i40e_main.c 340.4 KB
Newer Older
J
Jesse Brandeburg 已提交
1 2 3
/*******************************************************************************
 *
 * Intel Ethernet Controller XL710 Family Linux Driver
4
 * Copyright(c) 2013 - 2017 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
#include <linux/etherdevice.h>
#include <linux/of_net.h>
#include <linux/pci.h>
30
#include <linux/bpf.h>
31

J
Jesse Brandeburg 已提交
32 33
/* Local includes */
#include "i40e.h"
34
#include "i40e_diag.h"
35
#include <net/udp_tunnel.h>
S
Scott Peterson 已提交
36 37 38 39 40 41
/* All i40e tracepoints are defined by the include below, which
 * must be included exactly once across the whole kernel with
 * CREATE_TRACE_POINTS defined
 */
#define CREATE_TRACE_POINTS
#include "i40e_trace.h"
J
Jesse Brandeburg 已提交
42 43 44 45 46 47 48

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

#define DRV_KERN "-k"

49 50
#define DRV_VERSION_MAJOR 2
#define DRV_VERSION_MINOR 1
51
#define DRV_VERSION_BUILD 14
J
Jesse Brandeburg 已提交
52 53 54 55
#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
56
static const char i40e_copyright[] = "Copyright (c) 2013 - 2014 Intel Corporation.";
J
Jesse Brandeburg 已提交
57 58 59

/* a bit of forward declarations */
static void i40e_vsi_reinit_locked(struct i40e_vsi *vsi);
60
static void i40e_handle_reset_warning(struct i40e_pf *pf, bool lock_acquired);
J
Jesse Brandeburg 已提交
61 62
static int i40e_add_vsi(struct i40e_vsi *vsi);
static int i40e_add_veb(struct i40e_veb *veb, struct i40e_vsi *vsi);
63
static int i40e_setup_pf_switch(struct i40e_pf *pf, bool reinit);
J
Jesse Brandeburg 已提交
64 65 66
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);
67 68 69
static void i40e_prep_for_reset(struct i40e_pf *pf, bool lock_acquired);
static int i40e_reset(struct i40e_pf *pf);
static void i40e_rebuild(struct i40e_pf *pf, bool reinit, bool lock_acquired);
70
static void i40e_fdir_sb_setup(struct i40e_pf *pf);
N
Neerav Parikh 已提交
71
static int i40e_veb_get_bw_info(struct i40e_veb *veb);
J
Jesse Brandeburg 已提交
72 73 74 75 76 77 78 79

/* 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) }
 */
80
static const struct pci_device_id i40e_pci_tbl[] = {
81 82 83 84 85 86 87
	{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
88
	{PCI_VDEVICE(INTEL, I40E_DEV_ID_10G_BASE_T), 0},
89
	{PCI_VDEVICE(INTEL, I40E_DEV_ID_10G_BASE_T4), 0},
90 91
	{PCI_VDEVICE(INTEL, I40E_DEV_ID_KX_X722), 0},
	{PCI_VDEVICE(INTEL, I40E_DEV_ID_QSFP_X722), 0},
92 93 94
	{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},
95
	{PCI_VDEVICE(INTEL, I40E_DEV_ID_SFP_I_X722), 0},
96 97
	{PCI_VDEVICE(INTEL, I40E_DEV_ID_20G_KR2), 0},
	{PCI_VDEVICE(INTEL, I40E_DEV_ID_20G_KR2_A), 0},
98 99
	{PCI_VDEVICE(INTEL, I40E_DEV_ID_25G_B), 0},
	{PCI_VDEVICE(INTEL, I40E_DEV_ID_25G_SFP28), 0},
J
Jesse Brandeburg 已提交
100 101 102 103 104 105 106
	/* required last entry */
	{0, }
};
MODULE_DEVICE_TABLE(pci, i40e_pci_tbl);

#define I40E_MAX_VF_COUNT 128
static int debug = -1;
107 108
module_param(debug, uint, 0);
MODULE_PARM_DESC(debug, "Debug level (0=none,...,16=all), Debug mask (0x8XXXXXXX)");
J
Jesse Brandeburg 已提交
109 110 111 112 113 114

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);

115 116
static struct workqueue_struct *i40e_wq;

J
Jesse Brandeburg 已提交
117 118 119 120 121 122 123 124 125 126 127 128 129 130 131
/**
 * 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
132 133
	if (!mem->va)
		return -ENOMEM;
J
Jesse Brandeburg 已提交
134

J
Jesse Brandeburg 已提交
135
	return 0;
J
Jesse Brandeburg 已提交
136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166
}

/**
 * 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
167 168
	if (!mem->va)
		return -ENOMEM;
J
Jesse Brandeburg 已提交
169

J
Jesse Brandeburg 已提交
170
	return 0;
J
Jesse Brandeburg 已提交
171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204
}

/**
 * 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;
205
	int i, j;
J
Jesse Brandeburg 已提交
206 207 208 209 210 211 212 213 214 215

	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;
216
	while (i < pile->num_entries) {
J
Jesse Brandeburg 已提交
217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234
		/* 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;
235
			break;
J
Jesse Brandeburg 已提交
236
		}
J
Jesse Brandeburg 已提交
237 238 239

		/* not enough, so skip over it and continue looking */
		i += j;
J
Jesse Brandeburg 已提交
240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274
	}

	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;
}

275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290
/**
 * 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
291 292 293 294 295 296
/**
 * 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
 **/
297
void i40e_service_event_schedule(struct i40e_pf *pf)
J
Jesse Brandeburg 已提交
298
{
299
	if (!test_bit(__I40E_DOWN, pf->state) &&
300
	    !test_bit(__I40E_RESET_RECOVERY_PENDING, pf->state))
301
		queue_work(i40e_wq, &pf->service_task);
J
Jesse Brandeburg 已提交
302 303 304 305 306 307 308 309 310 311 312 313 314 315 316
}

/**
 * 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;
317 318 319
	struct i40e_ring *tx_ring = NULL;
	unsigned int i, hung_queue = 0;
	u32 head, val;
J
Jesse Brandeburg 已提交
320 321 322

	pf->tx_timeout_count++;

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

	switch (pf->tx_timeout_recovery_level) {
	case 1:
381
		set_bit(__I40E_PF_RESET_REQUESTED, pf->state);
J
Jesse Brandeburg 已提交
382 383
		break;
	case 2:
384
		set_bit(__I40E_CORE_RESET_REQUESTED, pf->state);
J
Jesse Brandeburg 已提交
385 386
		break;
	case 3:
387
		set_bit(__I40E_GLOBAL_RESET_REQUESTED, pf->state);
J
Jesse Brandeburg 已提交
388 389 390 391 392
		break;
	default:
		netdev_err(netdev, "tx_timeout recovery unsuccessful\n");
		break;
	}
393

J
Jesse Brandeburg 已提交
394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409
	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;
}

410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430
/**
 * i40e_get_netdev_stats_struct_tx - populate stats from a Tx ring
 * @ring: Tx ring to get statistics from
 * @stats: statistics entry to be updated
 **/
static void i40e_get_netdev_stats_struct_tx(struct i40e_ring *ring,
					    struct rtnl_link_stats64 *stats)
{
	u64 bytes, packets;
	unsigned int start;

	do {
		start = u64_stats_fetch_begin_irq(&ring->syncp);
		packets = ring->stats.packets;
		bytes   = ring->stats.bytes;
	} while (u64_stats_fetch_retry_irq(&ring->syncp, start));

	stats->tx_packets += packets;
	stats->tx_bytes   += bytes;
}

J
Jesse Brandeburg 已提交
431 432 433 434 435 436 437
/**
 * 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.
 **/
438
static void i40e_get_netdev_stats_struct(struct net_device *netdev,
439
				  struct rtnl_link_stats64 *stats)
J
Jesse Brandeburg 已提交
440 441
{
	struct i40e_netdev_priv *np = netdev_priv(netdev);
442
	struct i40e_ring *tx_ring, *rx_ring;
J
Jesse Brandeburg 已提交
443
	struct i40e_vsi *vsi = np->vsi;
444 445 446
	struct rtnl_link_stats64 *vsi_stats = i40e_get_vsi_stats_struct(vsi);
	int i;

447
	if (test_bit(__I40E_VSI_DOWN, vsi->state))
448
		return;
449

J
Jesse Brandeburg 已提交
450
	if (!vsi->tx_rings)
451
		return;
J
Jesse Brandeburg 已提交
452

453 454 455 456 457 458 459 460
	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;
461
		i40e_get_netdev_stats_struct_tx(tx_ring, stats);
462 463 464 465

		rx_ring = &tx_ring[1];

		do {
466
			start = u64_stats_fetch_begin_irq(&rx_ring->syncp);
467 468
			packets = rx_ring->stats.packets;
			bytes   = rx_ring->stats.bytes;
469
		} while (u64_stats_fetch_retry_irq(&rx_ring->syncp, start));
J
Jesse Brandeburg 已提交
470

471 472
		stats->rx_packets += packets;
		stats->rx_bytes   += bytes;
473 474 475

		if (i40e_enabled_xdp_vsi(vsi))
			i40e_get_netdev_stats_struct_tx(&rx_ring[1], stats);
476 477 478
	}
	rcu_read_unlock();

479
	/* following stats updated by i40e_watchdog_subtask() */
480 481 482 483
	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;
484
	stats->rx_dropped	= vsi_stats->rx_dropped;
485 486
	stats->rx_crc_errors	= vsi_stats->rx_crc_errors;
	stats->rx_length_errors	= vsi_stats->rx_length_errors;
J
Jesse Brandeburg 已提交
487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505
}

/**
 * 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));
506
	if (vsi->rx_rings && vsi->rx_rings[0]) {
J
Jesse Brandeburg 已提交
507
		for (i = 0; i < vsi->num_queue_pairs; i++) {
J
Jesse Brandeburg 已提交
508
			memset(&vsi->rx_rings[i]->stats, 0,
509
			       sizeof(vsi->rx_rings[i]->stats));
J
Jesse Brandeburg 已提交
510
			memset(&vsi->rx_rings[i]->rx_stats, 0,
511
			       sizeof(vsi->rx_rings[i]->rx_stats));
J
Jesse Brandeburg 已提交
512
			memset(&vsi->tx_rings[i]->stats, 0,
513 514 515
			       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
516
		}
517
	}
J
Jesse Brandeburg 已提交
518 519 520 521
	vsi->stat_offsets_loaded = false;
}

/**
522
 * i40e_pf_reset_stats - Reset all of the stats for the given PF
J
Jesse Brandeburg 已提交
523 524 525 526
 * @pf: the PF to be reset
 **/
void i40e_pf_reset_stats(struct i40e_pf *pf)
{
527 528
	int i;

J
Jesse Brandeburg 已提交
529 530 531
	memset(&pf->stats, 0, sizeof(pf->stats));
	memset(&pf->stats_offsets, 0, sizeof(pf->stats_offsets));
	pf->stat_offsets_loaded = false;
532 533 534 535 536 537 538 539 540 541

	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;
		}
	}
542
	pf->hw_csum_rx_error = 0;
J
Jesse Brandeburg 已提交
543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564
}

/**
 * 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;

565
	if (hw->device_id == I40E_DEV_ID_QEMU) {
J
Jesse Brandeburg 已提交
566 567 568 569 570 571 572 573 574 575
		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
576
		*stat = (new_data + BIT_ULL(48)) - *offset;
J
Jesse Brandeburg 已提交
577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598
	*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
599
		*stat = (u32)((new_data + BIT_ULL(32)) - *offset);
J
Jesse Brandeburg 已提交
600 601
}

602 603 604 605 606 607 608 609 610 611 612 613 614 615
/**
 * i40e_stat_update_and_clear32 - read and clear hw reg, update a 32 bit stat
 * @hw: ptr to the hardware info
 * @reg: the hw reg to read and clear
 * @stat: ptr to the stat
 **/
static void i40e_stat_update_and_clear32(struct i40e_hw *hw, u32 reg, u64 *stat)
{
	u32 new_data = rd32(hw, reg);

	wr32(hw, reg, 1); /* must write a nonzero value to clear register */
	*stat += new_data;
}

J
Jesse Brandeburg 已提交
616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637
/**
 * 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);
638 639 640 641 642 643
	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
644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690

	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 */
691 692 693
	struct i40e_veb_tc_stats *veb_oes;
	struct i40e_veb_tc_stats *veb_es;
	int i, idx = 0;
J
Jesse Brandeburg 已提交
694 695 696 697

	idx = veb->stats_idx;
	es = &veb->stats;
	oes = &veb->stats_offsets;
698 699
	veb_es = &veb->tc_stats;
	veb_oes = &veb->tc_stats_offsets;
J
Jesse Brandeburg 已提交
700 701 702 703 704

	/* 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);
705 706 707 708 709
	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
710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734
	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);
735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756
	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
757 758 759 760
	veb->stat_offsets_loaded = true;
}

/**
S
Shannon Nelson 已提交
761
 * i40e_update_vsi_stats - Update the vsi statistics counters.
J
Jesse Brandeburg 已提交
762 763 764 765 766 767
 * @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
768
 * VF communications.  We sort it out here.
J
Jesse Brandeburg 已提交
769
 **/
S
Shannon Nelson 已提交
770
static void i40e_update_vsi_stats(struct i40e_vsi *vsi)
J
Jesse Brandeburg 已提交
771 772 773 774 775 776 777
{
	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;
778
	struct i40e_ring *p;
J
Jesse Brandeburg 已提交
779
	u32 rx_page, rx_buf;
780 781
	u64 bytes, packets;
	unsigned int start;
782
	u64 tx_linearize;
783
	u64 tx_force_wb;
J
Jesse Brandeburg 已提交
784 785 786 787
	u64 rx_p, rx_b;
	u64 tx_p, tx_b;
	u16 q;

788 789
	if (test_bit(__I40E_VSI_DOWN, vsi->state) ||
	    test_bit(__I40E_CONFIG_BUSY, pf->state))
J
Jesse Brandeburg 已提交
790 791 792 793 794 795 796 797 798 799 800 801
		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;
802
	tx_restart = tx_busy = tx_linearize = tx_force_wb = 0;
J
Jesse Brandeburg 已提交
803 804
	rx_page = 0;
	rx_buf = 0;
805
	rcu_read_lock();
J
Jesse Brandeburg 已提交
806
	for (q = 0; q < vsi->num_queue_pairs; q++) {
807 808 809 810
		/* locate Tx ring */
		p = ACCESS_ONCE(vsi->tx_rings[q]);

		do {
811
			start = u64_stats_fetch_begin_irq(&p->syncp);
812 813
			packets = p->stats.packets;
			bytes = p->stats.bytes;
814
		} while (u64_stats_fetch_retry_irq(&p->syncp, start));
815 816 817 818
		tx_b += bytes;
		tx_p += packets;
		tx_restart += p->tx_stats.restart_queue;
		tx_busy += p->tx_stats.tx_busy;
819
		tx_linearize += p->tx_stats.tx_linearize;
820
		tx_force_wb += p->tx_stats.tx_force_wb;
J
Jesse Brandeburg 已提交
821

822 823 824
		/* Rx queue is part of the same block as Tx queue */
		p = &p[1];
		do {
825
			start = u64_stats_fetch_begin_irq(&p->syncp);
826 827
			packets = p->stats.packets;
			bytes = p->stats.bytes;
828
		} while (u64_stats_fetch_retry_irq(&p->syncp, start));
829 830
		rx_b += bytes;
		rx_p += packets;
M
Mitch Williams 已提交
831 832
		rx_buf += p->rx_stats.alloc_buff_failed;
		rx_page += p->rx_stats.alloc_page_failed;
J
Jesse Brandeburg 已提交
833
	}
834
	rcu_read_unlock();
J
Jesse Brandeburg 已提交
835 836
	vsi->tx_restart = tx_restart;
	vsi->tx_busy = tx_busy;
837
	vsi->tx_linearize = tx_linearize;
838
	vsi->tx_force_wb = tx_force_wb;
J
Jesse Brandeburg 已提交
839 840 841 842 843 844 845 846 847
	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
848
	i40e_update_eth_stats(vsi);
J
Jesse Brandeburg 已提交
849 850 851 852
	ons->tx_errors = oes->tx_errors;
	ns->tx_errors = es->tx_errors;
	ons->multicast = oes->rx_multicast;
	ns->multicast = es->rx_multicast;
853 854
	ons->rx_dropped = oes->rx_discards;
	ns->rx_dropped = es->rx_discards;
J
Jesse Brandeburg 已提交
855 856 857
	ons->tx_dropped = oes->tx_discards;
	ns->tx_dropped = es->tx_discards;

S
Shannon Nelson 已提交
858
	/* pull in a couple PF stats if this is the main vsi */
J
Jesse Brandeburg 已提交
859
	if (vsi == pf->vsi[pf->lan_vsi]) {
S
Shannon Nelson 已提交
860 861 862 863 864
		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
865

S
Shannon Nelson 已提交
866
/**
867
 * i40e_update_pf_stats - Update the PF statistics counters.
S
Shannon Nelson 已提交
868 869 870 871 872 873 874 875 876
 * @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
877

S
Shannon Nelson 已提交
878 879 880 881 882 883 884 885 886 887 888 889
	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
890 891 892 893 894
	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
895 896 897 898 899
	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
900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919
	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
920

S
Shannon Nelson 已提交
921 922 923 924
	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
925

S
Shannon Nelson 已提交
926 927 928
	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
929

S
Shannon Nelson 已提交
930 931 932
	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
933

S
Shannon Nelson 已提交
934 935 936 937 938 939 940 941
	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
942

S
Shannon Nelson 已提交
943 944 945 946
	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
947

S
Shannon Nelson 已提交
948 949 950 951 952 953
	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);
954 955 956
	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
957 958 959
	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
960

S
Shannon Nelson 已提交
961
	for (i = 0; i < 8; i++) {
962 963 964 965
		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
966
		i40e_stat_update32(hw, I40E_GLPRT_PXONRXC(hw->port, i),
J
Jesse Brandeburg 已提交
967
				   pf->stat_offsets_loaded,
S
Shannon Nelson 已提交
968 969 970
				   &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
971
				   pf->stat_offsets_loaded,
S
Shannon Nelson 已提交
972 973 974
				   &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
975
				   pf->stat_offsets_loaded,
S
Shannon Nelson 已提交
976 977 978 979
				   &osd->priority_xoff_tx[i],
				   &nsd->priority_xoff_tx[i]);
		i40e_stat_update32(hw,
				   I40E_GLPRT_RXON2OFFCNT(hw->port, i),
980
				   pf->stat_offsets_loaded,
S
Shannon Nelson 已提交
981 982
				   &osd->priority_xon_2_xoff[i],
				   &nsd->priority_xon_2_xoff[i]);
J
Jesse Brandeburg 已提交
983 984
	}

S
Shannon Nelson 已提交
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 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055
	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);

1056
	/* FDIR stats */
1057 1058 1059 1060 1061 1062 1063 1064 1065
	i40e_stat_update_and_clear32(hw,
			I40E_GLQF_PCNT(I40E_FD_ATR_STAT_IDX(hw->pf_id)),
			&nsd->fd_atr_match);
	i40e_stat_update_and_clear32(hw,
			I40E_GLQF_PCNT(I40E_FD_SB_STAT_IDX(hw->pf_id)),
			&nsd->fd_sb_match);
	i40e_stat_update_and_clear32(hw,
			I40E_GLQF_PCNT(I40E_FD_ATR_TUNNEL_STAT_IDX(hw->pf_id)),
			&nsd->fd_atr_tunnel_match);
1066

S
Shannon Nelson 已提交
1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080
	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);

1081
	if (pf->flags & I40E_FLAG_FD_SB_ENABLED &&
1082
	    !(pf->flags & I40E_FLAG_FD_SB_AUTO_DISABLED))
1083 1084 1085 1086 1087
		nsd->fd_sb_status = true;
	else
		nsd->fd_sb_status = false;

	if (pf->flags & I40E_FLAG_FD_ATR_ENABLED &&
1088
	    !(pf->flags & I40E_FLAG_FD_ATR_AUTO_DISABLED))
1089 1090 1091 1092
		nsd->fd_atr_status = true;
	else
		nsd->fd_atr_status = false;

J
Jesse Brandeburg 已提交
1093 1094 1095
	pf->stat_offsets_loaded = true;
}

S
Shannon Nelson 已提交
1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111
/**
 * 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
1112 1113 1114 1115 1116 1117 1118 1119 1120
/**
 * 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,
1121
						const u8 *macaddr, s16 vlan)
J
Jesse Brandeburg 已提交
1122 1123
{
	struct i40e_mac_filter *f;
1124
	u64 key;
J
Jesse Brandeburg 已提交
1125 1126 1127 1128

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

1129 1130
	key = i40e_addr_to_hkey(macaddr);
	hash_for_each_possible(vsi->mac_filter_hash, f, hlist, key) {
J
Jesse Brandeburg 已提交
1131
		if ((ether_addr_equal(macaddr, f->macaddr)) &&
1132
		    (vlan == f->vlan))
J
Jesse Brandeburg 已提交
1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145
			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
 **/
1146
struct i40e_mac_filter *i40e_find_mac(struct i40e_vsi *vsi, const u8 *macaddr)
J
Jesse Brandeburg 已提交
1147 1148
{
	struct i40e_mac_filter *f;
1149
	u64 key;
J
Jesse Brandeburg 已提交
1150 1151 1152 1153

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

1154 1155
	key = i40e_addr_to_hkey(macaddr);
	hash_for_each_possible(vsi->mac_filter_hash, f, hlist, key) {
1156
		if ((ether_addr_equal(macaddr, f->macaddr)))
J
Jesse Brandeburg 已提交
1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169
			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)
{
1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182
	/* 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
1183
	 *    i40e_sync_filters_subtask.
1184 1185 1186 1187
	 *
	 * 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
1188
	 * i40e_sync_filters_subtask, but we already have to search
1189 1190 1191 1192
	 * 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
1193
	 */
1194
	return vsi->has_vlan_filter;
J
Jesse Brandeburg 已提交
1195 1196
}

1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230
/**
 * 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)
{
1231
	s16 pvid = le16_to_cpu(vsi->info.pvid);
1232
	struct i40e_mac_filter *f, *add_head;
1233
	struct i40e_new_mac_filter *new;
1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251
	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 */
1252
	hlist_for_each_entry(new, tmp_add_list, hlist) {
1253 1254
		if (pvid && new->f->vlan != pvid)
			new->f->vlan = pvid;
1255 1256 1257 1258
		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;
1259 1260 1261 1262 1263 1264 1265 1266 1267
	}

	/* 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.
		 */
1268
		if ((pvid && f->vlan != pvid) ||
1269 1270 1271
		    (vlan_filters && f->vlan == I40E_VLAN_ANY) ||
		    (!vlan_filters && f->vlan == 0)) {
			/* Determine the new vlan we will be adding */
1272 1273
			if (pvid)
				new_vlan = pvid;
1274 1275 1276 1277 1278 1279 1280 1281 1282 1283
			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;

1284 1285 1286 1287 1288 1289 1290 1291 1292 1293
			/* 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);
1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306

			/* 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;
}

1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339
/**
 * 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
1340 1341 1342 1343 1344 1345 1346
/**
 * 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
1347
 *
1348
 * NOTE: This function is expected to be called with mac_filter_hash_lock
K
Kiran Patil 已提交
1349
 * being held.
J
Jesse Brandeburg 已提交
1350 1351
 **/
struct i40e_mac_filter *i40e_add_filter(struct i40e_vsi *vsi,
1352
					const u8 *macaddr, s16 vlan)
J
Jesse Brandeburg 已提交
1353 1354
{
	struct i40e_mac_filter *f;
1355
	u64 key;
J
Jesse Brandeburg 已提交
1356 1357 1358 1359

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

1360
	f = i40e_find_filter(vsi, macaddr, vlan);
J
Jesse Brandeburg 已提交
1361 1362 1363
	if (!f) {
		f = kzalloc(sizeof(*f), GFP_ATOMIC);
		if (!f)
1364
			return NULL;
J
Jesse Brandeburg 已提交
1365

1366 1367 1368 1369 1370 1371
		/* Update the boolean indicating if we need to function in
		 * VLAN mode.
		 */
		if (vlan >= 0)
			vsi->has_vlan_filter = true;

1372
		ether_addr_copy(f->macaddr, macaddr);
J
Jesse Brandeburg 已提交
1373
		f->vlan = vlan;
1374 1375 1376 1377
		/* 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.
		 */
1378
		if (test_bit(__I40E_VSI_OVERFLOW_PROMISC, vsi->state))
1379 1380 1381
			f->state = I40E_FILTER_FAILED;
		else
			f->state = I40E_FILTER_NEW;
1382 1383 1384 1385
		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
1386 1387 1388 1389 1390

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

1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401
	/* 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
1402 1403 1404 1405
	return f;
}

/**
1406 1407 1408 1409 1410 1411 1412
 * __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
1413
 *
1414
 * NOTE: This function is expected to be called with mac_filter_hash_lock
K
Kiran Patil 已提交
1415
 * being held.
1416 1417 1418
 * 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
1419
 **/
1420
void __i40e_del_filter(struct i40e_vsi *vsi, struct i40e_mac_filter *f)
J
Jesse Brandeburg 已提交
1421
{
1422
	if (!f)
J
Jesse Brandeburg 已提交
1423 1424
		return;

1425 1426 1427 1428
	/* 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.
	 */
1429 1430
	if ((f->state == I40E_FILTER_FAILED) ||
	    (f->state == I40E_FILTER_NEW)) {
1431
		hash_del(&f->hlist);
1432
		kfree(f);
J
Jesse Brandeburg 已提交
1433
	} else {
1434
		f->state = I40E_FILTER_REMOVE;
J
Jesse Brandeburg 已提交
1435
	}
1436 1437 1438

	vsi->flags |= I40E_VSI_FLAG_FILTER_CHANGED;
	vsi->back->flags |= I40E_FLAG_FILTER_SYNC;
J
Jesse Brandeburg 已提交
1439 1440
}

1441 1442 1443 1444 1445 1446
/**
 * i40e_del_filter - Remove a MAC/VLAN filter from the VSI
 * @vsi: the VSI to be searched
 * @macaddr: the MAC address
 * @vlan: the VLAN
 *
1447
 * NOTE: This function is expected to be called with mac_filter_hash_lock
1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463
 * 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);
}

1464
/**
1465
 * i40e_add_mac_filter - Add a MAC filter for all active VLANs
1466 1467 1468
 * @vsi: the VSI to be searched
 * @macaddr: the mac address to be filtered
 *
1469 1470
 * 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
1471 1472
 * unique vlan that already exists. If a PVID has been assigned, instead only
 * add the macaddr to that VLAN.
1473
 *
1474
 * Returns last filter added on success, else NULL
1475
 **/
1476 1477
struct i40e_mac_filter *i40e_add_mac_filter(struct i40e_vsi *vsi,
					    const u8 *macaddr)
1478
{
1479
	struct i40e_mac_filter *f, *add = NULL;
1480 1481
	struct hlist_node *h;
	int bkt;
1482 1483 1484 1485

	if (vsi->info.pvid)
		return i40e_add_filter(vsi, macaddr,
				       le16_to_cpu(vsi->info.pvid));
1486

1487 1488 1489
	if (!i40e_is_vsi_in_vlan(vsi))
		return i40e_add_filter(vsi, macaddr, I40E_VLAN_ANY);

1490
	hash_for_each_safe(vsi->mac_filter_hash, bkt, h, f, hlist) {
1491 1492
		if (f->state == I40E_FILTER_REMOVE)
			continue;
1493 1494 1495
		add = i40e_add_filter(vsi, macaddr, f->vlan);
		if (!add)
			return NULL;
1496 1497
	}

1498
	return add;
1499 1500 1501
}

/**
1502
 * i40e_del_mac_filter - Remove a MAC filter from all VLANs
1503 1504 1505
 * @vsi: the VSI to be searched
 * @macaddr: the mac address to be removed
 *
1506 1507
 * Removes a given MAC address from a VSI regardless of what VLAN it has been
 * associated with.
1508 1509 1510
 *
 * Returns 0 for success, or error
 **/
1511
int i40e_del_mac_filter(struct i40e_vsi *vsi, const u8 *macaddr)
1512
{
1513 1514
	struct i40e_mac_filter *f;
	struct hlist_node *h;
1515
	bool found = false;
1516
	int bkt;
1517

1518 1519 1520
	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) {
1521 1522 1523 1524
		if (ether_addr_equal(macaddr, f->macaddr)) {
			__i40e_del_filter(vsi, f);
			found = true;
		}
1525
	}
1526 1527

	if (found)
1528
		return 0;
1529 1530
	else
		return -ENOENT;
1531 1532
}

J
Jesse Brandeburg 已提交
1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543
/**
 * 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;
1544 1545
	struct i40e_pf *pf = vsi->back;
	struct i40e_hw *hw = &pf->hw;
J
Jesse Brandeburg 已提交
1546 1547 1548 1549 1550
	struct sockaddr *addr = p;

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

1551 1552 1553 1554 1555
	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
1556

1557 1558
	if (test_bit(__I40E_VSI_DOWN, vsi->back->state) ||
	    test_bit(__I40E_RESET_RECOVERY_PENDING, vsi->back->state))
1559 1560
		return -EADDRNOTAVAIL;

1561 1562 1563 1564 1565 1566
	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);

1567
	spin_lock_bh(&vsi->mac_filter_hash_lock);
1568 1569
	i40e_del_mac_filter(vsi, netdev->dev_addr);
	i40e_add_mac_filter(vsi, addr->sa_data);
1570
	spin_unlock_bh(&vsi->mac_filter_hash_lock);
1571
	ether_addr_copy(netdev->dev_addr, addr->sa_data);
J
Jesse Brandeburg 已提交
1572 1573
	if (vsi->type == I40E_VSI_MAIN) {
		i40e_status ret;
J
Jesse Brandeburg 已提交
1574

J
Jesse Brandeburg 已提交
1575
		ret = i40e_aq_mac_address_write(&vsi->back->hw,
1576
						I40E_AQC_WRITE_TYPE_LAA_WOL,
J
Jesse Brandeburg 已提交
1577
						addr->sa_data, NULL);
1578 1579 1580 1581
		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));
1582 1583
	}

1584 1585 1586 1587 1588
	/* 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
1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612
}

/**
 * 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
1613
	u16 num_tc_qps = 0;
J
Jesse Brandeburg 已提交
1614 1615 1616 1617 1618 1619 1620

	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
1621
			if (enabled_tc & BIT(i)) /* TC is enabled */
J
Jesse Brandeburg 已提交
1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634
				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
1635
	/* Number of queues per enabled TC */
1636 1637
	qcount = vsi->alloc_queue_pairs;

1638
	num_tc_qps = qcount / numtc;
1639
	num_tc_qps = min_t(int, num_tc_qps, i40e_pf_get_max_q_per_tc(pf));
J
Jesse Brandeburg 已提交
1640 1641 1642 1643

	/* 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
1644
		if (vsi->tc_config.enabled_tc & BIT(i)) {
1645
			/* TC is enabled */
J
Jesse Brandeburg 已提交
1646 1647 1648 1649
			int pow, num_qps;

			switch (vsi->type) {
			case I40E_VSI_MAIN:
1650 1651
				qcount = min_t(int, pf->alloc_rss_size,
					       num_tc_qps);
J
Jesse Brandeburg 已提交
1652 1653 1654 1655 1656
				break;
			case I40E_VSI_FDIR:
			case I40E_VSI_SRIOV:
			case I40E_VSI_VMDQ2:
			default:
N
Neerav Parikh 已提交
1657
				qcount = num_tc_qps;
J
Jesse Brandeburg 已提交
1658 1659 1660
				WARN_ON(i != 0);
				break;
			}
N
Neerav Parikh 已提交
1661 1662
			vsi->tc_config.tc_info[i].qoffset = offset;
			vsi->tc_config.tc_info[i].qcount = qcount;
J
Jesse Brandeburg 已提交
1663

S
Shannon Nelson 已提交
1664
			/* find the next higher power-of-2 of num queue pairs */
N
Neerav Parikh 已提交
1665
			num_qps = qcount;
J
Jesse Brandeburg 已提交
1666
			pow = 0;
1667
			while (num_qps && (BIT_ULL(pow) < qcount)) {
J
Jesse Brandeburg 已提交
1668 1669 1670 1671 1672 1673 1674 1675 1676
				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
1677
			offset += qcount;
J
Jesse Brandeburg 已提交
1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693
		} 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;
1694 1695 1696
	if ((vsi->type == I40E_VSI_MAIN) && (numtc == 1)) {
		if (vsi->req_queue_pairs > 0)
			vsi->num_queue_pairs = vsi->req_queue_pairs;
1697
		else if (pf->flags & I40E_FLAG_MSIX_ENABLED)
1698 1699
			vsi->num_queue_pairs = pf->num_lan_msix;
	}
J
Jesse Brandeburg 已提交
1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720

	/* 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);
}

1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733
/**
 * 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;

1734
	if (i40e_add_mac_filter(vsi, addr))
1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752
		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;

1753
	i40e_del_mac_filter(vsi, addr);
1754 1755 1756 1757

	return 0;
}

J
Jesse Brandeburg 已提交
1758 1759 1760 1761 1762 1763 1764 1765 1766
/**
 * 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;

1767
	spin_lock_bh(&vsi->mac_filter_hash_lock);
K
Kiran Patil 已提交
1768

1769 1770
	__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
1771

1772
	spin_unlock_bh(&vsi->mac_filter_hash_lock);
J
Jesse Brandeburg 已提交
1773 1774 1775 1776 1777 1778 1779 1780

	/* 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;
	}
}

K
Kiran Patil 已提交
1781
/**
1782
 * i40e_undo_del_filter_entries - Undo the changes made to MAC filter entries
1783
 * @vsi: Pointer to VSI struct
K
Kiran Patil 已提交
1784 1785 1786
 * @from: Pointer to list which contains MAC filter entries - changes to
 *        those entries needs to be undone.
 *
1787
 * MAC filter entries from this list were slated for deletion.
K
Kiran Patil 已提交
1788
 **/
1789 1790
static void i40e_undo_del_filter_entries(struct i40e_vsi *vsi,
					 struct hlist_head *from)
K
Kiran Patil 已提交
1791
{
1792 1793 1794 1795 1796
	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
1797 1798

		/* Move the element back into MAC filter list*/
1799 1800
		hlist_del(&f->hlist);
		hash_add(vsi->mac_filter_hash, &f->hlist, key);
K
Kiran Patil 已提交
1801 1802 1803
	}
}

1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824
/**
 * 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);
	}
}

1825 1826
/**
 * i40e_next_entry - Get the next non-broadcast filter from a list
1827
 * @next: pointer to filter in list
1828 1829 1830 1831 1832
 *
 * 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.
 */
1833 1834
static
struct i40e_new_mac_filter *i40e_next_filter(struct i40e_new_mac_filter *next)
1835
{
1836 1837 1838
	hlist_for_each_entry_continue(next, hlist) {
		if (!is_broadcast_ether_addr(next->f->macaddr))
			return next;
1839 1840
	}

1841
	return NULL;
1842 1843
}

K
Kiran Patil 已提交
1844
/**
1845 1846 1847 1848 1849
 * 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
1850
 *
1851 1852
 * 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
1853
 **/
1854 1855 1856
static int
i40e_update_filter_state(int count,
			 struct i40e_aqc_add_macvlan_element_data *add_list,
1857
			 struct i40e_new_mac_filter *add_head)
K
Kiran Patil 已提交
1858
{
1859 1860
	int retval = 0;
	int i;
K
Kiran Patil 已提交
1861

1862 1863 1864 1865 1866 1867
	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.
1868
		 */
1869
		if (add_list[i].match_method == I40E_AQC_MM_ERR_NO_RES) {
1870
			add_head->state = I40E_FILTER_FAILED;
1871 1872 1873
		} else {
			add_head->state = I40E_FILTER_ACTIVE;
			retval++;
1874
		}
1875

1876 1877 1878
		add_head = i40e_next_filter(add_head);
		if (!add_head)
			break;
K
Kiran Patil 已提交
1879
	}
1880

1881
	return retval;
K
Kiran Patil 已提交
1882 1883
}

1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934
/**
 * 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,
1935
			  struct i40e_new_mac_filter *add_head,
1936 1937 1938 1939 1940
			  int num_add, bool *promisc_changed)
{
	struct i40e_hw *hw = &vsi->back->hw;
	int aq_err, fcnt;

1941
	i40e_aq_add_macvlan(hw, vsi->seid, list, num_add, NULL);
1942
	aq_err = hw->aq.asq_last_status;
1943
	fcnt = i40e_update_filter_state(num_add, list, add_head);
1944 1945 1946

	if (fcnt != num_add) {
		*promisc_changed = true;
1947
		set_bit(__I40E_VSI_OVERFLOW_PROMISC, vsi->state);
1948 1949 1950 1951 1952 1953 1954
		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);
	}
}

1955 1956 1957 1958 1959 1960 1961 1962
/**
 * 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.
1963 1964
 *
 * Returns status indicating success or failure;
1965
 **/
1966 1967 1968
static i40e_status
i40e_aqc_broadcast_filter(struct i40e_vsi *vsi, const char *vsi_name,
			  struct i40e_mac_filter *f)
1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986
{
	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);
	}

1987
	if (aq_ret)
1988 1989 1990 1991
		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);
1992 1993

	return aq_ret;
1994 1995
}

J
Jesse Brandeburg 已提交
1996 1997 1998 1999 2000 2001 2002 2003
/**
 * 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
2004
int i40e_sync_vsi_filters(struct i40e_vsi *vsi)
J
Jesse Brandeburg 已提交
2005
{
2006
	struct hlist_head tmp_add_list, tmp_del_list;
2007 2008
	struct i40e_mac_filter *f;
	struct i40e_new_mac_filter *new, *add_head = NULL;
2009
	struct i40e_hw *hw = &vsi->back->hw;
2010
	unsigned int failed_filters = 0;
2011
	unsigned int vlan_filters = 0;
2012
	bool promisc_changed = false;
2013
	char vsi_name[16] = "PF";
J
Jesse Brandeburg 已提交
2014
	int filter_list_len = 0;
M
Mitch Williams 已提交
2015
	i40e_status aq_ret = 0;
2016
	u32 changed_flags = 0;
2017
	struct hlist_node *h;
J
Jesse Brandeburg 已提交
2018 2019 2020
	struct i40e_pf *pf;
	int num_add = 0;
	int num_del = 0;
2021
	int retval = 0;
J
Jesse Brandeburg 已提交
2022
	u16 cmd_flags;
2023
	int list_size;
2024
	int bkt;
J
Jesse Brandeburg 已提交
2025 2026 2027 2028 2029

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

2030
	while (test_and_set_bit(__I40E_VSI_SYNCING_FILTERS, vsi->state))
J
Jesse Brandeburg 已提交
2031 2032 2033 2034 2035 2036 2037 2038
		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;
	}

2039 2040
	INIT_HLIST_HEAD(&tmp_add_list);
	INIT_HLIST_HEAD(&tmp_del_list);
K
Kiran Patil 已提交
2041

2042 2043 2044 2045 2046
	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
2047 2048 2049
	if (vsi->flags & I40E_VSI_FLAG_FILTER_CHANGED) {
		vsi->flags &= ~I40E_VSI_FLAG_FILTER_CHANGED;

2050
		spin_lock_bh(&vsi->mac_filter_hash_lock);
2051
		/* Create a list of filters to delete. */
2052
		hash_for_each_safe(vsi->mac_filter_hash, bkt, h, f, hlist) {
2053 2054
			if (f->state == I40E_FILTER_REMOVE) {
				/* Move the element into temporary del_list */
2055 2056
				hash_del(&f->hlist);
				hlist_add_head(&f->hlist, &tmp_del_list);
2057 2058 2059

				/* Avoid counting removed filters */
				continue;
2060 2061
			}
			if (f->state == I40E_FILTER_NEW) {
2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072
				/* 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
2073
			}
2074

2075 2076 2077
			/* Count the number of active (current and new) VLAN
			 * filters we have now. Does not count filters which
			 * are marked for deletion.
2078 2079 2080 2081 2082
			 */
			if (f->vlan > 0)
				vlan_filters++;
		}

2083 2084 2085 2086 2087 2088
		retval = i40e_correct_mac_vlan_filters(vsi,
						       &tmp_add_list,
						       &tmp_del_list,
						       vlan_filters);
		if (retval)
			goto err_no_memory_locked;
2089

2090
		spin_unlock_bh(&vsi->mac_filter_hash_lock);
K
Kiran Patil 已提交
2091 2092 2093
	}

	/* Now process 'del_list' outside the lock */
2094
	if (!hlist_empty(&tmp_del_list)) {
2095
		filter_list_len = hw->aq.asq_buf_size /
K
Kiran Patil 已提交
2096
			    sizeof(struct i40e_aqc_remove_macvlan_element_data);
2097
		list_size = filter_list_len *
S
Shannon Nelson 已提交
2098
			    sizeof(struct i40e_aqc_remove_macvlan_element_data);
2099
		del_list = kzalloc(list_size, GFP_ATOMIC);
2100 2101
		if (!del_list)
			goto err_no_memory;
K
Kiran Patil 已提交
2102

2103
		hlist_for_each_entry_safe(f, h, &tmp_del_list, hlist) {
J
Jesse Brandeburg 已提交
2104 2105
			cmd_flags = 0;

2106
			/* handle broadcast filters by updating the broadcast
2107
			 * promiscuous flag and release filter list.
2108 2109 2110 2111 2112 2113 2114 2115 2116
			 */
			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
2117
			/* add to delete list */
2118
			ether_addr_copy(del_list[num_del].mac_addr, f->macaddr);
2119 2120
			if (f->vlan == I40E_VLAN_ANY) {
				del_list[num_del].vlan_tag = 0;
A
Alan Brady 已提交
2121
				cmd_flags |= I40E_AQC_MACVLAN_DEL_IGNORE_VLAN;
2122 2123 2124 2125
			} else {
				del_list[num_del].vlan_tag =
					cpu_to_le16((u16)(f->vlan));
			}
J
Jesse Brandeburg 已提交
2126 2127 2128 2129 2130 2131 2132

			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) {
2133 2134
				i40e_aqc_del_filters(vsi, vsi_name, del_list,
						     num_del, &retval);
2135
				memset(del_list, 0, list_size);
2136
				num_del = 0;
J
Jesse Brandeburg 已提交
2137
			}
K
Kiran Patil 已提交
2138 2139 2140
			/* Release memory for MAC filter entries which were
			 * synced up with HW.
			 */
2141
			hlist_del(&f->hlist);
K
Kiran Patil 已提交
2142
			kfree(f);
J
Jesse Brandeburg 已提交
2143
		}
K
Kiran Patil 已提交
2144

J
Jesse Brandeburg 已提交
2145
		if (num_del) {
2146 2147
			i40e_aqc_del_filters(vsi, vsi_name, del_list,
					     num_del, &retval);
J
Jesse Brandeburg 已提交
2148 2149 2150 2151
		}

		kfree(del_list);
		del_list = NULL;
K
Kiran Patil 已提交
2152 2153
	}

2154
	if (!hlist_empty(&tmp_add_list)) {
2155
		/* Do all the adds now. */
2156
		filter_list_len = hw->aq.asq_buf_size /
S
Shannon Nelson 已提交
2157
			       sizeof(struct i40e_aqc_add_macvlan_element_data);
2158 2159 2160
		list_size = filter_list_len *
			       sizeof(struct i40e_aqc_add_macvlan_element_data);
		add_list = kzalloc(list_size, GFP_ATOMIC);
2161 2162 2163
		if (!add_list)
			goto err_no_memory;

2164
		num_add = 0;
2165
		hlist_for_each_entry_safe(new, h, &tmp_add_list, hlist) {
2166
			if (test_bit(__I40E_VSI_OVERFLOW_PROMISC,
2167
				     vsi->state)) {
2168
				new->state = I40E_FILTER_FAILED;
2169 2170
				continue;
			}
2171 2172 2173 2174

			/* handle broadcast filters by updating the broadcast
			 * promiscuous flag instead of adding a MAC filter.
			 */
2175 2176 2177 2178 2179 2180
			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;
2181 2182 2183
				continue;
			}

J
Jesse Brandeburg 已提交
2184
			/* add to add array */
2185
			if (num_add == 0)
2186
				add_head = new;
2187
			cmd_flags = 0;
2188 2189 2190
			ether_addr_copy(add_list[num_add].mac_addr,
					new->f->macaddr);
			if (new->f->vlan == I40E_VLAN_ANY) {
2191 2192 2193 2194
				add_list[num_add].vlan_tag = 0;
				cmd_flags |= I40E_AQC_MACVLAN_ADD_IGNORE_VLAN;
			} else {
				add_list[num_add].vlan_tag =
2195
					cpu_to_le16((u16)(new->f->vlan));
2196
			}
J
Jesse Brandeburg 已提交
2197
			add_list[num_add].queue_number = 0;
2198
			/* set invalid match method for later detection */
2199
			add_list[num_add].match_method = I40E_AQC_MM_ERR_NO_RES;
J
Jesse Brandeburg 已提交
2200 2201 2202 2203 2204 2205
			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) {
2206 2207 2208
				i40e_aqc_add_filters(vsi, vsi_name, add_list,
						     add_head, num_add,
						     &promisc_changed);
2209
				memset(add_list, 0, list_size);
J
Jesse Brandeburg 已提交
2210 2211 2212 2213
				num_add = 0;
			}
		}
		if (num_add) {
2214 2215
			i40e_aqc_add_filters(vsi, vsi_name, add_list, add_head,
					     num_add, &promisc_changed);
J
Jesse Brandeburg 已提交
2216
		}
2217 2218 2219
		/* Now move all of the filters from the temp add list back to
		 * the VSI's list.
		 */
2220
		spin_lock_bh(&vsi->mac_filter_hash_lock);
2221 2222 2223 2224 2225 2226
		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);
2227
		}
2228
		spin_unlock_bh(&vsi->mac_filter_hash_lock);
J
Jesse Brandeburg 已提交
2229 2230
		kfree(add_list);
		add_list = NULL;
2231
	}
J
Jesse Brandeburg 已提交
2232

2233 2234 2235 2236 2237 2238 2239 2240 2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253
	/* 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.
	 */
2254
	if (test_bit(__I40E_VSI_OVERFLOW_PROMISC, vsi->state) &&
2255
	    !promisc_changed && !failed_filters &&
2256
	    (vsi->active_filters < vsi->promisc_threshold)) {
2257 2258 2259
		dev_info(&pf->pdev->dev,
			 "filter logjam cleared on %s, leaving overflow promiscuous mode\n",
			 vsi_name);
2260
		clear_bit(__I40E_VSI_OVERFLOW_PROMISC, vsi->state);
2261 2262
		promisc_changed = true;
		vsi->promisc_threshold = 0;
J
Jesse Brandeburg 已提交
2263 2264
	}

2265 2266
	/* if the VF is not trusted do not do promisc */
	if ((vsi->type == I40E_VSI_SRIOV) && !pf->vf[vsi->vf_id].trusted) {
2267
		clear_bit(__I40E_VSI_OVERFLOW_PROMISC, vsi->state);
2268 2269 2270
		goto out;
	}

J
Jesse Brandeburg 已提交
2271 2272 2273
	/* check for changes in promiscuous modes */
	if (changed_flags & IFF_ALLMULTI) {
		bool cur_multipromisc;
J
Jesse Brandeburg 已提交
2274

J
Jesse Brandeburg 已提交
2275
		cur_multipromisc = !!(vsi->current_netdev_flags & IFF_ALLMULTI);
M
Mitch Williams 已提交
2276 2277 2278 2279 2280 2281
		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,
2282
						     hw->aq.asq_last_status);
J
Jesse Brandeburg 已提交
2283
			dev_info(&pf->pdev->dev,
2284 2285
				 "set multi promisc failed on %s, err %s aq_err %s\n",
				 vsi_name,
2286 2287
				 i40e_stat_str(hw, aq_ret),
				 i40e_aq_str(hw, hw->aq.asq_last_status));
M
Mitch Williams 已提交
2288
		}
J
Jesse Brandeburg 已提交
2289
	}
2290 2291

	if ((changed_flags & IFF_PROMISC) || promisc_changed) {
J
Jesse Brandeburg 已提交
2292
		bool cur_promisc;
J
Jesse Brandeburg 已提交
2293

J
Jesse Brandeburg 已提交
2294
		cur_promisc = (!!(vsi->current_netdev_flags & IFF_PROMISC) ||
2295
			       test_bit(__I40E_VSI_OVERFLOW_PROMISC,
2296
					vsi->state));
2297 2298 2299
		if ((vsi->type == I40E_VSI_MAIN) &&
		    (pf->lan_veb != I40E_NO_VEB) &&
		    !(pf->flags & I40E_FLAG_MFP_ENABLED)) {
2300 2301 2302 2303 2304 2305 2306
			/* 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;
2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320
				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,
2321 2322
						 "Set default VSI failed on %s, err %s, aq_err %s\n",
						 vsi_name,
2323 2324 2325 2326
						 i40e_stat_str(hw, aq_ret),
						 i40e_aq_str(hw,
						     hw->aq.asq_last_status));
				}
2327 2328
			}
		} else {
M
Mitch Williams 已提交
2329
			aq_ret = i40e_aq_set_vsi_unicast_promiscuous(
2330
							  hw,
2331
							  vsi->seid,
2332 2333
							  cur_promisc, NULL,
							  true);
M
Mitch Williams 已提交
2334 2335 2336
			if (aq_ret) {
				retval =
				i40e_aq_rc_to_posix(aq_ret,
2337
						    hw->aq.asq_last_status);
2338
				dev_info(&pf->pdev->dev,
2339 2340
					 "set unicast promisc failed on %s, err %s, aq_err %s\n",
					 vsi_name,
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 2345
			}
			aq_ret = i40e_aq_set_vsi_multicast_promiscuous(
2346
							  hw,
2347 2348
							  vsi->seid,
							  cur_promisc, NULL);
M
Mitch Williams 已提交
2349 2350 2351
			if (aq_ret) {
				retval =
				i40e_aq_rc_to_posix(aq_ret,
2352
						    hw->aq.asq_last_status);
2353
				dev_info(&pf->pdev->dev,
2354 2355
					 "set multicast promisc failed on %s, err %s, aq_err %s\n",
					 vsi_name,
2356 2357 2358
					 i40e_stat_str(hw, aq_ret),
					 i40e_aq_str(hw,
						     hw->aq.asq_last_status));
M
Mitch Williams 已提交
2359
			}
2360
		}
M
Mitch Williams 已提交
2361 2362 2363 2364 2365 2366
		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);
2367
			dev_info(&pf->pdev->dev,
2368
				 "set brdcast promisc failed, err %s, aq_err %s\n",
2369 2370 2371
					 i40e_stat_str(hw, aq_ret),
					 i40e_aq_str(hw,
						     hw->aq.asq_last_status));
M
Mitch Williams 已提交
2372
		}
J
Jesse Brandeburg 已提交
2373
	}
M
Mitch Williams 已提交
2374
out:
2375 2376 2377 2378
	/* if something went wrong then set the changed flag so we try again */
	if (retval)
		vsi->flags |= I40E_VSI_FLAG_FILTER_CHANGED;

2379
	clear_bit(__I40E_VSI_SYNCING_FILTERS, vsi->state);
M
Mitch Williams 已提交
2380
	return retval;
2381 2382 2383 2384

err_no_memory:
	/* Restore elements on the temporary add and delete lists */
	spin_lock_bh(&vsi->mac_filter_hash_lock);
2385
err_no_memory_locked:
2386 2387
	i40e_undo_del_filter_entries(vsi, &tmp_del_list);
	i40e_undo_add_filter_entries(vsi, &tmp_add_list);
2388 2389 2390
	spin_unlock_bh(&vsi->mac_filter_hash_lock);

	vsi->flags |= I40E_VSI_FLAG_FILTER_CHANGED;
2391
	clear_bit(__I40E_VSI_SYNCING_FILTERS, vsi->state);
2392
	return -ENOMEM;
J
Jesse Brandeburg 已提交
2393 2394 2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405 2406
}

/**
 * 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
2407
	for (v = 0; v < pf->num_alloc_vsi; v++) {
J
Jesse Brandeburg 已提交
2408
		if (pf->vsi[v] &&
J
Jesse Brandeburg 已提交
2409 2410 2411 2412 2413 2414 2415 2416 2417
		    (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
2418 2419 2420
	}
}

2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432
/**
 * i40e_max_xdp_frame_size - returns the maximum allowed frame size for XDP
 * @vsi: the vsi
 **/
static int i40e_max_xdp_frame_size(struct i40e_vsi *vsi)
{
	if (PAGE_SIZE >= 8192 || (vsi->back->flags & I40E_FLAG_LEGACY_RX))
		return I40E_RXBUFFER_2048;
	else
		return I40E_RXBUFFER_3072;
}

J
Jesse Brandeburg 已提交
2433 2434 2435 2436 2437 2438 2439 2440 2441 2442 2443
/**
 * 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
2444
	struct i40e_pf *pf = vsi->back;
J
Jesse Brandeburg 已提交
2445

2446 2447 2448 2449 2450 2451 2452
	if (i40e_enabled_xdp_vsi(vsi)) {
		int frame_size = new_mtu + ETH_HLEN + ETH_FCS_LEN + VLAN_HLEN;

		if (frame_size > i40e_max_xdp_frame_size(vsi))
			return -EINVAL;
	}

J
Jesse Brandeburg 已提交
2453 2454 2455 2456 2457
	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
2458 2459
	pf->flags |= (I40E_FLAG_SERVICE_CLIENT_REQUESTED |
		      I40E_FLAG_CLIENT_L2_CHANGE);
J
Jesse Brandeburg 已提交
2460 2461 2462
	return 0;
}

J
Jacob Keller 已提交
2463 2464 2465 2466 2467 2468 2469 2470 2471 2472 2473 2474 2475 2476 2477 2478 2479 2480 2481 2482 2483
/**
 * 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
2484 2485 2486 2487 2488 2489 2490 2491 2492 2493 2494 2495 2496 2497 2498 2499 2500 2501 2502
/**
 * 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;
2503
	ctxt.info = vsi->info;
J
Jesse Brandeburg 已提交
2504 2505 2506
	ret = i40e_aq_update_vsi_params(&vsi->back->hw, &ctxt, NULL);
	if (ret) {
		dev_info(&vsi->back->pdev->dev,
2507 2508 2509 2510
			 "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
2511 2512 2513 2514 2515 2516 2517 2518 2519 2520 2521 2522 2523 2524 2525 2526 2527 2528 2529 2530 2531 2532 2533
	}
}

/**
 * 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;
2534
	ctxt.info = vsi->info;
J
Jesse Brandeburg 已提交
2535 2536 2537
	ret = i40e_aq_update_vsi_params(&vsi->back->hw, &ctxt, NULL);
	if (ret) {
		dev_info(&vsi->back->pdev->dev,
2538 2539 2540 2541
			 "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
2542 2543 2544 2545 2546 2547 2548 2549 2550 2551 2552 2553 2554 2555 2556 2557 2558 2559 2560 2561
	}
}

/**
 * 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);
}

/**
2562
 * i40e_add_vlan_all_mac - Add a MAC/VLAN filter for each existing MAC address
J
Jesse Brandeburg 已提交
2563 2564
 * @vsi: the vsi being configured
 * @vid: vlan id to be added (0 = untagged only , -1 = any)
2565 2566 2567 2568 2569 2570 2571 2572
 *
 * 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
2573
 **/
2574
int i40e_add_vlan_all_mac(struct i40e_vsi *vsi, s16 vid)
J
Jesse Brandeburg 已提交
2575
{
2576
	struct i40e_mac_filter *f, *add_f;
2577 2578
	struct hlist_node *h;
	int bkt;
J
Jesse Brandeburg 已提交
2579

2580
	hash_for_each_safe(vsi->mac_filter_hash, bkt, h, f, hlist) {
2581 2582
		if (f->state == I40E_FILTER_REMOVE)
			continue;
2583
		add_f = i40e_add_filter(vsi, f->macaddr, vid);
J
Jesse Brandeburg 已提交
2584 2585 2586 2587 2588 2589 2590 2591
		if (!add_f) {
			dev_info(&vsi->back->pdev->dev,
				 "Could not add vlan filter %d for %pM\n",
				 vid, f->macaddr);
			return -ENOMEM;
		}
	}

2592 2593 2594 2595 2596 2597
	return 0;
}

/**
 * i40e_vsi_add_vlan - Add VSI membership for given VLAN
 * @vsi: the VSI being configured
2598
 * @vid: VLAN id to be added
2599
 **/
2600
int i40e_vsi_add_vlan(struct i40e_vsi *vsi, u16 vid)
2601
{
2602
	int err;
2603

2604
	if (vsi->info.pvid)
2605 2606
		return -EINVAL;

2607 2608 2609 2610 2611 2612 2613 2614 2615 2616 2617
	/* The network stack will attempt to add VID=0, with the intention to
	 * receive priority tagged packets with a VLAN of 0. Our HW receives
	 * these packets by default when configured to receive untagged
	 * packets, so we don't need to add a filter for this case.
	 * Additionally, HW interprets adding a VID=0 filter as meaning to
	 * receive *only* tagged traffic and stops receiving untagged traffic.
	 * Thus, we do not want to actually add a filter for VID=0
	 */
	if (!vid)
		return 0;

2618 2619 2620
	/* 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);
2621
	spin_unlock_bh(&vsi->mac_filter_hash_lock);
2622 2623
	if (err)
		return err;
K
Kiran Patil 已提交
2624

2625 2626 2627 2628 2629
	/* 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
2630 2631 2632
}

/**
2633
 * i40e_rm_vlan_all_mac - Remove MAC/VLAN pair for all MAC with the given VLAN
J
Jesse Brandeburg 已提交
2634 2635
 * @vsi: the vsi being configured
 * @vid: vlan id to be removed (0 = untagged only , -1 = any)
2636 2637 2638 2639 2640 2641 2642 2643 2644
 *
 * 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
 */
2645
void i40e_rm_vlan_all_mac(struct i40e_vsi *vsi, s16 vid)
J
Jesse Brandeburg 已提交
2646
{
2647
	struct i40e_mac_filter *f;
2648 2649
	struct hlist_node *h;
	int bkt;
J
Jesse Brandeburg 已提交
2650

2651
	hash_for_each_safe(vsi->mac_filter_hash, bkt, h, f, hlist) {
2652 2653 2654
		if (f->vlan == vid)
			__i40e_del_filter(vsi, f);
	}
2655
}
J
Jesse Brandeburg 已提交
2656

2657 2658 2659
/**
 * i40e_vsi_kill_vlan - Remove VSI membership for given VLAN
 * @vsi: the VSI being configured
2660
 * @vid: VLAN id to be removed
2661
 **/
2662
void i40e_vsi_kill_vlan(struct i40e_vsi *vsi, u16 vid)
2663
{
2664 2665 2666
	if (!vid || vsi->info.pvid)
		return;

2667 2668
	spin_lock_bh(&vsi->mac_filter_hash_lock);
	i40e_rm_vlan_all_mac(vsi, vid);
2669
	spin_unlock_bh(&vsi->mac_filter_hash_lock);
K
Kiran Patil 已提交
2670

2671 2672 2673 2674
	/* 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
2675 2676 2677 2678 2679 2680
}

/**
 * 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
2681 2682
 *
 * net_device_ops implementation for adding vlan ids
J
Jesse Brandeburg 已提交
2683 2684 2685 2686 2687 2688
 **/
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
2689
	int ret = 0;
J
Jesse Brandeburg 已提交
2690

2691
	if (vid >= VLAN_N_VID)
J
Jesse Brandeburg 已提交
2692 2693
		return -EINVAL;

2694
	ret = i40e_vsi_add_vlan(vsi, vid);
2695
	if (!ret)
J
Jesse Brandeburg 已提交
2696
		set_bit(vid, vsi->active_vlans);
J
Jesse Brandeburg 已提交
2697

J
Jesse Brandeburg 已提交
2698
	return ret;
J
Jesse Brandeburg 已提交
2699 2700 2701 2702 2703 2704
}

/**
 * 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
2705
 *
A
Akeem G Abodunrin 已提交
2706
 * net_device_ops implementation for removing vlan ids
J
Jesse Brandeburg 已提交
2707 2708 2709 2710 2711 2712 2713 2714 2715
 **/
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
2716
	 * already printed from the other function
J
Jesse Brandeburg 已提交
2717 2718 2719 2720
	 */
	i40e_vsi_kill_vlan(vsi, vid);

	clear_bit(vid, vsi->active_vlans);
J
Jesse Brandeburg 已提交
2721

J
Jesse Brandeburg 已提交
2722 2723 2724 2725 2726 2727 2728 2729 2730 2731 2732 2733 2734 2735 2736 2737 2738 2739 2740 2741 2742 2743 2744 2745 2746 2747
	return 0;
}

/**
 * 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
2748
int i40e_vsi_add_pvid(struct i40e_vsi *vsi, u16 vid)
J
Jesse Brandeburg 已提交
2749 2750
{
	struct i40e_vsi_context ctxt;
2751
	i40e_status ret;
J
Jesse Brandeburg 已提交
2752 2753 2754

	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
2755 2756
	vsi->info.port_vlan_flags = I40E_AQ_VSI_PVLAN_MODE_TAGGED |
				    I40E_AQ_VSI_PVLAN_INSERT_PVID |
G
Greg Rose 已提交
2757
				    I40E_AQ_VSI_PVLAN_EMOD_STR;
J
Jesse Brandeburg 已提交
2758 2759

	ctxt.seid = vsi->seid;
2760
	ctxt.info = vsi->info;
2761 2762
	ret = i40e_aq_update_vsi_params(&vsi->back->hw, &ctxt, NULL);
	if (ret) {
J
Jesse Brandeburg 已提交
2763
		dev_info(&vsi->back->pdev->dev,
2764 2765 2766 2767
			 "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
2768
		return -ENOENT;
J
Jesse Brandeburg 已提交
2769 2770
	}

J
Jesse Brandeburg 已提交
2771
	return 0;
J
Jesse Brandeburg 已提交
2772 2773 2774 2775 2776 2777 2778 2779 2780 2781
}

/**
 * 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
2782 2783
	i40e_vlan_stripping_disable(vsi);

J
Jesse Brandeburg 已提交
2784 2785 2786 2787 2788 2789 2790 2791 2792 2793 2794 2795 2796 2797 2798 2799 2800 2801
	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++)
2802
		err = i40e_setup_tx_descriptors(vsi->tx_rings[i]);
J
Jesse Brandeburg 已提交
2803

2804 2805 2806 2807 2808 2809
	if (!i40e_enabled_xdp_vsi(vsi))
		return err;

	for (i = 0; i < vsi->num_queue_pairs && !err; i++)
		err = i40e_setup_tx_descriptors(vsi->xdp_rings[i]);

J
Jesse Brandeburg 已提交
2810 2811 2812 2813 2814 2815 2816 2817 2818 2819 2820 2821 2822
	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;

2823 2824 2825 2826 2827
	if (vsi->tx_rings) {
		for (i = 0; i < vsi->num_queue_pairs; i++)
			if (vsi->tx_rings[i] && vsi->tx_rings[i]->desc)
				i40e_free_tx_resources(vsi->tx_rings[i]);
	}
2828

2829 2830 2831 2832 2833
	if (vsi->xdp_rings) {
		for (i = 0; i < vsi->num_queue_pairs; i++)
			if (vsi->xdp_rings[i] && vsi->xdp_rings[i]->desc)
				i40e_free_tx_resources(vsi->xdp_rings[i]);
	}
J
Jesse Brandeburg 已提交
2834 2835 2836 2837 2838 2839 2840 2841 2842 2843 2844 2845 2846 2847 2848 2849 2850
}

/**
 * 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++)
2851
		err = i40e_setup_rx_descriptors(vsi->rx_rings[i]);
J
Jesse Brandeburg 已提交
2852 2853 2854 2855 2856 2857 2858 2859 2860 2861 2862 2863 2864
	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;

2865 2866 2867
	if (!vsi->rx_rings)
		return;

J
Jesse Brandeburg 已提交
2868
	for (i = 0; i < vsi->num_queue_pairs; i++)
2869
		if (vsi->rx_rings[i] && vsi->rx_rings[i]->desc)
2870
			i40e_free_rx_resources(vsi->rx_rings[i]);
J
Jesse Brandeburg 已提交
2871 2872
}

2873 2874 2875 2876 2877 2878 2879 2880 2881 2882
/**
 * 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;
2883
	int cpu;
2884

2885 2886 2887
	if (!ring->q_vector || !ring->netdev)
		return;

2888
	if ((vsi->tc_config.numtc <= 1) &&
2889
	    !test_and_set_bit(__I40E_TX_XPS_INIT_DONE, ring->state)) {
2890 2891
		cpu = cpumask_local_spread(ring->q_vector->v_idx, -1);
		netif_set_xps_queue(ring->netdev, get_cpu_mask(cpu),
2892
				    ring->queue_index);
2893
	}
2894 2895 2896 2897 2898

	/* schedule our worker thread which will take care of
	 * applying the new filter changes
	 */
	i40e_service_event_schedule(vsi->back);
2899 2900
}

J
Jesse Brandeburg 已提交
2901 2902 2903 2904 2905 2906 2907 2908 2909 2910 2911 2912 2913 2914 2915 2916
/**
 * 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
2917
	if (vsi->back->flags & I40E_FLAG_FD_ATR_ENABLED) {
J
Jesse Brandeburg 已提交
2918 2919 2920 2921 2922 2923
		ring->atr_sample_rate = vsi->back->atr_sample_rate;
		ring->atr_count = 0;
	} else {
		ring->atr_sample_rate = 0;
	}

2924 2925
	/* configure XPS */
	i40e_config_xps_tx_ring(ring);
J
Jesse Brandeburg 已提交
2926 2927 2928 2929 2930 2931 2932

	/* 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
2933 2934
	tx_ctx.fd_ena = !!(vsi->back->flags & (I40E_FLAG_FD_SB_ENABLED |
					       I40E_FLAG_FD_ATR_ENABLED));
J
Jacob Keller 已提交
2935
	tx_ctx.timesync_ena = !!(vsi->back->flags & I40E_FLAG_PTP);
2936 2937 2938 2939 2940
	/* 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
2941 2942 2943 2944 2945 2946 2947 2948 2949 2950 2951 2952 2953 2954 2955 2956 2957 2958 2959 2960 2961 2962 2963 2964 2965 2966 2967 2968 2969 2970 2971 2972 2973

	/* 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
2974
	if (vsi->type == I40E_VSI_VMDQ2) {
2975
		qtx_ctl = I40E_QTX_CTL_VM_QUEUE;
M
Mitch Williams 已提交
2976 2977 2978
		qtx_ctl |= ((vsi->id) << I40E_QTX_CTL_VFVM_INDX_SHIFT) &
			   I40E_QTX_CTL_VFVM_INDX_MASK;
	} else {
2979
		qtx_ctl = I40E_QTX_CTL_PF_QUEUE;
M
Mitch Williams 已提交
2980 2981
	}

2982 2983
	qtx_ctl |= ((hw->pf_id << I40E_QTX_CTL_PF_INDX_SHIFT) &
		    I40E_QTX_CTL_PF_INDX_MASK);
J
Jesse Brandeburg 已提交
2984 2985 2986 2987 2988 2989 2990 2991 2992 2993 2994 2995 2996 2997 2998 2999 3000 3001 3002 3003 3004 3005 3006 3007
	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;

3008
	bitmap_zero(ring->state, __I40E_RING_STATE_NBITS);
J
Jesse Brandeburg 已提交
3009 3010 3011 3012 3013 3014

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

	ring->rx_buf_len = vsi->rx_buf_len;

3015 3016
	rx_ctx.dbuff = DIV_ROUND_UP(ring->rx_buf_len,
				    BIT_ULL(I40E_RXQ_CTX_DBUFF_SHIFT));
J
Jesse Brandeburg 已提交
3017 3018 3019 3020

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

3021 3022
	/* use 32 byte descriptors */
	rx_ctx.dsize = 1;
J
Jesse Brandeburg 已提交
3023

3024 3025 3026
	/* descriptor type is always zero
	 * rx_ctx.dtype = 0;
	 */
3027
	rx_ctx.hsplit_0 = 0;
J
Jesse Brandeburg 已提交
3028

3029
	rx_ctx.rxmax = min_t(u16, vsi->max_frame, chain_len * ring->rx_buf_len);
3030 3031 3032 3033
	if (hw->revision_id == 0)
		rx_ctx.lrxqthresh = 0;
	else
		rx_ctx.lrxqthresh = 2;
J
Jesse Brandeburg 已提交
3034 3035
	rx_ctx.crcstrip = 1;
	rx_ctx.l2tsel = 1;
J
Jesse Brandeburg 已提交
3036 3037
	/* this controls whether VLAN is stripped from inner headers */
	rx_ctx.showiv = 0;
3038 3039
	/* set the prefena field to 1 because the manual says to */
	rx_ctx.prefena = 1;
J
Jesse Brandeburg 已提交
3040 3041 3042 3043 3044 3045 3046 3047 3048 3049 3050 3051 3052 3053 3054 3055 3056 3057 3058

	/* 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;
	}

3059 3060 3061 3062 3063 3064
	/* configure Rx buffer alignment */
	if (!vsi->netdev || (vsi->back->flags & I40E_FLAG_LEGACY_RX))
		clear_ring_build_skb_enabled(ring);
	else
		set_ring_build_skb_enabled(ring);

J
Jesse Brandeburg 已提交
3065 3066 3067 3068
	/* 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);

3069
	i40e_alloc_rx_buffers(ring, I40E_DESC_UNUSED(ring));
J
Jesse Brandeburg 已提交
3070 3071 3072 3073 3074 3075 3076 3077 3078 3079 3080 3081 3082 3083 3084

	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;

3085 3086
	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
3087

3088 3089 3090 3091 3092 3093
	if (!i40e_enabled_xdp_vsi(vsi))
		return err;

	for (i = 0; (i < vsi->num_queue_pairs) && !err; i++)
		err = i40e_configure_tx_ring(vsi->xdp_rings[i]);

J
Jesse Brandeburg 已提交
3094 3095 3096 3097 3098 3099 3100 3101 3102 3103 3104 3105 3106 3107
	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;

3108 3109 3110 3111
	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)
3112 3113
	} else if (!I40E_2K_TOO_SMALL_WITH_PADDING &&
		   (vsi->netdev->mtu <= ETH_DATA_LEN)) {
3114 3115 3116 3117 3118
		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;
3119 3120
		vsi->rx_buf_len = (PAGE_SIZE < 8192) ? I40E_RXBUFFER_3072 :
						       I40E_RXBUFFER_2048;
3121
	}
J
Jesse Brandeburg 已提交
3122 3123 3124

	/* set up individual rings */
	for (i = 0; i < vsi->num_queue_pairs && !err; i++)
3125
		err = i40e_configure_rx_ring(vsi->rx_rings[i]);
J
Jesse Brandeburg 已提交
3126 3127 3128 3129 3130 3131 3132 3133 3134 3135

	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)
{
3136
	struct i40e_ring *tx_ring, *rx_ring;
J
Jesse Brandeburg 已提交
3137 3138 3139
	u16 qoffset, qcount;
	int i, n;

3140 3141 3142 3143 3144 3145 3146 3147 3148
	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
3149 3150

	for (n = 0; n < I40E_MAX_TRAFFIC_CLASS; n++) {
3151
		if (!(vsi->tc_config.enabled_tc & BIT_ULL(n)))
J
Jesse Brandeburg 已提交
3152 3153 3154 3155 3156
			continue;

		qoffset = vsi->tc_config.tc_info[n].qoffset;
		qcount = vsi->tc_config.tc_info[n].qcount;
		for (i = qoffset; i < (qoffset + qcount); i++) {
3157 3158
			rx_ring = vsi->rx_rings[i];
			tx_ring = vsi->tx_rings[i];
J
Jesse Brandeburg 已提交
3159 3160 3161 3162 3163 3164 3165 3166 3167 3168 3169 3170 3171 3172 3173 3174
			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)
{
	if (vsi->netdev)
		i40e_set_rx_mode(vsi->netdev);
}

3175 3176 3177 3178 3179 3180 3181 3182 3183 3184 3185 3186 3187
/**
 * 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;

3188 3189 3190
	if (!(pf->flags & I40E_FLAG_FD_SB_ENABLED))
		return;

3191
	/* Reset FDir counters as we're replaying all existing filters */
3192 3193
	pf->fd_tcp4_filter_cnt = 0;
	pf->fd_udp4_filter_cnt = 0;
3194
	pf->fd_sctp4_filter_cnt = 0;
3195
	pf->fd_ip4_filter_cnt = 0;
3196

3197 3198 3199 3200 3201 3202
	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
3203 3204 3205 3206 3207 3208 3209 3210 3211 3212 3213 3214 3215 3216 3217 3218 3219 3220 3221 3222 3223 3224 3225 3226
/**
 * 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)
{
3227
	bool has_xdp = i40e_enabled_xdp_vsi(vsi);
J
Jesse Brandeburg 已提交
3228 3229 3230 3231 3232 3233 3234 3235 3236 3237 3238 3239
	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;
3240
	for (i = 0; i < vsi->num_q_vectors; i++, vector++) {
3241 3242
		struct i40e_q_vector *q_vector = vsi->q_vectors[i];

3243
		q_vector->itr_countdown = ITR_COUNTDOWN_START;
3244
		q_vector->rx.itr = ITR_TO_REG(vsi->rx_rings[i]->rx_itr_setting);
J
Jesse Brandeburg 已提交
3245 3246 3247
		q_vector->rx.latency_range = I40E_LOW_LATENCY;
		wr32(hw, I40E_PFINT_ITRN(I40E_RX_ITR, vector - 1),
		     q_vector->rx.itr);
3248
		q_vector->tx.itr = ITR_TO_REG(vsi->tx_rings[i]->tx_itr_setting);
J
Jesse Brandeburg 已提交
3249 3250 3251
		q_vector->tx.latency_range = I40E_LOW_LATENCY;
		wr32(hw, I40E_PFINT_ITRN(I40E_TX_ITR, vector - 1),
		     q_vector->tx.itr);
3252
		wr32(hw, I40E_PFINT_RATEN(vector - 1),
3253
		     i40e_intrl_usec_to_reg(vsi->int_rate_limit));
J
Jesse Brandeburg 已提交
3254 3255 3256 3257

		/* 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++) {
3258
			u32 nextqp = has_xdp ? qp + vsi->alloc_queue_pairs : qp;
3259 3260
			u32 val;

J
Jesse Brandeburg 已提交
3261
			val = I40E_QINT_RQCTL_CAUSE_ENA_MASK |
3262 3263 3264 3265 3266
			      (I40E_RX_ITR << I40E_QINT_RQCTL_ITR_INDX_SHIFT) |
			      (vector << I40E_QINT_RQCTL_MSIX_INDX_SHIFT) |
			      (nextqp << I40E_QINT_RQCTL_NEXTQ_INDX_SHIFT) |
			      (I40E_QUEUE_TYPE_TX <<
			       I40E_QINT_RQCTL_NEXTQ_TYPE_SHIFT);
J
Jesse Brandeburg 已提交
3267 3268 3269

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

3270 3271 3272 3273 3274 3275 3276 3277 3278 3279 3280
			if (has_xdp) {
				val = I40E_QINT_TQCTL_CAUSE_ENA_MASK |
				      (I40E_TX_ITR << I40E_QINT_TQCTL_ITR_INDX_SHIFT) |
				      (vector << I40E_QINT_TQCTL_MSIX_INDX_SHIFT) |
				      (qp << I40E_QINT_TQCTL_NEXTQ_INDX_SHIFT) |
				      (I40E_QUEUE_TYPE_TX <<
				       I40E_QINT_TQCTL_NEXTQ_TYPE_SHIFT);

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

J
Jesse Brandeburg 已提交
3281
			val = I40E_QINT_TQCTL_CAUSE_ENA_MASK |
3282 3283 3284 3285 3286
			      (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);
J
Jesse Brandeburg 已提交
3287 3288 3289

			/* Terminate the linked list */
			if (q == (q_vector->num_ringpairs - 1))
3290 3291
				val |= (I40E_QUEUE_END_OF_LIST <<
					I40E_QINT_TQCTL_NEXTQ_INDX_SHIFT);
J
Jesse Brandeburg 已提交
3292 3293 3294 3295 3296 3297 3298 3299 3300 3301 3302 3303 3304

			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
 **/
3305
static void i40e_enable_misc_int_causes(struct i40e_pf *pf)
J
Jesse Brandeburg 已提交
3306
{
3307
	struct i40e_hw *hw = &pf->hw;
J
Jesse Brandeburg 已提交
3308 3309 3310 3311 3312 3313 3314 3315 3316 3317 3318 3319 3320 3321 3322
	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;

3323 3324 3325
	if (pf->flags & I40E_FLAG_IWARP_ENABLED)
		val |= I40E_PFINT_ICR0_ENA_PE_CRITERR_MASK;

3326 3327 3328
	if (pf->flags & I40E_FLAG_PTP)
		val |= I40E_PFINT_ICR0_ENA_TIMESYNC_MASK;

J
Jesse Brandeburg 已提交
3329 3330 3331
	wr32(hw, I40E_PFINT_ICR0_ENA, val);

	/* SW_ITR_IDX = 0, but don't change INTENA */
3332 3333
	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
3334 3335 3336 3337 3338 3339 3340 3341 3342 3343 3344

	/* 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)
{
3345
	u32 nextqp = i40e_enabled_xdp_vsi(vsi) ? vsi->alloc_queue_pairs : 0;
3346
	struct i40e_q_vector *q_vector = vsi->q_vectors[0];
J
Jesse Brandeburg 已提交
3347 3348 3349 3350 3351
	struct i40e_pf *pf = vsi->back;
	struct i40e_hw *hw = &pf->hw;
	u32 val;

	/* set the ITR configuration */
3352
	q_vector->itr_countdown = ITR_COUNTDOWN_START;
3353
	q_vector->rx.itr = ITR_TO_REG(vsi->rx_rings[0]->rx_itr_setting);
J
Jesse Brandeburg 已提交
3354 3355
	q_vector->rx.latency_range = I40E_LOW_LATENCY;
	wr32(hw, I40E_PFINT_ITR0(I40E_RX_ITR), q_vector->rx.itr);
3356
	q_vector->tx.itr = ITR_TO_REG(vsi->tx_rings[0]->tx_itr_setting);
J
Jesse Brandeburg 已提交
3357 3358 3359
	q_vector->tx.latency_range = I40E_LOW_LATENCY;
	wr32(hw, I40E_PFINT_ITR0(I40E_TX_ITR), q_vector->tx.itr);

3360
	i40e_enable_misc_int_causes(pf);
J
Jesse Brandeburg 已提交
3361 3362 3363 3364

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

3365
	/* Associate the queue pair to the vector and enable the queue int */
3366 3367 3368
	val = I40E_QINT_RQCTL_CAUSE_ENA_MASK		       |
	      (I40E_RX_ITR << I40E_QINT_RQCTL_ITR_INDX_SHIFT)  |
	      (nextqp	   << I40E_QINT_RQCTL_NEXTQ_INDX_SHIFT)|
J
Jesse Brandeburg 已提交
3369 3370 3371 3372
	      (I40E_QUEUE_TYPE_TX << I40E_QINT_TQCTL_NEXTQ_TYPE_SHIFT);

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

3373 3374 3375 3376 3377 3378 3379 3380 3381
	if (i40e_enabled_xdp_vsi(vsi)) {
		val = I40E_QINT_TQCTL_CAUSE_ENA_MASK		     |
		      (I40E_TX_ITR << I40E_QINT_TQCTL_ITR_INDX_SHIFT)|
		      (I40E_QUEUE_TYPE_TX
		       << I40E_QINT_TQCTL_NEXTQ_TYPE_SHIFT);

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

J
Jesse Brandeburg 已提交
3382 3383 3384 3385 3386 3387 3388 3389
	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);
}

3390 3391 3392 3393 3394 3395 3396 3397 3398 3399 3400 3401 3402
/**
 * 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
3403 3404 3405
/**
 * i40e_irq_dynamic_enable_icr0 - Enable default interrupt generation for icr0
 * @pf: board private structure
3406
 * @clearpba: true when all pending interrupt events should be cleared
J
Jesse Brandeburg 已提交
3407
 **/
3408
void i40e_irq_dynamic_enable_icr0(struct i40e_pf *pf, bool clearpba)
J
Jesse Brandeburg 已提交
3409 3410 3411 3412 3413
{
	struct i40e_hw *hw = &pf->hw;
	u32 val;

	val = I40E_PFINT_DYN_CTL0_INTENA_MASK   |
3414
	      (clearpba ? I40E_PFINT_DYN_CTL0_CLEARPBA_MASK : 0) |
J
Jesse Brandeburg 已提交
3415 3416 3417 3418 3419 3420 3421 3422 3423 3424 3425 3426 3427 3428 3429
	      (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;

3430
	if (!q_vector->tx.ring && !q_vector->rx.ring)
J
Jesse Brandeburg 已提交
3431 3432
		return IRQ_HANDLED;

3433
	napi_schedule_irqoff(&q_vector->napi);
J
Jesse Brandeburg 已提交
3434 3435 3436 3437

	return IRQ_HANDLED;
}

3438 3439 3440 3441 3442 3443 3444 3445 3446 3447 3448 3449 3450 3451
/**
 * 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);

3452
	cpumask_copy(&q_vector->affinity_mask, mask);
3453 3454 3455 3456 3457 3458 3459 3460 3461 3462 3463 3464
}

/**
 * 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
3465 3466 3467 3468 3469 3470 3471 3472 3473 3474 3475 3476 3477 3478 3479
/**
 * 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;
3480
	int irq_num;
3481
	int cpu;
J
Jesse Brandeburg 已提交
3482 3483

	for (vector = 0; vector < q_vectors; vector++) {
3484
		struct i40e_q_vector *q_vector = vsi->q_vectors[vector];
J
Jesse Brandeburg 已提交
3485

3486 3487
		irq_num = pf->msix_entries[base + vector].vector;

3488
		if (q_vector->tx.ring && q_vector->rx.ring) {
J
Jesse Brandeburg 已提交
3489 3490 3491
			snprintf(q_vector->name, sizeof(q_vector->name) - 1,
				 "%s-%s-%d", basename, "TxRx", rx_int_idx++);
			tx_int_idx++;
3492
		} else if (q_vector->rx.ring) {
J
Jesse Brandeburg 已提交
3493 3494
			snprintf(q_vector->name, sizeof(q_vector->name) - 1,
				 "%s-%s-%d", basename, "rx", rx_int_idx++);
3495
		} else if (q_vector->tx.ring) {
J
Jesse Brandeburg 已提交
3496 3497 3498 3499 3500 3501
			snprintf(q_vector->name, sizeof(q_vector->name) - 1,
				 "%s-%s-%d", basename, "tx", tx_int_idx++);
		} else {
			/* skip this unused q_vector */
			continue;
		}
3502
		err = request_irq(irq_num,
J
Jesse Brandeburg 已提交
3503 3504 3505 3506 3507 3508
				  vsi->irq_handler,
				  0,
				  q_vector->name,
				  q_vector);
		if (err) {
			dev_info(&pf->pdev->dev,
S
Shannon Nelson 已提交
3509
				 "MSIX request_irq failed, error: %d\n", err);
J
Jesse Brandeburg 已提交
3510 3511
			goto free_queue_irqs;
		}
3512 3513 3514 3515 3516

		/* 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);
3517 3518 3519 3520 3521
		/* Spread affinity hints out across online CPUs.
		 *
		 * get_cpu_mask returns a static constant mask with
		 * a permanent lifetime so it's ok to pass to
		 * irq_set_affinity_hint without making a copy.
3522
		 */
3523 3524
		cpu = cpumask_local_spread(q_vector->v_idx, -1);
		irq_set_affinity_hint(irq_num, get_cpu_mask(cpu));
J
Jesse Brandeburg 已提交
3525 3526
	}

3527
	vsi->irqs_ready = true;
J
Jesse Brandeburg 已提交
3528 3529 3530 3531 3532
	return 0;

free_queue_irqs:
	while (vector) {
		vector--;
3533 3534 3535 3536
		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
3537 3538 3539 3540 3541 3542 3543 3544 3545 3546 3547 3548 3549 3550 3551
	}
	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;

S
Shannon Nelson 已提交
3552
	/* disable interrupt causation from each queue */
J
Jesse Brandeburg 已提交
3553
	for (i = 0; i < vsi->num_queue_pairs; i++) {
S
Shannon Nelson 已提交
3554 3555 3556 3557 3558 3559 3560 3561 3562 3563
		u32 val;

		val = rd32(hw, I40E_QINT_TQCTL(vsi->tx_rings[i]->reg_idx));
		val &= ~I40E_QINT_TQCTL_CAUSE_ENA_MASK;
		wr32(hw, I40E_QINT_TQCTL(vsi->tx_rings[i]->reg_idx), val);

		val = rd32(hw, I40E_QINT_RQCTL(vsi->rx_rings[i]->reg_idx));
		val &= ~I40E_QINT_RQCTL_CAUSE_ENA_MASK;
		wr32(hw, I40E_QINT_RQCTL(vsi->rx_rings[i]->reg_idx), val);

3564 3565 3566
		if (!i40e_enabled_xdp_vsi(vsi))
			continue;
		wr32(hw, I40E_QINT_TQCTL(vsi->xdp_rings[i]->reg_idx), 0);
J
Jesse Brandeburg 已提交
3567 3568
	}

S
Shannon Nelson 已提交
3569
	/* disable each interrupt */
J
Jesse Brandeburg 已提交
3570 3571 3572 3573 3574 3575 3576 3577 3578 3579 3580 3581 3582 3583 3584 3585 3586 3587 3588 3589 3590 3591 3592 3593 3594 3595 3596
	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) {
3597
		for (i = 0; i < vsi->num_q_vectors; i++)
J
Jesse Brandeburg 已提交
3598 3599
			i40e_irq_dynamic_enable(vsi, i);
	} else {
3600
		i40e_irq_dynamic_enable_icr0(pf, true);
J
Jesse Brandeburg 已提交
3601 3602
	}

3603
	i40e_flush(&pf->hw);
J
Jesse Brandeburg 已提交
3604 3605 3606 3607
	return 0;
}

/**
3608
 * i40e_free_misc_vector - Free the vector that handles non-queue events
J
Jesse Brandeburg 已提交
3609 3610
 * @pf: board private structure
 **/
3611
static void i40e_free_misc_vector(struct i40e_pf *pf)
J
Jesse Brandeburg 已提交
3612 3613 3614 3615
{
	/* Disable ICR 0 */
	wr32(&pf->hw, I40E_PFINT_ICR0_ENA, 0);
	i40e_flush(&pf->hw);
3616 3617 3618 3619 3620 3621

	if (pf->flags & I40E_FLAG_MSIX_ENABLED && pf->msix_entries) {
		synchronize_irq(pf->msix_entries[0].vector);
		free_irq(pf->msix_entries[0].vector, pf);
		clear_bit(__I40E_MISC_IRQ_REQUESTED, pf->state);
	}
J
Jesse Brandeburg 已提交
3622 3623 3624 3625 3626 3627 3628 3629 3630 3631 3632 3633 3634 3635 3636
}

/**
 * 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;
3637
	irqreturn_t ret = IRQ_NONE;
J
Jesse Brandeburg 已提交
3638 3639 3640 3641
	u32 icr0, icr0_remaining;
	u32 val, ena_mask;

	icr0 = rd32(hw, I40E_PFINT_ICR0);
3642
	ena_mask = rd32(hw, I40E_PFINT_ICR0_ENA);
J
Jesse Brandeburg 已提交
3643

3644 3645
	/* if sharing a legacy IRQ, we might get called w/o an intr pending */
	if ((icr0 & I40E_PFINT_ICR0_INTEVENT_MASK) == 0)
3646
		goto enable_intr;
J
Jesse Brandeburg 已提交
3647

S
Shannon Nelson 已提交
3648 3649 3650 3651 3652
	/* 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++;

3653
	if ((pf->flags & I40E_FLAG_IWARP_ENABLED) &&
3654
	    (icr0 & I40E_PFINT_ICR0_ENA_PE_CRITERR_MASK)) {
3655
		ena_mask &= ~I40E_PFINT_ICR0_ENA_PE_CRITERR_MASK;
3656
		dev_dbg(&pf->pdev->dev, "cleared PE_CRITERR\n");
3657
		set_bit(__I40E_CORE_RESET_REQUESTED, pf->state);
3658 3659
	}

J
Jesse Brandeburg 已提交
3660 3661
	/* only q0 is used in MSI/Legacy mode, and none are used in MSIX */
	if (icr0 & I40E_PFINT_ICR0_QUEUE_0_MASK) {
3662 3663
		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
3664

3665 3666 3667 3668 3669 3670
		/* 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.
		 */
3671
		if (!test_bit(__I40E_DOWN, pf->state))
3672
			napi_schedule_irqoff(&q_vector->napi);
J
Jesse Brandeburg 已提交
3673 3674 3675 3676
	}

	if (icr0 & I40E_PFINT_ICR0_ADMINQ_MASK) {
		ena_mask &= ~I40E_PFINT_ICR0_ENA_ADMINQ_MASK;
3677
		set_bit(__I40E_ADMINQ_EVENT_PENDING, pf->state);
3678
		i40e_debug(&pf->hw, I40E_DEBUG_NVM, "AdminQ event\n");
J
Jesse Brandeburg 已提交
3679 3680 3681 3682
	}

	if (icr0 & I40E_PFINT_ICR0_MAL_DETECT_MASK) {
		ena_mask &= ~I40E_PFINT_ICR0_ENA_MAL_DETECT_MASK;
3683
		set_bit(__I40E_MDD_EVENT_PENDING, pf->state);
J
Jesse Brandeburg 已提交
3684 3685 3686 3687
	}

	if (icr0 & I40E_PFINT_ICR0_VFLR_MASK) {
		ena_mask &= ~I40E_PFINT_ICR0_ENA_VFLR_MASK;
3688
		set_bit(__I40E_VFLR_EVENT_PENDING, pf->state);
J
Jesse Brandeburg 已提交
3689 3690 3691
	}

	if (icr0 & I40E_PFINT_ICR0_GRST_MASK) {
3692 3693
		if (!test_bit(__I40E_RESET_RECOVERY_PENDING, pf->state))
			set_bit(__I40E_RESET_INTR_RECEIVED, pf->state);
J
Jesse Brandeburg 已提交
3694 3695 3696 3697
		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;
3698
		if (val == I40E_RESET_CORER) {
J
Jesse Brandeburg 已提交
3699
			pf->corer_count++;
3700
		} else if (val == I40E_RESET_GLOBR) {
J
Jesse Brandeburg 已提交
3701
			pf->globr_count++;
3702
		} else if (val == I40E_RESET_EMPR) {
J
Jesse Brandeburg 已提交
3703
			pf->empr_count++;
3704
			set_bit(__I40E_EMP_RESET_INTR_RECEIVED, pf->state);
3705
		}
J
Jesse Brandeburg 已提交
3706 3707
	}

3708 3709 3710
	if (icr0 & I40E_PFINT_ICR0_HMC_ERR_MASK) {
		icr0 &= ~I40E_PFINT_ICR0_HMC_ERR_MASK;
		dev_info(&pf->pdev->dev, "HMC error interrupt\n");
3711 3712 3713
		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));
3714 3715
	}

J
Jacob Keller 已提交
3716 3717 3718 3719
	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) {
3720
			icr0 &= ~I40E_PFINT_ICR0_ENA_TIMESYNC_MASK;
J
Jacob Keller 已提交
3721 3722 3723 3724
			i40e_ptp_tx_hwtstamp(pf);
		}
	}

J
Jesse Brandeburg 已提交
3725 3726 3727 3728 3729 3730 3731 3732
	/* 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);
3733
		if ((icr0_remaining & I40E_PFINT_ICR0_PE_CRITERR_MASK) ||
J
Jesse Brandeburg 已提交
3734
		    (icr0_remaining & I40E_PFINT_ICR0_PCI_EXCEPTION_MASK) ||
3735
		    (icr0_remaining & I40E_PFINT_ICR0_ECC_ERR_MASK)) {
3736
			dev_info(&pf->pdev->dev, "device will be reset\n");
3737
			set_bit(__I40E_PF_RESET_REQUESTED, pf->state);
3738
			i40e_service_event_schedule(pf);
J
Jesse Brandeburg 已提交
3739 3740 3741
		}
		ena_mask &= ~icr0_remaining;
	}
3742
	ret = IRQ_HANDLED;
J
Jesse Brandeburg 已提交
3743

3744
enable_intr:
J
Jesse Brandeburg 已提交
3745 3746
	/* re-enable interrupt causes */
	wr32(hw, I40E_PFINT_ICR0_ENA, ena_mask);
3747
	if (!test_bit(__I40E_DOWN, pf->state)) {
J
Jesse Brandeburg 已提交
3748
		i40e_service_event_schedule(pf);
3749
		i40e_irq_dynamic_enable_icr0(pf, false);
J
Jesse Brandeburg 已提交
3750 3751
	}

3752
	return ret;
J
Jesse Brandeburg 已提交
3753 3754
}

3755 3756 3757 3758 3759 3760 3761 3762 3763 3764 3765 3766 3767 3768 3769 3770 3771 3772 3773 3774 3775 3776 3777 3778 3779 3780 3781 3782 3783 3784 3785 3786 3787 3788 3789 3790
/**
 * 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;

3791 3792 3793 3794 3795 3796 3797 3798 3799 3800 3801
		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);
		}
3802 3803 3804 3805 3806
		/* 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);
3807 3808
		if (tx_buf->tx_flags & I40E_TX_FLAGS_FD_SB)
			kfree(tx_buf->raw_buf);
3809

3810 3811 3812
		tx_buf->raw_buf = NULL;
		tx_buf->tx_flags = 0;
		tx_buf->next_to_watch = NULL;
3813
		dma_unmap_len_set(tx_buf, len, 0);
3814 3815
		tx_desc->buffer_addr = 0;
		tx_desc->cmd_type_offset_bsz = 0;
3816

3817
		/* move us past the eop_desc for start of next FD desc */
3818 3819 3820 3821 3822 3823 3824 3825 3826 3827 3828 3829 3830 3831 3832 3833
		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
3834
	if (vsi->back->flags & I40E_FLAG_MSIX_ENABLED)
3835
		i40e_irq_dynamic_enable(vsi, tx_ring->q_vector->v_idx);
J
Jesse Brandeburg 已提交
3836

3837 3838 3839 3840 3841 3842 3843 3844 3845 3846 3847 3848 3849 3850 3851 3852 3853 3854 3855 3856 3857 3858
	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
3859
/**
3860
 * i40e_map_vector_to_qp - Assigns the queue pair to the vector
J
Jesse Brandeburg 已提交
3861 3862
 * @vsi: the VSI being configured
 * @v_idx: vector index
3863
 * @qp_idx: queue pair index
J
Jesse Brandeburg 已提交
3864
 **/
3865
static void i40e_map_vector_to_qp(struct i40e_vsi *vsi, int v_idx, int qp_idx)
J
Jesse Brandeburg 已提交
3866
{
3867
	struct i40e_q_vector *q_vector = vsi->q_vectors[v_idx];
3868 3869
	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
3870 3871

	tx_ring->q_vector = q_vector;
3872 3873
	tx_ring->next = q_vector->tx.ring;
	q_vector->tx.ring = tx_ring;
J
Jesse Brandeburg 已提交
3874
	q_vector->tx.count++;
3875

3876 3877 3878 3879 3880 3881 3882 3883 3884 3885
	/* Place XDP Tx ring in the same q_vector ring list as regular Tx */
	if (i40e_enabled_xdp_vsi(vsi)) {
		struct i40e_ring *xdp_ring = vsi->xdp_rings[qp_idx];

		xdp_ring->q_vector = q_vector;
		xdp_ring->next = q_vector->tx.ring;
		q_vector->tx.ring = xdp_ring;
		q_vector->tx.count++;
	}

3886 3887 3888 3889
	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
3890 3891 3892 3893 3894 3895 3896 3897 3898 3899 3900 3901 3902 3903 3904
}

/**
 * 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;
3905
	int num_ringpairs;
J
Jesse Brandeburg 已提交
3906 3907 3908 3909 3910
	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.
3911 3912 3913 3914
	 * 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
3915
	 */
3916
	for (; v_start < q_vectors; v_start++) {
3917 3918 3919 3920 3921 3922 3923 3924 3925 3926 3927 3928
		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--) {
3929
			i40e_map_vector_to_qp(vsi, v_start, qp_idx);
3930 3931
			qp_idx++;
			qp_remaining--;
J
Jesse Brandeburg 已提交
3932 3933 3934 3935 3936 3937 3938 3939 3940 3941 3942 3943 3944 3945 3946 3947 3948 3949
		}
	}
}

/**
 * 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,
3950
				  pf->int_name, pf);
J
Jesse Brandeburg 已提交
3951 3952
	else
		err = request_irq(pf->pdev->irq, i40e_intr, IRQF_SHARED,
3953
				  pf->int_name, pf);
J
Jesse Brandeburg 已提交
3954 3955 3956 3957 3958 3959 3960 3961 3962

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

	return err;
}

#ifdef CONFIG_NET_POLL_CONTROLLER
/**
3963
 * i40e_netpoll - A Polling 'interrupt' handler
J
Jesse Brandeburg 已提交
3964 3965 3966 3967 3968 3969 3970 3971 3972 3973 3974 3975 3976
 * @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 */
3977
	if (test_bit(__I40E_VSI_DOWN, vsi->state))
J
Jesse Brandeburg 已提交
3978 3979 3980 3981
		return;

	if (pf->flags & I40E_FLAG_MSIX_ENABLED) {
		for (i = 0; i < vsi->num_q_vectors; i++)
3982
			i40e_msix_clean_rings(0, vsi->q_vectors[i]);
J
Jesse Brandeburg 已提交
3983 3984 3985 3986 3987 3988
	} else {
		i40e_intr(pf->pdev->irq, netdev);
	}
}
#endif

3989 3990
#define I40E_QTX_ENA_WAIT_COUNT 50

3991 3992 3993 3994 3995 3996 3997 3998 3999 4000 4001 4002 4003 4004 4005 4006 4007 4008 4009 4010 4011
/**
 * 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;

4012
		usleep_range(10, 20);
4013 4014 4015 4016 4017 4018 4019
	}
	if (i >= I40E_QUEUE_WAIT_RETRY_LIMIT)
		return -ETIMEDOUT;

	return 0;
}

4020 4021 4022 4023 4024 4025 4026 4027 4028 4029 4030 4031 4032 4033 4034 4035 4036 4037 4038 4039 4040 4041 4042 4043 4044 4045 4046 4047 4048 4049 4050 4051 4052 4053 4054 4055 4056 4057 4058 4059 4060 4061 4062 4063
/**
 * i40e_control_tx_q - Start or stop a particular Tx queue
 * @pf: the PF structure
 * @pf_q: the PF queue to configure
 * @enable: start or stop the queue
 *
 * This function enables or disables a single queue. Note that any delay
 * required after the operation is expected to be handled by the caller of
 * this function.
 **/
static void i40e_control_tx_q(struct i40e_pf *pf, int pf_q, bool enable)
{
	struct i40e_hw *hw = &pf->hw;
	u32 tx_reg;
	int i;

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

	for (i = 0; i < I40E_QTX_ENA_WAIT_COUNT; i++) {
		tx_reg = rd32(hw, I40E_QTX_ENA(pf_q));
		if (((tx_reg >> I40E_QTX_ENA_QENA_REQ_SHIFT) & 1) ==
		    ((tx_reg >> I40E_QTX_ENA_QENA_STAT_SHIFT) & 1))
			break;
		usleep_range(1000, 2000);
	}

	/* Skip if the queue is already in the requested state */
	if (enable == !!(tx_reg & I40E_QTX_ENA_QENA_STAT_MASK))
		return;

	/* turn on/off the queue */
	if (enable) {
		wr32(hw, I40E_QTX_HEAD(pf_q), 0);
		tx_reg |= I40E_QTX_ENA_QENA_REQ_MASK;
	} else {
		tx_reg &= ~I40E_QTX_ENA_QENA_REQ_MASK;
	}

	wr32(hw, I40E_QTX_ENA(pf_q), tx_reg);
}

4064 4065 4066 4067 4068 4069 4070 4071 4072 4073 4074 4075 4076 4077 4078 4079 4080 4081 4082 4083 4084 4085 4086 4087 4088 4089 4090
/**
 * i40e_control_wait_tx_q - Start/stop Tx queue and wait for completion
 * @seid: VSI SEID
 * @pf: the PF structure
 * @pf_q: the PF queue to configure
 * @is_xdp: true if the queue is used for XDP
 * @enable: start or stop the queue
 **/
static int i40e_control_wait_tx_q(int seid, struct i40e_pf *pf, int pf_q,
				  bool is_xdp, bool enable)
{
	int ret;

	i40e_control_tx_q(pf, pf_q, enable);

	/* wait for the change to finish */
	ret = i40e_pf_txq_wait(pf, pf_q, enable);
	if (ret) {
		dev_info(&pf->pdev->dev,
			 "VSI seid %d %sTx ring %d %sable timeout\n",
			 seid, (is_xdp ? "XDP " : ""), pf_q,
			 (enable ? "en" : "dis"));
	}

	return ret;
}

J
Jesse Brandeburg 已提交
4091 4092 4093 4094 4095 4096 4097 4098
/**
 * 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;
4099
	int i, pf_q, ret = 0;
J
Jesse Brandeburg 已提交
4100 4101 4102

	pf_q = vsi->base_queue;
	for (i = 0; i < vsi->num_queue_pairs; i++, pf_q++) {
4103 4104 4105 4106 4107
		ret = i40e_control_wait_tx_q(vsi->seid, pf,
					     pf_q,
					     false /*is xdp*/, enable);
		if (ret)
			break;
4108

4109 4110 4111 4112 4113 4114 4115
		if (!i40e_enabled_xdp_vsi(vsi))
			continue;

		ret = i40e_control_wait_tx_q(vsi->seid, pf,
					     pf_q + vsi->alloc_queue_pairs,
					     true /*is xdp*/, enable);
		if (ret)
4116
			break;
J
Jesse Brandeburg 已提交
4117 4118
	}

4119 4120 4121 4122 4123 4124 4125 4126 4127 4128 4129 4130 4131 4132 4133 4134 4135 4136 4137 4138 4139 4140 4141 4142
	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;

4143
		usleep_range(10, 20);
4144 4145 4146
	}
	if (i >= I40E_QUEUE_WAIT_RETRY_LIMIT)
		return -ETIMEDOUT;
4147

J
Jesse Brandeburg 已提交
4148 4149 4150
	return 0;
}

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
/**
 * i40e_control_rx_q - Start or stop a particular Rx queue
 * @pf: the PF structure
 * @pf_q: the PF queue to configure
 * @enable: start or stop the queue
 *
 * This function enables or disables a single queue. Note that any delay
 * required after the operation is expected to be handled by the caller of
 * this function.
 **/
static void i40e_control_rx_q(struct i40e_pf *pf, int pf_q, bool enable)
{
	struct i40e_hw *hw = &pf->hw;
	u32 rx_reg;
	int i;

	for (i = 0; i < I40E_QTX_ENA_WAIT_COUNT; i++) {
		rx_reg = rd32(hw, I40E_QRX_ENA(pf_q));
		if (((rx_reg >> I40E_QRX_ENA_QENA_REQ_SHIFT) & 1) ==
		    ((rx_reg >> I40E_QRX_ENA_QENA_STAT_SHIFT) & 1))
			break;
		usleep_range(1000, 2000);
	}

	/* Skip if the queue is already in the requested state */
	if (enable == !!(rx_reg & I40E_QRX_ENA_QENA_STAT_MASK))
		return;

	/* turn on/off the queue */
	if (enable)
		rx_reg |= I40E_QRX_ENA_QENA_REQ_MASK;
	else
		rx_reg &= ~I40E_QRX_ENA_QENA_REQ_MASK;

	wr32(hw, I40E_QRX_ENA(pf_q), rx_reg);
}

J
Jesse Brandeburg 已提交
4188 4189 4190 4191 4192 4193 4194 4195
/**
 * 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;
4196
	int i, pf_q, ret = 0;
J
Jesse Brandeburg 已提交
4197 4198 4199

	pf_q = vsi->base_queue;
	for (i = 0; i < vsi->num_queue_pairs; i++, pf_q++) {
4200
		i40e_control_rx_q(pf, pf_q, enable);
J
Jesse Brandeburg 已提交
4201 4202

		/* wait for the change to finish */
4203 4204 4205
		ret = i40e_pf_rxq_wait(pf, pf_q, enable);
		if (ret) {
			dev_info(&pf->pdev->dev,
S
Shannon Nelson 已提交
4206 4207
				 "VSI seid %d Rx ring %d %sable timeout\n",
				 vsi->seid, pf_q, (enable ? "en" : "dis"));
4208
			break;
J
Jesse Brandeburg 已提交
4209 4210 4211
		}
	}

4212 4213 4214 4215 4216 4217
	/* 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);

4218
	return ret;
J
Jesse Brandeburg 已提交
4219 4220 4221
}

/**
F
Filip Sadowski 已提交
4222
 * i40e_vsi_start_rings - Start a VSI's rings
J
Jesse Brandeburg 已提交
4223 4224
 * @vsi: the VSI being configured
 **/
F
Filip Sadowski 已提交
4225
int i40e_vsi_start_rings(struct i40e_vsi *vsi)
J
Jesse Brandeburg 已提交
4226
{
4227
	int ret = 0;
J
Jesse Brandeburg 已提交
4228 4229

	/* do rx first for enable and last for disable */
F
Filip Sadowski 已提交
4230 4231 4232 4233
	ret = i40e_vsi_control_rx(vsi, true);
	if (ret)
		return ret;
	ret = i40e_vsi_control_tx(vsi, true);
J
Jesse Brandeburg 已提交
4234 4235 4236 4237

	return ret;
}

F
Filip Sadowski 已提交
4238 4239 4240 4241 4242 4243
/**
 * i40e_vsi_stop_rings - Stop a VSI's rings
 * @vsi: the VSI being configured
 **/
void i40e_vsi_stop_rings(struct i40e_vsi *vsi)
{
4244
	/* When port TX is suspended, don't wait */
4245
	if (test_bit(__I40E_PORT_SUSPENDED, vsi->back->state))
4246 4247
		return i40e_vsi_stop_rings_no_wait(vsi);

F
Filip Sadowski 已提交
4248 4249 4250 4251 4252 4253 4254
	/* 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);
}

4255 4256 4257 4258 4259 4260 4261 4262 4263 4264 4265 4266 4267 4268 4269 4270 4271 4272 4273 4274 4275 4276 4277
/**
 * i40e_vsi_stop_rings_no_wait - Stop a VSI's rings and do not delay
 * @vsi: the VSI being shutdown
 *
 * This function stops all the rings for a VSI but does not delay to verify
 * that rings have been disabled. It is expected that the caller is shutting
 * down multiple VSIs at once and will delay together for all the VSIs after
 * initiating the shutdown. This is particularly useful for shutting down lots
 * of VFs together. Otherwise, a large delay can be incurred while configuring
 * each VSI in serial.
 **/
void i40e_vsi_stop_rings_no_wait(struct i40e_vsi *vsi)
{
	struct i40e_pf *pf = vsi->back;
	int i, pf_q;

	pf_q = vsi->base_queue;
	for (i = 0; i < vsi->num_queue_pairs; i++, pf_q++) {
		i40e_control_tx_q(pf, pf_q, false);
		i40e_control_rx_q(pf, pf_q, false);
	}
}

J
Jesse Brandeburg 已提交
4278 4279 4280 4281 4282 4283 4284 4285 4286 4287 4288 4289 4290 4291 4292 4293
/**
 * 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;

4294 4295 4296 4297
		if (!vsi->irqs_ready)
			return;

		vsi->irqs_ready = false;
J
Jesse Brandeburg 已提交
4298
		for (i = 0; i < vsi->num_q_vectors; i++) {
4299 4300 4301 4302 4303
			int irq_num;
			u16 vector;

			vector = i + base;
			irq_num = pf->msix_entries[vector].vector;
J
Jesse Brandeburg 已提交
4304 4305

			/* free only the irqs that were actually requested */
S
Shannon Nelson 已提交
4306 4307
			if (!vsi->q_vectors[i] ||
			    !vsi->q_vectors[i]->num_ringpairs)
J
Jesse Brandeburg 已提交
4308 4309
				continue;

4310 4311
			/* clear the affinity notifier in the IRQ descriptor */
			irq_set_affinity_notifier(irq_num, NULL);
4312
			/* remove our suggested affinity mask for this IRQ */
4313 4314 4315
			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
4316 4317 4318 4319 4320 4321 4322 4323 4324 4325 4326 4327 4328 4329 4330 4331 4332 4333 4334 4335 4336 4337 4338 4339 4340 4341 4342 4343 4344 4345 4346 4347 4348 4349 4350 4351 4352 4353 4354 4355 4356 4357 4358 4359 4360 4361 4362 4363 4364 4365 4366 4367 4368 4369 4370 4371 4372 4373 4374 4375 4376 4377 4378 4379 4380 4381 4382 4383 4384 4385 4386 4387 4388 4389 4390 4391 4392 4393 4394 4395 4396 4397

			/* 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);
	}
}

4398 4399 4400 4401 4402 4403 4404 4405 4406 4407 4408 4409
/**
 * 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];
4410
	struct i40e_ring *ring;
4411 4412 4413 4414 4415

	if (!q_vector)
		return;

	/* disassociate q_vector from rings */
4416 4417 4418 4419 4420
	i40e_for_each_ring(ring, q_vector->tx)
		ring->q_vector = NULL;

	i40e_for_each_ring(ring, q_vector->rx)
		ring->q_vector = NULL;
4421 4422 4423 4424 4425 4426 4427 4428 4429 4430

	/* 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
4431 4432 4433 4434 4435 4436 4437 4438 4439 4440 4441
/**
 * 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;

4442 4443
	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
4444 4445 4446 4447 4448 4449 4450 4451 4452 4453 4454 4455 4456
}

/**
 * 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;
4457 4458
		kfree(pf->irq_pile);
		pf->irq_pile = NULL;
J
Jesse Brandeburg 已提交
4459 4460 4461 4462 4463 4464 4465 4466 4467 4468 4469 4470 4471 4472 4473 4474 4475
	} 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;

4476
	i40e_free_misc_vector(pf);
4477

4478 4479 4480
	i40e_put_lump(pf->irq_pile, pf->iwarp_base_vector,
		      I40E_IWARP_IRQ_PILE_ID);

J
Jesse Brandeburg 已提交
4481
	i40e_put_lump(pf->irq_pile, 0, I40E_PILE_VALID_BIT-1);
M
Mitch Williams 已提交
4482
	for (i = 0; i < pf->num_alloc_vsi; i++)
J
Jesse Brandeburg 已提交
4483 4484 4485 4486 4487 4488 4489 4490 4491 4492 4493 4494 4495 4496 4497 4498
		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;

4499 4500 4501 4502 4503 4504
	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
4505 4506 4507 4508 4509 4510 4511 4512 4513 4514 4515 4516 4517
}

/**
 * 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;

4518 4519 4520 4521 4522 4523
	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
4524 4525
}

4526 4527 4528 4529 4530 4531
/**
 * 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
4532
	struct i40e_pf *pf = vsi->back;
4533
	if (!test_and_set_bit(__I40E_VSI_DOWN, vsi->state))
4534 4535 4536 4537
		i40e_down(vsi);
	i40e_vsi_free_irq(vsi);
	i40e_vsi_free_tx_resources(vsi);
	i40e_vsi_free_rx_resources(vsi);
4538
	vsi->current_netdev_flags = 0;
M
Mitch Williams 已提交
4539
	pf->flags |= I40E_FLAG_SERVICE_CLIENT_REQUESTED;
4540
	if (test_bit(__I40E_RESET_RECOVERY_PENDING, pf->state))
M
Mitch Williams 已提交
4541
		pf->flags |=  I40E_FLAG_CLIENT_RESET;
4542 4543
}

J
Jesse Brandeburg 已提交
4544 4545 4546 4547 4548 4549
/**
 * i40e_quiesce_vsi - Pause a given VSI
 * @vsi: the VSI being paused
 **/
static void i40e_quiesce_vsi(struct i40e_vsi *vsi)
{
4550
	if (test_bit(__I40E_VSI_DOWN, vsi->state))
J
Jesse Brandeburg 已提交
4551 4552
		return;

4553
	set_bit(__I40E_VSI_NEEDS_RESTART, vsi->state);
J
Jesse Brandeburg 已提交
4554
	if (vsi->netdev && netif_running(vsi->netdev))
J
Jesse Brandeburg 已提交
4555
		vsi->netdev->netdev_ops->ndo_stop(vsi->netdev);
J
Jesse Brandeburg 已提交
4556
	else
4557
		i40e_vsi_close(vsi);
J
Jesse Brandeburg 已提交
4558 4559 4560 4561 4562 4563 4564 4565
}

/**
 * i40e_unquiesce_vsi - Resume a given VSI
 * @vsi: the VSI being resumed
 **/
static void i40e_unquiesce_vsi(struct i40e_vsi *vsi)
{
4566
	if (!test_and_clear_bit(__I40E_VSI_NEEDS_RESTART, vsi->state))
J
Jesse Brandeburg 已提交
4567 4568 4569 4570 4571
		return;

	if (vsi->netdev && netif_running(vsi->netdev))
		vsi->netdev->netdev_ops->ndo_open(vsi->netdev);
	else
4572
		i40e_vsi_open(vsi);   /* this clears the DOWN bit */
J
Jesse Brandeburg 已提交
4573 4574 4575 4576 4577 4578 4579 4580 4581 4582
}

/**
 * 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
4583
	for (v = 0; v < pf->num_alloc_vsi; v++) {
J
Jesse Brandeburg 已提交
4584 4585 4586 4587 4588 4589 4590 4591 4592 4593 4594 4595 4596
		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
4597
	for (v = 0; v < pf->num_alloc_vsi; v++) {
J
Jesse Brandeburg 已提交
4598 4599 4600 4601 4602
		if (pf->vsi[v])
			i40e_unquiesce_vsi(pf->vsi[v]);
	}
}

4603
/**
4604
 * i40e_vsi_wait_queues_disabled - Wait for VSI's queues to be disabled
4605 4606
 * @vsi: the VSI being configured
 *
4607
 * Wait until all queues on a given VSI have been disabled.
4608
 **/
4609
int i40e_vsi_wait_queues_disabled(struct i40e_vsi *vsi)
4610 4611 4612 4613 4614 4615
{
	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++) {
4616
		/* Check and wait for the Tx queue */
4617 4618 4619
		ret = i40e_pf_txq_wait(pf, pf_q, false);
		if (ret) {
			dev_info(&pf->pdev->dev,
S
Shannon Nelson 已提交
4620 4621
				 "VSI seid %d Tx ring %d disable timeout\n",
				 vsi->seid, pf_q);
4622 4623
			return ret;
		}
4624 4625 4626 4627 4628 4629 4630 4631 4632 4633 4634 4635 4636 4637 4638

		if (!i40e_enabled_xdp_vsi(vsi))
			goto wait_rx;

		/* Check and wait for the XDP Tx queue */
		ret = i40e_pf_txq_wait(pf, pf_q + vsi->alloc_queue_pairs,
				       false);
		if (ret) {
			dev_info(&pf->pdev->dev,
				 "VSI seid %d XDP Tx ring %d disable timeout\n",
				 vsi->seid, pf_q);
			return ret;
		}
wait_rx:
		/* Check and wait for the Rx queue */
4639 4640 4641 4642 4643 4644 4645 4646 4647
		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;
		}
	}

4648 4649 4650
	return 0;
}

4651
#ifdef CONFIG_I40E_DCB
4652
/**
4653
 * i40e_pf_wait_queues_disabled - Wait for all queues of PF VSIs to be disabled
4654 4655
 * @pf: the PF
 *
4656
 * This function waits for the queues to be in disabled state for all the
4657 4658
 * VSIs that are managed by this PF.
 **/
4659
static int i40e_pf_wait_queues_disabled(struct i40e_pf *pf)
4660 4661 4662 4663
{
	int v, ret = 0;

	for (v = 0; v < pf->hw.func_caps.num_vsis; v++) {
4664
		if (pf->vsi[v]) {
4665
			ret = i40e_vsi_wait_queues_disabled(pf->vsi[v]);
4666 4667 4668 4669 4670 4671 4672 4673 4674
			if (ret)
				break;
		}
	}

	return ret;
}

#endif
4675 4676 4677 4678 4679 4680 4681

/**
 * 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.
4682 4683 4684
 * We proactively detect hung TX queues by checking if interrupts are disabled
 * but there are pending descriptors.  If it appears hung, attempt to recover
 * by triggering a SW interrupt.
4685 4686 4687 4688 4689
 **/
static void i40e_detect_recover_hung_queue(int q_idx, struct i40e_vsi *vsi)
{
	struct i40e_ring *tx_ring = NULL;
	struct i40e_pf	*pf;
4690
	u32 val, tx_pending;
4691 4692 4693 4694 4695 4696 4697 4698 4699 4700 4701 4702 4703 4704 4705 4706 4707 4708 4709 4710 4711 4712 4713 4714 4715
	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);

4716
	tx_pending = i40e_get_tx_pending(tx_ring);
4717

4718 4719 4720 4721
	/* Interrupts are disabled and TX pending is non-zero,
	 * trigger the SW interrupt (don't wait). Worst case
	 * there will be one extra interrupt which may result
	 * into not cleaning any queues because queues are cleaned.
4722
	 */
4723 4724
	if (tx_pending && (!(val & I40E_PFINT_DYN_CTLN_INTENA_MASK)))
		i40e_force_wb(vsi, tx_ring->q_vector);
4725 4726 4727 4728 4729 4730 4731 4732 4733 4734 4735 4736 4737 4738
}

/**
 * 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;
4739
	unsigned int i;
4740 4741 4742 4743 4744 4745 4746 4747

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

	if (!vsi)
		return;

	/* Make sure, VSI state is not DOWN/RECOVERY_PENDING */
4748 4749
	if (test_bit(__I40E_VSI_DOWN, vsi->back->state) ||
	    test_bit(__I40E_RESET_RECOVERY_PENDING, vsi->back->state))
4750 4751 4752 4753 4754 4755 4756 4757 4758 4759 4760 4761 4762 4763 4764 4765 4766 4767 4768 4769 4770 4771 4772 4773
		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);
	}
}

4774 4775
/**
 * i40e_get_iscsi_tc_map - Return TC map for iSCSI APP
4776
 * @pf: pointer to PF
4777 4778 4779 4780 4781 4782 4783 4784 4785 4786 4787 4788 4789 4790 4791 4792 4793 4794
 *
 * 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
4795
			enabled_tc |= BIT(tc);
4796 4797 4798 4799 4800 4801 4802
			break;
		}
	}

	return enabled_tc;
}

J
Jesse Brandeburg 已提交
4803 4804 4805 4806 4807 4808 4809 4810
/**
 * 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)
{
4811
	int i, tc_unused = 0;
J
Jesse Brandeburg 已提交
4812
	u8 num_tc = 0;
4813
	u8 ret = 0;
J
Jesse Brandeburg 已提交
4814 4815 4816

	/* Scan the ETS Config Priority Table to find
	 * traffic class enabled for a given priority
4817
	 * and create a bitmask of enabled TCs
J
Jesse Brandeburg 已提交
4818
	 */
4819 4820
	for (i = 0; i < I40E_MAX_USER_PRIORITY; i++)
		num_tc |= BIT(dcbcfg->etscfg.prioritytable[i]);
J
Jesse Brandeburg 已提交
4821

4822 4823
	/* Now scan the bitmask to check for
	 * contiguous TCs starting with TC0
J
Jesse Brandeburg 已提交
4824
	 */
4825 4826 4827 4828 4829 4830 4831 4832 4833 4834 4835 4836 4837 4838 4839 4840 4841 4842
	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
4843 4844 4845 4846 4847 4848 4849 4850 4851 4852 4853 4854 4855 4856 4857 4858
}

/**
 * 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++)
4859
		enabled_tc |= BIT(i);
J
Jesse Brandeburg 已提交
4860 4861 4862 4863 4864 4865 4866 4867 4868 4869 4870 4871 4872

	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;
4873
	u8 i, enabled_tc = 1;
J
Jesse Brandeburg 已提交
4874 4875 4876 4877 4878 4879 4880
	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;

4881 4882 4883 4884
	/* 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
4885
	/* MFP mode return count of enabled TCs for this PF */
4886 4887 4888
	if (pf->hw.func_caps.iscsi)
		enabled_tc =  i40e_get_iscsi_tc_map(pf);
	else
4889
		return 1; /* Only TC0 */
J
Jesse Brandeburg 已提交
4890

4891
	for (i = 0; i < I40E_MAX_TRAFFIC_CLASS; i++) {
J
Jesse Brandeburg 已提交
4892
		if (enabled_tc & BIT(i))
4893 4894 4895
			num_tc++;
	}
	return num_tc;
J
Jesse Brandeburg 已提交
4896 4897 4898 4899 4900 4901 4902 4903 4904 4905 4906 4907
}

/**
 * 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))
4908
		return I40E_DEFAULT_TRAFFIC_CLASS;
J
Jesse Brandeburg 已提交
4909 4910

	/* SFP mode we want PF to be enabled for all TCs */
4911 4912 4913
	if (!(pf->flags & I40E_FLAG_MFP_ENABLED))
		return i40e_dcb_get_enabled_tc(&pf->hw.local_dcbx_config);

4914
	/* MFP enabled and iSCSI PF type */
4915 4916 4917
	if (pf->hw.func_caps.iscsi)
		return i40e_get_iscsi_tc_map(pf);
	else
4918
		return I40E_DEFAULT_TRAFFIC_CLASS;
J
Jesse Brandeburg 已提交
4919 4920 4921 4922 4923 4924 4925 4926 4927 4928 4929 4930 4931 4932
}

/**
 * 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;
4933
	i40e_status ret;
J
Jesse Brandeburg 已提交
4934 4935 4936 4937
	u32 tc_bw_max;
	int i;

	/* Get the VSI level BW configuration */
4938 4939
	ret = i40e_aq_query_vsi_bw_config(hw, vsi->seid, &bw_config, NULL);
	if (ret) {
J
Jesse Brandeburg 已提交
4940
		dev_info(&pf->pdev->dev,
4941 4942 4943
			 "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
4944
		return -EINVAL;
J
Jesse Brandeburg 已提交
4945 4946 4947
	}

	/* Get the VSI level BW configuration per TC */
4948 4949 4950
	ret = i40e_aq_query_vsi_ets_sla_config(hw, vsi->seid, &bw_ets_config,
					       NULL);
	if (ret) {
J
Jesse Brandeburg 已提交
4951
		dev_info(&pf->pdev->dev,
4952 4953 4954
			 "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
4955
		return -EINVAL;
J
Jesse Brandeburg 已提交
4956 4957 4958 4959 4960 4961 4962 4963 4964 4965 4966 4967 4968 4969 4970 4971 4972 4973 4974 4975 4976
	}

	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
4977

J
Jesse Brandeburg 已提交
4978
	return 0;
J
Jesse Brandeburg 已提交
4979 4980 4981 4982 4983 4984 4985 4986 4987 4988
}

/**
 * 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
4989
static int i40e_vsi_configure_bw_alloc(struct i40e_vsi *vsi, u8 enabled_tc,
J
Jesse Brandeburg 已提交
4990 4991 4992
				       u8 *bw_share)
{
	struct i40e_aqc_configure_vsi_tc_bw_data bw_data;
4993
	i40e_status ret;
J
Jesse Brandeburg 已提交
4994
	int i;
J
Jesse Brandeburg 已提交
4995 4996 4997 4998 4999

	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];

5000 5001 5002
	ret = i40e_aq_config_vsi_tc_bw(&vsi->back->hw, vsi->seid, &bw_data,
				       NULL);
	if (ret) {
J
Jesse Brandeburg 已提交
5003
		dev_info(&vsi->back->pdev->dev,
J
Jesse Brandeburg 已提交
5004 5005
			 "AQ command Config VSI BW allocation per TC failed = %d\n",
			 vsi->back->hw.aq.asq_last_status);
J
Jesse Brandeburg 已提交
5006
		return -EINVAL;
J
Jesse Brandeburg 已提交
5007 5008 5009 5010 5011
	}

	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
5012
	return 0;
J
Jesse Brandeburg 已提交
5013 5014 5015 5016 5017 5018 5019 5020 5021 5022 5023 5024 5025 5026 5027 5028 5029 5030 5031 5032 5033 5034 5035 5036 5037 5038 5039 5040 5041 5042 5043 5044 5045 5046 5047 5048 5049 5050
}

/**
 * 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
5051
		if (vsi->tc_config.enabled_tc & BIT(i))
J
Jesse Brandeburg 已提交
5052 5053 5054 5055 5056 5057 5058 5059 5060 5061 5062 5063 5064 5065 5066 5067 5068 5069 5070 5071 5072 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 5109 5110 5111 5112
			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
5113
		if (enabled_tc & BIT(i))
J
Jesse Brandeburg 已提交
5114 5115 5116 5117 5118 5119 5120 5121 5122 5123 5124 5125 5126 5127 5128 5129
			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;
5130
	ctxt.info = vsi->info;
J
Jesse Brandeburg 已提交
5131 5132
	i40e_vsi_setup_queue_map(vsi, &ctxt, enabled_tc, false);

5133 5134 5135 5136 5137 5138
	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
5139 5140 5141 5142
	/* 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,
5143 5144 5145 5146
			 "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
5147 5148 5149 5150 5151 5152 5153 5154 5155 5156
		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,
5157 5158 5159 5160
			 "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
5161 5162 5163 5164 5165 5166 5167 5168 5169
		goto out;
	}

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

N
Neerav Parikh 已提交
5170 5171 5172 5173 5174 5175 5176 5177 5178 5179 5180 5181 5182 5183 5184 5185 5186 5187 5188 5189 5190 5191 5192
/**
 * 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
5193
		if (enabled_tc & BIT(i))
N
Neerav Parikh 已提交
5194 5195 5196 5197 5198 5199 5200
			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,
5201 5202 5203
			 "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
5204 5205 5206 5207 5208 5209 5210
		goto out;
	}

	/* Update the BW information */
	ret = i40e_veb_get_bw_info(veb);
	if (ret) {
		dev_info(&pf->pdev->dev,
5211 5212 5213
			 "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
5214 5215 5216 5217 5218 5219 5220 5221 5222 5223 5224 5225 5226 5227 5228 5229 5230 5231 5232 5233 5234 5235 5236 5237 5238 5239 5240 5241 5242 5243 5244 5245 5246 5247 5248 5249
	}

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
5250
	for (v = 0; v < pf->num_alloc_vsi; v++) {
N
Neerav Parikh 已提交
5251 5252 5253 5254 5255 5256 5257 5258 5259
		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
5260
			tc_map = I40E_DEFAULT_TRAFFIC_CLASS;
N
Neerav Parikh 已提交
5261 5262 5263 5264 5265 5266 5267 5268

		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 {
5269 5270
			/* Re-configure VSI vectors based on updated TC map */
			i40e_vsi_map_rings_to_vectors(pf->vsi[v]);
N
Neerav Parikh 已提交
5271 5272 5273 5274 5275 5276
			if (pf->vsi[v]->netdev)
				i40e_dcbnl_set_all(pf->vsi[v]);
		}
	}
}

5277 5278 5279 5280 5281 5282 5283 5284 5285 5286 5287 5288 5289 5290 5291
/**
 * 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,
5292 5293 5294
			 "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));
5295
		/* Schedule PF reset to recover */
5296
		set_bit(__I40E_PF_RESET_REQUESTED, pf->state);
5297 5298 5299 5300 5301 5302
		i40e_service_event_schedule(pf);
	}

	return ret;
}

N
Neerav Parikh 已提交
5303 5304 5305 5306 5307 5308 5309 5310 5311 5312 5313 5314
/**
 * 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;

5315
	/* Do not enable DCB for SW1 and SW2 images even if the FW is capable */
5316
	if (pf->hw_features & I40E_HW_NO_DCB_SUPPORT)
5317 5318
		goto out;

N
Neerav Parikh 已提交
5319 5320 5321 5322 5323 5324 5325 5326 5327 5328 5329 5330
	/* 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;
5331 5332

			pf->flags |= I40E_FLAG_DCB_CAPABLE;
5333 5334 5335
			/* Enable DCB tagging only when more than one TC
			 * or explicitly disable if only one TC
			 */
5336 5337
			if (i40e_dcb_get_num_tc(&hw->local_dcbx_config) > 1)
				pf->flags |= I40E_FLAG_DCB_ENABLED;
5338 5339
			else
				pf->flags &= ~I40E_FLAG_DCB_ENABLED;
5340 5341
			dev_dbg(&pf->pdev->dev,
				"DCBX offload is supported for this PF.\n");
N
Neerav Parikh 已提交
5342
		}
5343
	} else {
5344
		dev_info(&pf->pdev->dev,
5345 5346 5347
			 "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
5348 5349 5350 5351 5352 5353
	}

out:
	return err;
}
#endif /* CONFIG_I40E_DCB */
J
Jesse Brandeburg 已提交
5354 5355 5356 5357 5358 5359
#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
5360
void i40e_print_link_message(struct i40e_vsi *vsi, bool isup)
J
Jesse Brandeburg 已提交
5361
{
5362
	enum i40e_aq_link_speed new_speed;
5363
	struct i40e_pf *pf = vsi->back;
5364 5365
	char *speed = "Unknown";
	char *fc = "Unknown";
5366
	char *fec = "";
5367
	char *req_fec = "";
5368
	char *an = "";
J
Jesse Brandeburg 已提交
5369

5370
	new_speed = pf->hw.phy.link_info.link_speed;
5371 5372

	if ((vsi->current_isup == isup) && (vsi->current_speed == new_speed))
M
Matt Jared 已提交
5373 5374
		return;
	vsi->current_isup = isup;
5375
	vsi->current_speed = new_speed;
J
Jesse Brandeburg 已提交
5376 5377 5378 5379 5380
	if (!isup) {
		netdev_info(vsi->netdev, "NIC Link is Down\n");
		return;
	}

5381 5382 5383
	/* Warn user if link speed on NPAR enabled partition is not at
	 * least 10GB
	 */
5384 5385 5386
	if (pf->hw.func_caps.npar_enable &&
	    (pf->hw.phy.link_info.link_speed == I40E_LINK_SPEED_1GB ||
	     pf->hw.phy.link_info.link_speed == I40E_LINK_SPEED_100MB))
5387 5388 5389
		netdev_warn(vsi->netdev,
			    "The partition detected link speed that is less than 10Gbps\n");

5390
	switch (pf->hw.phy.link_info.link_speed) {
J
Jesse Brandeburg 已提交
5391
	case I40E_LINK_SPEED_40GB:
5392
		speed = "40 G";
J
Jesse Brandeburg 已提交
5393
		break;
5394
	case I40E_LINK_SPEED_20GB:
5395
		speed = "20 G";
5396
		break;
5397 5398 5399
	case I40E_LINK_SPEED_25GB:
		speed = "25 G";
		break;
J
Jesse Brandeburg 已提交
5400
	case I40E_LINK_SPEED_10GB:
5401
		speed = "10 G";
J
Jesse Brandeburg 已提交
5402 5403
		break;
	case I40E_LINK_SPEED_1GB:
5404
		speed = "1000 M";
J
Jesse Brandeburg 已提交
5405
		break;
M
Mitch Williams 已提交
5406
	case I40E_LINK_SPEED_100MB:
5407
		speed = "100 M";
M
Mitch Williams 已提交
5408
		break;
J
Jesse Brandeburg 已提交
5409 5410 5411 5412
	default:
		break;
	}

5413
	switch (pf->hw.fc.current_mode) {
J
Jesse Brandeburg 已提交
5414
	case I40E_FC_FULL:
5415
		fc = "RX/TX";
J
Jesse Brandeburg 已提交
5416 5417
		break;
	case I40E_FC_TX_PAUSE:
5418
		fc = "TX";
J
Jesse Brandeburg 已提交
5419 5420
		break;
	case I40E_FC_RX_PAUSE:
5421
		fc = "RX";
J
Jesse Brandeburg 已提交
5422 5423
		break;
	default:
5424
		fc = "None";
J
Jesse Brandeburg 已提交
5425 5426 5427
		break;
	}

5428
	if (pf->hw.phy.link_info.link_speed == I40E_LINK_SPEED_25GB) {
5429
		req_fec = ", Requested FEC: None";
5430 5431 5432
		fec = ", FEC: None";
		an = ", Autoneg: False";

5433
		if (pf->hw.phy.link_info.an_info & I40E_AQ_AN_COMPLETED)
5434 5435
			an = ", Autoneg: True";

5436
		if (pf->hw.phy.link_info.fec_info &
5437 5438
		    I40E_AQ_CONFIG_FEC_KR_ENA)
			fec = ", FEC: CL74 FC-FEC/BASE-R";
5439
		else if (pf->hw.phy.link_info.fec_info &
5440 5441
			 I40E_AQ_CONFIG_FEC_RS_ENA)
			fec = ", FEC: CL108 RS-FEC";
5442 5443 5444 5445 5446 5447 5448 5449 5450 5451 5452 5453

		/* 'CL108 RS-FEC' should be displayed when RS is requested, or
		 * both RS and FC are requested
		 */
		if (vsi->back->hw.phy.link_info.req_fec_info &
		    (I40E_AQ_REQUEST_FEC_KR | I40E_AQ_REQUEST_FEC_RS)) {
			if (vsi->back->hw.phy.link_info.req_fec_info &
			    I40E_AQ_REQUEST_FEC_RS)
				req_fec = ", Requested FEC: CL108 RS-FEC";
			else
				req_fec = ", Requested FEC: CL74 FC-FEC/BASE-R";
		}
5454 5455
	}

5456 5457
	netdev_info(vsi->netdev, "NIC Link is Up, %sbps Full Duplex%s%s%s, Flow Control: %s\n",
		    speed, req_fec, fec, an, fc);
J
Jesse Brandeburg 已提交
5458
}
N
Neerav Parikh 已提交
5459

J
Jesse Brandeburg 已提交
5460 5461 5462 5463 5464 5465 5466 5467 5468 5469 5470 5471 5472 5473 5474
/**
 * 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
5475
	err = i40e_vsi_start_rings(vsi);
J
Jesse Brandeburg 已提交
5476 5477 5478
	if (err)
		return err;

5479
	clear_bit(__I40E_VSI_DOWN, vsi->state);
J
Jesse Brandeburg 已提交
5480 5481 5482 5483 5484
	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
5485
		i40e_print_link_message(vsi, true);
J
Jesse Brandeburg 已提交
5486 5487 5488
		netif_tx_start_all_queues(vsi->netdev);
		netif_carrier_on(vsi->netdev);
	}
5489 5490

	/* replay FDIR SB filters */
5491 5492
	if (vsi->type == I40E_VSI_FDIR) {
		/* reset fd counters */
5493 5494
		pf->fd_add_err = 0;
		pf->fd_atr_cnt = 0;
5495
		i40e_fdir_filter_restore(vsi);
5496
	}
5497 5498 5499 5500 5501

	/* 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
5502 5503 5504 5505 5506 5507 5508 5509 5510 5511 5512 5513 5514 5515 5516 5517 5518
	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());
5519
	while (test_and_set_bit(__I40E_CONFIG_BUSY, pf->state))
J
Jesse Brandeburg 已提交
5520 5521 5522 5523
		usleep_range(1000, 2000);
	i40e_down(vsi);

	i40e_up(vsi);
5524
	clear_bit(__I40E_CONFIG_BUSY, pf->state);
J
Jesse Brandeburg 已提交
5525 5526 5527 5528 5529 5530 5531 5532 5533 5534 5535 5536 5537 5538 5539 5540 5541 5542 5543 5544 5545 5546 5547 5548 5549 5550
}

/**
 * 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
5551
	 * sets the vsi->state __I40E_VSI_DOWN bit.
J
Jesse Brandeburg 已提交
5552 5553 5554 5555 5556 5557
	 */
	if (vsi->netdev) {
		netif_carrier_off(vsi->netdev);
		netif_tx_disable(vsi->netdev);
	}
	i40e_vsi_disable_irq(vsi);
F
Filip Sadowski 已提交
5558
	i40e_vsi_stop_rings(vsi);
J
Jesse Brandeburg 已提交
5559 5560 5561
	i40e_napi_disable_all(vsi);

	for (i = 0; i < vsi->num_queue_pairs; i++) {
5562
		i40e_clean_tx_ring(vsi->tx_rings[i]);
5563 5564
		if (i40e_enabled_xdp_vsi(vsi))
			i40e_clean_tx_ring(vsi->xdp_rings[i]);
5565
		i40e_clean_rx_ring(vsi->rx_rings[i]);
J
Jesse Brandeburg 已提交
5566
	}
5567

J
Jesse Brandeburg 已提交
5568 5569 5570 5571 5572 5573 5574 5575 5576 5577 5578 5579 5580 5581 5582 5583 5584 5585 5586 5587 5588 5589 5590 5591 5592 5593 5594 5595 5596 5597 5598 5599 5600 5601 5602 5603
}

/**
 * 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
5604
		enabled_tc |= BIT(i);
J
Jesse Brandeburg 已提交
5605 5606 5607 5608 5609 5610 5611 5612 5613 5614 5615 5616 5617 5618 5619 5620 5621 5622 5623 5624 5625 5626 5627

	/* 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;
}

5628
static int __i40e_setup_tc(struct net_device *netdev, enum tc_setup_type type,
5629
			   void *type_data)
5630
{
5631 5632
	struct tc_mqprio_qopt *mqprio = type_data;

5633
	if (type != TC_SETUP_MQPRIO)
5634
		return -EOPNOTSUPP;
5635

5636
	mqprio->hw = TC_MQPRIO_HW_OFFLOAD_TCS;
5637

5638
	return i40e_setup_tc(netdev, mqprio->num_tc);
5639 5640
}

J
Jesse Brandeburg 已提交
5641 5642 5643 5644 5645 5646 5647 5648 5649 5650 5651 5652
/**
 * 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
 **/
5653
int i40e_open(struct net_device *netdev)
J
Jesse Brandeburg 已提交
5654 5655 5656 5657 5658 5659
{
	struct i40e_netdev_priv *np = netdev_priv(netdev);
	struct i40e_vsi *vsi = np->vsi;
	struct i40e_pf *pf = vsi->back;
	int err;

5660
	/* disallow open during test or if eeprom is broken */
5661 5662
	if (test_bit(__I40E_TESTING, pf->state) ||
	    test_bit(__I40E_BAD_EEPROM, pf->state))
J
Jesse Brandeburg 已提交
5663 5664 5665 5666
		return -EBUSY;

	netif_carrier_off(netdev);

5667 5668 5669 5670
	err = i40e_vsi_open(vsi);
	if (err)
		return err;

5671 5672 5673 5674 5675 5676 5677 5678
	/* 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);

5679
	udp_tunnel_get_rx_info(netdev);
5680

5681 5682 5683 5684 5685 5686 5687 5688 5689 5690
	return 0;
}

/**
 * i40e_vsi_open -
 * @vsi: the VSI to open
 *
 * Finish initialization of the VSI.
 *
 * Returns 0 on success, negative value on failure
5691 5692
 *
 * Note: expects to be called while under rtnl_lock()
5693 5694 5695 5696
 **/
int i40e_vsi_open(struct i40e_vsi *vsi)
{
	struct i40e_pf *pf = vsi->back;
5697
	char int_name[I40E_INT_NAME_STR_LEN];
5698 5699
	int err;

J
Jesse Brandeburg 已提交
5700 5701 5702 5703 5704 5705 5706 5707 5708 5709 5710 5711
	/* 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;

5712 5713 5714 5715 5716 5717
	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
5718

5719 5720 5721 5722 5723
		/* 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;
5724

5725 5726 5727 5728
		err = netif_set_real_num_rx_queues(vsi->netdev,
						   vsi->num_queue_pairs);
		if (err)
			goto err_set_queues;
5729 5730

	} else if (vsi->type == I40E_VSI_FDIR) {
5731
		snprintf(int_name, sizeof(int_name) - 1, "%s-%s:fdir",
5732 5733
			 dev_driver_string(&pf->pdev->dev),
			 dev_name(&pf->pdev->dev));
5734
		err = i40e_vsi_request_irq(vsi, int_name);
5735

5736
	} else {
5737
		err = -EINVAL;
5738 5739
		goto err_setup_rx;
	}
5740

J
Jesse Brandeburg 已提交
5741 5742 5743 5744 5745 5746 5747 5748
	err = i40e_up_complete(vsi);
	if (err)
		goto err_up_complete;

	return 0;

err_up_complete:
	i40e_down(vsi);
5749
err_set_queues:
J
Jesse Brandeburg 已提交
5750 5751 5752 5753 5754 5755
	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])
5756
		i40e_do_reset(pf, BIT_ULL(__I40E_PF_RESET_REQUESTED), true);
J
Jesse Brandeburg 已提交
5757 5758 5759 5760

	return err;
}

5761 5762
/**
 * i40e_fdir_filter_exit - Cleans up the Flow Director accounting
5763
 * @pf: Pointer to PF
5764 5765 5766 5767 5768 5769 5770
 *
 * 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;
5771
	struct i40e_flex_pit *pit_entry, *tmp;
5772 5773 5774 5775 5776 5777 5778
	struct hlist_node *node2;

	hlist_for_each_entry_safe(filter, node2,
				  &pf->fdir_filter_list, fdir_node) {
		hlist_del(&filter->fdir_node);
		kfree(filter);
	}
5779

5780 5781 5782 5783 5784 5785 5786 5787 5788 5789 5790 5791
	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);

5792
	pf->fdir_pf_active_filters = 0;
5793 5794
	pf->fd_tcp4_filter_cnt = 0;
	pf->fd_udp4_filter_cnt = 0;
5795
	pf->fd_sctp4_filter_cnt = 0;
5796
	pf->fd_ip4_filter_cnt = 0;
5797 5798 5799 5800 5801 5802 5803 5804 5805 5806 5807 5808 5809 5810 5811 5812 5813 5814 5815

	/* 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);
5816 5817
}

J
Jesse Brandeburg 已提交
5818 5819 5820 5821 5822 5823 5824 5825 5826 5827
/**
 * 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
 **/
5828
int i40e_close(struct net_device *netdev)
J
Jesse Brandeburg 已提交
5829 5830 5831 5832
{
	struct i40e_netdev_priv *np = netdev_priv(netdev);
	struct i40e_vsi *vsi = np->vsi;

5833
	i40e_vsi_close(vsi);
J
Jesse Brandeburg 已提交
5834 5835 5836 5837 5838 5839 5840 5841

	return 0;
}

/**
 * i40e_do_reset - Start a PF or Core Reset sequence
 * @pf: board private structure
 * @reset_flags: which reset is requested
5842 5843
 * @lock_acquired: indicates whether or not the lock has been acquired
 * before this function was called.
J
Jesse Brandeburg 已提交
5844 5845 5846 5847 5848
 *
 * 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.
 **/
5849
void i40e_do_reset(struct i40e_pf *pf, u32 reset_flags, bool lock_acquired)
J
Jesse Brandeburg 已提交
5850 5851 5852 5853 5854
{
	u32 val;

	WARN_ON(in_interrupt());

5855

J
Jesse Brandeburg 已提交
5856
	/* do the biggest reset indicated */
5857
	if (reset_flags & BIT_ULL(__I40E_GLOBAL_RESET_REQUESTED)) {
J
Jesse Brandeburg 已提交
5858 5859 5860 5861 5862 5863 5864 5865 5866

		/* 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
5867
		dev_dbg(&pf->pdev->dev, "GlobalR requested\n");
J
Jesse Brandeburg 已提交
5868 5869 5870 5871
		val = rd32(&pf->hw, I40E_GLGEN_RTRIG);
		val |= I40E_GLGEN_RTRIG_GLOBR_MASK;
		wr32(&pf->hw, I40E_GLGEN_RTRIG, val);

5872
	} else if (reset_flags & BIT_ULL(__I40E_CORE_RESET_REQUESTED)) {
J
Jesse Brandeburg 已提交
5873 5874 5875 5876 5877

		/* Request a Core Reset
		 *
		 * Same as Global Reset, except does *not* include the MAC/PHY
		 */
J
Jesse Brandeburg 已提交
5878
		dev_dbg(&pf->pdev->dev, "CoreR requested\n");
J
Jesse Brandeburg 已提交
5879 5880 5881 5882 5883
		val = rd32(&pf->hw, I40E_GLGEN_RTRIG);
		val |= I40E_GLGEN_RTRIG_CORER_MASK;
		wr32(&pf->hw, I40E_GLGEN_RTRIG, val);
		i40e_flush(&pf->hw);

5884
	} else if (reset_flags & BIT_ULL(__I40E_PF_RESET_REQUESTED)) {
J
Jesse Brandeburg 已提交
5885 5886 5887 5888 5889 5890 5891 5892 5893

		/* 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
5894
		dev_dbg(&pf->pdev->dev, "PFR requested\n");
5895
		i40e_handle_reset_warning(pf, lock_acquired);
J
Jesse Brandeburg 已提交
5896

5897
	} else if (reset_flags & BIT_ULL(__I40E_REINIT_REQUESTED)) {
J
Jesse Brandeburg 已提交
5898 5899 5900 5901 5902
		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
5903
		for (v = 0; v < pf->num_alloc_vsi; v++) {
J
Jesse Brandeburg 已提交
5904
			struct i40e_vsi *vsi = pf->vsi[v];
J
Jesse Brandeburg 已提交
5905

J
Jesse Brandeburg 已提交
5906
			if (vsi != NULL &&
5907
			    test_and_clear_bit(__I40E_VSI_REINIT_REQUESTED,
5908
					       vsi->state))
J
Jesse Brandeburg 已提交
5909 5910
				i40e_vsi_reinit_locked(pf->vsi[v]);
		}
5911
	} else if (reset_flags & BIT_ULL(__I40E_DOWN_REQUESTED)) {
5912 5913 5914 5915 5916 5917
		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
5918

5919
			if (vsi != NULL &&
5920
			    test_and_clear_bit(__I40E_VSI_DOWN_REQUESTED,
5921 5922
					       vsi->state)) {
				set_bit(__I40E_VSI_DOWN, vsi->state);
5923 5924 5925
				i40e_down(vsi);
			}
		}
J
Jesse Brandeburg 已提交
5926 5927 5928 5929 5930 5931
	} else {
		dev_info(&pf->pdev->dev,
			 "bad reset request 0x%08x\n", reset_flags);
	}
}

N
Neerav Parikh 已提交
5932 5933 5934 5935 5936 5937 5938 5939 5940 5941 5942 5943 5944 5945 5946 5947 5948 5949 5950 5951 5952 5953
#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
5954
			dev_dbg(&pf->pdev->dev, "ETS UP2TC changed.\n");
N
Neerav Parikh 已提交
5955 5956 5957 5958 5959
		}

		if (memcmp(&new_cfg->etscfg.tcbwtable,
			   &old_cfg->etscfg.tcbwtable,
			   sizeof(new_cfg->etscfg.tcbwtable)))
J
Jesse Brandeburg 已提交
5960
			dev_dbg(&pf->pdev->dev, "ETS TC BW Table changed.\n");
N
Neerav Parikh 已提交
5961 5962 5963 5964

		if (memcmp(&new_cfg->etscfg.tsatable,
			   &old_cfg->etscfg.tsatable,
			   sizeof(new_cfg->etscfg.tsatable)))
J
Jesse Brandeburg 已提交
5965
			dev_dbg(&pf->pdev->dev, "ETS TSA Table changed.\n");
N
Neerav Parikh 已提交
5966 5967 5968 5969 5970 5971 5972
	}

	/* 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
5973
		dev_dbg(&pf->pdev->dev, "PFC config change detected.\n");
N
Neerav Parikh 已提交
5974 5975 5976 5977 5978
	}

	/* Check if APP Table has changed */
	if (memcmp(&new_cfg->app,
		   &old_cfg->app,
5979
		   sizeof(new_cfg->app))) {
N
Neerav Parikh 已提交
5980
		need_reconfig = true;
J
Jesse Brandeburg 已提交
5981
		dev_dbg(&pf->pdev->dev, "APP Table change detected.\n");
5982
	}
N
Neerav Parikh 已提交
5983

S
Shannon Nelson 已提交
5984
	dev_dbg(&pf->pdev->dev, "dcb need_reconfig=%d\n", need_reconfig);
N
Neerav Parikh 已提交
5985 5986 5987 5988 5989 5990 5991 5992 5993 5994 5995 5996 5997 5998 5999 6000 6001 6002 6003
	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;

6004
	/* Not DCB capable or capability disabled */
6005
	if (!(pf->flags & I40E_FLAG_DCB_CAPABLE))
6006 6007
		return ret;

N
Neerav Parikh 已提交
6008 6009 6010
	/* 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
6011
	dev_dbg(&pf->pdev->dev, "LLDP event mib bridge type 0x%x\n", type);
N
Neerav Parikh 已提交
6012 6013 6014 6015 6016
	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;
6017
	dev_dbg(&pf->pdev->dev,
S
Shannon Nelson 已提交
6018
		"LLDP event mib type %s\n", type ? "remote" : "local");
N
Neerav Parikh 已提交
6019 6020 6021 6022 6023 6024 6025 6026
	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;
	}

6027
	/* Store the old configuration */
6028
	tmp_dcbx_cfg = hw->local_dcbx_config;
6029

6030 6031
	/* Reset the old DCBx configuration data */
	memset(&hw->local_dcbx_config, 0, sizeof(hw->local_dcbx_config));
6032 6033
	/* Get updated DCBX data from firmware */
	ret = i40e_get_dcb_config(&pf->hw);
N
Neerav Parikh 已提交
6034
	if (ret) {
6035 6036 6037 6038
		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
6039 6040 6041 6042
		goto exit;
	}

	/* No change detected in DCBX configs */
6043 6044
	if (!memcmp(&tmp_dcbx_cfg, &hw->local_dcbx_config,
		    sizeof(tmp_dcbx_cfg))) {
J
Jesse Brandeburg 已提交
6045
		dev_dbg(&pf->pdev->dev, "No change detected in DCBX configuration.\n");
N
Neerav Parikh 已提交
6046 6047 6048
		goto exit;
	}

6049 6050
	need_reconfig = i40e_dcb_need_reconfig(pf, &tmp_dcbx_cfg,
					       &hw->local_dcbx_config);
N
Neerav Parikh 已提交
6051

6052
	i40e_dcbnl_flush_apps(pf, &tmp_dcbx_cfg, &hw->local_dcbx_config);
N
Neerav Parikh 已提交
6053 6054 6055 6056

	if (!need_reconfig)
		goto exit;

6057
	/* Enable DCB tagging only when more than one TC */
6058
	if (i40e_dcb_get_num_tc(&hw->local_dcbx_config) > 1)
6059 6060 6061 6062
		pf->flags |= I40E_FLAG_DCB_ENABLED;
	else
		pf->flags &= ~I40E_FLAG_DCB_ENABLED;

6063
	set_bit(__I40E_PORT_SUSPENDED, pf->state);
N
Neerav Parikh 已提交
6064 6065 6066 6067 6068 6069
	/* Reconfiguration needed quiesce all VSIs */
	i40e_pf_quiesce_all_vsi(pf);

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

6070 6071
	ret = i40e_resume_port_tx(pf);

6072
	clear_bit(__I40E_PORT_SUSPENDED, pf->state);
6073
	/* In case of error no point in resuming VSIs */
6074 6075 6076
	if (ret)
		goto exit;

6077 6078
	/* Wait for the PF's queues to be disabled */
	ret = i40e_pf_wait_queues_disabled(pf);
6079 6080
	if (ret) {
		/* Schedule PF reset to recover */
6081
		set_bit(__I40E_PF_RESET_REQUESTED, pf->state);
6082 6083
		i40e_service_event_schedule(pf);
	} else {
6084
		i40e_pf_unquiesce_all_vsi(pf);
M
Mitch Williams 已提交
6085 6086
	pf->flags |= (I40E_FLAG_SERVICE_CLIENT_REQUESTED |
		      I40E_FLAG_CLIENT_L2_CHANGE);
6087 6088
	}

N
Neerav Parikh 已提交
6089 6090 6091 6092 6093
exit:
	return ret;
}
#endif /* CONFIG_I40E_DCB */

6094 6095 6096 6097 6098 6099 6100 6101 6102
/**
 * 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();
6103
	i40e_do_reset(pf, reset_flags, true);
6104 6105 6106
	rtnl_unlock();
}

J
Jesse Brandeburg 已提交
6107 6108 6109 6110 6111 6112 6113 6114 6115 6116 6117 6118 6119 6120 6121 6122 6123 6124 6125
/**
 * 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
6126 6127
	dev_dbg(&pf->pdev->dev, "overflow Rx Queue Number = %d QTX_CTL=0x%08x\n",
		queue, qtx_ctl);
J
Jesse Brandeburg 已提交
6128 6129 6130 6131 6132 6133 6134 6135 6136 6137 6138 6139 6140 6141 6142

	/* 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);
	}
}

6143
/**
6144 6145 6146
 * i40e_get_cur_guaranteed_fd_count - Get the consumed guaranteed FD filters
 * @pf: board private structure
 **/
6147
u32 i40e_get_cur_guaranteed_fd_count(struct i40e_pf *pf)
6148
{
6149
	u32 val, fcnt_prog;
6150 6151 6152 6153 6154 6155 6156

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

/**
6157
 * i40e_get_current_fd_count - Get total FD filters programmed for this PF
6158 6159
 * @pf: board private structure
 **/
6160
u32 i40e_get_current_fd_count(struct i40e_pf *pf)
6161
{
6162 6163
	u32 val, fcnt_prog;

6164 6165 6166 6167 6168 6169
	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;
}
6170

6171 6172 6173 6174 6175 6176 6177 6178 6179 6180 6181 6182 6183 6184 6185
/**
 * 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;
}

6186 6187 6188 6189 6190 6191
/**
 * 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)
{
6192
	struct i40e_fdir_filter *filter;
6193
	u32 fcnt_prog, fcnt_avail;
6194
	struct hlist_node *node;
6195

6196
	if (test_bit(__I40E_FD_FLUSH_REQUESTED, pf->state))
6197 6198
		return;

6199
	/* Check if we have enough room to re-enable FDir SB capability. */
6200
	fcnt_prog = i40e_get_global_fd_count(pf);
6201
	fcnt_avail = pf->fdir_pf_filter_count;
6202 6203 6204
	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)) {
6205 6206 6207 6208
		if (pf->flags & I40E_FLAG_FD_SB_AUTO_DISABLED) {
			pf->flags &= ~I40E_FLAG_FD_SB_AUTO_DISABLED;
			if ((pf->flags & I40E_FLAG_FD_SB_ENABLED) &&
			    (I40E_DEBUG_FD & pf->hw.debug_mask))
6209
				dev_info(&pf->pdev->dev, "FD Sideband/ntuple is being enabled since we have space in the table now\n");
6210 6211
		}
	}
6212

6213 6214 6215
	/* We should wait for even more space before re-enabling ATR.
	 * Additionally, we cannot enable ATR as long as we still have TCP SB
	 * rules active.
6216
	 */
6217 6218 6219 6220 6221 6222
	if ((fcnt_prog < (fcnt_avail - I40E_FDIR_BUFFER_HEAD_ROOM_FOR_ATR)) &&
	    (pf->fd_tcp4_filter_cnt == 0)) {
		if (pf->flags & I40E_FLAG_FD_ATR_AUTO_DISABLED) {
			pf->flags &= ~I40E_FLAG_FD_ATR_AUTO_DISABLED;
			if ((pf->flags & I40E_FLAG_FD_ATR_ENABLED) &&
			    (I40E_DEBUG_FD & pf->hw.debug_mask))
6223
				dev_info(&pf->pdev->dev, "ATR is being enabled since we have space in the table and there are no conflicting ntuple rules\n");
6224 6225
		}
	}
6226 6227 6228 6229 6230 6231 6232 6233 6234

	/* 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--;
6235
				pf->fd_inv = 0;
6236 6237 6238
			}
		}
	}
6239 6240
}

6241
#define I40E_MIN_FD_FLUSH_INTERVAL 10
6242
#define I40E_MIN_FD_FLUSH_SB_ATR_UNSTABLE 30
6243 6244 6245 6246 6247 6248
/**
 * 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)
{
6249
	unsigned long min_flush_time;
6250
	int flush_wait_retry = 50;
6251 6252
	bool disable_atr = false;
	int fd_room;
6253 6254
	int reg;

6255 6256 6257
	if (!time_after(jiffies, pf->fd_flush_timestamp +
				 (I40E_MIN_FD_FLUSH_INTERVAL * HZ)))
		return;
6258

6259 6260 6261 6262 6263 6264 6265 6266 6267 6268 6269 6270 6271 6272 6273
	/* 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;
6274
	pf->flags |= I40E_FLAG_FD_ATR_AUTO_DISABLED;
6275 6276 6277 6278 6279 6280 6281 6282 6283 6284 6285 6286 6287 6288 6289 6290 6291 6292
	/* 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]);
6293
		if (!disable_atr && !pf->fd_tcp4_filter_cnt)
6294
			pf->flags &= ~I40E_FLAG_FD_ATR_AUTO_DISABLED;
6295
		clear_bit(__I40E_FD_FLUSH_REQUESTED, pf->state);
6296 6297
		if (I40E_DEBUG_FD & pf->hw.debug_mask)
			dev_info(&pf->pdev->dev, "FD Filter table flushed and FD-SB replayed.\n");
6298 6299 6300 6301 6302 6303 6304
	}
}

/**
 * i40e_get_current_atr_count - Get the count of total FD ATR filters programmed
 * @pf: board private structure
 **/
6305
u32 i40e_get_current_atr_cnt(struct i40e_pf *pf)
6306 6307 6308 6309 6310 6311 6312 6313 6314 6315 6316
{
	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
6317 6318 6319 6320 6321 6322 6323 6324
/**
 * 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 */
6325
	if (test_bit(__I40E_DOWN, pf->state))
J
Jesse Brandeburg 已提交
6326
		return;
6327

6328
	if (test_bit(__I40E_FD_FLUSH_REQUESTED, pf->state))
6329 6330
		i40e_fdir_flush_and_replay(pf);

6331 6332
	i40e_fdir_check_and_reenable(pf);

J
Jesse Brandeburg 已提交
6333 6334 6335 6336 6337 6338 6339 6340 6341
}

/**
 * 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)
{
6342
	if (!vsi || test_bit(__I40E_VSI_DOWN, vsi->state))
J
Jesse Brandeburg 已提交
6343 6344 6345 6346 6347 6348 6349 6350 6351 6352 6353 6354 6355 6356 6357 6358 6359 6360 6361
		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:
6362
	case I40E_VSI_IWARP:
J
Jesse Brandeburg 已提交
6363 6364 6365 6366 6367 6368 6369 6370 6371 6372 6373 6374 6375 6376 6377 6378 6379 6380 6381 6382 6383 6384 6385 6386 6387 6388 6389
	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
6390
	for (i = 0; i < pf->num_alloc_vsi; i++)
J
Jesse Brandeburg 已提交
6391 6392 6393 6394 6395 6396 6397 6398 6399 6400
		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
6401
	struct i40e_vsi *vsi = pf->vsi[pf->lan_vsi];
6402
	u8 new_link_speed, old_link_speed;
6403 6404
	i40e_status status;
	bool new_link, old_link;
J
Jesse Brandeburg 已提交
6405

6406 6407 6408
	/* save off old link status information */
	pf->hw.phy.link_info_old = pf->hw.phy.link_info;

6409 6410 6411
	/* 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
6412
	old_link = (pf->hw.phy.link_info_old.link_info & I40E_AQ_LINK_UP);
6413 6414

	status = i40e_get_link_status(&pf->hw, &new_link);
6415 6416 6417 6418 6419 6420 6421 6422 6423 6424

	/* 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;
6425 6426 6427 6428 6429
		dev_dbg(&pf->pdev->dev, "couldn't get link state, status: %d\n",
			status);
		return;
	}

6430 6431
	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
6432

6433
	if (new_link == old_link &&
6434
	    new_link_speed == old_link_speed &&
6435
	    (test_bit(__I40E_VSI_DOWN, vsi->state) ||
M
Mitch Williams 已提交
6436
	     new_link == netif_carrier_ok(vsi->netdev)))
J
Jesse Brandeburg 已提交
6437
		return;
M
Mitch Williams 已提交
6438

6439
	i40e_print_link_message(vsi, new_link);
J
Jesse Brandeburg 已提交
6440 6441 6442 6443 6444 6445 6446

	/* 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
6447
		i40e_vsi_link_event(vsi, new_link);
J
Jesse Brandeburg 已提交
6448 6449 6450

	if (pf->vf)
		i40e_vc_notify_link_state(pf);
J
Jacob Keller 已提交
6451 6452 6453

	if (pf->flags & I40E_FLAG_PTP)
		i40e_ptp_set_increment(pf);
J
Jesse Brandeburg 已提交
6454 6455 6456
}

/**
6457
 * i40e_watchdog_subtask - periodic checks not using event driven response
J
Jesse Brandeburg 已提交
6458 6459 6460 6461 6462 6463 6464
 * @pf: board private structure
 **/
static void i40e_watchdog_subtask(struct i40e_pf *pf)
{
	int i;

	/* if interface is down do nothing */
6465
	if (test_bit(__I40E_DOWN, pf->state) ||
6466
	    test_bit(__I40E_CONFIG_BUSY, pf->state))
J
Jesse Brandeburg 已提交
6467 6468
		return;

6469 6470 6471 6472 6473 6474
	/* 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;

6475 6476
	if ((pf->flags & I40E_FLAG_LINK_POLLING_ENABLED) ||
	    (pf->flags & I40E_FLAG_TEMP_LINK_POLLING))
S
Shannon Nelson 已提交
6477
		i40e_link_event(pf);
6478

J
Jesse Brandeburg 已提交
6479 6480 6481
	/* 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
6482
	for (i = 0; i < pf->num_alloc_vsi; i++)
J
Jesse Brandeburg 已提交
6483 6484 6485
		if (pf->vsi[i] && pf->vsi[i]->netdev)
			i40e_update_stats(pf->vsi[i]);

6486 6487 6488 6489 6490 6491
	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
6492

6493
	i40e_ptp_rx_hang(pf);
6494
	i40e_ptp_tx_hang(pf);
J
Jesse Brandeburg 已提交
6495 6496 6497 6498 6499 6500 6501 6502 6503 6504
}

/**
 * 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;

6505
	if (test_bit(__I40E_REINIT_REQUESTED, pf->state)) {
J
Jesse Brandeburg 已提交
6506
		reset_flags |= BIT(__I40E_REINIT_REQUESTED);
6507
		clear_bit(__I40E_REINIT_REQUESTED, pf->state);
J
Jesse Brandeburg 已提交
6508
	}
6509
	if (test_bit(__I40E_PF_RESET_REQUESTED, pf->state)) {
J
Jesse Brandeburg 已提交
6510
		reset_flags |= BIT(__I40E_PF_RESET_REQUESTED);
6511
		clear_bit(__I40E_PF_RESET_REQUESTED, pf->state);
J
Jesse Brandeburg 已提交
6512
	}
6513
	if (test_bit(__I40E_CORE_RESET_REQUESTED, pf->state)) {
J
Jesse Brandeburg 已提交
6514
		reset_flags |= BIT(__I40E_CORE_RESET_REQUESTED);
6515
		clear_bit(__I40E_CORE_RESET_REQUESTED, pf->state);
J
Jesse Brandeburg 已提交
6516
	}
6517
	if (test_bit(__I40E_GLOBAL_RESET_REQUESTED, pf->state)) {
J
Jesse Brandeburg 已提交
6518
		reset_flags |= BIT(__I40E_GLOBAL_RESET_REQUESTED);
6519
		clear_bit(__I40E_GLOBAL_RESET_REQUESTED, pf->state);
J
Jesse Brandeburg 已提交
6520
	}
6521 6522 6523
	if (test_bit(__I40E_DOWN_REQUESTED, pf->state)) {
		reset_flags |= BIT(__I40E_DOWN_REQUESTED);
		clear_bit(__I40E_DOWN_REQUESTED, pf->state);
6524
	}
J
Jesse Brandeburg 已提交
6525 6526 6527 6528

	/* If there's a recovery already waiting, it takes
	 * precedence before starting a new reset sequence.
	 */
6529
	if (test_bit(__I40E_RESET_INTR_RECEIVED, pf->state)) {
6530 6531 6532
		i40e_prep_for_reset(pf, false);
		i40e_reset(pf);
		i40e_rebuild(pf, false, false);
J
Jesse Brandeburg 已提交
6533 6534 6535 6536
	}

	/* If we're already down or resetting, just bail */
	if (reset_flags &&
6537
	    !test_bit(__I40E_DOWN, pf->state) &&
6538
	    !test_bit(__I40E_CONFIG_BUSY, pf->state)) {
6539
		i40e_do_reset(pf, reset_flags, false);
6540
	}
J
Jesse Brandeburg 已提交
6541 6542 6543 6544 6545 6546 6547 6548 6549 6550 6551 6552 6553
}

/**
 * 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;

6554 6555 6556 6557 6558 6559 6560 6561
	/* 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);

6562 6563
	/* Check if module meets thermal requirements */
	if (status->phy_type == I40E_PHY_TYPE_NOT_SUPPORTED_HIGH_TEMP) {
6564
		dev_err(&pf->pdev->dev,
6565 6566 6567 6568 6569 6570 6571 6572 6573 6574 6575 6576 6577 6578 6579 6580 6581
			"Rx/Tx is disabled on this device because the module does not meet thermal requirements.\n");
		dev_err(&pf->pdev->dev,
			"Refer to the Intel(R) Ethernet Adapters and Devices User Guide for a list of supported modules.\n");
	} else {
		/* check for unqualified module, if link is down, suppress
		 * the message if link was forced to be down.
		 */
		if ((status->link_info & I40E_AQ_MEDIA_AVAILABLE) &&
		    (!(status->an_info & I40E_AQ_QUALIFIED_MODULE)) &&
		    (!(status->link_info & I40E_AQ_LINK_UP)) &&
		    (!(pf->flags & I40E_FLAG_LINK_DOWN_ON_CLOSE_ENABLED))) {
			dev_err(&pf->pdev->dev,
				"Rx/Tx is disabled on this device because an unsupported SFP module type was detected.\n");
			dev_err(&pf->pdev->dev,
				"Refer to the Intel(R) Ethernet Adapters and Devices User Guide for a list of supported modules.\n");
		}
	}
J
Jesse Brandeburg 已提交
6582 6583 6584 6585 6586 6587 6588 6589 6590 6591 6592 6593 6594
}

/**
 * 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;
6595
	u32 oldval;
J
Jesse Brandeburg 已提交
6596 6597
	u32 val;

6598
	/* Do not run clean AQ when PF reset fails */
6599
	if (test_bit(__I40E_RESET_FAILED, pf->state))
6600 6601
		return;

6602 6603 6604 6605
	/* 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
6606 6607
		if (hw->debug_mask & I40E_DEBUG_AQ)
			dev_info(&pf->pdev->dev, "ARQ VF Error detected\n");
6608 6609 6610
		val &= ~I40E_PF_ARQLEN_ARQVFE_MASK;
	}
	if (val & I40E_PF_ARQLEN_ARQOVFL_MASK) {
M
Mitch Williams 已提交
6611 6612
		if (hw->debug_mask & I40E_DEBUG_AQ)
			dev_info(&pf->pdev->dev, "ARQ Overflow Error detected\n");
6613
		val &= ~I40E_PF_ARQLEN_ARQOVFL_MASK;
6614
		pf->arq_overflows++;
6615 6616
	}
	if (val & I40E_PF_ARQLEN_ARQCRIT_MASK) {
M
Mitch Williams 已提交
6617 6618
		if (hw->debug_mask & I40E_DEBUG_AQ)
			dev_info(&pf->pdev->dev, "ARQ Critical Error detected\n");
6619 6620 6621 6622 6623 6624 6625 6626
		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
6627 6628
		if (pf->hw.debug_mask & I40E_DEBUG_AQ)
			dev_info(&pf->pdev->dev, "ASQ VF Error detected\n");
6629 6630 6631
		val &= ~I40E_PF_ATQLEN_ATQVFE_MASK;
	}
	if (val & I40E_PF_ATQLEN_ATQOVFL_MASK) {
M
Mitch Williams 已提交
6632 6633
		if (pf->hw.debug_mask & I40E_DEBUG_AQ)
			dev_info(&pf->pdev->dev, "ASQ Overflow Error detected\n");
6634 6635 6636
		val &= ~I40E_PF_ATQLEN_ATQOVFL_MASK;
	}
	if (val & I40E_PF_ATQLEN_ATQCRIT_MASK) {
M
Mitch Williams 已提交
6637 6638
		if (pf->hw.debug_mask & I40E_DEBUG_AQ)
			dev_info(&pf->pdev->dev, "ASQ Critical Error detected\n");
6639 6640 6641 6642 6643
		val &= ~I40E_PF_ATQLEN_ATQCRIT_MASK;
	}
	if (oldval != val)
		wr32(&pf->hw, pf->hw.aq.asq.len, val);

M
Mitch Williams 已提交
6644 6645
	event.buf_len = I40E_MAX_AQ_BUF_SIZE;
	event.msg_buf = kzalloc(event.buf_len, GFP_KERNEL);
J
Jesse Brandeburg 已提交
6646 6647 6648 6649 6650
	if (!event.msg_buf)
		return;

	do {
		ret = i40e_clean_arq_element(hw, &event, &pending);
M
Mitch Williams 已提交
6651
		if (ret == I40E_ERR_ADMIN_QUEUE_NO_WORK)
J
Jesse Brandeburg 已提交
6652
			break;
M
Mitch Williams 已提交
6653
		else if (ret) {
J
Jesse Brandeburg 已提交
6654 6655 6656 6657 6658 6659 6660 6661 6662 6663 6664 6665 6666 6667 6668 6669
			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
6670
					event.msg_len);
J
Jesse Brandeburg 已提交
6671 6672
			break;
		case i40e_aqc_opc_lldp_update_mib:
J
Jesse Brandeburg 已提交
6673
			dev_dbg(&pf->pdev->dev, "ARQ: Update LLDP MIB event received\n");
N
Neerav Parikh 已提交
6674 6675 6676 6677 6678
#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
6679 6680
			break;
		case i40e_aqc_opc_event_lan_overflow:
J
Jesse Brandeburg 已提交
6681
			dev_dbg(&pf->pdev->dev, "ARQ LAN queue overflow event received\n");
J
Jesse Brandeburg 已提交
6682 6683
			i40e_handle_lan_overflow_event(pf, &event);
			break;
6684 6685 6686
		case i40e_aqc_opc_send_msg_to_peer:
			dev_info(&pf->pdev->dev, "ARQ: Msg from other pf\n");
			break;
6687 6688
		case i40e_aqc_opc_nvm_erase:
		case i40e_aqc_opc_nvm_update:
6689
		case i40e_aqc_opc_oem_post_update:
6690 6691 6692
			i40e_debug(&pf->hw, I40E_DEBUG_NVM,
				   "ARQ NVM operation 0x%04x completed\n",
				   opcode);
6693
			break;
J
Jesse Brandeburg 已提交
6694 6695
		default:
			dev_info(&pf->pdev->dev,
6696
				 "ARQ: Unknown event 0x%04x ignored\n",
6697
				 opcode);
J
Jesse Brandeburg 已提交
6698 6699
			break;
		}
6700 6701 6702
	} while (i++ < pf->adminq_work_limit);

	if (i < pf->adminq_work_limit)
6703
		clear_bit(__I40E_ADMINQ_EVENT_PENDING, pf->state);
J
Jesse Brandeburg 已提交
6704 6705 6706 6707 6708 6709 6710 6711 6712 6713

	/* 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);
}

6714 6715 6716 6717 6718 6719 6720 6721 6722 6723 6724 6725 6726 6727 6728
/**
 * 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);
6729
			set_bit(__I40E_BAD_EEPROM, pf->state);
6730 6731 6732
		}
	}

6733
	if (!err && test_bit(__I40E_BAD_EEPROM, pf->state)) {
6734
		dev_info(&pf->pdev->dev, "eeprom check passed, Tx/Rx traffic enabled\n");
6735
		clear_bit(__I40E_BAD_EEPROM, pf->state);
6736 6737 6738
	}
}

6739 6740
/**
 * i40e_enable_pf_switch_lb
6741
 * @pf: pointer to the PF structure
6742 6743 6744 6745 6746 6747 6748
 *
 * 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;
6749
	int ret;
6750 6751 6752 6753

	ctxt.seid = pf->main_vsi_seid;
	ctxt.pf_num = pf->hw.pf_id;
	ctxt.vf_num = 0;
6754 6755
	ret = i40e_aq_get_vsi_params(&pf->hw, &ctxt, NULL);
	if (ret) {
6756
		dev_info(&pf->pdev->dev,
6757 6758 6759
			 "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));
6760 6761 6762 6763 6764 6765
		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);

6766 6767
	ret = i40e_aq_update_vsi_params(&vsi->back->hw, &ctxt, NULL);
	if (ret) {
6768
		dev_info(&pf->pdev->dev,
6769 6770 6771
			 "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));
6772 6773 6774 6775 6776
	}
}

/**
 * i40e_disable_pf_switch_lb
6777
 * @pf: pointer to the PF structure
6778 6779 6780 6781 6782 6783 6784
 *
 * 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;
6785
	int ret;
6786 6787 6788 6789

	ctxt.seid = pf->main_vsi_seid;
	ctxt.pf_num = pf->hw.pf_id;
	ctxt.vf_num = 0;
6790 6791
	ret = i40e_aq_get_vsi_params(&pf->hw, &ctxt, NULL);
	if (ret) {
6792
		dev_info(&pf->pdev->dev,
6793 6794 6795
			 "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));
6796 6797 6798 6799 6800 6801
		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);

6802 6803
	ret = i40e_aq_update_vsi_params(&vsi->back->hw, &ctxt, NULL);
	if (ret) {
6804
		dev_info(&pf->pdev->dev,
6805 6806 6807
			 "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));
6808 6809 6810
	}
}

6811 6812 6813 6814 6815 6816 6817 6818 6819 6820 6821 6822
/**
 * 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;

6823 6824 6825
	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");
6826 6827 6828 6829 6830 6831
	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
6832 6833 6834 6835 6836 6837 6838 6839 6840 6841 6842 6843 6844 6845 6846 6847 6848
/**
 * 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
6849
	for (v = 0; v < pf->num_alloc_vsi && !ctl_vsi; v++) {
J
Jesse Brandeburg 已提交
6850 6851 6852 6853 6854 6855 6856 6857 6858 6859 6860 6861 6862 6863 6864 6865 6866 6867
		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,
6868 6869
			 "rebuild of veb_idx %d owner VSI failed: %d\n",
			 veb->idx, ret);
J
Jesse Brandeburg 已提交
6870 6871 6872 6873 6874 6875 6876 6877 6878
		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;

6879 6880 6881 6882
	if (pf->flags & I40E_FLAG_VEB_MODE_ENABLED)
		veb->bridge_mode = BRIDGE_MODE_VEB;
	else
		veb->bridge_mode = BRIDGE_MODE_VEPA;
6883
	i40e_config_bridge_mode(veb);
6884

J
Jesse Brandeburg 已提交
6885
	/* create the remaining VSIs attached to this VEB */
M
Mitch Williams 已提交
6886
	for (v = 0; v < pf->num_alloc_vsi; v++) {
J
Jesse Brandeburg 已提交
6887 6888 6889 6890 6891
		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
6892

J
Jesse Brandeburg 已提交
6893 6894 6895 6896 6897 6898 6899 6900 6901 6902 6903 6904 6905 6906 6907 6908 6909 6910 6911 6912 6913 6914 6915 6916 6917 6918 6919 6920 6921 6922 6923 6924 6925 6926 6927 6928 6929 6930 6931 6932 6933 6934 6935 6936 6937 6938 6939 6940 6941 6942 6943 6944 6945 6946 6947 6948
			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,
6949 6950 6951 6952
				 "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
6953 6954 6955 6956 6957 6958 6959 6960 6961 6962 6963 6964 6965 6966 6967
			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);

6968 6969 6970 6971 6972 6973 6974 6975 6976
#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
6977 6978 6979
	return 0;
}

6980 6981
static int i40e_vsi_clear(struct i40e_vsi *vsi);

J
Jesse Brandeburg 已提交
6982
/**
6983
 * i40e_fdir_sb_setup - initialize the Flow Director resources for Sideband
J
Jesse Brandeburg 已提交
6984 6985
 * @pf: board private structure
 **/
6986
static void i40e_fdir_sb_setup(struct i40e_pf *pf)
J
Jesse Brandeburg 已提交
6987 6988 6989
{
	struct i40e_vsi *vsi;

6990 6991 6992 6993 6994 6995 6996 6997 6998
	/* 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};
6999
		int i;
7000 7001 7002 7003 7004

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

7005
	if (!(pf->flags & I40E_FLAG_FD_SB_ENABLED))
J
Jesse Brandeburg 已提交
7006 7007
		return;

7008
	/* find existing VSI and see if it needs configuring */
7009
	vsi = i40e_find_vsi_by_type(pf, I40E_VSI_FDIR);
7010 7011

	/* create a new VSI if none exists */
J
Jesse Brandeburg 已提交
7012
	if (!vsi) {
7013 7014
		vsi = i40e_vsi_setup(pf, I40E_VSI_FDIR,
				     pf->vsi[pf->lan_vsi]->seid, 0);
J
Jesse Brandeburg 已提交
7015 7016
		if (!vsi) {
			dev_info(&pf->pdev->dev, "Couldn't create FDir VSI\n");
7017 7018
			pf->flags &= ~I40E_FLAG_FD_SB_ENABLED;
			return;
J
Jesse Brandeburg 已提交
7019
		}
7020
	}
J
Jesse Brandeburg 已提交
7021

7022
	i40e_vsi_setup_irqhandler(vsi, i40e_fdir_clean_ring);
J
Jesse Brandeburg 已提交
7023 7024 7025 7026 7027 7028 7029 7030
}

/**
 * i40e_fdir_teardown - release the Flow Director resources
 * @pf: board private structure
 **/
static void i40e_fdir_teardown(struct i40e_pf *pf)
{
7031
	struct i40e_vsi *vsi;
J
Jesse Brandeburg 已提交
7032

7033
	i40e_fdir_filter_exit(pf);
7034 7035 7036
	vsi = i40e_find_vsi_by_type(pf, I40E_VSI_FDIR);
	if (vsi)
		i40e_vsi_release(vsi);
J
Jesse Brandeburg 已提交
7037 7038 7039
}

/**
A
Anjali Singhai Jain 已提交
7040
 * i40e_prep_for_reset - prep for the core to reset
J
Jesse Brandeburg 已提交
7041
 * @pf: board private structure
7042 7043
 * @lock_acquired: indicates whether or not the lock has been acquired
 * before this function was called.
J
Jesse Brandeburg 已提交
7044
 *
7045
 * Close up the VFs and other things in prep for PF Reset.
A
Anjali Singhai Jain 已提交
7046
  **/
7047
static void i40e_prep_for_reset(struct i40e_pf *pf, bool lock_acquired)
J
Jesse Brandeburg 已提交
7048 7049
{
	struct i40e_hw *hw = &pf->hw;
7050
	i40e_status ret = 0;
J
Jesse Brandeburg 已提交
7051 7052
	u32 v;

7053 7054
	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
7055
		return;
M
Mitch Williams 已提交
7056 7057
	if (i40e_check_asq_alive(&pf->hw))
		i40e_vc_notify_reset(pf);
J
Jesse Brandeburg 已提交
7058

J
Jesse Brandeburg 已提交
7059
	dev_dbg(&pf->pdev->dev, "Tearing down internal switch for reset\n");
J
Jesse Brandeburg 已提交
7060 7061

	/* quiesce the VSIs and their queues that are not already DOWN */
7062 7063 7064
	/* pf_quiesce_all_vsi modifies netdev structures -rtnl_lock needed */
	if (!lock_acquired)
		rtnl_lock();
J
Jesse Brandeburg 已提交
7065
	i40e_pf_quiesce_all_vsi(pf);
7066 7067
	if (!lock_acquired)
		rtnl_unlock();
J
Jesse Brandeburg 已提交
7068

M
Mitch Williams 已提交
7069
	for (v = 0; v < pf->num_alloc_vsi; v++) {
J
Jesse Brandeburg 已提交
7070 7071 7072 7073 7074 7075
		if (pf->vsi[v])
			pf->vsi[v]->seid = 0;
	}

	i40e_shutdown_adminq(&pf->hw);

A
Anjali Singhai Jain 已提交
7076
	/* call shutdown HMC */
7077 7078
	if (hw->hmc.hmc_obj) {
		ret = i40e_shutdown_lan_hmc(hw);
S
Shannon Nelson 已提交
7079
		if (ret)
7080 7081
			dev_warn(&pf->pdev->dev,
				 "shutdown_lan_hmc failed: %d\n", ret);
A
Anjali Singhai Jain 已提交
7082 7083 7084
	}
}

J
Jesse Brandeburg 已提交
7085 7086 7087 7088 7089 7090 7091 7092 7093 7094 7095 7096
/**
 * 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;
7097
	strlcpy(dv.driver_string, DRV_VERSION, sizeof(dv.driver_string));
J
Jesse Brandeburg 已提交
7098 7099 7100
	i40e_aq_send_driver_version(&pf->hw, &dv, NULL);
}

7101 7102 7103 7104 7105 7106 7107 7108 7109 7110 7111 7112 7113 7114 7115 7116 7117 7118 7119 7120 7121 7122 7123 7124 7125 7126 7127 7128 7129 7130 7131 7132 7133 7134 7135 7136 7137 7138 7139 7140 7141 7142 7143 7144 7145
/**
 * i40e_get_oem_version - get OEM specific version information
 * @hw: pointer to the hardware structure
 **/
static void i40e_get_oem_version(struct i40e_hw *hw)
{
	u16 block_offset = 0xffff;
	u16 block_length = 0;
	u16 capabilities = 0;
	u16 gen_snap = 0;
	u16 release = 0;

#define I40E_SR_NVM_OEM_VERSION_PTR		0x1B
#define I40E_NVM_OEM_LENGTH_OFFSET		0x00
#define I40E_NVM_OEM_CAPABILITIES_OFFSET	0x01
#define I40E_NVM_OEM_GEN_OFFSET			0x02
#define I40E_NVM_OEM_RELEASE_OFFSET		0x03
#define I40E_NVM_OEM_CAPABILITIES_MASK		0x000F
#define I40E_NVM_OEM_LENGTH			3

	/* Check if pointer to OEM version block is valid. */
	i40e_read_nvm_word(hw, I40E_SR_NVM_OEM_VERSION_PTR, &block_offset);
	if (block_offset == 0xffff)
		return;

	/* Check if OEM version block has correct length. */
	i40e_read_nvm_word(hw, block_offset + I40E_NVM_OEM_LENGTH_OFFSET,
			   &block_length);
	if (block_length < I40E_NVM_OEM_LENGTH)
		return;

	/* Check if OEM version format is as expected. */
	i40e_read_nvm_word(hw, block_offset + I40E_NVM_OEM_CAPABILITIES_OFFSET,
			   &capabilities);
	if ((capabilities & I40E_NVM_OEM_CAPABILITIES_MASK) != 0)
		return;

	i40e_read_nvm_word(hw, block_offset + I40E_NVM_OEM_GEN_OFFSET,
			   &gen_snap);
	i40e_read_nvm_word(hw, block_offset + I40E_NVM_OEM_RELEASE_OFFSET,
			   &release);
	hw->nvm.oem_ver = (gen_snap << I40E_OEM_SNAP_SHIFT) | release;
	hw->nvm.eetrack = I40E_OEM_EETRACK_ID;
}

A
Anjali Singhai Jain 已提交
7146
/**
7147
 * i40e_reset - wait for core reset to finish reset, reset pf if corer not seen
A
Anjali Singhai Jain 已提交
7148 7149
 * @pf: board private structure
 **/
7150
static int i40e_reset(struct i40e_pf *pf)
A
Anjali Singhai Jain 已提交
7151 7152 7153 7154
{
	struct i40e_hw *hw = &pf->hw;
	i40e_status ret;

J
Jesse Brandeburg 已提交
7155
	ret = i40e_pf_reset(hw);
7156
	if (ret) {
J
Jesse Brandeburg 已提交
7157
		dev_info(&pf->pdev->dev, "PF reset failed, %d\n", ret);
7158 7159
		set_bit(__I40E_RESET_FAILED, pf->state);
		clear_bit(__I40E_RESET_RECOVERY_PENDING, pf->state);
7160 7161
	} else {
		pf->pfr_count++;
7162
	}
7163 7164 7165 7166 7167 7168 7169 7170 7171 7172 7173 7174 7175 7176 7177 7178 7179
	return ret;
}

/**
 * i40e_rebuild - rebuild using a saved config
 * @pf: board private structure
 * @reinit: if the Main VSI needs to re-initialized.
 * @lock_acquired: indicates whether or not the lock has been acquired
 * before this function was called.
 **/
static void i40e_rebuild(struct i40e_pf *pf, bool reinit, bool lock_acquired)
{
	struct i40e_hw *hw = &pf->hw;
	u8 set_fc_aq_fail = 0;
	i40e_status ret;
	u32 val;
	int v;
J
Jesse Brandeburg 已提交
7180

7181
	if (test_bit(__I40E_DOWN, pf->state))
7182
		goto clear_recovery;
J
Jesse Brandeburg 已提交
7183
	dev_dbg(&pf->pdev->dev, "Rebuilding internal switch\n");
J
Jesse Brandeburg 已提交
7184 7185 7186 7187

	/* rebuild the basics for the AdminQ, HMC, and initial HW switch */
	ret = i40e_init_adminq(&pf->hw);
	if (ret) {
7188 7189 7190
		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));
7191
		goto clear_recovery;
J
Jesse Brandeburg 已提交
7192
	}
7193
	i40e_get_oem_version(&pf->hw);
J
Jesse Brandeburg 已提交
7194

7195
	/* re-verify the eeprom if we just had an EMP reset */
7196
	if (test_and_clear_bit(__I40E_EMP_RESET_INTR_RECEIVED, pf->state))
7197 7198
		i40e_verify_eeprom(pf);

7199
	i40e_clear_pxe_mode(hw);
J
Jesse Brandeburg 已提交
7200
	ret = i40e_get_capabilities(pf);
7201
	if (ret)
J
Jesse Brandeburg 已提交
7202 7203 7204
		goto end_core_reset;

	ret = i40e_init_lan_hmc(hw, hw->func_caps.num_tx_qp,
7205
				hw->func_caps.num_rx_qp, 0, 0);
J
Jesse Brandeburg 已提交
7206 7207 7208 7209 7210 7211 7212 7213 7214 7215
	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
7216 7217 7218
#ifdef CONFIG_I40E_DCB
	ret = i40e_init_pf_dcb(pf);
	if (ret) {
7219 7220 7221
		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
7222 7223
	}
#endif /* CONFIG_I40E_DCB */
J
Jesse Brandeburg 已提交
7224
	/* do basic switch setup */
7225 7226
	if (!lock_acquired)
		rtnl_lock();
7227
	ret = i40e_setup_pf_switch(pf, reinit);
J
Jesse Brandeburg 已提交
7228
	if (ret)
7229
		goto end_unlock;
J
Jesse Brandeburg 已提交
7230

S
Shannon Nelson 已提交
7231 7232
	/* The driver only wants link up/down and module qualification
	 * reports from firmware.  Note the negative logic.
J
Jesse Brandeburg 已提交
7233 7234
	 */
	ret = i40e_aq_set_phy_int_mask(&pf->hw,
S
Shannon Nelson 已提交
7235
				       ~(I40E_AQ_EVENT_LINK_UPDOWN |
7236
					 I40E_AQ_EVENT_MEDIA_NA |
S
Shannon Nelson 已提交
7237
					 I40E_AQ_EVENT_MODULE_QUAL_FAIL), NULL);
J
Jesse Brandeburg 已提交
7238
	if (ret)
7239 7240 7241
		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
7242

7243 7244 7245
	/* make sure our flow control settings are restored */
	ret = i40e_set_fc(&pf->hw, &set_fc_aq_fail, true);
	if (ret)
7246 7247 7248
		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));
7249

J
Jesse Brandeburg 已提交
7250 7251 7252 7253 7254 7255 7256 7257
	/* 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
7258
		dev_dbg(&pf->pdev->dev, "attempting to rebuild switch\n");
J
Jesse Brandeburg 已提交
7259 7260 7261 7262 7263 7264 7265 7266 7267 7268 7269 7270 7271 7272 7273 7274 7275 7276 7277 7278 7279 7280 7281 7282 7283 7284 7285 7286 7287 7288 7289 7290 7291 7292 7293
		/* 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
7294
		dev_dbg(&pf->pdev->dev, "attempting to rebuild PF VSI\n");
J
Jesse Brandeburg 已提交
7295 7296 7297 7298 7299
		/* 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);
7300
			goto end_unlock;
J
Jesse Brandeburg 已提交
7301 7302 7303
		}
	}

7304 7305 7306 7307 7308 7309 7310 7311 7312 7313 7314 7315 7316 7317
	/* 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);
	}

7318
	if (pf->hw_features & I40E_HW_RESTART_AUTONEG) {
7319 7320 7321
		msleep(75);
		ret = i40e_aq_set_link_restart_an(&pf->hw, true, NULL);
		if (ret)
7322 7323 7324 7325
			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));
7326
	}
J
Jesse Brandeburg 已提交
7327 7328 7329 7330
	/* reinit the misc interrupt */
	if (pf->flags & I40E_FLAG_MSIX_ENABLED)
		ret = i40e_setup_misc_vector(pf);

7331 7332 7333 7334 7335 7336 7337 7338 7339
	/* 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
7340 7341 7342
	/* restart the VSIs that were rebuilt and running before the reset */
	i40e_pf_unquiesce_all_vsi(pf);

7343 7344 7345 7346
	/* Release the RTNL lock before we start resetting VFs */
	if (!lock_acquired)
		rtnl_unlock();

7347
	i40e_reset_all_vfs(pf, true);
M
Mitch Williams 已提交
7348

J
Jesse Brandeburg 已提交
7349
	/* tell the firmware that we're starting */
J
Jesse Brandeburg 已提交
7350
	i40e_send_version(pf);
J
Jesse Brandeburg 已提交
7351

7352 7353 7354
	/* We've already released the lock, so don't do it again */
	goto end_core_reset;

7355
end_unlock:
7356 7357
	if (!lock_acquired)
		rtnl_unlock();
J
Jesse Brandeburg 已提交
7358
end_core_reset:
7359
	clear_bit(__I40E_RESET_FAILED, pf->state);
7360
clear_recovery:
7361
	clear_bit(__I40E_RESET_RECOVERY_PENDING, pf->state);
J
Jesse Brandeburg 已提交
7362 7363
}

7364 7365 7366 7367 7368 7369 7370 7371 7372 7373 7374 7375 7376 7377 7378 7379 7380 7381 7382 7383
/**
 * i40e_reset_and_rebuild - reset and rebuild using a saved config
 * @pf: board private structure
 * @reinit: if the Main VSI needs to re-initialized.
 * @lock_acquired: indicates whether or not the lock has been acquired
 * before this function was called.
 **/
static void i40e_reset_and_rebuild(struct i40e_pf *pf, bool reinit,
				   bool lock_acquired)
{
	int ret;
	/* 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_reset(pf);
	if (!ret)
		i40e_rebuild(pf, reinit, lock_acquired);
}

A
Anjali Singhai Jain 已提交
7384
/**
7385
 * i40e_handle_reset_warning - prep for the PF to reset, reset and rebuild
A
Anjali Singhai Jain 已提交
7386 7387 7388 7389
 * @pf: board private structure
 *
 * Close up the VFs and other things in prep for a Core Reset,
 * then get ready to rebuild the world.
7390 7391
 * @lock_acquired: indicates whether or not the lock has been acquired
 * before this function was called.
A
Anjali Singhai Jain 已提交
7392
 **/
7393
static void i40e_handle_reset_warning(struct i40e_pf *pf, bool lock_acquired)
A
Anjali Singhai Jain 已提交
7394
{
7395 7396
	i40e_prep_for_reset(pf, lock_acquired);
	i40e_reset_and_rebuild(pf, false, lock_acquired);
A
Anjali Singhai Jain 已提交
7397 7398
}

J
Jesse Brandeburg 已提交
7399 7400
/**
 * i40e_handle_mdd_event
7401
 * @pf: pointer to the PF structure
J
Jesse Brandeburg 已提交
7402 7403 7404 7405 7406 7407 7408
 *
 * 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;
7409
	bool pf_mdd_detected = false;
J
Jesse Brandeburg 已提交
7410 7411 7412 7413
	struct i40e_vf *vf;
	u32 reg;
	int i;

7414
	if (!test_bit(__I40E_MDD_EVENT_PENDING, pf->state))
J
Jesse Brandeburg 已提交
7415 7416 7417 7418 7419
		return;

	/* find what triggered the MDD event */
	reg = rd32(hw, I40E_GL_MDET_TX);
	if (reg & I40E_GL_MDET_TX_VALID_MASK) {
7420 7421
		u8 pf_num = (reg & I40E_GL_MDET_TX_PF_NUM_MASK) >>
				I40E_GL_MDET_TX_PF_NUM_SHIFT;
7422
		u16 vf_num = (reg & I40E_GL_MDET_TX_VF_NUM_MASK) >>
7423
				I40E_GL_MDET_TX_VF_NUM_SHIFT;
7424
		u8 event = (reg & I40E_GL_MDET_TX_EVENT_MASK) >>
7425
				I40E_GL_MDET_TX_EVENT_SHIFT;
7426 7427 7428
		u16 queue = ((reg & I40E_GL_MDET_TX_QUEUE_MASK) >>
				I40E_GL_MDET_TX_QUEUE_SHIFT) -
				pf->hw.func_caps.base_queue;
7429
		if (netif_msg_tx_err(pf))
7430
			dev_info(&pf->pdev->dev, "Malicious Driver Detection event 0x%02x on TX queue %d PF number 0x%02x VF number 0x%02x\n",
7431
				 event, queue, pf_num, vf_num);
J
Jesse Brandeburg 已提交
7432 7433 7434 7435 7436
		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) {
7437 7438
		u8 func = (reg & I40E_GL_MDET_RX_FUNCTION_MASK) >>
				I40E_GL_MDET_RX_FUNCTION_SHIFT;
7439
		u8 event = (reg & I40E_GL_MDET_RX_EVENT_MASK) >>
7440
				I40E_GL_MDET_RX_EVENT_SHIFT;
7441 7442 7443
		u16 queue = ((reg & I40E_GL_MDET_RX_QUEUE_MASK) >>
				I40E_GL_MDET_RX_QUEUE_SHIFT) -
				pf->hw.func_caps.base_queue;
7444 7445 7446
		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
7447 7448 7449 7450
		wr32(hw, I40E_GL_MDET_RX, 0xffffffff);
		mdd_detected = true;
	}

7451 7452 7453 7454
	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);
7455
			dev_info(&pf->pdev->dev, "TX driver issue detected, PF reset issued\n");
7456 7457 7458 7459 7460
			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);
7461
			dev_info(&pf->pdev->dev, "RX driver issue detected, PF reset issued\n");
7462 7463 7464 7465
			pf_mdd_detected = true;
		}
		/* Queue belongs to the PF, initiate a reset */
		if (pf_mdd_detected) {
7466
			set_bit(__I40E_PF_RESET_REQUESTED, pf->state);
7467 7468 7469 7470
			i40e_service_event_schedule(pf);
		}
	}

J
Jesse Brandeburg 已提交
7471 7472 7473 7474 7475 7476 7477
	/* 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++;
7478 7479
			dev_info(&pf->pdev->dev, "TX driver issue detected on VF %d\n",
				 i);
J
Jesse Brandeburg 已提交
7480 7481 7482 7483 7484 7485
		}

		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++;
7486 7487
			dev_info(&pf->pdev->dev, "RX driver issue detected on VF %d\n",
				 i);
J
Jesse Brandeburg 已提交
7488 7489 7490 7491 7492 7493 7494
		}

		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");
7495
			set_bit(I40E_VF_STATE_DISABLED, &vf->vf_states);
J
Jesse Brandeburg 已提交
7496 7497 7498 7499
		}
	}

	/* re-enable mdd interrupt cause */
7500
	clear_bit(__I40E_MDD_EVENT_PENDING, pf->state);
J
Jesse Brandeburg 已提交
7501 7502 7503 7504 7505 7506
	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);
}

7507 7508 7509 7510 7511 7512 7513 7514 7515 7516 7517 7518
static const char *i40e_tunnel_name(struct i40e_udp_port_config *port)
{
	switch (port->type) {
	case UDP_TUNNEL_TYPE_VXLAN:
		return "vxlan";
	case UDP_TUNNEL_TYPE_GENEVE:
		return "geneve";
	default:
		return "unknown";
	}
}

7519 7520 7521 7522 7523 7524 7525 7526 7527 7528 7529 7530 7531 7532 7533 7534 7535
/**
 * i40e_sync_udp_filters - Trigger a sync event for existing UDP filters
 * @pf: board private structure
 **/
static void i40e_sync_udp_filters(struct i40e_pf *pf)
{
	int i;

	/* loop through and set pending bit for all active UDP filters */
	for (i = 0; i < I40E_MAX_PF_UDP_OFFLOAD_PORTS; i++) {
		if (pf->udp_ports[i].port)
			pf->pending_udp_bitmap |= BIT_ULL(i);
	}

	pf->flags |= I40E_FLAG_UDP_FILTER_SYNC;
}

7536
/**
7537
 * i40e_sync_udp_filters_subtask - Sync the VSI filter list with HW
7538 7539
 * @pf: board private structure
 **/
7540
static void i40e_sync_udp_filters_subtask(struct i40e_pf *pf)
7541 7542 7543
{
	struct i40e_hw *hw = &pf->hw;
	i40e_status ret;
7544
	u16 port;
7545 7546
	int i;

7547
	if (!(pf->flags & I40E_FLAG_UDP_FILTER_SYNC))
7548 7549
		return;

7550
	pf->flags &= ~I40E_FLAG_UDP_FILTER_SYNC;
7551 7552

	for (i = 0; i < I40E_MAX_PF_UDP_OFFLOAD_PORTS; i++) {
7553 7554
		if (pf->pending_udp_bitmap & BIT_ULL(i)) {
			pf->pending_udp_bitmap &= ~BIT_ULL(i);
7555
			port = pf->udp_ports[i].port;
S
Shannon Nelson 已提交
7556
			if (port)
7557 7558 7559
				ret = i40e_aq_add_udp_tunnel(hw, port,
							pf->udp_ports[i].type,
							NULL, NULL);
S
Shannon Nelson 已提交
7560 7561
			else
				ret = i40e_aq_del_udp_tunnel(hw, i, NULL);
7562 7563

			if (ret) {
7564 7565 7566 7567 7568 7569 7570 7571
				dev_info(&pf->pdev->dev,
					 "%s %s port %d, index %d failed, err %s aq_err %s\n",
					 i40e_tunnel_name(&pf->udp_ports[i]),
					 port ? "add" : "delete",
					 port, i,
					 i40e_stat_str(&pf->hw, ret),
					 i40e_aq_str(&pf->hw,
						     pf->hw.aq.asq_last_status));
7572
				pf->udp_ports[i].port = 0;
7573 7574 7575 7576 7577
			}
		}
	}
}

J
Jesse Brandeburg 已提交
7578 7579 7580 7581 7582 7583 7584 7585 7586 7587 7588
/**
 * 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;

7589
	/* don't bother with service tasks if a reset is in progress */
7590
	if (test_bit(__I40E_RESET_RECOVERY_PENDING, pf->state))
7591 7592
		return;

7593
	if (test_and_set_bit(__I40E_SERVICE_SCHED, pf->state))
M
Mitch Williams 已提交
7594 7595
		return;

7596
	i40e_detect_recover_hung(pf);
7597
	i40e_sync_filters_subtask(pf);
J
Jesse Brandeburg 已提交
7598 7599 7600 7601 7602
	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
7603 7604 7605 7606 7607 7608 7609 7610 7611 7612 7613 7614
	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
7615
	i40e_sync_filters_subtask(pf);
7616
	i40e_sync_udp_filters_subtask(pf);
J
Jesse Brandeburg 已提交
7617 7618
	i40e_clean_adminq_subtask(pf);

M
Mitch Williams 已提交
7619 7620
	/* flush memory to make sure state is correct before next watchdog */
	smp_mb__before_atomic();
7621
	clear_bit(__I40E_SERVICE_SCHED, pf->state);
J
Jesse Brandeburg 已提交
7622 7623 7624 7625 7626 7627

	/* 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)) ||
7628 7629 7630
	    test_bit(__I40E_ADMINQ_EVENT_PENDING, pf->state)		 ||
	    test_bit(__I40E_MDD_EVENT_PENDING, pf->state)		 ||
	    test_bit(__I40E_VFLR_EVENT_PENDING, pf->state))
J
Jesse Brandeburg 已提交
7631 7632 7633 7634 7635 7636 7637 7638 7639 7640 7641 7642 7643 7644 7645 7646 7647 7648 7649 7650 7651 7652 7653 7654 7655 7656 7657 7658 7659 7660 7661 7662 7663 7664 7665 7666 7667 7668 7669 7670
		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);
7671
		vsi->num_q_vectors = pf->num_fdsb_msix;
J
Jesse Brandeburg 已提交
7672 7673 7674 7675 7676 7677 7678 7679 7680 7681 7682 7683 7684 7685 7686 7687 7688 7689 7690 7691 7692 7693 7694
		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
7695 7696 7697
/**
 * i40e_vsi_alloc_arrays - Allocate queue and vector pointer arrays for the vsi
 * @type: VSI pointer
7698
 * @alloc_qvectors: a bool to specify if q_vectors need to be allocated.
A
Anjali Singhai Jain 已提交
7699 7700 7701 7702
 *
 * On error: returns error code (negative)
 * On success: returns 0
 **/
7703
static int i40e_vsi_alloc_arrays(struct i40e_vsi *vsi, bool alloc_qvectors)
A
Anjali Singhai Jain 已提交
7704
{
7705
	struct i40e_ring **next_rings;
A
Anjali Singhai Jain 已提交
7706 7707 7708
	int size;
	int ret = 0;

7709 7710 7711
	/* allocate memory for both Tx, XDP Tx and Rx ring pointers */
	size = sizeof(struct i40e_ring *) * vsi->alloc_queue_pairs *
	       (i40e_enabled_xdp_vsi(vsi) ? 3 : 2);
A
Anjali Singhai Jain 已提交
7712 7713 7714
	vsi->tx_rings = kzalloc(size, GFP_KERNEL);
	if (!vsi->tx_rings)
		return -ENOMEM;
7715 7716 7717 7718 7719 7720
	next_rings = vsi->tx_rings + vsi->alloc_queue_pairs;
	if (i40e_enabled_xdp_vsi(vsi)) {
		vsi->xdp_rings = next_rings;
		next_rings += vsi->alloc_queue_pairs;
	}
	vsi->rx_rings = next_rings;
A
Anjali Singhai Jain 已提交
7721

7722 7723
	if (alloc_qvectors) {
		/* allocate memory for q_vector pointers */
7724
		size = sizeof(struct i40e_q_vector *) * vsi->num_q_vectors;
7725 7726 7727 7728 7729
		vsi->q_vectors = kzalloc(size, GFP_KERNEL);
		if (!vsi->q_vectors) {
			ret = -ENOMEM;
			goto err_vectors;
		}
A
Anjali Singhai Jain 已提交
7730 7731 7732 7733 7734 7735 7736 7737
	}
	return ret;

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

J
Jesse Brandeburg 已提交
7738 7739 7740 7741 7742 7743 7744 7745 7746 7747 7748 7749 7750 7751 7752 7753 7754 7755 7756 7757 7758 7759 7760 7761 7762
/**
 * 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
7763
	while (i < pf->num_alloc_vsi && pf->vsi[i])
J
Jesse Brandeburg 已提交
7764
		i++;
M
Mitch Williams 已提交
7765
	if (i >= pf->num_alloc_vsi) {
J
Jesse Brandeburg 已提交
7766 7767 7768 7769 7770
		i = 0;
		while (i < pf->next_vsi && pf->vsi[i])
			i++;
	}

M
Mitch Williams 已提交
7771
	if (i < pf->num_alloc_vsi && !pf->vsi[i]) {
J
Jesse Brandeburg 已提交
7772 7773 7774
		vsi_idx = i;             /* Found one! */
	} else {
		ret = -ENODEV;
7775
		goto unlock_pf;  /* out of VSI slots! */
J
Jesse Brandeburg 已提交
7776 7777 7778 7779 7780 7781
	}
	pf->next_vsi = ++i;

	vsi = kzalloc(sizeof(*vsi), GFP_KERNEL);
	if (!vsi) {
		ret = -ENOMEM;
7782
		goto unlock_pf;
J
Jesse Brandeburg 已提交
7783 7784 7785
	}
	vsi->type = type;
	vsi->back = pf;
7786
	set_bit(__I40E_VSI_DOWN, vsi->state);
J
Jesse Brandeburg 已提交
7787 7788
	vsi->flags = 0;
	vsi->idx = vsi_idx;
7789
	vsi->int_rate_limit = 0;
7790 7791
	vsi->rss_table_size = (vsi->type == I40E_VSI_MAIN) ?
				pf->rss_table_size : 64;
J
Jesse Brandeburg 已提交
7792 7793
	vsi->netdev_registered = false;
	vsi->work_limit = I40E_DEFAULT_IRQ_WORK;
7794
	hash_init(vsi->mac_filter_hash);
7795
	vsi->irqs_ready = false;
J
Jesse Brandeburg 已提交
7796

7797 7798 7799 7800
	ret = i40e_set_num_rings_in_vsi(vsi);
	if (ret)
		goto err_rings;

7801
	ret = i40e_vsi_alloc_arrays(vsi, true);
A
Anjali Singhai Jain 已提交
7802
	if (ret)
7803
		goto err_rings;
7804

J
Jesse Brandeburg 已提交
7805 7806 7807
	/* Setup default MSIX irq handler for VSI */
	i40e_vsi_setup_irqhandler(vsi, i40e_msix_clean_rings);

K
Kiran Patil 已提交
7808
	/* Initialize VSI lock */
7809
	spin_lock_init(&vsi->mac_filter_hash_lock);
J
Jesse Brandeburg 已提交
7810 7811
	pf->vsi[vsi_idx] = vsi;
	ret = vsi_idx;
7812 7813
	goto unlock_pf;

7814
err_rings:
7815 7816 7817
	pf->next_vsi = i - 1;
	kfree(vsi);
unlock_pf:
J
Jesse Brandeburg 已提交
7818 7819 7820 7821
	mutex_unlock(&pf->switch_mutex);
	return ret;
}

A
Anjali Singhai Jain 已提交
7822 7823 7824
/**
 * i40e_vsi_free_arrays - Free queue and vector pointer arrays for the VSI
 * @type: VSI pointer
7825
 * @free_qvectors: a bool to specify if q_vectors need to be freed.
A
Anjali Singhai Jain 已提交
7826 7827 7828 7829
 *
 * On error: returns error code (negative)
 * On success: returns 0
 **/
7830
static void i40e_vsi_free_arrays(struct i40e_vsi *vsi, bool free_qvectors)
A
Anjali Singhai Jain 已提交
7831 7832
{
	/* free the ring and vector containers */
7833 7834 7835 7836
	if (free_qvectors) {
		kfree(vsi->q_vectors);
		vsi->q_vectors = NULL;
	}
A
Anjali Singhai Jain 已提交
7837 7838 7839
	kfree(vsi->tx_rings);
	vsi->tx_rings = NULL;
	vsi->rx_rings = NULL;
7840
	vsi->xdp_rings = NULL;
A
Anjali Singhai Jain 已提交
7841 7842
}

7843 7844 7845 7846 7847 7848 7849 7850 7851 7852 7853 7854 7855 7856 7857 7858 7859
/**
 * 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
7860 7861 7862 7863 7864 7865 7866 7867 7868 7869 7870 7871 7872 7873 7874 7875 7876 7877 7878 7879 7880 7881 7882 7883 7884 7885 7886 7887 7888 7889 7890 7891
/**
 * 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;
	}

7892
	/* updates the PF for this cleared vsi */
J
Jesse Brandeburg 已提交
7893 7894 7895
	i40e_put_lump(pf->qp_pile, vsi->base_queue, vsi->idx);
	i40e_put_lump(pf->irq_pile, vsi->base_vector, vsi->idx);

7896
	i40e_vsi_free_arrays(vsi, true);
7897
	i40e_clear_rss_config_user(vsi);
7898

J
Jesse Brandeburg 已提交
7899 7900 7901 7902 7903 7904 7905 7906 7907 7908 7909 7910
	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;
}

7911 7912 7913 7914
/**
 * i40e_vsi_clear_rings - Deallocates the Rx and Tx rings for the provided VSI
 * @vsi: the VSI being cleaned
 **/
7915
static void i40e_vsi_clear_rings(struct i40e_vsi *vsi)
7916 7917 7918
{
	int i;

7919
	if (vsi->tx_rings && vsi->tx_rings[0]) {
N
Neerav Parikh 已提交
7920
		for (i = 0; i < vsi->alloc_queue_pairs; i++) {
7921 7922 7923
			kfree_rcu(vsi->tx_rings[i], rcu);
			vsi->tx_rings[i] = NULL;
			vsi->rx_rings[i] = NULL;
7924 7925
			if (vsi->xdp_rings)
				vsi->xdp_rings[i] = NULL;
7926
		}
7927
	}
7928 7929
}

J
Jesse Brandeburg 已提交
7930 7931 7932 7933 7934 7935
/**
 * 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)
{
7936
	int i, qpv = i40e_enabled_xdp_vsi(vsi) ? 3 : 2;
J
Jesse Brandeburg 已提交
7937
	struct i40e_pf *pf = vsi->back;
7938
	struct i40e_ring *ring;
J
Jesse Brandeburg 已提交
7939 7940

	/* Set basic values in the rings to be used later during open() */
N
Neerav Parikh 已提交
7941
	for (i = 0; i < vsi->alloc_queue_pairs; i++) {
7942
		/* allocate space for both Tx and Rx in one shot */
7943 7944
		ring = kcalloc(qpv, sizeof(struct i40e_ring), GFP_KERNEL);
		if (!ring)
7945
			goto err_out;
J
Jesse Brandeburg 已提交
7946

7947 7948 7949 7950 7951 7952 7953 7954 7955
		ring->queue_index = i;
		ring->reg_idx = vsi->base_queue + i;
		ring->ring_active = false;
		ring->vsi = vsi;
		ring->netdev = vsi->netdev;
		ring->dev = &pf->pdev->dev;
		ring->count = vsi->num_desc;
		ring->size = 0;
		ring->dcb_tc = 0;
7956
		if (vsi->back->hw_features & I40E_HW_WB_ON_ITR_CAPABLE)
7957 7958 7959 7960 7961 7962 7963 7964 7965 7966 7967 7968 7969 7970 7971 7972
			ring->flags = I40E_TXR_FLAGS_WB_ON_ITR;
		ring->tx_itr_setting = pf->tx_itr_default;
		vsi->tx_rings[i] = ring++;

		if (!i40e_enabled_xdp_vsi(vsi))
			goto setup_rx;

		ring->queue_index = vsi->alloc_queue_pairs + i;
		ring->reg_idx = vsi->base_queue + ring->queue_index;
		ring->ring_active = false;
		ring->vsi = vsi;
		ring->netdev = NULL;
		ring->dev = &pf->pdev->dev;
		ring->count = vsi->num_desc;
		ring->size = 0;
		ring->dcb_tc = 0;
7973
		if (vsi->back->hw_features & I40E_HW_WB_ON_ITR_CAPABLE)
7974 7975 7976 7977 7978 7979 7980 7981 7982 7983 7984 7985 7986 7987 7988 7989 7990
			ring->flags = I40E_TXR_FLAGS_WB_ON_ITR;
		set_ring_xdp(ring);
		ring->tx_itr_setting = pf->tx_itr_default;
		vsi->xdp_rings[i] = ring++;

setup_rx:
		ring->queue_index = i;
		ring->reg_idx = vsi->base_queue + i;
		ring->ring_active = false;
		ring->vsi = vsi;
		ring->netdev = vsi->netdev;
		ring->dev = &pf->pdev->dev;
		ring->count = vsi->num_desc;
		ring->size = 0;
		ring->dcb_tc = 0;
		ring->rx_itr_setting = pf->rx_itr_default;
		vsi->rx_rings[i] = ring;
J
Jesse Brandeburg 已提交
7991 7992 7993
	}

	return 0;
7994 7995 7996 7997

err_out:
	i40e_vsi_clear_rings(vsi);
	return -ENOMEM;
J
Jesse Brandeburg 已提交
7998 7999 8000 8001 8002 8003 8004 8005 8006 8007 8008
}

/**
 * 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)
{
8009 8010 8011
	vectors = pci_enable_msix_range(pf->pdev, pf->msix_entries,
					I40E_MIN_MSIX, vectors);
	if (vectors < 0) {
J
Jesse Brandeburg 已提交
8012
		dev_info(&pf->pdev->dev,
8013
			 "MSI-X vector reservation failed: %d\n", vectors);
J
Jesse Brandeburg 已提交
8014 8015 8016 8017 8018 8019 8020 8021 8022 8023 8024 8025
		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.
 *
8026
 * Returns the number of vectors reserved or negative on failure
J
Jesse Brandeburg 已提交
8027 8028 8029 8030
 **/
static int i40e_init_msix(struct i40e_pf *pf)
{
	struct i40e_hw *hw = &pf->hw;
8031
	int cpus, extra_vectors;
S
Shannon Nelson 已提交
8032
	int vectors_left;
J
Jesse Brandeburg 已提交
8033
	int v_budget, i;
8034
	int v_actual;
8035
	int iwarp_requested = 0;
J
Jesse Brandeburg 已提交
8036 8037 8038 8039 8040 8041 8042

	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
8043 8044 8045 8046 8047
	 *	- 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
8048
	 *   - The number of VMDq pairs
8049
	 *   - The CPU count within the NUMA node if iWARP is enabled
J
Jesse Brandeburg 已提交
8050 8051 8052 8053 8054
	 * 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
8055 8056 8057 8058 8059 8060 8061 8062 8063
	vectors_left = hw->func_caps.num_msix_vectors;
	v_budget = 0;

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

8064 8065 8066 8067 8068 8069 8070 8071 8072
	/* 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
8073 8074 8075 8076 8077
	vectors_left -= pf->num_lan_msix;

	/* reserve one vector for sideband flow director */
	if (pf->flags & I40E_FLAG_FD_SB_ENABLED) {
		if (vectors_left) {
8078
			pf->num_fdsb_msix = 1;
S
Shannon Nelson 已提交
8079 8080 8081
			v_budget++;
			vectors_left--;
		} else {
8082
			pf->num_fdsb_msix = 0;
S
Shannon Nelson 已提交
8083 8084
		}
	}
8085

8086 8087
	/* can we reserve enough for iWARP? */
	if (pf->flags & I40E_FLAG_IWARP_ENABLED) {
8088 8089
		iwarp_requested = pf->num_iwarp_msix;

8090 8091 8092 8093 8094 8095 8096 8097
		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
8098 8099 8100 8101 8102
	/* 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);

8103 8104 8105 8106 8107 8108 8109 8110 8111 8112 8113 8114 8115
		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
8116

8117 8118 8119
			v_budget += vmdq_vecs;
			vectors_left -= vmdq_vecs;
		}
S
Shannon Nelson 已提交
8120
	}
J
Jesse Brandeburg 已提交
8121

8122 8123 8124 8125 8126 8127 8128 8129 8130 8131 8132 8133 8134 8135 8136 8137 8138
	/* 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
8139 8140 8141 8142 8143 8144 8145
	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;
8146
	v_actual = i40e_reserve_msix_vectors(pf, v_budget);
8147

8148
	if (v_actual < I40E_MIN_MSIX) {
J
Jesse Brandeburg 已提交
8149 8150 8151
		pf->flags &= ~I40E_FLAG_MSIX_ENABLED;
		kfree(pf->msix_entries);
		pf->msix_entries = NULL;
8152
		pci_disable_msix(pf->pdev);
J
Jesse Brandeburg 已提交
8153 8154
		return -ENODEV;

8155
	} else if (v_actual == I40E_MIN_MSIX) {
J
Jesse Brandeburg 已提交
8156 8157 8158 8159 8160 8161
		/* 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;

8162 8163 8164 8165 8166 8167
	} 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.
		 */
8168 8169
		int vec;

8170 8171
		dev_info(&pf->pdev->dev,
			 "MSI-X vector limit reached, attempting to redistribute vectors\n");
8172
		/* reserve the misc vector */
8173
		vec = v_actual - 1;
8174

J
Jesse Brandeburg 已提交
8175 8176
		/* Scale vector usage down */
		pf->num_vmdq_msix = 1;    /* force VMDqs to only one vector */
8177
		pf->num_vmdq_vsis = 1;
S
Shannon Nelson 已提交
8178
		pf->num_vmdq_qps = 1;
J
Jesse Brandeburg 已提交
8179 8180 8181 8182 8183 8184 8185

		/* partition out the remaining vectors */
		switch (vec) {
		case 2:
			pf->num_lan_msix = 1;
			break;
		case 3:
8186 8187 8188 8189 8190 8191
			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
8192 8193
			break;
		default:
8194 8195 8196 8197 8198 8199 8200 8201 8202
			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);
			}
8203 8204 8205 8206
			if (pf->flags & I40E_FLAG_FD_SB_ENABLED) {
				pf->num_fdsb_msix = 1;
				vec--;
			}
8207 8208 8209
			pf->num_lan_msix = min_t(int,
			       (vec - (pf->num_iwarp_msix + pf->num_vmdq_vsis)),
							      pf->num_lan_msix);
8210
			pf->num_lan_qps = pf->num_lan_msix;
J
Jesse Brandeburg 已提交
8211 8212 8213 8214
			break;
		}
	}

8215 8216 8217 8218 8219
	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;
	}
8220 8221 8222 8223 8224
	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;
	}
8225 8226 8227 8228 8229 8230

	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;
	}
8231 8232 8233 8234 8235 8236 8237
	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);

8238
	return v_actual;
J
Jesse Brandeburg 已提交
8239 8240
}

8241
/**
8242
 * i40e_vsi_alloc_q_vector - Allocate memory for a single interrupt vector
8243 8244
 * @vsi: the VSI being configured
 * @v_idx: index of the vector in the vsi struct
8245
 * @cpu: cpu to be used on affinity_mask
8246 8247 8248
 *
 * We allocate one q_vector.  If allocation fails we return -ENOMEM.
 **/
8249
static int i40e_vsi_alloc_q_vector(struct i40e_vsi *vsi, int v_idx, int cpu)
8250 8251 8252 8253 8254 8255 8256 8257 8258 8259
{
	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;
8260
	cpumask_copy(&q_vector->affinity_mask, cpu_possible_mask);
8261

8262 8263
	if (vsi->netdev)
		netif_napi_add(vsi->netdev, &q_vector->napi,
J
Jesse Brandeburg 已提交
8264
			       i40e_napi_poll, NAPI_POLL_WEIGHT);
8265

8266 8267 8268
	q_vector->rx.latency_range = I40E_LOW_LATENCY;
	q_vector->tx.latency_range = I40E_LOW_LATENCY;

8269 8270 8271 8272 8273 8274
	/* tie q_vector and vsi together */
	vsi->q_vectors[v_idx] = q_vector;

	return 0;
}

J
Jesse Brandeburg 已提交
8275
/**
8276
 * i40e_vsi_alloc_q_vectors - Allocate memory for interrupt vectors
J
Jesse Brandeburg 已提交
8277 8278 8279 8280 8281
 * @vsi: the VSI being configured
 *
 * We allocate one q_vector per queue interrupt.  If allocation fails we
 * return -ENOMEM.
 **/
8282
static int i40e_vsi_alloc_q_vectors(struct i40e_vsi *vsi)
J
Jesse Brandeburg 已提交
8283 8284
{
	struct i40e_pf *pf = vsi->back;
8285
	int err, v_idx, num_q_vectors, current_cpu;
J
Jesse Brandeburg 已提交
8286 8287 8288 8289 8290 8291 8292 8293 8294

	/* 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;

8295 8296
	current_cpu = cpumask_first(cpu_online_mask);

J
Jesse Brandeburg 已提交
8297
	for (v_idx = 0; v_idx < num_q_vectors; v_idx++) {
8298
		err = i40e_vsi_alloc_q_vector(vsi, v_idx, current_cpu);
8299 8300
		if (err)
			goto err_out;
8301 8302 8303
		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
8304 8305 8306
	}

	return 0;
8307 8308 8309 8310 8311 8312

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

	return err;
J
Jesse Brandeburg 已提交
8313 8314 8315 8316 8317 8318
}

/**
 * i40e_init_interrupt_scheme - Determine proper interrupt scheme
 * @pf: board private structure to initialize
 **/
8319
static int i40e_init_interrupt_scheme(struct i40e_pf *pf)
J
Jesse Brandeburg 已提交
8320
{
8321 8322
	int vectors = 0;
	ssize_t size;
J
Jesse Brandeburg 已提交
8323 8324

	if (pf->flags & I40E_FLAG_MSIX_ENABLED) {
8325 8326
		vectors = i40e_init_msix(pf);
		if (vectors < 0) {
J
Jesse Brandeburg 已提交
8327
			pf->flags &= ~(I40E_FLAG_MSIX_ENABLED	|
8328
				       I40E_FLAG_IWARP_ENABLED	|
J
Jesse Brandeburg 已提交
8329
				       I40E_FLAG_RSS_ENABLED	|
8330
				       I40E_FLAG_DCB_CAPABLE	|
8331
				       I40E_FLAG_DCB_ENABLED	|
J
Jesse Brandeburg 已提交
8332 8333 8334 8335
				       I40E_FLAG_SRIOV_ENABLED	|
				       I40E_FLAG_FD_SB_ENABLED	|
				       I40E_FLAG_FD_ATR_ENABLED	|
				       I40E_FLAG_VMDQ_ENABLED);
J
Jesse Brandeburg 已提交
8336 8337 8338 8339 8340 8341 8342 8343

			/* 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
8344
		dev_info(&pf->pdev->dev, "MSI-X not available, trying MSI\n");
8345 8346 8347 8348
		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
8349 8350
			pf->flags &= ~I40E_FLAG_MSI_ENABLED;
		}
8351
		vectors = 1;  /* one MSI or Legacy vector */
J
Jesse Brandeburg 已提交
8352 8353
	}

8354
	if (!(pf->flags & (I40E_FLAG_MSIX_ENABLED | I40E_FLAG_MSI_ENABLED)))
C
Catherine Sullivan 已提交
8355
		dev_info(&pf->pdev->dev, "MSI-X and MSI not available, falling back to Legacy IRQ\n");
8356

8357 8358 8359
	/* set up vector assignment tracking */
	size = sizeof(struct i40e_lump_tracking) + (sizeof(u16) * vectors);
	pf->irq_pile = kzalloc(size, GFP_KERNEL);
8360 8361 8362 8363
	if (!pf->irq_pile) {
		dev_err(&pf->pdev->dev, "error allocating irq_pile memory\n");
		return -ENOMEM;
	}
8364 8365 8366
	pf->irq_pile->num_entries = vectors;
	pf->irq_pile->search_hint = 0;

8367
	/* track first vector for misc interrupts, ignore return */
8368
	(void)i40e_get_lump(pf, pf->irq_pile, 1, I40E_PILE_VALID_BIT - 1);
8369 8370

	return 0;
J
Jesse Brandeburg 已提交
8371 8372
}

8373 8374 8375 8376 8377 8378 8379 8380 8381 8382 8383 8384 8385 8386 8387 8388 8389 8390 8391 8392 8393 8394 8395 8396 8397 8398 8399 8400 8401 8402 8403 8404 8405 8406 8407 8408 8409 8410 8411 8412 8413 8414 8415 8416 8417 8418 8419 8420 8421 8422 8423
#ifdef CONFIG_PM
/**
 * i40e_restore_interrupt_scheme - Restore the interrupt scheme
 * @pf: private board data structure
 *
 * Restore the interrupt scheme that was cleared when we suspended the
 * device. This should be called during resume to re-allocate the q_vectors
 * and reacquire IRQs.
 */
static int i40e_restore_interrupt_scheme(struct i40e_pf *pf)
{
	int err, i;

	/* We cleared the MSI and MSI-X flags when disabling the old interrupt
	 * scheme. We need to re-enabled them here in order to attempt to
	 * re-acquire the MSI or MSI-X vectors
	 */
	pf->flags |= (I40E_FLAG_MSIX_ENABLED | I40E_FLAG_MSI_ENABLED);

	err = i40e_init_interrupt_scheme(pf);
	if (err)
		return err;

	/* Now that we've re-acquired IRQs, we need to remap the vectors and
	 * rings together again.
	 */
	for (i = 0; i < pf->num_alloc_vsi; i++) {
		if (pf->vsi[i]) {
			err = i40e_vsi_alloc_q_vectors(pf->vsi[i]);
			if (err)
				goto err_unwind;
			i40e_vsi_map_rings_to_vectors(pf->vsi[i]);
		}
	}

	err = i40e_setup_misc_vector(pf);
	if (err)
		goto err_unwind;

	return 0;

err_unwind:
	while (i--) {
		if (pf->vsi[i])
			i40e_vsi_free_q_vectors(pf->vsi[i]);
	}

	return err;
}
#endif /* CONFIG_PM */

J
Jesse Brandeburg 已提交
8424 8425 8426 8427 8428 8429 8430 8431 8432 8433 8434 8435 8436
/**
 * 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;

8437 8438
	/* Only request the IRQ once, the first time through. */
	if (!test_and_set_bit(__I40E_MISC_IRQ_REQUESTED, pf->state)) {
J
Jesse Brandeburg 已提交
8439
		err = request_irq(pf->msix_entries[0].vector,
8440
				  i40e_intr, 0, pf->int_name, pf);
J
Jesse Brandeburg 已提交
8441
		if (err) {
8442
			clear_bit(__I40E_MISC_IRQ_REQUESTED, pf->state);
J
Jesse Brandeburg 已提交
8443
			dev_info(&pf->pdev->dev,
C
Catherine Sullivan 已提交
8444
				 "request_irq for %s failed: %d\n",
8445
				 pf->int_name, err);
J
Jesse Brandeburg 已提交
8446 8447 8448 8449
			return -EFAULT;
		}
	}

8450
	i40e_enable_misc_int_causes(pf);
J
Jesse Brandeburg 已提交
8451 8452 8453 8454 8455 8456 8457

	/* 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);

8458
	i40e_irq_dynamic_enable_icr0(pf, true);
J
Jesse Brandeburg 已提交
8459 8460 8461 8462 8463

	return err;
}

/**
8464 8465 8466 8467
 * i40e_config_rss_aq - Prepare for RSS using AQ commands
 * @vsi: vsi structure
 * @seed: RSS hash seed
 **/
8468 8469
static int i40e_config_rss_aq(struct i40e_vsi *vsi, const u8 *seed,
			      u8 *lut, u16 lut_size)
8470 8471 8472
{
	struct i40e_pf *pf = vsi->back;
	struct i40e_hw *hw = &pf->hw;
8473
	int ret = 0;
8474

8475 8476 8477 8478 8479 8480 8481 8482 8483 8484 8485
	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;
		}
8486
	}
8487 8488
	if (lut) {
		bool pf_lut = vsi->type == I40E_VSI_MAIN ? true : false;
8489

8490 8491 8492 8493 8494 8495 8496 8497 8498
		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;
		}
	}
8499 8500 8501
	return ret;
}

8502 8503 8504 8505 8506 8507 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 8541 8542 8543 8544 8545 8546 8547
/**
 * 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;
}

8548 8549 8550 8551 8552 8553 8554 8555 8556 8557 8558
/**
 * 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;

8559
	if (!(pf->hw_features & I40E_HW_RSS_AQ_CAPABLE))
8560 8561
		return 0;

8562 8563 8564 8565 8566 8567
	if (!vsi->rss_size)
		vsi->rss_size = min_t(int, pf->alloc_rss_size,
				      vsi->num_queue_pairs);
	if (!vsi->rss_size)
		return -EINVAL;

8568 8569 8570
	lut = kzalloc(vsi->rss_table_size, GFP_KERNEL);
	if (!lut)
		return -ENOMEM;
8571 8572 8573 8574 8575 8576 8577 8578 8579 8580 8581
	/* 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);
8582 8583 8584 8585 8586 8587
	ret = i40e_config_rss_aq(vsi, seed, lut, vsi->rss_table_size);
	kfree(lut);

	return ret;
}

8588
/**
8589
 * i40e_config_rss_reg - Configure RSS keys and lut by writing registers
8590
 * @vsi: Pointer to vsi structure
8591
 * @seed: RSS hash seed
8592 8593 8594 8595
 * @lut: Lookup table
 * @lut_size: Lookup table size
 *
 * Returns 0 on success, negative on failure
J
Jesse Brandeburg 已提交
8596
 **/
8597 8598
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
8599
{
8600
	struct i40e_pf *pf = vsi->back;
8601
	struct i40e_hw *hw = &pf->hw;
8602
	u16 vf_id = vsi->vf_id;
8603
	u8 i;
J
Jesse Brandeburg 已提交
8604

8605
	/* Fill out hash function seed */
8606 8607 8608
	if (seed) {
		u32 *seed_dw = (u32 *)seed;

8609 8610
		if (vsi->type == I40E_VSI_MAIN) {
			for (i = 0; i <= I40E_PFQF_HKEY_MAX_INDEX; i++)
8611
				wr32(hw, I40E_PFQF_HKEY(i), seed_dw[i]);
8612 8613
		} else if (vsi->type == I40E_VSI_SRIOV) {
			for (i = 0; i <= I40E_VFQF_HKEY1_MAX_INDEX; i++)
8614
				wr32(hw, I40E_VFQF_HKEY1(i, vf_id), seed_dw[i]);
8615 8616 8617
		} else {
			dev_err(&pf->pdev->dev, "Cannot set RSS seed - invalid VSI type\n");
		}
8618 8619 8620 8621 8622
	}

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

8623 8624 8625 8626 8627 8628 8629 8630 8631
		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++)
8632
				wr32(hw, I40E_VFQF_HLUT1(i, vf_id), lut_dw[i]);
8633 8634 8635
		} else {
			dev_err(&pf->pdev->dev, "Cannot set RSS LUT - invalid VSI type\n");
		}
8636 8637 8638 8639 8640 8641
	}
	i40e_flush(hw);

	return 0;
}

8642 8643 8644 8645 8646 8647 8648 8649 8650 8651 8652 8653 8654 8655 8656 8657 8658 8659 8660 8661
/**
 * 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++)
8662
			seed_dw[i] = i40e_read_rx_ctl(hw, I40E_PFQF_HKEY(i));
8663 8664 8665 8666 8667 8668 8669 8670 8671 8672 8673 8674 8675 8676 8677 8678 8679 8680 8681 8682 8683 8684 8685 8686 8687 8688
	}
	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;

8689
	if (pf->hw_features & I40E_HW_RSS_AQ_CAPABLE)
8690 8691 8692 8693 8694 8695 8696 8697 8698 8699 8700 8701 8702 8703 8704 8705
		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)
{
8706 8707
	struct i40e_pf *pf = vsi->back;

8708
	if (pf->hw_features & I40E_HW_RSS_AQ_CAPABLE)
8709 8710 8711
		return i40e_get_rss_aq(vsi, seed, lut, lut_size);
	else
		return i40e_get_rss_reg(vsi, seed, lut, lut_size);
8712 8713
}

8714 8715 8716 8717 8718 8719 8720
/**
 * 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
 */
8721 8722
void i40e_fill_rss_lut(struct i40e_pf *pf, u8 *lut,
		       u16 rss_table_size, u16 rss_size)
8723 8724 8725 8726 8727 8728 8729
{
	u16 i;

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

8730
/**
8731
 * i40e_pf_config_rss - Prepare for RSS if used
8732 8733
 * @pf: board private structure
 **/
8734
static int i40e_pf_config_rss(struct i40e_pf *pf)
8735 8736 8737
{
	struct i40e_vsi *vsi = pf->vsi[pf->lan_vsi];
	u8 seed[I40E_HKEY_ARRAY_SIZE];
8738
	u8 *lut;
8739 8740 8741
	struct i40e_hw *hw = &pf->hw;
	u32 reg_val;
	u64 hena;
8742
	int ret;
8743

J
Jesse Brandeburg 已提交
8744
	/* By default we enable TCP/UDP with IPv4/IPv6 ptypes */
8745 8746
	hena = (u64)i40e_read_rx_ctl(hw, I40E_PFQF_HENA(0)) |
		((u64)i40e_read_rx_ctl(hw, I40E_PFQF_HENA(1)) << 32);
8747 8748
	hena |= i40e_pf_get_default_rss_hena(pf);

8749 8750
	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
8751

8752
	/* Determine the RSS table size based on the hardware capabilities */
8753
	reg_val = i40e_read_rx_ctl(hw, I40E_PFQF_CTL_0);
8754 8755 8756
	reg_val = (pf->rss_table_size == 512) ?
			(reg_val | I40E_PFQF_CTL_0_HASHLUTSIZE_512) :
			(reg_val & ~I40E_PFQF_CTL_0_HASHLUTSIZE_512);
8757
	i40e_write_rx_ctl(hw, I40E_PFQF_CTL_0, reg_val);
8758

8759
	/* Determine the RSS size of the VSI */
8760 8761 8762 8763 8764 8765
	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);
	}
8766 8767
	if (!vsi->rss_size)
		return -EINVAL;
8768

8769 8770 8771 8772
	lut = kzalloc(vsi->rss_table_size, GFP_KERNEL);
	if (!lut)
		return -ENOMEM;

8773 8774 8775 8776 8777
	/* 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);
8778

8779 8780 8781 8782 8783 8784 8785
	/* 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);
8786
	ret = i40e_config_rss(vsi, seed, lut, vsi->rss_table_size);
8787 8788 8789
	kfree(lut);

	return ret;
J
Jesse Brandeburg 已提交
8790 8791
}

8792 8793 8794 8795 8796 8797 8798
/**
 * 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.
8799
 * Note: expects to be called while under rtnl_lock()
8800 8801 8802
 **/
int i40e_reconfig_rss_queues(struct i40e_pf *pf, int queue_count)
{
8803 8804 8805
	struct i40e_vsi *vsi = pf->vsi[pf->lan_vsi];
	int new_rss_size;

8806 8807 8808
	if (!(pf->flags & I40E_FLAG_RSS_ENABLED))
		return 0;

8809
	new_rss_size = min_t(int, queue_count, pf->rss_size_max);
8810

8811
	if (queue_count != vsi->num_queue_pairs) {
8812 8813
		u16 qcount;

8814
		vsi->req_queue_pairs = queue_count;
8815
		i40e_prep_for_reset(pf, true);
8816

8817
		pf->alloc_rss_size = new_rss_size;
8818

8819
		i40e_reset_and_rebuild(pf, true, true);
8820 8821 8822 8823 8824 8825 8826 8827 8828 8829 8830

		/* 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 */
8831 8832
		qcount = vsi->num_queue_pairs / vsi->tc_config.numtc;
		vsi->rss_size = min_t(int, pf->alloc_rss_size, qcount);
8833

8834
		i40e_pf_config_rss(pf);
8835
	}
8836 8837
	dev_info(&pf->pdev->dev, "User requested queue count/HW max RSS count:  %d/%d\n",
		 vsi->req_queue_pairs, pf->rss_size_max);
8838
	return pf->alloc_rss_size;
8839 8840
}

8841
/**
8842
 * i40e_get_partition_bw_setting - Retrieve BW settings for this PF partition
8843 8844
 * @pf: board private structure
 **/
8845
i40e_status i40e_get_partition_bw_setting(struct i40e_pf *pf)
8846 8847 8848 8849 8850 8851 8852 8853 8854 8855
{
	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)
8856
			pf->min_bw = min_bw;
8857
		if (max_valid)
8858
			pf->max_bw = max_bw;
8859 8860 8861 8862 8863 8864
	}

	return status;
}

/**
8865
 * i40e_set_partition_bw_setting - Set BW settings for this PF partition
8866 8867
 * @pf: board private structure
 **/
8868
i40e_status i40e_set_partition_bw_setting(struct i40e_pf *pf)
8869 8870 8871 8872
{
	struct i40e_aqc_configure_partition_bw_data bw_data;
	i40e_status status;

8873
	/* Set the valid bit for this PF */
8874
	bw_data.pf_valid_bits = cpu_to_le16(BIT(pf->hw.pf_id));
8875 8876
	bw_data.max_bw[pf->hw.pf_id] = pf->max_bw & I40E_ALT_BW_VALUE_MASK;
	bw_data.min_bw[pf->hw.pf_id] = pf->min_bw & I40E_ALT_BW_VALUE_MASK;
8877 8878 8879 8880 8881 8882 8883 8884

	/* Set the new bandwidths */
	status = i40e_aq_configure_partition_bw(&pf->hw, &bw_data, NULL);

	return status;
}

/**
8885
 * i40e_commit_partition_bw_setting - Commit BW settings for this PF partition
8886 8887
 * @pf: board private structure
 **/
8888
i40e_status i40e_commit_partition_bw_setting(struct i40e_pf *pf)
8889 8890 8891 8892 8893 8894 8895 8896 8897 8898 8899 8900 8901 8902 8903 8904 8905 8906 8907
{
	/* 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,
8908 8909 8910
			 "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));
8911 8912 8913 8914 8915 8916 8917 8918 8919 8920 8921 8922 8923 8924
		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) {
8925 8926 8927
		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));
8928 8929 8930 8931 8932 8933 8934 8935 8936 8937 8938
		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,
8939 8940 8941
			 "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));
8942 8943 8944 8945 8946 8947 8948 8949 8950 8951 8952 8953 8954 8955 8956 8957 8958
		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,
8959 8960 8961
			 "BW settings NOT SAVED, err %s aq_err %s\n",
			 i40e_stat_str(&pf->hw, ret),
			 i40e_aq_str(&pf->hw, last_aq_status));
8962 8963 8964 8965 8966
bw_commit_out:

	return ret;
}

J
Jesse Brandeburg 已提交
8967 8968 8969 8970 8971 8972 8973 8974 8975 8976 8977 8978 8979 8980 8981 8982
/**
 * 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     |
8983 8984
		    I40E_FLAG_MSIX_ENABLED;

8985 8986 8987 8988
	/* 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;

8989 8990 8991
	/* Depending on PF configurations, it is possible that the RSS
	 * maximum might end up larger than the available queues
	 */
8992
	pf->rss_size_max = BIT(pf->hw.func_caps.rss_table_entry_width);
8993
	pf->alloc_rss_size = 1;
8994
	pf->rss_table_size = pf->hw.func_caps.rss_table_size;
8995 8996
	pf->rss_size_max = min_t(int, pf->rss_size_max,
				 pf->hw.func_caps.num_tx_qp);
J
Jesse Brandeburg 已提交
8997 8998
	if (pf->hw.func_caps.rss) {
		pf->flags |= I40E_FLAG_RSS_ENABLED;
8999 9000
		pf->alloc_rss_size = min_t(int, pf->rss_size_max,
					   num_online_cpus());
J
Jesse Brandeburg 已提交
9001 9002
	}

9003
	/* MFP mode enabled */
9004
	if (pf->hw.func_caps.npar_enable || pf->hw.func_caps.flex10_enable) {
9005 9006
		pf->flags |= I40E_FLAG_MFP_ENABLED;
		dev_info(&pf->pdev->dev, "MFP mode Enabled\n");
9007
		if (i40e_get_partition_bw_setting(pf)) {
9008
			dev_warn(&pf->pdev->dev,
9009 9010
				 "Could not get partition bw settings\n");
		} else {
9011
			dev_info(&pf->pdev->dev,
9012 9013 9014 9015 9016 9017
				 "Partition BW Min = %8.8x, Max = %8.8x\n",
				 pf->min_bw, pf->max_bw);

			/* nudge the Tx scheduler */
			i40e_set_partition_bw_setting(pf);
		}
9018 9019
	}

9020 9021 9022 9023
	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;
9024 9025
		if (pf->flags & I40E_FLAG_MFP_ENABLED &&
		    pf->hw.num_partitions > 1)
9026
			dev_info(&pf->pdev->dev,
A
Anjali Singhai Jain 已提交
9027
				 "Flow Director Sideband mode Disabled in MFP mode\n");
9028 9029
		else
			pf->flags |= I40E_FLAG_FD_SB_ENABLED;
9030 9031 9032 9033
		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
9034 9035
	}

9036
	if (pf->hw.mac.type == I40E_MAC_X722) {
9037 9038 9039 9040 9041 9042 9043 9044 9045 9046 9047
		pf->hw_features |= (I40E_HW_RSS_AQ_CAPABLE |
				    I40E_HW_128_QP_RSS_CAPABLE |
				    I40E_HW_ATR_EVICT_CAPABLE |
				    I40E_HW_WB_ON_ITR_CAPABLE |
				    I40E_HW_MULTIPLE_TCP_UDP_RSS_PCTYPE |
				    I40E_HW_NO_PCI_LINK_CHECK |
				    I40E_HW_USE_SET_LLDP_MIB |
				    I40E_HW_GENEVE_OFFLOAD_CAPABLE |
				    I40E_HW_PTP_L4_CAPABLE |
				    I40E_HW_WOL_MC_MAGIC_PKT_WAKE |
				    I40E_HW_OUTER_UDP_CSUM_CAPABLE);
9048 9049 9050 9051 9052 9053 9054 9055

#define I40E_FDEVICT_PCTYPE_DEFAULT 0xc03
		if (rd32(&pf->hw, I40E_GLQF_FDEVICTENA(1)) !=
		    I40E_FDEVICT_PCTYPE_DEFAULT) {
			dev_warn(&pf->pdev->dev,
				 "FD EVICT PCTYPES are not right, disable FD HW EVICT\n");
			pf->hw_features &= ~I40E_HW_ATR_EVICT_CAPABLE;
		}
9056 9057 9058 9059
	} 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 */
9060
		pf->hw_features |= I40E_HW_GENEVE_OFFLOAD_CAPABLE;
9061 9062 9063
	}

	/* Enable HW ATR eviction if possible */
9064
	if (pf->hw_features & I40E_HW_ATR_EVICT_CAPABLE)
9065 9066
		pf->flags |= I40E_FLAG_HW_ATR_EVICT_ENABLED;

B
Bimmy Pujari 已提交
9067
	if ((pf->hw.mac.type == I40E_MAC_XL710) &&
9068
	    (((pf->hw.aq.fw_maj_ver == 4) && (pf->hw.aq.fw_min_ver < 33)) ||
9069
	    (pf->hw.aq.fw_maj_ver < 4))) {
9070
		pf->hw_features |= I40E_HW_RESTART_AUTONEG;
9071
		/* No DCB support  for FW < v4.33 */
9072
		pf->hw_features |= I40E_HW_NO_DCB_SUPPORT;
9073 9074 9075
	}

	/* Disable FW LLDP if FW < v4.3 */
B
Bimmy Pujari 已提交
9076
	if ((pf->hw.mac.type == I40E_MAC_XL710) &&
9077 9078
	    (((pf->hw.aq.fw_maj_ver == 4) && (pf->hw.aq.fw_min_ver < 3)) ||
	    (pf->hw.aq.fw_maj_ver < 4)))
9079
		pf->hw_features |= I40E_HW_STOP_FW_LLDP;
9080 9081

	/* Use the FW Set LLDP MIB API if FW > v4.40 */
B
Bimmy Pujari 已提交
9082
	if ((pf->hw.mac.type == I40E_MAC_XL710) &&
9083 9084
	    (((pf->hw.aq.fw_maj_ver == 4) && (pf->hw.aq.fw_min_ver >= 40)) ||
	    (pf->hw.aq.fw_maj_ver >= 5)))
9085
		pf->hw_features |= I40E_HW_USE_SET_LLDP_MIB;
9086

9087 9088 9089 9090 9091
	/* Enable PTP L4 if FW > v6.0 */
	if (pf->hw.mac.type == I40E_MAC_XL710 &&
	    pf->hw.aq.fw_maj_ver >= 6)
		pf->hw_features |= I40E_HW_PTP_L4_CAPABLE;

J
Jesse Brandeburg 已提交
9092 9093
	if (pf->hw.func_caps.vmdq) {
		pf->num_vmdq_vsis = I40E_DEFAULT_NUM_VMDQ_VSI;
9094
		pf->flags |= I40E_FLAG_VMDQ_ENABLED;
J
Jesse Brandeburg 已提交
9095
		pf->num_vmdq_qps = i40e_default_queues_per_vmdq(pf);
J
Jesse Brandeburg 已提交
9096 9097
	}

9098 9099 9100 9101 9102 9103
	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
9104
#ifdef CONFIG_PCI_IOV
9105
	if (pf->hw.func_caps.num_vfs && pf->hw.partition_id == 1) {
J
Jesse Brandeburg 已提交
9106 9107 9108 9109 9110 9111 9112 9113 9114 9115 9116
		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 */
	pf->eeprom_version = 0xDEAD;
	pf->lan_veb = I40E_NO_VEB;
	pf->lan_vsi = I40E_NO_VSI;

9117 9118 9119
	/* By default FW has this off for performance reasons */
	pf->flags &= ~I40E_FLAG_VEB_STATS_ENABLED;

J
Jesse Brandeburg 已提交
9120 9121 9122 9123 9124 9125 9126 9127 9128 9129 9130
	/* 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;

9131 9132
	pf->tx_timeout_recovery_level = 1;

J
Jesse Brandeburg 已提交
9133 9134 9135 9136 9137 9138
	mutex_init(&pf->switch_mutex);

sw_init_done:
	return err;
}

9139 9140 9141 9142 9143 9144 9145 9146 9147 9148 9149 9150 9151 9152 9153 9154 9155 9156
/**
 * 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;
9157 9158 9159
		/* enable FD_SB only if there is MSI-X vector */
		if (pf->num_fdsb_msix > 0)
			pf->flags |= I40E_FLAG_FD_SB_ENABLED;
9160 9161 9162 9163 9164 9165
	} 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);
		}
9166 9167
		pf->flags &= ~(I40E_FLAG_FD_SB_ENABLED |
			       I40E_FLAG_FD_SB_AUTO_DISABLED);
9168
		/* reset fd counters */
9169 9170
		pf->fd_add_err = 0;
		pf->fd_atr_cnt = 0;
9171
		/* if ATR was auto disabled it can be re-enabled. */
9172 9173 9174 9175
		if (pf->flags & I40E_FLAG_FD_ATR_AUTO_DISABLED) {
			pf->flags &= ~I40E_FLAG_FD_ATR_AUTO_DISABLED;
			if ((pf->flags & I40E_FLAG_FD_ATR_ENABLED) &&
			    (I40E_DEBUG_FD & pf->hw.debug_mask))
9176 9177
				dev_info(&pf->pdev->dev, "ATR re-enabled.\n");
		}
9178 9179 9180 9181
	}
	return need_reset;
}

9182 9183 9184 9185 9186 9187 9188 9189 9190 9191 9192 9193 9194 9195 9196 9197 9198 9199 9200 9201 9202 9203
/**
 * 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
9204 9205 9206 9207
/**
 * i40e_set_features - set the netdev feature flags
 * @netdev: ptr to the netdev being adjusted
 * @features: the feature set that the stack is suggesting
9208
 * Note: expects to be called while under rtnl_lock()
J
Jesse Brandeburg 已提交
9209 9210 9211 9212 9213 9214
 **/
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;
9215 9216
	struct i40e_pf *pf = vsi->back;
	bool need_reset;
J
Jesse Brandeburg 已提交
9217

9218 9219 9220 9221 9222 9223
	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
9224 9225 9226 9227 9228
	if (features & NETIF_F_HW_VLAN_CTAG_RX)
		i40e_vlan_stripping_enable(vsi);
	else
		i40e_vlan_stripping_disable(vsi);

9229 9230 9231
	need_reset = i40e_set_ntuple(pf, features);

	if (need_reset)
9232
		i40e_do_reset(pf, BIT_ULL(__I40E_PF_RESET_REQUESTED), true);
9233

J
Jesse Brandeburg 已提交
9234 9235 9236
	return 0;
}

9237
/**
9238
 * i40e_get_udp_port_idx - Lookup a possibly offloaded for Rx UDP port
9239 9240 9241 9242 9243
 * @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
 **/
9244
static u8 i40e_get_udp_port_idx(struct i40e_pf *pf, u16 port)
9245 9246 9247 9248
{
	u8 i;

	for (i = 0; i < I40E_MAX_PF_UDP_OFFLOAD_PORTS; i++) {
9249
		if (pf->udp_ports[i].port == port)
9250 9251 9252 9253 9254 9255 9256
			return i;
	}

	return i;
}

/**
9257
 * i40e_udp_tunnel_add - Get notifications about UDP tunnel ports that come up
9258
 * @netdev: This physical port's netdev
9259
 * @ti: Tunnel endpoint information
9260
 **/
9261 9262
static void i40e_udp_tunnel_add(struct net_device *netdev,
				struct udp_tunnel_info *ti)
9263 9264 9265 9266
{
	struct i40e_netdev_priv *np = netdev_priv(netdev);
	struct i40e_vsi *vsi = np->vsi;
	struct i40e_pf *pf = vsi->back;
9267
	u16 port = ntohs(ti->port);
9268 9269 9270
	u8 next_idx;
	u8 idx;

9271
	idx = i40e_get_udp_port_idx(pf, port);
9272 9273 9274

	/* Check if port already exists */
	if (idx < I40E_MAX_PF_UDP_OFFLOAD_PORTS) {
9275
		netdev_info(netdev, "port %d already offloaded\n", port);
9276 9277 9278 9279
		return;
	}

	/* Now check if there is space to add the new port */
9280
	next_idx = i40e_get_udp_port_idx(pf, 0);
9281 9282

	if (next_idx == I40E_MAX_PF_UDP_OFFLOAD_PORTS) {
9283
		netdev_info(netdev, "maximum number of offloaded UDP ports reached, not adding port %d\n",
9284
			    port);
9285 9286 9287
		return;
	}

9288 9289 9290 9291 9292
	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:
9293
		if (!(pf->hw_features & I40E_HW_GENEVE_OFFLOAD_CAPABLE))
9294 9295 9296 9297
			return;
		pf->udp_ports[next_idx].type = I40E_AQC_TUNNEL_TYPE_NGE;
		break;
	default:
9298 9299 9300 9301
		return;
	}

	/* New port: add it and mark its index in the bitmap */
9302
	pf->udp_ports[next_idx].port = port;
9303 9304
	pf->pending_udp_bitmap |= BIT_ULL(next_idx);
	pf->flags |= I40E_FLAG_UDP_FILTER_SYNC;
9305 9306
}

9307
/**
9308
 * i40e_udp_tunnel_del - Get notifications about UDP tunnel ports that go away
9309
 * @netdev: This physical port's netdev
9310
 * @ti: Tunnel endpoint information
9311
 **/
9312 9313
static void i40e_udp_tunnel_del(struct net_device *netdev,
				struct udp_tunnel_info *ti)
9314 9315 9316 9317
{
	struct i40e_netdev_priv *np = netdev_priv(netdev);
	struct i40e_vsi *vsi = np->vsi;
	struct i40e_pf *pf = vsi->back;
9318
	u16 port = ntohs(ti->port);
9319 9320 9321 9322 9323
	u8 idx;

	idx = i40e_get_udp_port_idx(pf, port);

	/* Check if port already exists */
9324 9325
	if (idx >= I40E_MAX_PF_UDP_OFFLOAD_PORTS)
		goto not_found;
9326

9327 9328 9329 9330 9331 9332 9333 9334 9335 9336 9337
	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;
9338
	}
9339 9340 9341 9342

	/* if port exists, set it to 0 (mark for deletion)
	 * and make it pending
	 */
9343
	pf->udp_ports[idx].port = 0;
9344 9345 9346 9347 9348 9349
	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",
9350
		    port);
9351 9352
}

9353
static int i40e_get_phys_port_id(struct net_device *netdev,
9354
				 struct netdev_phys_item_id *ppid)
9355 9356 9357 9358 9359
{
	struct i40e_netdev_priv *np = netdev_priv(netdev);
	struct i40e_pf *pf = np->vsi->back;
	struct i40e_hw *hw = &pf->hw;

9360
	if (!(pf->hw_features & I40E_HW_PORT_ID_VALID))
9361 9362 9363 9364 9365 9366 9367 9368
		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;
}

9369 9370 9371 9372 9373 9374 9375 9376
/**
 * 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
 */
9377 9378
static int i40e_ndo_fdb_add(struct ndmsg *ndm, struct nlattr *tb[],
			    struct net_device *dev,
9379
			    const unsigned char *addr, u16 vid,
9380 9381 9382 9383 9384 9385 9386 9387 9388
			    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;

9389 9390 9391 9392 9393
	if (vid) {
		pr_info("%s: vlans aren't supported yet for dev_uc|mc_add()\n", dev->name);
		return -EINVAL;
	}

9394 9395 9396 9397 9398 9399 9400 9401 9402 9403 9404 9405 9406 9407 9408 9409 9410 9411 9412 9413 9414 9415
	/* 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;
}

9416 9417 9418 9419 9420 9421 9422 9423 9424 9425 9426
/**
 * 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.
9427 9428
 *
 * Note: expects to be called while under rtnl_lock()
9429 9430
 **/
static int i40e_ndo_bridge_setlink(struct net_device *dev,
9431 9432
				   struct nlmsghdr *nlh,
				   u16 flags)
9433 9434 9435 9436 9437 9438 9439 9440 9441 9442 9443 9444 9445 9446 9447 9448 9449 9450 9451 9452 9453 9454 9455 9456 9457 9458 9459 9460 9461 9462 9463 9464 9465 9466 9467 9468 9469 9470 9471 9472 9473 9474 9475 9476 9477 9478
{
	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;
9479 9480 9481 9482 9483
			/* 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;
9484 9485
			i40e_do_reset(pf, BIT_ULL(__I40E_PF_RESET_REQUESTED),
				      true);
9486 9487 9488 9489 9490 9491 9492 9493 9494 9495 9496 9497 9498 9499
			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
9500
 * @nlflags: netlink flags passed in
9501 9502 9503 9504 9505 9506
 *
 * 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,
9507 9508
				   u32 __always_unused filter_mask,
				   int nlflags)
9509 9510 9511 9512 9513 9514 9515 9516 9517 9518 9519 9520 9521 9522 9523 9524 9525 9526 9527 9528
{
	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;

9529
	return ndo_dflt_bridge_getlink(skb, pid, seq, dev, veb->bridge_mode,
9530
				       0, 0, nlflags, filter_mask, NULL);
9531 9532
}

9533 9534 9535
/**
 * i40e_features_check - Validate encapsulated packet conforms to limits
 * @skb: skb buff
J
Jean Sacren 已提交
9536
 * @dev: This physical port's netdev
9537 9538 9539 9540 9541 9542
 * @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)
{
9543 9544 9545 9546 9547 9548 9549 9550 9551 9552 9553 9554 9555 9556 9557 9558 9559 9560 9561 9562 9563 9564 9565 9566 9567 9568 9569 9570 9571 9572 9573 9574 9575 9576 9577 9578 9579 9580 9581 9582 9583 9584
	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
	 */
9585 9586

	return features;
9587 9588
out_err:
	return features & ~(NETIF_F_CSUM_MASK | NETIF_F_GSO_MASK);
9589 9590
}

9591 9592 9593 9594 9595 9596 9597 9598 9599 9600 9601 9602 9603 9604 9605 9606 9607 9608 9609 9610 9611 9612 9613 9614 9615 9616 9617 9618 9619 9620 9621 9622 9623 9624 9625 9626 9627 9628 9629 9630 9631 9632 9633 9634 9635 9636 9637 9638 9639 9640 9641 9642 9643 9644 9645 9646 9647 9648 9649 9650
/**
 * i40e_xdp_setup - add/remove an XDP program
 * @vsi: VSI to changed
 * @prog: XDP program
 **/
static int i40e_xdp_setup(struct i40e_vsi *vsi,
			  struct bpf_prog *prog)
{
	int frame_size = vsi->netdev->mtu + ETH_HLEN + ETH_FCS_LEN + VLAN_HLEN;
	struct i40e_pf *pf = vsi->back;
	struct bpf_prog *old_prog;
	bool need_reset;
	int i;

	/* Don't allow frames that span over multiple buffers */
	if (frame_size > vsi->rx_buf_len)
		return -EINVAL;

	if (!i40e_enabled_xdp_vsi(vsi) && !prog)
		return 0;

	/* When turning XDP on->off/off->on we reset and rebuild the rings. */
	need_reset = (i40e_enabled_xdp_vsi(vsi) != !!prog);

	if (need_reset)
		i40e_prep_for_reset(pf, true);

	old_prog = xchg(&vsi->xdp_prog, prog);

	if (need_reset)
		i40e_reset_and_rebuild(pf, true, true);

	for (i = 0; i < vsi->num_queue_pairs; i++)
		WRITE_ONCE(vsi->rx_rings[i]->xdp_prog, vsi->xdp_prog);

	if (old_prog)
		bpf_prog_put(old_prog);

	return 0;
}

/**
 * i40e_xdp - implements ndo_xdp for i40e
 * @dev: netdevice
 * @xdp: XDP command
 **/
static int i40e_xdp(struct net_device *dev,
		    struct netdev_xdp *xdp)
{
	struct i40e_netdev_priv *np = netdev_priv(dev);
	struct i40e_vsi *vsi = np->vsi;

	if (vsi->type != I40E_VSI_MAIN)
		return -EINVAL;

	switch (xdp->command) {
	case XDP_SETUP_PROG:
		return i40e_xdp_setup(vsi, xdp->prog);
	case XDP_QUERY_PROG:
		xdp->prog_attached = i40e_enabled_xdp_vsi(vsi);
9651
		xdp->prog_id = vsi->xdp_prog ? vsi->xdp_prog->aux->id : 0;
9652 9653 9654 9655 9656 9657
		return 0;
	default:
		return -EINVAL;
	}
}

9658
static const struct net_device_ops i40e_netdev_ops = {
J
Jesse Brandeburg 已提交
9659 9660 9661 9662 9663 9664 9665 9666
	.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
9667
	.ndo_do_ioctl		= i40e_ioctl,
J
Jesse Brandeburg 已提交
9668 9669 9670 9671 9672 9673
	.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
9674
	.ndo_setup_tc		= __i40e_setup_tc,
J
Jesse Brandeburg 已提交
9675 9676 9677
	.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,
9678
	.ndo_set_vf_rate	= i40e_ndo_set_vf_bw,
J
Jesse Brandeburg 已提交
9679
	.ndo_get_vf_config	= i40e_ndo_get_vf_config,
M
Mitch Williams 已提交
9680
	.ndo_set_vf_link_state	= i40e_ndo_set_vf_link_state,
9681
	.ndo_set_vf_spoofchk	= i40e_ndo_set_vf_spoofchk,
9682
	.ndo_set_vf_trust	= i40e_ndo_set_vf_trust,
9683 9684
	.ndo_udp_tunnel_add	= i40e_udp_tunnel_add,
	.ndo_udp_tunnel_del	= i40e_udp_tunnel_del,
9685
	.ndo_get_phys_port_id	= i40e_get_phys_port_id,
9686
	.ndo_fdb_add		= i40e_ndo_fdb_add,
9687
	.ndo_features_check	= i40e_features_check,
9688 9689
	.ndo_bridge_getlink	= i40e_ndo_bridge_getlink,
	.ndo_bridge_setlink	= i40e_ndo_bridge_setlink,
9690
	.ndo_xdp		= i40e_xdp,
J
Jesse Brandeburg 已提交
9691 9692 9693 9694 9695 9696 9697 9698 9699 9700 9701 9702 9703 9704
};

/**
 * 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;
9705
	u8 broadcast[ETH_ALEN];
J
Jesse Brandeburg 已提交
9706 9707
	u8 mac_addr[ETH_ALEN];
	int etherdev_size;
9708 9709
	netdev_features_t hw_enc_features;
	netdev_features_t hw_features;
J
Jesse Brandeburg 已提交
9710 9711

	etherdev_size = sizeof(struct i40e_netdev_priv);
9712
	netdev = alloc_etherdev_mq(etherdev_size, vsi->alloc_queue_pairs);
J
Jesse Brandeburg 已提交
9713 9714 9715 9716 9717 9718 9719
	if (!netdev)
		return -ENOMEM;

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

9720 9721 9722 9723 9724 9725 9726 9727 9728 9729 9730 9731 9732 9733 9734 9735 9736
	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
9737

9738
	if (!(pf->hw_features & I40E_HW_OUTER_UDP_CSUM_CAPABLE))
9739 9740 9741
		netdev->gso_partial_features |= NETIF_F_GSO_UDP_TUNNEL_CSUM;

	netdev->gso_partial_features |= NETIF_F_GSO_GRE_CSUM;
9742

9743 9744
	netdev->hw_enc_features |= hw_enc_features;

9745
	/* record features VLANs can make use of */
9746
	netdev->vlan_features |= hw_enc_features | NETIF_F_TSO_MANGLEID;
J
Jesse Brandeburg 已提交
9747

9748
	if (!(pf->flags & I40E_FLAG_MFP_ENABLED))
9749
		netdev->hw_features |= NETIF_F_NTUPLE;
9750 9751 9752
	hw_features = hw_enc_features		|
		      NETIF_F_HW_VLAN_CTAG_TX	|
		      NETIF_F_HW_VLAN_CTAG_RX;
9753

9754
	netdev->hw_features |= hw_features;
9755

9756
	netdev->features |= hw_features | NETIF_F_HW_VLAN_CTAG_FILTER;
9757
	netdev->hw_enc_features |= NETIF_F_TSO_MANGLEID;
J
Jesse Brandeburg 已提交
9758 9759 9760

	if (vsi->type == I40E_VSI_MAIN) {
		SET_NETDEV_DEV(netdev, &pf->pdev->dev);
9761
		ether_addr_copy(mac_addr, hw->mac.perm_addr);
9762 9763 9764 9765 9766 9767 9768 9769 9770
		/* The following steps are necessary for two reasons. First,
		 * some older NVM configurations load a default MAC-VLAN
		 * filter that will accept any tagged packet, and we want to
		 * replace this with a normal filter. Additionally, it is
		 * possible our MAC address was provided by the platform using
		 * Open Firmware or similar.
		 *
		 * Thus, we need to remove the default filter and install one
		 * specific to the MAC address.
9771 9772
		 */
		i40e_rm_default_mac_filter(vsi, mac_addr);
9773
		spin_lock_bh(&vsi->mac_filter_hash_lock);
9774
		i40e_add_mac_filter(vsi, mac_addr);
9775
		spin_unlock_bh(&vsi->mac_filter_hash_lock);
J
Jesse Brandeburg 已提交
9776
	} else {
9777 9778 9779 9780 9781 9782 9783
		/* Relate the VSI_VMDQ name to the VSI_MAIN name. Note that we
		 * are still limited by IFNAMSIZ, but we're adding 'v%d\0' to
		 * the end, which is 4 bytes long, so force truncation of the
		 * original name by IFNAMSIZ - 4
		 */
		snprintf(netdev->name, IFNAMSIZ, "%.*sv%%d",
			 IFNAMSIZ - 4,
J
Jesse Brandeburg 已提交
9784 9785
			 pf->vsi[pf->lan_vsi]->netdev->name);
		random_ether_addr(mac_addr);
K
Kiran Patil 已提交
9786

9787
		spin_lock_bh(&vsi->mac_filter_hash_lock);
9788
		i40e_add_mac_filter(vsi, mac_addr);
9789
		spin_unlock_bh(&vsi->mac_filter_hash_lock);
J
Jesse Brandeburg 已提交
9790
	}
K
Kiran Patil 已提交
9791

9792 9793 9794 9795 9796 9797 9798 9799 9800 9801 9802 9803 9804 9805 9806
	/* 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);
9807
	i40e_add_mac_filter(vsi, broadcast);
9808 9809
	spin_unlock_bh(&vsi->mac_filter_hash_lock);

9810 9811
	ether_addr_copy(netdev->dev_addr, mac_addr);
	ether_addr_copy(netdev->perm_addr, mac_addr);
9812

J
Jesse Brandeburg 已提交
9813 9814 9815 9816 9817 9818 9819 9820 9821
	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);

9822 9823
	/* MTU range: 68 - 9706 */
	netdev->min_mtu = ETH_MIN_MTU;
9824
	netdev->max_mtu = I40E_MAX_RXBUFFER - I40E_PACKET_HDR_PAD;
9825

J
Jesse Brandeburg 已提交
9826 9827 9828 9829 9830 9831 9832 9833 9834 9835 9836 9837 9838 9839 9840 9841 9842 9843
	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);
}

9844 9845 9846 9847 9848 9849 9850 9851 9852 9853 9854 9855 9856 9857 9858 9859
/**
 * 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];
9860 9861 9862 9863 9864 9865
	if (!veb) {
		dev_info(&pf->pdev->dev,
			 "There is no veb associated with the bridge\n");
		return -ENOENT;
	}

9866
	/* Uplink is a bridge in VEPA mode */
9867
	if (veb->bridge_mode & BRIDGE_MODE_VEPA) {
9868
		return 0;
9869 9870 9871 9872
	} else {
		/* Uplink is a bridge in VEB mode */
		return 1;
	}
9873

9874 9875
	/* VEPA is now default bridge, so return 0 */
	return 0;
9876 9877
}

J
Jesse Brandeburg 已提交
9878 9879 9880 9881 9882 9883 9884 9885 9886 9887 9888 9889 9890
/**
 * 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;
9891 9892 9893
	struct i40e_mac_filter *f;
	struct hlist_node *h;
	int bkt;
K
Kiran Patil 已提交
9894

J
Jesse Brandeburg 已提交
9895 9896 9897 9898 9899 9900 9901 9902 9903 9904 9905 9906 9907 9908 9909 9910 9911 9912
	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,
9913 9914 9915 9916
				 "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
9917 9918
			return -ENOENT;
		}
9919
		vsi->info = ctxt.info;
J
Jesse Brandeburg 已提交
9920 9921 9922 9923 9924 9925 9926
		vsi->info.valid_sections = 0;

		vsi->seid = ctxt.seid;
		vsi->id = ctxt.vsi_number;

		enabled_tc = i40e_pf_get_tc_map(pf);

9927 9928 9929 9930 9931 9932 9933 9934 9935 9936 9937 9938 9939 9940 9941 9942 9943 9944 9945 9946 9947 9948 9949 9950 9951
		/* Source pruning is enabled by default, so the flag is
		 * negative logic - if it's set, we need to fiddle with
		 * the VSI to disable source pruning.
		 */
		if (pf->flags & I40E_FLAG_SOURCE_PRUNING_DISABLED) {
			memset(&ctxt, 0, sizeof(ctxt));
			ctxt.seid = pf->main_vsi_seid;
			ctxt.pf_num = pf->hw.pf_id;
			ctxt.vf_num = 0;
			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_LOCAL_LB);
			ret = i40e_aq_update_vsi_params(hw, &ctxt, NULL);
			if (ret) {
				dev_info(&pf->pdev->dev,
					 "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));
				ret = -ENOENT;
				goto err;
			}
		}

J
Jesse Brandeburg 已提交
9952
		/* MFP mode setup queue map and update VSI */
9953 9954
		if ((pf->flags & I40E_FLAG_MFP_ENABLED) &&
		    !(pf->hw.func_caps.iscsi)) { /* NIC type PF */
J
Jesse Brandeburg 已提交
9955 9956 9957 9958 9959 9960 9961 9962
			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,
9963 9964 9965 9966
					 "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
9967 9968 9969 9970 9971 9972 9973 9974 9975 9976
				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.
9977 9978
			 * For MFP case the iSCSI PF would use this
			 * flow to enable LAN+iSCSI TC.
J
Jesse Brandeburg 已提交
9979 9980 9981
			 */
			ret = i40e_vsi_config_tc(vsi, enabled_tc);
			if (ret) {
9982 9983 9984
				/* Single TC condition is not fatal,
				 * message and continue
				 */
J
Jesse Brandeburg 已提交
9985
				dev_info(&pf->pdev->dev,
9986 9987 9988 9989 9990
					 "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
9991 9992 9993 9994 9995
			}
		}
		break;

	case I40E_VSI_FDIR:
9996 9997 9998
		ctxt.pf_num = hw->pf_id;
		ctxt.vf_num = 0;
		ctxt.uplink_seid = vsi->uplink_seid;
9999
		ctxt.connection_type = I40E_AQ_VSI_CONN_TYPE_NORMAL;
10000
		ctxt.flags = I40E_AQ_VSI_TYPE_PF;
10001 10002
		if ((pf->flags & I40E_FLAG_VEB_MODE_ENABLED) &&
		    (i40e_is_vsi_uplink_mode_veb(vsi))) {
10003
			ctxt.info.valid_sections |=
10004
			     cpu_to_le16(I40E_AQ_VSI_PROP_SWITCH_VALID);
10005
			ctxt.info.switch_id =
10006
			   cpu_to_le16(I40E_AQ_VSI_SW_ID_FLAG_ALLOW_LB);
10007
		}
J
Jesse Brandeburg 已提交
10008 10009 10010 10011 10012 10013 10014
		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;
10015
		ctxt.connection_type = I40E_AQ_VSI_CONN_TYPE_NORMAL;
J
Jesse Brandeburg 已提交
10016 10017 10018 10019 10020
		ctxt.flags = I40E_AQ_VSI_TYPE_VMDQ2;

		/* This VSI is connected to VEB so the switch_id
		 * should be set to zero by default.
		 */
10021 10022 10023 10024 10025 10026
		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
10027 10028 10029 10030 10031 10032 10033 10034 10035

		/* 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;
10036
		ctxt.connection_type = I40E_AQ_VSI_CONN_TYPE_NORMAL;
J
Jesse Brandeburg 已提交
10037 10038 10039 10040 10041
		ctxt.flags = I40E_AQ_VSI_TYPE_VF;

		/* This VSI is connected to VEB so the switch_id
		 * should be set to zero by default.
		 */
10042 10043 10044 10045 10046 10047
		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
10048

10049 10050 10051 10052
		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 |=
10053 10054
				(I40E_AQ_VSI_QUE_OPT_TCP_ENA |
				 I40E_AQ_VSI_QUE_OPT_RSS_LUT_VSI);
10055 10056
		}

J
Jesse Brandeburg 已提交
10057 10058
		ctxt.info.valid_sections |= cpu_to_le16(I40E_AQ_VSI_PROP_VLAN_VALID);
		ctxt.info.port_vlan_flags |= I40E_AQ_VSI_PVLAN_MODE_ALL;
10059 10060 10061 10062 10063 10064 10065
		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
10066 10067 10068 10069
		/* Setup the VSI tx/rx queue map for TC0 only for now */
		i40e_vsi_setup_queue_map(vsi, &ctxt, enabled_tc, true);
		break;

10070 10071 10072 10073
	case I40E_VSI_IWARP:
		/* send down message to iWARP */
		break;

J
Jesse Brandeburg 已提交
10074 10075 10076 10077 10078 10079 10080 10081
	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,
10082 10083 10084 10085
				 "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
10086 10087 10088
			ret = -ENOENT;
			goto err;
		}
10089
		vsi->info = ctxt.info;
J
Jesse Brandeburg 已提交
10090 10091 10092 10093 10094
		vsi->info.valid_sections = 0;
		vsi->seid = ctxt.seid;
		vsi->id = ctxt.vsi_number;
	}

10095
	vsi->active_filters = 0;
10096
	clear_bit(__I40E_VSI_OVERFLOW_PROMISC, vsi->state);
10097
	spin_lock_bh(&vsi->mac_filter_hash_lock);
J
Jesse Brandeburg 已提交
10098
	/* If macvlan filters already exist, force them to get loaded */
10099
	hash_for_each_safe(vsi->mac_filter_hash, bkt, h, f, hlist) {
10100
		f->state = I40E_FILTER_NEW;
J
Jesse Brandeburg 已提交
10101
		f_count++;
K
Kiran Patil 已提交
10102
	}
10103
	spin_unlock_bh(&vsi->mac_filter_hash_lock);
10104

J
Jesse Brandeburg 已提交
10105 10106 10107 10108 10109 10110 10111 10112 10113
	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,
10114 10115 10116
			 "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
10117 10118 10119 10120 10121 10122 10123 10124 10125 10126 10127 10128 10129 10130 10131 10132
		/* 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)
{
10133 10134
	struct i40e_mac_filter *f;
	struct hlist_node *h;
J
Jesse Brandeburg 已提交
10135 10136 10137
	struct i40e_veb *veb = NULL;
	struct i40e_pf *pf;
	u16 uplink_seid;
10138
	int i, n, bkt;
J
Jesse Brandeburg 已提交
10139 10140 10141 10142 10143 10144 10145 10146 10147 10148

	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] &&
10149
	    !test_bit(__I40E_DOWN, pf->state)) {
J
Jesse Brandeburg 已提交
10150 10151 10152 10153 10154 10155 10156 10157 10158 10159 10160 10161 10162
		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 {
10163
			i40e_vsi_close(vsi);
J
Jesse Brandeburg 已提交
10164 10165 10166 10167
		}
		i40e_vsi_disable_irq(vsi);
	}

10168
	spin_lock_bh(&vsi->mac_filter_hash_lock);
10169 10170 10171 10172 10173 10174 10175 10176

	/* 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 */
10177
	hash_for_each_safe(vsi->mac_filter_hash, bkt, h, f, hlist)
10178
		__i40e_del_filter(vsi, f);
10179

10180
	spin_unlock_bh(&vsi->mac_filter_hash_lock);
K
Kiran Patil 已提交
10181

J
Jesse Brandeburg 已提交
10182
	i40e_sync_vsi_filters(vsi);
J
Jesse Brandeburg 已提交
10183 10184 10185

	i40e_vsi_delete(vsi);
	i40e_vsi_free_q_vectors(vsi);
10186 10187 10188 10189
	if (vsi->netdev) {
		free_netdev(vsi->netdev);
		vsi->netdev = NULL;
	}
J
Jesse Brandeburg 已提交
10190 10191 10192 10193 10194 10195 10196 10197 10198 10199 10200
	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
10201
	for (n = 0, i = 0; i < pf->num_alloc_vsi; i++) {
J
Jesse Brandeburg 已提交
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 10234 10235 10236
		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;

10237
	if (vsi->q_vectors[0]) {
J
Jesse Brandeburg 已提交
10238 10239 10240 10241 10242 10243
		dev_info(&pf->pdev->dev, "VSI %d has existing q_vectors\n",
			 vsi->seid);
		return -EEXIST;
	}

	if (vsi->base_vector) {
10244
		dev_info(&pf->pdev->dev, "VSI %d has non-zero base vector %d\n",
J
Jesse Brandeburg 已提交
10245 10246 10247 10248
			 vsi->seid, vsi->base_vector);
		return -EEXIST;
	}

10249
	ret = i40e_vsi_alloc_q_vectors(vsi);
J
Jesse Brandeburg 已提交
10250 10251 10252 10253 10254 10255 10256 10257
	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;
	}

10258 10259 10260 10261 10262
	/* 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;
10263 10264 10265
	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
10266 10267
	if (vsi->base_vector < 0) {
		dev_info(&pf->pdev->dev,
10268 10269
			 "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
10270 10271 10272 10273 10274 10275 10276 10277 10278
		i40e_vsi_free_q_vectors(vsi);
		ret = -ENOENT;
		goto vector_setup_out;
	}

vector_setup_out:
	return ret;
}

10279 10280 10281 10282 10283 10284 10285 10286 10287 10288 10289
/**
 * 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)
{
10290
	u16 alloc_queue_pairs;
J
John Underwood 已提交
10291
	struct i40e_pf *pf;
10292 10293 10294
	u8 enabled_tc;
	int ret;

J
John Underwood 已提交
10295 10296 10297 10298 10299
	if (!vsi)
		return NULL;

	pf = vsi->back;

10300 10301 10302 10303 10304 10305 10306 10307 10308
	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;

10309 10310 10311 10312
	alloc_queue_pairs = vsi->alloc_queue_pairs *
			    (i40e_enabled_xdp_vsi(vsi) ? 2 : 1);

	ret = i40e_get_lump(pf, pf->qp_pile, alloc_queue_pairs, vsi->idx);
10313
	if (ret < 0) {
10314
		dev_info(&pf->pdev->dev,
10315
			 "failed to get tracking for %d queues for VSI %d err %d\n",
10316
			 alloc_queue_pairs, vsi->seid, ret);
10317 10318 10319 10320 10321 10322 10323 10324 10325 10326 10327
		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);
10328 10329
	if (vsi->type == I40E_VSI_MAIN)
		i40e_rm_default_mac_filter(vsi, pf->hw.mac.perm_addr);
10330 10331 10332 10333 10334 10335 10336 10337 10338 10339 10340 10341 10342 10343 10344 10345 10346 10347 10348 10349 10350 10351 10352 10353

	/* 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
10354 10355 10356 10357 10358 10359 10360 10361 10362 10363 10364 10365 10366 10367 10368 10369 10370 10371
/**
 * 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;
10372
	u16 alloc_queue_pairs;
J
Jesse Brandeburg 已提交
10373 10374 10375 10376 10377 10378 10379 10380 10381 10382 10383 10384 10385 10386 10387 10388 10389 10390 10391 10392 10393 10394 10395 10396 10397
	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
10398
		for (i = 0; i < pf->num_alloc_vsi; i++) {
J
Jesse Brandeburg 已提交
10399 10400 10401 10402 10403 10404 10405 10406 10407 10408 10409 10410 10411 10412 10413 10414 10415
			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);
10416 10417 10418
		if (veb) {
			if (vsi->seid != pf->vsi[pf->lan_vsi]->seid) {
				dev_info(&vsi->back->pdev->dev,
S
Shannon Nelson 已提交
10419
					 "New VSI creation error, uplink seid of LAN VSI expected.\n");
10420 10421
				return NULL;
			}
10422 10423 10424 10425 10426 10427 10428 10429
			/* 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;
			}
10430
			i40e_config_bridge_mode(veb);
10431
		}
J
Jesse Brandeburg 已提交
10432 10433 10434 10435 10436 10437 10438 10439 10440 10441 10442 10443 10444 10445 10446 10447 10448 10449
		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];
10450 10451
	if (!vsi)
		goto err_alloc;
J
Jesse Brandeburg 已提交
10452 10453 10454 10455 10456 10457 10458 10459
	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 */
10460 10461 10462 10463
	alloc_queue_pairs = vsi->alloc_queue_pairs *
			    (i40e_enabled_xdp_vsi(vsi) ? 2 : 1);

	ret = i40e_get_lump(pf, pf->qp_pile, alloc_queue_pairs, vsi->idx);
J
Jesse Brandeburg 已提交
10464
	if (ret < 0) {
10465 10466
		dev_info(&pf->pdev->dev,
			 "failed to get tracking for %d queues for VSI %d err=%d\n",
10467
			 alloc_queue_pairs, vsi->seid, ret);
J
Jesse Brandeburg 已提交
10468 10469 10470 10471 10472 10473 10474 10475 10476 10477 10478 10479 10480 10481 10482 10483 10484 10485 10486 10487 10488 10489
		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:
	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
10490 10491 10492 10493
#ifdef CONFIG_I40E_DCB
		/* Setup DCB netlink interface */
		i40e_dcbnl_setup(vsi);
#endif /* CONFIG_I40E_DCB */
J
Jesse Brandeburg 已提交
10494 10495 10496 10497 10498 10499 10500 10501 10502 10503 10504 10505 10506 10507 10508 10509 10510 10511 10512 10513 10514 10515 10516
		/* 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;
	}

10517
	if ((pf->hw_features & I40E_HW_RSS_AQ_CAPABLE) &&
10518 10519 10520
	    (vsi->type == I40E_VSI_VMDQ2)) {
		ret = i40e_vsi_config_rss(vsi);
	}
J
Jesse Brandeburg 已提交
10521 10522 10523 10524 10525 10526 10527 10528 10529 10530 10531 10532 10533 10534 10535 10536 10537 10538 10539 10540 10541 10542 10543 10544 10545 10546 10547 10548 10549 10550 10551 10552 10553 10554 10555 10556 10557 10558 10559
	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,
10560 10561 10562
			 "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
10563 10564 10565 10566 10567 10568 10569
		goto out;
	}

	ret = i40e_aq_query_switch_comp_ets_config(hw, veb->seid,
						   &ets_data, NULL);
	if (ret) {
		dev_info(&pf->pdev->dev,
10570 10571 10572
			 "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
10573 10574 10575 10576 10577 10578
		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;
10579
	veb->enabled_tc = ets_data.tc_valid_bits;
J
Jesse Brandeburg 已提交
10580 10581 10582 10583 10584 10585 10586 10587 10588 10589 10590 10591 10592 10593 10594 10595 10596 10597 10598 10599 10600 10601 10602 10603 10604 10605 10606 10607 10608 10609 10610 10611 10612 10613 10614 10615 10616 10617 10618 10619 10620 10621 10622 10623 10624 10625 10626 10627 10628 10629 10630 10631 10632 10633 10634 10635 10636 10637 10638 10639 10640 10641 10642 10643 10644 10645 10646 10647 10648 10649 10650 10651 10652 10653 10654 10655 10656 10657 10658 10659 10660 10661 10662 10663 10664 10665
	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
10666
	for (i = 0; i < pf->num_alloc_vsi; i++) {
J
Jesse Brandeburg 已提交
10667 10668 10669 10670 10671 10672 10673 10674 10675 10676 10677 10678 10679 10680 10681 10682 10683 10684 10685 10686 10687 10688 10689 10690 10691 10692 10693 10694 10695 10696 10697 10698 10699 10700 10701 10702 10703 10704 10705 10706 10707 10708 10709 10710 10711 10712 10713 10714 10715 10716 10717
		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
10718
	for (i = 0; i < pf->num_alloc_vsi; i++) {
J
Jesse Brandeburg 已提交
10719 10720 10721 10722 10723 10724 10725 10726 10727 10728 10729 10730 10731 10732 10733 10734 10735 10736 10737 10738 10739 10740 10741 10742 10743 10744 10745 10746 10747 10748 10749 10750 10751 10752 10753 10754 10755
		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)
{
10756
	struct i40e_pf *pf = veb->pf;
10757
	bool enable_stats = !!(pf->flags & I40E_FLAG_VEB_STATS_ENABLED);
J
Jesse Brandeburg 已提交
10758 10759
	int ret;

10760
	ret = i40e_aq_add_veb(&pf->hw, veb->uplink_seid, vsi->seid,
10761
			      veb->enabled_tc, false,
10762
			      &veb->seid, enable_stats, NULL);
10763 10764

	/* get a VEB from the hardware */
J
Jesse Brandeburg 已提交
10765
	if (ret) {
10766 10767 10768 10769
		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
10770 10771 10772 10773
		return -EPERM;
	}

	/* get statistics counter */
10774
	ret = i40e_aq_get_veb_parameters(&pf->hw, veb->seid, NULL, NULL,
J
Jesse Brandeburg 已提交
10775 10776
					 &veb->stats_idx, NULL, NULL, NULL);
	if (ret) {
10777 10778 10779 10780
		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
10781 10782 10783 10784
		return -EPERM;
	}
	ret = i40e_veb_get_bw_info(veb);
	if (ret) {
10785 10786 10787 10788 10789
		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
10790 10791 10792 10793 10794 10795 10796 10797 10798 10799 10800 10801 10802 10803 10804 10805 10806 10807 10808 10809 10810 10811 10812 10813 10814 10815 10816 10817 10818 10819 10820 10821 10822 10823 10824 10825 10826 10827 10828 10829 10830 10831 10832 10833
		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
10834
	for (vsi_idx = 0; vsi_idx < pf->num_alloc_vsi; vsi_idx++)
J
Jesse Brandeburg 已提交
10835 10836
		if (pf->vsi[vsi_idx] && pf->vsi[vsi_idx]->seid == vsi_seid)
			break;
M
Mitch Williams 已提交
10837
	if (vsi_idx >= pf->num_alloc_vsi && vsi_seid != 0) {
J
Jesse Brandeburg 已提交
10838 10839 10840 10841 10842 10843 10844 10845 10846 10847 10848 10849 10850 10851 10852 10853 10854 10855 10856 10857 10858 10859 10860 10861 10862 10863 10864 10865 10866 10867 10868 10869 10870 10871
		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
10872 10873
	if (vsi_idx == pf->lan_vsi)
		pf->lan_veb = veb->idx;
J
Jesse Brandeburg 已提交
10874 10875 10876 10877 10878 10879 10880 10881 10882 10883

	return veb;

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

/**
10884
 * i40e_setup_pf_switch_element - set PF vars based on switch type
J
Jesse Brandeburg 已提交
10885 10886 10887 10888 10889 10890 10891 10892 10893 10894 10895 10896 10897 10898 10899 10900 10901 10902 10903 10904 10905 10906 10907 10908 10909 10910 10911 10912 10913 10914 10915 10916 10917 10918 10919 10920 10921 10922 10923 10924 10925 10926 10927 10928 10929 10930 10931 10932 10933 10934 10935 10936 10937 10938 10939 10940 10941 10942 10943 10944 10945 10946 10947 10948 10949 10950 10951 10952 10953 10954 10955 10956 10957 10958 10959 10960 10961 10962 10963 10964 10965 10966 10967 10968 10969 10970 10971 10972 10973 10974 10975 10976 10977 10978 10979 10980 10981 10982 10983 10984 10985 10986 10987 10988 10989 10990 10991 10992 10993 10994
 * @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,
10995 10996 10997 10998
				 "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
10999 11000 11001 11002 11003 11004 11005 11006 11007 11008 11009 11010 11011 11012 11013 11014 11015 11016 11017 11018 11019 11020 11021 11022 11023 11024 11025 11026
			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
11027
 * @reinit: if the Main VSI needs to re-initialized.
J
Jesse Brandeburg 已提交
11028 11029 11030
 *
 * Returns 0 on success, negative value on failure
 **/
11031
static int i40e_setup_pf_switch(struct i40e_pf *pf, bool reinit)
J
Jesse Brandeburg 已提交
11032
{
11033
	u16 flags = 0;
J
Jesse Brandeburg 已提交
11034 11035 11036 11037 11038 11039
	int ret;

	/* find out what's out there already */
	ret = i40e_fetch_switch_configuration(pf, false);
	if (ret) {
		dev_info(&pf->pdev->dev,
11040 11041 11042
			 "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
11043 11044 11045 11046
		return ret;
	}
	i40e_pf_reset_stats(pf);

11047 11048 11049 11050 11051 11052 11053 11054 11055 11056 11057 11058 11059 11060 11061 11062 11063 11064 11065 11066 11067 11068 11069 11070 11071 11072
	/* 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
11073
	/* first time setup */
11074
	if (pf->lan_vsi == I40E_NO_VSI || reinit) {
J
Jesse Brandeburg 已提交
11075 11076 11077 11078 11079 11080 11081 11082 11083 11084
		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;
11085 11086 11087 11088
		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
11089 11090 11091 11092 11093 11094 11095 11096
		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
11097

J
Jesse Brandeburg 已提交
11098 11099 11100 11101 11102 11103
		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]);

11104 11105
	i40e_fdir_sb_setup(pf);

J
Jesse Brandeburg 已提交
11106 11107 11108 11109 11110 11111 11112 11113 11114 11115 11116 11117
	/* 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))
11118
		i40e_pf_config_rss(pf);
J
Jesse Brandeburg 已提交
11119 11120

	/* fill in link information and enable LSE reporting */
11121 11122
	i40e_link_event(pf);

11123
	/* Initialize user-specific link properties */
J
Jesse Brandeburg 已提交
11124 11125
	pf->fc_autoneg_status = ((pf->hw.phy.link_info.an_info &
				  I40E_AQ_AN_COMPLETED) ? true : false);
11126

J
Jacob Keller 已提交
11127 11128
	i40e_ptp_init(pf);

11129 11130 11131
	/* repopulate tunnel port filters */
	i40e_sync_udp_filters(pf);

J
Jesse Brandeburg 已提交
11132 11133 11134 11135 11136 11137 11138 11139 11140 11141
	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;
11142
	int q_max;
J
Jesse Brandeburg 已提交
11143 11144 11145 11146 11147 11148 11149 11150 11151

	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;

11152
	if ((queues_left == 1) ||
11153
	    !(pf->flags & I40E_FLAG_MSIX_ENABLED)) {
J
Jesse Brandeburg 已提交
11154 11155
		/* one qp for PF, no queues for anything else */
		queues_left = 0;
11156
		pf->alloc_rss_size = pf->num_lan_qps = 1;
J
Jesse Brandeburg 已提交
11157 11158

		/* make sure all the fancies are disabled */
J
Jesse Brandeburg 已提交
11159
		pf->flags &= ~(I40E_FLAG_RSS_ENABLED	|
11160
			       I40E_FLAG_IWARP_ENABLED	|
J
Jesse Brandeburg 已提交
11161 11162
			       I40E_FLAG_FD_SB_ENABLED	|
			       I40E_FLAG_FD_ATR_ENABLED	|
11163
			       I40E_FLAG_DCB_CAPABLE	|
11164
			       I40E_FLAG_DCB_ENABLED	|
J
Jesse Brandeburg 已提交
11165 11166
			       I40E_FLAG_SRIOV_ENABLED	|
			       I40E_FLAG_VMDQ_ENABLED);
11167 11168
	} else if (!(pf->flags & (I40E_FLAG_RSS_ENABLED |
				  I40E_FLAG_FD_SB_ENABLED |
11169
				  I40E_FLAG_FD_ATR_ENABLED |
11170
				  I40E_FLAG_DCB_CAPABLE))) {
11171
		/* one qp for PF */
11172
		pf->alloc_rss_size = pf->num_lan_qps = 1;
11173 11174 11175
		queues_left -= pf->num_lan_qps;

		pf->flags &= ~(I40E_FLAG_RSS_ENABLED	|
11176
			       I40E_FLAG_IWARP_ENABLED	|
11177 11178 11179 11180
			       I40E_FLAG_FD_SB_ENABLED	|
			       I40E_FLAG_FD_ATR_ENABLED	|
			       I40E_FLAG_DCB_ENABLED	|
			       I40E_FLAG_VMDQ_ENABLED);
J
Jesse Brandeburg 已提交
11181
	} else {
11182
		/* Not enough queues for all TCs */
11183
		if ((pf->flags & I40E_FLAG_DCB_CAPABLE) &&
11184
		    (queues_left < I40E_MAX_TRAFFIC_CLASS)) {
11185 11186
			pf->flags &= ~(I40E_FLAG_DCB_CAPABLE |
					I40E_FLAG_DCB_ENABLED);
11187 11188
			dev_info(&pf->pdev->dev, "not enough queues for DCB. DCB is disabled.\n");
		}
11189 11190 11191 11192 11193 11194

		/* limit lan qps to the smaller of qps, cpus or msix */
		q_max = max_t(int, pf->rss_size_max, num_online_cpus());
		q_max = min_t(int, q_max, pf->hw.func_caps.num_tx_qp);
		q_max = min_t(int, q_max, pf->hw.func_caps.num_msix_vectors);
		pf->num_lan_qps = q_max;
11195

11196 11197 11198 11199 11200 11201 11202 11203 11204 11205
		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
11206 11207 11208 11209
	}

	if ((pf->flags & I40E_FLAG_SRIOV_ENABLED) &&
	    pf->num_vf_qps && pf->num_req_vfs && queues_left) {
11210 11211
		pf->num_req_vfs = min_t(int, pf->num_req_vfs,
					(queues_left / pf->num_vf_qps));
J
Jesse Brandeburg 已提交
11212 11213 11214 11215 11216 11217 11218 11219 11220 11221
		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);
	}

11222
	pf->queues_left = queues_left;
11223 11224 11225 11226
	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),
11227 11228 11229
		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
11230 11231 11232 11233 11234 11235
}

/**
 * i40e_setup_pf_filter_control - Setup PF static filter control
 * @pf: PF to be setup
 *
11236
 * i40e_setup_pf_filter_control sets up a PF's initial filter control
J
Jesse Brandeburg 已提交
11237 11238 11239 11240 11241 11242 11243 11244 11245 11246 11247 11248 11249
 * 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
11250
	if (pf->flags & (I40E_FLAG_FD_SB_ENABLED | I40E_FLAG_FD_ATR_ENABLED))
J
Jesse Brandeburg 已提交
11251 11252 11253 11254 11255 11256 11257 11258 11259 11260 11261 11262
		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
11263
#define INFO_STRING_LEN 255
11264
#define REMAIN(__x) (INFO_STRING_LEN - (__x))
J
Jesse Brandeburg 已提交
11265 11266 11267
static void i40e_print_features(struct i40e_pf *pf)
{
	struct i40e_hw *hw = &pf->hw;
11268 11269
	char *buf;
	int i;
J
Jesse Brandeburg 已提交
11270

11271 11272
	buf = kmalloc(INFO_STRING_LEN, GFP_KERNEL);
	if (!buf)
J
Jesse Brandeburg 已提交
11273 11274
		return;

11275
	i = snprintf(buf, INFO_STRING_LEN, "Features: PF-id[%d]", hw->pf_id);
J
Jesse Brandeburg 已提交
11276
#ifdef CONFIG_PCI_IOV
11277
	i += snprintf(&buf[i], REMAIN(i), " VFs: %d", pf->num_req_vfs);
J
Jesse Brandeburg 已提交
11278
#endif
11279
	i += snprintf(&buf[i], REMAIN(i), " VSIs: %d QP: %d",
11280
		      pf->hw.func_caps.num_vsis,
11281
		      pf->vsi[pf->lan_vsi]->num_queue_pairs);
J
Jesse Brandeburg 已提交
11282
	if (pf->flags & I40E_FLAG_RSS_ENABLED)
11283
		i += snprintf(&buf[i], REMAIN(i), " RSS");
J
Jesse Brandeburg 已提交
11284
	if (pf->flags & I40E_FLAG_FD_ATR_ENABLED)
11285
		i += snprintf(&buf[i], REMAIN(i), " FD_ATR");
11286
	if (pf->flags & I40E_FLAG_FD_SB_ENABLED) {
11287 11288
		i += snprintf(&buf[i], REMAIN(i), " FD_SB");
		i += snprintf(&buf[i], REMAIN(i), " NTUPLE");
11289
	}
11290
	if (pf->flags & I40E_FLAG_DCB_CAPABLE)
11291 11292
		i += snprintf(&buf[i], REMAIN(i), " DCB");
	i += snprintf(&buf[i], REMAIN(i), " VxLAN");
11293
	i += snprintf(&buf[i], REMAIN(i), " Geneve");
J
Jesse Brandeburg 已提交
11294
	if (pf->flags & I40E_FLAG_PTP)
11295
		i += snprintf(&buf[i], REMAIN(i), " PTP");
11296
	if (pf->flags & I40E_FLAG_VEB_MODE_ENABLED)
11297
		i += snprintf(&buf[i], REMAIN(i), " VEB");
11298
	else
11299
		i += snprintf(&buf[i], REMAIN(i), " VEPA");
J
Jesse Brandeburg 已提交
11300

11301 11302
	dev_info(&pf->pdev->dev, "%s\n", buf);
	kfree(buf);
11303
	WARN_ON(i > INFO_STRING_LEN);
J
Jesse Brandeburg 已提交
11304 11305
}

11306 11307 11308 11309 11310
/**
 * i40e_get_platform_mac_addr - get platform-specific MAC address
 * @pdev: PCI device information struct
 * @pf: board private structure
 *
11311 11312 11313 11314
 * Look up the MAC address for the device. First we'll try
 * eth_platform_get_mac_address, which will check Open Firmware, or arch
 * specific fallback. Otherwise, we'll default to the stored value in
 * firmware.
11315 11316 11317
 **/
static void i40e_get_platform_mac_addr(struct pci_dev *pdev, struct i40e_pf *pf)
{
11318 11319
	if (eth_platform_get_mac_address(&pdev->dev, pf->hw.mac.addr))
		i40e_get_mac_addr(&pf->hw, pf->hw.mac.addr);
11320 11321
}

J
Jesse Brandeburg 已提交
11322 11323 11324 11325 11326
/**
 * i40e_probe - Device initialization routine
 * @pdev: PCI device information struct
 * @ent: entry in i40e_pci_tbl
 *
11327 11328
 * 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
11329 11330 11331 11332 11333 11334
 * 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)
{
11335
	struct i40e_aq_get_phy_abilities_resp abilities;
J
Jesse Brandeburg 已提交
11336 11337
	struct i40e_pf *pf;
	struct i40e_hw *hw;
11338
	static u16 pfs_found;
11339
	u16 wol_nvm_bits;
11340
	u16 link_status;
11341
	int err;
11342
	u32 val;
11343
	u32 i;
11344
	u8 set_fc_aq_fail;
J
Jesse Brandeburg 已提交
11345 11346 11347 11348 11349 11350

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

	/* set up for high or low dma */
11351 11352
	err = dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(64));
	if (err) {
11353 11354 11355 11356 11357 11358
		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
11359 11360 11361
	}

	/* set up pci connections */
11362
	err = pci_request_mem_regions(pdev, i40e_driver_name);
J
Jesse Brandeburg 已提交
11363 11364 11365 11366 11367 11368 11369 11370 11371 11372 11373 11374 11375 11376 11377 11378 11379 11380 11381 11382 11383
	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;
11384
	set_bit(__I40E_DOWN, pf->state);
J
Jesse Brandeburg 已提交
11385 11386 11387

	hw = &pf->hw;
	hw->back = pf;
A
Anjali Singhai 已提交
11388

S
Shannon Nelson 已提交
11389 11390
	pf->ioremap_len = min_t(int, pci_resource_len(pdev, 0),
				I40E_MAX_CSR_SPACE);
A
Anjali Singhai 已提交
11391

S
Shannon Nelson 已提交
11392
	hw->hw_addr = ioremap(pci_resource_start(pdev, 0), pf->ioremap_len);
J
Jesse Brandeburg 已提交
11393 11394 11395 11396
	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
11397
			 pf->ioremap_len, err);
J
Jesse Brandeburg 已提交
11398 11399 11400 11401 11402 11403 11404 11405 11406
		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);
11407
	hw->bus.bus_id = pdev->bus->number;
11408
	pf->instance = pfs_found;
J
Jesse Brandeburg 已提交
11409

11410 11411 11412 11413 11414 11415 11416
	/* Select something other than the 802.1ad ethertype for the
	 * switch to use internally and drop on ingress.
	 */
	hw->switch_tag = 0xffff;
	hw->first_tag = ETH_P_8021AD;
	hw->second_tag = ETH_P_8021Q;

11417 11418 11419
	INIT_LIST_HEAD(&pf->l3_flex_pit_list);
	INIT_LIST_HEAD(&pf->l4_flex_pit_list);

11420 11421 11422 11423 11424 11425
	/* 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);

11426 11427 11428 11429 11430 11431
	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
11432

11433 11434 11435 11436 11437 11438 11439 11440 11441 11442 11443
	/* 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
11444
	/* Reset here to make sure all is clean and to define PF 'n' */
11445
	i40e_clear_hw(hw);
J
Jesse Brandeburg 已提交
11446 11447 11448 11449 11450 11451 11452 11453 11454 11455 11456 11457
	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;
11458

11459
	snprintf(pf->int_name, sizeof(pf->int_name) - 1,
11460 11461
		 "%s-%s:misc",
		 dev_driver_string(&pf->pdev->dev), dev_name(&pdev->dev));
J
Jesse Brandeburg 已提交
11462 11463 11464

	err = i40e_init_shared_code(hw);
	if (err) {
11465 11466
		dev_warn(&pdev->dev, "unidentified MAC or BLANK NVM: %d\n",
			 err);
J
Jesse Brandeburg 已提交
11467 11468 11469
		goto err_pf_reset;
	}

11470 11471 11472
	/* set up a default setting for link flow control */
	pf->hw.fc.requested_mode = I40E_FC_NONE;

J
Jesse Brandeburg 已提交
11473
	err = i40e_init_adminq(hw);
C
Carolyn Wyborny 已提交
11474 11475 11476 11477 11478 11479 11480 11481 11482 11483
	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;
	}
11484
	i40e_get_oem_version(hw);
11485

11486 11487 11488 11489 11490
	/* 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));
11491

11492
	if (hw->aq.api_maj_ver == I40E_FW_API_VERSION_MAJOR &&
11493
	    hw->aq.api_min_ver > I40E_FW_MINOR_VERSION(hw))
11494
		dev_info(&pdev->dev,
11495
			 "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");
11496
	else if (hw->aq.api_maj_ver == 1 && hw->aq.api_min_ver < 4)
11497
		dev_info(&pdev->dev,
11498
			 "The driver for the device detected an older version of the NVM image than expected. Please update the NVM image.\n");
11499

11500 11501
	i40e_verify_eeprom(pf);

11502 11503 11504 11505
	/* 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");

11506
	i40e_clear_pxe_mode(hw);
J
Jesse Brandeburg 已提交
11507 11508 11509 11510 11511 11512 11513 11514 11515 11516 11517
	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,
11518
				hw->func_caps.num_rx_qp, 0, 0);
J
Jesse Brandeburg 已提交
11519 11520 11521 11522 11523 11524 11525 11526 11527 11528 11529 11530
	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;
	}

11531 11532 11533 11534
	/* 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
	 */
11535
	if (pf->hw_features & I40E_HW_STOP_FW_LLDP) {
11536 11537 11538 11539
		dev_info(&pdev->dev, "Stopping firmware LLDP agent.\n");
		i40e_aq_stop_lldp(hw, true, NULL);
	}

11540 11541
	/* allow a platform config to override the HW addr */
	i40e_get_platform_mac_addr(pdev, pf);
11542

11543
	if (!is_valid_ether_addr(hw->mac.addr)) {
J
Jesse Brandeburg 已提交
11544 11545 11546 11547 11548
		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);
11549
	ether_addr_copy(hw->mac.perm_addr, hw->mac.addr);
11550 11551
	i40e_get_port_mac_addr(hw, hw->mac.port_addr);
	if (is_valid_ether_addr(hw->mac.port_addr))
11552
		pf->hw_features |= I40E_HW_PORT_ID_VALID;
J
Jesse Brandeburg 已提交
11553 11554 11555

	pci_set_drvdata(pdev, pf);
	pci_save_state(pdev);
N
Neerav Parikh 已提交
11556 11557 11558
#ifdef CONFIG_I40E_DCB
	err = i40e_init_pf_dcb(pf);
	if (err) {
11559
		dev_info(&pdev->dev, "DCB init failed %d, disabled\n", err);
11560
		pf->flags &= ~(I40E_FLAG_DCB_CAPABLE | I40E_FLAG_DCB_ENABLED);
11561
		/* Continue without DCB enabled */
N
Neerav Parikh 已提交
11562 11563
	}
#endif /* CONFIG_I40E_DCB */
J
Jesse Brandeburg 已提交
11564 11565 11566 11567 11568 11569

	/* 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);
11570
	clear_bit(__I40E_SERVICE_SCHED, pf->state);
J
Jesse Brandeburg 已提交
11571

11572 11573
	/* 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
11574
	if (BIT (hw->port) & wol_nvm_bits || hw->partition_id != 1)
11575 11576 11577
		pf->wol_en = false;
	else
		pf->wol_en = true;
S
Shannon Nelson 已提交
11578 11579
	device_set_wakeup_enable(&pf->pdev->dev, pf->wol_en);

J
Jesse Brandeburg 已提交
11580 11581
	/* set up the main switch operations */
	i40e_determine_queue_usage(pf);
11582 11583 11584
	err = i40e_init_interrupt_scheme(pf);
	if (err)
		goto err_switch_setup;
J
Jesse Brandeburg 已提交
11585

M
Mitch Williams 已提交
11586 11587 11588 11589
	/* 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
11590
	 */
M
Mitch Williams 已提交
11591 11592 11593 11594 11595 11596
	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. */
11597 11598
	pf->vsi = kcalloc(pf->num_alloc_vsi, sizeof(struct i40e_vsi *),
			  GFP_KERNEL);
11599 11600
	if (!pf->vsi) {
		err = -ENOMEM;
J
Jesse Brandeburg 已提交
11601
		goto err_switch_setup;
11602
	}
J
Jesse Brandeburg 已提交
11603

11604 11605 11606 11607
#ifdef CONFIG_PCI_IOV
	/* prep for VF support */
	if ((pf->flags & I40E_FLAG_SRIOV_ENABLED) &&
	    (pf->flags & I40E_FLAG_MSIX_ENABLED) &&
11608
	    !test_bit(__I40E_BAD_EEPROM, pf->state)) {
11609 11610 11611 11612
		if (pci_num_vf(pdev))
			pf->flags |= I40E_FLAG_VEB_MODE_ENABLED;
	}
#endif
11613
	err = i40e_setup_pf_switch(pf, false);
J
Jesse Brandeburg 已提交
11614 11615 11616 11617
	if (err) {
		dev_info(&pdev->dev, "setup_pf_switch failed: %d\n", err);
		goto err_vsis;
	}
11618 11619 11620 11621 11622 11623 11624 11625 11626 11627 11628 11629 11630 11631 11632 11633 11634 11635 11636

	/* 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));

11637
	/* if FDIR VSI was set up, start it now */
M
Mitch Williams 已提交
11638
	for (i = 0; i < pf->num_alloc_vsi; i++) {
11639 11640 11641 11642 11643
		if (pf->vsi[i] && pf->vsi[i]->type == I40E_VSI_FDIR) {
			i40e_vsi_open(pf->vsi[i]);
			break;
		}
	}
J
Jesse Brandeburg 已提交
11644

S
Shannon Nelson 已提交
11645 11646
	/* The driver only wants link up/down and module qualification
	 * reports from firmware.  Note the negative logic.
J
Jesse Brandeburg 已提交
11647 11648
	 */
	err = i40e_aq_set_phy_int_mask(&pf->hw,
S
Shannon Nelson 已提交
11649
				       ~(I40E_AQ_EVENT_LINK_UPDOWN |
11650
					 I40E_AQ_EVENT_MEDIA_NA |
S
Shannon Nelson 已提交
11651
					 I40E_AQ_EVENT_MODULE_QUAL_FAIL), NULL);
J
Jesse Brandeburg 已提交
11652
	if (err)
11653 11654 11655
		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
11656

11657 11658 11659 11660 11661 11662 11663 11664 11665 11666 11667
	/* 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);
	}

11668
	if (pf->hw_features & I40E_HW_RESTART_AUTONEG) {
11669 11670 11671
		msleep(75);
		err = i40e_aq_set_link_restart_an(&pf->hw, true, NULL);
		if (err)
11672 11673 11674 11675
			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));
11676
	}
J
Jesse Brandeburg 已提交
11677 11678 11679 11680
	/* 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.
	 */
11681
	clear_bit(__I40E_DOWN, pf->state);
J
Jesse Brandeburg 已提交
11682 11683 11684 11685 11686 11687 11688 11689 11690 11691 11692 11693 11694 11695 11696

	/* 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;
		}
	}

11697
#ifdef CONFIG_PCI_IOV
J
Jesse Brandeburg 已提交
11698 11699
	/* prep for VF support */
	if ((pf->flags & I40E_FLAG_SRIOV_ENABLED) &&
11700
	    (pf->flags & I40E_FLAG_MSIX_ENABLED) &&
11701
	    !test_bit(__I40E_BAD_EEPROM, pf->state)) {
J
Jesse Brandeburg 已提交
11702 11703 11704 11705 11706
		/* 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
11707 11708 11709 11710 11711 11712 11713 11714 11715 11716

		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
11717
	}
11718
#endif /* CONFIG_PCI_IOV */
J
Jesse Brandeburg 已提交
11719

11720 11721 11722 11723 11724 11725 11726 11727 11728 11729 11730
	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;
		}
	}
11731

J
Jesse Brandeburg 已提交
11732 11733 11734
	i40e_dbg_pf_init(pf);

	/* tell the firmware that we're starting */
J
Jesse Brandeburg 已提交
11735
	i40e_send_version(pf);
J
Jesse Brandeburg 已提交
11736 11737 11738 11739 11740

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

11741
	/* add this PF to client device list and launch a client service task */
11742 11743 11744 11745 11746 11747
	if (pf->flags & I40E_FLAG_IWARP_ENABLED) {
		err = i40e_lan_add_device(pf);
		if (err)
			dev_info(&pdev->dev, "Failed to add PF to client API service list: %d\n",
				 err);
	}
11748

11749 11750 11751 11752 11753 11754
#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.
	 */
11755
	if (!(pf->hw_features & I40E_HW_NO_PCI_LINK_CHECK)) {
11756 11757 11758 11759 11760 11761 11762 11763 11764 11765 11766 11767 11768 11769 11770 11771 11772 11773 11774 11775 11776 11777 11778 11779 11780 11781 11782 11783 11784 11785 11786 11787 11788 11789 11790 11791 11792 11793 11794 11795 11796 11797
		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");
		}
11798 11799
	}

11800 11801 11802
	/* get the requested speeds from the fw */
	err = i40e_aq_get_phy_capabilities(hw, false, false, &abilities, NULL);
	if (err)
11803 11804 11805
		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));
11806 11807
	pf->hw.phy.link_info.requested_speeds = abilities.link_speed;

11808 11809 11810 11811 11812 11813 11814
	/* 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));

11815 11816 11817 11818 11819 11820 11821 11822 11823
	/* 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);

11824
	if ((pf->hw.device_id == I40E_DEV_ID_10G_BASE_T) ||
H
Henry Tieman 已提交
11825
		(pf->hw.device_id == I40E_DEV_ID_10G_BASE_T4))
11826
		pf->hw_features |= I40E_HW_PHY_CONTROLS_LEDS;
11827
	if (pf->hw.device_id == I40E_DEV_ID_SFP_I_X722)
11828
		pf->hw_features |= I40E_HW_HAVE_CRT_RETIMER;
J
Jesse Brandeburg 已提交
11829 11830 11831
	/* print a string summarizing features */
	i40e_print_features(pf);

J
Jesse Brandeburg 已提交
11832 11833 11834 11835
	return 0;

	/* Unwind what we've done if something failed in the setup */
err_vsis:
11836
	set_bit(__I40E_DOWN, pf->state);
J
Jesse Brandeburg 已提交
11837 11838
	i40e_clear_interrupt_scheme(pf);
	kfree(pf->vsi);
11839 11840
err_switch_setup:
	i40e_reset_interrupt_capability(pf);
J
Jesse Brandeburg 已提交
11841 11842 11843 11844 11845 11846 11847 11848 11849 11850 11851 11852 11853 11854
	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);
11855
	pci_release_mem_regions(pdev);
J
Jesse Brandeburg 已提交
11856 11857 11858 11859 11860 11861 11862 11863 11864 11865 11866 11867 11868 11869 11870 11871 11872 11873
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);
11874
	struct i40e_hw *hw = &pf->hw;
J
Jesse Brandeburg 已提交
11875 11876 11877 11878 11879
	i40e_status ret_code;
	int i;

	i40e_dbg_pf_exit(pf);

J
Jacob Keller 已提交
11880 11881
	i40e_ptp_stop(pf);

11882
	/* Disable RSS in hw */
11883 11884
	i40e_write_rx_ctl(hw, I40E_PFQF_HENA(0), 0);
	i40e_write_rx_ctl(hw, I40E_PFQF_HENA(1), 0);
11885

J
Jesse Brandeburg 已提交
11886
	/* no more scheduling of any task */
11887
	set_bit(__I40E_SUSPENDED, pf->state);
11888
	set_bit(__I40E_DOWN, pf->state);
11889 11890 11891 11892
	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
11893

11894 11895 11896 11897 11898
	/* 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);

11899 11900 11901 11902 11903
	if (pf->flags & I40E_FLAG_SRIOV_ENABLED) {
		i40e_free_vfs(pf);
		pf->flags &= ~I40E_FLAG_SRIOV_ENABLED;
	}

J
Jesse Brandeburg 已提交
11904 11905 11906 11907 11908 11909 11910 11911 11912 11913 11914 11915 11916 11917 11918 11919 11920 11921 11922 11923
	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]);

11924
	/* remove attached clients */
11925 11926 11927 11928 11929
	if (pf->flags & I40E_FLAG_IWARP_ENABLED) {
		ret_code = i40e_lan_del_device(pf);
		if (ret_code)
			dev_warn(&pdev->dev, "Failed to delete client device: %d\n",
				 ret_code);
11930 11931
	}

J
Jesse Brandeburg 已提交
11932
	/* shutdown and destroy the HMC */
11933 11934
	if (hw->hmc.hmc_obj) {
		ret_code = i40e_shutdown_lan_hmc(hw);
11935 11936 11937 11938 11939
		if (ret_code)
			dev_warn(&pdev->dev,
				 "Failed to destroy the HMC resources: %d\n",
				 ret_code);
	}
J
Jesse Brandeburg 已提交
11940 11941

	/* shutdown the adminq */
H
Henry Tieman 已提交
11942
	i40e_shutdown_adminq(hw);
J
Jesse Brandeburg 已提交
11943

11944 11945 11946 11947
	/* destroy the locks only once, here */
	mutex_destroy(&hw->aq.arq_mutex);
	mutex_destroy(&hw->aq.asq_mutex);

J
Jesse Brandeburg 已提交
11948 11949
	/* Clear all dynamic memory lists of rings, q_vectors, and VSIs */
	i40e_clear_interrupt_scheme(pf);
M
Mitch Williams 已提交
11950
	for (i = 0; i < pf->num_alloc_vsi; i++) {
J
Jesse Brandeburg 已提交
11951 11952 11953 11954 11955 11956 11957 11958 11959 11960 11961 11962 11963 11964 11965
		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);

11966
	iounmap(hw->hw_addr);
J
Jesse Brandeburg 已提交
11967
	kfree(pf);
11968
	pci_release_mem_regions(pdev);
J
Jesse Brandeburg 已提交
11969 11970 11971 11972 11973 11974 11975 11976 11977 11978 11979 11980 11981 11982 11983 11984 11985 11986 11987 11988

	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);

11989 11990 11991 11992 11993 11994
	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
11995
	/* shutdown all operations */
11996 11997
	if (!test_bit(__I40E_SUSPENDED, pf->state))
		i40e_prep_for_reset(pf, false);
J
Jesse Brandeburg 已提交
11998 11999 12000 12001 12002 12003 12004 12005 12006 12007 12008 12009 12010 12011 12012 12013 12014 12015 12016 12017 12018

	/* 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
12019
	dev_dbg(&pdev->dev, "%s\n", __func__);
J
Jesse Brandeburg 已提交
12020 12021 12022 12023 12024 12025 12026 12027 12028 12029 12030 12031 12032 12033 12034 12035 12036 12037 12038 12039 12040 12041 12042 12043 12044 12045 12046 12047
	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;
}

12048 12049 12050 12051 12052 12053 12054 12055 12056 12057 12058 12059 12060 12061 12062 12063 12064 12065 12066 12067 12068 12069
/**
 * i40e_pci_error_reset_prepare - prepare device driver for pci reset
 * @pdev: PCI device information struct
 */
static void i40e_pci_error_reset_prepare(struct pci_dev *pdev)
{
	struct i40e_pf *pf = pci_get_drvdata(pdev);

	i40e_prep_for_reset(pf, false);
}

/**
 * i40e_pci_error_reset_done - pci reset done, device driver reset can begin
 * @pdev: PCI device information struct
 */
static void i40e_pci_error_reset_done(struct pci_dev *pdev)
{
	struct i40e_pf *pf = pci_get_drvdata(pdev);

	i40e_reset_and_rebuild(pf, false, false);
}

J
Jesse Brandeburg 已提交
12070 12071 12072 12073 12074 12075 12076 12077 12078 12079 12080
/**
 * 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
12081
	dev_dbg(&pdev->dev, "%s\n", __func__);
12082
	if (test_bit(__I40E_SUSPENDED, pf->state))
12083 12084
		return;

12085
	i40e_handle_reset_warning(pf, false);
12086 12087
}

12088 12089 12090 12091 12092 12093 12094 12095 12096 12097 12098 12099 12100 12101 12102 12103 12104 12105 12106 12107 12108 12109 12110 12111 12112 12113 12114 12115 12116 12117 12118 12119 12120 12121 12122 12123 12124 12125 12126 12127 12128 12129 12130 12131 12132 12133 12134
/**
 * 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");
}

12135 12136 12137 12138 12139 12140 12141
/**
 * 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
12142
	struct i40e_hw *hw = &pf->hw;
12143

12144
	set_bit(__I40E_SUSPENDED, pf->state);
12145
	set_bit(__I40E_DOWN, pf->state);
12146
	rtnl_lock();
12147
	i40e_prep_for_reset(pf, true);
12148 12149
	rtnl_unlock();

S
Shannon Nelson 已提交
12150 12151 12152
	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));

12153 12154 12155 12156
	del_timer_sync(&pf->service_timer);
	cancel_work_sync(&pf->service_task);
	i40e_fdir_teardown(pf);

12157 12158 12159 12160 12161
	/* 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);

12162
	if (pf->wol_en && (pf->hw_features & I40E_HW_WOL_MC_MAGIC_PKT_WAKE))
12163 12164
		i40e_enable_mc_magic_wake(pf);

12165
	i40e_prep_for_reset(pf, false);
12166 12167 12168 12169 12170 12171

	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));

12172 12173
	i40e_clear_interrupt_scheme(pf);

12174
	if (system_state == SYSTEM_POWER_OFF) {
S
Shannon Nelson 已提交
12175
		pci_wake_from_d3(pdev, pf->wol_en);
12176 12177 12178 12179 12180 12181
		pci_set_power_state(pdev, PCI_D3hot);
	}
}

#ifdef CONFIG_PM
/**
12182 12183
 * i40e_suspend - PM callback for moving to D3
 * @dev: generic device information structure
12184
 **/
12185
static int i40e_suspend(struct device *dev)
12186
{
12187
	struct pci_dev *pdev = to_pci_dev(dev);
12188
	struct i40e_pf *pf = pci_get_drvdata(pdev);
S
Shannon Nelson 已提交
12189
	struct i40e_hw *hw = &pf->hw;
12190

12191 12192 12193 12194
	/* If we're already suspended, then there is nothing to do */
	if (test_and_set_bit(__I40E_SUSPENDED, pf->state))
		return 0;

12195
	set_bit(__I40E_DOWN, pf->state);
12196

12197 12198 12199 12200
	/* Ensure service task will not be running */
	del_timer_sync(&pf->service_timer);
	cancel_work_sync(&pf->service_task);

12201
	if (pf->wol_en && (pf->hw_features & I40E_HW_WOL_MC_MAGIC_PKT_WAKE))
12202 12203
		i40e_enable_mc_magic_wake(pf);

12204
	i40e_prep_for_reset(pf, false);
12205

S
Shannon Nelson 已提交
12206 12207 12208
	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));

12209 12210 12211 12212 12213 12214
	/* Clear the interrupt scheme and release our IRQs so that the system
	 * can safely hibernate even when there are a large number of CPUs.
	 * Otherwise hibernation might fail when mapping all the vectors back
	 * to CPU0.
	 */
	i40e_clear_interrupt_scheme(pf);
12215

12216
	return 0;
J
Jesse Brandeburg 已提交
12217 12218
}

12219
/**
12220 12221
 * i40e_resume - PM callback for waking up from D3
 * @dev: generic device information structure
12222
 **/
12223
static int i40e_resume(struct device *dev)
12224
{
12225
	struct pci_dev *pdev = to_pci_dev(dev);
12226
	struct i40e_pf *pf = pci_get_drvdata(pdev);
12227
	int err;
12228

12229 12230 12231 12232
	/* If we're not suspended, then there is nothing to do */
	if (!test_bit(__I40E_SUSPENDED, pf->state))
		return 0;

12233 12234 12235 12236 12237 12238 12239 12240 12241
	/* We cleared the interrupt scheme when we suspended, so we need to
	 * restore it now to resume device functionality.
	 */
	err = i40e_restore_interrupt_scheme(pf);
	if (err) {
		dev_err(&pdev->dev, "Cannot restore interrupt scheme: %d\n",
			err);
	}

12242 12243 12244 12245 12246
	clear_bit(__I40E_DOWN, pf->state);
	i40e_reset_and_rebuild(pf, false, false);

	/* Clear suspended state last after everything is recovered */
	clear_bit(__I40E_SUSPENDED, pf->state);
12247

12248 12249 12250 12251
	/* Restart the service task */
	mod_timer(&pf->service_timer,
		  round_jiffies(jiffies + pf->service_timer_period));

12252 12253 12254
	return 0;
}

12255 12256
#endif /* CONFIG_PM */

J
Jesse Brandeburg 已提交
12257 12258 12259
static const struct pci_error_handlers i40e_err_handler = {
	.error_detected = i40e_pci_error_detected,
	.slot_reset = i40e_pci_error_slot_reset,
12260 12261
	.reset_prepare = i40e_pci_error_reset_prepare,
	.reset_done = i40e_pci_error_reset_done,
J
Jesse Brandeburg 已提交
12262 12263 12264
	.resume = i40e_pci_error_resume,
};

12265 12266
static SIMPLE_DEV_PM_OPS(i40e_pm_ops, i40e_suspend, i40e_resume);

J
Jesse Brandeburg 已提交
12267 12268 12269 12270 12271
static struct pci_driver i40e_driver = {
	.name     = i40e_driver_name,
	.id_table = i40e_pci_tbl,
	.probe    = i40e_probe,
	.remove   = i40e_remove,
12272
#ifdef CONFIG_PM
12273 12274 12275 12276
	.driver   = {
		.pm = &i40e_pm_ops,
	},
#endif /* CONFIG_PM */
12277
	.shutdown = i40e_shutdown,
J
Jesse Brandeburg 已提交
12278 12279 12280 12281 12282 12283 12284 12285 12286 12287 12288 12289 12290 12291 12292
	.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);
12293

12294 12295 12296 12297 12298 12299
	/* There is no need to throttle the number of active tasks because
	 * each device limits its own task using a state bit for scheduling
	 * the service task, and the device tasks do not interfere with each
	 * other, so we don't set a max task limit. We must set WQ_MEM_RECLAIM
	 * since we need to be able to guarantee forward progress even under
	 * memory pressure.
12300
	 */
12301
	i40e_wq = alloc_workqueue("%s", WQ_MEM_RECLAIM, 0, i40e_driver_name);
12302 12303 12304 12305 12306
	if (!i40e_wq) {
		pr_err("%s: Failed to create workqueue\n", i40e_driver_name);
		return -ENOMEM;
	}

J
Jesse Brandeburg 已提交
12307 12308 12309 12310 12311 12312 12313 12314 12315 12316 12317 12318 12319 12320
	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);
12321
	destroy_workqueue(i40e_wq);
J
Jesse Brandeburg 已提交
12322 12323 12324
	i40e_dbg_exit();
}
module_exit(i40e_exit_module);