under the
 * mac_filter_hash_lock
 */
static int i40e_correct_mac_vlan_filters(struct i40e_vsi *vsi,
					 struct hlist_head *tmp_add_list,
					 struct hlist_head *tmp_del_list,
					 int vlan_filters)
{
1197
	s16 pvid = le16_to_cpu(vsi->info.pvid);
1198
	struct i40e_mac_filter *f, *add_head;
1199
	struct i40e_new_mac_filter *new;
1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217
	struct hlist_node *h;
	int bkt, new_vlan;

	/* To determine if a particular filter needs to be replaced we
	 * have the three following conditions:
	 *
	 * a) if we have a PVID assigned, then all filters which are
	 *    not marked as VLAN=PVID must be replaced with filters that
	 *    are.
	 * b) otherwise, if we have any active VLANS, all filters
	 *    which are marked as VLAN=-1 must be replaced with
	 *    filters marked as VLAN=0
	 * c) finally, if we do not have any active VLANS, all filters
	 *    which are marked as VLAN=0 must be replaced with filters
	 *    marked as VLAN=-1
	 */

	/* Update the filters about to be added in place */
1218
	hlist_for_each_entry(new, tmp_add_list, hlist) {
1219 1220
		if (pvid && new->f->vlan != pvid)
			new->f->vlan = pvid;
1221 1222 1223 1224
		else if (vlan_filters && new->f->vlan == I40E_VLAN_ANY)
			new->f->vlan = 0;
		else if (!vlan_filters && new->f->vlan == 0)
			new->f->vlan = I40E_VLAN_ANY;
1225 1226 1227 1228 1229 1230 1231 1232 1233
	}

	/* Update the remaining active filters */
	hash_for_each_safe(vsi->mac_filter_hash, bkt, h, f, hlist) {
		/* Combine the checks for whether a filter needs to be changed
		 * and then determine the new VLAN inside the if block, in
		 * order to avoid duplicating code for adding the new filter
		 * then deleting the old filter.
		 */
1234
		if ((pvid && f->vlan != pvid) ||
1235 1236 1237
		    (vlan_filters && f->vlan == I40E_VLAN_ANY) ||
		    (!vlan_filters && f->vlan == 0)) {
			/* Determine the new vlan we will be adding */
1238 1239
			if (pvid)
				new_vlan = pvid;
1240 1241 1242 1243 1244 1245 1246 1247 1248 1249
			else if (vlan_filters)
				new_vlan = 0;
			else
				new_vlan = I40E_VLAN_ANY;

			/* Create the new filter */
			add_head = i40e_add_filter(vsi, f->macaddr, new_vlan);
			if (!add_head)
				return -ENOMEM;

1250 1251 1252 1253 1254 1255 1256 1257 1258 1259
			/* Create a temporary i40e_new_mac_filter */
			new = kzalloc(sizeof(*new), GFP_ATOMIC);
			if (!new)
				return -ENOMEM;

			new->f = add_head;
			new->state = add_head->state;

			/* Add the new filter to the tmp list */
			hlist_add_head(&new->hlist, tmp_add_list);
1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272

			/* Put the original filter into the delete list */
			f->state = I40E_FILTER_REMOVE;
			hash_del(&f->hlist);
			hlist_add_head(&f->hlist, tmp_del_list);
		}
	}

	vsi->has_vlan_filter = !!vlan_filters;

	return 0;
}

1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305
/**
 * i40e_rm_default_mac_filter - Remove the default MAC filter set by NVM
 * @vsi: the PF Main VSI - inappropriate for any other VSI
 * @macaddr: the MAC address
 *
 * Remove whatever filter the firmware set up so the driver can manage
 * its own filtering intelligently.
 **/
static void i40e_rm_default_mac_filter(struct i40e_vsi *vsi, u8 *macaddr)
{
	struct i40e_aqc_remove_macvlan_element_data element;
	struct i40e_pf *pf = vsi->back;

	/* Only appropriate for the PF main VSI */
	if (vsi->type != I40E_VSI_MAIN)
		return;

	memset(&element, 0, sizeof(element));
	ether_addr_copy(element.mac_addr, macaddr);
	element.vlan_tag = 0;
	/* Ignore error returns, some firmware does it this way... */
	element.flags = I40E_AQC_MACVLAN_DEL_PERFECT_MATCH;
	i40e_aq_remove_macvlan(&pf->hw, vsi->seid, &element, 1, NULL);

	memset(&element, 0, sizeof(element));
	ether_addr_copy(element.mac_addr, macaddr);
	element.vlan_tag = 0;
	/* ...and some firmware does it this way. */
	element.flags = I40E_AQC_MACVLAN_DEL_PERFECT_MATCH |
			I40E_AQC_MACVLAN_DEL_IGNORE_VLAN;
	i40e_aq_remove_macvlan(&pf->hw, vsi->seid, &element, 1, NULL);
}

J
Jesse Brandeburg 已提交
1306 1307 1308 1309 1310 1311 1312
/**
 * 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
1313
 *
1314
 * NOTE: This function is expected to be called with mac_filter_hash_lock
K
Kiran Patil 已提交
1315
 * being held.
J
Jesse Brandeburg 已提交
1316 1317
 **/
struct i40e_mac_filter *i40e_add_filter(struct i40e_vsi *vsi,
1318
					const u8 *macaddr, s16 vlan)
J
Jesse Brandeburg 已提交
1319 1320
{
	struct i40e_mac_filter *f;
1321
	u64 key;
J
Jesse Brandeburg 已提交
1322 1323 1324 1325

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

1326
	f = i40e_find_filter(vsi, macaddr, vlan);
J
Jesse Brandeburg 已提交
1327 1328 1329
	if (!f) {
		f = kzalloc(sizeof(*f), GFP_ATOMIC);
		if (!f)
1330
			return NULL;
J
Jesse Brandeburg 已提交
1331

1332 1333 1334 1335 1336 1337
		/* Update the boolean indicating if we need to function in
		 * VLAN mode.
		 */
		if (vlan >= 0)
			vsi->has_vlan_filter = true;

1338
		ether_addr_copy(f->macaddr, macaddr);
J
Jesse Brandeburg 已提交
1339
		f->vlan = vlan;
1340 1341 1342 1343 1344 1345 1346 1347
		/* 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;
1348 1349 1350 1351
		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
1352 1353 1354 1355 1356

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

1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367
	/* 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
1368 1369 1370 1371
	return f;
}

/**
1372 1373 1374 1375 1376 1377 1378
 * __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
1379
 *
1380
 * NOTE: This function is expected to be called with mac_filter_hash_lock
K
Kiran Patil 已提交
1381
 * being held.
1382 1383 1384
 * 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
1385
 **/
1386
void __i40e_del_filter(struct i40e_vsi *vsi, struct i40e_mac_filter *f)
J
Jesse Brandeburg 已提交
1387
{
1388
	if (!f)
J
Jesse Brandeburg 已提交
1389 1390
		return;

1391 1392 1393 1394
	/* If the filter was never added to firmware then we can just delete it
	 * directly and we don't want to set the status to remove or else an
	 * admin queue command will unnecessarily fire.
	 */
1395 1396
	if ((f->state == I40E_FILTER_FAILED) ||
	    (f->state == I40E_FILTER_NEW)) {
1397
		hash_del(&f->hlist);
1398
		kfree(f);
J
Jesse Brandeburg 已提交
1399
	} else {
1400
		f->state = I40E_FILTER_REMOVE;
J
Jesse Brandeburg 已提交
1401
	}
1402 1403 1404

	vsi->flags |= I40E_VSI_FLAG_FILTER_CHANGED;
	vsi->back->flags |= I40E_FLAG_FILTER_SYNC;
J
Jesse Brandeburg 已提交
1405 1406
}

1407 1408 1409 1410 1411 1412
/**
 * i40e_del_filter - Remove a MAC/VLAN filter from the VSI
 * @vsi: the VSI to be searched
 * @macaddr: the MAC address
 * @vlan: the VLAN
 *
1413
 * NOTE: This function is expected to be called with mac_filter_hash_lock
1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429
 * 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);
}

1430
/**
1431
 * i40e_add_mac_filter - Add a MAC filter for all active VLANs
1432 1433 1434
 * @vsi: the VSI to be searched
 * @macaddr: the mac address to be filtered
 *
1435 1436
 * If we're not in VLAN mode, just add the filter to I40E_VLAN_ANY. Otherwise,
 * go through all the macvlan filters and add a macvlan filter for each
1437 1438
 * unique vlan that already exists. If a PVID has been assigned, instead only
 * add the macaddr to that VLAN.
1439
 *
1440
 * Returns last filter added on success, else NULL
1441
 **/
1442 1443
struct i40e_mac_filter *i40e_add_mac_filter(struct i40e_vsi *vsi,
					    const u8 *macaddr)
1444
{
1445
	struct i40e_mac_filter *f, *add = NULL;
1446 1447
	struct hlist_node *h;
	int bkt;
1448 1449 1450 1451

	if (vsi->info.pvid)
		return i40e_add_filter(vsi, macaddr,
				       le16_to_cpu(vsi->info.pvid));
1452

1453 1454 1455
	if (!i40e_is_vsi_in_vlan(vsi))
		return i40e_add_filter(vsi, macaddr, I40E_VLAN_ANY);

1456
	hash_for_each_safe(vsi->mac_filter_hash, bkt, h, f, hlist) {
1457 1458
		if (f->state == I40E_FILTER_REMOVE)
			continue;
1459 1460 1461
		add = i40e_add_filter(vsi, macaddr, f->vlan);
		if (!add)
			return NULL;
1462 1463
	}

1464
	return add;
1465 1466 1467
}

/**
1468
 * i40e_del_mac_filter - Remove a MAC filter from all VLANs
1469 1470 1471
 * @vsi: the VSI to be searched
 * @macaddr: the mac address to be removed
 *
1472 1473
 * Removes a given MAC address from a VSI regardless of what VLAN it has been
 * associated with.
1474 1475 1476
 *
 * Returns 0 for success, or error
 **/
1477
int i40e_del_mac_filter(struct i40e_vsi *vsi, const u8 *macaddr)
1478
{
1479 1480
	struct i40e_mac_filter *f;
	struct hlist_node *h;
1481
	bool found = false;
1482
	int bkt;
1483

1484 1485 1486
	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) {
1487 1488 1489 1490
		if (ether_addr_equal(macaddr, f->macaddr)) {
			__i40e_del_filter(vsi, f);
			found = true;
		}
1491
	}
1492 1493

	if (found)
1494
		return 0;
1495 1496
	else
		return -ENOENT;
1497 1498
}

J
Jesse Brandeburg 已提交
1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509
/**
 * 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
 **/
static int i40e_set_mac(struct net_device *netdev, void *p)
{
	struct i40e_netdev_priv *np = netdev_priv(netdev);
	struct i40e_vsi *vsi = np->vsi;
1510 1511
	struct i40e_pf *pf = vsi->back;
	struct i40e_hw *hw = &pf->hw;
J
Jesse Brandeburg 已提交
1512 1513 1514 1515 1516
	struct sockaddr *addr = p;

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

1517 1518 1519 1520 1521
	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
1522

1523 1524 1525 1526
	if (test_bit(__I40E_DOWN, &vsi->back->state) ||
	    test_bit(__I40E_RESET_RECOVERY_PENDING, &vsi->back->state))
		return -EADDRNOTAVAIL;

1527 1528 1529 1530 1531 1532
	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);

1533
	spin_lock_bh(&vsi->mac_filter_hash_lock);
1534 1535
	i40e_del_mac_filter(vsi, netdev->dev_addr);
	i40e_add_mac_filter(vsi, addr->sa_data);
1536
	spin_unlock_bh(&vsi->mac_filter_hash_lock);
1537
	ether_addr_copy(netdev->dev_addr, addr->sa_data);
J
Jesse Brandeburg 已提交
1538 1539
	if (vsi->type == I40E_VSI_MAIN) {
		i40e_status ret;
J
Jesse Brandeburg 已提交
1540

J
Jesse Brandeburg 已提交
1541
		ret = i40e_aq_mac_address_write(&vsi->back->hw,
1542
						I40E_AQC_WRITE_TYPE_LAA_WOL,
J
Jesse Brandeburg 已提交
1543
						addr->sa_data, NULL);
1544 1545 1546 1547
		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));
1548 1549
	}

1550 1551 1552 1553 1554
	/* 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
1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578
}

/**
 * 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.
 **/
static void i40e_vsi_setup_queue_map(struct i40e_vsi *vsi,
				     struct i40e_vsi_context *ctxt,
				     u8 enabled_tc,
				     bool is_add)
{
	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
1579
	u16 num_tc_qps = 0;
J
Jesse Brandeburg 已提交
1580 1581 1582 1583 1584 1585 1586

	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
1587
			if (enabled_tc & BIT(i)) /* TC is enabled */
J
Jesse Brandeburg 已提交
1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600
				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
1601
	/* Number of queues per enabled TC */
1602 1603
	qcount = vsi->alloc_queue_pairs;

1604
	num_tc_qps = qcount / numtc;
1605
	num_tc_qps = min_t(int, num_tc_qps, i40e_pf_get_max_q_per_tc(pf));
J
Jesse Brandeburg 已提交
1606 1607 1608 1609

	/* 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
1610
		if (vsi->tc_config.enabled_tc & BIT(i)) {
1611
			/* TC is enabled */
J
Jesse Brandeburg 已提交
1612 1613 1614 1615
			int pow, num_qps;

			switch (vsi->type) {
			case I40E_VSI_MAIN:
1616 1617
				qcount = min_t(int, pf->alloc_rss_size,
					       num_tc_qps);
J
Jesse Brandeburg 已提交
1618 1619 1620 1621 1622
				break;
			case I40E_VSI_FDIR:
			case I40E_VSI_SRIOV:
			case I40E_VSI_VMDQ2:
			default:
N
Neerav Parikh 已提交
1623
				qcount = num_tc_qps;
J
Jesse Brandeburg 已提交
1624 1625 1626
				WARN_ON(i != 0);
				break;
			}
N
Neerav Parikh 已提交
1627 1628
			vsi->tc_config.tc_info[i].qoffset = offset;
			vsi->tc_config.tc_info[i].qcount = qcount;
J
Jesse Brandeburg 已提交
1629

S
Shannon Nelson 已提交
1630
			/* find the next higher power-of-2 of num queue pairs */
N
Neerav Parikh 已提交
1631
			num_qps = qcount;
J
Jesse Brandeburg 已提交
1632
			pow = 0;
1633
			while (num_qps && (BIT_ULL(pow) < qcount)) {
J
Jesse Brandeburg 已提交
1634 1635 1636 1637 1638 1639 1640 1641 1642
				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
1643
			offset += qcount;
J
Jesse Brandeburg 已提交
1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659
		} 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;
1660 1661 1662
	if ((vsi->type == I40E_VSI_MAIN) && (numtc == 1)) {
		if (vsi->req_queue_pairs > 0)
			vsi->num_queue_pairs = vsi->req_queue_pairs;
1663
		else if (pf->flags & I40E_FLAG_MSIX_ENABLED)
1664 1665
			vsi->num_queue_pairs = pf->num_lan_msix;
	}
J
Jesse Brandeburg 已提交
1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686

	/* 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);
}

1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699
/**
 * 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;

1700
	if (i40e_add_mac_filter(vsi, addr))
1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718
		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;

1719
	i40e_del_mac_filter(vsi, addr);
1720 1721 1722 1723

	return 0;
}

J
Jesse Brandeburg 已提交
1724 1725 1726 1727 1728 1729 1730 1731 1732
/**
 * i40e_set_rx_mode - NDO callback to set the netdev filters
 * @netdev: network interface device structure
 **/
static void i40e_set_rx_mode(struct net_device *netdev)
{
	struct i40e_netdev_priv *np = netdev_priv(netdev);
	struct i40e_vsi *vsi = np->vsi;

1733
	spin_lock_bh(&vsi->mac_filter_hash_lock);
K
Kiran Patil 已提交
1734

1735 1736
	__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
1737

1738
	spin_unlock_bh(&vsi->mac_filter_hash_lock);
J
Jesse Brandeburg 已提交
1739 1740 1741 1742 1743 1744

	/* 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;
	}
1745 1746 1747 1748 1749

	/* 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
1750 1751
}

K
Kiran Patil 已提交
1752
/**
1753
 * i40e_undo_del_filter_entries - Undo the changes made to MAC filter entries
1754
 * @vsi: Pointer to VSI struct
K
Kiran Patil 已提交
1755 1756 1757
 * @from: Pointer to list which contains MAC filter entries - changes to
 *        those entries needs to be undone.
 *
1758
 * MAC filter entries from this list were slated for deletion.
K
Kiran Patil 已提交
1759
 **/
1760 1761
static void i40e_undo_del_filter_entries(struct i40e_vsi *vsi,
					 struct hlist_head *from)
K
Kiran Patil 已提交
1762
{
1763 1764 1765 1766 1767
	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
1768 1769

		/* Move the element back into MAC filter list*/
1770 1771
		hlist_del(&f->hlist);
		hash_add(vsi->mac_filter_hash, &f->hlist, key);
K
Kiran Patil 已提交
1772 1773 1774
	}
}

1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795
/**
 * i40e_undo_add_filter_entries - Undo the changes made to MAC filter entries
 * @vsi: Pointer to vsi struct
 * @from: Pointer to list which contains MAC filter entries - changes to
 *        those entries needs to be undone.
 *
 * MAC filter entries from this list were slated for addition.
 **/
static void i40e_undo_add_filter_entries(struct i40e_vsi *vsi,
					 struct hlist_head *from)
{
	struct i40e_new_mac_filter *new;
	struct hlist_node *h;

	hlist_for_each_entry_safe(new, h, from, hlist) {
		/* We can simply free the wrapper structure */
		hlist_del(&new->hlist);
		kfree(new);
	}
}

1796 1797
/**
 * i40e_next_entry - Get the next non-broadcast filter from a list
1798
 * @next: pointer to filter in list
1799 1800 1801 1802 1803
 *
 * Returns the next non-broadcast filter in the list. Required so that we
 * ignore broadcast filters within the list, since these are not handled via
 * the normal firmware update path.
 */
1804 1805
static
struct i40e_new_mac_filter *i40e_next_filter(struct i40e_new_mac_filter *next)
1806
{
1807 1808 1809
	hlist_for_each_entry_continue(next, hlist) {
		if (!is_broadcast_ether_addr(next->f->macaddr))
			return next;
1810 1811
	}

1812
	return NULL;
1813 1814
}

K
Kiran Patil 已提交
1815
/**
1816 1817 1818 1819 1820
 * 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
K
Kiran Patil 已提交
1821
 *
1822 1823
 * 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
1824
 **/
1825 1826 1827
static int
i40e_update_filter_state(int count,
			 struct i40e_aqc_add_macvlan_element_data *add_list,
1828
			 struct i40e_new_mac_filter *add_head)
K
Kiran Patil 已提交
1829
{
1830 1831
	int retval = 0;
	int i;
K
Kiran Patil 已提交
1832

1833 1834 1835 1836 1837 1838
	for (i = 0; i < count; i++) {
		/* Always check status of each filter. We don't need to check
		 * the firmware return status because we pre-set the filter
		 * status to I40E_AQC_MM_ERR_NO_RES when sending the filter
		 * request to the adminq. Thus, if it no longer matches then
		 * we know the filter is active.
1839
		 */
1840
		if (add_list[i].match_method == I40E_AQC_MM_ERR_NO_RES) {
1841
			add_head->state = I40E_FILTER_FAILED;
1842 1843 1844
		} else {
			add_head->state = I40E_FILTER_ACTIVE;
			retval++;
1845
		}
1846

1847 1848 1849
		add_head = i40e_next_filter(add_head);
		if (!add_head)
			break;
K
Kiran Patil 已提交
1850
	}
1851

1852
	return retval;
K
Kiran Patil 已提交
1853 1854
}

1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905
/**
 * 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,
1906
			  struct i40e_new_mac_filter *add_head,
1907 1908 1909 1910 1911
			  int num_add, bool *promisc_changed)
{
	struct i40e_hw *hw = &vsi->back->hw;
	int aq_err, fcnt;

1912
	i40e_aq_add_macvlan(hw, vsi->seid, list, num_add, NULL);
1913
	aq_err = hw->aq.asq_last_status;
1914
	fcnt = i40e_update_filter_state(num_add, list, add_head);
1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925

	if (fcnt != num_add) {
		*promisc_changed = true;
		set_bit(__I40E_FILTER_OVERFLOW_PROMISC, &vsi->state);
		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);
	}
}

1926 1927 1928 1929 1930 1931 1932 1933
/**
 * i40e_aqc_broadcast_filter - Set promiscuous broadcast flags
 * @vsi: pointer to the VSI
 * @f: filter data
 *
 * This function sets or clears the promiscuous broadcast flags for VLAN
 * filters in order to properly receive broadcast frames. Assumes that only
 * broadcast filters are passed.
1934 1935
 *
 * Returns status indicating success or failure;
1936
 **/
1937 1938 1939
static i40e_status
i40e_aqc_broadcast_filter(struct i40e_vsi *vsi, const char *vsi_name,
			  struct i40e_mac_filter *f)
1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957
{
	bool enable = f->state == I40E_FILTER_NEW;
	struct i40e_hw *hw = &vsi->back->hw;
	i40e_status aq_ret;

	if (f->vlan == I40E_VLAN_ANY) {
		aq_ret = i40e_aq_set_vsi_broadcast(hw,
						   vsi->seid,
						   enable,
						   NULL);
	} else {
		aq_ret = i40e_aq_set_vsi_bc_promisc_on_vlan(hw,
							    vsi->seid,
							    enable,
							    f->vlan,
							    NULL);
	}

1958
	if (aq_ret)
1959 1960 1961 1962
		dev_warn(&vsi->back->pdev->dev,
			 "Error %s setting broadcast promiscuous mode on %s\n",
			 i40e_aq_str(hw, hw->aq.asq_last_status),
			 vsi_name);
1963 1964

	return aq_ret;
1965 1966
}

J
Jesse Brandeburg 已提交
1967 1968 1969 1970 1971 1972 1973 1974
/**
 * 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
1975
int i40e_sync_vsi_filters(struct i40e_vsi *vsi)
J
Jesse Brandeburg 已提交
1976
{
1977
	struct hlist_head tmp_add_list, tmp_del_list;
1978 1979
	struct i40e_mac_filter *f;
	struct i40e_new_mac_filter *new, *add_head = NULL;
1980
	struct i40e_hw *hw = &vsi->back->hw;
1981
	unsigned int failed_filters = 0;
1982
	unsigned int vlan_filters = 0;
1983
	bool promisc_changed = false;
1984
	char vsi_name[16] = "PF";
J
Jesse Brandeburg 已提交
1985
	int filter_list_len = 0;
M
Mitch Williams 已提交
1986
	i40e_status aq_ret = 0;
1987
	u32 changed_flags = 0;
1988
	struct hlist_node *h;
J
Jesse Brandeburg 已提交
1989 1990 1991
	struct i40e_pf *pf;
	int num_add = 0;
	int num_del = 0;
1992
	int retval = 0;
J
Jesse Brandeburg 已提交
1993
	u16 cmd_flags;
1994
	int list_size;
1995
	int bkt;
J
Jesse Brandeburg 已提交
1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009

	/* 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;
	}

2010 2011
	INIT_HLIST_HEAD(&tmp_add_list);
	INIT_HLIST_HEAD(&tmp_del_list);
K
Kiran Patil 已提交
2012

2013 2014 2015 2016 2017
	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
2018 2019 2020
	if (vsi->flags & I40E_VSI_FLAG_FILTER_CHANGED) {
		vsi->flags &= ~I40E_VSI_FLAG_FILTER_CHANGED;

2021
		spin_lock_bh(&vsi->mac_filter_hash_lock);
2022
		/* Create a list of filters to delete. */
2023
		hash_for_each_safe(vsi->mac_filter_hash, bkt, h, f, hlist) {
2024 2025
			if (f->state == I40E_FILTER_REMOVE) {
				/* Move the element into temporary del_list */
2026 2027
				hash_del(&f->hlist);
				hlist_add_head(&f->hlist, &tmp_del_list);
2028 2029 2030

				/* Avoid counting removed filters */
				continue;
2031 2032
			}
			if (f->state == I40E_FILTER_NEW) {
2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043
				/* Create a temporary i40e_new_mac_filter */
				new = kzalloc(sizeof(*new), GFP_ATOMIC);
				if (!new)
					goto err_no_memory_locked;

				/* Store pointer to the real filter */
				new->f = f;
				new->state = f->state;

				/* Add it to the hash list */
				hlist_add_head(&new->hlist, &tmp_add_list);
K
Kiran Patil 已提交
2044
			}
2045

2046 2047 2048
			/* Count the number of active (current and new) VLAN
			 * filters we have now. Does not count filters which
			 * are marked for deletion.
2049 2050 2051 2052 2053
			 */
			if (f->vlan > 0)
				vlan_filters++;
		}

2054 2055 2056 2057 2058 2059
		retval = i40e_correct_mac_vlan_filters(vsi,
						       &tmp_add_list,
						       &tmp_del_list,
						       vlan_filters);
		if (retval)
			goto err_no_memory_locked;
2060

2061
		spin_unlock_bh(&vsi->mac_filter_hash_lock);
K
Kiran Patil 已提交
2062 2063 2064
	}

	/* Now process 'del_list' outside the lock */
2065
	if (!hlist_empty(&tmp_del_list)) {
2066
		filter_list_len = hw->aq.asq_buf_size /
K
Kiran Patil 已提交
2067
			    sizeof(struct i40e_aqc_remove_macvlan_element_data);
2068
		list_size = filter_list_len *
S
Shannon Nelson 已提交
2069
			    sizeof(struct i40e_aqc_remove_macvlan_element_data);
2070
		del_list = kzalloc(list_size, GFP_ATOMIC);
2071 2072
		if (!del_list)
			goto err_no_memory;
K
Kiran Patil 已提交
2073

2074
		hlist_for_each_entry_safe(f, h, &tmp_del_list, hlist) {
J
Jesse Brandeburg 已提交
2075 2076
			cmd_flags = 0;

2077
			/* handle broadcast filters by updating the broadcast
2078
			 * promiscuous flag and release filter list.
2079 2080 2081 2082 2083 2084 2085 2086 2087
			 */
			if (is_broadcast_ether_addr(f->macaddr)) {
				i40e_aqc_broadcast_filter(vsi, vsi_name, f);

				hlist_del(&f->hlist);
				kfree(f);
				continue;
			}

J
Jesse Brandeburg 已提交
2088
			/* add to delete list */
2089
			ether_addr_copy(del_list[num_del].mac_addr, f->macaddr);
2090 2091
			if (f->vlan == I40E_VLAN_ANY) {
				del_list[num_del].vlan_tag = 0;
A
Alan Brady 已提交
2092
				cmd_flags |= I40E_AQC_MACVLAN_DEL_IGNORE_VLAN;
2093 2094 2095 2096
			} else {
				del_list[num_del].vlan_tag =
					cpu_to_le16((u16)(f->vlan));
			}
J
Jesse Brandeburg 已提交
2097 2098 2099 2100 2101 2102 2103

			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) {
2104 2105
				i40e_aqc_del_filters(vsi, vsi_name, del_list,
						     num_del, &retval);
2106
				memset(del_list, 0, list_size);
2107
				num_del = 0;
J
Jesse Brandeburg 已提交
2108
			}
K
Kiran Patil 已提交
2109 2110 2111
			/* Release memory for MAC filter entries which were
			 * synced up with HW.
			 */
2112
			hlist_del(&f->hlist);
K
Kiran Patil 已提交
2113
			kfree(f);
J
Jesse Brandeburg 已提交
2114
		}
K
Kiran Patil 已提交
2115

J
Jesse Brandeburg 已提交
2116
		if (num_del) {
2117 2118
			i40e_aqc_del_filters(vsi, vsi_name, del_list,
					     num_del, &retval);
J
Jesse Brandeburg 已提交
2119 2120 2121 2122
		}

		kfree(del_list);
		del_list = NULL;
K
Kiran Patil 已提交
2123 2124
	}

2125
	if (!hlist_empty(&tmp_add_list)) {
2126
		/* Do all the adds now. */
2127
		filter_list_len = hw->aq.asq_buf_size /
S
Shannon Nelson 已提交
2128
			       sizeof(struct i40e_aqc_add_macvlan_element_data);
2129 2130 2131
		list_size = filter_list_len *
			       sizeof(struct i40e_aqc_add_macvlan_element_data);
		add_list = kzalloc(list_size, GFP_ATOMIC);
2132 2133 2134
		if (!add_list)
			goto err_no_memory;

2135
		num_add = 0;
2136
		hlist_for_each_entry_safe(new, h, &tmp_add_list, hlist) {
2137 2138
			if (test_bit(__I40E_FILTER_OVERFLOW_PROMISC,
				     &vsi->state)) {
2139
				new->state = I40E_FILTER_FAILED;
2140 2141
				continue;
			}
2142 2143 2144 2145

			/* handle broadcast filters by updating the broadcast
			 * promiscuous flag instead of adding a MAC filter.
			 */
2146 2147 2148 2149 2150 2151
			if (is_broadcast_ether_addr(new->f->macaddr)) {
				if (i40e_aqc_broadcast_filter(vsi, vsi_name,
							      new->f))
					new->state = I40E_FILTER_FAILED;
				else
					new->state = I40E_FILTER_ACTIVE;
2152 2153 2154
				continue;
			}

J
Jesse Brandeburg 已提交
2155
			/* add to add array */
2156
			if (num_add == 0)
2157
				add_head = new;
2158
			cmd_flags = 0;
2159 2160 2161
			ether_addr_copy(add_list[num_add].mac_addr,
					new->f->macaddr);
			if (new->f->vlan == I40E_VLAN_ANY) {
2162 2163 2164 2165
				add_list[num_add].vlan_tag = 0;
				cmd_flags |= I40E_AQC_MACVLAN_ADD_IGNORE_VLAN;
			} else {
				add_list[num_add].vlan_tag =
2166
					cpu_to_le16((u16)(new->f->vlan));
2167
			}
J
Jesse Brandeburg 已提交
2168
			add_list[num_add].queue_number = 0;
2169
			/* set invalid match method for later detection */
2170
			add_list[num_add].match_method = I40E_AQC_MM_ERR_NO_RES;
J
Jesse Brandeburg 已提交
2171 2172 2173 2174 2175 2176
			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) {
2177 2178 2179
				i40e_aqc_add_filters(vsi, vsi_name, add_list,
						     add_head, num_add,
						     &promisc_changed);
2180
				memset(add_list, 0, list_size);
J
Jesse Brandeburg 已提交
2181 2182 2183 2184
				num_add = 0;
			}
		}
		if (num_add) {
2185 2186
			i40e_aqc_add_filters(vsi, vsi_name, add_list, add_head,
					     num_add, &promisc_changed);
J
Jesse Brandeburg 已提交
2187
		}
2188 2189 2190
		/* Now move all of the filters from the temp add list back to
		 * the VSI's list.
		 */
2191
		spin_lock_bh(&vsi->mac_filter_hash_lock);
2192 2193 2194 2195 2196 2197
		hlist_for_each_entry_safe(new, h, &tmp_add_list, hlist) {
			/* Only update the state if we're still NEW */
			if (new->f->state == I40E_FILTER_NEW)
				new->f->state = new->state;
			hlist_del(&new->hlist);
			kfree(new);
2198
		}
2199
		spin_unlock_bh(&vsi->mac_filter_hash_lock);
J
Jesse Brandeburg 已提交
2200 2201
		kfree(add_list);
		add_list = NULL;
2202
	}
J
Jesse Brandeburg 已提交
2203

2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 2224
	/* Determine the number of active and failed filters. */
	spin_lock_bh(&vsi->mac_filter_hash_lock);
	vsi->active_filters = 0;
	hash_for_each(vsi->mac_filter_hash, bkt, f, hlist) {
		if (f->state == I40E_FILTER_ACTIVE)
			vsi->active_filters++;
		else if (f->state == I40E_FILTER_FAILED)
			failed_filters++;
	}
	spin_unlock_bh(&vsi->mac_filter_hash_lock);

	/* If promiscuous mode has changed, we need to calculate a new
	 * threshold for when we are safe to exit
	 */
	if (promisc_changed)
		vsi->promisc_threshold = (vsi->active_filters * 3) / 4;

	/* Check if we are able to exit overflow promiscuous mode. We can
	 * safely exit if we didn't just enter, we no longer have any failed
	 * filters, and we have reduced filters below the threshold value.
	 */
2225
	if (test_bit(__I40E_FILTER_OVERFLOW_PROMISC, &vsi->state) &&
2226
	    !promisc_changed && !failed_filters &&
2227
	    (vsi->active_filters < vsi->promisc_threshold)) {
2228 2229 2230 2231 2232 2233
		dev_info(&pf->pdev->dev,
			 "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
2234 2235
	}

2236 2237 2238 2239 2240 2241
	/* 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
2242 2243 2244
	/* check for changes in promiscuous modes */
	if (changed_flags & IFF_ALLMULTI) {
		bool cur_multipromisc;
J
Jesse Brandeburg 已提交
2245

J
Jesse Brandeburg 已提交
2246
		cur_multipromisc = !!(vsi->current_netdev_flags & IFF_ALLMULTI);
M
Mitch Williams 已提交
2247 2248 2249 2250 2251 2252
		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,
2253
						     hw->aq.asq_last_status);
J
Jesse Brandeburg 已提交
2254
			dev_info(&pf->pdev->dev,
2255 2256
				 "set multi promisc failed on %s, err %s aq_err %s\n",
				 vsi_name,
2257 2258
				 i40e_stat_str(hw, aq_ret),
				 i40e_aq_str(hw, hw->aq.asq_last_status));
M
Mitch Williams 已提交
2259
		}
J
Jesse Brandeburg 已提交
2260
	}
2261 2262 2263
	if ((changed_flags & IFF_PROMISC) ||
	    (promisc_changed &&
	     test_bit(__I40E_FILTER_OVERFLOW_PROMISC, &vsi->state))) {
J
Jesse Brandeburg 已提交
2264
		bool cur_promisc;
J
Jesse Brandeburg 已提交
2265

J
Jesse Brandeburg 已提交
2266 2267 2268
		cur_promisc = (!!(vsi->current_netdev_flags & IFF_PROMISC) ||
			       test_bit(__I40E_FILTER_OVERFLOW_PROMISC,
					&vsi->state));
2269 2270 2271
		if ((vsi->type == I40E_VSI_MAIN) &&
		    (pf->lan_veb != I40E_NO_VEB) &&
		    !(pf->flags & I40E_FLAG_MFP_ENABLED)) {
2272 2273 2274 2275 2276 2277 2278
			/* 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;
2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292
				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,
2293 2294
						 "Set default VSI failed on %s, err %s, aq_err %s\n",
						 vsi_name,
2295 2296 2297 2298
						 i40e_stat_str(hw, aq_ret),
						 i40e_aq_str(hw,
						     hw->aq.asq_last_status));
				}
2299 2300
			}
		} else {
M
Mitch Williams 已提交
2301
			aq_ret = i40e_aq_set_vsi_unicast_promiscuous(
2302
							  hw,
2303
							  vsi->seid,
2304 2305
							  cur_promisc, NULL,
							  true);
M
Mitch Williams 已提交
2306 2307 2308
			if (aq_ret) {
				retval =
				i40e_aq_rc_to_posix(aq_ret,
2309
						    hw->aq.asq_last_status);
2310
				dev_info(&pf->pdev->dev,
2311 2312
					 "set unicast promisc failed on %s, err %s, aq_err %s\n",
					 vsi_name,
2313 2314 2315
					 i40e_stat_str(hw, aq_ret),
					 i40e_aq_str(hw,
						     hw->aq.asq_last_status));
M
Mitch Williams 已提交
2316 2317
			}
			aq_ret = i40e_aq_set_vsi_multicast_promiscuous(
2318
							  hw,
2319 2320
							  vsi->seid,
							  cur_promisc, NULL);
M
Mitch Williams 已提交
2321 2322 2323
			if (aq_ret) {
				retval =
				i40e_aq_rc_to_posix(aq_ret,
2324
						    hw->aq.asq_last_status);
2325
				dev_info(&pf->pdev->dev,
2326 2327
					 "set multicast promisc failed on %s, err %s, aq_err %s\n",
					 vsi_name,
2328 2329 2330
					 i40e_stat_str(hw, aq_ret),
					 i40e_aq_str(hw,
						     hw->aq.asq_last_status));
M
Mitch Williams 已提交
2331
			}
2332
		}
M
Mitch Williams 已提交
2333 2334 2335 2336 2337 2338
		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);
2339
			dev_info(&pf->pdev->dev,
2340
				 "set brdcast promisc failed, err %s, aq_err %s\n",
2341 2342 2343
					 i40e_stat_str(hw, aq_ret),
					 i40e_aq_str(hw,
						     hw->aq.asq_last_status));
M
Mitch Williams 已提交
2344
		}
J
Jesse Brandeburg 已提交
2345
	}
M
Mitch Williams 已提交
2346
out:
2347 2348 2349 2350
	/* 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
2351
	clear_bit(__I40E_CONFIG_BUSY, &vsi->state);
M
Mitch Williams 已提交
2352
	return retval;
2353 2354 2355 2356

err_no_memory:
	/* Restore elements on the temporary add and delete lists */
	spin_lock_bh(&vsi->mac_filter_hash_lock);
2357
err_no_memory_locked:
2358 2359
	i40e_undo_del_filter_entries(vsi, &tmp_del_list);
	i40e_undo_add_filter_entries(vsi, &tmp_add_list);
2360 2361 2362 2363 2364
	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
2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378
}

/**
 * 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
2379
	for (v = 0; v < pf->num_alloc_vsi; v++) {
J
Jesse Brandeburg 已提交
2380
		if (pf->vsi[v] &&
J
Jesse Brandeburg 已提交
2381 2382 2383 2384 2385 2386 2387 2388 2389
		    (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
2390 2391 2392 2393 2394 2395 2396 2397 2398 2399 2400 2401 2402 2403
	}
}

/**
 * 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;
M
Mitch Williams 已提交
2404
	struct i40e_pf *pf = vsi->back;
J
Jesse Brandeburg 已提交
2405 2406 2407 2408 2409 2410

	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);
M
Mitch Williams 已提交
2411 2412
	pf->flags |= (I40E_FLAG_SERVICE_CLIENT_REQUESTED |
		      I40E_FLAG_CLIENT_L2_CHANGE);
J
Jesse Brandeburg 已提交
2413 2414 2415
	return 0;
}

J
Jacob Keller 已提交
2416 2417 2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435 2436
/**
 * 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
2437 2438 2439 2440 2441 2442 2443 2444 2445 2446 2447 2448 2449 2450 2451 2452 2453 2454 2455
/**
 * 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;
2456
	ctxt.info = vsi->info;
J
Jesse Brandeburg 已提交
2457 2458 2459
	ret = i40e_aq_update_vsi_params(&vsi->back->hw, &ctxt, NULL);
	if (ret) {
		dev_info(&vsi->back->pdev->dev,
2460 2461 2462 2463
			 "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
2464 2465 2466 2467 2468 2469 2470 2471 2472 2473 2474 2475 2476 2477 2478 2479 2480 2481 2482 2483 2484 2485 2486
	}
}

/**
 * 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;
2487
	ctxt.info = vsi->info;
J
Jesse Brandeburg 已提交
2488 2489 2490
	ret = i40e_aq_update_vsi_params(&vsi->back->hw, &ctxt, NULL);
	if (ret) {
		dev_info(&vsi->back->pdev->dev,
2491 2492 2493 2494
			 "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
2495 2496 2497 2498 2499 2500 2501 2502 2503 2504 2505 2506 2507 2508 2509 2510 2511 2512 2513 2514
	}
}

/**
 * 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);
}

/**
2515
 * i40e_add_vlan_all_mac - Add a MAC/VLAN filter for each existing MAC address
J
Jesse Brandeburg 已提交
2516 2517
 * @vsi: the vsi being configured
 * @vid: vlan id to be added (0 = untagged only , -1 = any)
2518 2519 2520 2521 2522 2523 2524 2525
 *
 * This is a helper function for adding a new MAC/VLAN filter with the
 * specified VLAN for each existing MAC address already in the hash table.
 * This function does *not* perform any accounting to update filters based on
 * VLAN mode.
 *
 * NOTE: this function expects to be called while under the
 * mac_filter_hash_lock
J
Jesse Brandeburg 已提交
2526
 **/
2527
int i40e_add_vlan_all_mac(struct i40e_vsi *vsi, s16 vid)
J
Jesse Brandeburg 已提交
2528
{
2529
	struct i40e_mac_filter *f, *add_f;
2530 2531
	struct hlist_node *h;
	int bkt;
J
Jesse Brandeburg 已提交
2532

2533
	hash_for_each_safe(vsi->mac_filter_hash, bkt, h, f, hlist) {
2534 2535
		if (f->state == I40E_FILTER_REMOVE)
			continue;
2536
		add_f = i40e_add_filter(vsi, f->macaddr, vid);
J
Jesse Brandeburg 已提交
2537 2538 2539 2540 2541 2542 2543 2544
		if (!add_f) {
			dev_info(&vsi->back->pdev->dev,
				 "Could not add vlan filter %d for %pM\n",
				 vid, f->macaddr);
			return -ENOMEM;
		}
	}

2545 2546 2547 2548 2549 2550
	return 0;
}

/**
 * i40e_vsi_add_vlan - Add VSI membership for given VLAN
 * @vsi: the VSI being configured
2551
 * @vid: VLAN id to be added
2552
 **/
2553
int i40e_vsi_add_vlan(struct i40e_vsi *vsi, u16 vid)
2554
{
2555
	int err;
2556

2557 2558 2559
	if (!vid || vsi->info.pvid)
		return -EINVAL;

2560 2561 2562
	/* Locked once because all functions invoked below iterates list*/
	spin_lock_bh(&vsi->mac_filter_hash_lock);
	err = i40e_add_vlan_all_mac(vsi, vid);
2563
	spin_unlock_bh(&vsi->mac_filter_hash_lock);
2564 2565
	if (err)
		return err;
K
Kiran Patil 已提交
2566

2567 2568 2569 2570 2571
	/* 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
2572 2573 2574
}

/**
2575
 * i40e_rm_vlan_all_mac - Remove MAC/VLAN pair for all MAC with the given VLAN
J
Jesse Brandeburg 已提交
2576 2577
 * @vsi: the vsi being configured
 * @vid: vlan id to be removed (0 = untagged only , -1 = any)
2578 2579 2580 2581 2582 2583 2584 2585 2586
 *
 * This function should be used to remove all VLAN filters which match the
 * given VID. It does not schedule the service event and does not take the
 * mac_filter_hash_lock so it may be combined with other operations under
 * a single invocation of the mac_filter_hash_lock.
 *
 * NOTE: this function expects to be called while under the
 * mac_filter_hash_lock
 */
2587
void i40e_rm_vlan_all_mac(struct i40e_vsi *vsi, s16 vid)
J
Jesse Brandeburg 已提交
2588
{
2589
	struct i40e_mac_filter *f;
2590 2591
	struct hlist_node *h;
	int bkt;
J
Jesse Brandeburg 已提交
2592

2593
	hash_for_each_safe(vsi->mac_filter_hash, bkt, h, f, hlist) {
2594 2595 2596
		if (f->vlan == vid)
			__i40e_del_filter(vsi, f);
	}
2597
}
J
Jesse Brandeburg 已提交
2598

2599 2600 2601
/**
 * i40e_vsi_kill_vlan - Remove VSI membership for given VLAN
 * @vsi: the VSI being configured
2602
 * @vid: VLAN id to be removed
2603
 **/
2604
void i40e_vsi_kill_vlan(struct i40e_vsi *vsi, u16 vid)
2605
{
2606 2607 2608
	if (!vid || vsi->info.pvid)
		return;

2609 2610
	spin_lock_bh(&vsi->mac_filter_hash_lock);
	i40e_rm_vlan_all_mac(vsi, vid);
2611
	spin_unlock_bh(&vsi->mac_filter_hash_lock);
K
Kiran Patil 已提交
2612

2613 2614 2615 2616
	/* 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
2617 2618 2619 2620 2621 2622
}

/**
 * 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
2623 2624
 *
 * net_device_ops implementation for adding vlan ids
J
Jesse Brandeburg 已提交
2625 2626 2627 2628 2629 2630
 **/
static int i40e_vlan_rx_add_vid(struct net_device *netdev,
				__always_unused __be16 proto, u16 vid)
{
	struct i40e_netdev_priv *np = netdev_priv(netdev);
	struct i40e_vsi *vsi = np->vsi;
J
Jesse Brandeburg 已提交
2631
	int ret = 0;
J
Jesse Brandeburg 已提交
2632

2633
	if (vid >= VLAN_N_VID)
J
Jesse Brandeburg 已提交
2634 2635
		return -EINVAL;

2636 2637 2638 2639 2640
	/* 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
2641
	 */
2642 2643
	if (vid)
		ret = i40e_vsi_add_vlan(vsi, vid);
J
Jesse Brandeburg 已提交
2644

2645
	if (!ret)
J
Jesse Brandeburg 已提交
2646
		set_bit(vid, vsi->active_vlans);
J
Jesse Brandeburg 已提交
2647

J
Jesse Brandeburg 已提交
2648
	return ret;
J
Jesse Brandeburg 已提交
2649 2650 2651 2652 2653 2654
}

/**
 * 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
2655
 *
A
Akeem G Abodunrin 已提交
2656
 * net_device_ops implementation for removing vlan ids
J
Jesse Brandeburg 已提交
2657 2658 2659 2660 2661 2662 2663 2664 2665
 **/
static int i40e_vlan_rx_kill_vid(struct net_device *netdev,
				 __always_unused __be16 proto, u16 vid)
{
	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
2666
	 * already printed from the other function
J
Jesse Brandeburg 已提交
2667 2668 2669 2670
	 */
	i40e_vsi_kill_vlan(vsi, vid);

	clear_bit(vid, vsi->active_vlans);
J
Jesse Brandeburg 已提交
2671

J
Jesse Brandeburg 已提交
2672 2673 2674
	return 0;
}

2675 2676 2677 2678 2679 2680 2681 2682 2683 2684 2685 2686 2687 2688 2689 2690 2691 2692 2693 2694 2695 2696 2697 2698 2699 2700 2701 2702 2703 2704 2705 2706 2707 2708 2709 2710 2711 2712
/**
 * 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
2713 2714 2715 2716 2717 2718 2719 2720 2721 2722 2723 2724 2725 2726 2727 2728 2729 2730 2731 2732 2733 2734 2735
/**
 * 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
2736
int i40e_vsi_add_pvid(struct i40e_vsi *vsi, u16 vid)
J
Jesse Brandeburg 已提交
2737 2738
{
	struct i40e_vsi_context ctxt;
2739
	i40e_status ret;
J
Jesse Brandeburg 已提交
2740 2741 2742

	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
2743 2744
	vsi->info.port_vlan_flags = I40E_AQ_VSI_PVLAN_MODE_TAGGED |
				    I40E_AQ_VSI_PVLAN_INSERT_PVID |
G
Greg Rose 已提交
2745
				    I40E_AQ_VSI_PVLAN_EMOD_STR;
J
Jesse Brandeburg 已提交
2746 2747

	ctxt.seid = vsi->seid;
2748
	ctxt.info = vsi->info;
2749 2750
	ret = i40e_aq_update_vsi_params(&vsi->back->hw, &ctxt, NULL);
	if (ret) {
J
Jesse Brandeburg 已提交
2751
		dev_info(&vsi->back->pdev->dev,
2752 2753 2754 2755
			 "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
2756
		return -ENOENT;
J
Jesse Brandeburg 已提交
2757 2758
	}

J
Jesse Brandeburg 已提交
2759
	return 0;
J
Jesse Brandeburg 已提交
2760 2761 2762 2763 2764 2765 2766 2767 2768 2769
}

/**
 * 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
2770 2771
	i40e_vlan_stripping_disable(vsi);

J
Jesse Brandeburg 已提交
2772 2773 2774 2775 2776 2777 2778 2779 2780 2781 2782 2783 2784 2785 2786 2787 2788 2789
	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++)
2790
		err = i40e_setup_tx_descriptors(vsi->tx_rings[i]);
J
Jesse Brandeburg 已提交
2791 2792 2793 2794 2795 2796 2797 2798 2799 2800 2801 2802 2803 2804

	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;

2805 2806 2807
	if (!vsi->tx_rings)
		return;

J
Jesse Brandeburg 已提交
2808
	for (i = 0; i < vsi->num_queue_pairs; i++)
2809
		if (vsi->tx_rings[i] && vsi->tx_rings[i]->desc)
2810
			i40e_free_tx_resources(vsi->tx_rings[i]);
J
Jesse Brandeburg 已提交
2811 2812 2813 2814 2815 2816 2817 2818 2819 2820 2821 2822 2823 2824 2825 2826 2827
}

/**
 * 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++)
2828
		err = i40e_setup_rx_descriptors(vsi->rx_rings[i]);
J
Jesse Brandeburg 已提交
2829 2830 2831 2832 2833 2834 2835 2836 2837 2838 2839 2840 2841
	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;

2842 2843 2844
	if (!vsi->rx_rings)
		return;

J
Jesse Brandeburg 已提交
2845
	for (i = 0; i < vsi->num_queue_pairs; i++)
2846
		if (vsi->rx_rings[i] && vsi->rx_rings[i]->desc)
2847
			i40e_free_rx_resources(vsi->rx_rings[i]);
J
Jesse Brandeburg 已提交
2848 2849
}

2850 2851 2852 2853 2854 2855 2856 2857 2858 2859 2860 2861
/**
 * 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;

2862 2863 2864 2865 2866 2867
	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))
2868 2869 2870
			netif_set_xps_queue(ring->netdev,
					    &ring->q_vector->affinity_mask,
					    ring->queue_index);
2871 2872 2873 2874 2875
	} 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);
2876
	}
2877 2878 2879 2880 2881

	/* schedule our worker thread which will take care of
	 * applying the new filter changes
	 */
	i40e_service_event_schedule(vsi->back);
2882 2883
}

J
Jesse Brandeburg 已提交
2884 2885 2886 2887 2888 2889 2890 2891 2892 2893 2894 2895 2896 2897 2898 2899
/**
 * 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
2900
	if (vsi->back->flags & I40E_FLAG_FD_ATR_ENABLED) {
J
Jesse Brandeburg 已提交
2901 2902 2903 2904 2905 2906
		ring->atr_sample_rate = vsi->back->atr_sample_rate;
		ring->atr_count = 0;
	} else {
		ring->atr_sample_rate = 0;
	}

2907 2908
	/* configure XPS */
	i40e_config_xps_tx_ring(ring);
J
Jesse Brandeburg 已提交
2909 2910 2911 2912 2913 2914 2915

	/* 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
2916 2917
	tx_ctx.fd_ena = !!(vsi->back->flags & (I40E_FLAG_FD_SB_ENABLED |
					       I40E_FLAG_FD_ATR_ENABLED));
J
Jacob Keller 已提交
2918
	tx_ctx.timesync_ena = !!(vsi->back->flags & I40E_FLAG_PTP);
2919 2920 2921 2922 2923
	/* 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
2924 2925 2926 2927 2928 2929 2930 2931 2932 2933 2934 2935 2936 2937 2938 2939 2940 2941 2942 2943 2944 2945 2946 2947 2948 2949 2950 2951 2952 2953 2954 2955 2956

	/* 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
2957
	if (vsi->type == I40E_VSI_VMDQ2) {
2958
		qtx_ctl = I40E_QTX_CTL_VM_QUEUE;
M
Mitch Williams 已提交
2959 2960 2961
		qtx_ctl |= ((vsi->id) << I40E_QTX_CTL_VFVM_INDX_SHIFT) &
			   I40E_QTX_CTL_VFVM_INDX_MASK;
	} else {
2962
		qtx_ctl = I40E_QTX_CTL_PF_QUEUE;
M
Mitch Williams 已提交
2963 2964
	}

2965 2966
	qtx_ctl |= ((hw->pf_id << I40E_QTX_CTL_PF_INDX_SHIFT) &
		    I40E_QTX_CTL_PF_INDX_MASK);
J
Jesse Brandeburg 已提交
2967 2968 2969 2970 2971 2972 2973 2974 2975 2976 2977 2978 2979 2980 2981 2982 2983 2984 2985 2986 2987 2988 2989 2990 2991 2992 2993 2994 2995 2996 2997
	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;

2998 2999
	rx_ctx.dbuff = DIV_ROUND_UP(ring->rx_buf_len,
				    BIT_ULL(I40E_RXQ_CTX_DBUFF_SHIFT));
J
Jesse Brandeburg 已提交
3000 3001 3002 3003

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

3004 3005
	/* use 32 byte descriptors */
	rx_ctx.dsize = 1;
J
Jesse Brandeburg 已提交
3006

3007 3008 3009
	/* descriptor type is always zero
	 * rx_ctx.dtype = 0;
	 */
3010
	rx_ctx.hsplit_0 = 0;
J
Jesse Brandeburg 已提交
3011

3012
	rx_ctx.rxmax = min_t(u16, vsi->max_frame, chain_len * ring->rx_buf_len);
3013 3014 3015 3016
	if (hw->revision_id == 0)
		rx_ctx.lrxqthresh = 0;
	else
		rx_ctx.lrxqthresh = 2;
J
Jesse Brandeburg 已提交
3017 3018
	rx_ctx.crcstrip = 1;
	rx_ctx.l2tsel = 1;
J
Jesse Brandeburg 已提交
3019 3020
	/* this controls whether VLAN is stripped from inner headers */
	rx_ctx.showiv = 0;
3021 3022
	/* set the prefena field to 1 because the manual says to */
	rx_ctx.prefena = 1;
J
Jesse Brandeburg 已提交
3023 3024 3025 3026 3027 3028 3029 3030 3031 3032 3033 3034 3035 3036 3037 3038 3039 3040 3041 3042 3043 3044 3045

	/* 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);

3046
	i40e_alloc_rx_buffers(ring, I40E_DESC_UNUSED(ring));
J
Jesse Brandeburg 已提交
3047 3048 3049 3050 3051 3052 3053 3054 3055 3056 3057 3058 3059 3060 3061

	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;

3062 3063
	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
3064 3065 3066 3067 3068 3069 3070 3071 3072 3073 3074 3075 3076 3077 3078

	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;

3079 3080 3081 3082 3083 3084 3085 3086 3087 3088 3089 3090
	if (!vsi->netdev || (vsi->back->flags & I40E_FLAG_LEGACY_RX)) {
		vsi->max_frame = I40E_MAX_RXBUFFER;
		vsi->rx_buf_len = I40E_RXBUFFER_2048;
#if (PAGE_SIZE < 8192)
	} else if (vsi->netdev->mtu <= ETH_DATA_LEN) {
		vsi->max_frame = I40E_RXBUFFER_1536 - NET_IP_ALIGN;
		vsi->rx_buf_len = I40E_RXBUFFER_1536 - NET_IP_ALIGN;
#endif
	} else {
		vsi->max_frame = I40E_MAX_RXBUFFER;
		vsi->rx_buf_len = I40E_RXBUFFER_2048;
	}
J
Jesse Brandeburg 已提交
3091 3092 3093

	/* set up individual rings */
	for (i = 0; i < vsi->num_queue_pairs && !err; i++)
3094
		err = i40e_configure_rx_ring(vsi->rx_rings[i]);
J
Jesse Brandeburg 已提交
3095 3096 3097 3098 3099 3100 3101 3102 3103 3104

	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)
{
3105
	struct i40e_ring *tx_ring, *rx_ring;
J
Jesse Brandeburg 已提交
3106 3107 3108
	u16 qoffset, qcount;
	int i, n;

3109 3110 3111 3112 3113 3114 3115 3116 3117
	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
3118 3119

	for (n = 0; n < I40E_MAX_TRAFFIC_CLASS; n++) {
3120
		if (!(vsi->tc_config.enabled_tc & BIT_ULL(n)))
J
Jesse Brandeburg 已提交
3121 3122 3123 3124 3125
			continue;

		qoffset = vsi->tc_config.tc_info[n].qoffset;
		qcount = vsi->tc_config.tc_info[n].qcount;
		for (i = qoffset; i < (qoffset + qcount); i++) {
3126 3127
			rx_ring = vsi->rx_rings[i];
			tx_ring = vsi->tx_rings[i];
J
Jesse Brandeburg 已提交
3128 3129 3130 3131 3132 3133 3134 3135 3136 3137 3138 3139
			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)
{
3140 3141 3142
	struct i40e_pf *pf = vsi->back;
	int err;

J
Jesse Brandeburg 已提交
3143 3144
	if (vsi->netdev)
		i40e_set_rx_mode(vsi->netdev);
3145 3146 3147 3148 3149 3150 3151 3152

	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
3153 3154
}

3155 3156 3157 3158 3159 3160 3161 3162 3163 3164 3165 3166 3167
/**
 * 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;

3168 3169 3170
	if (!(pf->flags & I40E_FLAG_FD_SB_ENABLED))
		return;

3171
	/* Reset FDir counters as we're replaying all existing filters */
3172 3173
	pf->fd_tcp4_filter_cnt = 0;
	pf->fd_udp4_filter_cnt = 0;
3174
	pf->fd_sctp4_filter_cnt = 0;
3175
	pf->fd_ip4_filter_cnt = 0;
3176

3177 3178 3179 3180 3181 3182
	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
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 3209 3210 3211 3212 3213 3214 3215 3216 3217 3218
/**
 * 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;
3219
	for (i = 0; i < vsi->num_q_vectors; i++, vector++) {
3220 3221
		struct i40e_q_vector *q_vector = vsi->q_vectors[i];

3222
		q_vector->itr_countdown = ITR_COUNTDOWN_START;
3223
		q_vector->rx.itr = ITR_TO_REG(vsi->rx_rings[i]->rx_itr_setting);
J
Jesse Brandeburg 已提交
3224 3225 3226
		q_vector->rx.latency_range = I40E_LOW_LATENCY;
		wr32(hw, I40E_PFINT_ITRN(I40E_RX_ITR, vector - 1),
		     q_vector->rx.itr);
3227
		q_vector->tx.itr = ITR_TO_REG(vsi->tx_rings[i]->tx_itr_setting);
J
Jesse Brandeburg 已提交
3228 3229 3230
		q_vector->tx.latency_range = I40E_LOW_LATENCY;
		wr32(hw, I40E_PFINT_ITRN(I40E_TX_ITR, vector - 1),
		     q_vector->tx.itr);
3231
		wr32(hw, I40E_PFINT_RATEN(vector - 1),
3232
		     i40e_intrl_usec_to_reg(vsi->int_rate_limit));
J
Jesse Brandeburg 已提交
3233 3234 3235 3236

		/* 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++) {
3237 3238
			u32 val;

J
Jesse Brandeburg 已提交
3239 3240 3241 3242 3243 3244 3245 3246 3247 3248 3249 3250 3251 3252 3253 3254 3255 3256 3257 3258 3259 3260 3261 3262 3263 3264 3265 3266 3267 3268 3269 3270 3271
			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
 **/
3272
static void i40e_enable_misc_int_causes(struct i40e_pf *pf)
J
Jesse Brandeburg 已提交
3273
{
3274
	struct i40e_hw *hw = &pf->hw;
J
Jesse Brandeburg 已提交
3275 3276 3277 3278 3279 3280 3281 3282 3283 3284 3285 3286 3287 3288 3289
	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;

3290 3291 3292
	if (pf->flags & I40E_FLAG_IWARP_ENABLED)
		val |= I40E_PFINT_ICR0_ENA_PE_CRITERR_MASK;

3293 3294 3295
	if (pf->flags & I40E_FLAG_PTP)
		val |= I40E_PFINT_ICR0_ENA_TIMESYNC_MASK;

J
Jesse Brandeburg 已提交
3296 3297 3298
	wr32(hw, I40E_PFINT_ICR0_ENA, val);

	/* SW_ITR_IDX = 0, but don't change INTENA */
3299 3300
	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
3301 3302 3303 3304 3305 3306 3307 3308 3309 3310 3311

	/* 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)
{
3312
	struct i40e_q_vector *q_vector = vsi->q_vectors[0];
J
Jesse Brandeburg 已提交
3313 3314 3315 3316 3317
	struct i40e_pf *pf = vsi->back;
	struct i40e_hw *hw = &pf->hw;
	u32 val;

	/* set the ITR configuration */
3318
	q_vector->itr_countdown = ITR_COUNTDOWN_START;
3319
	q_vector->rx.itr = ITR_TO_REG(vsi->rx_rings[0]->rx_itr_setting);
J
Jesse Brandeburg 已提交
3320 3321
	q_vector->rx.latency_range = I40E_LOW_LATENCY;
	wr32(hw, I40E_PFINT_ITR0(I40E_RX_ITR), q_vector->rx.itr);
3322
	q_vector->tx.itr = ITR_TO_REG(vsi->tx_rings[0]->tx_itr_setting);
J
Jesse Brandeburg 已提交
3323 3324 3325
	q_vector->tx.latency_range = I40E_LOW_LATENCY;
	wr32(hw, I40E_PFINT_ITR0(I40E_TX_ITR), q_vector->tx.itr);

3326
	i40e_enable_misc_int_causes(pf);
J
Jesse Brandeburg 已提交
3327 3328 3329 3330

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

3331
	/* Associate the queue pair to the vector and enable the queue int */
J
Jesse Brandeburg 已提交
3332 3333 3334 3335 3336 3337 3338 3339 3340 3341 3342 3343 3344 3345
	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);
}

3346 3347 3348 3349 3350 3351 3352 3353 3354 3355 3356 3357 3358
/**
 * 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
3359 3360 3361
/**
 * i40e_irq_dynamic_enable_icr0 - Enable default interrupt generation for icr0
 * @pf: board private structure
3362
 * @clearpba: true when all pending interrupt events should be cleared
J
Jesse Brandeburg 已提交
3363
 **/
3364
void i40e_irq_dynamic_enable_icr0(struct i40e_pf *pf, bool clearpba)
J
Jesse Brandeburg 已提交
3365 3366 3367 3368 3369
{
	struct i40e_hw *hw = &pf->hw;
	u32 val;

	val = I40E_PFINT_DYN_CTL0_INTENA_MASK   |
3370
	      (clearpba ? I40E_PFINT_DYN_CTL0_CLEARPBA_MASK : 0) |
J
Jesse Brandeburg 已提交
3371 3372 3373 3374 3375 3376 3377 3378 3379 3380 3381 3382 3383 3384 3385
	      (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;

3386
	if (!q_vector->tx.ring && !q_vector->rx.ring)
J
Jesse Brandeburg 已提交
3387 3388
		return IRQ_HANDLED;

3389
	napi_schedule_irqoff(&q_vector->napi);
J
Jesse Brandeburg 已提交
3390 3391 3392 3393

	return IRQ_HANDLED;
}

3394 3395 3396 3397 3398 3399 3400 3401 3402 3403 3404 3405 3406 3407 3408 3409 3410 3411 3412 3413 3414 3415 3416 3417 3418 3419 3420
/**
 * 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
3421 3422 3423 3424 3425 3426 3427 3428 3429 3430 3431 3432 3433 3434 3435
/**
 * 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;
3436
	int irq_num;
J
Jesse Brandeburg 已提交
3437 3438

	for (vector = 0; vector < q_vectors; vector++) {
3439
		struct i40e_q_vector *q_vector = vsi->q_vectors[vector];
J
Jesse Brandeburg 已提交
3440

3441 3442
		irq_num = pf->msix_entries[base + vector].vector;

3443
		if (q_vector->tx.ring && q_vector->rx.ring) {
J
Jesse Brandeburg 已提交
3444 3445 3446
			snprintf(q_vector->name, sizeof(q_vector->name) - 1,
				 "%s-%s-%d", basename, "TxRx", rx_int_idx++);
			tx_int_idx++;
3447
		} else if (q_vector->rx.ring) {
J
Jesse Brandeburg 已提交
3448 3449
			snprintf(q_vector->name, sizeof(q_vector->name) - 1,
				 "%s-%s-%d", basename, "rx", rx_int_idx++);
3450
		} else if (q_vector->tx.ring) {
J
Jesse Brandeburg 已提交
3451 3452 3453 3454 3455 3456
			snprintf(q_vector->name, sizeof(q_vector->name) - 1,
				 "%s-%s-%d", basename, "tx", tx_int_idx++);
		} else {
			/* skip this unused q_vector */
			continue;
		}
3457
		err = request_irq(irq_num,
J
Jesse Brandeburg 已提交
3458 3459 3460 3461 3462 3463
				  vsi->irq_handler,
				  0,
				  q_vector->name,
				  q_vector);
		if (err) {
			dev_info(&pf->pdev->dev,
S
Shannon Nelson 已提交
3464
				 "MSIX request_irq failed, error: %d\n", err);
J
Jesse Brandeburg 已提交
3465 3466
			goto free_queue_irqs;
		}
3467 3468 3469 3470 3471

		/* 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
3472
		/* assign the mask for this irq */
3473
		irq_set_affinity_hint(irq_num, &q_vector->affinity_mask);
J
Jesse Brandeburg 已提交
3474 3475
	}

3476
	vsi->irqs_ready = true;
J
Jesse Brandeburg 已提交
3477 3478 3479 3480 3481
	return 0;

free_queue_irqs:
	while (vector) {
		vector--;
3482 3483 3484 3485
		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
3486 3487 3488 3489 3490 3491 3492 3493 3494 3495 3496 3497 3498 3499 3500 3501
	}
	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++) {
3502 3503
		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
3504 3505 3506 3507 3508 3509 3510 3511 3512 3513 3514 3515 3516 3517 3518 3519 3520 3521 3522 3523 3524 3525 3526 3527 3528 3529 3530 3531 3532
	}

	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) {
3533
		for (i = 0; i < vsi->num_q_vectors; i++)
J
Jesse Brandeburg 已提交
3534 3535
			i40e_irq_dynamic_enable(vsi, i);
	} else {
3536
		i40e_irq_dynamic_enable_icr0(pf, true);
J
Jesse Brandeburg 已提交
3537 3538
	}

3539
	i40e_flush(&pf->hw);
J
Jesse Brandeburg 已提交
3540 3541 3542 3543 3544 3545 3546 3547 3548 3549 3550 3551 3552 3553 3554 3555 3556 3557 3558 3559 3560 3561 3562 3563 3564 3565 3566
	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;
3567
	irqreturn_t ret = IRQ_NONE;
J
Jesse Brandeburg 已提交
3568 3569 3570 3571
	u32 icr0, icr0_remaining;
	u32 val, ena_mask;

	icr0 = rd32(hw, I40E_PFINT_ICR0);
3572
	ena_mask = rd32(hw, I40E_PFINT_ICR0_ENA);
J
Jesse Brandeburg 已提交
3573

3574 3575
	/* if sharing a legacy IRQ, we might get called w/o an intr pending */
	if ((icr0 & I40E_PFINT_ICR0_INTEVENT_MASK) == 0)
3576
		goto enable_intr;
J
Jesse Brandeburg 已提交
3577

S
Shannon Nelson 已提交
3578 3579 3580 3581 3582
	/* 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++;

3583 3584 3585 3586
	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;
3587
		dev_dbg(&pf->pdev->dev, "cleared PE_CRITERR\n");
3588 3589
	}

J
Jesse Brandeburg 已提交
3590 3591
	/* only q0 is used in MSI/Legacy mode, and none are used in MSIX */
	if (icr0 & I40E_PFINT_ICR0_QUEUE_0_MASK) {
3592 3593
		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
3594

3595 3596 3597 3598 3599 3600
		/* 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
3601
		if (!test_bit(__I40E_DOWN, &pf->state))
3602
			napi_schedule_irqoff(&q_vector->napi);
J
Jesse Brandeburg 已提交
3603 3604 3605 3606 3607
	}

	if (icr0 & I40E_PFINT_ICR0_ADMINQ_MASK) {
		ena_mask &= ~I40E_PFINT_ICR0_ENA_ADMINQ_MASK;
		set_bit(__I40E_ADMINQ_EVENT_PENDING, &pf->state);
3608
		i40e_debug(&pf->hw, I40E_DEBUG_NVM, "AdminQ event\n");
J
Jesse Brandeburg 已提交
3609 3610 3611 3612 3613 3614 3615 3616 3617 3618 3619 3620 3621 3622 3623 3624 3625 3626 3627
	}

	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;
3628
		if (val == I40E_RESET_CORER) {
J
Jesse Brandeburg 已提交
3629
			pf->corer_count++;
3630
		} else if (val == I40E_RESET_GLOBR) {
J
Jesse Brandeburg 已提交
3631
			pf->globr_count++;
3632
		} else if (val == I40E_RESET_EMPR) {
J
Jesse Brandeburg 已提交
3633
			pf->empr_count++;
3634
			set_bit(__I40E_EMP_RESET_INTR_RECEIVED, &pf->state);
3635
		}
J
Jesse Brandeburg 已提交
3636 3637
	}

3638 3639 3640
	if (icr0 & I40E_PFINT_ICR0_HMC_ERR_MASK) {
		icr0 &= ~I40E_PFINT_ICR0_HMC_ERR_MASK;
		dev_info(&pf->pdev->dev, "HMC error interrupt\n");
3641 3642 3643
		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));
3644 3645
	}

J
Jacob Keller 已提交
3646 3647 3648 3649
	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) {
3650
			icr0 &= ~I40E_PFINT_ICR0_ENA_TIMESYNC_MASK;
J
Jacob Keller 已提交
3651 3652 3653 3654
			i40e_ptp_tx_hwtstamp(pf);
		}
	}

J
Jesse Brandeburg 已提交
3655 3656 3657 3658 3659 3660 3661 3662
	/* 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);
3663
		if ((icr0_remaining & I40E_PFINT_ICR0_PE_CRITERR_MASK) ||
J
Jesse Brandeburg 已提交
3664
		    (icr0_remaining & I40E_PFINT_ICR0_PCI_EXCEPTION_MASK) ||
3665
		    (icr0_remaining & I40E_PFINT_ICR0_ECC_ERR_MASK)) {
3666 3667 3668
			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
3669 3670 3671
		}
		ena_mask &= ~icr0_remaining;
	}
3672
	ret = IRQ_HANDLED;
J
Jesse Brandeburg 已提交
3673

3674
enable_intr:
J
Jesse Brandeburg 已提交
3675 3676 3677 3678
	/* re-enable interrupt causes */
	wr32(hw, I40E_PFINT_ICR0_ENA, ena_mask);
	if (!test_bit(__I40E_DOWN, &pf->state)) {
		i40e_service_event_schedule(pf);
3679
		i40e_irq_dynamic_enable_icr0(pf, false);
J
Jesse Brandeburg 已提交
3680 3681
	}

3682
	return ret;
J
Jesse Brandeburg 已提交
3683 3684
}

3685 3686 3687 3688 3689 3690 3691 3692 3693 3694 3695 3696 3697 3698 3699 3700 3701 3702 3703 3704 3705 3706 3707 3708 3709 3710 3711 3712 3713 3714 3715 3716 3717 3718 3719 3720
/**
 * 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;

3721 3722 3723 3724 3725 3726 3727 3728 3729 3730 3731
		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);
		}
3732 3733 3734 3735 3736
		/* 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);
3737 3738
		if (tx_buf->tx_flags & I40E_TX_FLAGS_FD_SB)
			kfree(tx_buf->raw_buf);
3739

3740 3741 3742
		tx_buf->raw_buf = NULL;
		tx_buf->tx_flags = 0;
		tx_buf->next_to_watch = NULL;
3743
		dma_unmap_len_set(tx_buf, len, 0);
3744 3745
		tx_desc->buffer_addr = 0;
		tx_desc->cmd_type_offset_bsz = 0;
3746

3747
		/* move us past the eop_desc for start of next FD desc */
3748 3749 3750 3751 3752 3753 3754 3755 3756 3757 3758 3759 3760 3761 3762 3763
		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
3764
	if (vsi->back->flags & I40E_FLAG_MSIX_ENABLED)
3765
		i40e_irq_dynamic_enable(vsi, tx_ring->q_vector->v_idx);
J
Jesse Brandeburg 已提交
3766

3767 3768 3769 3770 3771 3772 3773 3774 3775 3776 3777 3778 3779 3780 3781 3782 3783 3784 3785 3786 3787 3788
	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
3789
/**
3790
 * i40e_map_vector_to_qp - Assigns the queue pair to the vector
J
Jesse Brandeburg 已提交
3791 3792
 * @vsi: the VSI being configured
 * @v_idx: vector index
3793
 * @qp_idx: queue pair index
J
Jesse Brandeburg 已提交
3794
 **/
3795
static void i40e_map_vector_to_qp(struct i40e_vsi *vsi, int v_idx, int qp_idx)
J
Jesse Brandeburg 已提交
3796
{
3797
	struct i40e_q_vector *q_vector = vsi->q_vectors[v_idx];
3798 3799
	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
3800 3801

	tx_ring->q_vector = q_vector;
3802 3803
	tx_ring->next = q_vector->tx.ring;
	q_vector->tx.ring = tx_ring;
J
Jesse Brandeburg 已提交
3804
	q_vector->tx.count++;
3805 3806 3807 3808 3809

	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
3810 3811 3812 3813 3814 3815 3816 3817 3818 3819 3820 3821 3822 3823 3824
}

/**
 * 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;
3825
	int num_ringpairs;
J
Jesse Brandeburg 已提交
3826 3827 3828 3829 3830
	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.
3831 3832 3833 3834
	 * 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
3835
	 */
3836
	for (; v_start < q_vectors; v_start++) {
3837 3838 3839 3840 3841 3842 3843 3844 3845 3846 3847 3848
		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--) {
3849
			i40e_map_vector_to_qp(vsi, v_start, qp_idx);
3850 3851
			qp_idx++;
			qp_remaining--;
J
Jesse Brandeburg 已提交
3852 3853 3854 3855 3856 3857 3858 3859 3860 3861 3862 3863 3864 3865 3866 3867 3868 3869
		}
	}
}

/**
 * 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,
3870
				  pf->int_name, pf);
J
Jesse Brandeburg 已提交
3871 3872
	else
		err = request_irq(pf->pdev->irq, i40e_intr, IRQF_SHARED,
3873
				  pf->int_name, pf);
J
Jesse Brandeburg 已提交
3874 3875 3876 3877 3878 3879 3880 3881 3882

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

	return err;
}

#ifdef CONFIG_NET_POLL_CONTROLLER
/**
3883
 * i40e_netpoll - A Polling 'interrupt' handler
J
Jesse Brandeburg 已提交
3884 3885 3886 3887 3888 3889 3890 3891 3892 3893 3894 3895 3896 3897 3898 3899 3900 3901
 * @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.
 **/
static void i40e_netpoll(struct net_device *netdev)
{
	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++)
3902
			i40e_msix_clean_rings(0, vsi->q_vectors[i]);
J
Jesse Brandeburg 已提交
3903 3904 3905 3906 3907 3908
	} else {
		i40e_intr(pf->pdev->irq, netdev);
	}
}
#endif

3909 3910 3911 3912 3913 3914 3915 3916 3917 3918 3919 3920 3921 3922 3923 3924 3925 3926 3927 3928 3929
/**
 * 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;

3930
		usleep_range(10, 20);
3931 3932 3933 3934 3935 3936 3937
	}
	if (i >= I40E_QUEUE_WAIT_RETRY_LIMIT)
		return -ETIMEDOUT;

	return 0;
}

J
Jesse Brandeburg 已提交
3938 3939 3940 3941 3942 3943 3944 3945 3946
/**
 * 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;
3947
	int i, j, pf_q, ret = 0;
J
Jesse Brandeburg 已提交
3948 3949 3950 3951
	u32 tx_reg;

	pf_q = vsi->base_queue;
	for (i = 0; i < vsi->num_queue_pairs; i++, pf_q++) {
3952 3953 3954 3955

		/* warn the TX unit of coming changes */
		i40e_pre_tx_queue_cfg(&pf->hw, pf_q, enable);
		if (!enable)
3956
			usleep_range(10, 20);
3957

3958
		for (j = 0; j < 50; j++) {
J
Jesse Brandeburg 已提交
3959
			tx_reg = rd32(hw, I40E_QTX_ENA(pf_q));
3960 3961 3962 3963 3964
			if (((tx_reg >> I40E_QTX_ENA_QENA_REQ_SHIFT) & 1) ==
			    ((tx_reg >> I40E_QTX_ENA_QENA_STAT_SHIFT) & 1))
				break;
			usleep_range(1000, 2000);
		}
3965
		/* Skip if the queue is already in the requested state */
3966
		if (enable == !!(tx_reg & I40E_QTX_ENA_QENA_STAT_MASK))
3967
			continue;
J
Jesse Brandeburg 已提交
3968 3969

		/* turn on/off the queue */
3970 3971
		if (enable) {
			wr32(hw, I40E_QTX_HEAD(pf_q), 0);
3972
			tx_reg |= I40E_QTX_ENA_QENA_REQ_MASK;
3973
		} else {
J
Jesse Brandeburg 已提交
3974
			tx_reg &= ~I40E_QTX_ENA_QENA_REQ_MASK;
3975
		}
J
Jesse Brandeburg 已提交
3976 3977

		wr32(hw, I40E_QTX_ENA(pf_q), tx_reg);
3978 3979 3980
		/* 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
3981 3982

		/* wait for the change to finish */
3983 3984 3985
		ret = i40e_pf_txq_wait(pf, pf_q, enable);
		if (ret) {
			dev_info(&pf->pdev->dev,
S
Shannon Nelson 已提交
3986 3987
				 "VSI seid %d Tx ring %d %sable timeout\n",
				 vsi->seid, pf_q, (enable ? "en" : "dis"));
3988
			break;
J
Jesse Brandeburg 已提交
3989 3990 3991
		}
	}

3992 3993 3994 3995 3996 3997 3998 3999 4000 4001 4002 4003 4004 4005 4006 4007 4008 4009 4010 4011 4012 4013 4014 4015
	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;

4016
		usleep_range(10, 20);
4017 4018 4019
	}
	if (i >= I40E_QUEUE_WAIT_RETRY_LIMIT)
		return -ETIMEDOUT;
4020

J
Jesse Brandeburg 已提交
4021 4022 4023 4024 4025 4026 4027 4028 4029 4030 4031 4032
	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;
4033
	int i, j, pf_q, ret = 0;
J
Jesse Brandeburg 已提交
4034 4035 4036 4037
	u32 rx_reg;

	pf_q = vsi->base_queue;
	for (i = 0; i < vsi->num_queue_pairs; i++, pf_q++) {
4038
		for (j = 0; j < 50; j++) {
J
Jesse Brandeburg 已提交
4039
			rx_reg = rd32(hw, I40E_QRX_ENA(pf_q));
4040 4041 4042 4043 4044
			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
4045

4046 4047 4048
		/* 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
4049 4050 4051

		/* turn on/off the queue */
		if (enable)
4052
			rx_reg |= I40E_QRX_ENA_QENA_REQ_MASK;
J
Jesse Brandeburg 已提交
4053
		else
4054
			rx_reg &= ~I40E_QRX_ENA_QENA_REQ_MASK;
J
Jesse Brandeburg 已提交
4055
		wr32(hw, I40E_QRX_ENA(pf_q), rx_reg);
4056 4057 4058
		/* 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
4059 4060

		/* wait for the change to finish */
4061 4062 4063
		ret = i40e_pf_rxq_wait(pf, pf_q, enable);
		if (ret) {
			dev_info(&pf->pdev->dev,
S
Shannon Nelson 已提交
4064 4065
				 "VSI seid %d Rx ring %d %sable timeout\n",
				 vsi->seid, pf_q, (enable ? "en" : "dis"));
4066
			break;
J
Jesse Brandeburg 已提交
4067 4068 4069
		}
	}

4070 4071 4072 4073 4074 4075
	/* Due to HW errata, on Rx disable only, the register can indicate done
	 * before it really is. Needs 50ms to be sure
	 */
	if (!enable)
		mdelay(50);

4076
	return ret;
J
Jesse Brandeburg 已提交
4077 4078 4079
}

/**
F
Filip Sadowski 已提交
4080
 * i40e_vsi_start_rings - Start a VSI's rings
J
Jesse Brandeburg 已提交
4081 4082
 * @vsi: the VSI being configured
 **/
F
Filip Sadowski 已提交
4083
int i40e_vsi_start_rings(struct i40e_vsi *vsi)
J
Jesse Brandeburg 已提交
4084
{
4085
	int ret = 0;
J
Jesse Brandeburg 已提交
4086 4087

	/* do rx first for enable and last for disable */
F
Filip Sadowski 已提交
4088 4089 4090 4091
	ret = i40e_vsi_control_rx(vsi, true);
	if (ret)
		return ret;
	ret = i40e_vsi_control_tx(vsi, true);
J
Jesse Brandeburg 已提交
4092 4093 4094 4095

	return ret;
}

F
Filip Sadowski 已提交
4096 4097 4098 4099 4100 4101 4102 4103 4104 4105 4106 4107 4108
/**
 * 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
4109 4110 4111 4112 4113 4114 4115 4116 4117 4118 4119 4120 4121 4122 4123 4124
/**
 * 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;

4125 4126 4127 4128
		if (!vsi->irqs_ready)
			return;

		vsi->irqs_ready = false;
J
Jesse Brandeburg 已提交
4129
		for (i = 0; i < vsi->num_q_vectors; i++) {
4130 4131 4132 4133 4134
			int irq_num;
			u16 vector;

			vector = i + base;
			irq_num = pf->msix_entries[vector].vector;
J
Jesse Brandeburg 已提交
4135 4136

			/* free only the irqs that were actually requested */
S
Shannon Nelson 已提交
4137 4138
			if (!vsi->q_vectors[i] ||
			    !vsi->q_vectors[i]->num_ringpairs)
J
Jesse Brandeburg 已提交
4139 4140
				continue;

4141 4142
			/* clear the affinity notifier in the IRQ descriptor */
			irq_set_affinity_notifier(irq_num, NULL);
J
Jesse Brandeburg 已提交
4143
			/* clear the affinity_mask in the IRQ descriptor */
4144 4145 4146
			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
4147 4148 4149 4150 4151 4152 4153 4154 4155 4156 4157 4158 4159 4160 4161 4162 4163 4164 4165 4166 4167 4168 4169 4170 4171 4172 4173 4174 4175 4176 4177 4178 4179 4180 4181 4182 4183 4184 4185 4186 4187 4188 4189 4190 4191 4192 4193 4194 4195 4196 4197 4198 4199 4200 4201 4202 4203 4204 4205 4206 4207 4208 4209 4210 4211 4212 4213 4214 4215 4216 4217 4218 4219 4220 4221 4222 4223 4224 4225 4226 4227 4228

			/* 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);
	}
}

4229 4230 4231 4232 4233 4234 4235 4236 4237 4238 4239 4240
/**
 * 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];
4241
	struct i40e_ring *ring;
4242 4243 4244 4245 4246

	if (!q_vector)
		return;

	/* disassociate q_vector from rings */
4247 4248 4249 4250 4251
	i40e_for_each_ring(ring, q_vector->tx)
		ring->q_vector = NULL;

	i40e_for_each_ring(ring, q_vector->rx)
		ring->q_vector = NULL;
4252 4253 4254 4255 4256 4257 4258 4259 4260 4261

	/* 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
4262 4263 4264 4265 4266 4267 4268 4269 4270 4271 4272
/**
 * 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;

4273 4274
	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
4275 4276 4277 4278 4279 4280 4281 4282 4283 4284 4285 4286 4287
}

/**
 * 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;
4288 4289
		kfree(pf->irq_pile);
		pf->irq_pile = NULL;
J
Jesse Brandeburg 已提交
4290 4291 4292 4293 4294 4295 4296 4297 4298 4299 4300 4301 4302 4303 4304 4305 4306
	} 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;

4307
	i40e_stop_misc_vector(pf);
4308
	if (pf->flags & I40E_FLAG_MSIX_ENABLED && pf->msix_entries) {
4309 4310 4311 4312
		synchronize_irq(pf->msix_entries[0].vector);
		free_irq(pf->msix_entries[0].vector, pf);
	}

4313 4314 4315
	i40e_put_lump(pf->irq_pile, pf->iwarp_base_vector,
		      I40E_IWARP_IRQ_PILE_ID);

J
Jesse Brandeburg 已提交
4316
	i40e_put_lump(pf->irq_pile, 0, I40E_PILE_VALID_BIT-1);
M
Mitch Williams 已提交
4317
	for (i = 0; i < pf->num_alloc_vsi; i++)
J
Jesse Brandeburg 已提交
4318 4319 4320 4321 4322 4323 4324 4325 4326 4327 4328 4329 4330 4331 4332 4333
		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;

4334 4335 4336 4337 4338 4339
	for (q_idx = 0; q_idx < vsi->num_q_vectors; q_idx++) {
		struct i40e_q_vector *q_vector = vsi->q_vectors[q_idx];

		if (q_vector->rx.ring || q_vector->tx.ring)
			napi_enable(&q_vector->napi);
	}
J
Jesse Brandeburg 已提交
4340 4341 4342 4343 4344 4345 4346 4347 4348 4349 4350 4351 4352
}

/**
 * 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;

4353 4354 4355 4356 4357 4358
	for (q_idx = 0; q_idx < vsi->num_q_vectors; q_idx++) {
		struct i40e_q_vector *q_vector = vsi->q_vectors[q_idx];

		if (q_vector->rx.ring || q_vector->tx.ring)
			napi_disable(&q_vector->napi);
	}
J
Jesse Brandeburg 已提交
4359 4360
}

4361 4362 4363 4364 4365 4366
/**
 * i40e_vsi_close - Shut down a VSI
 * @vsi: the vsi to be quelled
 **/
static void i40e_vsi_close(struct i40e_vsi *vsi)
{
M
Mitch Williams 已提交
4367
	struct i40e_pf *pf = vsi->back;
4368 4369 4370 4371 4372
	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);
4373
	vsi->current_netdev_flags = 0;
M
Mitch Williams 已提交
4374 4375 4376
	pf->flags |= I40E_FLAG_SERVICE_CLIENT_REQUESTED;
	if (test_bit(__I40E_RESET_RECOVERY_PENDING, &pf->state))
		pf->flags |=  I40E_FLAG_CLIENT_RESET;
4377 4378
}

J
Jesse Brandeburg 已提交
4379 4380 4381 4382 4383 4384 4385 4386 4387 4388
/**
 * 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;

	set_bit(__I40E_NEEDS_RESTART, &vsi->state);
J
Jesse Brandeburg 已提交
4389
	if (vsi->netdev && netif_running(vsi->netdev))
J
Jesse Brandeburg 已提交
4390
		vsi->netdev->netdev_ops->ndo_stop(vsi->netdev);
J
Jesse Brandeburg 已提交
4391
	else
4392
		i40e_vsi_close(vsi);
J
Jesse Brandeburg 已提交
4393 4394 4395 4396 4397 4398 4399 4400 4401 4402 4403 4404 4405 4406 4407
}

/**
 * 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
4408
		i40e_vsi_open(vsi);   /* this clears the DOWN bit */
J
Jesse Brandeburg 已提交
4409 4410 4411 4412 4413 4414 4415 4416 4417 4418
}

/**
 * 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
4419
	for (v = 0; v < pf->num_alloc_vsi; v++) {
J
Jesse Brandeburg 已提交
4420 4421 4422 4423 4424 4425 4426 4427 4428 4429 4430 4431 4432
		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
4433
	for (v = 0; v < pf->num_alloc_vsi; v++) {
J
Jesse Brandeburg 已提交
4434 4435 4436 4437 4438
		if (pf->vsi[v])
			i40e_unquiesce_vsi(pf->vsi[v]);
	}
}

4439 4440
#ifdef CONFIG_I40E_DCB
/**
4441
 * i40e_vsi_wait_queues_disabled - Wait for VSI's queues to be disabled
4442 4443
 * @vsi: the VSI being configured
 *
4444
 * This function waits for the given VSI's queues to be disabled.
4445
 **/
4446
static int i40e_vsi_wait_queues_disabled(struct i40e_vsi *vsi)
4447 4448 4449 4450 4451 4452 4453 4454 4455 4456
{
	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
4457 4458
				 "VSI seid %d Tx ring %d disable timeout\n",
				 vsi->seid, pf_q);
4459 4460 4461 4462
			return ret;
		}
	}

4463 4464 4465 4466 4467 4468 4469 4470 4471 4472 4473 4474
	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;
		}
	}

4475 4476 4477 4478
	return 0;
}

/**
4479
 * i40e_pf_wait_queues_disabled - Wait for all queues of PF VSIs to be disabled
4480 4481
 * @pf: the PF
 *
4482
 * This function waits for the queues to be in disabled state for all the
4483 4484
 * VSIs that are managed by this PF.
 **/
4485
static int i40e_pf_wait_queues_disabled(struct i40e_pf *pf)
4486 4487 4488 4489
{
	int v, ret = 0;

	for (v = 0; v < pf->hw.func_caps.num_vsis; v++) {
4490
		if (pf->vsi[v]) {
4491
			ret = i40e_vsi_wait_queues_disabled(pf->vsi[v]);
4492 4493 4494 4495 4496 4497 4498 4499 4500
			if (ret)
				break;
		}
	}

	return ret;
}

#endif
4501 4502 4503 4504 4505 4506 4507 4508 4509 4510 4511 4512 4513 4514 4515 4516

/**
 * 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;
4517
	u32 head, val, tx_pending_hw;
4518 4519 4520 4521 4522 4523 4524 4525 4526 4527 4528 4529 4530 4531 4532 4533 4534 4535 4536 4537 4538 4539 4540 4541 4542 4543 4544
	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);

4545
	tx_pending_hw = i40e_get_tx_pending(tx_ring, false);
4546

4547 4548 4549 4550 4551
	/* 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.
4552
	 */
4553
	if (tx_pending_hw && (!(val & I40E_PFINT_DYN_CTLN_INTENA_MASK))) {
4554 4555 4556
		/* 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)) {
4557 4558
			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,
4559 4560 4561 4562 4563 4564 4565 4566 4567 4568 4569 4570
				    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);
		}
	}
4571 4572 4573 4574 4575 4576 4577 4578

	/* 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))) {
4579
		local_bh_disable();
4580 4581
		if (napi_reschedule(&tx_ring->q_vector->napi))
			tx_ring->tx_stats.tx_lost_interrupt++;
4582
		local_bh_enable();
4583
	}
4584 4585 4586 4587 4588 4589 4590 4591 4592 4593 4594 4595 4596 4597 4598 4599 4600 4601 4602 4603 4604 4605 4606 4607 4608 4609 4610 4611 4612 4613 4614 4615 4616 4617 4618 4619 4620 4621 4622 4623 4624 4625 4626 4627 4628 4629 4630 4631 4632
}

/**
 * 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);
	}
}

4633 4634
/**
 * i40e_get_iscsi_tc_map - Return TC map for iSCSI APP
4635
 * @pf: pointer to PF
4636 4637 4638 4639 4640 4641 4642 4643 4644 4645 4646 4647 4648 4649 4650 4651 4652 4653
 *
 * 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
4654
			enabled_tc |= BIT(tc);
4655 4656 4657 4658 4659 4660 4661
			break;
		}
	}

	return enabled_tc;
}

J
Jesse Brandeburg 已提交
4662 4663 4664 4665 4666 4667 4668 4669
/**
 * 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)
{
4670
	int i, tc_unused = 0;
J
Jesse Brandeburg 已提交
4671
	u8 num_tc = 0;
4672
	u8 ret = 0;
J
Jesse Brandeburg 已提交
4673 4674 4675

	/* Scan the ETS Config Priority Table to find
	 * traffic class enabled for a given priority
4676
	 * and create a bitmask of enabled TCs
J
Jesse Brandeburg 已提交
4677
	 */
4678 4679
	for (i = 0; i < I40E_MAX_USER_PRIORITY; i++)
		num_tc |= BIT(dcbcfg->etscfg.prioritytable[i]);
J
Jesse Brandeburg 已提交
4680

4681 4682
	/* Now scan the bitmask to check for
	 * contiguous TCs starting with TC0
J
Jesse Brandeburg 已提交
4683
	 */
4684 4685 4686 4687 4688 4689 4690 4691 4692 4693 4694 4695 4696 4697 4698 4699 4700 4701
	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
4702 4703 4704 4705 4706 4707 4708 4709 4710 4711 4712 4713 4714 4715 4716 4717
}

/**
 * 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++)
4718
		enabled_tc |= BIT(i);
J
Jesse Brandeburg 已提交
4719 4720 4721 4722 4723 4724 4725 4726 4727 4728 4729 4730 4731

	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;
4732
	u8 i, enabled_tc = 1;
J
Jesse Brandeburg 已提交
4733 4734 4735 4736 4737 4738 4739
	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;

4740 4741 4742 4743
	/* 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
4744
	/* MFP mode return count of enabled TCs for this PF */
4745 4746 4747
	if (pf->hw.func_caps.iscsi)
		enabled_tc =  i40e_get_iscsi_tc_map(pf);
	else
4748
		return 1; /* Only TC0 */
J
Jesse Brandeburg 已提交
4749

4750
	for (i = 0; i < I40E_MAX_TRAFFIC_CLASS; i++) {
J
Jesse Brandeburg 已提交
4751
		if (enabled_tc & BIT(i))
4752 4753 4754
			num_tc++;
	}
	return num_tc;
J
Jesse Brandeburg 已提交
4755 4756 4757 4758 4759 4760 4761 4762 4763 4764 4765 4766
}

/**
 * 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))
4767
		return I40E_DEFAULT_TRAFFIC_CLASS;
J
Jesse Brandeburg 已提交
4768 4769

	/* SFP mode we want PF to be enabled for all TCs */
4770 4771 4772
	if (!(pf->flags & I40E_FLAG_MFP_ENABLED))
		return i40e_dcb_get_enabled_tc(&pf->hw.local_dcbx_config);

4773
	/* MFP enabled and iSCSI PF type */
4774 4775 4776
	if (pf->hw.func_caps.iscsi)
		return i40e_get_iscsi_tc_map(pf);
	else
4777
		return I40E_DEFAULT_TRAFFIC_CLASS;
J
Jesse Brandeburg 已提交
4778 4779 4780 4781 4782 4783 4784 4785 4786 4787 4788 4789 4790 4791
}

/**
 * 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;
4792
	i40e_status ret;
J
Jesse Brandeburg 已提交
4793 4794 4795 4796
	u32 tc_bw_max;
	int i;

	/* Get the VSI level BW configuration */
4797 4798
	ret = i40e_aq_query_vsi_bw_config(hw, vsi->seid, &bw_config, NULL);
	if (ret) {
J
Jesse Brandeburg 已提交
4799
		dev_info(&pf->pdev->dev,
4800 4801 4802
			 "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
4803
		return -EINVAL;
J
Jesse Brandeburg 已提交
4804 4805 4806
	}

	/* Get the VSI level BW configuration per TC */
4807 4808 4809
	ret = i40e_aq_query_vsi_ets_sla_config(hw, vsi->seid, &bw_ets_config,
					       NULL);
	if (ret) {
J
Jesse Brandeburg 已提交
4810
		dev_info(&pf->pdev->dev,
4811 4812 4813
			 "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
4814
		return -EINVAL;
J
Jesse Brandeburg 已提交
4815 4816 4817 4818 4819 4820 4821 4822 4823 4824 4825 4826 4827 4828 4829 4830 4831 4832 4833 4834 4835
	}

	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
4836

J
Jesse Brandeburg 已提交
4837
	return 0;
J
Jesse Brandeburg 已提交
4838 4839 4840 4841 4842 4843 4844 4845 4846 4847
}

/**
 * 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
4848
static int i40e_vsi_configure_bw_alloc(struct i40e_vsi *vsi, u8 enabled_tc,
J
Jesse Brandeburg 已提交
4849 4850 4851
				       u8 *bw_share)
{
	struct i40e_aqc_configure_vsi_tc_bw_data bw_data;
4852
	i40e_status ret;
J
Jesse Brandeburg 已提交
4853
	int i;
J
Jesse Brandeburg 已提交
4854 4855 4856 4857 4858

	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];

4859 4860 4861
	ret = i40e_aq_config_vsi_tc_bw(&vsi->back->hw, vsi->seid, &bw_data,
				       NULL);
	if (ret) {
J
Jesse Brandeburg 已提交
4862
		dev_info(&vsi->back->pdev->dev,
J
Jesse Brandeburg 已提交
4863 4864
			 "AQ command Config VSI BW allocation per TC failed = %d\n",
			 vsi->back->hw.aq.asq_last_status);
J
Jesse Brandeburg 已提交
4865
		return -EINVAL;
J
Jesse Brandeburg 已提交
4866 4867 4868 4869 4870
	}

	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
4871
	return 0;
J
Jesse Brandeburg 已提交
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 4909
}

/**
 * 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
4910
		if (vsi->tc_config.enabled_tc & BIT(i))
J
Jesse Brandeburg 已提交
4911 4912 4913 4914 4915 4916 4917 4918 4919 4920 4921 4922 4923 4924 4925 4926 4927 4928 4929 4930 4931 4932 4933 4934 4935 4936 4937 4938 4939 4940 4941 4942 4943 4944 4945 4946 4947 4948 4949 4950 4951 4952 4953 4954 4955 4956 4957 4958 4959 4960 4961 4962 4963 4964 4965 4966 4967 4968 4969 4970 4971
			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
4972
		if (enabled_tc & BIT(i))
J
Jesse Brandeburg 已提交
4973 4974 4975 4976 4977 4978 4979 4980 4981 4982 4983 4984 4985 4986 4987 4988
			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;
4989
	ctxt.info = vsi->info;
J
Jesse Brandeburg 已提交
4990 4991
	i40e_vsi_setup_queue_map(vsi, &ctxt, enabled_tc, false);

4992 4993 4994 4995 4996 4997
	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
4998 4999 5000 5001
	/* 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,
5002 5003 5004 5005
			 "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
5006 5007 5008 5009 5010 5011 5012 5013 5014 5015
		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,
5016 5017 5018 5019
			 "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
5020 5021 5022 5023 5024 5025 5026 5027 5028
		goto out;
	}

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

N
Neerav Parikh 已提交
5029 5030 5031 5032 5033 5034 5035 5036 5037 5038 5039 5040 5041 5042 5043 5044 5045 5046 5047 5048 5049 5050 5051
/**
 * 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
5052
		if (enabled_tc & BIT(i))
N
Neerav Parikh 已提交
5053 5054 5055 5056 5057 5058 5059
			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,
5060 5061 5062
			 "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
5063 5064 5065 5066 5067 5068 5069
		goto out;
	}

	/* Update the BW information */
	ret = i40e_veb_get_bw_info(veb);
	if (ret) {
		dev_info(&pf->pdev->dev,
5070 5071 5072
			 "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
5073 5074 5075 5076 5077 5078 5079 5080 5081 5082 5083 5084 5085 5086 5087 5088 5089 5090 5091 5092 5093 5094 5095 5096 5097 5098 5099 5100 5101 5102 5103 5104 5105 5106 5107 5108
	}

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
5109
	for (v = 0; v < pf->num_alloc_vsi; v++) {
N
Neerav Parikh 已提交
5110 5111 5112 5113 5114 5115 5116 5117 5118
		if (!pf->vsi[v])
			continue;

		/* - Enable all TCs for the LAN VSI
		 * - For all others keep them at TC0 for now
		 */
		if (v == pf->lan_vsi)
			tc_map = i40e_pf_get_tc_map(pf);
		else
5119
			tc_map = I40E_DEFAULT_TRAFFIC_CLASS;
N
Neerav Parikh 已提交
5120 5121 5122 5123 5124 5125 5126 5127

		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 {
5128 5129
			/* Re-configure VSI vectors based on updated TC map */
			i40e_vsi_map_rings_to_vectors(pf->vsi[v]);
N
Neerav Parikh 已提交
5130 5131 5132 5133 5134 5135
			if (pf->vsi[v]->netdev)
				i40e_dcbnl_set_all(pf->vsi[v]);
		}
	}
}

5136 5137 5138 5139 5140 5141 5142 5143 5144 5145 5146 5147 5148 5149 5150
/**
 * 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,
5151 5152 5153
			 "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));
5154 5155 5156 5157 5158 5159 5160 5161
		/* 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
5162 5163 5164 5165 5166 5167 5168 5169 5170 5171 5172 5173
/**
 * 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;

5174
	/* Do not enable DCB for SW1 and SW2 images even if the FW is capable */
5175
	if (pf->flags & I40E_FLAG_NO_DCB_SUPPORT)
5176 5177
		goto out;

N
Neerav Parikh 已提交
5178 5179 5180 5181 5182 5183 5184 5185 5186 5187 5188 5189
	/* 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");
		} else {
			/* When status is not DISABLED then DCBX in FW */
			pf->dcbx_cap = DCB_CAP_DCBX_LLD_MANAGED |
				       DCB_CAP_DCBX_VER_IEEE;
5190 5191

			pf->flags |= I40E_FLAG_DCB_CAPABLE;
5192 5193 5194
			/* Enable DCB tagging only when more than one TC
			 * or explicitly disable if only one TC
			 */
5195 5196
			if (i40e_dcb_get_num_tc(&hw->local_dcbx_config) > 1)
				pf->flags |= I40E_FLAG_DCB_ENABLED;
5197 5198
			else
				pf->flags &= ~I40E_FLAG_DCB_ENABLED;
5199 5200
			dev_dbg(&pf->pdev->dev,
				"DCBX offload is supported for this PF.\n");
N
Neerav Parikh 已提交
5201
		}
5202
	} else {
5203
		dev_info(&pf->pdev->dev,
5204 5205 5206
			 "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
5207 5208 5209 5210 5211 5212
	}

out:
	return err;
}
#endif /* CONFIG_I40E_DCB */
J
Jesse Brandeburg 已提交
5213 5214 5215 5216 5217 5218
#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
5219
void i40e_print_link_message(struct i40e_vsi *vsi, bool isup)
J
Jesse Brandeburg 已提交
5220
{
5221
	enum i40e_aq_link_speed new_speed;
5222 5223
	char *speed = "Unknown";
	char *fc = "Unknown";
5224 5225
	char *fec = "";
	char *an = "";
J
Jesse Brandeburg 已提交
5226

5227 5228 5229
	new_speed = vsi->back->hw.phy.link_info.link_speed;

	if ((vsi->current_isup == isup) && (vsi->current_speed == new_speed))
M
Matt Jared 已提交
5230 5231
		return;
	vsi->current_isup = isup;
5232
	vsi->current_speed = new_speed;
J
Jesse Brandeburg 已提交
5233 5234 5235 5236 5237
	if (!isup) {
		netdev_info(vsi->netdev, "NIC Link is Down\n");
		return;
	}

5238 5239 5240 5241 5242 5243 5244 5245 5246
	/* 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
5247 5248
	switch (vsi->back->hw.phy.link_info.link_speed) {
	case I40E_LINK_SPEED_40GB:
5249
		speed = "40 G";
J
Jesse Brandeburg 已提交
5250
		break;
5251
	case I40E_LINK_SPEED_20GB:
5252
		speed = "20 G";
5253
		break;
5254 5255 5256
	case I40E_LINK_SPEED_25GB:
		speed = "25 G";
		break;
J
Jesse Brandeburg 已提交
5257
	case I40E_LINK_SPEED_10GB:
5258
		speed = "10 G";
J
Jesse Brandeburg 已提交
5259 5260
		break;
	case I40E_LINK_SPEED_1GB:
5261
		speed = "1000 M";
J
Jesse Brandeburg 已提交
5262
		break;
M
Mitch Williams 已提交
5263
	case I40E_LINK_SPEED_100MB:
5264
		speed = "100 M";
M
Mitch Williams 已提交
5265
		break;
J
Jesse Brandeburg 已提交
5266 5267 5268 5269 5270 5271
	default:
		break;
	}

	switch (vsi->back->hw.fc.current_mode) {
	case I40E_FC_FULL:
5272
		fc = "RX/TX";
J
Jesse Brandeburg 已提交
5273 5274
		break;
	case I40E_FC_TX_PAUSE:
5275
		fc = "TX";
J
Jesse Brandeburg 已提交
5276 5277
		break;
	case I40E_FC_RX_PAUSE:
5278
		fc = "RX";
J
Jesse Brandeburg 已提交
5279 5280
		break;
	default:
5281
		fc = "None";
J
Jesse Brandeburg 已提交
5282 5283 5284
		break;
	}

5285 5286 5287 5288 5289 5290 5291 5292 5293 5294 5295 5296 5297 5298 5299 5300 5301
	if (vsi->back->hw.phy.link_info.link_speed == I40E_LINK_SPEED_25GB) {
		fec = ", FEC: None";
		an = ", Autoneg: False";

		if (vsi->back->hw.phy.link_info.an_info & I40E_AQ_AN_COMPLETED)
			an = ", Autoneg: True";

		if (vsi->back->hw.phy.link_info.fec_info &
		    I40E_AQ_CONFIG_FEC_KR_ENA)
			fec = ", FEC: CL74 FC-FEC/BASE-R";
		else if (vsi->back->hw.phy.link_info.fec_info &
			 I40E_AQ_CONFIG_FEC_RS_ENA)
			fec = ", FEC: CL108 RS-FEC";
	}

	netdev_info(vsi->netdev, "NIC Link is Up, %sbps Full Duplex%s%s, Flow Control: %s\n",
		    speed, fec, an, fc);
J
Jesse Brandeburg 已提交
5302
}
N
Neerav Parikh 已提交
5303

J
Jesse Brandeburg 已提交
5304 5305 5306 5307 5308 5309 5310 5311 5312 5313 5314 5315 5316 5317 5318
/**
 * 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
5319
	err = i40e_vsi_start_rings(vsi);
J
Jesse Brandeburg 已提交
5320 5321 5322 5323 5324 5325 5326 5327 5328
	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
5329
		i40e_print_link_message(vsi, true);
J
Jesse Brandeburg 已提交
5330 5331
		netif_tx_start_all_queues(vsi->netdev);
		netif_carrier_on(vsi->netdev);
A
Anjali Singhai 已提交
5332
	} else if (vsi->netdev) {
J
Jesse Brandeburg 已提交
5333
		i40e_print_link_message(vsi, false);
5334 5335 5336 5337 5338 5339 5340
		/* 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
5341
	}
5342 5343

	/* replay FDIR SB filters */
5344 5345
	if (vsi->type == I40E_VSI_FDIR) {
		/* reset fd counters */
5346 5347
		pf->fd_add_err = 0;
		pf->fd_atr_cnt = 0;
5348
		i40e_fdir_filter_restore(vsi);
5349
	}
5350 5351 5352 5353 5354

	/* 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
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 5387 5388 5389 5390 5391 5392 5393 5394 5395 5396 5397 5398 5399 5400 5401 5402 5403 5404 5405 5406 5407 5408 5409 5410
	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
5411
	i40e_vsi_stop_rings(vsi);
J
Jesse Brandeburg 已提交
5412 5413 5414
	i40e_napi_disable_all(vsi);

	for (i = 0; i < vsi->num_queue_pairs; i++) {
5415 5416
		i40e_clean_tx_ring(vsi->tx_rings[i]);
		i40e_clean_rx_ring(vsi->rx_rings[i]);
J
Jesse Brandeburg 已提交
5417
	}
5418

J
Jesse Brandeburg 已提交
5419 5420 5421 5422 5423 5424 5425 5426 5427 5428 5429 5430 5431 5432 5433 5434 5435 5436 5437 5438 5439 5440 5441 5442 5443 5444 5445 5446 5447 5448 5449 5450 5451 5452 5453 5454
}

/**
 * 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
5455
		enabled_tc |= BIT(i);
J
Jesse Brandeburg 已提交
5456 5457 5458 5459 5460 5461 5462 5463 5464 5465 5466 5467 5468 5469 5470 5471 5472 5473 5474 5475 5476 5477 5478

	/* 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;
}

5479 5480
static int __i40e_setup_tc(struct net_device *netdev, u32 handle, __be16 proto,
			   struct tc_to_netdev *tc)
5481
{
5482
	if (tc->type != TC_SETUP_MQPRIO)
5483
		return -EINVAL;
5484 5485 5486 5487

	tc->mqprio->hw = TC_MQPRIO_HW_OFFLOAD_TCS;

	return i40e_setup_tc(netdev, tc->mqprio->num_tc);
5488 5489
}

J
Jesse Brandeburg 已提交
5490 5491 5492 5493 5494 5495 5496 5497 5498 5499 5500 5501
/**
 * 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
 **/
5502
int i40e_open(struct net_device *netdev)
J
Jesse Brandeburg 已提交
5503 5504 5505 5506 5507 5508
{
	struct i40e_netdev_priv *np = netdev_priv(netdev);
	struct i40e_vsi *vsi = np->vsi;
	struct i40e_pf *pf = vsi->back;
	int err;

5509 5510 5511
	/* 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
5512 5513 5514 5515
		return -EBUSY;

	netif_carrier_off(netdev);

5516 5517 5518 5519
	err = i40e_vsi_open(vsi);
	if (err)
		return err;

5520 5521 5522 5523 5524 5525 5526 5527
	/* 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);

5528
	udp_tunnel_get_rx_info(netdev);
5529

5530 5531 5532 5533 5534 5535 5536 5537 5538 5539 5540 5541 5542 5543
	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;
5544
	char int_name[I40E_INT_NAME_STR_LEN];
5545 5546
	int err;

J
Jesse Brandeburg 已提交
5547 5548 5549 5550 5551 5552 5553 5554 5555 5556 5557 5558
	/* 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;

5559 5560 5561 5562 5563 5564
	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
5565

5566 5567 5568 5569 5570
		/* 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;
5571

5572 5573 5574 5575
		err = netif_set_real_num_rx_queues(vsi->netdev,
						   vsi->num_queue_pairs);
		if (err)
			goto err_set_queues;
5576 5577

	} else if (vsi->type == I40E_VSI_FDIR) {
5578
		snprintf(int_name, sizeof(int_name) - 1, "%s-%s:fdir",
5579 5580
			 dev_driver_string(&pf->pdev->dev),
			 dev_name(&pf->pdev->dev));
5581
		err = i40e_vsi_request_irq(vsi, int_name);
5582

5583
	} else {
5584
		err = -EINVAL;
5585 5586
		goto err_setup_rx;
	}
5587

J
Jesse Brandeburg 已提交
5588 5589 5590 5591 5592 5593 5594 5595
	err = i40e_up_complete(vsi);
	if (err)
		goto err_up_complete;

	return 0;

err_up_complete:
	i40e_down(vsi);
5596
err_set_queues:
J
Jesse Brandeburg 已提交
5597 5598 5599 5600 5601 5602
	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])
5603
		i40e_do_reset(pf, BIT_ULL(__I40E_PF_RESET_REQUESTED));
J
Jesse Brandeburg 已提交
5604 5605 5606 5607

	return err;
}

5608 5609
/**
 * i40e_fdir_filter_exit - Cleans up the Flow Director accounting
5610
 * @pf: Pointer to PF
5611 5612 5613 5614 5615 5616 5617
 *
 * 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;
5618
	struct i40e_flex_pit *pit_entry, *tmp;
5619 5620 5621 5622 5623 5624 5625
	struct hlist_node *node2;

	hlist_for_each_entry_safe(filter, node2,
				  &pf->fdir_filter_list, fdir_node) {
		hlist_del(&filter->fdir_node);
		kfree(filter);
	}
5626

5627 5628 5629 5630 5631 5632 5633 5634 5635 5636 5637 5638
	list_for_each_entry_safe(pit_entry, tmp, &pf->l3_flex_pit_list, list) {
		list_del(&pit_entry->list);
		kfree(pit_entry);
	}
	INIT_LIST_HEAD(&pf->l3_flex_pit_list);

	list_for_each_entry_safe(pit_entry, tmp, &pf->l4_flex_pit_list, list) {
		list_del(&pit_entry->list);
		kfree(pit_entry);
	}
	INIT_LIST_HEAD(&pf->l4_flex_pit_list);

5639
	pf->fdir_pf_active_filters = 0;
5640 5641
	pf->fd_tcp4_filter_cnt = 0;
	pf->fd_udp4_filter_cnt = 0;
5642
	pf->fd_sctp4_filter_cnt = 0;
5643
	pf->fd_ip4_filter_cnt = 0;
5644 5645 5646 5647 5648 5649 5650 5651 5652 5653 5654 5655 5656 5657 5658 5659 5660 5661 5662

	/* Reprogram the default input set for TCP/IPv4 */
	i40e_write_fd_input_set(pf, I40E_FILTER_PCTYPE_NONF_IPV4_TCP,
				I40E_L3_SRC_MASK | I40E_L3_DST_MASK |
				I40E_L4_SRC_MASK | I40E_L4_DST_MASK);

	/* Reprogram the default input set for UDP/IPv4 */
	i40e_write_fd_input_set(pf, I40E_FILTER_PCTYPE_NONF_IPV4_UDP,
				I40E_L3_SRC_MASK | I40E_L3_DST_MASK |
				I40E_L4_SRC_MASK | I40E_L4_DST_MASK);

	/* Reprogram the default input set for SCTP/IPv4 */
	i40e_write_fd_input_set(pf, I40E_FILTER_PCTYPE_NONF_IPV4_SCTP,
				I40E_L3_SRC_MASK | I40E_L3_DST_MASK |
				I40E_L4_SRC_MASK | I40E_L4_DST_MASK);

	/* Reprogram the default input set for Other/IPv4 */
	i40e_write_fd_input_set(pf, I40E_FILTER_PCTYPE_NONF_IPV4_OTHER,
				I40E_L3_SRC_MASK | I40E_L3_DST_MASK);
5663 5664
}

J
Jesse Brandeburg 已提交
5665 5666 5667 5668 5669 5670 5671 5672 5673 5674
/**
 * 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
 **/
5675
int i40e_close(struct net_device *netdev)
J
Jesse Brandeburg 已提交
5676 5677 5678 5679
{
	struct i40e_netdev_priv *np = netdev_priv(netdev);
	struct i40e_vsi *vsi = np->vsi;

5680
	i40e_vsi_close(vsi);
J
Jesse Brandeburg 已提交
5681 5682 5683 5684 5685 5686 5687 5688 5689 5690 5691 5692 5693 5694 5695 5696 5697 5698 5699

	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());

5700

J
Jesse Brandeburg 已提交
5701
	/* do the biggest reset indicated */
5702
	if (reset_flags & BIT_ULL(__I40E_GLOBAL_RESET_REQUESTED)) {
J
Jesse Brandeburg 已提交
5703 5704 5705 5706 5707 5708 5709 5710 5711

		/* 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
5712
		dev_dbg(&pf->pdev->dev, "GlobalR requested\n");
J
Jesse Brandeburg 已提交
5713 5714 5715 5716
		val = rd32(&pf->hw, I40E_GLGEN_RTRIG);
		val |= I40E_GLGEN_RTRIG_GLOBR_MASK;
		wr32(&pf->hw, I40E_GLGEN_RTRIG, val);

5717
	} else if (reset_flags & BIT_ULL(__I40E_CORE_RESET_REQUESTED)) {
J
Jesse Brandeburg 已提交
5718 5719 5720 5721 5722

		/* Request a Core Reset
		 *
		 * Same as Global Reset, except does *not* include the MAC/PHY
		 */
J
Jesse Brandeburg 已提交
5723
		dev_dbg(&pf->pdev->dev, "CoreR requested\n");
J
Jesse Brandeburg 已提交
5724 5725 5726 5727 5728
		val = rd32(&pf->hw, I40E_GLGEN_RTRIG);
		val |= I40E_GLGEN_RTRIG_CORER_MASK;
		wr32(&pf->hw, I40E_GLGEN_RTRIG, val);
		i40e_flush(&pf->hw);

5729
	} else if (reset_flags & BIT_ULL(__I40E_PF_RESET_REQUESTED)) {
J
Jesse Brandeburg 已提交
5730 5731 5732 5733 5734 5735 5736 5737 5738

		/* 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
5739
		dev_dbg(&pf->pdev->dev, "PFR requested\n");
J
Jesse Brandeburg 已提交
5740 5741
		i40e_handle_reset_warning(pf);

5742
	} else if (reset_flags & BIT_ULL(__I40E_REINIT_REQUESTED)) {
J
Jesse Brandeburg 已提交
5743 5744 5745 5746 5747
		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
5748
		for (v = 0; v < pf->num_alloc_vsi; v++) {
J
Jesse Brandeburg 已提交
5749
			struct i40e_vsi *vsi = pf->vsi[v];
J
Jesse Brandeburg 已提交
5750

J
Jesse Brandeburg 已提交
5751 5752 5753 5754 5755 5756
			if (vsi != NULL &&
			    test_bit(__I40E_REINIT_REQUESTED, &vsi->state)) {
				i40e_vsi_reinit_locked(pf->vsi[v]);
				clear_bit(__I40E_REINIT_REQUESTED, &vsi->state);
			}
		}
5757
	} else if (reset_flags & BIT_ULL(__I40E_DOWN_REQUESTED)) {
5758 5759 5760 5761 5762 5763
		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
5764

5765 5766 5767 5768 5769 5770 5771
			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
5772 5773 5774 5775 5776 5777
	} else {
		dev_info(&pf->pdev->dev,
			 "bad reset request 0x%08x\n", reset_flags);
	}
}

N
Neerav Parikh 已提交
5778 5779 5780 5781 5782 5783 5784 5785 5786 5787 5788 5789 5790 5791 5792 5793 5794 5795 5796 5797 5798 5799
#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
5800
			dev_dbg(&pf->pdev->dev, "ETS UP2TC changed.\n");
N
Neerav Parikh 已提交
5801 5802 5803 5804 5805
		}

		if (memcmp(&new_cfg->etscfg.tcbwtable,
			   &old_cfg->etscfg.tcbwtable,
			   sizeof(new_cfg->etscfg.tcbwtable)))
J
Jesse Brandeburg 已提交
5806
			dev_dbg(&pf->pdev->dev, "ETS TC BW Table changed.\n");
N
Neerav Parikh 已提交
5807 5808 5809 5810

		if (memcmp(&new_cfg->etscfg.tsatable,
			   &old_cfg->etscfg.tsatable,
			   sizeof(new_cfg->etscfg.tsatable)))
J
Jesse Brandeburg 已提交
5811
			dev_dbg(&pf->pdev->dev, "ETS TSA Table changed.\n");
N
Neerav Parikh 已提交
5812 5813 5814 5815 5816 5817 5818
	}

	/* 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
5819
		dev_dbg(&pf->pdev->dev, "PFC config change detected.\n");
N
Neerav Parikh 已提交
5820 5821 5822 5823 5824
	}

	/* Check if APP Table has changed */
	if (memcmp(&new_cfg->app,
		   &old_cfg->app,
5825
		   sizeof(new_cfg->app))) {
N
Neerav Parikh 已提交
5826
		need_reconfig = true;
J
Jesse Brandeburg 已提交
5827
		dev_dbg(&pf->pdev->dev, "APP Table change detected.\n");
5828
	}
N
Neerav Parikh 已提交
5829

S
Shannon Nelson 已提交
5830
	dev_dbg(&pf->pdev->dev, "dcb need_reconfig=%d\n", need_reconfig);
N
Neerav Parikh 已提交
5831 5832 5833 5834 5835 5836 5837 5838 5839 5840 5841 5842 5843 5844 5845 5846 5847 5848 5849
	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;

5850
	/* Not DCB capable or capability disabled */
5851
	if (!(pf->flags & I40E_FLAG_DCB_CAPABLE))
5852 5853
		return ret;

N
Neerav Parikh 已提交
5854 5855 5856
	/* 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
5857
	dev_dbg(&pf->pdev->dev, "LLDP event mib bridge type 0x%x\n", type);
N
Neerav Parikh 已提交
5858 5859 5860 5861 5862
	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;
5863
	dev_dbg(&pf->pdev->dev,
S
Shannon Nelson 已提交
5864
		"LLDP event mib type %s\n", type ? "remote" : "local");
N
Neerav Parikh 已提交
5865 5866 5867 5868 5869 5870 5871 5872
	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;
	}

5873
	/* Store the old configuration */
5874
	tmp_dcbx_cfg = hw->local_dcbx_config;
5875

5876 5877
	/* Reset the old DCBx configuration data */
	memset(&hw->local_dcbx_config, 0, sizeof(hw->local_dcbx_config));
5878 5879
	/* Get updated DCBX data from firmware */
	ret = i40e_get_dcb_config(&pf->hw);
N
Neerav Parikh 已提交
5880
	if (ret) {
5881 5882 5883 5884
		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
5885 5886 5887 5888
		goto exit;
	}

	/* No change detected in DCBX configs */
5889 5890
	if (!memcmp(&tmp_dcbx_cfg, &hw->local_dcbx_config,
		    sizeof(tmp_dcbx_cfg))) {
J
Jesse Brandeburg 已提交
5891
		dev_dbg(&pf->pdev->dev, "No change detected in DCBX configuration.\n");
N
Neerav Parikh 已提交
5892 5893 5894
		goto exit;
	}

5895 5896
	need_reconfig = i40e_dcb_need_reconfig(pf, &tmp_dcbx_cfg,
					       &hw->local_dcbx_config);
N
Neerav Parikh 已提交
5897

5898
	i40e_dcbnl_flush_apps(pf, &tmp_dcbx_cfg, &hw->local_dcbx_config);
N
Neerav Parikh 已提交
5899 5900 5901 5902

	if (!need_reconfig)
		goto exit;

5903
	/* Enable DCB tagging only when more than one TC */
5904
	if (i40e_dcb_get_num_tc(&hw->local_dcbx_config) > 1)
5905 5906 5907 5908
		pf->flags |= I40E_FLAG_DCB_ENABLED;
	else
		pf->flags &= ~I40E_FLAG_DCB_ENABLED;

5909
	set_bit(__I40E_PORT_TX_SUSPENDED, &pf->state);
N
Neerav Parikh 已提交
5910 5911 5912 5913 5914 5915
	/* Reconfiguration needed quiesce all VSIs */
	i40e_pf_quiesce_all_vsi(pf);

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

5916 5917
	ret = i40e_resume_port_tx(pf);

5918
	clear_bit(__I40E_PORT_TX_SUSPENDED, &pf->state);
5919
	/* In case of error no point in resuming VSIs */
5920 5921 5922
	if (ret)
		goto exit;

5923 5924
	/* Wait for the PF's queues to be disabled */
	ret = i40e_pf_wait_queues_disabled(pf);
5925 5926 5927 5928 5929
	if (ret) {
		/* Schedule PF reset to recover */
		set_bit(__I40E_PF_RESET_REQUESTED, &pf->state);
		i40e_service_event_schedule(pf);
	} else {
5930
		i40e_pf_unquiesce_all_vsi(pf);
M
Mitch Williams 已提交
5931 5932
	pf->flags |= (I40E_FLAG_SERVICE_CLIENT_REQUESTED |
		      I40E_FLAG_CLIENT_L2_CHANGE);
5933 5934
	}

N
Neerav Parikh 已提交
5935 5936 5937 5938 5939
exit:
	return ret;
}
#endif /* CONFIG_I40E_DCB */

5940 5941 5942 5943 5944 5945 5946 5947 5948 5949 5950 5951 5952
/**
 * 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
5953 5954 5955 5956 5957 5958 5959 5960 5961 5962 5963 5964 5965 5966 5967 5968 5969 5970 5971
/**
 * 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
5972 5973
	dev_dbg(&pf->pdev->dev, "overflow Rx Queue Number = %d QTX_CTL=0x%08x\n",
		queue, qtx_ctl);
J
Jesse Brandeburg 已提交
5974 5975 5976 5977 5978 5979 5980 5981 5982 5983 5984 5985 5986 5987 5988

	/* 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);
	}
}

5989
/**
5990 5991 5992
 * i40e_get_cur_guaranteed_fd_count - Get the consumed guaranteed FD filters
 * @pf: board private structure
 **/
5993
u32 i40e_get_cur_guaranteed_fd_count(struct i40e_pf *pf)
5994
{
5995
	u32 val, fcnt_prog;
5996 5997 5998 5999 6000 6001 6002

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

/**
6003
 * i40e_get_current_fd_count - Get total FD filters programmed for this PF
6004 6005
 * @pf: board private structure
 **/
6006
u32 i40e_get_current_fd_count(struct i40e_pf *pf)
6007
{
6008 6009
	u32 val, fcnt_prog;

6010 6011 6012 6013 6014 6015
	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;
}
6016

6017 6018 6019 6020 6021 6022 6023 6024 6025 6026 6027 6028 6029 6030 6031
/**
 * 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;
}

6032 6033 6034 6035 6036 6037
/**
 * 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)
{
6038
	struct i40e_fdir_filter *filter;
6039
	u32 fcnt_prog, fcnt_avail;
6040
	struct hlist_node *node;
6041

6042 6043 6044
	if (test_bit(__I40E_FD_FLUSH_REQUESTED, &pf->state))
		return;

6045 6046 6047
	/* Check if, FD SB or ATR was auto disabled and if there is enough room
	 * to re-enable
	 */
6048
	fcnt_prog = i40e_get_global_fd_count(pf);
6049
	fcnt_avail = pf->fdir_pf_filter_count;
6050 6051 6052
	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)) {
6053
		if ((pf->flags & I40E_FLAG_FD_SB_ENABLED) &&
6054 6055
		    (pf->hw_disabled_flags & I40E_FLAG_FD_SB_ENABLED)) {
			pf->hw_disabled_flags &= ~I40E_FLAG_FD_SB_ENABLED;
6056 6057
			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");
6058 6059
		}
	}
6060 6061 6062 6063

	/* 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
	 */
6064 6065
	if (fcnt_prog < (fcnt_avail - I40E_FDIR_BUFFER_HEAD_ROOM * 2)) {
		if ((pf->flags & I40E_FLAG_FD_ATR_ENABLED) &&
6066
		    (pf->hw_disabled_flags & I40E_FLAG_FD_ATR_ENABLED) &&
6067
		    (pf->fd_tcp4_filter_cnt == 0)) {
6068
			pf->hw_disabled_flags &= ~I40E_FLAG_FD_ATR_ENABLED;
6069
			if (I40E_DEBUG_FD & pf->hw.debug_mask)
6070
				dev_info(&pf->pdev->dev, "ATR is being enabled since we have space in the table and there are no conflicting ntuple rules\n");
6071 6072
		}
	}
6073 6074 6075 6076 6077 6078 6079 6080 6081 6082 6083 6084

	/* 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--;
			}
		}
	}
6085 6086
}

6087
#define I40E_MIN_FD_FLUSH_INTERVAL 10
6088
#define I40E_MIN_FD_FLUSH_SB_ATR_UNSTABLE 30
6089 6090 6091 6092 6093 6094
/**
 * 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)
{
6095
	unsigned long min_flush_time;
6096
	int flush_wait_retry = 50;
6097 6098
	bool disable_atr = false;
	int fd_room;
6099 6100
	int reg;

6101 6102 6103
	if (!time_after(jiffies, pf->fd_flush_timestamp +
				 (I40E_MIN_FD_FLUSH_INTERVAL * HZ)))
		return;
6104

6105 6106 6107 6108 6109 6110 6111 6112 6113 6114 6115 6116 6117 6118 6119
	/* 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;
6120
	pf->hw_disabled_flags |= I40E_FLAG_FD_ATR_ENABLED;
6121 6122 6123 6124 6125 6126 6127 6128 6129 6130 6131 6132 6133 6134 6135 6136 6137 6138
	/* 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]);
6139
		if (!disable_atr && !pf->fd_tcp4_filter_cnt)
6140
			pf->hw_disabled_flags &= ~I40E_FLAG_FD_ATR_ENABLED;
6141 6142 6143
		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");
6144 6145 6146 6147 6148 6149 6150
	}
}

/**
 * i40e_get_current_atr_count - Get the count of total FD ATR filters programmed
 * @pf: board private structure
 **/
6151
u32 i40e_get_current_atr_cnt(struct i40e_pf *pf)
6152 6153 6154 6155 6156 6157 6158 6159 6160 6161 6162
{
	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
6163 6164 6165 6166 6167 6168 6169 6170 6171 6172
/**
 * 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;
6173

6174
	if (test_bit(__I40E_FD_FLUSH_REQUESTED, &pf->state))
6175 6176
		i40e_fdir_flush_and_replay(pf);

6177 6178
	i40e_fdir_check_and_reenable(pf);

J
Jesse Brandeburg 已提交
6179 6180 6181 6182 6183 6184 6185 6186 6187
}

/**
 * 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)
{
6188
	if (!vsi || test_bit(__I40E_DOWN, &vsi->state))
J
Jesse Brandeburg 已提交
6189 6190 6191 6192 6193 6194 6195 6196 6197 6198 6199 6200 6201 6202 6203 6204 6205 6206 6207
		return;

	switch (vsi->type) {
	case I40E_VSI_MAIN:
		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:
6208
	case I40E_VSI_IWARP:
J
Jesse Brandeburg 已提交
6209 6210 6211 6212 6213 6214 6215 6216 6217 6218 6219 6220 6221 6222 6223 6224 6225 6226 6227 6228 6229 6230 6231 6232 6233 6234 6235
	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
6236
	for (i = 0; i < pf->num_alloc_vsi; i++)
J
Jesse Brandeburg 已提交
6237 6238 6239 6240 6241 6242 6243 6244 6245 6246
		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
6247
	struct i40e_vsi *vsi = pf->vsi[pf->lan_vsi];
6248
	u8 new_link_speed, old_link_speed;
6249 6250
	i40e_status status;
	bool new_link, old_link;
J
Jesse Brandeburg 已提交
6251

6252 6253 6254
	/* save off old link status information */
	pf->hw.phy.link_info_old = pf->hw.phy.link_info;

6255 6256 6257
	/* 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
6258
	old_link = (pf->hw.phy.link_info_old.link_info & I40E_AQ_LINK_UP);
6259 6260

	status = i40e_get_link_status(&pf->hw, &new_link);
6261 6262 6263 6264 6265 6266 6267 6268 6269 6270

	/* On success, disable temp link polling */
	if (status == I40E_SUCCESS) {
		if (pf->flags & I40E_FLAG_TEMP_LINK_POLLING)
			pf->flags &= ~I40E_FLAG_TEMP_LINK_POLLING;
	} else {
		/* Enable link polling temporarily until i40e_get_link_status
		 * returns I40E_SUCCESS
		 */
		pf->flags |= I40E_FLAG_TEMP_LINK_POLLING;
6271 6272 6273 6274 6275
		dev_dbg(&pf->pdev->dev, "couldn't get link state, status: %d\n",
			status);
		return;
	}

6276 6277
	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
6278

6279
	if (new_link == old_link &&
6280
	    new_link_speed == old_link_speed &&
M
Mitch Williams 已提交
6281 6282
	    (test_bit(__I40E_DOWN, &vsi->state) ||
	     new_link == netif_carrier_ok(vsi->netdev)))
J
Jesse Brandeburg 已提交
6283
		return;
M
Mitch Williams 已提交
6284 6285 6286

	if (!test_bit(__I40E_DOWN, &vsi->state))
		i40e_print_link_message(vsi, new_link);
J
Jesse Brandeburg 已提交
6287 6288 6289 6290 6291 6292 6293

	/* 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
6294
		i40e_vsi_link_event(vsi, new_link);
J
Jesse Brandeburg 已提交
6295 6296 6297

	if (pf->vf)
		i40e_vc_notify_link_state(pf);
J
Jacob Keller 已提交
6298 6299 6300

	if (pf->flags & I40E_FLAG_PTP)
		i40e_ptp_set_increment(pf);
J
Jesse Brandeburg 已提交
6301 6302 6303
}

/**
6304
 * i40e_watchdog_subtask - periodic checks not using event driven response
J
Jesse Brandeburg 已提交
6305 6306 6307 6308 6309 6310 6311 6312 6313 6314 6315
 * @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;

6316 6317 6318 6319 6320 6321
	/* 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;

6322 6323
	if ((pf->flags & I40E_FLAG_LINK_POLLING_ENABLED) ||
	    (pf->flags & I40E_FLAG_TEMP_LINK_POLLING))
S
Shannon Nelson 已提交
6324
		i40e_link_event(pf);
6325

J
Jesse Brandeburg 已提交
6326 6327 6328
	/* 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
6329
	for (i = 0; i < pf->num_alloc_vsi; i++)
J
Jesse Brandeburg 已提交
6330 6331 6332
		if (pf->vsi[i] && pf->vsi[i]->netdev)
			i40e_update_stats(pf->vsi[i]);

6333 6334 6335 6336 6337 6338
	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
6339 6340

	i40e_ptp_rx_hang(pf->vsi[pf->lan_vsi]);
J
Jesse Brandeburg 已提交
6341 6342 6343 6344 6345 6346 6347 6348 6349 6350
}

/**
 * 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;

6351
	rtnl_lock();
J
Jesse Brandeburg 已提交
6352
	if (test_bit(__I40E_REINIT_REQUESTED, &pf->state)) {
J
Jesse Brandeburg 已提交
6353
		reset_flags |= BIT(__I40E_REINIT_REQUESTED);
J
Jesse Brandeburg 已提交
6354 6355 6356
		clear_bit(__I40E_REINIT_REQUESTED, &pf->state);
	}
	if (test_bit(__I40E_PF_RESET_REQUESTED, &pf->state)) {
J
Jesse Brandeburg 已提交
6357
		reset_flags |= BIT(__I40E_PF_RESET_REQUESTED);
J
Jesse Brandeburg 已提交
6358 6359 6360
		clear_bit(__I40E_PF_RESET_REQUESTED, &pf->state);
	}
	if (test_bit(__I40E_CORE_RESET_REQUESTED, &pf->state)) {
J
Jesse Brandeburg 已提交
6361
		reset_flags |= BIT(__I40E_CORE_RESET_REQUESTED);
J
Jesse Brandeburg 已提交
6362 6363 6364
		clear_bit(__I40E_CORE_RESET_REQUESTED, &pf->state);
	}
	if (test_bit(__I40E_GLOBAL_RESET_REQUESTED, &pf->state)) {
J
Jesse Brandeburg 已提交
6365
		reset_flags |= BIT(__I40E_GLOBAL_RESET_REQUESTED);
J
Jesse Brandeburg 已提交
6366 6367
		clear_bit(__I40E_GLOBAL_RESET_REQUESTED, &pf->state);
	}
6368
	if (test_bit(__I40E_DOWN_REQUESTED, &pf->state)) {
J
Jesse Brandeburg 已提交
6369
		reset_flags |= BIT(__I40E_DOWN_REQUESTED);
6370 6371
		clear_bit(__I40E_DOWN_REQUESTED, &pf->state);
	}
J
Jesse Brandeburg 已提交
6372 6373 6374 6375 6376 6377

	/* 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);
6378
		goto unlock;
J
Jesse Brandeburg 已提交
6379 6380 6381 6382 6383 6384 6385
	}

	/* 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);
6386 6387 6388

unlock:
	rtnl_unlock();
J
Jesse Brandeburg 已提交
6389 6390 6391 6392 6393 6394 6395 6396 6397 6398 6399 6400 6401
}

/**
 * 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;

6402 6403 6404 6405 6406 6407 6408 6409
	/* 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);

6410 6411 6412 6413 6414 6415
	/* 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
6416 6417 6418 6419 6420 6421 6422 6423 6424 6425 6426 6427 6428
}

/**
 * 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;
6429
	u32 oldval;
J
Jesse Brandeburg 已提交
6430 6431
	u32 val;

6432 6433 6434 6435
	/* Do not run clean AQ when PF reset fails */
	if (test_bit(__I40E_RESET_FAILED, &pf->state))
		return;

6436 6437 6438 6439
	/* 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
6440 6441
		if (hw->debug_mask & I40E_DEBUG_AQ)
			dev_info(&pf->pdev->dev, "ARQ VF Error detected\n");
6442 6443 6444
		val &= ~I40E_PF_ARQLEN_ARQVFE_MASK;
	}
	if (val & I40E_PF_ARQLEN_ARQOVFL_MASK) {
M
Mitch Williams 已提交
6445 6446
		if (hw->debug_mask & I40E_DEBUG_AQ)
			dev_info(&pf->pdev->dev, "ARQ Overflow Error detected\n");
6447
		val &= ~I40E_PF_ARQLEN_ARQOVFL_MASK;
6448
		pf->arq_overflows++;
6449 6450
	}
	if (val & I40E_PF_ARQLEN_ARQCRIT_MASK) {
M
Mitch Williams 已提交
6451 6452
		if (hw->debug_mask & I40E_DEBUG_AQ)
			dev_info(&pf->pdev->dev, "ARQ Critical Error detected\n");
6453 6454 6455 6456 6457 6458 6459 6460
		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
6461 6462
		if (pf->hw.debug_mask & I40E_DEBUG_AQ)
			dev_info(&pf->pdev->dev, "ASQ VF Error detected\n");
6463 6464 6465
		val &= ~I40E_PF_ATQLEN_ATQVFE_MASK;
	}
	if (val & I40E_PF_ATQLEN_ATQOVFL_MASK) {
M
Mitch Williams 已提交
6466 6467
		if (pf->hw.debug_mask & I40E_DEBUG_AQ)
			dev_info(&pf->pdev->dev, "ASQ Overflow Error detected\n");
6468 6469 6470
		val &= ~I40E_PF_ATQLEN_ATQOVFL_MASK;
	}
	if (val & I40E_PF_ATQLEN_ATQCRIT_MASK) {
M
Mitch Williams 已提交
6471 6472
		if (pf->hw.debug_mask & I40E_DEBUG_AQ)
			dev_info(&pf->pdev->dev, "ASQ Critical Error detected\n");
6473 6474 6475 6476 6477
		val &= ~I40E_PF_ATQLEN_ATQCRIT_MASK;
	}
	if (oldval != val)
		wr32(&pf->hw, pf->hw.aq.asq.len, val);

M
Mitch Williams 已提交
6478 6479
	event.buf_len = I40E_MAX_AQ_BUF_SIZE;
	event.msg_buf = kzalloc(event.buf_len, GFP_KERNEL);
J
Jesse Brandeburg 已提交
6480 6481 6482 6483 6484
	if (!event.msg_buf)
		return;

	do {
		ret = i40e_clean_arq_element(hw, &event, &pending);
M
Mitch Williams 已提交
6485
		if (ret == I40E_ERR_ADMIN_QUEUE_NO_WORK)
J
Jesse Brandeburg 已提交
6486
			break;
M
Mitch Williams 已提交
6487
		else if (ret) {
J
Jesse Brandeburg 已提交
6488 6489 6490 6491 6492 6493 6494 6495 6496 6497 6498 6499 6500 6501 6502 6503
			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
6504
					event.msg_len);
J
Jesse Brandeburg 已提交
6505 6506
			break;
		case i40e_aqc_opc_lldp_update_mib:
J
Jesse Brandeburg 已提交
6507
			dev_dbg(&pf->pdev->dev, "ARQ: Update LLDP MIB event received\n");
N
Neerav Parikh 已提交
6508 6509 6510 6511 6512
#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
6513 6514
			break;
		case i40e_aqc_opc_event_lan_overflow:
J
Jesse Brandeburg 已提交
6515
			dev_dbg(&pf->pdev->dev, "ARQ LAN queue overflow event received\n");
J
Jesse Brandeburg 已提交
6516 6517
			i40e_handle_lan_overflow_event(pf, &event);
			break;
6518 6519 6520
		case i40e_aqc_opc_send_msg_to_peer:
			dev_info(&pf->pdev->dev, "ARQ: Msg from other pf\n");
			break;
6521 6522
		case i40e_aqc_opc_nvm_erase:
		case i40e_aqc_opc_nvm_update:
6523
		case i40e_aqc_opc_oem_post_update:
6524 6525 6526
			i40e_debug(&pf->hw, I40E_DEBUG_NVM,
				   "ARQ NVM operation 0x%04x completed\n",
				   opcode);
6527
			break;
J
Jesse Brandeburg 已提交
6528 6529
		default:
			dev_info(&pf->pdev->dev,
6530
				 "ARQ: Unknown event 0x%04x ignored\n",
6531
				 opcode);
J
Jesse Brandeburg 已提交
6532 6533
			break;
		}
6534 6535 6536 6537
	} while (i++ < pf->adminq_work_limit);

	if (i < pf->adminq_work_limit)
		clear_bit(__I40E_ADMINQ_EVENT_PENDING, &pf->state);
J
Jesse Brandeburg 已提交
6538 6539 6540 6541 6542 6543 6544 6545 6546 6547

	/* 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);
}

6548 6549 6550 6551 6552 6553 6554 6555 6556 6557 6558 6559 6560 6561 6562 6563 6564 6565 6566 6567 6568 6569 6570 6571 6572
/**
 * 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);
	}
}

6573 6574
/**
 * i40e_enable_pf_switch_lb
6575
 * @pf: pointer to the PF structure
6576 6577 6578 6579 6580 6581 6582
 *
 * 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;
6583
	int ret;
6584 6585 6586 6587

	ctxt.seid = pf->main_vsi_seid;
	ctxt.pf_num = pf->hw.pf_id;
	ctxt.vf_num = 0;
6588 6589
	ret = i40e_aq_get_vsi_params(&pf->hw, &ctxt, NULL);
	if (ret) {
6590
		dev_info(&pf->pdev->dev,
6591 6592 6593
			 "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));
6594 6595 6596 6597 6598 6599
		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);

6600 6601
	ret = i40e_aq_update_vsi_params(&vsi->back->hw, &ctxt, NULL);
	if (ret) {
6602
		dev_info(&pf->pdev->dev,
6603 6604 6605
			 "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));
6606 6607 6608 6609 6610
	}
}

/**
 * i40e_disable_pf_switch_lb
6611
 * @pf: pointer to the PF structure
6612 6613 6614 6615 6616 6617 6618
 *
 * 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;
6619
	int ret;
6620 6621 6622 6623

	ctxt.seid = pf->main_vsi_seid;
	ctxt.pf_num = pf->hw.pf_id;
	ctxt.vf_num = 0;
6624 6625
	ret = i40e_aq_get_vsi_params(&pf->hw, &ctxt, NULL);
	if (ret) {
6626
		dev_info(&pf->pdev->dev,
6627 6628 6629
			 "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));
6630 6631 6632 6633 6634 6635
		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);

6636 6637
	ret = i40e_aq_update_vsi_params(&vsi->back->hw, &ctxt, NULL);
	if (ret) {
6638
		dev_info(&pf->pdev->dev,
6639 6640 6641
			 "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));
6642 6643 6644
	}
}

6645 6646 6647 6648 6649 6650 6651 6652 6653 6654 6655 6656
/**
 * 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;

6657 6658 6659
	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");
6660 6661 6662 6663 6664 6665
	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
6666 6667 6668 6669 6670 6671 6672 6673 6674 6675 6676 6677 6678 6679 6680 6681 6682
/**
 * 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
6683
	for (v = 0; v < pf->num_alloc_vsi && !ctl_vsi; v++) {
J
Jesse Brandeburg 已提交
6684 6685 6686 6687 6688 6689 6690 6691 6692 6693 6694 6695 6696 6697 6698 6699 6700 6701
		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,
6702 6703
			 "rebuild of veb_idx %d owner VSI failed: %d\n",
			 veb->idx, ret);
J
Jesse Brandeburg 已提交
6704 6705 6706 6707 6708 6709 6710 6711 6712
		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;

6713 6714 6715 6716
	if (pf->flags & I40E_FLAG_VEB_MODE_ENABLED)
		veb->bridge_mode = BRIDGE_MODE_VEB;
	else
		veb->bridge_mode = BRIDGE_MODE_VEPA;
6717
	i40e_config_bridge_mode(veb);
6718

J
Jesse Brandeburg 已提交
6719
	/* create the remaining VSIs attached to this VEB */
M
Mitch Williams 已提交
6720
	for (v = 0; v < pf->num_alloc_vsi; v++) {
J
Jesse Brandeburg 已提交
6721 6722 6723 6724 6725
		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
6726

J
Jesse Brandeburg 已提交
6727 6728 6729 6730 6731 6732 6733 6734 6735 6736 6737 6738 6739 6740 6741 6742 6743 6744 6745 6746 6747 6748 6749 6750 6751 6752 6753 6754 6755 6756 6757 6758 6759 6760 6761 6762 6763 6764 6765 6766 6767 6768 6769 6770 6771 6772 6773 6774 6775 6776 6777 6778 6779 6780 6781 6782
			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,
6783 6784 6785 6786
				 "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
6787 6788 6789 6790 6791 6792 6793 6794 6795 6796 6797 6798 6799 6800 6801
			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);

6802 6803 6804 6805 6806 6807 6808 6809 6810
#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
6811 6812 6813
	return 0;
}

6814 6815
static int i40e_vsi_clear(struct i40e_vsi *vsi);

J
Jesse Brandeburg 已提交
6816
/**
6817
 * i40e_fdir_sb_setup - initialize the Flow Director resources for Sideband
J
Jesse Brandeburg 已提交
6818 6819
 * @pf: board private structure
 **/
6820
static void i40e_fdir_sb_setup(struct i40e_pf *pf)
J
Jesse Brandeburg 已提交
6821 6822 6823
{
	struct i40e_vsi *vsi;

6824 6825 6826 6827 6828 6829 6830 6831 6832
	/* 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};
6833
		int i;
6834 6835 6836 6837 6838

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

6839
	if (!(pf->flags & I40E_FLAG_FD_SB_ENABLED))
J
Jesse Brandeburg 已提交
6840 6841
		return;

6842
	/* find existing VSI and see if it needs configuring */
6843
	vsi = i40e_find_vsi_by_type(pf, I40E_VSI_FDIR);
6844 6845

	/* create a new VSI if none exists */
J
Jesse Brandeburg 已提交
6846
	if (!vsi) {
6847 6848
		vsi = i40e_vsi_setup(pf, I40E_VSI_FDIR,
				     pf->vsi[pf->lan_vsi]->seid, 0);
J
Jesse Brandeburg 已提交
6849 6850
		if (!vsi) {
			dev_info(&pf->pdev->dev, "Couldn't create FDir VSI\n");
6851 6852
			pf->flags &= ~I40E_FLAG_FD_SB_ENABLED;
			return;
J
Jesse Brandeburg 已提交
6853
		}
6854
	}
J
Jesse Brandeburg 已提交
6855

6856
	i40e_vsi_setup_irqhandler(vsi, i40e_fdir_clean_ring);
J
Jesse Brandeburg 已提交
6857 6858 6859 6860 6861 6862 6863 6864
}

/**
 * i40e_fdir_teardown - release the Flow Director resources
 * @pf: board private structure
 **/
static void i40e_fdir_teardown(struct i40e_pf *pf)
{
6865
	struct i40e_vsi *vsi;
J
Jesse Brandeburg 已提交
6866

6867
	i40e_fdir_filter_exit(pf);
6868 6869 6870
	vsi = i40e_find_vsi_by_type(pf, I40E_VSI_FDIR);
	if (vsi)
		i40e_vsi_release(vsi);
J
Jesse Brandeburg 已提交
6871 6872 6873
}

/**
A
Anjali Singhai Jain 已提交
6874
 * i40e_prep_for_reset - prep for the core to reset
J
Jesse Brandeburg 已提交
6875 6876
 * @pf: board private structure
 *
6877
 * Close up the VFs and other things in prep for PF Reset.
A
Anjali Singhai Jain 已提交
6878
  **/
S
Shannon Nelson 已提交
6879
static void i40e_prep_for_reset(struct i40e_pf *pf)
J
Jesse Brandeburg 已提交
6880 6881
{
	struct i40e_hw *hw = &pf->hw;
6882
	i40e_status ret = 0;
J
Jesse Brandeburg 已提交
6883 6884 6885 6886
	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
6887
		return;
M
Mitch Williams 已提交
6888 6889
	if (i40e_check_asq_alive(&pf->hw))
		i40e_vc_notify_reset(pf);
J
Jesse Brandeburg 已提交
6890

J
Jesse Brandeburg 已提交
6891
	dev_dbg(&pf->pdev->dev, "Tearing down internal switch for reset\n");
J
Jesse Brandeburg 已提交
6892 6893 6894 6895

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

M
Mitch Williams 已提交
6896
	for (v = 0; v < pf->num_alloc_vsi; v++) {
J
Jesse Brandeburg 已提交
6897 6898 6899 6900 6901 6902
		if (pf->vsi[v])
			pf->vsi[v]->seid = 0;
	}

	i40e_shutdown_adminq(&pf->hw);

A
Anjali Singhai Jain 已提交
6903
	/* call shutdown HMC */
6904 6905
	if (hw->hmc.hmc_obj) {
		ret = i40e_shutdown_lan_hmc(hw);
S
Shannon Nelson 已提交
6906
		if (ret)
6907 6908
			dev_warn(&pf->pdev->dev,
				 "shutdown_lan_hmc failed: %d\n", ret);
A
Anjali Singhai Jain 已提交
6909 6910 6911
	}
}

J
Jesse Brandeburg 已提交
6912 6913 6914 6915 6916 6917 6918 6919 6920 6921 6922 6923
/**
 * 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;
6924
	strlcpy(dv.driver_string, DRV_VERSION, sizeof(dv.driver_string));
J
Jesse Brandeburg 已提交
6925 6926 6927
	i40e_aq_send_driver_version(&pf->hw, &dv, NULL);
}

A
Anjali Singhai Jain 已提交
6928
/**
J
Jesse Brandeburg 已提交
6929
 * i40e_reset_and_rebuild - reset and rebuild using a saved config
A
Anjali Singhai Jain 已提交
6930
 * @pf: board private structure
6931
 * @reinit: if the Main VSI needs to re-initialized.
A
Anjali Singhai Jain 已提交
6932
 **/
6933
static void i40e_reset_and_rebuild(struct i40e_pf *pf, bool reinit)
A
Anjali Singhai Jain 已提交
6934 6935
{
	struct i40e_hw *hw = &pf->hw;
6936
	u8 set_fc_aq_fail = 0;
A
Anjali Singhai Jain 已提交
6937
	i40e_status ret;
6938
	u32 val;
A
Anjali Singhai Jain 已提交
6939 6940
	u32 v;

J
Jesse Brandeburg 已提交
6941 6942 6943 6944 6945
	/* 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);
6946
	if (ret) {
J
Jesse Brandeburg 已提交
6947
		dev_info(&pf->pdev->dev, "PF reset failed, %d\n", ret);
6948 6949
		set_bit(__I40E_RESET_FAILED, &pf->state);
		goto clear_recovery;
6950
	}
J
Jesse Brandeburg 已提交
6951 6952 6953
	pf->pfr_count++;

	if (test_bit(__I40E_DOWN, &pf->state))
6954
		goto clear_recovery;
J
Jesse Brandeburg 已提交
6955
	dev_dbg(&pf->pdev->dev, "Rebuilding internal switch\n");
J
Jesse Brandeburg 已提交
6956 6957 6958 6959

	/* rebuild the basics for the AdminQ, HMC, and initial HW switch */
	ret = i40e_init_adminq(&pf->hw);
	if (ret) {
6960 6961 6962
		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));
6963
		goto clear_recovery;
J
Jesse Brandeburg 已提交
6964 6965
	}

6966
	/* re-verify the eeprom if we just had an EMP reset */
6967
	if (test_and_clear_bit(__I40E_EMP_RESET_INTR_RECEIVED, &pf->state))
6968 6969
		i40e_verify_eeprom(pf);

6970
	i40e_clear_pxe_mode(hw);
J
Jesse Brandeburg 已提交
6971
	ret = i40e_get_capabilities(pf);
6972
	if (ret)
J
Jesse Brandeburg 已提交
6973 6974 6975
		goto end_core_reset;

	ret = i40e_init_lan_hmc(hw, hw->func_caps.num_tx_qp,
6976
				hw->func_caps.num_rx_qp, 0, 0);
J
Jesse Brandeburg 已提交
6977 6978 6979 6980 6981 6982 6983 6984 6985 6986
	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
6987 6988 6989
#ifdef CONFIG_I40E_DCB
	ret = i40e_init_pf_dcb(pf);
	if (ret) {
6990 6991 6992
		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
6993 6994
	}
#endif /* CONFIG_I40E_DCB */
J
Jesse Brandeburg 已提交
6995
	/* do basic switch setup */
6996
	ret = i40e_setup_pf_switch(pf, reinit);
J
Jesse Brandeburg 已提交
6997 6998 6999
	if (ret)
		goto end_core_reset;

S
Shannon Nelson 已提交
7000 7001
	/* The driver only wants link up/down and module qualification
	 * reports from firmware.  Note the negative logic.
J
Jesse Brandeburg 已提交
7002 7003
	 */
	ret = i40e_aq_set_phy_int_mask(&pf->hw,
S
Shannon Nelson 已提交
7004
				       ~(I40E_AQ_EVENT_LINK_UPDOWN |
7005
					 I40E_AQ_EVENT_MEDIA_NA |
S
Shannon Nelson 已提交
7006
					 I40E_AQ_EVENT_MODULE_QUAL_FAIL), NULL);
J
Jesse Brandeburg 已提交
7007
	if (ret)
7008 7009 7010
		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
7011

7012 7013 7014
	/* make sure our flow control settings are restored */
	ret = i40e_set_fc(&pf->hw, &set_fc_aq_fail, true);
	if (ret)
7015 7016 7017
		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));
7018

J
Jesse Brandeburg 已提交
7019 7020 7021 7022 7023 7024 7025 7026
	/* 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
7027
		dev_dbg(&pf->pdev->dev, "attempting to rebuild switch\n");
J
Jesse Brandeburg 已提交
7028 7029 7030 7031 7032 7033 7034 7035 7036 7037 7038 7039 7040 7041 7042 7043 7044 7045 7046 7047 7048 7049 7050 7051 7052 7053 7054 7055 7056 7057 7058 7059 7060 7061 7062
		/* 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
7063
		dev_dbg(&pf->pdev->dev, "attempting to rebuild PF VSI\n");
J
Jesse Brandeburg 已提交
7064 7065 7066 7067 7068 7069 7070 7071 7072
		/* 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;
		}
	}

7073 7074 7075 7076 7077 7078 7079 7080 7081 7082 7083 7084 7085 7086
	/* 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);
	}

7087
	if (pf->flags & I40E_FLAG_RESTART_AUTONEG) {
7088 7089 7090
		msleep(75);
		ret = i40e_aq_set_link_restart_an(&pf->hw, true, NULL);
		if (ret)
7091 7092 7093 7094
			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));
7095
	}
J
Jesse Brandeburg 已提交
7096 7097 7098 7099
	/* reinit the misc interrupt */
	if (pf->flags & I40E_FLAG_MSIX_ENABLED)
		ret = i40e_setup_misc_vector(pf);

7100 7101 7102 7103 7104 7105 7106 7107 7108
	/* 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
7109 7110 7111
	/* restart the VSIs that were rebuilt and running before the reset */
	i40e_pf_unquiesce_all_vsi(pf);

M
Mitch Williams 已提交
7112 7113 7114 7115 7116
	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
7117
	/* tell the firmware that we're starting */
J
Jesse Brandeburg 已提交
7118
	i40e_send_version(pf);
J
Jesse Brandeburg 已提交
7119 7120

end_core_reset:
7121 7122
	clear_bit(__I40E_RESET_FAILED, &pf->state);
clear_recovery:
J
Jesse Brandeburg 已提交
7123 7124 7125
	clear_bit(__I40E_RESET_RECOVERY_PENDING, &pf->state);
}

A
Anjali Singhai Jain 已提交
7126
/**
7127
 * i40e_handle_reset_warning - prep for the PF to reset, reset and rebuild
A
Anjali Singhai Jain 已提交
7128 7129 7130 7131 7132 7133 7134
 * @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
7135 7136
	i40e_prep_for_reset(pf);
	i40e_reset_and_rebuild(pf, false);
A
Anjali Singhai Jain 已提交
7137 7138
}

J
Jesse Brandeburg 已提交
7139 7140
/**
 * i40e_handle_mdd_event
7141
 * @pf: pointer to the PF structure
J
Jesse Brandeburg 已提交
7142 7143 7144 7145 7146 7147 7148
 *
 * 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;
7149
	bool pf_mdd_detected = false;
J
Jesse Brandeburg 已提交
7150 7151 7152 7153 7154 7155 7156 7157 7158 7159
	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) {
7160 7161
		u8 pf_num = (reg & I40E_GL_MDET_TX_PF_NUM_MASK) >>
				I40E_GL_MDET_TX_PF_NUM_SHIFT;
7162
		u16 vf_num = (reg & I40E_GL_MDET_TX_VF_NUM_MASK) >>
7163
				I40E_GL_MDET_TX_VF_NUM_SHIFT;
7164
		u8 event = (reg & I40E_GL_MDET_TX_EVENT_MASK) >>
7165
				I40E_GL_MDET_TX_EVENT_SHIFT;
7166 7167 7168
		u16 queue = ((reg & I40E_GL_MDET_TX_QUEUE_MASK) >>
				I40E_GL_MDET_TX_QUEUE_SHIFT) -
				pf->hw.func_caps.base_queue;
7169
		if (netif_msg_tx_err(pf))
7170
			dev_info(&pf->pdev->dev, "Malicious Driver Detection event 0x%02x on TX queue %d PF number 0x%02x VF number 0x%02x\n",
7171
				 event, queue, pf_num, vf_num);
J
Jesse Brandeburg 已提交
7172 7173 7174 7175 7176
		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) {
7177 7178
		u8 func = (reg & I40E_GL_MDET_RX_FUNCTION_MASK) >>
				I40E_GL_MDET_RX_FUNCTION_SHIFT;
7179
		u8 event = (reg & I40E_GL_MDET_RX_EVENT_MASK) >>
7180
				I40E_GL_MDET_RX_EVENT_SHIFT;
7181 7182 7183
		u16 queue = ((reg & I40E_GL_MDET_RX_QUEUE_MASK) >>
				I40E_GL_MDET_RX_QUEUE_SHIFT) -
				pf->hw.func_caps.base_queue;
7184 7185 7186
		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
7187 7188 7189 7190
		wr32(hw, I40E_GL_MDET_RX, 0xffffffff);
		mdd_detected = true;
	}

7191 7192 7193 7194
	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);
7195
			dev_info(&pf->pdev->dev, "TX driver issue detected, PF reset issued\n");
7196 7197 7198 7199 7200
			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);
7201
			dev_info(&pf->pdev->dev, "RX driver issue detected, PF reset issued\n");
7202 7203 7204 7205 7206 7207 7208 7209 7210
			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
7211 7212 7213 7214 7215 7216 7217
	/* 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++;
7218 7219
			dev_info(&pf->pdev->dev, "TX driver issue detected on VF %d\n",
				 i);
J
Jesse Brandeburg 已提交
7220 7221 7222 7223 7224 7225
		}

		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++;
7226 7227
			dev_info(&pf->pdev->dev, "RX driver issue detected on VF %d\n",
				 i);
J
Jesse Brandeburg 已提交
7228 7229 7230 7231 7232 7233 7234 7235 7236 7237 7238 7239 7240 7241 7242 7243 7244 7245 7246
		}

		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);
}

7247
/**
7248
 * i40e_sync_udp_filters_subtask - Sync the VSI filter list with HW
7249 7250
 * @pf: board private structure
 **/
7251
static void i40e_sync_udp_filters_subtask(struct i40e_pf *pf)
7252 7253 7254
{
	struct i40e_hw *hw = &pf->hw;
	i40e_status ret;
7255
	u16 port;
7256 7257
	int i;

7258
	if (!(pf->flags & I40E_FLAG_UDP_FILTER_SYNC))
7259 7260
		return;

7261
	pf->flags &= ~I40E_FLAG_UDP_FILTER_SYNC;
7262 7263

	for (i = 0; i < I40E_MAX_PF_UDP_OFFLOAD_PORTS; i++) {
7264 7265 7266
		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
7267
			if (port)
7268 7269 7270
				ret = i40e_aq_add_udp_tunnel(hw, port,
							pf->udp_ports[i].type,
							NULL, NULL);
S
Shannon Nelson 已提交
7271 7272
			else
				ret = i40e_aq_del_udp_tunnel(hw, i, NULL);
7273 7274

			if (ret) {
7275 7276 7277 7278
				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",
7279
					port, i,
7280 7281
					i40e_stat_str(&pf->hw, ret),
					i40e_aq_str(&pf->hw,
7282
						    pf->hw.aq.asq_last_status));
7283
				pf->udp_ports[i].index = 0;
7284 7285 7286 7287 7288
			}
		}
	}
}

J
Jesse Brandeburg 已提交
7289 7290 7291 7292 7293 7294 7295 7296 7297 7298 7299
/**
 * 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;

7300 7301 7302 7303 7304
	/* don't bother with service tasks if a reset is in progress */
	if (test_bit(__I40E_RESET_RECOVERY_PENDING, &pf->state)) {
		return;
	}

M
Mitch Williams 已提交
7305 7306 7307
	if (test_and_set_bit(__I40E_SERVICE_SCHED, &pf->state))
		return;

7308
	i40e_detect_recover_hung(pf);
7309
	i40e_sync_filters_subtask(pf);
J
Jesse Brandeburg 已提交
7310 7311 7312 7313 7314
	i40e_reset_subtask(pf);
	i40e_handle_mdd_event(pf);
	i40e_vc_process_vflr_event(pf);
	i40e_watchdog_subtask(pf);
	i40e_fdir_reinit_subtask(pf);
M
Mitch Williams 已提交
7315 7316 7317 7318 7319 7320 7321 7322 7323 7324 7325 7326
	if (pf->flags & I40E_FLAG_CLIENT_RESET) {
		/* Client subtask will reopen next time through. */
		i40e_notify_client_of_netdev_close(pf->vsi[pf->lan_vsi], true);
		pf->flags &= ~I40E_FLAG_CLIENT_RESET;
	} else {
		i40e_client_subtask(pf);
		if (pf->flags & I40E_FLAG_CLIENT_L2_CHANGE) {
			i40e_notify_client_of_l2_param_changes(
							pf->vsi[pf->lan_vsi]);
			pf->flags &= ~I40E_FLAG_CLIENT_L2_CHANGE;
		}
	}
J
Jesse Brandeburg 已提交
7327
	i40e_sync_filters_subtask(pf);
7328
	i40e_sync_udp_filters_subtask(pf);
J
Jesse Brandeburg 已提交
7329 7330
	i40e_clean_adminq_subtask(pf);

M
Mitch Williams 已提交
7331 7332 7333
	/* flush memory to make sure state is correct before next watchdog */
	smp_mb__before_atomic();
	clear_bit(__I40E_SERVICE_SCHED, &pf->state);
J
Jesse Brandeburg 已提交
7334 7335 7336 7337 7338 7339 7340 7341 7342 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 7368 7369 7370 7371 7372 7373 7374 7375 7376 7377 7378 7379 7380 7381 7382

	/* 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);
7383
		vsi->num_q_vectors = pf->num_fdsb_msix;
J
Jesse Brandeburg 已提交
7384 7385 7386 7387 7388 7389 7390 7391 7392 7393 7394 7395 7396 7397 7398 7399 7400 7401 7402 7403 7404 7405 7406
		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;

	default:
		WARN_ON(1);
		return -ENODATA;
	}

	return 0;
}

A
Anjali Singhai Jain 已提交
7407 7408 7409
/**
 * i40e_vsi_alloc_arrays - Allocate queue and vector pointer arrays for the vsi
 * @type: VSI pointer
7410
 * @alloc_qvectors: a bool to specify if q_vectors need to be allocated.
A
Anjali Singhai Jain 已提交
7411 7412 7413 7414
 *
 * On error: returns error code (negative)
 * On success: returns 0
 **/
7415
static int i40e_vsi_alloc_arrays(struct i40e_vsi *vsi, bool alloc_qvectors)
A
Anjali Singhai Jain 已提交
7416 7417 7418 7419
{
	int size;
	int ret = 0;

7420
	/* allocate memory for both Tx and Rx ring pointers */
A
Anjali Singhai Jain 已提交
7421 7422 7423 7424 7425 7426
	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];

7427 7428
	if (alloc_qvectors) {
		/* allocate memory for q_vector pointers */
7429
		size = sizeof(struct i40e_q_vector *) * vsi->num_q_vectors;
7430 7431 7432 7433 7434
		vsi->q_vectors = kzalloc(size, GFP_KERNEL);
		if (!vsi->q_vectors) {
			ret = -ENOMEM;
			goto err_vectors;
		}
A
Anjali Singhai Jain 已提交
7435 7436 7437 7438 7439 7440 7441 7442
	}
	return ret;

err_vectors:
	kfree(vsi->tx_rings);
	return ret;
}

J
Jesse Brandeburg 已提交
7443 7444 7445 7446 7447 7448 7449 7450 7451 7452 7453 7454 7455 7456 7457 7458 7459 7460 7461 7462 7463 7464 7465 7466 7467
/**
 * 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
7468
	while (i < pf->num_alloc_vsi && pf->vsi[i])
J
Jesse Brandeburg 已提交
7469
		i++;
M
Mitch Williams 已提交
7470
	if (i >= pf->num_alloc_vsi) {
J
Jesse Brandeburg 已提交
7471 7472 7473 7474 7475
		i = 0;
		while (i < pf->next_vsi && pf->vsi[i])
			i++;
	}

M
Mitch Williams 已提交
7476
	if (i < pf->num_alloc_vsi && !pf->vsi[i]) {
J
Jesse Brandeburg 已提交
7477 7478 7479
		vsi_idx = i;             /* Found one! */
	} else {
		ret = -ENODEV;
7480
		goto unlock_pf;  /* out of VSI slots! */
J
Jesse Brandeburg 已提交
7481 7482 7483 7484 7485 7486
	}
	pf->next_vsi = ++i;

	vsi = kzalloc(sizeof(*vsi), GFP_KERNEL);
	if (!vsi) {
		ret = -ENOMEM;
7487
		goto unlock_pf;
J
Jesse Brandeburg 已提交
7488 7489 7490 7491 7492 7493
	}
	vsi->type = type;
	vsi->back = pf;
	set_bit(__I40E_DOWN, &vsi->state);
	vsi->flags = 0;
	vsi->idx = vsi_idx;
7494
	vsi->int_rate_limit = 0;
7495 7496
	vsi->rss_table_size = (vsi->type == I40E_VSI_MAIN) ?
				pf->rss_table_size : 64;
J
Jesse Brandeburg 已提交
7497 7498
	vsi->netdev_registered = false;
	vsi->work_limit = I40E_DEFAULT_IRQ_WORK;
7499
	hash_init(vsi->mac_filter_hash);
7500
	vsi->irqs_ready = false;
J
Jesse Brandeburg 已提交
7501

7502 7503 7504 7505
	ret = i40e_set_num_rings_in_vsi(vsi);
	if (ret)
		goto err_rings;

7506
	ret = i40e_vsi_alloc_arrays(vsi, true);
A
Anjali Singhai Jain 已提交
7507
	if (ret)
7508
		goto err_rings;
7509

J
Jesse Brandeburg 已提交
7510 7511 7512
	/* Setup default MSIX irq handler for VSI */
	i40e_vsi_setup_irqhandler(vsi, i40e_msix_clean_rings);

K
Kiran Patil 已提交
7513
	/* Initialize VSI lock */
7514
	spin_lock_init(&vsi->mac_filter_hash_lock);
J
Jesse Brandeburg 已提交
7515 7516
	pf->vsi[vsi_idx] = vsi;
	ret = vsi_idx;
7517 7518
	goto unlock_pf;

7519
err_rings:
7520 7521 7522
	pf->next_vsi = i - 1;
	kfree(vsi);
unlock_pf:
J
Jesse Brandeburg 已提交
7523 7524 7525 7526
	mutex_unlock(&pf->switch_mutex);
	return ret;
}

A
Anjali Singhai Jain 已提交
7527 7528 7529
/**
 * i40e_vsi_free_arrays - Free queue and vector pointer arrays for the VSI
 * @type: VSI pointer
7530
 * @free_qvectors: a bool to specify if q_vectors need to be freed.
A
Anjali Singhai Jain 已提交
7531 7532 7533 7534
 *
 * On error: returns error code (negative)
 * On success: returns 0
 **/
7535
static void i40e_vsi_free_arrays(struct i40e_vsi *vsi, bool free_qvectors)
A
Anjali Singhai Jain 已提交
7536 7537
{
	/* free the ring and vector containers */
7538 7539 7540 7541
	if (free_qvectors) {
		kfree(vsi->q_vectors);
		vsi->q_vectors = NULL;
	}
A
Anjali Singhai Jain 已提交
7542 7543 7544 7545 7546
	kfree(vsi->tx_rings);
	vsi->tx_rings = NULL;
	vsi->rx_rings = NULL;
}

7547 7548 7549 7550 7551 7552 7553 7554 7555 7556 7557 7558 7559 7560 7561 7562 7563
/**
 * 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
7564 7565 7566 7567 7568 7569 7570 7571 7572 7573 7574 7575 7576 7577 7578 7579 7580 7581 7582 7583 7584 7585 7586 7587 7588 7589 7590 7591 7592 7593 7594 7595
/**
 * 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;
	}

7596
	/* updates the PF for this cleared vsi */
J
Jesse Brandeburg 已提交
7597 7598 7599
	i40e_put_lump(pf->qp_pile, vsi->base_queue, vsi->idx);
	i40e_put_lump(pf->irq_pile, vsi->base_vector, vsi->idx);

7600
	i40e_vsi_free_arrays(vsi, true);
7601
	i40e_clear_rss_config_user(vsi);
7602

J
Jesse Brandeburg 已提交
7603 7604 7605 7606 7607 7608 7609 7610 7611 7612 7613 7614
	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;
}

7615 7616 7617 7618
/**
 * i40e_vsi_clear_rings - Deallocates the Rx and Tx rings for the provided VSI
 * @vsi: the VSI being cleaned
 **/
7619
static void i40e_vsi_clear_rings(struct i40e_vsi *vsi)
7620 7621 7622
{
	int i;

7623
	if (vsi->tx_rings && vsi->tx_rings[0]) {
N
Neerav Parikh 已提交
7624
		for (i = 0; i < vsi->alloc_queue_pairs; i++) {
7625 7626 7627 7628
			kfree_rcu(vsi->tx_rings[i], rcu);
			vsi->tx_rings[i] = NULL;
			vsi->rx_rings[i] = NULL;
		}
7629
	}
7630 7631
}

J
Jesse Brandeburg 已提交
7632 7633 7634 7635 7636 7637
/**
 * 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)
{
7638
	struct i40e_ring *tx_ring, *rx_ring;
J
Jesse Brandeburg 已提交
7639 7640 7641 7642
	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
7643
	for (i = 0; i < vsi->alloc_queue_pairs; i++) {
7644
		/* allocate space for both Tx and Rx in one shot */
7645 7646 7647
		tx_ring = kzalloc(sizeof(struct i40e_ring) * 2, GFP_KERNEL);
		if (!tx_ring)
			goto err_out;
J
Jesse Brandeburg 已提交
7648 7649 7650 7651 7652 7653 7654 7655 7656 7657

		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;
7658 7659
		if (vsi->back->flags & I40E_FLAG_WB_ON_ITR_CAPABLE)
			tx_ring->flags = I40E_TXR_FLAGS_WB_ON_ITR;
7660
		tx_ring->tx_itr_setting = pf->tx_itr_default;
7661
		vsi->tx_rings[i] = tx_ring;
J
Jesse Brandeburg 已提交
7662

7663
		rx_ring = &tx_ring[1];
J
Jesse Brandeburg 已提交
7664 7665 7666 7667 7668 7669 7670 7671 7672
		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;
7673
		rx_ring->rx_itr_setting = pf->rx_itr_default;
7674
		vsi->rx_rings[i] = rx_ring;
J
Jesse Brandeburg 已提交
7675 7676 7677
	}

	return 0;
7678 7679 7680 7681

err_out:
	i40e_vsi_clear_rings(vsi);
	return -ENOMEM;
J
Jesse Brandeburg 已提交
7682 7683 7684 7685 7686 7687 7688 7689 7690 7691 7692
}

/**
 * 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)
{
7693 7694 7695
	vectors = pci_enable_msix_range(pf->pdev, pf->msix_entries,
					I40E_MIN_MSIX, vectors);
	if (vectors < 0) {
J
Jesse Brandeburg 已提交
7696
		dev_info(&pf->pdev->dev,
7697
			 "MSI-X vector reservation failed: %d\n", vectors);
J
Jesse Brandeburg 已提交
7698 7699 7700 7701 7702 7703 7704 7705 7706 7707 7708 7709
		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.
 *
7710
 * Returns the number of vectors reserved or negative on failure
J
Jesse Brandeburg 已提交
7711 7712 7713 7714
 **/
static int i40e_init_msix(struct i40e_pf *pf)
{
	struct i40e_hw *hw = &pf->hw;
7715
	int cpus, extra_vectors;
S
Shannon Nelson 已提交
7716
	int vectors_left;
J
Jesse Brandeburg 已提交
7717
	int v_budget, i;
7718
	int v_actual;
7719
	int iwarp_requested = 0;
J
Jesse Brandeburg 已提交
7720 7721 7722 7723 7724 7725 7726

	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
7727 7728 7729 7730 7731
	 *	- 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
7732
	 *   - The number of VMDq pairs
7733
	 *   - The CPU count within the NUMA node if iWARP is enabled
J
Jesse Brandeburg 已提交
7734 7735 7736 7737 7738
	 * 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
7739 7740 7741 7742 7743 7744 7745 7746 7747
	vectors_left = hw->func_caps.num_msix_vectors;
	v_budget = 0;

	/* reserve one vector for miscellaneous handler */
	if (vectors_left) {
		v_budget++;
		vectors_left--;
	}

7748 7749 7750 7751 7752 7753 7754 7755 7756
	/* reserve some vectors for the main PF traffic queues. Initially we
	 * only reserve at most 50% of the available vectors, in the case that
	 * the number of online CPUs is large. This ensures that we can enable
	 * extra features as well. Once we've enabled the other features, we
	 * will use any remaining vectors to reach as close as we can to the
	 * number of online CPUs.
	 */
	cpus = num_online_cpus();
	pf->num_lan_msix = min_t(int, cpus, vectors_left / 2);
S
Shannon Nelson 已提交
7757 7758 7759 7760 7761
	vectors_left -= pf->num_lan_msix;

	/* reserve one vector for sideband flow director */
	if (pf->flags & I40E_FLAG_FD_SB_ENABLED) {
		if (vectors_left) {
7762
			pf->num_fdsb_msix = 1;
S
Shannon Nelson 已提交
7763 7764 7765
			v_budget++;
			vectors_left--;
		} else {
7766
			pf->num_fdsb_msix = 0;
S
Shannon Nelson 已提交
7767 7768
		}
	}
7769

7770 7771
	/* can we reserve enough for iWARP? */
	if (pf->flags & I40E_FLAG_IWARP_ENABLED) {
7772 7773
		iwarp_requested = pf->num_iwarp_msix;

7774 7775 7776 7777 7778 7779 7780 7781
		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
7782 7783 7784 7785 7786
	/* 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);

7787 7788 7789 7790 7791 7792 7793 7794 7795 7796 7797 7798 7799
		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
7800

7801 7802 7803
			v_budget += vmdq_vecs;
			vectors_left -= vmdq_vecs;
		}
S
Shannon Nelson 已提交
7804
	}
J
Jesse Brandeburg 已提交
7805

7806 7807 7808 7809 7810 7811 7812 7813 7814 7815 7816 7817 7818 7819 7820 7821 7822
	/* On systems with a large number of SMP cores, we previously limited
	 * the number of vectors for num_lan_msix to be at most 50% of the
	 * available vectors, to allow for other features. Now, we add back
	 * the remaining vectors. However, we ensure that the total
	 * num_lan_msix will not exceed num_online_cpus(). To do this, we
	 * calculate the number of vectors we can add without going over the
	 * cap of CPUs. For systems with a small number of CPUs this will be
	 * zero.
	 */
	extra_vectors = min_t(int, cpus - pf->num_lan_msix, vectors_left);
	pf->num_lan_msix += extra_vectors;
	vectors_left -= extra_vectors;

	WARN(vectors_left < 0,
	     "Calculation of remaining vectors underflowed. This is an accounting bug when determining total MSI-X vectors.\n");

	v_budget += pf->num_lan_msix;
J
Jesse Brandeburg 已提交
7823 7824 7825 7826 7827 7828 7829
	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;
7830
	v_actual = i40e_reserve_msix_vectors(pf, v_budget);
7831

7832
	if (v_actual < I40E_MIN_MSIX) {
J
Jesse Brandeburg 已提交
7833 7834 7835
		pf->flags &= ~I40E_FLAG_MSIX_ENABLED;
		kfree(pf->msix_entries);
		pf->msix_entries = NULL;
7836
		pci_disable_msix(pf->pdev);
J
Jesse Brandeburg 已提交
7837 7838
		return -ENODEV;

7839
	} else if (v_actual == I40E_MIN_MSIX) {
J
Jesse Brandeburg 已提交
7840 7841 7842 7843 7844 7845
		/* 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;

7846 7847 7848 7849 7850 7851
	} 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.
		 */
7852 7853
		int vec;

7854 7855
		dev_info(&pf->pdev->dev,
			 "MSI-X vector limit reached, attempting to redistribute vectors\n");
7856
		/* reserve the misc vector */
7857
		vec = v_actual - 1;
7858

J
Jesse Brandeburg 已提交
7859 7860
		/* Scale vector usage down */
		pf->num_vmdq_msix = 1;    /* force VMDqs to only one vector */
7861
		pf->num_vmdq_vsis = 1;
S
Shannon Nelson 已提交
7862
		pf->num_vmdq_qps = 1;
J
Jesse Brandeburg 已提交
7863 7864 7865 7866 7867 7868 7869

		/* partition out the remaining vectors */
		switch (vec) {
		case 2:
			pf->num_lan_msix = 1;
			break;
		case 3:
7870 7871 7872 7873 7874 7875
			if (pf->flags & I40E_FLAG_IWARP_ENABLED) {
				pf->num_lan_msix = 1;
				pf->num_iwarp_msix = 1;
			} else {
				pf->num_lan_msix = 2;
			}
J
Jesse Brandeburg 已提交
7876 7877
			break;
		default:
7878 7879 7880 7881 7882 7883 7884 7885 7886
			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);
			}
7887 7888 7889 7890
			if (pf->flags & I40E_FLAG_FD_SB_ENABLED) {
				pf->num_fdsb_msix = 1;
				vec--;
			}
7891 7892 7893
			pf->num_lan_msix = min_t(int,
			       (vec - (pf->num_iwarp_msix + pf->num_vmdq_vsis)),
							      pf->num_lan_msix);
7894
			pf->num_lan_qps = pf->num_lan_msix;
J
Jesse Brandeburg 已提交
7895 7896 7897 7898
			break;
		}
	}

7899 7900 7901 7902 7903
	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;
	}
7904 7905 7906 7907 7908
	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;
	}
7909 7910 7911 7912 7913 7914

	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;
	}
7915 7916 7917 7918 7919 7920 7921
	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);

7922
	return v_actual;
J
Jesse Brandeburg 已提交
7923 7924
}

7925
/**
7926
 * i40e_vsi_alloc_q_vector - Allocate memory for a single interrupt vector
7927 7928
 * @vsi: the VSI being configured
 * @v_idx: index of the vector in the vsi struct
7929
 * @cpu: cpu to be used on affinity_mask
7930 7931 7932
 *
 * We allocate one q_vector.  If allocation fails we return -ENOMEM.
 **/
7933
static int i40e_vsi_alloc_q_vector(struct i40e_vsi *vsi, int v_idx, int cpu)
7934 7935 7936 7937 7938 7939 7940 7941 7942 7943
{
	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;
7944 7945
	cpumask_set_cpu(cpu, &q_vector->affinity_mask);

7946 7947
	if (vsi->netdev)
		netif_napi_add(vsi->netdev, &q_vector->napi,
J
Jesse Brandeburg 已提交
7948
			       i40e_napi_poll, NAPI_POLL_WEIGHT);
7949

7950 7951 7952
	q_vector->rx.latency_range = I40E_LOW_LATENCY;
	q_vector->tx.latency_range = I40E_LOW_LATENCY;

7953 7954 7955 7956 7957 7958
	/* tie q_vector and vsi together */
	vsi->q_vectors[v_idx] = q_vector;

	return 0;
}

J
Jesse Brandeburg 已提交
7959
/**
7960
 * i40e_vsi_alloc_q_vectors - Allocate memory for interrupt vectors
J
Jesse Brandeburg 已提交
7961 7962 7963 7964 7965
 * @vsi: the VSI being configured
 *
 * We allocate one q_vector per queue interrupt.  If allocation fails we
 * return -ENOMEM.
 **/
7966
static int i40e_vsi_alloc_q_vectors(struct i40e_vsi *vsi)
J
Jesse Brandeburg 已提交
7967 7968
{
	struct i40e_pf *pf = vsi->back;
7969
	int err, v_idx, num_q_vectors, current_cpu;
J
Jesse Brandeburg 已提交
7970 7971 7972 7973 7974 7975 7976 7977 7978

	/* 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;

7979 7980
	current_cpu = cpumask_first(cpu_online_mask);

J
Jesse Brandeburg 已提交
7981
	for (v_idx = 0; v_idx < num_q_vectors; v_idx++) {
7982
		err = i40e_vsi_alloc_q_vector(vsi, v_idx, current_cpu);
7983 7984
		if (err)
			goto err_out;
7985 7986 7987
		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
7988 7989 7990
	}

	return 0;
7991 7992 7993 7994 7995 7996

err_out:
	while (v_idx--)
		i40e_free_q_vector(vsi, v_idx);

	return err;
J
Jesse Brandeburg 已提交
7997 7998 7999 8000 8001 8002
}

/**
 * i40e_init_interrupt_scheme - Determine proper interrupt scheme
 * @pf: board private structure to initialize
 **/
8003
static int i40e_init_interrupt_scheme(struct i40e_pf *pf)
J
Jesse Brandeburg 已提交
8004
{
8005 8006
	int vectors = 0;
	ssize_t size;
J
Jesse Brandeburg 已提交
8007 8008

	if (pf->flags & I40E_FLAG_MSIX_ENABLED) {
8009 8010
		vectors = i40e_init_msix(pf);
		if (vectors < 0) {
J
Jesse Brandeburg 已提交
8011
			pf->flags &= ~(I40E_FLAG_MSIX_ENABLED	|
8012
				       I40E_FLAG_IWARP_ENABLED	|
J
Jesse Brandeburg 已提交
8013
				       I40E_FLAG_RSS_ENABLED	|
8014
				       I40E_FLAG_DCB_CAPABLE	|
8015
				       I40E_FLAG_DCB_ENABLED	|
J
Jesse Brandeburg 已提交
8016 8017 8018 8019
				       I40E_FLAG_SRIOV_ENABLED	|
				       I40E_FLAG_FD_SB_ENABLED	|
				       I40E_FLAG_FD_ATR_ENABLED	|
				       I40E_FLAG_VMDQ_ENABLED);
J
Jesse Brandeburg 已提交
8020 8021 8022 8023 8024 8025 8026 8027

			/* 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
8028
		dev_info(&pf->pdev->dev, "MSI-X not available, trying MSI\n");
8029 8030 8031 8032
		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
8033 8034
			pf->flags &= ~I40E_FLAG_MSI_ENABLED;
		}
8035
		vectors = 1;  /* one MSI or Legacy vector */
J
Jesse Brandeburg 已提交
8036 8037
	}

8038
	if (!(pf->flags & (I40E_FLAG_MSIX_ENABLED | I40E_FLAG_MSI_ENABLED)))
C
Catherine Sullivan 已提交
8039
		dev_info(&pf->pdev->dev, "MSI-X and MSI not available, falling back to Legacy IRQ\n");
8040

8041 8042 8043
	/* set up vector assignment tracking */
	size = sizeof(struct i40e_lump_tracking) + (sizeof(u16) * vectors);
	pf->irq_pile = kzalloc(size, GFP_KERNEL);
8044 8045 8046 8047
	if (!pf->irq_pile) {
		dev_err(&pf->pdev->dev, "error allocating irq_pile memory\n");
		return -ENOMEM;
	}
8048 8049 8050
	pf->irq_pile->num_entries = vectors;
	pf->irq_pile->search_hint = 0;

8051
	/* track first vector for misc interrupts, ignore return */
8052
	(void)i40e_get_lump(pf, pf->irq_pile, 1, I40E_PILE_VALID_BIT - 1);
8053 8054

	return 0;
J
Jesse Brandeburg 已提交
8055 8056 8057 8058 8059 8060 8061 8062 8063 8064 8065 8066 8067 8068 8069 8070 8071 8072 8073 8074
}

/**
 * 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,
8075
				  i40e_intr, 0, pf->int_name, pf);
J
Jesse Brandeburg 已提交
8076 8077
		if (err) {
			dev_info(&pf->pdev->dev,
C
Catherine Sullivan 已提交
8078
				 "request_irq for %s failed: %d\n",
8079
				 pf->int_name, err);
J
Jesse Brandeburg 已提交
8080 8081 8082 8083
			return -EFAULT;
		}
	}

8084
	i40e_enable_misc_int_causes(pf);
J
Jesse Brandeburg 已提交
8085 8086 8087 8088 8089 8090 8091

	/* 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);

8092
	i40e_irq_dynamic_enable_icr0(pf, true);
J
Jesse Brandeburg 已提交
8093 8094 8095 8096 8097

	return err;
}

/**
8098 8099 8100 8101
 * i40e_config_rss_aq - Prepare for RSS using AQ commands
 * @vsi: vsi structure
 * @seed: RSS hash seed
 **/
8102 8103
static int i40e_config_rss_aq(struct i40e_vsi *vsi, const u8 *seed,
			      u8 *lut, u16 lut_size)
8104 8105 8106
{
	struct i40e_pf *pf = vsi->back;
	struct i40e_hw *hw = &pf->hw;
8107
	int ret = 0;
8108

8109 8110 8111 8112 8113 8114 8115 8116 8117 8118 8119
	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;
		}
8120
	}
8121 8122
	if (lut) {
		bool pf_lut = vsi->type == I40E_VSI_MAIN ? true : false;
8123

8124 8125 8126 8127 8128 8129 8130 8131 8132
		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;
		}
	}
8133 8134 8135
	return ret;
}

8136 8137 8138 8139 8140 8141 8142 8143 8144 8145 8146 8147 8148 8149 8150 8151 8152 8153 8154 8155 8156 8157 8158 8159 8160 8161 8162 8163 8164 8165 8166 8167 8168 8169 8170 8171 8172 8173 8174 8175 8176 8177 8178 8179 8180 8181
/**
 * 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;
}

8182 8183 8184 8185 8186 8187 8188 8189 8190 8191 8192 8193 8194 8195
/**
 * 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;

8196 8197 8198 8199 8200 8201
	if (!vsi->rss_size)
		vsi->rss_size = min_t(int, pf->alloc_rss_size,
				      vsi->num_queue_pairs);
	if (!vsi->rss_size)
		return -EINVAL;

8202 8203 8204
	lut = kzalloc(vsi->rss_table_size, GFP_KERNEL);
	if (!lut)
		return -ENOMEM;
8205 8206 8207 8208 8209 8210 8211 8212 8213 8214 8215
	/* 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);
8216 8217 8218 8219 8220 8221
	ret = i40e_config_rss_aq(vsi, seed, lut, vsi->rss_table_size);
	kfree(lut);

	return ret;
}

8222
/**
8223
 * i40e_config_rss_reg - Configure RSS keys and lut by writing registers
8224
 * @vsi: Pointer to vsi structure
8225
 * @seed: RSS hash seed
8226 8227 8228 8229
 * @lut: Lookup table
 * @lut_size: Lookup table size
 *
 * Returns 0 on success, negative on failure
J
Jesse Brandeburg 已提交
8230
 **/
8231 8232
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
8233
{
8234
	struct i40e_pf *pf = vsi->back;
8235
	struct i40e_hw *hw = &pf->hw;
8236
	u16 vf_id = vsi->vf_id;
8237
	u8 i;
J
Jesse Brandeburg 已提交
8238

8239
	/* Fill out hash function seed */
8240 8241 8242
	if (seed) {
		u32 *seed_dw = (u32 *)seed;

8243 8244
		if (vsi->type == I40E_VSI_MAIN) {
			for (i = 0; i <= I40E_PFQF_HKEY_MAX_INDEX; i++)
8245
				wr32(hw, I40E_PFQF_HKEY(i), seed_dw[i]);
8246 8247
		} else if (vsi->type == I40E_VSI_SRIOV) {
			for (i = 0; i <= I40E_VFQF_HKEY1_MAX_INDEX; i++)
8248
				wr32(hw, I40E_VFQF_HKEY1(i, vf_id), seed_dw[i]);
8249 8250 8251
		} else {
			dev_err(&pf->pdev->dev, "Cannot set RSS seed - invalid VSI type\n");
		}
8252 8253 8254 8255 8256
	}

	if (lut) {
		u32 *lut_dw = (u32 *)lut;

8257 8258 8259 8260 8261 8262 8263 8264 8265
		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++)
8266
				wr32(hw, I40E_VFQF_HLUT1(i, vf_id), lut_dw[i]);
8267 8268 8269
		} else {
			dev_err(&pf->pdev->dev, "Cannot set RSS LUT - invalid VSI type\n");
		}
8270 8271 8272 8273 8274 8275
	}
	i40e_flush(hw);

	return 0;
}

8276 8277 8278 8279 8280 8281 8282 8283 8284 8285 8286 8287 8288 8289 8290 8291 8292 8293 8294 8295
/**
 * 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++)
8296
			seed_dw[i] = i40e_read_rx_ctl(hw, I40E_PFQF_HKEY(i));
8297 8298 8299 8300 8301 8302 8303 8304 8305 8306 8307 8308 8309 8310 8311 8312 8313 8314 8315 8316 8317 8318 8319 8320 8321 8322 8323 8324 8325 8326 8327 8328 8329 8330 8331 8332 8333 8334 8335 8336 8337 8338 8339
	}
	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)
{
8340 8341 8342 8343 8344 8345
	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);
8346 8347
}

8348 8349 8350 8351 8352 8353 8354
/**
 * 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
 */
8355 8356
void i40e_fill_rss_lut(struct i40e_pf *pf, u8 *lut,
		       u16 rss_table_size, u16 rss_size)
8357 8358 8359 8360 8361 8362 8363
{
	u16 i;

	for (i = 0; i < rss_table_size; i++)
		lut[i] = i % rss_size;
}

8364
/**
8365
 * i40e_pf_config_rss - Prepare for RSS if used
8366 8367
 * @pf: board private structure
 **/
8368
static int i40e_pf_config_rss(struct i40e_pf *pf)
8369 8370 8371
{
	struct i40e_vsi *vsi = pf->vsi[pf->lan_vsi];
	u8 seed[I40E_HKEY_ARRAY_SIZE];
8372
	u8 *lut;
8373 8374 8375
	struct i40e_hw *hw = &pf->hw;
	u32 reg_val;
	u64 hena;
8376
	int ret;
8377

J
Jesse Brandeburg 已提交
8378
	/* By default we enable TCP/UDP with IPv4/IPv6 ptypes */
8379 8380
	hena = (u64)i40e_read_rx_ctl(hw, I40E_PFQF_HENA(0)) |
		((u64)i40e_read_rx_ctl(hw, I40E_PFQF_HENA(1)) << 32);
8381 8382
	hena |= i40e_pf_get_default_rss_hena(pf);

8383 8384
	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
8385

8386
	/* Determine the RSS table size based on the hardware capabilities */
8387
	reg_val = i40e_read_rx_ctl(hw, I40E_PFQF_CTL_0);
8388 8389 8390
	reg_val = (pf->rss_table_size == 512) ?
			(reg_val | I40E_PFQF_CTL_0_HASHLUTSIZE_512) :
			(reg_val & ~I40E_PFQF_CTL_0_HASHLUTSIZE_512);
8391
	i40e_write_rx_ctl(hw, I40E_PFQF_CTL_0, reg_val);
8392

8393
	/* Determine the RSS size of the VSI */
8394 8395 8396 8397 8398 8399
	if (!vsi->rss_size) {
		u16 qcount;

		qcount = vsi->num_queue_pairs / vsi->tc_config.numtc;
		vsi->rss_size = min_t(int, pf->alloc_rss_size, qcount);
	}
8400 8401
	if (!vsi->rss_size)
		return -EINVAL;
8402

8403 8404 8405 8406
	lut = kzalloc(vsi->rss_table_size, GFP_KERNEL);
	if (!lut)
		return -ENOMEM;

8407 8408 8409 8410 8411
	/* 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);
8412

8413 8414 8415 8416 8417 8418 8419
	/* 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);
8420
	ret = i40e_config_rss(vsi, seed, lut, vsi->rss_table_size);
8421 8422 8423
	kfree(lut);

	return ret;
J
Jesse Brandeburg 已提交
8424 8425
}

8426 8427 8428 8429 8430 8431 8432 8433 8434 8435
/**
 * 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)
{
8436 8437 8438
	struct i40e_vsi *vsi = pf->vsi[pf->lan_vsi];
	int new_rss_size;

8439 8440 8441
	if (!(pf->flags & I40E_FLAG_RSS_ENABLED))
		return 0;

8442
	new_rss_size = min_t(int, queue_count, pf->rss_size_max);
8443

8444
	if (queue_count != vsi->num_queue_pairs) {
8445 8446
		u16 qcount;

8447
		vsi->req_queue_pairs = queue_count;
8448 8449
		i40e_prep_for_reset(pf);

8450
		pf->alloc_rss_size = new_rss_size;
8451 8452

		i40e_reset_and_rebuild(pf, true);
8453 8454 8455 8456 8457 8458 8459 8460 8461 8462 8463

		/* 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 */
8464 8465
		qcount = vsi->num_queue_pairs / vsi->tc_config.numtc;
		vsi->rss_size = min_t(int, pf->alloc_rss_size, qcount);
8466

8467
		i40e_pf_config_rss(pf);
8468
	}
8469 8470
	dev_info(&pf->pdev->dev, "User requested queue count/HW max RSS count:  %d/%d\n",
		 vsi->req_queue_pairs, pf->rss_size_max);
8471
	return pf->alloc_rss_size;
8472 8473
}

8474 8475 8476 8477 8478 8479 8480 8481 8482 8483 8484 8485 8486 8487 8488 8489 8490 8491 8492 8493 8494 8495 8496 8497 8498 8499 8500 8501 8502 8503 8504 8505
/**
 * 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;

8506
	/* Set the valid bit for this PF */
8507
	bw_data.pf_valid_bits = cpu_to_le16(BIT(pf->hw.pf_id));
8508 8509 8510 8511 8512 8513 8514 8515 8516 8517 8518 8519 8520 8521 8522 8523 8524 8525 8526 8527 8528 8529 8530 8531 8532 8533 8534 8535 8536 8537 8538 8539 8540
	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,
8541 8542 8543
			 "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));
8544 8545 8546 8547 8548 8549 8550 8551 8552 8553 8554 8555 8556 8557
		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) {
8558 8559 8560
		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));
8561 8562 8563 8564 8565 8566 8567 8568 8569 8570 8571
		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,
8572 8573 8574
			 "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));
8575 8576 8577 8578 8579 8580 8581 8582 8583 8584 8585 8586 8587 8588 8589 8590 8591
		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,
8592 8593 8594
			 "BW settings NOT SAVED, err %s aq_err %s\n",
			 i40e_stat_str(&pf->hw, ret),
			 i40e_aq_str(&pf->hw, last_aq_status));
8595 8596 8597 8598 8599
bw_commit_out:

	return ret;
}

J
Jesse Brandeburg 已提交
8600 8601 8602 8603 8604 8605 8606 8607 8608 8609 8610 8611 8612 8613 8614 8615
/**
 * 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     |
8616 8617
		    I40E_FLAG_MSIX_ENABLED;

8618 8619 8620 8621
	/* 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;

8622 8623 8624
	/* Depending on PF configurations, it is possible that the RSS
	 * maximum might end up larger than the available queues
	 */
8625
	pf->rss_size_max = BIT(pf->hw.func_caps.rss_table_entry_width);
8626
	pf->alloc_rss_size = 1;
8627
	pf->rss_table_size = pf->hw.func_caps.rss_table_size;
8628 8629
	pf->rss_size_max = min_t(int, pf->rss_size_max,
				 pf->hw.func_caps.num_tx_qp);
J
Jesse Brandeburg 已提交
8630 8631
	if (pf->hw.func_caps.rss) {
		pf->flags |= I40E_FLAG_RSS_ENABLED;
8632 8633
		pf->alloc_rss_size = min_t(int, pf->rss_size_max,
					   num_online_cpus());
J
Jesse Brandeburg 已提交
8634 8635
	}

8636
	/* MFP mode enabled */
8637
	if (pf->hw.func_caps.npar_enable || pf->hw.func_caps.flex10_enable) {
8638 8639
		pf->flags |= I40E_FLAG_MFP_ENABLED;
		dev_info(&pf->pdev->dev, "MFP mode Enabled\n");
8640 8641 8642 8643 8644 8645 8646
		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);
8647 8648
	}

8649 8650 8651 8652 8653
	/* 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;
8654 8655
		if (pf->flags & I40E_FLAG_MFP_ENABLED &&
		    pf->hw.num_partitions > 1)
8656
			dev_info(&pf->pdev->dev,
A
Anjali Singhai Jain 已提交
8657
				 "Flow Director Sideband mode Disabled in MFP mode\n");
8658 8659
		else
			pf->flags |= I40E_FLAG_FD_SB_ENABLED;
8660 8661 8662 8663
		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
8664 8665
	}

B
Bimmy Pujari 已提交
8666
	if ((pf->hw.mac.type == I40E_MAC_XL710) &&
8667
	    (((pf->hw.aq.fw_maj_ver == 4) && (pf->hw.aq.fw_min_ver < 33)) ||
8668
	    (pf->hw.aq.fw_maj_ver < 4))) {
8669
		pf->flags |= I40E_FLAG_RESTART_AUTONEG;
8670 8671 8672 8673 8674
		/* No DCB support  for FW < v4.33 */
		pf->flags |= I40E_FLAG_NO_DCB_SUPPORT;
	}

	/* Disable FW LLDP if FW < v4.3 */
B
Bimmy Pujari 已提交
8675
	if ((pf->hw.mac.type == I40E_MAC_XL710) &&
8676 8677 8678 8679 8680
	    (((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 */
B
Bimmy Pujari 已提交
8681
	if ((pf->hw.mac.type == I40E_MAC_XL710) &&
8682 8683 8684
	    (((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;
8685

J
Jesse Brandeburg 已提交
8686 8687
	if (pf->hw.func_caps.vmdq) {
		pf->num_vmdq_vsis = I40E_DEFAULT_NUM_VMDQ_VSI;
8688
		pf->flags |= I40E_FLAG_VMDQ_ENABLED;
J
Jesse Brandeburg 已提交
8689
		pf->num_vmdq_qps = i40e_default_queues_per_vmdq(pf);
J
Jesse Brandeburg 已提交
8690 8691
	}

8692 8693 8694 8695 8696 8697
	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;
	}

J
Jesse Brandeburg 已提交
8698
#ifdef CONFIG_PCI_IOV
8699
	if (pf->hw.func_caps.num_vfs && pf->hw.partition_id == 1) {
J
Jesse Brandeburg 已提交
8700 8701 8702 8703 8704 8705 8706
		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 */
8707
	if (pf->hw.mac.type == I40E_MAC_X722) {
8708 8709 8710 8711 8712 8713 8714 8715 8716 8717 8718
		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
			     | I40E_FLAG_MULTIPLE_TCP_UDP_RSS_PCTYPE
			     | I40E_FLAG_NO_PCI_LINK_CHECK
			     | I40E_FLAG_USE_SET_LLDP_MIB
			     | I40E_FLAG_GENEVE_OFFLOAD_CAPABLE
			     | I40E_FLAG_PTP_L4_CAPABLE
			     | I40E_FLAG_WOL_MC_MAGIC_PKT_WAKE;
8719 8720 8721 8722 8723
	} 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;
8724
		pf->hw_disabled_flags = I40E_FLAG_HW_ATR_EVICT_CAPABLE;
8725
	} else {
8726
		pf->hw_disabled_flags = I40E_FLAG_HW_ATR_EVICT_CAPABLE;
8727
	}
8728

J
Jesse Brandeburg 已提交
8729 8730 8731 8732
	pf->eeprom_version = 0xDEAD;
	pf->lan_veb = I40E_NO_VEB;
	pf->lan_vsi = I40E_NO_VSI;

8733 8734 8735
	/* By default FW has this off for performance reasons */
	pf->flags &= ~I40E_FLAG_VEB_STATS_ENABLED;

J
Jesse Brandeburg 已提交
8736 8737 8738 8739 8740 8741 8742 8743 8744 8745 8746
	/* 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;

8747 8748
	pf->tx_timeout_recovery_level = 1;

J
Jesse Brandeburg 已提交
8749 8750
	mutex_init(&pf->switch_mutex);

G
Greg Rose 已提交
8751 8752 8753 8754
	/* 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
8755 8756 8757 8758
sw_init_done:
	return err;
}

8759 8760 8761 8762 8763 8764 8765 8766 8767 8768 8769 8770 8771 8772 8773 8774 8775 8776
/**
 * 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;
8777 8778 8779
		/* enable FD_SB only if there is MSI-X vector */
		if (pf->num_fdsb_msix > 0)
			pf->flags |= I40E_FLAG_FD_SB_ENABLED;
8780 8781 8782 8783 8784 8785 8786
	} 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;
8787
		pf->hw_disabled_flags &= ~I40E_FLAG_FD_SB_ENABLED;
8788
		/* reset fd counters */
8789 8790
		pf->fd_add_err = 0;
		pf->fd_atr_cnt = 0;
8791 8792
		/* if ATR was auto disabled it can be re-enabled. */
		if ((pf->flags & I40E_FLAG_FD_ATR_ENABLED) &&
8793 8794
		    (pf->hw_disabled_flags & I40E_FLAG_FD_ATR_ENABLED)) {
			pf->hw_disabled_flags &= ~I40E_FLAG_FD_ATR_ENABLED;
8795 8796 8797
			if (I40E_DEBUG_FD & pf->hw.debug_mask)
				dev_info(&pf->pdev->dev, "ATR re-enabled.\n");
		}
8798 8799 8800 8801
	}
	return need_reset;
}

8802 8803 8804 8805 8806 8807 8808 8809 8810 8811 8812 8813 8814 8815 8816 8817 8818 8819 8820 8821 8822 8823
/**
 * 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
8824 8825 8826 8827 8828 8829 8830 8831 8832 8833
/**
 * 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;
8834 8835
	struct i40e_pf *pf = vsi->back;
	bool need_reset;
J
Jesse Brandeburg 已提交
8836

8837 8838 8839 8840 8841 8842
	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
8843 8844 8845 8846 8847
	if (features & NETIF_F_HW_VLAN_CTAG_RX)
		i40e_vlan_stripping_enable(vsi);
	else
		i40e_vlan_stripping_disable(vsi);

8848 8849 8850
	need_reset = i40e_set_ntuple(pf, features);

	if (need_reset)
8851
		i40e_do_reset(pf, BIT_ULL(__I40E_PF_RESET_REQUESTED));
8852

J
Jesse Brandeburg 已提交
8853 8854 8855
	return 0;
}

8856
/**
8857
 * i40e_get_udp_port_idx - Lookup a possibly offloaded for Rx UDP port
8858 8859 8860 8861 8862
 * @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
 **/
8863
static u8 i40e_get_udp_port_idx(struct i40e_pf *pf, u16 port)
8864 8865 8866 8867
{
	u8 i;

	for (i = 0; i < I40E_MAX_PF_UDP_OFFLOAD_PORTS; i++) {
8868
		if (pf->udp_ports[i].index == port)
8869 8870 8871 8872 8873 8874 8875
			return i;
	}

	return i;
}

/**
8876
 * i40e_udp_tunnel_add - Get notifications about UDP tunnel ports that come up
8877
 * @netdev: This physical port's netdev
8878
 * @ti: Tunnel endpoint information
8879
 **/
8880 8881
static void i40e_udp_tunnel_add(struct net_device *netdev,
				struct udp_tunnel_info *ti)
8882 8883 8884 8885
{
	struct i40e_netdev_priv *np = netdev_priv(netdev);
	struct i40e_vsi *vsi = np->vsi;
	struct i40e_pf *pf = vsi->back;
8886
	u16 port = ntohs(ti->port);
8887 8888 8889
	u8 next_idx;
	u8 idx;

8890
	idx = i40e_get_udp_port_idx(pf, port);
8891 8892 8893

	/* Check if port already exists */
	if (idx < I40E_MAX_PF_UDP_OFFLOAD_PORTS) {
8894
		netdev_info(netdev, "port %d already offloaded\n", port);
8895 8896 8897 8898
		return;
	}

	/* Now check if there is space to add the new port */
8899
	next_idx = i40e_get_udp_port_idx(pf, 0);
8900 8901

	if (next_idx == I40E_MAX_PF_UDP_OFFLOAD_PORTS) {
8902
		netdev_info(netdev, "maximum number of offloaded UDP ports reached, not adding port %d\n",
8903
			    port);
8904 8905 8906
		return;
	}

8907 8908 8909 8910 8911 8912 8913 8914 8915 8916
	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:
8917 8918 8919 8920 8921 8922 8923
		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;
8924 8925
}

8926
/**
8927
 * i40e_udp_tunnel_del - Get notifications about UDP tunnel ports that go away
8928
 * @netdev: This physical port's netdev
8929
 * @ti: Tunnel endpoint information
8930
 **/
8931 8932
static void i40e_udp_tunnel_del(struct net_device *netdev,
				struct udp_tunnel_info *ti)
8933 8934 8935 8936
{
	struct i40e_netdev_priv *np = netdev_priv(netdev);
	struct i40e_vsi *vsi = np->vsi;
	struct i40e_pf *pf = vsi->back;
8937
	u16 port = ntohs(ti->port);
8938 8939 8940 8941 8942
	u8 idx;

	idx = i40e_get_udp_port_idx(pf, port);

	/* Check if port already exists */
8943 8944
	if (idx >= I40E_MAX_PF_UDP_OFFLOAD_PORTS)
		goto not_found;
8945

8946 8947 8948 8949 8950 8951 8952 8953 8954 8955 8956
	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;
8957
	}
8958 8959 8960 8961 8962 8963 8964 8965 8966 8967 8968

	/* 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",
8969
		    port);
8970 8971
}

8972
static int i40e_get_phys_port_id(struct net_device *netdev,
8973
				 struct netdev_phys_item_id *ppid)
8974 8975 8976 8977 8978 8979 8980 8981 8982 8983 8984 8985 8986 8987
{
	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;
}

8988 8989 8990 8991 8992 8993 8994 8995
/**
 * 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
 */
8996 8997
static int i40e_ndo_fdb_add(struct ndmsg *ndm, struct nlattr *tb[],
			    struct net_device *dev,
8998
			    const unsigned char *addr, u16 vid,
8999 9000 9001 9002 9003 9004 9005 9006 9007
			    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;

9008 9009 9010 9011 9012
	if (vid) {
		pr_info("%s: vlans aren't supported yet for dev_uc|mc_add()\n", dev->name);
		return -EINVAL;
	}

9013 9014 9015 9016 9017 9018 9019 9020 9021 9022 9023 9024 9025 9026 9027 9028 9029 9030 9031 9032 9033 9034
	/* 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;
}

9035 9036 9037 9038 9039 9040 9041 9042 9043 9044 9045 9046 9047
/**
 * 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,
9048 9049
				   struct nlmsghdr *nlh,
				   u16 flags)
9050 9051 9052 9053 9054 9055 9056 9057 9058 9059 9060 9061 9062 9063 9064 9065 9066 9067 9068 9069 9070 9071 9072 9073 9074 9075 9076 9077 9078 9079 9080 9081 9082 9083 9084 9085 9086 9087 9088 9089 9090 9091 9092 9093 9094 9095
{
	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;
9096 9097 9098 9099 9100 9101
			/* 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));
9102 9103 9104 9105 9106 9107 9108 9109 9110 9111 9112 9113 9114 9115
			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
9116
 * @nlflags: netlink flags passed in
9117 9118 9119 9120 9121 9122
 *
 * 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,
9123 9124
				   u32 __always_unused filter_mask,
				   int nlflags)
9125 9126 9127 9128 9129 9130 9131 9132 9133 9134 9135 9136 9137 9138 9139 9140 9141 9142 9143 9144
{
	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;

9145
	return ndo_dflt_bridge_getlink(skb, pid, seq, dev, veb->bridge_mode,
9146
				       0, 0, nlflags, filter_mask, NULL);
9147 9148
}

9149 9150 9151
/**
 * i40e_features_check - Validate encapsulated packet conforms to limits
 * @skb: skb buff
J
Jean Sacren 已提交
9152
 * @dev: This physical port's netdev
9153 9154 9155 9156 9157 9158
 * @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)
{
9159 9160 9161 9162 9163 9164 9165 9166 9167 9168 9169 9170 9171 9172 9173 9174 9175 9176 9177 9178 9179 9180 9181 9182 9183 9184 9185 9186 9187 9188 9189 9190 9191 9192 9193 9194 9195 9196 9197 9198 9199 9200
	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
	 */
9201 9202

	return features;
9203 9204
out_err:
	return features & ~(NETIF_F_CSUM_MASK | NETIF_F_GSO_MASK);
9205 9206
}

9207
static const struct net_device_ops i40e_netdev_ops = {
J
Jesse Brandeburg 已提交
9208 9209 9210 9211 9212 9213 9214 9215
	.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
9216
	.ndo_do_ioctl		= i40e_ioctl,
J
Jesse Brandeburg 已提交
9217 9218 9219 9220 9221 9222
	.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
9223
	.ndo_setup_tc		= __i40e_setup_tc,
J
Jesse Brandeburg 已提交
9224 9225 9226
	.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,
9227
	.ndo_set_vf_rate	= i40e_ndo_set_vf_bw,
J
Jesse Brandeburg 已提交
9228
	.ndo_get_vf_config	= i40e_ndo_get_vf_config,
M
Mitch Williams 已提交
9229
	.ndo_set_vf_link_state	= i40e_ndo_set_vf_link_state,
9230
	.ndo_set_vf_spoofchk	= i40e_ndo_set_vf_spoofchk,
9231
	.ndo_set_vf_trust	= i40e_ndo_set_vf_trust,
9232 9233
	.ndo_udp_tunnel_add	= i40e_udp_tunnel_add,
	.ndo_udp_tunnel_del	= i40e_udp_tunnel_del,
9234
	.ndo_get_phys_port_id	= i40e_get_phys_port_id,
9235
	.ndo_fdb_add		= i40e_ndo_fdb_add,
9236
	.ndo_features_check	= i40e_features_check,
9237 9238
	.ndo_bridge_getlink	= i40e_ndo_bridge_getlink,
	.ndo_bridge_setlink	= i40e_ndo_bridge_setlink,
J
Jesse Brandeburg 已提交
9239 9240 9241 9242 9243 9244 9245 9246 9247 9248 9249 9250 9251 9252
};

/**
 * 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;
9253
	u8 broadcast[ETH_ALEN];
J
Jesse Brandeburg 已提交
9254 9255
	u8 mac_addr[ETH_ALEN];
	int etherdev_size;
9256 9257
	netdev_features_t hw_enc_features;
	netdev_features_t hw_features;
J
Jesse Brandeburg 已提交
9258 9259

	etherdev_size = sizeof(struct i40e_netdev_priv);
9260
	netdev = alloc_etherdev_mq(etherdev_size, vsi->alloc_queue_pairs);
J
Jesse Brandeburg 已提交
9261 9262 9263 9264 9265 9266 9267
	if (!netdev)
		return -ENOMEM;

	vsi->netdev = netdev;
	np = netdev_priv(netdev);
	np->vsi = vsi;

9268 9269 9270 9271 9272 9273 9274 9275 9276 9277 9278 9279 9280 9281 9282 9283 9284
	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		|
			  NETIF_F_GSO_GRE_CSUM		|
			  NETIF_F_GSO_PARTIAL		|
			  NETIF_F_GSO_UDP_TUNNEL	|
			  NETIF_F_GSO_UDP_TUNNEL_CSUM	|
			  NETIF_F_SCTP_CRC		|
			  NETIF_F_RXHASH		|
			  NETIF_F_RXCSUM		|
			  0;
J
Jesse Brandeburg 已提交
9285

9286
	if (!(pf->flags & I40E_FLAG_OUTER_UDP_CSUM_CAPABLE))
9287 9288 9289
		netdev->gso_partial_features |= NETIF_F_GSO_UDP_TUNNEL_CSUM;

	netdev->gso_partial_features |= NETIF_F_GSO_GRE_CSUM;
9290

9291 9292
	netdev->hw_enc_features |= hw_enc_features;

9293
	/* record features VLANs can make use of */
9294
	netdev->vlan_features |= hw_enc_features | NETIF_F_TSO_MANGLEID;
J
Jesse Brandeburg 已提交
9295

9296
	if (!(pf->flags & I40E_FLAG_MFP_ENABLED))
9297
		netdev->hw_features |= NETIF_F_NTUPLE;
9298 9299 9300
	hw_features = hw_enc_features		|
		      NETIF_F_HW_VLAN_CTAG_TX	|
		      NETIF_F_HW_VLAN_CTAG_RX;
9301

9302
	netdev->hw_features |= hw_features;
9303

9304
	netdev->features |= hw_features | NETIF_F_HW_VLAN_CTAG_FILTER;
9305
	netdev->hw_enc_features |= NETIF_F_TSO_MANGLEID;
J
Jesse Brandeburg 已提交
9306 9307 9308

	if (vsi->type == I40E_VSI_MAIN) {
		SET_NETDEV_DEV(netdev, &pf->pdev->dev);
9309
		ether_addr_copy(mac_addr, hw->mac.perm_addr);
9310 9311 9312 9313
		/* The following steps are necessary to properly keep track of
		 * MAC-VLAN filters loaded into firmware - first we remove
		 * filter that is automatically generated by firmware and then
		 * add new filter both to the driver hash table and firmware.
9314 9315
		 */
		i40e_rm_default_mac_filter(vsi, mac_addr);
9316
		spin_lock_bh(&vsi->mac_filter_hash_lock);
9317
		i40e_add_mac_filter(vsi, mac_addr);
9318
		spin_unlock_bh(&vsi->mac_filter_hash_lock);
J
Jesse Brandeburg 已提交
9319 9320 9321 9322 9323
	} 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
9324

9325
		spin_lock_bh(&vsi->mac_filter_hash_lock);
9326
		i40e_add_mac_filter(vsi, mac_addr);
9327
		spin_unlock_bh(&vsi->mac_filter_hash_lock);
J
Jesse Brandeburg 已提交
9328
	}
K
Kiran Patil 已提交
9329

9330 9331 9332 9333 9334 9335 9336 9337 9338 9339 9340 9341 9342 9343 9344
	/* Add the broadcast filter so that we initially will receive
	 * broadcast packets. Note that when a new VLAN is first added the
	 * driver will convert all filters marked I40E_VLAN_ANY into VLAN
	 * specific filters as part of transitioning into "vlan" operation.
	 * When more VLANs are added, the driver will copy each existing MAC
	 * filter and add it for the new VLAN.
	 *
	 * Broadcast filters are handled specially by
	 * i40e_sync_filters_subtask, as the driver must to set the broadcast
	 * promiscuous bit instead of adding this directly as a MAC/VLAN
	 * filter. The subtask will update the correct broadcast promiscuous
	 * bits as VLANs become active or inactive.
	 */
	eth_broadcast_addr(broadcast);
	spin_lock_bh(&vsi->mac_filter_hash_lock);
9345
	i40e_add_mac_filter(vsi, broadcast);
9346 9347
	spin_unlock_bh(&vsi->mac_filter_hash_lock);

9348 9349
	ether_addr_copy(netdev->dev_addr, mac_addr);
	ether_addr_copy(netdev->perm_addr, mac_addr);
9350

J
Jesse Brandeburg 已提交
9351 9352 9353 9354 9355 9356 9357 9358 9359
	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);

9360 9361 9362 9363 9364
	/* 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
9365 9366 9367 9368 9369 9370 9371 9372 9373 9374 9375 9376 9377 9378 9379 9380 9381 9382
	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);
}

9383 9384 9385 9386 9387 9388 9389 9390 9391 9392 9393 9394 9395 9396 9397 9398
/**
 * 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];
9399 9400 9401 9402 9403 9404
	if (!veb) {
		dev_info(&pf->pdev->dev,
			 "There is no veb associated with the bridge\n");
		return -ENOENT;
	}

9405
	/* Uplink is a bridge in VEPA mode */
9406
	if (veb->bridge_mode & BRIDGE_MODE_VEPA) {
9407
		return 0;
9408 9409 9410 9411
	} else {
		/* Uplink is a bridge in VEB mode */
		return 1;
	}
9412

9413 9414
	/* VEPA is now default bridge, so return 0 */
	return 0;
9415 9416
}

J
Jesse Brandeburg 已提交
9417 9418 9419 9420 9421 9422 9423 9424 9425 9426 9427 9428 9429
/**
 * 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;
	struct i40e_pf *pf = vsi->back;
	struct i40e_hw *hw = &pf->hw;
	struct i40e_vsi_context ctxt;
9430 9431 9432
	struct i40e_mac_filter *f;
	struct hlist_node *h;
	int bkt;
K
Kiran Patil 已提交
9433

J
Jesse Brandeburg 已提交
9434 9435 9436 9437 9438 9439 9440 9441 9442 9443 9444 9445 9446 9447 9448 9449 9450 9451
	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,
9452 9453 9454 9455
				 "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
9456 9457
			return -ENOENT;
		}
9458
		vsi->info = ctxt.info;
J
Jesse Brandeburg 已提交
9459 9460 9461 9462 9463 9464 9465 9466
		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 */
9467 9468
		if ((pf->flags & I40E_FLAG_MFP_ENABLED) &&
		    !(pf->hw.func_caps.iscsi)) { /* NIC type PF */
J
Jesse Brandeburg 已提交
9469 9470 9471 9472 9473 9474 9475 9476
			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,
9477 9478 9479 9480
					 "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
9481 9482 9483 9484 9485 9486 9487 9488 9489 9490
				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.
9491 9492
			 * For MFP case the iSCSI PF would use this
			 * flow to enable LAN+iSCSI TC.
J
Jesse Brandeburg 已提交
9493 9494 9495 9496
			 */
			ret = i40e_vsi_config_tc(vsi, enabled_tc);
			if (ret) {
				dev_info(&pf->pdev->dev,
9497 9498 9499 9500 9501
					 "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
9502 9503 9504 9505 9506 9507
				ret = -ENOENT;
			}
		}
		break;

	case I40E_VSI_FDIR:
9508 9509 9510
		ctxt.pf_num = hw->pf_id;
		ctxt.vf_num = 0;
		ctxt.uplink_seid = vsi->uplink_seid;
9511
		ctxt.connection_type = I40E_AQ_VSI_CONN_TYPE_NORMAL;
9512
		ctxt.flags = I40E_AQ_VSI_TYPE_PF;
9513 9514
		if ((pf->flags & I40E_FLAG_VEB_MODE_ENABLED) &&
		    (i40e_is_vsi_uplink_mode_veb(vsi))) {
9515
			ctxt.info.valid_sections |=
9516
			     cpu_to_le16(I40E_AQ_VSI_PROP_SWITCH_VALID);
9517
			ctxt.info.switch_id =
9518
			   cpu_to_le16(I40E_AQ_VSI_SW_ID_FLAG_ALLOW_LB);
9519
		}
J
Jesse Brandeburg 已提交
9520 9521 9522 9523 9524 9525 9526
		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;
9527
		ctxt.connection_type = I40E_AQ_VSI_CONN_TYPE_NORMAL;
J
Jesse Brandeburg 已提交
9528 9529 9530 9531 9532
		ctxt.flags = I40E_AQ_VSI_TYPE_VMDQ2;

		/* This VSI is connected to VEB so the switch_id
		 * should be set to zero by default.
		 */
9533 9534 9535 9536 9537 9538
		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
9539 9540 9541 9542 9543 9544 9545 9546 9547

		/* 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;
9548
		ctxt.connection_type = I40E_AQ_VSI_CONN_TYPE_NORMAL;
J
Jesse Brandeburg 已提交
9549 9550 9551 9552 9553
		ctxt.flags = I40E_AQ_VSI_TYPE_VF;

		/* This VSI is connected to VEB so the switch_id
		 * should be set to zero by default.
		 */
9554 9555 9556 9557 9558 9559
		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
9560

9561 9562 9563 9564
		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 |=
9565 9566
				(I40E_AQ_VSI_QUE_OPT_TCP_ENA |
				 I40E_AQ_VSI_QUE_OPT_RSS_LUT_VSI);
9567 9568
		}

J
Jesse Brandeburg 已提交
9569 9570
		ctxt.info.valid_sections |= cpu_to_le16(I40E_AQ_VSI_PROP_VLAN_VALID);
		ctxt.info.port_vlan_flags |= I40E_AQ_VSI_PVLAN_MODE_ALL;
9571 9572 9573 9574 9575 9576 9577
		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
9578 9579 9580 9581
		/* Setup the VSI tx/rx queue map for TC0 only for now */
		i40e_vsi_setup_queue_map(vsi, &ctxt, enabled_tc, true);
		break;

9582 9583 9584 9585
	case I40E_VSI_IWARP:
		/* send down message to iWARP */
		break;

J
Jesse Brandeburg 已提交
9586 9587 9588 9589 9590 9591 9592 9593
	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,
9594 9595 9596 9597
				 "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
9598 9599 9600
			ret = -ENOENT;
			goto err;
		}
9601
		vsi->info = ctxt.info;
J
Jesse Brandeburg 已提交
9602 9603 9604 9605 9606
		vsi->info.valid_sections = 0;
		vsi->seid = ctxt.seid;
		vsi->id = ctxt.vsi_number;
	}

9607 9608
	vsi->active_filters = 0;
	clear_bit(__I40E_FILTER_OVERFLOW_PROMISC, &vsi->state);
9609
	spin_lock_bh(&vsi->mac_filter_hash_lock);
J
Jesse Brandeburg 已提交
9610
	/* If macvlan filters already exist, force them to get loaded */
9611
	hash_for_each_safe(vsi->mac_filter_hash, bkt, h, f, hlist) {
9612
		f->state = I40E_FILTER_NEW;
J
Jesse Brandeburg 已提交
9613
		f_count++;
K
Kiran Patil 已提交
9614
	}
9615
	spin_unlock_bh(&vsi->mac_filter_hash_lock);
9616

J
Jesse Brandeburg 已提交
9617 9618 9619 9620 9621 9622 9623 9624 9625
	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,
9626 9627 9628
			 "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
9629 9630 9631 9632 9633 9634 9635 9636 9637 9638 9639 9640 9641 9642 9643 9644
		/* 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)
{
9645 9646
	struct i40e_mac_filter *f;
	struct hlist_node *h;
J
Jesse Brandeburg 已提交
9647 9648 9649
	struct i40e_veb *veb = NULL;
	struct i40e_pf *pf;
	u16 uplink_seid;
9650
	int i, n, bkt;
J
Jesse Brandeburg 已提交
9651 9652 9653 9654 9655 9656 9657 9658 9659 9660 9661 9662 9663 9664 9665 9666 9667 9668 9669 9670 9671 9672 9673 9674

	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 {
9675
			i40e_vsi_close(vsi);
J
Jesse Brandeburg 已提交
9676 9677 9678 9679
		}
		i40e_vsi_disable_irq(vsi);
	}

9680
	spin_lock_bh(&vsi->mac_filter_hash_lock);
9681 9682 9683 9684 9685 9686 9687 9688

	/* 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 */
9689
	hash_for_each_safe(vsi->mac_filter_hash, bkt, h, f, hlist)
9690
		__i40e_del_filter(vsi, f);
9691

9692
	spin_unlock_bh(&vsi->mac_filter_hash_lock);
K
Kiran Patil 已提交
9693

J
Jesse Brandeburg 已提交
9694
	i40e_sync_vsi_filters(vsi);
J
Jesse Brandeburg 已提交
9695 9696 9697

	i40e_vsi_delete(vsi);
	i40e_vsi_free_q_vectors(vsi);
9698 9699 9700 9701
	if (vsi->netdev) {
		free_netdev(vsi->netdev);
		vsi->netdev = NULL;
	}
J
Jesse Brandeburg 已提交
9702 9703 9704 9705 9706 9707 9708 9709 9710 9711 9712
	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
9713
	for (n = 0, i = 0; i < pf->num_alloc_vsi; i++) {
J
Jesse Brandeburg 已提交
9714 9715 9716 9717 9718 9719 9720 9721 9722 9723 9724 9725 9726 9727 9728 9729 9730 9731 9732 9733 9734 9735 9736 9737 9738 9739 9740 9741 9742 9743 9744 9745 9746 9747 9748
		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;

9749
	if (vsi->q_vectors[0]) {
J
Jesse Brandeburg 已提交
9750 9751 9752 9753 9754 9755
		dev_info(&pf->pdev->dev, "VSI %d has existing q_vectors\n",
			 vsi->seid);
		return -EEXIST;
	}

	if (vsi->base_vector) {
9756
		dev_info(&pf->pdev->dev, "VSI %d has non-zero base vector %d\n",
J
Jesse Brandeburg 已提交
9757 9758 9759 9760
			 vsi->seid, vsi->base_vector);
		return -EEXIST;
	}

9761
	ret = i40e_vsi_alloc_q_vectors(vsi);
J
Jesse Brandeburg 已提交
9762 9763 9764 9765 9766 9767 9768 9769
	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;
	}

9770 9771 9772 9773 9774
	/* 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;
9775 9776 9777
	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
9778 9779
	if (vsi->base_vector < 0) {
		dev_info(&pf->pdev->dev,
9780 9781
			 "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
9782 9783 9784 9785 9786 9787 9788 9789 9790
		i40e_vsi_free_q_vectors(vsi);
		ret = -ENOENT;
		goto vector_setup_out;
	}

vector_setup_out:
	return ret;
}

9791 9792 9793 9794 9795 9796 9797 9798 9799 9800 9801
/**
 * 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
9802
	struct i40e_pf *pf;
9803 9804 9805
	u8 enabled_tc;
	int ret;

J
John Underwood 已提交
9806 9807 9808 9809 9810
	if (!vsi)
		return NULL;

	pf = vsi->back;

9811 9812 9813 9814 9815 9816 9817 9818 9819 9820 9821
	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) {
9822
		dev_info(&pf->pdev->dev,
9823
			 "failed to get tracking for %d queues for VSI %d err %d\n",
9824
			 vsi->alloc_queue_pairs, vsi->seid, ret);
9825 9826 9827 9828 9829 9830 9831 9832 9833 9834 9835
		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);
9836 9837
	if (vsi->type == I40E_VSI_MAIN)
		i40e_rm_default_mac_filter(vsi, pf->hw.mac.perm_addr);
9838 9839 9840 9841 9842 9843 9844 9845 9846 9847 9848 9849 9850 9851 9852 9853 9854 9855 9856 9857 9858 9859 9860 9861

	/* 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
9862 9863 9864 9865 9866 9867 9868 9869 9870 9871 9872 9873 9874 9875 9876 9877 9878 9879 9880 9881 9882 9883 9884 9885 9886 9887 9888 9889 9890 9891 9892 9893 9894 9895 9896 9897 9898 9899 9900 9901 9902 9903 9904
/**
 * 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
9905
		for (i = 0; i < pf->num_alloc_vsi; i++) {
J
Jesse Brandeburg 已提交
9906 9907 9908 9909 9910 9911 9912 9913 9914 9915 9916 9917 9918 9919 9920 9921 9922
			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);
9923 9924 9925
		if (veb) {
			if (vsi->seid != pf->vsi[pf->lan_vsi]->seid) {
				dev_info(&vsi->back->pdev->dev,
S
Shannon Nelson 已提交
9926
					 "New VSI creation error, uplink seid of LAN VSI expected.\n");
9927 9928
				return NULL;
			}
9929 9930 9931 9932 9933 9934 9935 9936
			/* 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;
			}
9937
			i40e_config_bridge_mode(veb);
9938
		}
J
Jesse Brandeburg 已提交
9939 9940 9941 9942 9943 9944 9945 9946 9947 9948 9949 9950 9951 9952 9953 9954 9955 9956
		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];
9957 9958
	if (!vsi)
		goto err_alloc;
J
Jesse Brandeburg 已提交
9959 9960 9961 9962 9963 9964 9965 9966
	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 */
9967 9968
	ret = i40e_get_lump(pf, pf->qp_pile, vsi->alloc_queue_pairs,
				vsi->idx);
J
Jesse Brandeburg 已提交
9969
	if (ret < 0) {
9970 9971 9972
		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
9973 9974 9975 9976 9977 9978 9979 9980 9981 9982 9983 9984 9985
		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:
9986 9987 9988 9989 9990 9991 9992 9993 9994 9995 9996
		/* 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
9997 9998 9999 10000 10001 10002 10003 10004 10005
	case I40E_VSI_VMDQ2:
		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
10006 10007 10008 10009
#ifdef CONFIG_I40E_DCB
		/* Setup DCB netlink interface */
		i40e_dcbnl_setup(vsi);
#endif /* CONFIG_I40E_DCB */
J
Jesse Brandeburg 已提交
10010 10011 10012 10013 10014 10015 10016 10017 10018 10019 10020 10021 10022 10023 10024 10025 10026 10027 10028 10029 10030 10031 10032
		/* 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;
	}

10033 10034 10035 10036
	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
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
	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,
10076 10077 10078
			 "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
10079 10080 10081 10082 10083 10084 10085
		goto out;
	}

	ret = i40e_aq_query_switch_comp_ets_config(hw, veb->seid,
						   &ets_data, NULL);
	if (ret) {
		dev_info(&pf->pdev->dev,
10086 10087 10088
			 "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
10089 10090 10091 10092 10093 10094
		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;
10095
	veb->enabled_tc = ets_data.tc_valid_bits;
J
Jesse Brandeburg 已提交
10096 10097 10098 10099 10100 10101 10102 10103 10104 10105 10106 10107 10108 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 10160 10161 10162 10163 10164 10165 10166 10167 10168 10169 10170 10171 10172 10173 10174 10175 10176 10177 10178 10179 10180 10181
	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
10182
	for (i = 0; i < pf->num_alloc_vsi; i++) {
J
Jesse Brandeburg 已提交
10183 10184 10185 10186 10187 10188 10189 10190 10191 10192 10193 10194 10195 10196 10197 10198 10199 10200 10201 10202 10203 10204 10205 10206 10207 10208 10209 10210 10211 10212 10213 10214 10215 10216 10217 10218 10219 10220 10221 10222 10223 10224 10225 10226 10227 10228 10229 10230 10231 10232 10233
		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
10234
	for (i = 0; i < pf->num_alloc_vsi; i++) {
J
Jesse Brandeburg 已提交
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
		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)
{
10272
	struct i40e_pf *pf = veb->pf;
10273
	bool enable_stats = !!(pf->flags & I40E_FLAG_VEB_STATS_ENABLED);
J
Jesse Brandeburg 已提交
10274 10275
	int ret;

10276
	ret = i40e_aq_add_veb(&pf->hw, veb->uplink_seid, vsi->seid,
10277
			      veb->enabled_tc, false,
10278
			      &veb->seid, enable_stats, NULL);
10279 10280

	/* get a VEB from the hardware */
J
Jesse Brandeburg 已提交
10281
	if (ret) {
10282 10283 10284 10285
		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
10286 10287 10288 10289
		return -EPERM;
	}

	/* get statistics counter */
10290
	ret = i40e_aq_get_veb_parameters(&pf->hw, veb->seid, NULL, NULL,
J
Jesse Brandeburg 已提交
10291 10292
					 &veb->stats_idx, NULL, NULL, NULL);
	if (ret) {
10293 10294 10295 10296
		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
10297 10298 10299 10300
		return -EPERM;
	}
	ret = i40e_veb_get_bw_info(veb);
	if (ret) {
10301 10302 10303 10304 10305
		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
10306 10307 10308 10309 10310 10311 10312 10313 10314 10315 10316 10317 10318 10319 10320 10321 10322 10323 10324 10325 10326 10327 10328 10329 10330 10331 10332 10333 10334 10335 10336 10337 10338 10339 10340 10341 10342 10343 10344 10345 10346 10347 10348 10349
		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
10350
	for (vsi_idx = 0; vsi_idx < pf->num_alloc_vsi; vsi_idx++)
J
Jesse Brandeburg 已提交
10351 10352
		if (pf->vsi[vsi_idx] && pf->vsi[vsi_idx]->seid == vsi_seid)
			break;
M
Mitch Williams 已提交
10353
	if (vsi_idx >= pf->num_alloc_vsi && vsi_seid != 0) {
J
Jesse Brandeburg 已提交
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
		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
10388 10389
	if (vsi_idx == pf->lan_vsi)
		pf->lan_veb = veb->idx;
J
Jesse Brandeburg 已提交
10390 10391 10392 10393 10394 10395 10396 10397 10398 10399

	return veb;

err_veb:
	i40e_veb_clear(veb);
err_alloc:
	return NULL;
}

/**
10400
 * i40e_setup_pf_switch_element - set PF vars based on switch type
J
Jesse Brandeburg 已提交
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 10437 10438 10439 10440 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 10469 10470 10471 10472 10473 10474 10475 10476 10477 10478 10479 10480 10481 10482 10483 10484 10485 10486 10487 10488 10489 10490 10491 10492 10493 10494 10495 10496 10497 10498 10499 10500 10501 10502 10503 10504 10505 10506 10507 10508 10509 10510
 * @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,
10511 10512 10513 10514
				 "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
10515 10516 10517 10518 10519 10520 10521 10522 10523 10524 10525 10526 10527 10528 10529 10530 10531 10532 10533 10534 10535 10536 10537 10538 10539 10540 10541 10542
			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
10543
 * @reinit: if the Main VSI needs to re-initialized.
J
Jesse Brandeburg 已提交
10544 10545 10546
 *
 * Returns 0 on success, negative value on failure
 **/
10547
static int i40e_setup_pf_switch(struct i40e_pf *pf, bool reinit)
J
Jesse Brandeburg 已提交
10548
{
10549
	u16 flags = 0;
J
Jesse Brandeburg 已提交
10550 10551 10552 10553 10554 10555
	int ret;

	/* find out what's out there already */
	ret = i40e_fetch_switch_configuration(pf, false);
	if (ret) {
		dev_info(&pf->pdev->dev,
10556 10557 10558
			 "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
10559 10560 10561 10562
		return ret;
	}
	i40e_pf_reset_stats(pf);

10563 10564 10565 10566 10567 10568 10569 10570 10571 10572 10573 10574 10575 10576 10577 10578 10579 10580 10581 10582 10583 10584 10585 10586 10587 10588
	/* 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
10589
	/* first time setup */
10590
	if (pf->lan_vsi == I40E_NO_VSI || reinit) {
J
Jesse Brandeburg 已提交
10591 10592 10593 10594 10595 10596 10597 10598 10599 10600
		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;
10601 10602 10603 10604
		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
10605 10606 10607 10608 10609 10610 10611 10612
		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
10613

J
Jesse Brandeburg 已提交
10614 10615 10616 10617 10618 10619
		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]);

10620 10621
	i40e_fdir_sb_setup(pf);

J
Jesse Brandeburg 已提交
10622 10623 10624 10625 10626 10627 10628 10629 10630 10631 10632 10633
	/* 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))
10634
		i40e_pf_config_rss(pf);
J
Jesse Brandeburg 已提交
10635 10636

	/* fill in link information and enable LSE reporting */
10637 10638
	i40e_link_event(pf);

10639
	/* Initialize user-specific link properties */
J
Jesse Brandeburg 已提交
10640 10641
	pf->fc_autoneg_status = ((pf->hw.phy.link_info.an_info &
				  I40E_AQ_AN_COMPLETED) ? true : false);
10642

J
Jacob Keller 已提交
10643 10644
	i40e_ptp_init(pf);

J
Jesse Brandeburg 已提交
10645 10646 10647 10648 10649 10650 10651 10652 10653 10654 10655 10656 10657 10658 10659 10660 10661 10662 10663
	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;

	/* 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;

10664
	if ((queues_left == 1) ||
10665
	    !(pf->flags & I40E_FLAG_MSIX_ENABLED)) {
J
Jesse Brandeburg 已提交
10666 10667
		/* one qp for PF, no queues for anything else */
		queues_left = 0;
10668
		pf->alloc_rss_size = pf->num_lan_qps = 1;
J
Jesse Brandeburg 已提交
10669 10670

		/* make sure all the fancies are disabled */
J
Jesse Brandeburg 已提交
10671
		pf->flags &= ~(I40E_FLAG_RSS_ENABLED	|
10672
			       I40E_FLAG_IWARP_ENABLED	|
J
Jesse Brandeburg 已提交
10673 10674
			       I40E_FLAG_FD_SB_ENABLED	|
			       I40E_FLAG_FD_ATR_ENABLED	|
10675
			       I40E_FLAG_DCB_CAPABLE	|
10676
			       I40E_FLAG_DCB_ENABLED	|
J
Jesse Brandeburg 已提交
10677 10678
			       I40E_FLAG_SRIOV_ENABLED	|
			       I40E_FLAG_VMDQ_ENABLED);
10679 10680
	} else if (!(pf->flags & (I40E_FLAG_RSS_ENABLED |
				  I40E_FLAG_FD_SB_ENABLED |
10681
				  I40E_FLAG_FD_ATR_ENABLED |
10682
				  I40E_FLAG_DCB_CAPABLE))) {
10683
		/* one qp for PF */
10684
		pf->alloc_rss_size = pf->num_lan_qps = 1;
10685 10686 10687
		queues_left -= pf->num_lan_qps;

		pf->flags &= ~(I40E_FLAG_RSS_ENABLED	|
10688
			       I40E_FLAG_IWARP_ENABLED	|
10689 10690 10691 10692
			       I40E_FLAG_FD_SB_ENABLED	|
			       I40E_FLAG_FD_ATR_ENABLED	|
			       I40E_FLAG_DCB_ENABLED	|
			       I40E_FLAG_VMDQ_ENABLED);
J
Jesse Brandeburg 已提交
10693
	} else {
10694
		/* Not enough queues for all TCs */
10695
		if ((pf->flags & I40E_FLAG_DCB_CAPABLE) &&
10696
		    (queues_left < I40E_MAX_TRAFFIC_CLASS)) {
10697 10698
			pf->flags &= ~(I40E_FLAG_DCB_CAPABLE |
					I40E_FLAG_DCB_ENABLED);
10699 10700
			dev_info(&pf->pdev->dev, "not enough queues for DCB. DCB is disabled.\n");
		}
10701 10702 10703 10704 10705
		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);

10706 10707 10708 10709 10710 10711 10712 10713 10714 10715
		queues_left -= pf->num_lan_qps;
	}

	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
10716 10717 10718 10719
	}

	if ((pf->flags & I40E_FLAG_SRIOV_ENABLED) &&
	    pf->num_vf_qps && pf->num_req_vfs && queues_left) {
10720 10721
		pf->num_req_vfs = min_t(int, pf->num_req_vfs,
					(queues_left / pf->num_vf_qps));
J
Jesse Brandeburg 已提交
10722 10723 10724 10725 10726 10727 10728 10729 10730 10731
		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);
	}

10732
	pf->queues_left = queues_left;
10733 10734 10735 10736
	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),
10737 10738 10739
		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);
J
Jesse Brandeburg 已提交
10740 10741 10742 10743 10744 10745
}

/**
 * i40e_setup_pf_filter_control - Setup PF static filter control
 * @pf: PF to be setup
 *
10746
 * i40e_setup_pf_filter_control sets up a PF's initial filter control
J
Jesse Brandeburg 已提交
10747 10748 10749 10750 10751 10752 10753 10754 10755 10756 10757 10758 10759
 * 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
10760
	if (pf->flags & (I40E_FLAG_FD_SB_ENABLED | I40E_FLAG_FD_ATR_ENABLED))
J
Jesse Brandeburg 已提交
10761 10762 10763 10764 10765 10766 10767 10768 10769 10770 10771 10772
		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
10773
#define INFO_STRING_LEN 255
10774
#define REMAIN(__x) (INFO_STRING_LEN - (__x))
J
Jesse Brandeburg 已提交
10775 10776 10777
static void i40e_print_features(struct i40e_pf *pf)
{
	struct i40e_hw *hw = &pf->hw;
10778 10779
	char *buf;
	int i;
J
Jesse Brandeburg 已提交
10780

10781 10782
	buf = kmalloc(INFO_STRING_LEN, GFP_KERNEL);
	if (!buf)
J
Jesse Brandeburg 已提交
10783 10784
		return;

10785
	i = snprintf(buf, INFO_STRING_LEN, "Features: PF-id[%d]", hw->pf_id);
J
Jesse Brandeburg 已提交
10786
#ifdef CONFIG_PCI_IOV
10787
	i += snprintf(&buf[i], REMAIN(i), " VFs: %d", pf->num_req_vfs);
J
Jesse Brandeburg 已提交
10788
#endif
10789
	i += snprintf(&buf[i], REMAIN(i), " VSIs: %d QP: %d",
10790
		      pf->hw.func_caps.num_vsis,
10791
		      pf->vsi[pf->lan_vsi]->num_queue_pairs);
J
Jesse Brandeburg 已提交
10792
	if (pf->flags & I40E_FLAG_RSS_ENABLED)
10793
		i += snprintf(&buf[i], REMAIN(i), " RSS");
J
Jesse Brandeburg 已提交
10794
	if (pf->flags & I40E_FLAG_FD_ATR_ENABLED)
10795
		i += snprintf(&buf[i], REMAIN(i), " FD_ATR");
10796
	if (pf->flags & I40E_FLAG_FD_SB_ENABLED) {
10797 10798
		i += snprintf(&buf[i], REMAIN(i), " FD_SB");
		i += snprintf(&buf[i], REMAIN(i), " NTUPLE");
10799
	}
10800
	if (pf->flags & I40E_FLAG_DCB_CAPABLE)
10801 10802
		i += snprintf(&buf[i], REMAIN(i), " DCB");
	i += snprintf(&buf[i], REMAIN(i), " VxLAN");
10803
	i += snprintf(&buf[i], REMAIN(i), " Geneve");
J
Jesse Brandeburg 已提交
10804
	if (pf->flags & I40E_FLAG_PTP)
10805
		i += snprintf(&buf[i], REMAIN(i), " PTP");
10806
	if (pf->flags & I40E_FLAG_VEB_MODE_ENABLED)
10807
		i += snprintf(&buf[i], REMAIN(i), " VEB");
10808
	else
10809
		i += snprintf(&buf[i], REMAIN(i), " VEPA");
J
Jesse Brandeburg 已提交
10810

10811 10812
	dev_info(&pf->pdev->dev, "%s\n", buf);
	kfree(buf);
10813
	WARN_ON(i > INFO_STRING_LEN);
J
Jesse Brandeburg 已提交
10814 10815
}

10816 10817 10818 10819 10820 10821 10822 10823 10824 10825 10826 10827 10828 10829
/**
 * 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;
10830
	if (!eth_platform_get_mac_address(&pdev->dev, pf->hw.mac.addr))
10831 10832 10833
		pf->flags |= I40E_FLAG_PF_MAC;
}

J
Jesse Brandeburg 已提交
10834 10835 10836 10837 10838
/**
 * i40e_probe - Device initialization routine
 * @pdev: PCI device information struct
 * @ent: entry in i40e_pci_tbl
 *
10839 10840
 * 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
10841 10842 10843 10844 10845 10846
 * 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)
{
10847
	struct i40e_aq_get_phy_abilities_resp abilities;
J
Jesse Brandeburg 已提交
10848 10849
	struct i40e_pf *pf;
	struct i40e_hw *hw;
10850
	static u16 pfs_found;
10851
	u16 wol_nvm_bits;
10852
	u16 link_status;
10853
	int err;
10854
	u32 val;
10855
	u32 i;
10856
	u8 set_fc_aq_fail;
J
Jesse Brandeburg 已提交
10857 10858 10859 10860 10861 10862

	err = pci_enable_device_mem(pdev);
	if (err)
		return err;

	/* set up for high or low dma */
10863 10864
	err = dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(64));
	if (err) {
10865 10866 10867 10868 10869 10870
		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
10871 10872 10873
	}

	/* set up pci connections */
10874
	err = pci_request_mem_regions(pdev, i40e_driver_name);
J
Jesse Brandeburg 已提交
10875 10876 10877 10878 10879 10880 10881 10882 10883 10884 10885 10886 10887 10888 10889 10890 10891 10892 10893 10894 10895 10896 10897 10898 10899
	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
10900

S
Shannon Nelson 已提交
10901 10902
	pf->ioremap_len = min_t(int, pci_resource_len(pdev, 0),
				I40E_MAX_CSR_SPACE);
A
Anjali Singhai 已提交
10903

S
Shannon Nelson 已提交
10904
	hw->hw_addr = ioremap(pci_resource_start(pdev, 0), pf->ioremap_len);
J
Jesse Brandeburg 已提交
10905 10906 10907 10908
	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
10909
			 pf->ioremap_len, err);
J
Jesse Brandeburg 已提交
10910 10911 10912 10913 10914 10915 10916 10917 10918
		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);
10919
	hw->bus.bus_id = pdev->bus->number;
10920
	pf->instance = pfs_found;
J
Jesse Brandeburg 已提交
10921

10922 10923 10924
	INIT_LIST_HEAD(&pf->l3_flex_pit_list);
	INIT_LIST_HEAD(&pf->l4_flex_pit_list);

10925 10926 10927 10928 10929 10930
	/* 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);

10931 10932 10933 10934 10935 10936
	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
10937

10938 10939 10940 10941 10942 10943 10944 10945 10946 10947 10948
	/* 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
10949
	/* Reset here to make sure all is clean and to define PF 'n' */
10950
	i40e_clear_hw(hw);
J
Jesse Brandeburg 已提交
10951 10952 10953 10954 10955 10956 10957 10958 10959 10960 10961 10962
	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;
10963

10964
	snprintf(pf->int_name, sizeof(pf->int_name) - 1,
10965 10966
		 "%s-%s:misc",
		 dev_driver_string(&pf->pdev->dev), dev_name(&pdev->dev));
J
Jesse Brandeburg 已提交
10967 10968 10969

	err = i40e_init_shared_code(hw);
	if (err) {
10970 10971
		dev_warn(&pdev->dev, "unidentified MAC or BLANK NVM: %d\n",
			 err);
J
Jesse Brandeburg 已提交
10972 10973 10974
		goto err_pf_reset;
	}

10975 10976 10977
	/* set up a default setting for link flow control */
	pf->hw.fc.requested_mode = I40E_FC_NONE;

J
Jesse Brandeburg 已提交
10978
	err = i40e_init_adminq(hw);
C
Carolyn Wyborny 已提交
10979 10980 10981 10982 10983 10984 10985 10986 10987 10988
	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;
	}
10989

10990 10991 10992 10993 10994
	/* 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));
10995

10996 10997
	if (hw->aq.api_maj_ver == I40E_FW_API_VERSION_MAJOR &&
	    hw->aq.api_min_ver > I40E_FW_API_VERSION_MINOR)
10998
		dev_info(&pdev->dev,
10999 11000 11001
			 "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))
11002
		dev_info(&pdev->dev,
11003
			 "The driver for the device detected an older version of the NVM image than expected. Please update the NVM image.\n");
11004

11005 11006
	i40e_verify_eeprom(pf);

11007 11008 11009 11010
	/* 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");

11011
	i40e_clear_pxe_mode(hw);
J
Jesse Brandeburg 已提交
11012 11013 11014 11015 11016 11017 11018 11019 11020 11021 11022
	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,
11023
				hw->func_caps.num_rx_qp, 0, 0);
J
Jesse Brandeburg 已提交
11024 11025 11026 11027 11028 11029 11030 11031 11032 11033 11034 11035
	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;
	}

11036 11037 11038 11039
	/* 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
	 */
11040
	if (pf->flags & I40E_FLAG_STOP_FW_LLDP) {
11041 11042 11043 11044
		dev_info(&pdev->dev, "Stopping firmware LLDP agent.\n");
		i40e_aq_stop_lldp(hw, true, NULL);
	}

J
Jesse Brandeburg 已提交
11045
	i40e_get_mac_addr(hw, hw->mac.addr);
11046 11047
	/* allow a platform config to override the HW addr */
	i40e_get_platform_mac_addr(pdev, pf);
11048
	if (!is_valid_ether_addr(hw->mac.addr)) {
J
Jesse Brandeburg 已提交
11049 11050 11051 11052 11053
		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);
11054
	ether_addr_copy(hw->mac.perm_addr, hw->mac.addr);
11055 11056 11057
	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;
J
Jesse Brandeburg 已提交
11058 11059 11060

	pci_set_drvdata(pdev, pf);
	pci_save_state(pdev);
N
Neerav Parikh 已提交
11061 11062 11063
#ifdef CONFIG_I40E_DCB
	err = i40e_init_pf_dcb(pf);
	if (err) {
11064
		dev_info(&pdev->dev, "DCB init failed %d, disabled\n", err);
11065
		pf->flags &= ~(I40E_FLAG_DCB_CAPABLE | I40E_FLAG_DCB_ENABLED);
11066
		/* Continue without DCB enabled */
N
Neerav Parikh 已提交
11067 11068
	}
#endif /* CONFIG_I40E_DCB */
J
Jesse Brandeburg 已提交
11069 11070 11071 11072 11073 11074 11075 11076 11077

	/* 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;

11078 11079
	/* 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
11080
	if (BIT (hw->port) & wol_nvm_bits || hw->partition_id != 1)
11081 11082 11083
		pf->wol_en = false;
	else
		pf->wol_en = true;
S
Shannon Nelson 已提交
11084 11085
	device_set_wakeup_enable(&pf->pdev->dev, pf->wol_en);

J
Jesse Brandeburg 已提交
11086 11087
	/* set up the main switch operations */
	i40e_determine_queue_usage(pf);
11088 11089 11090
	err = i40e_init_interrupt_scheme(pf);
	if (err)
		goto err_switch_setup;
J
Jesse Brandeburg 已提交
11091

M
Mitch Williams 已提交
11092 11093 11094 11095
	/* 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
11096
	 */
M
Mitch Williams 已提交
11097 11098 11099 11100 11101 11102
	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. */
11103 11104
	pf->vsi = kcalloc(pf->num_alloc_vsi, sizeof(struct i40e_vsi *),
			  GFP_KERNEL);
11105 11106
	if (!pf->vsi) {
		err = -ENOMEM;
J
Jesse Brandeburg 已提交
11107
		goto err_switch_setup;
11108
	}
J
Jesse Brandeburg 已提交
11109

11110 11111 11112 11113 11114 11115 11116 11117 11118
#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
11119
	err = i40e_setup_pf_switch(pf, false);
J
Jesse Brandeburg 已提交
11120 11121 11122 11123
	if (err) {
		dev_info(&pdev->dev, "setup_pf_switch failed: %d\n", err);
		goto err_vsis;
	}
11124 11125 11126 11127 11128 11129 11130 11131 11132 11133 11134 11135 11136 11137 11138 11139 11140 11141 11142

	/* 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));

11143
	/* if FDIR VSI was set up, start it now */
M
Mitch Williams 已提交
11144
	for (i = 0; i < pf->num_alloc_vsi; i++) {
11145 11146 11147 11148 11149
		if (pf->vsi[i] && pf->vsi[i]->type == I40E_VSI_FDIR) {
			i40e_vsi_open(pf->vsi[i]);
			break;
		}
	}
J
Jesse Brandeburg 已提交
11150

S
Shannon Nelson 已提交
11151 11152
	/* The driver only wants link up/down and module qualification
	 * reports from firmware.  Note the negative logic.
J
Jesse Brandeburg 已提交
11153 11154
	 */
	err = i40e_aq_set_phy_int_mask(&pf->hw,
S
Shannon Nelson 已提交
11155
				       ~(I40E_AQ_EVENT_LINK_UPDOWN |
11156
					 I40E_AQ_EVENT_MEDIA_NA |
S
Shannon Nelson 已提交
11157
					 I40E_AQ_EVENT_MODULE_QUAL_FAIL), NULL);
J
Jesse Brandeburg 已提交
11158
	if (err)
11159 11160 11161
		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
11162

11163 11164 11165 11166 11167 11168 11169 11170 11171 11172 11173
	/* 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);
	}

11174
	if (pf->flags & I40E_FLAG_RESTART_AUTONEG) {
11175 11176 11177
		msleep(75);
		err = i40e_aq_set_link_restart_an(&pf->hw, true, NULL);
		if (err)
11178 11179 11180 11181
			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));
11182
	}
J
Jesse Brandeburg 已提交
11183 11184 11185 11186 11187 11188 11189 11190 11191 11192 11193 11194 11195 11196 11197 11198 11199 11200 11201 11202
	/* 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;
		}
	}

11203
#ifdef CONFIG_PCI_IOV
J
Jesse Brandeburg 已提交
11204 11205
	/* prep for VF support */
	if ((pf->flags & I40E_FLAG_SRIOV_ENABLED) &&
11206 11207
	    (pf->flags & I40E_FLAG_MSIX_ENABLED) &&
	    !test_bit(__I40E_BAD_EEPROM, &pf->state)) {
J
Jesse Brandeburg 已提交
11208 11209 11210 11211 11212
		/* 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
11213 11214 11215 11216 11217 11218 11219 11220 11221 11222

		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
11223
	}
11224
#endif /* CONFIG_PCI_IOV */
J
Jesse Brandeburg 已提交
11225

11226 11227 11228 11229 11230 11231 11232 11233 11234 11235 11236
	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;
		}
	}
11237

J
Jesse Brandeburg 已提交
11238 11239 11240
	i40e_dbg_pf_init(pf);

	/* tell the firmware that we're starting */
J
Jesse Brandeburg 已提交
11241
	i40e_send_version(pf);
J
Jesse Brandeburg 已提交
11242 11243 11244 11245 11246

	/* since everything's happy, start the service_task timer */
	mod_timer(&pf->service_timer,
		  round_jiffies(jiffies + pf->service_timer_period));

11247 11248 11249 11250 11251 11252
	/* 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);

11253 11254 11255 11256 11257 11258 11259 11260 11261 11262 11263 11264 11265 11266 11267 11268 11269 11270 11271 11272 11273 11274 11275 11276 11277 11278 11279 11280 11281 11282 11283 11284 11285 11286 11287 11288 11289 11290 11291 11292 11293 11294 11295 11296 11297 11298 11299 11300 11301
#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");
		}
11302 11303
	}

11304 11305 11306
	/* get the requested speeds from the fw */
	err = i40e_aq_get_phy_capabilities(hw, false, false, &abilities, NULL);
	if (err)
11307 11308 11309
		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));
11310 11311
	pf->hw.phy.link_info.requested_speeds = abilities.link_speed;

11312 11313 11314 11315 11316 11317 11318
	/* 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));

11319 11320 11321 11322 11323 11324 11325 11326 11327
	/* 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);

11328
	if ((pf->hw.device_id == I40E_DEV_ID_10G_BASE_T) ||
H
Henry Tieman 已提交
11329 11330
		(pf->hw.device_id == I40E_DEV_ID_10G_BASE_T4))
		pf->flags |= I40E_FLAG_PHY_CONTROLS_LEDS;
11331 11332
	if (pf->hw.device_id == I40E_DEV_ID_SFP_I_X722)
		pf->flags |= I40E_FLAG_HAVE_CRT_RETIMER;
J
Jesse Brandeburg 已提交
11333 11334 11335
	/* print a string summarizing features */
	i40e_print_features(pf);

J
Jesse Brandeburg 已提交
11336 11337 11338 11339 11340 11341 11342
	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);
11343 11344
err_switch_setup:
	i40e_reset_interrupt_capability(pf);
J
Jesse Brandeburg 已提交
11345 11346 11347 11348 11349 11350 11351 11352 11353 11354 11355 11356 11357 11358
	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);
11359
	pci_release_mem_regions(pdev);
J
Jesse Brandeburg 已提交
11360 11361 11362 11363 11364 11365 11366 11367 11368 11369 11370 11371 11372 11373 11374 11375 11376 11377
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);
11378
	struct i40e_hw *hw = &pf->hw;
J
Jesse Brandeburg 已提交
11379 11380 11381 11382 11383
	i40e_status ret_code;
	int i;

	i40e_dbg_pf_exit(pf);

J
Jacob Keller 已提交
11384 11385
	i40e_ptp_stop(pf);

11386
	/* Disable RSS in hw */
11387 11388
	i40e_write_rx_ctl(hw, I40E_PFQF_HENA(0), 0);
	i40e_write_rx_ctl(hw, I40E_PFQF_HENA(1), 0);
11389

J
Jesse Brandeburg 已提交
11390
	/* no more scheduling of any task */
11391
	set_bit(__I40E_SUSPENDED, &pf->state);
J
Jesse Brandeburg 已提交
11392
	set_bit(__I40E_DOWN, &pf->state);
11393 11394 11395 11396
	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
11397

11398 11399 11400 11401 11402
	/* Client close must be called explicitly here because the timer
	 * has been stopped.
	 */
	i40e_notify_client_of_netdev_close(pf->vsi[pf->lan_vsi], false);

11403 11404 11405 11406 11407
	if (pf->flags & I40E_FLAG_SRIOV_ENABLED) {
		i40e_free_vfs(pf);
		pf->flags &= ~I40E_FLAG_SRIOV_ENABLED;
	}

J
Jesse Brandeburg 已提交
11408 11409 11410 11411 11412 11413 11414 11415 11416 11417 11418 11419 11420 11421 11422 11423 11424 11425 11426 11427
	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]);

11428 11429 11430 11431 11432 11433 11434
	/* 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
11435
	/* shutdown and destroy the HMC */
11436 11437
	if (hw->hmc.hmc_obj) {
		ret_code = i40e_shutdown_lan_hmc(hw);
11438 11439 11440 11441 11442
		if (ret_code)
			dev_warn(&pdev->dev,
				 "Failed to destroy the HMC resources: %d\n",
				 ret_code);
	}
J
Jesse Brandeburg 已提交
11443 11444

	/* shutdown the adminq */
H
Henry Tieman 已提交
11445
	i40e_shutdown_adminq(hw);
J
Jesse Brandeburg 已提交
11446

11447 11448 11449 11450
	/* destroy the locks only once, here */
	mutex_destroy(&hw->aq.arq_mutex);
	mutex_destroy(&hw->aq.asq_mutex);

J
Jesse Brandeburg 已提交
11451 11452
	/* Clear all dynamic memory lists of rings, q_vectors, and VSIs */
	i40e_clear_interrupt_scheme(pf);
M
Mitch Williams 已提交
11453
	for (i = 0; i < pf->num_alloc_vsi; i++) {
J
Jesse Brandeburg 已提交
11454 11455 11456 11457 11458 11459 11460 11461 11462 11463 11464 11465 11466 11467 11468
		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);

11469
	iounmap(hw->hw_addr);
J
Jesse Brandeburg 已提交
11470
	kfree(pf);
11471
	pci_release_mem_regions(pdev);
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

	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);

11492 11493 11494 11495 11496 11497
	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
11498
	/* shutdown all operations */
11499 11500 11501 11502 11503
	if (!test_bit(__I40E_SUSPENDED, &pf->state)) {
		rtnl_lock();
		i40e_prep_for_reset(pf);
		rtnl_unlock();
	}
J
Jesse Brandeburg 已提交
11504 11505 11506 11507 11508 11509 11510 11511 11512 11513 11514 11515 11516 11517 11518 11519 11520 11521 11522 11523 11524

	/* 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
11525
	dev_dbg(&pdev->dev, "%s\n", __func__);
J
Jesse Brandeburg 已提交
11526 11527 11528 11529 11530 11531 11532 11533 11534 11535 11536 11537 11538 11539 11540 11541 11542 11543 11544 11545 11546 11547 11548 11549 11550 11551 11552 11553 11554 11555 11556 11557 11558 11559 11560 11561 11562 11563 11564
	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
11565
	dev_dbg(&pdev->dev, "%s\n", __func__);
11566 11567 11568 11569
	if (test_bit(__I40E_SUSPENDED, &pf->state))
		return;

	rtnl_lock();
J
Jesse Brandeburg 已提交
11570
	i40e_handle_reset_warning(pf);
11571
	rtnl_unlock();
11572 11573
}

11574 11575 11576 11577 11578 11579 11580 11581 11582 11583 11584 11585 11586 11587 11588 11589 11590 11591 11592 11593 11594 11595 11596 11597 11598 11599 11600 11601 11602 11603 11604 11605 11606 11607 11608 11609 11610 11611 11612 11613 11614 11615 11616 11617 11618 11619 11620
/**
 * i40e_enable_mc_magic_wake - enable multicast magic packet wake up
 * using the mac_address_write admin q function
 * @pf: pointer to i40e_pf struct
 **/
static void i40e_enable_mc_magic_wake(struct i40e_pf *pf)
{
	struct i40e_hw *hw = &pf->hw;
	i40e_status ret;
	u8 mac_addr[6];
	u16 flags = 0;

	/* Get current MAC address in case it's an LAA */
	if (pf->vsi[pf->lan_vsi] && pf->vsi[pf->lan_vsi]->netdev) {
		ether_addr_copy(mac_addr,
				pf->vsi[pf->lan_vsi]->netdev->dev_addr);
	} else {
		dev_err(&pf->pdev->dev,
			"Failed to retrieve MAC address; using default\n");
		ether_addr_copy(mac_addr, hw->mac.addr);
	}

	/* The FW expects the mac address write cmd to first be called with
	 * one of these flags before calling it again with the multicast
	 * enable flags.
	 */
	flags = I40E_AQC_WRITE_TYPE_LAA_WOL;

	if (hw->func_caps.flex10_enable && hw->partition_id != 1)
		flags = I40E_AQC_WRITE_TYPE_LAA_ONLY;

	ret = i40e_aq_mac_address_write(hw, flags, mac_addr, NULL);
	if (ret) {
		dev_err(&pf->pdev->dev,
			"Failed to update MAC address registers; cannot enable Multicast Magic packet wake up");
		return;
	}

	flags = I40E_AQC_MC_MAG_EN
			| I40E_AQC_WOL_PRESERVE_ON_PFR
			| I40E_AQC_WRITE_TYPE_UPDATE_MC_MAG;
	ret = i40e_aq_mac_address_write(hw, flags, mac_addr, NULL);
	if (ret)
		dev_err(&pf->pdev->dev,
			"Failed to enable Multicast Magic Packet wake up\n");
}

11621 11622 11623 11624 11625 11626 11627
/**
 * 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
11628
	struct i40e_hw *hw = &pf->hw;
11629 11630 11631 11632 11633 11634 11635

	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
11636 11637 11638
	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));

11639 11640 11641 11642
	del_timer_sync(&pf->service_timer);
	cancel_work_sync(&pf->service_task);
	i40e_fdir_teardown(pf);

11643 11644 11645 11646 11647
	/* Client close must be called explicitly here because the timer
	 * has been stopped.
	 */
	i40e_notify_client_of_netdev_close(pf->vsi[pf->lan_vsi], false);

11648 11649 11650
	if (pf->wol_en && (pf->flags & I40E_FLAG_WOL_MC_MAGIC_PKT_WAKE))
		i40e_enable_mc_magic_wake(pf);

11651 11652 11653 11654 11655 11656 11657 11658 11659
	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));

11660 11661
	i40e_clear_interrupt_scheme(pf);

11662
	if (system_state == SYSTEM_POWER_OFF) {
S
Shannon Nelson 已提交
11663
		pci_wake_from_d3(pdev, pf->wol_en);
11664 11665 11666 11667 11668 11669 11670 11671 11672 11673 11674 11675
		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
11676
	struct i40e_hw *hw = &pf->hw;
G
Greg Rose 已提交
11677
	int retval = 0;
11678 11679 11680

	set_bit(__I40E_SUSPENDED, &pf->state);
	set_bit(__I40E_DOWN, &pf->state);
11681

11682 11683 11684
	if (pf->wol_en && (pf->flags & I40E_FLAG_WOL_MC_MAGIC_PKT_WAKE))
		i40e_enable_mc_magic_wake(pf);

11685 11686 11687 11688
	rtnl_lock();
	i40e_prep_for_reset(pf);
	rtnl_unlock();

S
Shannon Nelson 已提交
11689 11690 11691
	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));

11692 11693
	i40e_stop_misc_vector(pf);

G
Greg Rose 已提交
11694 11695 11696 11697
	retval = pci_save_state(pdev);
	if (retval)
		return retval;

S
Shannon Nelson 已提交
11698
	pci_wake_from_d3(pdev, pf->wol_en);
11699 11700
	pci_set_power_state(pdev, PCI_D3hot);

G
Greg Rose 已提交
11701
	return retval;
J
Jesse Brandeburg 已提交
11702 11703
}

11704 11705 11706 11707 11708 11709 11710 11711 11712 11713 11714 11715 11716 11717 11718 11719 11720 11721
/**
 * 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
11722
		dev_err(&pdev->dev, "Cannot enable PCI device from suspend\n");
11723 11724 11725 11726 11727 11728 11729 11730 11731 11732 11733 11734 11735 11736 11737 11738 11739 11740 11741
		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
11742 11743 11744 11745 11746 11747 11748 11749 11750 11751 11752
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,
11753 11754 11755 11756 11757
#ifdef CONFIG_PM
	.suspend  = i40e_suspend,
	.resume   = i40e_resume,
#endif
	.shutdown = i40e_shutdown,
J
Jesse Brandeburg 已提交
11758 11759 11760 11761 11762 11763 11764 11765 11766 11767 11768 11769 11770 11771 11772
	.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);
11773

11774 11775 11776 11777
	/* 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
	 */
11778 11779
	i40e_wq = alloc_workqueue("%s", WQ_UNBOUND | WQ_MEM_RECLAIM, 1,
				  i40e_driver_name);
11780 11781 11782 11783 11784
	if (!i40e_wq) {
		pr_err("%s: Failed to create workqueue\n", i40e_driver_name);
		return -ENOMEM;
	}

J
Jesse Brandeburg 已提交
11785 11786 11787 11788 11789 11790 11791 11792 11793 11794 11795 11796 11797 11798
	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);
11799
	destroy_workqueue(i40e_wq);
J
Jesse Brandeburg 已提交
11800 11801 11802
	i40e_dbg_exit();
}
module_exit(i40e_exit_module);