i40e_main.c 395.1 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
#define DRV_VERSION_MAJOR 2
50 51
#define DRV_VERSION_MINOR 3
#define DRV_VERSION_BUILD 2
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);
72 73 74
static int i40e_get_capabilities(struct i40e_pf *pf,
				 enum i40e_admin_queue_opc list_type);

J
Jesse Brandeburg 已提交
75 76 77 78 79 80 81 82

/* 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) }
 */
83
static const struct pci_device_id i40e_pci_tbl[] = {
84 85 86 87 88 89 90
	{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
91
	{PCI_VDEVICE(INTEL, I40E_DEV_ID_10G_BASE_T), 0},
92
	{PCI_VDEVICE(INTEL, I40E_DEV_ID_10G_BASE_T4), 0},
93 94
	{PCI_VDEVICE(INTEL, I40E_DEV_ID_KX_X722), 0},
	{PCI_VDEVICE(INTEL, I40E_DEV_ID_QSFP_X722), 0},
95 96 97
	{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},
98
	{PCI_VDEVICE(INTEL, I40E_DEV_ID_SFP_I_X722), 0},
99 100
	{PCI_VDEVICE(INTEL, I40E_DEV_ID_20G_KR2), 0},
	{PCI_VDEVICE(INTEL, I40E_DEV_ID_20G_KR2_A), 0},
101 102
	{PCI_VDEVICE(INTEL, I40E_DEV_ID_25G_B), 0},
	{PCI_VDEVICE(INTEL, I40E_DEV_ID_25G_SFP28), 0},
J
Jesse Brandeburg 已提交
103 104 105 106 107 108 109
	/* required last entry */
	{0, }
};
MODULE_DEVICE_TABLE(pci, i40e_pci_tbl);

#define I40E_MAX_VF_COUNT 128
static int debug = -1;
110 111
module_param(debug, uint, 0);
MODULE_PARM_DESC(debug, "Debug level (0=none,...,16=all), Debug mask (0x8XXXXXXX)");
J
Jesse Brandeburg 已提交
112 113 114 115 116 117

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);

118 119
static struct workqueue_struct *i40e_wq;

J
Jesse Brandeburg 已提交
120 121 122 123 124 125 126 127 128 129 130 131 132 133 134
/**
 * 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
135 136
	if (!mem->va)
		return -ENOMEM;
J
Jesse Brandeburg 已提交
137

J
Jesse Brandeburg 已提交
138
	return 0;
J
Jesse Brandeburg 已提交
139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169
}

/**
 * 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
170 171
	if (!mem->va)
		return -ENOMEM;
J
Jesse Brandeburg 已提交
172

J
Jesse Brandeburg 已提交
173
	return 0;
J
Jesse Brandeburg 已提交
174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207
}

/**
 * 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;
208
	int i, j;
J
Jesse Brandeburg 已提交
209 210 211

	if (!pile || needed == 0 || id >= I40E_PILE_VALID_BIT) {
		dev_info(&pf->pdev->dev,
212 213
			 "param err: pile=%s needed=%d id=0x%04x\n",
			 pile ? "<valid>" : "<null>", needed, id);
J
Jesse Brandeburg 已提交
214 215 216 217 218
		return -EINVAL;
	}

	/* start the linear search with an imperfect hint */
	i = pile->search_hint;
219
	while (i < pile->num_entries) {
J
Jesse Brandeburg 已提交
220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237
		/* 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;
238
			break;
J
Jesse Brandeburg 已提交
239
		}
J
Jesse Brandeburg 已提交
240 241 242

		/* not enough, so skip over it and continue looking */
		i += j;
J
Jesse Brandeburg 已提交
243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277
	}

	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;
}

278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293
/**
 * 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
294 295 296 297 298 299
/**
 * 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
 **/
300
void i40e_service_event_schedule(struct i40e_pf *pf)
J
Jesse Brandeburg 已提交
301
{
302
	if (!test_bit(__I40E_DOWN, pf->state) &&
303
	    !test_bit(__I40E_RESET_RECOVERY_PENDING, pf->state))
304
		queue_work(i40e_wq, &pf->service_task);
J
Jesse Brandeburg 已提交
305 306 307 308 309 310 311 312 313 314 315 316 317 318 319
}

/**
 * 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;
320 321 322
	struct i40e_ring *tx_ring = NULL;
	unsigned int i, hung_queue = 0;
	u32 head, val;
J
Jesse Brandeburg 已提交
323 324 325

	pf->tx_timeout_count++;

326 327 328 329 330 331
	/* 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
332
		trans_start = q->trans_start;
333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355
		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
356
	if (time_after(jiffies, (pf->tx_timeout_last_recovery + HZ*20)))
357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377
		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
378
	pf->tx_timeout_last_recovery = jiffies;
379 380
	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
381 382 383

	switch (pf->tx_timeout_recovery_level) {
	case 1:
384
		set_bit(__I40E_PF_RESET_REQUESTED, pf->state);
J
Jesse Brandeburg 已提交
385 386
		break;
	case 2:
387
		set_bit(__I40E_CORE_RESET_REQUESTED, pf->state);
J
Jesse Brandeburg 已提交
388 389
		break;
	case 3:
390
		set_bit(__I40E_GLOBAL_RESET_REQUESTED, pf->state);
J
Jesse Brandeburg 已提交
391 392 393 394 395
		break;
	default:
		netdev_err(netdev, "tx_timeout recovery unsuccessful\n");
		break;
	}
396

J
Jesse Brandeburg 已提交
397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412
	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;
}

413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433
/**
 * 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
434 435 436 437 438 439 440
/**
 * 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.
 **/
441
static void i40e_get_netdev_stats_struct(struct net_device *netdev,
442
				  struct rtnl_link_stats64 *stats)
J
Jesse Brandeburg 已提交
443 444
{
	struct i40e_netdev_priv *np = netdev_priv(netdev);
445
	struct i40e_ring *tx_ring, *rx_ring;
J
Jesse Brandeburg 已提交
446
	struct i40e_vsi *vsi = np->vsi;
447 448 449
	struct rtnl_link_stats64 *vsi_stats = i40e_get_vsi_stats_struct(vsi);
	int i;

450
	if (test_bit(__I40E_VSI_DOWN, vsi->state))
451
		return;
452

J
Jesse Brandeburg 已提交
453
	if (!vsi->tx_rings)
454
		return;
J
Jesse Brandeburg 已提交
455

456 457 458 459 460
	rcu_read_lock();
	for (i = 0; i < vsi->num_queue_pairs; i++) {
		u64 bytes, packets;
		unsigned int start;

461
		tx_ring = READ_ONCE(vsi->tx_rings[i]);
462 463
		if (!tx_ring)
			continue;
464
		i40e_get_netdev_stats_struct_tx(tx_ring, stats);
465 466 467 468

		rx_ring = &tx_ring[1];

		do {
469
			start = u64_stats_fetch_begin_irq(&rx_ring->syncp);
470 471
			packets = rx_ring->stats.packets;
			bytes   = rx_ring->stats.bytes;
472
		} while (u64_stats_fetch_retry_irq(&rx_ring->syncp, start));
J
Jesse Brandeburg 已提交
473

474 475
		stats->rx_packets += packets;
		stats->rx_bytes   += bytes;
476 477 478

		if (i40e_enabled_xdp_vsi(vsi))
			i40e_get_netdev_stats_struct_tx(&rx_ring[1], stats);
479 480 481
	}
	rcu_read_unlock();

482
	/* following stats updated by i40e_watchdog_subtask() */
483 484 485 486
	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;
487
	stats->rx_dropped	= vsi_stats->rx_dropped;
488 489
	stats->rx_crc_errors	= vsi_stats->rx_crc_errors;
	stats->rx_length_errors	= vsi_stats->rx_length_errors;
J
Jesse Brandeburg 已提交
490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508
}

/**
 * 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));
509
	if (vsi->rx_rings && vsi->rx_rings[0]) {
J
Jesse Brandeburg 已提交
510
		for (i = 0; i < vsi->num_queue_pairs; i++) {
J
Jesse Brandeburg 已提交
511
			memset(&vsi->rx_rings[i]->stats, 0,
512
			       sizeof(vsi->rx_rings[i]->stats));
J
Jesse Brandeburg 已提交
513
			memset(&vsi->rx_rings[i]->rx_stats, 0,
514
			       sizeof(vsi->rx_rings[i]->rx_stats));
J
Jesse Brandeburg 已提交
515
			memset(&vsi->tx_rings[i]->stats, 0,
516 517 518
			       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
519
		}
520
	}
J
Jesse Brandeburg 已提交
521 522 523 524
	vsi->stat_offsets_loaded = false;
}

/**
525
 * i40e_pf_reset_stats - Reset all of the stats for the given PF
J
Jesse Brandeburg 已提交
526 527 528 529
 * @pf: the PF to be reset
 **/
void i40e_pf_reset_stats(struct i40e_pf *pf)
{
530 531
	int i;

J
Jesse Brandeburg 已提交
532 533 534
	memset(&pf->stats, 0, sizeof(pf->stats));
	memset(&pf->stats_offsets, 0, sizeof(pf->stats_offsets));
	pf->stat_offsets_loaded = false;
535 536 537 538 539 540 541 542 543 544

	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;
		}
	}
545
	pf->hw_csum_rx_error = 0;
J
Jesse Brandeburg 已提交
546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567
}

/**
 * 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;

568
	if (hw->device_id == I40E_DEV_ID_QEMU) {
J
Jesse Brandeburg 已提交
569 570 571 572 573 574 575 576 577 578
		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
579
		*stat = (new_data + BIT_ULL(48)) - *offset;
J
Jesse Brandeburg 已提交
580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601
	*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
602
		*stat = (u32)((new_data + BIT_ULL(32)) - *offset);
J
Jesse Brandeburg 已提交
603 604
}

605 606 607 608 609 610 611 612 613 614 615 616 617 618
/**
 * 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
619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640
/**
 * 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);
641 642 643 644 645 646
	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
647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693

	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 */
694 695 696
	struct i40e_veb_tc_stats *veb_oes;
	struct i40e_veb_tc_stats *veb_es;
	int i, idx = 0;
J
Jesse Brandeburg 已提交
697 698 699 700

	idx = veb->stats_idx;
	es = &veb->stats;
	oes = &veb->stats_offsets;
701 702
	veb_es = &veb->tc_stats;
	veb_oes = &veb->tc_stats_offsets;
J
Jesse Brandeburg 已提交
703 704 705 706 707

	/* 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);
708 709 710 711 712
	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
713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737
	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);
738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759
	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
760 761 762 763
	veb->stat_offsets_loaded = true;
}

/**
S
Shannon Nelson 已提交
764
 * i40e_update_vsi_stats - Update the vsi statistics counters.
J
Jesse Brandeburg 已提交
765 766 767 768 769 770
 * @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
771
 * VF communications.  We sort it out here.
J
Jesse Brandeburg 已提交
772
 **/
S
Shannon Nelson 已提交
773
static void i40e_update_vsi_stats(struct i40e_vsi *vsi)
J
Jesse Brandeburg 已提交
774 775 776 777 778 779 780
{
	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;
781
	struct i40e_ring *p;
J
Jesse Brandeburg 已提交
782
	u32 rx_page, rx_buf;
783 784
	u64 bytes, packets;
	unsigned int start;
785
	u64 tx_linearize;
786
	u64 tx_force_wb;
J
Jesse Brandeburg 已提交
787 788 789 790
	u64 rx_p, rx_b;
	u64 tx_p, tx_b;
	u16 q;

791 792
	if (test_bit(__I40E_VSI_DOWN, vsi->state) ||
	    test_bit(__I40E_CONFIG_BUSY, pf->state))
J
Jesse Brandeburg 已提交
793 794 795 796 797 798 799 800 801 802 803 804
		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;
805
	tx_restart = tx_busy = tx_linearize = tx_force_wb = 0;
J
Jesse Brandeburg 已提交
806 807
	rx_page = 0;
	rx_buf = 0;
808
	rcu_read_lock();
J
Jesse Brandeburg 已提交
809
	for (q = 0; q < vsi->num_queue_pairs; q++) {
810
		/* locate Tx ring */
811
		p = READ_ONCE(vsi->tx_rings[q]);
812 813

		do {
814
			start = u64_stats_fetch_begin_irq(&p->syncp);
815 816
			packets = p->stats.packets;
			bytes = p->stats.bytes;
817
		} while (u64_stats_fetch_retry_irq(&p->syncp, start));
818 819 820 821
		tx_b += bytes;
		tx_p += packets;
		tx_restart += p->tx_stats.restart_queue;
		tx_busy += p->tx_stats.tx_busy;
822
		tx_linearize += p->tx_stats.tx_linearize;
823
		tx_force_wb += p->tx_stats.tx_force_wb;
J
Jesse Brandeburg 已提交
824

825 826 827
		/* Rx queue is part of the same block as Tx queue */
		p = &p[1];
		do {
828
			start = u64_stats_fetch_begin_irq(&p->syncp);
829 830
			packets = p->stats.packets;
			bytes = p->stats.bytes;
831
		} while (u64_stats_fetch_retry_irq(&p->syncp, start));
832 833
		rx_b += bytes;
		rx_p += packets;
M
Mitch Williams 已提交
834 835
		rx_buf += p->rx_stats.alloc_buff_failed;
		rx_page += p->rx_stats.alloc_page_failed;
J
Jesse Brandeburg 已提交
836
	}
837
	rcu_read_unlock();
J
Jesse Brandeburg 已提交
838 839
	vsi->tx_restart = tx_restart;
	vsi->tx_busy = tx_busy;
840
	vsi->tx_linearize = tx_linearize;
841
	vsi->tx_force_wb = tx_force_wb;
J
Jesse Brandeburg 已提交
842 843 844 845 846 847 848 849 850
	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
851
	i40e_update_eth_stats(vsi);
J
Jesse Brandeburg 已提交
852 853 854 855
	ons->tx_errors = oes->tx_errors;
	ns->tx_errors = es->tx_errors;
	ons->multicast = oes->rx_multicast;
	ns->multicast = es->rx_multicast;
856 857
	ons->rx_dropped = oes->rx_discards;
	ns->rx_dropped = es->rx_discards;
J
Jesse Brandeburg 已提交
858 859 860
	ons->tx_dropped = oes->tx_discards;
	ns->tx_dropped = es->tx_discards;

S
Shannon Nelson 已提交
861
	/* pull in a couple PF stats if this is the main vsi */
J
Jesse Brandeburg 已提交
862
	if (vsi == pf->vsi[pf->lan_vsi]) {
S
Shannon Nelson 已提交
863 864 865 866 867
		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
868

S
Shannon Nelson 已提交
869
/**
870
 * i40e_update_pf_stats - Update the PF statistics counters.
S
Shannon Nelson 已提交
871 872 873 874 875 876 877 878 879
 * @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
880

S
Shannon Nelson 已提交
881 882 883 884 885 886 887 888 889 890 891 892
	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
893 894 895 896 897
	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
898 899 900 901 902
	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
903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922
	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
923

S
Shannon Nelson 已提交
924 925 926 927
	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
928

S
Shannon Nelson 已提交
929 930 931
	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
932

S
Shannon Nelson 已提交
933 934 935
	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
936

S
Shannon Nelson 已提交
937 938 939 940 941 942 943 944
	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
945

S
Shannon Nelson 已提交
946 947 948 949
	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
950

S
Shannon Nelson 已提交
951 952 953 954 955 956
	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);
957 958 959
	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
960 961 962
	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
963

S
Shannon Nelson 已提交
964
	for (i = 0; i < 8; i++) {
965 966 967 968
		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
969
		i40e_stat_update32(hw, I40E_GLPRT_PXONRXC(hw->port, i),
J
Jesse Brandeburg 已提交
970
				   pf->stat_offsets_loaded,
S
Shannon Nelson 已提交
971 972 973
				   &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
974
				   pf->stat_offsets_loaded,
S
Shannon Nelson 已提交
975 976 977
				   &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
978
				   pf->stat_offsets_loaded,
S
Shannon Nelson 已提交
979 980 981 982
				   &osd->priority_xoff_tx[i],
				   &nsd->priority_xoff_tx[i]);
		i40e_stat_update32(hw,
				   I40E_GLPRT_RXON2OFFCNT(hw->port, i),
983
				   pf->stat_offsets_loaded,
S
Shannon Nelson 已提交
984 985
				   &osd->priority_xon_2_xoff[i],
				   &nsd->priority_xon_2_xoff[i]);
J
Jesse Brandeburg 已提交
986 987
	}

S
Shannon Nelson 已提交
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 1056 1057 1058
	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);

1059
	/* FDIR stats */
1060 1061 1062 1063 1064 1065 1066 1067 1068
	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);
1069

S
Shannon Nelson 已提交
1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083
	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);

1084
	if (pf->flags & I40E_FLAG_FD_SB_ENABLED &&
1085
	    !(pf->flags & I40E_FLAG_FD_SB_AUTO_DISABLED))
1086 1087 1088 1089 1090
		nsd->fd_sb_status = true;
	else
		nsd->fd_sb_status = false;

	if (pf->flags & I40E_FLAG_FD_ATR_ENABLED &&
1091
	    !(pf->flags & I40E_FLAG_FD_ATR_AUTO_DISABLED))
1092 1093 1094 1095
		nsd->fd_atr_status = true;
	else
		nsd->fd_atr_status = false;

J
Jesse Brandeburg 已提交
1096 1097 1098
	pf->stat_offsets_loaded = true;
}

S
Shannon Nelson 已提交
1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114
/**
 * 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
1115 1116 1117 1118 1119 1120 1121 1122 1123
/**
 * 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,
1124
						const u8 *macaddr, s16 vlan)
J
Jesse Brandeburg 已提交
1125 1126
{
	struct i40e_mac_filter *f;
1127
	u64 key;
J
Jesse Brandeburg 已提交
1128 1129 1130 1131

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

1132 1133
	key = i40e_addr_to_hkey(macaddr);
	hash_for_each_possible(vsi->mac_filter_hash, f, hlist, key) {
J
Jesse Brandeburg 已提交
1134
		if ((ether_addr_equal(macaddr, f->macaddr)) &&
1135
		    (vlan == f->vlan))
J
Jesse Brandeburg 已提交
1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148
			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
 **/
1149
struct i40e_mac_filter *i40e_find_mac(struct i40e_vsi *vsi, const u8 *macaddr)
J
Jesse Brandeburg 已提交
1150 1151
{
	struct i40e_mac_filter *f;
1152
	u64 key;
J
Jesse Brandeburg 已提交
1153 1154 1155 1156

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

1157 1158
	key = i40e_addr_to_hkey(macaddr);
	hash_for_each_possible(vsi->mac_filter_hash, f, hlist, key) {
1159
		if ((ether_addr_equal(macaddr, f->macaddr)))
J
Jesse Brandeburg 已提交
1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172
			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)
{
1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185
	/* 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
1186
	 *    i40e_sync_filters_subtask.
1187 1188 1189 1190
	 *
	 * 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
1191
	 * i40e_sync_filters_subtask, but we already have to search
1192 1193 1194 1195
	 * 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
1196
	 */
1197
	return vsi->has_vlan_filter;
J
Jesse Brandeburg 已提交
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 1231 1232 1233
/**
 * 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)
{
1234
	s16 pvid = le16_to_cpu(vsi->info.pvid);
1235
	struct i40e_mac_filter *f, *add_head;
1236
	struct i40e_new_mac_filter *new;
1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254
	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 */
1255
	hlist_for_each_entry(new, tmp_add_list, hlist) {
1256 1257
		if (pvid && new->f->vlan != pvid)
			new->f->vlan = pvid;
1258 1259 1260 1261
		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;
1262 1263 1264 1265 1266 1267 1268 1269 1270
	}

	/* 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.
		 */
1271
		if ((pvid && f->vlan != pvid) ||
1272 1273 1274
		    (vlan_filters && f->vlan == I40E_VLAN_ANY) ||
		    (!vlan_filters && f->vlan == 0)) {
			/* Determine the new vlan we will be adding */
1275 1276
			if (pvid)
				new_vlan = pvid;
1277 1278 1279 1280 1281 1282 1283 1284 1285 1286
			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;

1287 1288 1289 1290 1291 1292 1293 1294 1295 1296
			/* 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);
1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309

			/* 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;
}

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 1340 1341 1342
/**
 * 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
1343 1344 1345 1346 1347 1348 1349
/**
 * 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
1350
 *
1351
 * NOTE: This function is expected to be called with mac_filter_hash_lock
K
Kiran Patil 已提交
1352
 * being held.
J
Jesse Brandeburg 已提交
1353 1354
 **/
struct i40e_mac_filter *i40e_add_filter(struct i40e_vsi *vsi,
1355
					const u8 *macaddr, s16 vlan)
J
Jesse Brandeburg 已提交
1356 1357
{
	struct i40e_mac_filter *f;
1358
	u64 key;
J
Jesse Brandeburg 已提交
1359 1360 1361 1362

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

1363
	f = i40e_find_filter(vsi, macaddr, vlan);
J
Jesse Brandeburg 已提交
1364 1365 1366
	if (!f) {
		f = kzalloc(sizeof(*f), GFP_ATOMIC);
		if (!f)
1367
			return NULL;
J
Jesse Brandeburg 已提交
1368

1369 1370 1371 1372 1373 1374
		/* Update the boolean indicating if we need to function in
		 * VLAN mode.
		 */
		if (vlan >= 0)
			vsi->has_vlan_filter = true;

1375
		ether_addr_copy(f->macaddr, macaddr);
J
Jesse Brandeburg 已提交
1376
		f->vlan = vlan;
1377
		f->state = I40E_FILTER_NEW;
1378 1379 1380 1381
		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
1382 1383 1384 1385 1386

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

1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397
	/* 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
1398 1399 1400 1401
	return f;
}

/**
1402 1403 1404 1405 1406 1407 1408
 * __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
1409
 *
1410
 * NOTE: This function is expected to be called with mac_filter_hash_lock
K
Kiran Patil 已提交
1411
 * being held.
1412 1413 1414
 * 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
1415
 **/
1416
void __i40e_del_filter(struct i40e_vsi *vsi, struct i40e_mac_filter *f)
J
Jesse Brandeburg 已提交
1417
{
1418
	if (!f)
J
Jesse Brandeburg 已提交
1419 1420
		return;

1421 1422 1423 1424
	/* 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.
	 */
1425 1426
	if ((f->state == I40E_FILTER_FAILED) ||
	    (f->state == I40E_FILTER_NEW)) {
1427
		hash_del(&f->hlist);
1428
		kfree(f);
J
Jesse Brandeburg 已提交
1429
	} else {
1430
		f->state = I40E_FILTER_REMOVE;
J
Jesse Brandeburg 已提交
1431
	}
1432 1433 1434

	vsi->flags |= I40E_VSI_FLAG_FILTER_CHANGED;
	vsi->back->flags |= I40E_FLAG_FILTER_SYNC;
J
Jesse Brandeburg 已提交
1435 1436
}

1437 1438 1439 1440 1441 1442
/**
 * i40e_del_filter - Remove a MAC/VLAN filter from the VSI
 * @vsi: the VSI to be searched
 * @macaddr: the MAC address
 * @vlan: the VLAN
 *
1443
 * NOTE: This function is expected to be called with mac_filter_hash_lock
1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459
 * 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);
}

1460
/**
1461
 * i40e_add_mac_filter - Add a MAC filter for all active VLANs
1462 1463 1464
 * @vsi: the VSI to be searched
 * @macaddr: the mac address to be filtered
 *
1465 1466
 * 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
1467 1468
 * unique vlan that already exists. If a PVID has been assigned, instead only
 * add the macaddr to that VLAN.
1469
 *
1470
 * Returns last filter added on success, else NULL
1471
 **/
1472 1473
struct i40e_mac_filter *i40e_add_mac_filter(struct i40e_vsi *vsi,
					    const u8 *macaddr)
1474
{
1475
	struct i40e_mac_filter *f, *add = NULL;
1476 1477
	struct hlist_node *h;
	int bkt;
1478 1479 1480 1481

	if (vsi->info.pvid)
		return i40e_add_filter(vsi, macaddr,
				       le16_to_cpu(vsi->info.pvid));
1482

1483 1484 1485
	if (!i40e_is_vsi_in_vlan(vsi))
		return i40e_add_filter(vsi, macaddr, I40E_VLAN_ANY);

1486
	hash_for_each_safe(vsi->mac_filter_hash, bkt, h, f, hlist) {
1487 1488
		if (f->state == I40E_FILTER_REMOVE)
			continue;
1489 1490 1491
		add = i40e_add_filter(vsi, macaddr, f->vlan);
		if (!add)
			return NULL;
1492 1493
	}

1494
	return add;
1495 1496 1497
}

/**
1498
 * i40e_del_mac_filter - Remove a MAC filter from all VLANs
1499 1500 1501
 * @vsi: the VSI to be searched
 * @macaddr: the mac address to be removed
 *
1502 1503
 * Removes a given MAC address from a VSI regardless of what VLAN it has been
 * associated with.
1504 1505 1506
 *
 * Returns 0 for success, or error
 **/
1507
int i40e_del_mac_filter(struct i40e_vsi *vsi, const u8 *macaddr)
1508
{
1509 1510
	struct i40e_mac_filter *f;
	struct hlist_node *h;
1511
	bool found = false;
1512
	int bkt;
1513

1514 1515 1516
	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) {
1517 1518 1519 1520
		if (ether_addr_equal(macaddr, f->macaddr)) {
			__i40e_del_filter(vsi, f);
			found = true;
		}
1521
	}
1522 1523

	if (found)
1524
		return 0;
1525 1526
	else
		return -ENOENT;
1527 1528
}

J
Jesse Brandeburg 已提交
1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539
/**
 * 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;
1540 1541
	struct i40e_pf *pf = vsi->back;
	struct i40e_hw *hw = &pf->hw;
J
Jesse Brandeburg 已提交
1542 1543 1544 1545 1546
	struct sockaddr *addr = p;

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

1547 1548 1549 1550 1551
	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
1552

1553 1554
	if (test_bit(__I40E_VSI_DOWN, vsi->back->state) ||
	    test_bit(__I40E_RESET_RECOVERY_PENDING, vsi->back->state))
1555 1556
		return -EADDRNOTAVAIL;

1557 1558 1559 1560 1561 1562
	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);

1563 1564 1565 1566 1567 1568 1569 1570
	/* Copy the address first, so that we avoid a possible race with
	 * .set_rx_mode(). If we copy after changing the address in the filter
	 * list, we might open ourselves to a narrow race window where
	 * .set_rx_mode could delete our dev_addr filter and prevent traffic
	 * from passing.
	 */
	ether_addr_copy(netdev->dev_addr, addr->sa_data);

1571
	spin_lock_bh(&vsi->mac_filter_hash_lock);
1572 1573
	i40e_del_mac_filter(vsi, netdev->dev_addr);
	i40e_add_mac_filter(vsi, addr->sa_data);
1574
	spin_unlock_bh(&vsi->mac_filter_hash_lock);
J
Jesse Brandeburg 已提交
1575 1576
	if (vsi->type == I40E_VSI_MAIN) {
		i40e_status ret;
J
Jesse Brandeburg 已提交
1577

J
Jesse Brandeburg 已提交
1578
		ret = i40e_aq_mac_address_write(&vsi->back->hw,
1579
						I40E_AQC_WRITE_TYPE_LAA_WOL,
J
Jesse Brandeburg 已提交
1580
						addr->sa_data, NULL);
1581 1582 1583 1584
		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));
1585 1586
	}

1587 1588 1589 1590 1591
	/* 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
1592 1593
}

1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757
/**
 * i40e_config_rss_aq - Prepare for RSS using AQ commands
 * @vsi: vsi structure
 * @seed: RSS hash seed
 **/
static int i40e_config_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) {
		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;
		}
	}
	if (lut) {
		bool pf_lut = vsi->type == I40E_VSI_MAIN ? true : false;

		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;
		}
	}
	return ret;
}

/**
 * i40e_vsi_config_rss - Prepare for VSI(VMDq) RSS if used
 * @vsi: VSI structure
 **/
static int i40e_vsi_config_rss(struct i40e_vsi *vsi)
{
	struct i40e_pf *pf = vsi->back;
	u8 seed[I40E_HKEY_ARRAY_SIZE];
	u8 *lut;
	int ret;

	if (!(pf->hw_features & I40E_HW_RSS_AQ_CAPABLE))
		return 0;
	if (!vsi->rss_size)
		vsi->rss_size = min_t(int, pf->alloc_rss_size,
				      vsi->num_queue_pairs);
	if (!vsi->rss_size)
		return -EINVAL;
	lut = kzalloc(vsi->rss_table_size, GFP_KERNEL);
	if (!lut)
		return -ENOMEM;

	/* 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);
	ret = i40e_config_rss_aq(vsi, seed, lut, vsi->rss_table_size);
	kfree(lut);
	return ret;
}

/**
 * i40e_vsi_setup_queue_map_mqprio - Prepares mqprio based tc_config
 * @vsi: the VSI being configured,
 * @ctxt: VSI context structure
 * @enabled_tc: number of traffic classes to enable
 *
 * Prepares VSI tc_config to have queue configurations based on MQPRIO options.
 **/
static int i40e_vsi_setup_queue_map_mqprio(struct i40e_vsi *vsi,
					   struct i40e_vsi_context *ctxt,
					   u8 enabled_tc)
{
	u16 qcount = 0, max_qcount, qmap, sections = 0;
	int i, override_q, pow, num_qps, ret;
	u8 netdev_tc = 0, offset = 0;

	if (vsi->type != I40E_VSI_MAIN)
		return -EINVAL;
	sections = I40E_AQ_VSI_PROP_QUEUE_MAP_VALID;
	sections |= I40E_AQ_VSI_PROP_SCHED_VALID;
	vsi->tc_config.numtc = vsi->mqprio_qopt.qopt.num_tc;
	vsi->tc_config.enabled_tc = enabled_tc ? enabled_tc : 1;
	num_qps = vsi->mqprio_qopt.qopt.count[0];

	/* find the next higher power-of-2 of num queue pairs */
	pow = ilog2(num_qps);
	if (!is_power_of_2(num_qps))
		pow++;
	qmap = (offset << I40E_AQ_VSI_TC_QUE_OFFSET_SHIFT) |
		(pow << I40E_AQ_VSI_TC_QUE_NUMBER_SHIFT);

	/* Setup queue offset/count for all TCs for given VSI */
	max_qcount = vsi->mqprio_qopt.qopt.count[0];
	for (i = 0; i < I40E_MAX_TRAFFIC_CLASS; i++) {
		/* See if the given TC is enabled for the given VSI */
		if (vsi->tc_config.enabled_tc & BIT(i)) {
			offset = vsi->mqprio_qopt.qopt.offset[i];
			qcount = vsi->mqprio_qopt.qopt.count[i];
			if (qcount > max_qcount)
				max_qcount = qcount;
			vsi->tc_config.tc_info[i].qoffset = offset;
			vsi->tc_config.tc_info[i].qcount = qcount;
			vsi->tc_config.tc_info[i].netdev_tc = netdev_tc++;
		} 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;
		}
	}

	/* Set actual Tx/Rx queue pairs */
	vsi->num_queue_pairs = offset + qcount;

	/* Setup queue TC[0].qmap for given VSI context */
	ctxt->info.tc_mapping[0] = cpu_to_le16(qmap);
	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);

	/* Reconfigure RSS for main VSI with max queue count */
	vsi->rss_size = max_qcount;
	ret = i40e_vsi_config_rss(vsi);
	if (ret) {
		dev_info(&vsi->back->pdev->dev,
			 "Failed to reconfig rss for num_queues (%u)\n",
			 max_qcount);
		return ret;
	}
	vsi->reconfig_rss = true;
	dev_dbg(&vsi->back->pdev->dev,
		"Reconfigured rss with num_queues (%u)\n", max_qcount);

	/* Find queue count available for channel VSIs and starting offset
	 * for channel VSIs
	 */
	override_q = vsi->mqprio_qopt.qopt.count[0];
	if (override_q && override_q < vsi->num_queue_pairs) {
		vsi->cnt_q_avail = vsi->num_queue_pairs - override_q;
		vsi->next_base_queue = override_q;
	}
	return 0;
}

J
Jesse Brandeburg 已提交
1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774
/**
 * 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;
1775
	u16 numtc = 1;
J
Jesse Brandeburg 已提交
1776 1777 1778 1779
	u16 qcount;
	u8 offset;
	u16 qmap;
	int i;
N
Neerav Parikh 已提交
1780
	u16 num_tc_qps = 0;
J
Jesse Brandeburg 已提交
1781 1782 1783 1784

	sections = I40E_AQ_VSI_PROP_QUEUE_MAP_VALID;
	offset = 0;

1785 1786
	/* Number of queues per enabled TC */
	num_tc_qps = vsi->alloc_queue_pairs;
J
Jesse Brandeburg 已提交
1787 1788
	if (enabled_tc && (vsi->back->flags & I40E_FLAG_DCB_ENABLED)) {
		/* Find numtc from enabled TC bitmap */
1789
		for (i = 0, numtc = 0; i < I40E_MAX_TRAFFIC_CLASS; i++) {
J
Jesse Brandeburg 已提交
1790
			if (enabled_tc & BIT(i)) /* TC is enabled */
J
Jesse Brandeburg 已提交
1791 1792 1793 1794 1795 1796
				numtc++;
		}
		if (!numtc) {
			dev_warn(&pf->pdev->dev, "DCB is enabled but no TC enabled, forcing TC0\n");
			numtc = 1;
		}
1797 1798 1799
		num_tc_qps = num_tc_qps / numtc;
		num_tc_qps = min_t(int, num_tc_qps,
				   i40e_pf_get_max_q_per_tc(pf));
J
Jesse Brandeburg 已提交
1800 1801 1802 1803 1804
	}

	vsi->tc_config.numtc = numtc;
	vsi->tc_config.enabled_tc = enabled_tc ? enabled_tc : 1;

1805 1806 1807 1808
	/* Do not allow use more TC queue pairs than MSI-X vectors exist */
	if (pf->flags & I40E_FLAG_MSIX_ENABLED)
		num_tc_qps = min_t(int, num_tc_qps, pf->num_lan_msix);

J
Jesse Brandeburg 已提交
1809 1810 1811
	/* 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
1812
		if (vsi->tc_config.enabled_tc & BIT(i)) {
1813
			/* TC is enabled */
J
Jesse Brandeburg 已提交
1814 1815 1816 1817
			int pow, num_qps;

			switch (vsi->type) {
			case I40E_VSI_MAIN:
1818 1819 1820 1821 1822 1823 1824
				if (!(pf->flags & (I40E_FLAG_FD_SB_ENABLED |
				    I40E_FLAG_FD_ATR_ENABLED)) ||
				    vsi->tc_config.enabled_tc != 1) {
					qcount = min_t(int, pf->alloc_rss_size,
						       num_tc_qps);
					break;
				}
J
Jesse Brandeburg 已提交
1825 1826 1827 1828
			case I40E_VSI_FDIR:
			case I40E_VSI_SRIOV:
			case I40E_VSI_VMDQ2:
			default:
N
Neerav Parikh 已提交
1829
				qcount = num_tc_qps;
J
Jesse Brandeburg 已提交
1830 1831 1832
				WARN_ON(i != 0);
				break;
			}
N
Neerav Parikh 已提交
1833 1834
			vsi->tc_config.tc_info[i].qoffset = offset;
			vsi->tc_config.tc_info[i].qcount = qcount;
J
Jesse Brandeburg 已提交
1835

S
Shannon Nelson 已提交
1836
			/* find the next higher power-of-2 of num queue pairs */
N
Neerav Parikh 已提交
1837
			num_qps = qcount;
J
Jesse Brandeburg 已提交
1838
			pow = 0;
1839
			while (num_qps && (BIT_ULL(pow) < qcount)) {
J
Jesse Brandeburg 已提交
1840 1841 1842 1843 1844 1845 1846 1847 1848
				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
1849
			offset += qcount;
J
Jesse Brandeburg 已提交
1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865
		} 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;
1866 1867 1868
	if ((vsi->type == I40E_VSI_MAIN) && (numtc == 1)) {
		if (vsi->req_queue_pairs > 0)
			vsi->num_queue_pairs = vsi->req_queue_pairs;
1869
		else if (pf->flags & I40E_FLAG_MSIX_ENABLED)
1870 1871
			vsi->num_queue_pairs = pf->num_lan_msix;
	}
J
Jesse Brandeburg 已提交
1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892

	/* 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);
}

1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905
/**
 * 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;

1906
	if (i40e_add_mac_filter(vsi, addr))
1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924
		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;

1925 1926 1927 1928 1929 1930 1931 1932
	/* Under some circumstances, we might receive a request to delete
	 * our own device address from our uc list. Because we store the
	 * device address in the VSI's MAC/VLAN filter list, we need to ignore
	 * such requests and not delete our device address from this list.
	 */
	if (ether_addr_equal(addr, netdev->dev_addr))
		return 0;

1933
	i40e_del_mac_filter(vsi, addr);
1934 1935 1936 1937

	return 0;
}

J
Jesse Brandeburg 已提交
1938 1939 1940 1941 1942 1943 1944 1945 1946
/**
 * 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;

1947
	spin_lock_bh(&vsi->mac_filter_hash_lock);
K
Kiran Patil 已提交
1948

1949 1950
	__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
1951

1952
	spin_unlock_bh(&vsi->mac_filter_hash_lock);
J
Jesse Brandeburg 已提交
1953 1954 1955 1956 1957 1958 1959 1960

	/* 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
1961
/**
1962
 * i40e_undo_del_filter_entries - Undo the changes made to MAC filter entries
1963
 * @vsi: Pointer to VSI struct
K
Kiran Patil 已提交
1964 1965 1966
 * @from: Pointer to list which contains MAC filter entries - changes to
 *        those entries needs to be undone.
 *
1967
 * MAC filter entries from this list were slated for deletion.
K
Kiran Patil 已提交
1968
 **/
1969 1970
static void i40e_undo_del_filter_entries(struct i40e_vsi *vsi,
					 struct hlist_head *from)
K
Kiran Patil 已提交
1971
{
1972 1973 1974 1975 1976
	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
1977 1978

		/* Move the element back into MAC filter list*/
1979 1980
		hlist_del(&f->hlist);
		hash_add(vsi->mac_filter_hash, &f->hlist, key);
K
Kiran Patil 已提交
1981 1982 1983
	}
}

1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
/**
 * 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);
	}
}

2005 2006
/**
 * i40e_next_entry - Get the next non-broadcast filter from a list
2007
 * @next: pointer to filter in list
2008 2009 2010 2011 2012
 *
 * 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.
 */
2013 2014
static
struct i40e_new_mac_filter *i40e_next_filter(struct i40e_new_mac_filter *next)
2015
{
2016 2017 2018
	hlist_for_each_entry_continue(next, hlist) {
		if (!is_broadcast_ether_addr(next->f->macaddr))
			return next;
2019 2020
	}

2021
	return NULL;
2022 2023
}

K
Kiran Patil 已提交
2024
/**
2025 2026 2027 2028 2029
 * 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
2030
 *
2031 2032
 * 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
2033
 **/
2034 2035 2036
static int
i40e_update_filter_state(int count,
			 struct i40e_aqc_add_macvlan_element_data *add_list,
2037
			 struct i40e_new_mac_filter *add_head)
K
Kiran Patil 已提交
2038
{
2039 2040
	int retval = 0;
	int i;
K
Kiran Patil 已提交
2041

2042 2043 2044 2045 2046 2047
	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.
2048
		 */
2049
		if (add_list[i].match_method == I40E_AQC_MM_ERR_NO_RES) {
2050
			add_head->state = I40E_FILTER_FAILED;
2051 2052 2053
		} else {
			add_head->state = I40E_FILTER_ACTIVE;
			retval++;
2054
		}
2055

2056 2057 2058
		add_head = i40e_next_filter(add_head);
		if (!add_head)
			break;
K
Kiran Patil 已提交
2059
	}
2060

2061
	return retval;
K
Kiran Patil 已提交
2062 2063
}

2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107
/**
 * 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
 *
 * Send a request to firmware via AdminQ to add a chunk of filters. Will set
2108 2109
 * __I40E_VSI_OVERFLOW_PROMISC bit in vsi->state if the firmware has run out of
 * space for more filters.
2110 2111 2112 2113
 */
static
void i40e_aqc_add_filters(struct i40e_vsi *vsi, const char *vsi_name,
			  struct i40e_aqc_add_macvlan_element_data *list,
2114
			  struct i40e_new_mac_filter *add_head,
2115
			  int num_add)
2116 2117 2118 2119
{
	struct i40e_hw *hw = &vsi->back->hw;
	int aq_err, fcnt;

2120
	i40e_aq_add_macvlan(hw, vsi->seid, list, num_add, NULL);
2121
	aq_err = hw->aq.asq_last_status;
2122
	fcnt = i40e_update_filter_state(num_add, list, add_head);
2123 2124

	if (fcnt != num_add) {
2125
		set_bit(__I40E_VSI_OVERFLOW_PROMISC, vsi->state);
2126 2127 2128 2129 2130 2131 2132
		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);
	}
}

2133 2134 2135 2136 2137 2138 2139 2140
/**
 * 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.
2141 2142
 *
 * Returns status indicating success or failure;
2143
 **/
2144 2145 2146
static i40e_status
i40e_aqc_broadcast_filter(struct i40e_vsi *vsi, const char *vsi_name,
			  struct i40e_mac_filter *f)
2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164
{
	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);
	}

2165 2166
	if (aq_ret) {
		set_bit(__I40E_VSI_OVERFLOW_PROMISC, vsi->state);
2167
		dev_warn(&vsi->back->pdev->dev,
2168
			 "Error %s, forcing overflow promiscuous on %s\n",
2169 2170
			 i40e_aq_str(hw, hw->aq.asq_last_status),
			 vsi_name);
2171
	}
2172 2173

	return aq_ret;
2174 2175
}

A
Alan Brady 已提交
2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240 2241 2242
/**
 * i40e_set_promiscuous - set promiscuous mode
 * @pf: board private structure
 * @promisc: promisc on or off
 *
 * There are different ways of setting promiscuous mode on a PF depending on
 * what state/environment we're in.  This identifies and sets it appropriately.
 * Returns 0 on success.
 **/
static int i40e_set_promiscuous(struct i40e_pf *pf, bool promisc)
{
	struct i40e_vsi *vsi = pf->vsi[pf->lan_vsi];
	struct i40e_hw *hw = &pf->hw;
	i40e_status aq_ret;

	if (vsi->type == I40E_VSI_MAIN &&
	    pf->lan_veb != I40E_NO_VEB &&
	    !(pf->flags & I40E_FLAG_MFP_ENABLED)) {
		/* 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 (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) {
			dev_info(&pf->pdev->dev,
				 "Set default VSI failed, err %s, aq_err %s\n",
				 i40e_stat_str(hw, aq_ret),
				 i40e_aq_str(hw, hw->aq.asq_last_status));
		}
	} else {
		aq_ret = i40e_aq_set_vsi_unicast_promiscuous(
						  hw,
						  vsi->seid,
						  promisc, NULL,
						  true);
		if (aq_ret) {
			dev_info(&pf->pdev->dev,
				 "set unicast promisc failed, err %s, aq_err %s\n",
				 i40e_stat_str(hw, aq_ret),
				 i40e_aq_str(hw, hw->aq.asq_last_status));
		}
		aq_ret = i40e_aq_set_vsi_multicast_promiscuous(
						  hw,
						  vsi->seid,
						  promisc, NULL);
		if (aq_ret) {
			dev_info(&pf->pdev->dev,
				 "set multicast promisc failed, err %s, aq_err %s\n",
				 i40e_stat_str(hw, aq_ret),
				 i40e_aq_str(hw, hw->aq.asq_last_status));
		}
	}

	if (!aq_ret)
		pf->cur_promisc = promisc;

	return aq_ret;
}

J
Jesse Brandeburg 已提交
2243 2244 2245 2246 2247 2248 2249 2250
/**
 * 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
2251
int i40e_sync_vsi_filters(struct i40e_vsi *vsi)
J
Jesse Brandeburg 已提交
2252
{
2253
	struct hlist_head tmp_add_list, tmp_del_list;
2254 2255
	struct i40e_mac_filter *f;
	struct i40e_new_mac_filter *new, *add_head = NULL;
2256
	struct i40e_hw *hw = &vsi->back->hw;
2257
	bool old_overflow, new_overflow;
2258
	unsigned int failed_filters = 0;
2259
	unsigned int vlan_filters = 0;
2260
	char vsi_name[16] = "PF";
J
Jesse Brandeburg 已提交
2261
	int filter_list_len = 0;
M
Mitch Williams 已提交
2262
	i40e_status aq_ret = 0;
2263
	u32 changed_flags = 0;
2264
	struct hlist_node *h;
J
Jesse Brandeburg 已提交
2265 2266 2267
	struct i40e_pf *pf;
	int num_add = 0;
	int num_del = 0;
2268
	int retval = 0;
J
Jesse Brandeburg 已提交
2269
	u16 cmd_flags;
2270
	int list_size;
2271
	int bkt;
J
Jesse Brandeburg 已提交
2272 2273 2274 2275 2276

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

2277
	while (test_and_set_bit(__I40E_VSI_SYNCING_FILTERS, vsi->state))
J
Jesse Brandeburg 已提交
2278 2279 2280
		usleep_range(1000, 2000);
	pf = vsi->back;

2281 2282
	old_overflow = test_bit(__I40E_VSI_OVERFLOW_PROMISC, vsi->state);

J
Jesse Brandeburg 已提交
2283 2284 2285 2286 2287
	if (vsi->netdev) {
		changed_flags = vsi->current_netdev_flags ^ vsi->netdev->flags;
		vsi->current_netdev_flags = vsi->netdev->flags;
	}

2288 2289
	INIT_HLIST_HEAD(&tmp_add_list);
	INIT_HLIST_HEAD(&tmp_del_list);
K
Kiran Patil 已提交
2290

2291 2292 2293 2294 2295
	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
2296 2297 2298
	if (vsi->flags & I40E_VSI_FLAG_FILTER_CHANGED) {
		vsi->flags &= ~I40E_VSI_FLAG_FILTER_CHANGED;

2299
		spin_lock_bh(&vsi->mac_filter_hash_lock);
2300
		/* Create a list of filters to delete. */
2301
		hash_for_each_safe(vsi->mac_filter_hash, bkt, h, f, hlist) {
2302 2303
			if (f->state == I40E_FILTER_REMOVE) {
				/* Move the element into temporary del_list */
2304 2305
				hash_del(&f->hlist);
				hlist_add_head(&f->hlist, &tmp_del_list);
2306 2307 2308

				/* Avoid counting removed filters */
				continue;
2309 2310
			}
			if (f->state == I40E_FILTER_NEW) {
2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321
				/* 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
2322
			}
2323

2324 2325 2326
			/* Count the number of active (current and new) VLAN
			 * filters we have now. Does not count filters which
			 * are marked for deletion.
2327 2328 2329 2330 2331
			 */
			if (f->vlan > 0)
				vlan_filters++;
		}

2332 2333 2334 2335 2336 2337
		retval = i40e_correct_mac_vlan_filters(vsi,
						       &tmp_add_list,
						       &tmp_del_list,
						       vlan_filters);
		if (retval)
			goto err_no_memory_locked;
2338

2339
		spin_unlock_bh(&vsi->mac_filter_hash_lock);
K
Kiran Patil 已提交
2340 2341 2342
	}

	/* Now process 'del_list' outside the lock */
2343
	if (!hlist_empty(&tmp_del_list)) {
2344
		filter_list_len = hw->aq.asq_buf_size /
K
Kiran Patil 已提交
2345
			    sizeof(struct i40e_aqc_remove_macvlan_element_data);
2346
		list_size = filter_list_len *
S
Shannon Nelson 已提交
2347
			    sizeof(struct i40e_aqc_remove_macvlan_element_data);
2348
		del_list = kzalloc(list_size, GFP_ATOMIC);
2349 2350
		if (!del_list)
			goto err_no_memory;
K
Kiran Patil 已提交
2351

2352
		hlist_for_each_entry_safe(f, h, &tmp_del_list, hlist) {
J
Jesse Brandeburg 已提交
2353 2354
			cmd_flags = 0;

2355
			/* handle broadcast filters by updating the broadcast
2356
			 * promiscuous flag and release filter list.
2357 2358 2359 2360 2361 2362 2363 2364 2365
			 */
			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
2366
			/* add to delete list */
2367
			ether_addr_copy(del_list[num_del].mac_addr, f->macaddr);
2368 2369
			if (f->vlan == I40E_VLAN_ANY) {
				del_list[num_del].vlan_tag = 0;
A
Alan Brady 已提交
2370
				cmd_flags |= I40E_AQC_MACVLAN_DEL_IGNORE_VLAN;
2371 2372 2373 2374
			} else {
				del_list[num_del].vlan_tag =
					cpu_to_le16((u16)(f->vlan));
			}
J
Jesse Brandeburg 已提交
2375 2376 2377 2378 2379 2380 2381

			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) {
2382 2383
				i40e_aqc_del_filters(vsi, vsi_name, del_list,
						     num_del, &retval);
2384
				memset(del_list, 0, list_size);
2385
				num_del = 0;
J
Jesse Brandeburg 已提交
2386
			}
K
Kiran Patil 已提交
2387 2388 2389
			/* Release memory for MAC filter entries which were
			 * synced up with HW.
			 */
2390
			hlist_del(&f->hlist);
K
Kiran Patil 已提交
2391
			kfree(f);
J
Jesse Brandeburg 已提交
2392
		}
K
Kiran Patil 已提交
2393

J
Jesse Brandeburg 已提交
2394
		if (num_del) {
2395 2396
			i40e_aqc_del_filters(vsi, vsi_name, del_list,
					     num_del, &retval);
J
Jesse Brandeburg 已提交
2397 2398 2399 2400
		}

		kfree(del_list);
		del_list = NULL;
K
Kiran Patil 已提交
2401 2402
	}

2403
	if (!hlist_empty(&tmp_add_list)) {
2404
		/* Do all the adds now. */
2405
		filter_list_len = hw->aq.asq_buf_size /
S
Shannon Nelson 已提交
2406
			       sizeof(struct i40e_aqc_add_macvlan_element_data);
2407 2408 2409
		list_size = filter_list_len *
			       sizeof(struct i40e_aqc_add_macvlan_element_data);
		add_list = kzalloc(list_size, GFP_ATOMIC);
2410 2411 2412
		if (!add_list)
			goto err_no_memory;

2413
		num_add = 0;
2414
		hlist_for_each_entry_safe(new, h, &tmp_add_list, hlist) {
2415 2416 2417
			/* handle broadcast filters by updating the broadcast
			 * promiscuous flag instead of adding a MAC filter.
			 */
2418 2419 2420 2421 2422 2423
			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;
2424 2425 2426
				continue;
			}

J
Jesse Brandeburg 已提交
2427
			/* add to add array */
2428
			if (num_add == 0)
2429
				add_head = new;
2430
			cmd_flags = 0;
2431 2432 2433
			ether_addr_copy(add_list[num_add].mac_addr,
					new->f->macaddr);
			if (new->f->vlan == I40E_VLAN_ANY) {
2434 2435 2436 2437
				add_list[num_add].vlan_tag = 0;
				cmd_flags |= I40E_AQC_MACVLAN_ADD_IGNORE_VLAN;
			} else {
				add_list[num_add].vlan_tag =
2438
					cpu_to_le16((u16)(new->f->vlan));
2439
			}
J
Jesse Brandeburg 已提交
2440
			add_list[num_add].queue_number = 0;
2441
			/* set invalid match method for later detection */
2442
			add_list[num_add].match_method = I40E_AQC_MM_ERR_NO_RES;
J
Jesse Brandeburg 已提交
2443 2444 2445 2446 2447 2448
			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) {
2449
				i40e_aqc_add_filters(vsi, vsi_name, add_list,
2450
						     add_head, num_add);
2451
				memset(add_list, 0, list_size);
J
Jesse Brandeburg 已提交
2452 2453 2454 2455
				num_add = 0;
			}
		}
		if (num_add) {
2456
			i40e_aqc_add_filters(vsi, vsi_name, add_list, add_head,
2457
					     num_add);
J
Jesse Brandeburg 已提交
2458
		}
2459 2460 2461
		/* Now move all of the filters from the temp add list back to
		 * the VSI's list.
		 */
2462
		spin_lock_bh(&vsi->mac_filter_hash_lock);
2463 2464 2465 2466 2467 2468
		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);
2469
		}
2470
		spin_unlock_bh(&vsi->mac_filter_hash_lock);
J
Jesse Brandeburg 已提交
2471 2472
		kfree(add_list);
		add_list = NULL;
2473
	}
J
Jesse Brandeburg 已提交
2474

2475 2476 2477 2478 2479 2480 2481 2482 2483 2484 2485 2486 2487 2488 2489
	/* 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);

	/* 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.
	 */
2490 2491
	if (old_overflow && !failed_filters &&
	    vsi->active_filters < vsi->promisc_threshold) {
2492 2493 2494
		dev_info(&pf->pdev->dev,
			 "filter logjam cleared on %s, leaving overflow promiscuous mode\n",
			 vsi_name);
2495
		clear_bit(__I40E_VSI_OVERFLOW_PROMISC, vsi->state);
2496
		vsi->promisc_threshold = 0;
J
Jesse Brandeburg 已提交
2497 2498
	}

2499 2500
	/* if the VF is not trusted do not do promisc */
	if ((vsi->type == I40E_VSI_SRIOV) && !pf->vf[vsi->vf_id].trusted) {
2501
		clear_bit(__I40E_VSI_OVERFLOW_PROMISC, vsi->state);
2502 2503 2504
		goto out;
	}

2505 2506 2507 2508 2509 2510 2511 2512
	new_overflow = test_bit(__I40E_VSI_OVERFLOW_PROMISC, vsi->state);

	/* If we are entering overflow promiscuous, we need to calculate a new
	 * threshold for when we are safe to exit
	 */
	if (!old_overflow && new_overflow)
		vsi->promisc_threshold = (vsi->active_filters * 3) / 4;

J
Jesse Brandeburg 已提交
2513 2514 2515
	/* check for changes in promiscuous modes */
	if (changed_flags & IFF_ALLMULTI) {
		bool cur_multipromisc;
J
Jesse Brandeburg 已提交
2516

J
Jesse Brandeburg 已提交
2517
		cur_multipromisc = !!(vsi->current_netdev_flags & IFF_ALLMULTI);
M
Mitch Williams 已提交
2518 2519 2520 2521 2522 2523
		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,
2524
						     hw->aq.asq_last_status);
J
Jesse Brandeburg 已提交
2525
			dev_info(&pf->pdev->dev,
2526 2527
				 "set multi promisc failed on %s, err %s aq_err %s\n",
				 vsi_name,
2528 2529
				 i40e_stat_str(hw, aq_ret),
				 i40e_aq_str(hw, hw->aq.asq_last_status));
M
Mitch Williams 已提交
2530
		}
J
Jesse Brandeburg 已提交
2531
	}
2532

2533
	if ((changed_flags & IFF_PROMISC) || old_overflow != new_overflow) {
J
Jesse Brandeburg 已提交
2534
		bool cur_promisc;
J
Jesse Brandeburg 已提交
2535

J
Jesse Brandeburg 已提交
2536
		cur_promisc = (!!(vsi->current_netdev_flags & IFF_PROMISC) ||
2537
			       new_overflow);
A
Alan Brady 已提交
2538
		aq_ret = i40e_set_promiscuous(pf, cur_promisc);
M
Mitch Williams 已提交
2539 2540
		if (aq_ret) {
			retval = i40e_aq_rc_to_posix(aq_ret,
A
Alan Brady 已提交
2541
						     hw->aq.asq_last_status);
2542
			dev_info(&pf->pdev->dev,
A
Alan Brady 已提交
2543 2544 2545 2546 2547
				 "Setting promiscuous %s failed on %s, err %s aq_err %s\n",
				 cur_promisc ? "on" : "off",
				 vsi_name,
				 i40e_stat_str(hw, aq_ret),
				 i40e_aq_str(hw, hw->aq.asq_last_status));
M
Mitch Williams 已提交
2548
		}
J
Jesse Brandeburg 已提交
2549
	}
M
Mitch Williams 已提交
2550
out:
2551 2552 2553 2554
	/* if something went wrong then set the changed flag so we try again */
	if (retval)
		vsi->flags |= I40E_VSI_FLAG_FILTER_CHANGED;

2555
	clear_bit(__I40E_VSI_SYNCING_FILTERS, vsi->state);
M
Mitch Williams 已提交
2556
	return retval;
2557 2558 2559 2560

err_no_memory:
	/* Restore elements on the temporary add and delete lists */
	spin_lock_bh(&vsi->mac_filter_hash_lock);
2561
err_no_memory_locked:
2562 2563
	i40e_undo_del_filter_entries(vsi, &tmp_del_list);
	i40e_undo_add_filter_entries(vsi, &tmp_add_list);
2564 2565 2566
	spin_unlock_bh(&vsi->mac_filter_hash_lock);

	vsi->flags |= I40E_VSI_FLAG_FILTER_CHANGED;
2567
	clear_bit(__I40E_VSI_SYNCING_FILTERS, vsi->state);
2568
	return -ENOMEM;
J
Jesse Brandeburg 已提交
2569 2570 2571 2572 2573 2574 2575 2576 2577 2578 2579 2580 2581 2582
}

/**
 * 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
2583
	for (v = 0; v < pf->num_alloc_vsi; v++) {
J
Jesse Brandeburg 已提交
2584
		if (pf->vsi[v] &&
J
Jesse Brandeburg 已提交
2585 2586 2587 2588 2589 2590 2591 2592 2593
		    (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
2594 2595 2596
	}
}

2597 2598 2599 2600 2601 2602 2603 2604 2605 2606 2607 2608
/**
 * 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
2609 2610 2611 2612 2613 2614 2615 2616 2617 2618 2619
/**
 * 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
2620
	struct i40e_pf *pf = vsi->back;
J
Jesse Brandeburg 已提交
2621

2622 2623 2624 2625 2626 2627 2628
	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
2629 2630 2631 2632 2633
	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
2634 2635
	pf->flags |= (I40E_FLAG_SERVICE_CLIENT_REQUESTED |
		      I40E_FLAG_CLIENT_L2_CHANGE);
J
Jesse Brandeburg 已提交
2636 2637 2638
	return 0;
}

J
Jacob Keller 已提交
2639 2640 2641 2642 2643 2644 2645 2646 2647 2648 2649 2650 2651 2652 2653 2654 2655 2656 2657 2658 2659
/**
 * 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
2660 2661 2662 2663 2664 2665 2666 2667 2668 2669 2670 2671 2672 2673 2674 2675 2676 2677 2678
/**
 * 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;
2679
	ctxt.info = vsi->info;
J
Jesse Brandeburg 已提交
2680 2681 2682
	ret = i40e_aq_update_vsi_params(&vsi->back->hw, &ctxt, NULL);
	if (ret) {
		dev_info(&vsi->back->pdev->dev,
2683 2684 2685 2686
			 "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
2687 2688 2689 2690 2691 2692 2693 2694 2695 2696 2697 2698 2699 2700 2701 2702 2703 2704 2705 2706 2707 2708 2709
	}
}

/**
 * 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;
2710
	ctxt.info = vsi->info;
J
Jesse Brandeburg 已提交
2711 2712 2713
	ret = i40e_aq_update_vsi_params(&vsi->back->hw, &ctxt, NULL);
	if (ret) {
		dev_info(&vsi->back->pdev->dev,
2714 2715 2716 2717
			 "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
2718 2719 2720 2721 2722 2723 2724 2725 2726 2727 2728 2729 2730 2731 2732 2733 2734 2735 2736 2737
	}
}

/**
 * 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);
}

/**
2738
 * i40e_add_vlan_all_mac - Add a MAC/VLAN filter for each existing MAC address
J
Jesse Brandeburg 已提交
2739 2740
 * @vsi: the vsi being configured
 * @vid: vlan id to be added (0 = untagged only , -1 = any)
2741 2742 2743 2744 2745 2746 2747 2748
 *
 * 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
2749
 **/
2750
int i40e_add_vlan_all_mac(struct i40e_vsi *vsi, s16 vid)
J
Jesse Brandeburg 已提交
2751
{
2752
	struct i40e_mac_filter *f, *add_f;
2753 2754
	struct hlist_node *h;
	int bkt;
J
Jesse Brandeburg 已提交
2755

2756
	hash_for_each_safe(vsi->mac_filter_hash, bkt, h, f, hlist) {
2757 2758
		if (f->state == I40E_FILTER_REMOVE)
			continue;
2759
		add_f = i40e_add_filter(vsi, f->macaddr, vid);
J
Jesse Brandeburg 已提交
2760 2761 2762 2763 2764 2765 2766 2767
		if (!add_f) {
			dev_info(&vsi->back->pdev->dev,
				 "Could not add vlan filter %d for %pM\n",
				 vid, f->macaddr);
			return -ENOMEM;
		}
	}

2768 2769 2770 2771 2772 2773
	return 0;
}

/**
 * i40e_vsi_add_vlan - Add VSI membership for given VLAN
 * @vsi: the VSI being configured
2774
 * @vid: VLAN id to be added
2775
 **/
2776
int i40e_vsi_add_vlan(struct i40e_vsi *vsi, u16 vid)
2777
{
2778
	int err;
2779

2780
	if (vsi->info.pvid)
2781 2782
		return -EINVAL;

2783 2784 2785 2786 2787 2788 2789 2790 2791 2792 2793
	/* 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;

2794 2795 2796
	/* 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);
2797
	spin_unlock_bh(&vsi->mac_filter_hash_lock);
2798 2799
	if (err)
		return err;
K
Kiran Patil 已提交
2800

2801 2802 2803 2804 2805
	/* 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
2806 2807 2808
}

/**
2809
 * i40e_rm_vlan_all_mac - Remove MAC/VLAN pair for all MAC with the given VLAN
J
Jesse Brandeburg 已提交
2810 2811
 * @vsi: the vsi being configured
 * @vid: vlan id to be removed (0 = untagged only , -1 = any)
2812 2813 2814 2815 2816 2817 2818 2819 2820
 *
 * 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
 */
2821
void i40e_rm_vlan_all_mac(struct i40e_vsi *vsi, s16 vid)
J
Jesse Brandeburg 已提交
2822
{
2823
	struct i40e_mac_filter *f;
2824 2825
	struct hlist_node *h;
	int bkt;
J
Jesse Brandeburg 已提交
2826

2827
	hash_for_each_safe(vsi->mac_filter_hash, bkt, h, f, hlist) {
2828 2829 2830
		if (f->vlan == vid)
			__i40e_del_filter(vsi, f);
	}
2831
}
J
Jesse Brandeburg 已提交
2832

2833 2834 2835
/**
 * i40e_vsi_kill_vlan - Remove VSI membership for given VLAN
 * @vsi: the VSI being configured
2836
 * @vid: VLAN id to be removed
2837
 **/
2838
void i40e_vsi_kill_vlan(struct i40e_vsi *vsi, u16 vid)
2839
{
2840 2841 2842
	if (!vid || vsi->info.pvid)
		return;

2843 2844
	spin_lock_bh(&vsi->mac_filter_hash_lock);
	i40e_rm_vlan_all_mac(vsi, vid);
2845
	spin_unlock_bh(&vsi->mac_filter_hash_lock);
K
Kiran Patil 已提交
2846

2847 2848 2849 2850
	/* 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
2851 2852 2853 2854 2855 2856
}

/**
 * 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
2857 2858
 *
 * net_device_ops implementation for adding vlan ids
J
Jesse Brandeburg 已提交
2859 2860 2861 2862 2863 2864
 **/
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
2865
	int ret = 0;
J
Jesse Brandeburg 已提交
2866

2867
	if (vid >= VLAN_N_VID)
J
Jesse Brandeburg 已提交
2868 2869
		return -EINVAL;

2870
	ret = i40e_vsi_add_vlan(vsi, vid);
2871
	if (!ret)
J
Jesse Brandeburg 已提交
2872
		set_bit(vid, vsi->active_vlans);
J
Jesse Brandeburg 已提交
2873

J
Jesse Brandeburg 已提交
2874
	return ret;
J
Jesse Brandeburg 已提交
2875 2876 2877 2878 2879 2880
}

/**
 * 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
2881
 *
A
Akeem G Abodunrin 已提交
2882
 * net_device_ops implementation for removing vlan ids
J
Jesse Brandeburg 已提交
2883 2884 2885 2886 2887 2888 2889 2890 2891
 **/
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
2892
	 * already printed from the other function
J
Jesse Brandeburg 已提交
2893 2894 2895 2896
	 */
	i40e_vsi_kill_vlan(vsi, vid);

	clear_bit(vid, vsi->active_vlans);
J
Jesse Brandeburg 已提交
2897

J
Jesse Brandeburg 已提交
2898 2899 2900 2901 2902 2903 2904 2905 2906 2907 2908 2909 2910 2911 2912 2913 2914 2915 2916 2917 2918 2919 2920 2921 2922 2923
	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
2924
int i40e_vsi_add_pvid(struct i40e_vsi *vsi, u16 vid)
J
Jesse Brandeburg 已提交
2925 2926
{
	struct i40e_vsi_context ctxt;
2927
	i40e_status ret;
J
Jesse Brandeburg 已提交
2928 2929 2930

	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
2931 2932
	vsi->info.port_vlan_flags = I40E_AQ_VSI_PVLAN_MODE_TAGGED |
				    I40E_AQ_VSI_PVLAN_INSERT_PVID |
G
Greg Rose 已提交
2933
				    I40E_AQ_VSI_PVLAN_EMOD_STR;
J
Jesse Brandeburg 已提交
2934 2935

	ctxt.seid = vsi->seid;
2936
	ctxt.info = vsi->info;
2937 2938
	ret = i40e_aq_update_vsi_params(&vsi->back->hw, &ctxt, NULL);
	if (ret) {
J
Jesse Brandeburg 已提交
2939
		dev_info(&vsi->back->pdev->dev,
2940 2941 2942 2943
			 "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
2944
		return -ENOENT;
J
Jesse Brandeburg 已提交
2945 2946
	}

J
Jesse Brandeburg 已提交
2947
	return 0;
J
Jesse Brandeburg 已提交
2948 2949 2950 2951 2952 2953 2954 2955 2956 2957
}

/**
 * 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
2958 2959
	i40e_vlan_stripping_disable(vsi);

J
Jesse Brandeburg 已提交
2960 2961 2962 2963 2964 2965 2966 2967 2968 2969 2970 2971 2972 2973 2974 2975 2976 2977
	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++)
2978
		err = i40e_setup_tx_descriptors(vsi->tx_rings[i]);
J
Jesse Brandeburg 已提交
2979

2980 2981 2982 2983 2984 2985
	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
2986 2987 2988 2989 2990 2991 2992 2993 2994 2995 2996 2997 2998
	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;

2999 3000 3001 3002 3003
	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]);
	}
3004

3005 3006 3007 3008 3009
	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
3010 3011 3012 3013 3014 3015 3016 3017 3018 3019 3020 3021 3022 3023 3024 3025 3026
}

/**
 * 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++)
3027
		err = i40e_setup_rx_descriptors(vsi->rx_rings[i]);
J
Jesse Brandeburg 已提交
3028 3029 3030 3031 3032 3033 3034 3035 3036 3037 3038 3039 3040
	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;

3041 3042 3043
	if (!vsi->rx_rings)
		return;

J
Jesse Brandeburg 已提交
3044
	for (i = 0; i < vsi->num_queue_pairs; i++)
3045
		if (vsi->rx_rings[i] && vsi->rx_rings[i]->desc)
3046
			i40e_free_rx_resources(vsi->rx_rings[i]);
J
Jesse Brandeburg 已提交
3047 3048
}

3049 3050 3051 3052 3053 3054 3055 3056 3057
/**
 * 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)
{
3058
	int cpu;
3059

3060
	if (!ring->q_vector || !ring->netdev || ring->ch)
3061 3062
		return;

J
Jacob Keller 已提交
3063 3064 3065
	/* We only initialize XPS once, so as not to overwrite user settings */
	if (test_and_set_bit(__I40E_TX_XPS_INIT_DONE, ring->state))
		return;
3066

J
Jacob Keller 已提交
3067 3068 3069
	cpu = cpumask_local_spread(ring->q_vector->v_idx, -1);
	netif_set_xps_queue(ring->netdev, get_cpu_mask(cpu),
			    ring->queue_index);
3070 3071
}

J
Jesse Brandeburg 已提交
3072 3073 3074 3075 3076 3077 3078 3079 3080 3081 3082 3083 3084 3085 3086 3087
/**
 * 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
3088
	if (vsi->back->flags & I40E_FLAG_FD_ATR_ENABLED) {
J
Jesse Brandeburg 已提交
3089 3090 3091 3092 3093 3094
		ring->atr_sample_rate = vsi->back->atr_sample_rate;
		ring->atr_count = 0;
	} else {
		ring->atr_sample_rate = 0;
	}

3095 3096
	/* configure XPS */
	i40e_config_xps_tx_ring(ring);
J
Jesse Brandeburg 已提交
3097 3098 3099 3100 3101 3102 3103

	/* 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
3104 3105
	tx_ctx.fd_ena = !!(vsi->back->flags & (I40E_FLAG_FD_SB_ENABLED |
					       I40E_FLAG_FD_ATR_ENABLED));
J
Jacob Keller 已提交
3106
	tx_ctx.timesync_ena = !!(vsi->back->flags & I40E_FLAG_PTP);
3107 3108 3109 3110 3111
	/* 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
3112 3113 3114 3115 3116 3117 3118 3119 3120 3121 3122

	/* 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.
	 */
3123 3124 3125 3126 3127 3128 3129 3130

	if (ring->ch)
		tx_ctx.rdylist =
			le16_to_cpu(ring->ch->info.qs_handle[ring->dcb_tc]);

	else
		tx_ctx.rdylist = le16_to_cpu(vsi->info.qs_handle[ring->dcb_tc]);

J
Jesse Brandeburg 已提交
3131 3132 3133 3134 3135 3136 3137 3138 3139 3140 3141 3142 3143 3144 3145 3146 3147 3148 3149 3150 3151
	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 */
3152 3153 3154 3155 3156 3157 3158 3159 3160
	if (ring->ch) {
		if (ring->ch->type == I40E_VSI_VMDQ2)
			qtx_ctl = I40E_QTX_CTL_VM_QUEUE;
		else
			return -EINVAL;

		qtx_ctl |= (ring->ch->vsi_number <<
			    I40E_QTX_CTL_VFVM_INDX_SHIFT) &
			    I40E_QTX_CTL_VFVM_INDX_MASK;
M
Mitch Williams 已提交
3161
	} else {
3162 3163 3164 3165 3166 3167 3168
		if (vsi->type == I40E_VSI_VMDQ2) {
			qtx_ctl = I40E_QTX_CTL_VM_QUEUE;
			qtx_ctl |= ((vsi->id) << I40E_QTX_CTL_VFVM_INDX_SHIFT) &
				    I40E_QTX_CTL_VFVM_INDX_MASK;
		} else {
			qtx_ctl = I40E_QTX_CTL_PF_QUEUE;
		}
M
Mitch Williams 已提交
3169 3170
	}

3171 3172
	qtx_ctl |= ((hw->pf_id << I40E_QTX_CTL_PF_INDX_SHIFT) &
		    I40E_QTX_CTL_PF_INDX_MASK);
J
Jesse Brandeburg 已提交
3173 3174 3175 3176 3177 3178 3179 3180 3181 3182 3183 3184 3185 3186 3187 3188 3189 3190 3191 3192 3193 3194 3195 3196
	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;

3197
	bitmap_zero(ring->state, __I40E_RING_STATE_NBITS);
J
Jesse Brandeburg 已提交
3198 3199 3200 3201 3202 3203

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

	ring->rx_buf_len = vsi->rx_buf_len;

3204 3205
	rx_ctx.dbuff = DIV_ROUND_UP(ring->rx_buf_len,
				    BIT_ULL(I40E_RXQ_CTX_DBUFF_SHIFT));
J
Jesse Brandeburg 已提交
3206 3207 3208 3209

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

3210 3211
	/* use 32 byte descriptors */
	rx_ctx.dsize = 1;
J
Jesse Brandeburg 已提交
3212

3213 3214 3215
	/* descriptor type is always zero
	 * rx_ctx.dtype = 0;
	 */
3216
	rx_ctx.hsplit_0 = 0;
J
Jesse Brandeburg 已提交
3217

3218
	rx_ctx.rxmax = min_t(u16, vsi->max_frame, chain_len * ring->rx_buf_len);
3219 3220 3221
	if (hw->revision_id == 0)
		rx_ctx.lrxqthresh = 0;
	else
J
Jacob Keller 已提交
3222
		rx_ctx.lrxqthresh = 1;
J
Jesse Brandeburg 已提交
3223 3224
	rx_ctx.crcstrip = 1;
	rx_ctx.l2tsel = 1;
J
Jesse Brandeburg 已提交
3225 3226
	/* this controls whether VLAN is stripped from inner headers */
	rx_ctx.showiv = 0;
3227 3228
	/* set the prefena field to 1 because the manual says to */
	rx_ctx.prefena = 1;
J
Jesse Brandeburg 已提交
3229 3230 3231 3232 3233 3234 3235 3236 3237 3238 3239 3240 3241 3242 3243 3244 3245 3246 3247

	/* 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;
	}

3248 3249 3250 3251 3252 3253
	/* 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
3254 3255 3256 3257
	/* 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);

3258
	i40e_alloc_rx_buffers(ring, I40E_DESC_UNUSED(ring));
J
Jesse Brandeburg 已提交
3259 3260 3261 3262 3263 3264 3265 3266 3267 3268 3269 3270 3271 3272 3273

	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;

3274 3275
	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
3276

3277 3278 3279 3280 3281 3282
	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
3283 3284 3285 3286 3287 3288 3289 3290 3291 3292 3293 3294 3295 3296
	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;

3297 3298 3299 3300
	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)
3301 3302
	} else if (!I40E_2K_TOO_SMALL_WITH_PADDING &&
		   (vsi->netdev->mtu <= ETH_DATA_LEN)) {
3303 3304 3305 3306 3307
		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;
3308 3309
		vsi->rx_buf_len = (PAGE_SIZE < 8192) ? I40E_RXBUFFER_3072 :
						       I40E_RXBUFFER_2048;
3310
	}
J
Jesse Brandeburg 已提交
3311 3312 3313

	/* set up individual rings */
	for (i = 0; i < vsi->num_queue_pairs && !err; i++)
3314
		err = i40e_configure_rx_ring(vsi->rx_rings[i]);
J
Jesse Brandeburg 已提交
3315 3316 3317 3318 3319 3320 3321 3322 3323 3324

	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)
{
3325
	struct i40e_ring *tx_ring, *rx_ring;
J
Jesse Brandeburg 已提交
3326 3327 3328
	u16 qoffset, qcount;
	int i, n;

3329 3330 3331 3332 3333 3334 3335 3336
	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;
		}
3337
		return;
3338
	}
J
Jesse Brandeburg 已提交
3339 3340

	for (n = 0; n < I40E_MAX_TRAFFIC_CLASS; n++) {
3341
		if (!(vsi->tc_config.enabled_tc & BIT_ULL(n)))
J
Jesse Brandeburg 已提交
3342 3343 3344 3345 3346
			continue;

		qoffset = vsi->tc_config.tc_info[n].qoffset;
		qcount = vsi->tc_config.tc_info[n].qcount;
		for (i = qoffset; i < (qoffset + qcount); i++) {
3347 3348
			rx_ring = vsi->rx_rings[i];
			tx_ring = vsi->tx_rings[i];
J
Jesse Brandeburg 已提交
3349 3350 3351 3352 3353 3354 3355 3356 3357 3358 3359 3360 3361 3362 3363 3364
			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);
}

3365 3366 3367 3368 3369 3370 3371 3372 3373 3374 3375 3376 3377
/**
 * 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;

3378 3379 3380
	if (!(pf->flags & I40E_FLAG_FD_SB_ENABLED))
		return;

3381
	/* Reset FDir counters as we're replaying all existing filters */
3382 3383
	pf->fd_tcp4_filter_cnt = 0;
	pf->fd_udp4_filter_cnt = 0;
3384
	pf->fd_sctp4_filter_cnt = 0;
3385
	pf->fd_ip4_filter_cnt = 0;
3386

3387 3388 3389 3390 3391 3392
	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
3393 3394 3395 3396 3397 3398 3399 3400 3401 3402 3403 3404 3405 3406 3407 3408 3409 3410 3411 3412 3413 3414 3415 3416
/**
 * 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)
{
3417
	bool has_xdp = i40e_enabled_xdp_vsi(vsi);
J
Jesse Brandeburg 已提交
3418 3419 3420 3421 3422 3423 3424 3425 3426 3427 3428 3429
	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;
3430
	for (i = 0; i < vsi->num_q_vectors; i++, vector++) {
3431 3432
		struct i40e_q_vector *q_vector = vsi->q_vectors[i];

3433
		q_vector->rx.next_update = jiffies + 1;
3434 3435
		q_vector->rx.target_itr =
			ITR_TO_REG(vsi->rx_rings[i]->itr_setting);
J
Jesse Brandeburg 已提交
3436
		wr32(hw, I40E_PFINT_ITRN(I40E_RX_ITR, vector - 1),
3437 3438
		     q_vector->rx.target_itr);
		q_vector->rx.current_itr = q_vector->rx.target_itr;
3439 3440

		q_vector->tx.next_update = jiffies + 1;
3441 3442
		q_vector->tx.target_itr =
			ITR_TO_REG(vsi->tx_rings[i]->itr_setting);
J
Jesse Brandeburg 已提交
3443
		wr32(hw, I40E_PFINT_ITRN(I40E_TX_ITR, vector - 1),
3444 3445
		     q_vector->tx.target_itr);
		q_vector->tx.current_itr = q_vector->tx.target_itr;
3446

3447
		wr32(hw, I40E_PFINT_RATEN(vector - 1),
3448
		     i40e_intrl_usec_to_reg(vsi->int_rate_limit));
J
Jesse Brandeburg 已提交
3449 3450 3451 3452

		/* 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++) {
3453
			u32 nextqp = has_xdp ? qp + vsi->alloc_queue_pairs : qp;
3454 3455
			u32 val;

J
Jesse Brandeburg 已提交
3456
			val = I40E_QINT_RQCTL_CAUSE_ENA_MASK |
3457 3458 3459 3460 3461
			      (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
3462 3463 3464

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

3465 3466 3467 3468 3469 3470 3471 3472 3473 3474 3475
			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
3476
			val = I40E_QINT_TQCTL_CAUSE_ENA_MASK |
3477 3478 3479 3480 3481
			      (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
3482 3483 3484

			/* Terminate the linked list */
			if (q == (q_vector->num_ringpairs - 1))
3485 3486
				val |= (I40E_QUEUE_END_OF_LIST <<
					I40E_QINT_TQCTL_NEXTQ_INDX_SHIFT);
J
Jesse Brandeburg 已提交
3487 3488 3489 3490 3491 3492 3493 3494 3495 3496 3497 3498 3499

			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
 **/
3500
static void i40e_enable_misc_int_causes(struct i40e_pf *pf)
J
Jesse Brandeburg 已提交
3501
{
3502
	struct i40e_hw *hw = &pf->hw;
J
Jesse Brandeburg 已提交
3503 3504 3505 3506 3507 3508 3509 3510 3511 3512 3513 3514 3515 3516 3517
	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;

3518 3519 3520
	if (pf->flags & I40E_FLAG_IWARP_ENABLED)
		val |= I40E_PFINT_ICR0_ENA_PE_CRITERR_MASK;

3521 3522 3523
	if (pf->flags & I40E_FLAG_PTP)
		val |= I40E_PFINT_ICR0_ENA_TIMESYNC_MASK;

J
Jesse Brandeburg 已提交
3524 3525 3526
	wr32(hw, I40E_PFINT_ICR0_ENA, val);

	/* SW_ITR_IDX = 0, but don't change INTENA */
3527 3528
	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
3529 3530 3531 3532 3533 3534 3535 3536 3537 3538 3539

	/* 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)
{
3540
	u32 nextqp = i40e_enabled_xdp_vsi(vsi) ? vsi->alloc_queue_pairs : 0;
3541
	struct i40e_q_vector *q_vector = vsi->q_vectors[0];
J
Jesse Brandeburg 已提交
3542 3543 3544 3545 3546
	struct i40e_pf *pf = vsi->back;
	struct i40e_hw *hw = &pf->hw;
	u32 val;

	/* set the ITR configuration */
3547
	q_vector->rx.next_update = jiffies + 1;
3548 3549 3550
	q_vector->rx.target_itr = ITR_TO_REG(vsi->rx_rings[0]->itr_setting);
	wr32(hw, I40E_PFINT_ITR0(I40E_RX_ITR), q_vector->rx.target_itr);
	q_vector->rx.current_itr = q_vector->rx.target_itr;
3551
	q_vector->tx.next_update = jiffies + 1;
3552 3553 3554
	q_vector->tx.target_itr = ITR_TO_REG(vsi->tx_rings[0]->itr_setting);
	wr32(hw, I40E_PFINT_ITR0(I40E_TX_ITR), q_vector->tx.target_itr);
	q_vector->tx.current_itr = q_vector->tx.target_itr;
J
Jesse Brandeburg 已提交
3555

3556
	i40e_enable_misc_int_causes(pf);
J
Jesse Brandeburg 已提交
3557 3558 3559 3560

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

3561
	/* Associate the queue pair to the vector and enable the queue int */
3562 3563 3564
	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
3565 3566 3567 3568
	      (I40E_QUEUE_TYPE_TX << I40E_QINT_TQCTL_NEXTQ_TYPE_SHIFT);

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

3569 3570 3571 3572 3573 3574 3575 3576 3577
	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
3578 3579 3580 3581 3582 3583 3584 3585
	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);
}

3586 3587 3588 3589 3590 3591 3592 3593 3594 3595 3596 3597 3598
/**
 * 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
3599 3600 3601 3602
/**
 * i40e_irq_dynamic_enable_icr0 - Enable default interrupt generation for icr0
 * @pf: board private structure
 **/
3603
void i40e_irq_dynamic_enable_icr0(struct i40e_pf *pf)
J
Jesse Brandeburg 已提交
3604 3605 3606 3607 3608
{
	struct i40e_hw *hw = &pf->hw;
	u32 val;

	val = I40E_PFINT_DYN_CTL0_INTENA_MASK   |
3609
	      I40E_PFINT_DYN_CTL0_CLEARPBA_MASK |
J
Jesse Brandeburg 已提交
3610 3611 3612 3613 3614 3615 3616 3617 3618 3619 3620 3621 3622 3623 3624
	      (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;

3625
	if (!q_vector->tx.ring && !q_vector->rx.ring)
J
Jesse Brandeburg 已提交
3626 3627
		return IRQ_HANDLED;

3628
	napi_schedule_irqoff(&q_vector->napi);
J
Jesse Brandeburg 已提交
3629 3630 3631 3632

	return IRQ_HANDLED;
}

3633 3634 3635 3636 3637 3638 3639 3640 3641 3642 3643 3644 3645 3646
/**
 * 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);

3647
	cpumask_copy(&q_vector->affinity_mask, mask);
3648 3649 3650 3651 3652 3653 3654 3655 3656 3657 3658 3659
}

/**
 * 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
3660 3661 3662 3663 3664 3665 3666 3667 3668 3669 3670 3671 3672 3673 3674
/**
 * 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;
3675
	int irq_num;
3676
	int cpu;
J
Jesse Brandeburg 已提交
3677 3678

	for (vector = 0; vector < q_vectors; vector++) {
3679
		struct i40e_q_vector *q_vector = vsi->q_vectors[vector];
J
Jesse Brandeburg 已提交
3680

3681 3682
		irq_num = pf->msix_entries[base + vector].vector;

3683
		if (q_vector->tx.ring && q_vector->rx.ring) {
J
Jesse Brandeburg 已提交
3684 3685 3686
			snprintf(q_vector->name, sizeof(q_vector->name) - 1,
				 "%s-%s-%d", basename, "TxRx", rx_int_idx++);
			tx_int_idx++;
3687
		} else if (q_vector->rx.ring) {
J
Jesse Brandeburg 已提交
3688 3689
			snprintf(q_vector->name, sizeof(q_vector->name) - 1,
				 "%s-%s-%d", basename, "rx", rx_int_idx++);
3690
		} else if (q_vector->tx.ring) {
J
Jesse Brandeburg 已提交
3691 3692 3693 3694 3695 3696
			snprintf(q_vector->name, sizeof(q_vector->name) - 1,
				 "%s-%s-%d", basename, "tx", tx_int_idx++);
		} else {
			/* skip this unused q_vector */
			continue;
		}
3697
		err = request_irq(irq_num,
J
Jesse Brandeburg 已提交
3698 3699 3700 3701 3702 3703
				  vsi->irq_handler,
				  0,
				  q_vector->name,
				  q_vector);
		if (err) {
			dev_info(&pf->pdev->dev,
S
Shannon Nelson 已提交
3704
				 "MSIX request_irq failed, error: %d\n", err);
J
Jesse Brandeburg 已提交
3705 3706
			goto free_queue_irqs;
		}
3707 3708 3709 3710 3711

		/* 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);
3712 3713 3714 3715 3716
		/* 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.
3717
		 */
3718 3719
		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
3720 3721
	}

3722
	vsi->irqs_ready = true;
J
Jesse Brandeburg 已提交
3723 3724 3725 3726 3727
	return 0;

free_queue_irqs:
	while (vector) {
		vector--;
3728 3729 3730 3731
		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
3732 3733 3734 3735 3736 3737 3738 3739 3740 3741 3742 3743 3744 3745 3746
	}
	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
3747
	/* disable interrupt causation from each queue */
J
Jesse Brandeburg 已提交
3748
	for (i = 0; i < vsi->num_queue_pairs; i++) {
S
Shannon Nelson 已提交
3749 3750 3751 3752 3753 3754 3755 3756 3757 3758
		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);

3759 3760 3761
		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
3762 3763
	}

S
Shannon Nelson 已提交
3764
	/* disable each interrupt */
J
Jesse Brandeburg 已提交
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 3791
	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) {
3792
		for (i = 0; i < vsi->num_q_vectors; i++)
J
Jesse Brandeburg 已提交
3793 3794
			i40e_irq_dynamic_enable(vsi, i);
	} else {
3795
		i40e_irq_dynamic_enable_icr0(pf);
J
Jesse Brandeburg 已提交
3796 3797
	}

3798
	i40e_flush(&pf->hw);
J
Jesse Brandeburg 已提交
3799 3800 3801 3802
	return 0;
}

/**
3803
 * i40e_free_misc_vector - Free the vector that handles non-queue events
J
Jesse Brandeburg 已提交
3804 3805
 * @pf: board private structure
 **/
3806
static void i40e_free_misc_vector(struct i40e_pf *pf)
J
Jesse Brandeburg 已提交
3807 3808 3809 3810
{
	/* Disable ICR 0 */
	wr32(&pf->hw, I40E_PFINT_ICR0_ENA, 0);
	i40e_flush(&pf->hw);
3811 3812 3813 3814 3815 3816

	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
3817 3818 3819 3820 3821 3822 3823 3824 3825 3826 3827 3828 3829 3830 3831
}

/**
 * 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;
3832
	irqreturn_t ret = IRQ_NONE;
J
Jesse Brandeburg 已提交
3833 3834 3835 3836
	u32 icr0, icr0_remaining;
	u32 val, ena_mask;

	icr0 = rd32(hw, I40E_PFINT_ICR0);
3837
	ena_mask = rd32(hw, I40E_PFINT_ICR0_ENA);
J
Jesse Brandeburg 已提交
3838

3839 3840
	/* if sharing a legacy IRQ, we might get called w/o an intr pending */
	if ((icr0 & I40E_PFINT_ICR0_INTEVENT_MASK) == 0)
3841
		goto enable_intr;
J
Jesse Brandeburg 已提交
3842

S
Shannon Nelson 已提交
3843 3844 3845 3846 3847
	/* 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++;

3848
	if ((pf->flags & I40E_FLAG_IWARP_ENABLED) &&
3849
	    (icr0 & I40E_PFINT_ICR0_ENA_PE_CRITERR_MASK)) {
3850
		ena_mask &= ~I40E_PFINT_ICR0_ENA_PE_CRITERR_MASK;
3851
		dev_dbg(&pf->pdev->dev, "cleared PE_CRITERR\n");
3852
		set_bit(__I40E_CORE_RESET_REQUESTED, pf->state);
3853 3854
	}

J
Jesse Brandeburg 已提交
3855 3856
	/* only q0 is used in MSI/Legacy mode, and none are used in MSIX */
	if (icr0 & I40E_PFINT_ICR0_QUEUE_0_MASK) {
3857 3858
		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
3859

3860 3861 3862 3863 3864 3865
		/* 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.
		 */
3866
		if (!test_bit(__I40E_DOWN, pf->state))
3867
			napi_schedule_irqoff(&q_vector->napi);
J
Jesse Brandeburg 已提交
3868 3869 3870 3871
	}

	if (icr0 & I40E_PFINT_ICR0_ADMINQ_MASK) {
		ena_mask &= ~I40E_PFINT_ICR0_ENA_ADMINQ_MASK;
3872
		set_bit(__I40E_ADMINQ_EVENT_PENDING, pf->state);
3873
		i40e_debug(&pf->hw, I40E_DEBUG_NVM, "AdminQ event\n");
J
Jesse Brandeburg 已提交
3874 3875 3876 3877
	}

	if (icr0 & I40E_PFINT_ICR0_MAL_DETECT_MASK) {
		ena_mask &= ~I40E_PFINT_ICR0_ENA_MAL_DETECT_MASK;
3878
		set_bit(__I40E_MDD_EVENT_PENDING, pf->state);
J
Jesse Brandeburg 已提交
3879 3880 3881 3882
	}

	if (icr0 & I40E_PFINT_ICR0_VFLR_MASK) {
		ena_mask &= ~I40E_PFINT_ICR0_ENA_VFLR_MASK;
3883
		set_bit(__I40E_VFLR_EVENT_PENDING, pf->state);
J
Jesse Brandeburg 已提交
3884 3885 3886
	}

	if (icr0 & I40E_PFINT_ICR0_GRST_MASK) {
3887 3888
		if (!test_bit(__I40E_RESET_RECOVERY_PENDING, pf->state))
			set_bit(__I40E_RESET_INTR_RECEIVED, pf->state);
J
Jesse Brandeburg 已提交
3889 3890 3891 3892
		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;
3893
		if (val == I40E_RESET_CORER) {
J
Jesse Brandeburg 已提交
3894
			pf->corer_count++;
3895
		} else if (val == I40E_RESET_GLOBR) {
J
Jesse Brandeburg 已提交
3896
			pf->globr_count++;
3897
		} else if (val == I40E_RESET_EMPR) {
J
Jesse Brandeburg 已提交
3898
			pf->empr_count++;
3899
			set_bit(__I40E_EMP_RESET_INTR_RECEIVED, pf->state);
3900
		}
J
Jesse Brandeburg 已提交
3901 3902
	}

3903 3904 3905
	if (icr0 & I40E_PFINT_ICR0_HMC_ERR_MASK) {
		icr0 &= ~I40E_PFINT_ICR0_HMC_ERR_MASK;
		dev_info(&pf->pdev->dev, "HMC error interrupt\n");
3906 3907 3908
		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));
3909 3910
	}

J
Jacob Keller 已提交
3911 3912 3913 3914
	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) {
3915
			icr0 &= ~I40E_PFINT_ICR0_ENA_TIMESYNC_MASK;
J
Jacob Keller 已提交
3916 3917 3918 3919
			i40e_ptp_tx_hwtstamp(pf);
		}
	}

J
Jesse Brandeburg 已提交
3920 3921 3922 3923 3924 3925 3926 3927
	/* 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);
3928
		if ((icr0_remaining & I40E_PFINT_ICR0_PE_CRITERR_MASK) ||
J
Jesse Brandeburg 已提交
3929
		    (icr0_remaining & I40E_PFINT_ICR0_PCI_EXCEPTION_MASK) ||
3930
		    (icr0_remaining & I40E_PFINT_ICR0_ECC_ERR_MASK)) {
3931
			dev_info(&pf->pdev->dev, "device will be reset\n");
3932
			set_bit(__I40E_PF_RESET_REQUESTED, pf->state);
3933
			i40e_service_event_schedule(pf);
J
Jesse Brandeburg 已提交
3934 3935 3936
		}
		ena_mask &= ~icr0_remaining;
	}
3937
	ret = IRQ_HANDLED;
J
Jesse Brandeburg 已提交
3938

3939
enable_intr:
J
Jesse Brandeburg 已提交
3940 3941
	/* re-enable interrupt causes */
	wr32(hw, I40E_PFINT_ICR0_ENA, ena_mask);
3942
	if (!test_bit(__I40E_DOWN, pf->state)) {
J
Jesse Brandeburg 已提交
3943
		i40e_service_event_schedule(pf);
3944
		i40e_irq_dynamic_enable_icr0(pf);
J
Jesse Brandeburg 已提交
3945 3946
	}

3947
	return ret;
J
Jesse Brandeburg 已提交
3948 3949
}

3950 3951 3952 3953 3954 3955 3956 3957 3958 3959 3960 3961 3962 3963 3964 3965 3966 3967 3968 3969 3970 3971 3972 3973 3974 3975
/**
 * 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 */
3976
		smp_rmb();
3977 3978 3979 3980 3981 3982 3983 3984 3985

		/* 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;

3986 3987 3988 3989 3990 3991 3992 3993 3994 3995 3996
		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);
		}
3997 3998 3999 4000 4001
		/* 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);
4002 4003
		if (tx_buf->tx_flags & I40E_TX_FLAGS_FD_SB)
			kfree(tx_buf->raw_buf);
4004

4005 4006 4007
		tx_buf->raw_buf = NULL;
		tx_buf->tx_flags = 0;
		tx_buf->next_to_watch = NULL;
4008
		dma_unmap_len_set(tx_buf, len, 0);
4009 4010
		tx_desc->buffer_addr = 0;
		tx_desc->cmd_type_offset_bsz = 0;
4011

4012
		/* move us past the eop_desc for start of next FD desc */
4013 4014 4015 4016 4017 4018 4019 4020 4021 4022 4023 4024 4025 4026 4027 4028
		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
4029
	if (vsi->back->flags & I40E_FLAG_MSIX_ENABLED)
4030
		i40e_irq_dynamic_enable(vsi, tx_ring->q_vector->v_idx);
J
Jesse Brandeburg 已提交
4031

4032 4033 4034 4035 4036 4037 4038 4039 4040 4041 4042 4043 4044 4045 4046 4047 4048 4049 4050 4051 4052 4053
	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
4054
/**
4055
 * i40e_map_vector_to_qp - Assigns the queue pair to the vector
J
Jesse Brandeburg 已提交
4056 4057
 * @vsi: the VSI being configured
 * @v_idx: vector index
4058
 * @qp_idx: queue pair index
J
Jesse Brandeburg 已提交
4059
 **/
4060
static void i40e_map_vector_to_qp(struct i40e_vsi *vsi, int v_idx, int qp_idx)
J
Jesse Brandeburg 已提交
4061
{
4062
	struct i40e_q_vector *q_vector = vsi->q_vectors[v_idx];
4063 4064
	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
4065 4066

	tx_ring->q_vector = q_vector;
4067 4068
	tx_ring->next = q_vector->tx.ring;
	q_vector->tx.ring = tx_ring;
J
Jesse Brandeburg 已提交
4069
	q_vector->tx.count++;
4070

4071 4072 4073 4074 4075 4076 4077 4078 4079 4080
	/* 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++;
	}

4081 4082 4083 4084
	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
4085 4086 4087 4088 4089 4090 4091 4092 4093 4094 4095 4096 4097 4098 4099
}

/**
 * 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;
4100
	int num_ringpairs;
J
Jesse Brandeburg 已提交
4101 4102 4103 4104 4105
	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.
4106 4107 4108 4109
	 * 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
4110
	 */
4111
	for (; v_start < q_vectors; v_start++) {
4112 4113 4114 4115 4116
		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;
4117
		q_vector->reg_idx = q_vector->v_idx + vsi->base_vector - 1;
4118 4119 4120 4121 4122 4123 4124

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

		while (num_ringpairs--) {
4125
			i40e_map_vector_to_qp(vsi, v_start, qp_idx);
4126 4127
			qp_idx++;
			qp_remaining--;
J
Jesse Brandeburg 已提交
4128 4129 4130 4131 4132 4133 4134 4135 4136 4137 4138 4139 4140 4141 4142 4143 4144 4145
		}
	}
}

/**
 * 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,
4146
				  pf->int_name, pf);
J
Jesse Brandeburg 已提交
4147 4148
	else
		err = request_irq(pf->pdev->irq, i40e_intr, IRQF_SHARED,
4149
				  pf->int_name, pf);
J
Jesse Brandeburg 已提交
4150 4151 4152 4153 4154 4155 4156 4157 4158

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

	return err;
}

#ifdef CONFIG_NET_POLL_CONTROLLER
/**
4159
 * i40e_netpoll - A Polling 'interrupt' handler
J
Jesse Brandeburg 已提交
4160 4161 4162 4163 4164 4165 4166 4167 4168 4169 4170 4171 4172
 * @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 */
4173
	if (test_bit(__I40E_VSI_DOWN, vsi->state))
J
Jesse Brandeburg 已提交
4174 4175 4176 4177
		return;

	if (pf->flags & I40E_FLAG_MSIX_ENABLED) {
		for (i = 0; i < vsi->num_q_vectors; i++)
4178
			i40e_msix_clean_rings(0, vsi->q_vectors[i]);
J
Jesse Brandeburg 已提交
4179 4180 4181 4182 4183 4184
	} else {
		i40e_intr(pf->pdev->irq, netdev);
	}
}
#endif

4185 4186
#define I40E_QTX_ENA_WAIT_COUNT 50

4187 4188 4189 4190 4191 4192 4193 4194 4195 4196 4197 4198 4199 4200 4201 4202 4203 4204 4205 4206 4207
/**
 * 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;

4208
		usleep_range(10, 20);
4209 4210 4211 4212 4213 4214 4215
	}
	if (i >= I40E_QUEUE_WAIT_RETRY_LIMIT)
		return -ETIMEDOUT;

	return 0;
}

4216 4217 4218 4219 4220 4221 4222 4223 4224 4225 4226 4227 4228 4229 4230 4231 4232 4233 4234 4235 4236 4237 4238 4239 4240 4241 4242 4243 4244 4245 4246 4247 4248 4249 4250 4251 4252 4253 4254 4255 4256 4257 4258 4259
/**
 * 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);
}

4260 4261 4262 4263 4264 4265 4266 4267 4268 4269 4270 4271 4272 4273 4274 4275 4276 4277 4278 4279 4280 4281 4282 4283 4284 4285 4286
/**
 * 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
4287 4288 4289 4290 4291 4292 4293 4294
/**
 * 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;
4295
	int i, pf_q, ret = 0;
J
Jesse Brandeburg 已提交
4296 4297 4298

	pf_q = vsi->base_queue;
	for (i = 0; i < vsi->num_queue_pairs; i++, pf_q++) {
4299 4300 4301 4302 4303
		ret = i40e_control_wait_tx_q(vsi->seid, pf,
					     pf_q,
					     false /*is xdp*/, enable);
		if (ret)
			break;
4304

4305 4306 4307 4308 4309 4310 4311
		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)
4312
			break;
J
Jesse Brandeburg 已提交
4313 4314
	}

4315 4316 4317 4318 4319 4320 4321 4322 4323 4324 4325 4326 4327 4328 4329 4330 4331 4332 4333 4334 4335 4336 4337 4338
	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;

4339
		usleep_range(10, 20);
4340 4341 4342
	}
	if (i >= I40E_QUEUE_WAIT_RETRY_LIMIT)
		return -ETIMEDOUT;
4343

J
Jesse Brandeburg 已提交
4344 4345 4346
	return 0;
}

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
/**
 * 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
4384 4385 4386 4387 4388 4389 4390 4391
/**
 * 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;
4392
	int i, pf_q, ret = 0;
J
Jesse Brandeburg 已提交
4393 4394 4395

	pf_q = vsi->base_queue;
	for (i = 0; i < vsi->num_queue_pairs; i++, pf_q++) {
4396
		i40e_control_rx_q(pf, pf_q, enable);
J
Jesse Brandeburg 已提交
4397 4398

		/* wait for the change to finish */
4399 4400 4401
		ret = i40e_pf_rxq_wait(pf, pf_q, enable);
		if (ret) {
			dev_info(&pf->pdev->dev,
S
Shannon Nelson 已提交
4402 4403
				 "VSI seid %d Rx ring %d %sable timeout\n",
				 vsi->seid, pf_q, (enable ? "en" : "dis"));
4404
			break;
J
Jesse Brandeburg 已提交
4405 4406 4407
		}
	}

4408 4409 4410 4411 4412 4413
	/* 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);

4414
	return ret;
J
Jesse Brandeburg 已提交
4415 4416 4417
}

/**
F
Filip Sadowski 已提交
4418
 * i40e_vsi_start_rings - Start a VSI's rings
J
Jesse Brandeburg 已提交
4419 4420
 * @vsi: the VSI being configured
 **/
F
Filip Sadowski 已提交
4421
int i40e_vsi_start_rings(struct i40e_vsi *vsi)
J
Jesse Brandeburg 已提交
4422
{
4423
	int ret = 0;
J
Jesse Brandeburg 已提交
4424 4425

	/* do rx first for enable and last for disable */
F
Filip Sadowski 已提交
4426 4427 4428 4429
	ret = i40e_vsi_control_rx(vsi, true);
	if (ret)
		return ret;
	ret = i40e_vsi_control_tx(vsi, true);
J
Jesse Brandeburg 已提交
4430 4431 4432 4433

	return ret;
}

F
Filip Sadowski 已提交
4434 4435 4436 4437 4438 4439
/**
 * i40e_vsi_stop_rings - Stop a VSI's rings
 * @vsi: the VSI being configured
 **/
void i40e_vsi_stop_rings(struct i40e_vsi *vsi)
{
4440
	/* When port TX is suspended, don't wait */
4441
	if (test_bit(__I40E_PORT_SUSPENDED, vsi->back->state))
4442 4443
		return i40e_vsi_stop_rings_no_wait(vsi);

F
Filip Sadowski 已提交
4444 4445 4446 4447 4448 4449 4450
	/* 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);
}

4451 4452 4453 4454 4455 4456 4457 4458 4459 4460 4461 4462 4463 4464 4465 4466 4467 4468 4469 4470 4471 4472 4473
/**
 * 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
4474 4475 4476 4477 4478 4479 4480 4481 4482 4483 4484 4485 4486 4487 4488 4489
/**
 * 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;

4490 4491 4492 4493
		if (!vsi->irqs_ready)
			return;

		vsi->irqs_ready = false;
J
Jesse Brandeburg 已提交
4494
		for (i = 0; i < vsi->num_q_vectors; i++) {
4495 4496 4497 4498 4499
			int irq_num;
			u16 vector;

			vector = i + base;
			irq_num = pf->msix_entries[vector].vector;
J
Jesse Brandeburg 已提交
4500 4501

			/* free only the irqs that were actually requested */
S
Shannon Nelson 已提交
4502 4503
			if (!vsi->q_vectors[i] ||
			    !vsi->q_vectors[i]->num_ringpairs)
J
Jesse Brandeburg 已提交
4504 4505
				continue;

4506 4507
			/* clear the affinity notifier in the IRQ descriptor */
			irq_set_affinity_notifier(irq_num, NULL);
4508
			/* remove our suggested affinity mask for this IRQ */
4509 4510 4511
			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
4512 4513 4514 4515 4516 4517 4518 4519 4520 4521 4522 4523 4524 4525 4526 4527 4528 4529 4530 4531 4532 4533 4534 4535 4536 4537 4538 4539 4540 4541 4542 4543 4544 4545 4546 4547 4548 4549 4550 4551 4552 4553 4554 4555 4556 4557 4558 4559 4560 4561 4562 4563 4564 4565 4566 4567 4568 4569 4570 4571 4572 4573 4574 4575 4576 4577 4578 4579 4580 4581 4582 4583 4584 4585 4586 4587 4588 4589 4590 4591 4592 4593

			/* 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);
	}
}

4594 4595 4596 4597 4598 4599 4600 4601 4602 4603 4604 4605
/**
 * 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];
4606
	struct i40e_ring *ring;
4607 4608 4609 4610 4611

	if (!q_vector)
		return;

	/* disassociate q_vector from rings */
4612 4613 4614 4615 4616
	i40e_for_each_ring(ring, q_vector->tx)
		ring->q_vector = NULL;

	i40e_for_each_ring(ring, q_vector->rx)
		ring->q_vector = NULL;
4617 4618 4619 4620 4621 4622 4623 4624 4625 4626

	/* 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
4627 4628 4629 4630 4631 4632 4633 4634 4635 4636 4637
/**
 * 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;

4638 4639
	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
4640 4641 4642 4643 4644 4645 4646 4647 4648 4649 4650 4651 4652
}

/**
 * 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;
4653 4654
		kfree(pf->irq_pile);
		pf->irq_pile = NULL;
J
Jesse Brandeburg 已提交
4655 4656 4657 4658 4659 4660 4661 4662 4663 4664 4665 4666 4667 4668 4669 4670 4671
	} 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;

4672
	i40e_free_misc_vector(pf);
4673

4674 4675 4676
	i40e_put_lump(pf->irq_pile, pf->iwarp_base_vector,
		      I40E_IWARP_IRQ_PILE_ID);

J
Jesse Brandeburg 已提交
4677
	i40e_put_lump(pf->irq_pile, 0, I40E_PILE_VALID_BIT-1);
M
Mitch Williams 已提交
4678
	for (i = 0; i < pf->num_alloc_vsi; i++)
J
Jesse Brandeburg 已提交
4679 4680 4681 4682 4683 4684 4685 4686 4687 4688 4689 4690 4691 4692 4693 4694
		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;

4695 4696 4697 4698 4699 4700
	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
4701 4702 4703 4704 4705 4706 4707 4708 4709 4710 4711 4712 4713
}

/**
 * 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;

4714 4715 4716 4717 4718 4719
	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
4720 4721
}

4722 4723 4724 4725 4726 4727
/**
 * 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
4728
	struct i40e_pf *pf = vsi->back;
4729
	if (!test_and_set_bit(__I40E_VSI_DOWN, vsi->state))
4730 4731 4732 4733
		i40e_down(vsi);
	i40e_vsi_free_irq(vsi);
	i40e_vsi_free_tx_resources(vsi);
	i40e_vsi_free_rx_resources(vsi);
4734
	vsi->current_netdev_flags = 0;
M
Mitch Williams 已提交
4735
	pf->flags |= I40E_FLAG_SERVICE_CLIENT_REQUESTED;
4736
	if (test_bit(__I40E_RESET_RECOVERY_PENDING, pf->state))
M
Mitch Williams 已提交
4737
		pf->flags |=  I40E_FLAG_CLIENT_RESET;
4738 4739
}

J
Jesse Brandeburg 已提交
4740 4741 4742 4743 4744 4745
/**
 * i40e_quiesce_vsi - Pause a given VSI
 * @vsi: the VSI being paused
 **/
static void i40e_quiesce_vsi(struct i40e_vsi *vsi)
{
4746
	if (test_bit(__I40E_VSI_DOWN, vsi->state))
J
Jesse Brandeburg 已提交
4747 4748
		return;

4749
	set_bit(__I40E_VSI_NEEDS_RESTART, vsi->state);
J
Jesse Brandeburg 已提交
4750
	if (vsi->netdev && netif_running(vsi->netdev))
J
Jesse Brandeburg 已提交
4751
		vsi->netdev->netdev_ops->ndo_stop(vsi->netdev);
J
Jesse Brandeburg 已提交
4752
	else
4753
		i40e_vsi_close(vsi);
J
Jesse Brandeburg 已提交
4754 4755 4756 4757 4758 4759 4760 4761
}

/**
 * i40e_unquiesce_vsi - Resume a given VSI
 * @vsi: the VSI being resumed
 **/
static void i40e_unquiesce_vsi(struct i40e_vsi *vsi)
{
4762
	if (!test_and_clear_bit(__I40E_VSI_NEEDS_RESTART, vsi->state))
J
Jesse Brandeburg 已提交
4763 4764 4765 4766 4767
		return;

	if (vsi->netdev && netif_running(vsi->netdev))
		vsi->netdev->netdev_ops->ndo_open(vsi->netdev);
	else
4768
		i40e_vsi_open(vsi);   /* this clears the DOWN bit */
J
Jesse Brandeburg 已提交
4769 4770 4771 4772 4773 4774 4775 4776 4777 4778
}

/**
 * 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
4779
	for (v = 0; v < pf->num_alloc_vsi; v++) {
J
Jesse Brandeburg 已提交
4780 4781 4782 4783 4784 4785 4786 4787 4788 4789 4790 4791 4792
		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
4793
	for (v = 0; v < pf->num_alloc_vsi; v++) {
J
Jesse Brandeburg 已提交
4794 4795 4796 4797 4798
		if (pf->vsi[v])
			i40e_unquiesce_vsi(pf->vsi[v]);
	}
}

4799
/**
4800
 * i40e_vsi_wait_queues_disabled - Wait for VSI's queues to be disabled
4801 4802
 * @vsi: the VSI being configured
 *
4803
 * Wait until all queues on a given VSI have been disabled.
4804
 **/
4805
int i40e_vsi_wait_queues_disabled(struct i40e_vsi *vsi)
4806 4807 4808 4809 4810 4811
{
	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++) {
4812
		/* Check and wait for the Tx queue */
4813 4814 4815
		ret = i40e_pf_txq_wait(pf, pf_q, false);
		if (ret) {
			dev_info(&pf->pdev->dev,
S
Shannon Nelson 已提交
4816 4817
				 "VSI seid %d Tx ring %d disable timeout\n",
				 vsi->seid, pf_q);
4818 4819
			return ret;
		}
4820 4821 4822 4823 4824 4825 4826 4827 4828 4829 4830 4831 4832 4833 4834

		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 */
4835 4836 4837 4838 4839 4840 4841 4842 4843
		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;
		}
	}

4844 4845 4846
	return 0;
}

4847
#ifdef CONFIG_I40E_DCB
4848
/**
4849
 * i40e_pf_wait_queues_disabled - Wait for all queues of PF VSIs to be disabled
4850 4851
 * @pf: the PF
 *
4852
 * This function waits for the queues to be in disabled state for all the
4853 4854
 * VSIs that are managed by this PF.
 **/
4855
static int i40e_pf_wait_queues_disabled(struct i40e_pf *pf)
4856 4857 4858 4859
{
	int v, ret = 0;

	for (v = 0; v < pf->hw.func_caps.num_vsis; v++) {
4860
		if (pf->vsi[v]) {
4861
			ret = i40e_vsi_wait_queues_disabled(pf->vsi[v]);
4862 4863 4864 4865 4866 4867 4868 4869 4870
			if (ret)
				break;
		}
	}

	return ret;
}

#endif
4871

4872 4873
/**
 * i40e_get_iscsi_tc_map - Return TC map for iSCSI APP
4874
 * @pf: pointer to PF
4875 4876 4877 4878 4879 4880 4881 4882 4883 4884 4885 4886 4887 4888 4889 4890 4891 4892
 *
 * 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
4893
			enabled_tc |= BIT(tc);
4894 4895 4896 4897 4898 4899 4900
			break;
		}
	}

	return enabled_tc;
}

J
Jesse Brandeburg 已提交
4901 4902 4903 4904 4905 4906 4907 4908
/**
 * 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)
{
4909
	int i, tc_unused = 0;
J
Jesse Brandeburg 已提交
4910
	u8 num_tc = 0;
4911
	u8 ret = 0;
J
Jesse Brandeburg 已提交
4912 4913 4914

	/* Scan the ETS Config Priority Table to find
	 * traffic class enabled for a given priority
4915
	 * and create a bitmask of enabled TCs
J
Jesse Brandeburg 已提交
4916
	 */
4917 4918
	for (i = 0; i < I40E_MAX_USER_PRIORITY; i++)
		num_tc |= BIT(dcbcfg->etscfg.prioritytable[i]);
J
Jesse Brandeburg 已提交
4919

4920 4921
	/* Now scan the bitmask to check for
	 * contiguous TCs starting with TC0
J
Jesse Brandeburg 已提交
4922
	 */
4923 4924 4925 4926 4927 4928 4929 4930 4931 4932 4933 4934 4935 4936 4937 4938 4939 4940
	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
4941 4942 4943 4944 4945 4946 4947 4948 4949 4950 4951 4952 4953 4954 4955 4956
}

/**
 * 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++)
4957
		enabled_tc |= BIT(i);
J
Jesse Brandeburg 已提交
4958 4959 4960 4961

	return enabled_tc;
}

4962 4963 4964 4965 4966 4967 4968 4969 4970 4971 4972 4973 4974 4975 4976 4977 4978 4979
/**
 * i40e_mqprio_get_enabled_tc - Get enabled traffic classes
 * @pf: PF being queried
 *
 * Query the current MQPRIO configuration and return the number of
 * traffic classes enabled.
 **/
static u8 i40e_mqprio_get_enabled_tc(struct i40e_pf *pf)
{
	struct i40e_vsi *vsi = pf->vsi[pf->lan_vsi];
	u8 num_tc = vsi->mqprio_qopt.qopt.num_tc;
	u8 enabled_tc = 1, i;

	for (i = 1; i < num_tc; i++)
		enabled_tc |= BIT(i);
	return enabled_tc;
}

J
Jesse Brandeburg 已提交
4980 4981 4982 4983 4984 4985 4986 4987 4988
/**
 * 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;
4989
	u8 i, enabled_tc = 1;
J
Jesse Brandeburg 已提交
4990 4991 4992
	u8 num_tc = 0;
	struct i40e_dcbx_config *dcbcfg = &hw->local_dcbx_config;

4993 4994 4995 4996
	if (pf->flags & I40E_FLAG_TC_MQPRIO)
		return pf->vsi[pf->lan_vsi]->mqprio_qopt.qopt.num_tc;

	/* If neither MQPRIO nor DCB is enabled, then always use single TC */
J
Jesse Brandeburg 已提交
4997 4998 4999
	if (!(pf->flags & I40E_FLAG_DCB_ENABLED))
		return 1;

5000 5001 5002 5003
	/* 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
5004
	/* MFP mode return count of enabled TCs for this PF */
5005 5006 5007
	if (pf->hw.func_caps.iscsi)
		enabled_tc =  i40e_get_iscsi_tc_map(pf);
	else
5008
		return 1; /* Only TC0 */
J
Jesse Brandeburg 已提交
5009

5010
	for (i = 0; i < I40E_MAX_TRAFFIC_CLASS; i++) {
J
Jesse Brandeburg 已提交
5011
		if (enabled_tc & BIT(i))
5012 5013 5014
			num_tc++;
	}
	return num_tc;
J
Jesse Brandeburg 已提交
5015 5016 5017 5018 5019 5020 5021 5022 5023 5024
}

/**
 * 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)
{
5025 5026 5027 5028 5029 5030
	if (pf->flags & I40E_FLAG_TC_MQPRIO)
		return i40e_mqprio_get_enabled_tc(pf);

	/* If neither MQPRIO nor DCB is enabled for this PF then just return
	 * default TC
	 */
J
Jesse Brandeburg 已提交
5031
	if (!(pf->flags & I40E_FLAG_DCB_ENABLED))
5032
		return I40E_DEFAULT_TRAFFIC_CLASS;
J
Jesse Brandeburg 已提交
5033 5034

	/* SFP mode we want PF to be enabled for all TCs */
5035 5036 5037
	if (!(pf->flags & I40E_FLAG_MFP_ENABLED))
		return i40e_dcb_get_enabled_tc(&pf->hw.local_dcbx_config);

5038
	/* MFP enabled and iSCSI PF type */
5039 5040 5041
	if (pf->hw.func_caps.iscsi)
		return i40e_get_iscsi_tc_map(pf);
	else
5042
		return I40E_DEFAULT_TRAFFIC_CLASS;
J
Jesse Brandeburg 已提交
5043 5044 5045 5046 5047 5048 5049 5050 5051 5052 5053 5054 5055 5056
}

/**
 * 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;
5057
	i40e_status ret;
J
Jesse Brandeburg 已提交
5058 5059 5060 5061
	u32 tc_bw_max;
	int i;

	/* Get the VSI level BW configuration */
5062 5063
	ret = i40e_aq_query_vsi_bw_config(hw, vsi->seid, &bw_config, NULL);
	if (ret) {
J
Jesse Brandeburg 已提交
5064
		dev_info(&pf->pdev->dev,
5065 5066 5067
			 "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
5068
		return -EINVAL;
J
Jesse Brandeburg 已提交
5069 5070 5071
	}

	/* Get the VSI level BW configuration per TC */
5072 5073 5074
	ret = i40e_aq_query_vsi_ets_sla_config(hw, vsi->seid, &bw_ets_config,
					       NULL);
	if (ret) {
J
Jesse Brandeburg 已提交
5075
		dev_info(&pf->pdev->dev,
5076 5077 5078
			 "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
5079
		return -EINVAL;
J
Jesse Brandeburg 已提交
5080 5081 5082 5083 5084 5085 5086 5087 5088 5089 5090 5091 5092 5093 5094 5095 5096 5097 5098 5099 5100
	}

	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
5101

J
Jesse Brandeburg 已提交
5102
	return 0;
J
Jesse Brandeburg 已提交
5103 5104 5105 5106 5107 5108 5109 5110 5111 5112
}

/**
 * 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
5113
static int i40e_vsi_configure_bw_alloc(struct i40e_vsi *vsi, u8 enabled_tc,
J
Jesse Brandeburg 已提交
5114 5115 5116
				       u8 *bw_share)
{
	struct i40e_aqc_configure_vsi_tc_bw_data bw_data;
5117
	i40e_status ret;
J
Jesse Brandeburg 已提交
5118
	int i;
J
Jesse Brandeburg 已提交
5119

5120
	if (vsi->back->flags & I40E_FLAG_TC_MQPRIO)
5121
		return 0;
5122 5123 5124 5125 5126 5127 5128 5129
	if (!vsi->mqprio_qopt.qopt.hw) {
		ret = i40e_set_bw_limit(vsi, vsi->seid, 0);
		if (ret)
			dev_info(&vsi->back->pdev->dev,
				 "Failed to reset tx rate for vsi->seid %u\n",
				 vsi->seid);
		return ret;
	}
J
Jesse Brandeburg 已提交
5130 5131 5132 5133
	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];

5134 5135 5136
	ret = i40e_aq_config_vsi_tc_bw(&vsi->back->hw, vsi->seid, &bw_data,
				       NULL);
	if (ret) {
J
Jesse Brandeburg 已提交
5137
		dev_info(&vsi->back->pdev->dev,
J
Jesse Brandeburg 已提交
5138 5139
			 "AQ command Config VSI BW allocation per TC failed = %d\n",
			 vsi->back->hw.aq.asq_last_status);
J
Jesse Brandeburg 已提交
5140
		return -EINVAL;
J
Jesse Brandeburg 已提交
5141 5142 5143 5144 5145
	}

	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
5146
	return 0;
J
Jesse Brandeburg 已提交
5147 5148 5149 5150 5151 5152 5153 5154 5155 5156 5157 5158 5159 5160 5161 5162 5163 5164 5165 5166 5167 5168 5169 5170 5171 5172 5173 5174 5175 5176 5177 5178 5179 5180 5181 5182 5183 5184
}

/**
 * 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
5185
		if (vsi->tc_config.enabled_tc & BIT(i))
J
Jesse Brandeburg 已提交
5186 5187 5188 5189 5190 5191
			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);
	}

5192 5193 5194
	if (pf->flags & I40E_FLAG_TC_MQPRIO)
		return;

J
Jesse Brandeburg 已提交
5195 5196 5197 5198 5199 5200 5201 5202 5203 5204 5205 5206 5207 5208 5209 5210 5211 5212 5213 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
	/* 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};
5240 5241
	struct i40e_pf *pf = vsi->back;
	struct i40e_hw *hw = &pf->hw;
J
Jesse Brandeburg 已提交
5242 5243 5244 5245 5246
	struct i40e_vsi_context ctxt;
	int ret = 0;
	int i;

	/* Check if enabled_tc is same as existing or new TCs */
5247 5248
	if (vsi->tc_config.enabled_tc == enabled_tc &&
	    vsi->mqprio_qopt.mode != TC_MQPRIO_MODE_CHANNEL)
J
Jesse Brandeburg 已提交
5249 5250 5251 5252
		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
5253
		if (enabled_tc & BIT(i))
J
Jesse Brandeburg 已提交
5254 5255 5256 5257 5258
			bw_share[i] = 1;
	}

	ret = i40e_vsi_configure_bw_alloc(vsi, enabled_tc, bw_share);
	if (ret) {
5259 5260
		struct i40e_aqc_query_vsi_bw_config_resp bw_config = {0};

5261
		dev_info(&pf->pdev->dev,
J
Jesse Brandeburg 已提交
5262 5263
			 "Failed configuring TC map %d for VSI %d\n",
			 enabled_tc, vsi->seid);
5264 5265 5266 5267 5268 5269 5270 5271 5272 5273 5274 5275 5276 5277 5278 5279 5280 5281 5282 5283 5284 5285 5286 5287 5288 5289 5290 5291 5292
		ret = i40e_aq_query_vsi_bw_config(hw, vsi->seid,
						  &bw_config, NULL);
		if (ret) {
			dev_info(&pf->pdev->dev,
				 "Failed querying vsi bw info, err %s aq_err %s\n",
				 i40e_stat_str(hw, ret),
				 i40e_aq_str(hw, hw->aq.asq_last_status));
			goto out;
		}
		if ((bw_config.tc_valid_bits & enabled_tc) != enabled_tc) {
			u8 valid_tc = bw_config.tc_valid_bits & enabled_tc;

			if (!valid_tc)
				valid_tc = bw_config.tc_valid_bits;
			/* Always enable TC0, no matter what */
			valid_tc |= 1;
			dev_info(&pf->pdev->dev,
				 "Requested tc 0x%x, but FW reports 0x%x as valid. Attempting to use 0x%x.\n",
				 enabled_tc, bw_config.tc_valid_bits, valid_tc);
			enabled_tc = valid_tc;
		}

		ret = i40e_vsi_configure_bw_alloc(vsi, enabled_tc, bw_share);
		if (ret) {
			dev_err(&pf->pdev->dev,
				"Unable to  configure TC map %d for VSI %d\n",
				enabled_tc, vsi->seid);
			goto out;
		}
J
Jesse Brandeburg 已提交
5293 5294 5295 5296 5297 5298 5299
	}

	/* 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;
5300
	ctxt.info = vsi->info;
5301 5302 5303 5304 5305 5306 5307
	if (vsi->back->flags & I40E_FLAG_TC_MQPRIO) {
		ret = i40e_vsi_setup_queue_map_mqprio(vsi, &ctxt, enabled_tc);
		if (ret)
			goto out;
	} else {
		i40e_vsi_setup_queue_map(vsi, &ctxt, enabled_tc, false);
	}
J
Jesse Brandeburg 已提交
5308

5309 5310 5311 5312 5313 5314 5315 5316 5317 5318 5319 5320 5321 5322
	/* On destroying the qdisc, reset vsi->rss_size, as number of enabled
	 * queues changed.
	 */
	if (!vsi->mqprio_qopt.qopt.hw && vsi->reconfig_rss) {
		vsi->rss_size = min_t(int, vsi->back->alloc_rss_size,
				      vsi->num_queue_pairs);
		ret = i40e_vsi_config_rss(vsi);
		if (ret) {
			dev_info(&vsi->back->pdev->dev,
				 "Failed to reconfig rss for num_queues\n");
			return ret;
		}
		vsi->reconfig_rss = false;
	}
5323 5324 5325 5326 5327 5328
	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;
	}

5329 5330 5331
	/* Update the VSI after updating the VSI queue-mapping
	 * information
	 */
5332
	ret = i40e_aq_update_vsi_params(hw, &ctxt, NULL);
J
Jesse Brandeburg 已提交
5333
	if (ret) {
5334
		dev_info(&pf->pdev->dev,
5335
			 "Update vsi tc config failed, err %s aq_err %s\n",
5336 5337
			 i40e_stat_str(hw, ret),
			 i40e_aq_str(hw, hw->aq.asq_last_status));
J
Jesse Brandeburg 已提交
5338 5339 5340 5341 5342 5343 5344 5345 5346
		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) {
5347
		dev_info(&pf->pdev->dev,
5348
			 "Failed updating vsi bw info, err %s aq_err %s\n",
5349 5350
			 i40e_stat_str(hw, ret),
			 i40e_aq_str(hw, hw->aq.asq_last_status));
J
Jesse Brandeburg 已提交
5351 5352 5353 5354 5355 5356 5357 5358 5359
		goto out;
	}

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

N
Neerav Parikh 已提交
5360
/**
5361 5362
 * i40e_get_link_speed - Returns link speed for the interface
 * @vsi: VSI to be configured
N
Neerav Parikh 已提交
5363 5364
 *
 **/
5365
static int i40e_get_link_speed(struct i40e_vsi *vsi)
N
Neerav Parikh 已提交
5366
{
5367
	struct i40e_pf *pf = vsi->back;
N
Neerav Parikh 已提交
5368

5369 5370 5371 5372 5373 5374 5375 5376 5377 5378 5379 5380 5381
	switch (pf->hw.phy.link_info.link_speed) {
	case I40E_LINK_SPEED_40GB:
		return 40000;
	case I40E_LINK_SPEED_25GB:
		return 25000;
	case I40E_LINK_SPEED_20GB:
		return 20000;
	case I40E_LINK_SPEED_10GB:
		return 10000;
	case I40E_LINK_SPEED_1GB:
		return 1000;
	default:
		return -EINVAL;
N
Neerav Parikh 已提交
5382 5383 5384 5385
	}
}

/**
5386 5387 5388 5389
 * i40e_set_bw_limit - setup BW limit for Tx traffic based on max_tx_rate
 * @vsi: VSI to be configured
 * @seid: seid of the channel/VSI
 * @max_tx_rate: max TX rate to be configured as BW limit
N
Neerav Parikh 已提交
5390
 *
5391
 * Helper function to set BW limit for a given VSI
N
Neerav Parikh 已提交
5392
 **/
5393
int i40e_set_bw_limit(struct i40e_vsi *vsi, u16 seid, u64 max_tx_rate)
N
Neerav Parikh 已提交
5394
{
5395
	struct i40e_pf *pf = vsi->back;
A
Alan Brady 已提交
5396
	u64 credits = 0;
5397 5398
	int speed = 0;
	int ret = 0;
N
Neerav Parikh 已提交
5399

5400 5401 5402 5403 5404 5405 5406 5407 5408 5409 5410 5411 5412 5413
	speed = i40e_get_link_speed(vsi);
	if (max_tx_rate > speed) {
		dev_err(&pf->pdev->dev,
			"Invalid max tx rate %llu specified for VSI seid %d.",
			max_tx_rate, seid);
		return -EINVAL;
	}
	if (max_tx_rate && max_tx_rate < 50) {
		dev_warn(&pf->pdev->dev,
			 "Setting max tx rate to minimum usable value of 50Mbps.\n");
		max_tx_rate = 50;
	}

	/* Tx rate credits are in values of 50Mbps, 0 is disabled */
A
Alan Brady 已提交
5414 5415 5416
	credits = max_tx_rate;
	do_div(credits, I40E_BW_CREDIT_DIVISOR);
	ret = i40e_aq_config_vsi_bw_limit(&pf->hw, seid, credits,
5417 5418 5419 5420 5421 5422 5423 5424 5425
					  I40E_MAX_BW_INACTIVE_ACCUM, NULL);
	if (ret)
		dev_err(&pf->pdev->dev,
			"Failed set tx rate (%llu Mbps) for vsi->seid %u, err %s aq_err %s\n",
			max_tx_rate, seid, i40e_stat_str(&pf->hw, ret),
			i40e_aq_str(&pf->hw, pf->hw.aq.asq_last_status));
	return ret;
}

5426 5427 5428 5429 5430 5431 5432 5433
/**
 * i40e_remove_queue_channels - Remove queue channels for the TCs
 * @vsi: VSI to be configured
 *
 * Remove queue channels for the TCs
 **/
static void i40e_remove_queue_channels(struct i40e_vsi *vsi)
{
5434 5435
	enum i40e_admin_queue_err last_aq_status;
	struct i40e_cloud_filter *cfilter;
5436
	struct i40e_channel *ch, *ch_tmp;
5437 5438
	struct i40e_pf *pf = vsi->back;
	struct hlist_node *node;
5439 5440 5441 5442 5443 5444 5445 5446 5447 5448 5449 5450 5451 5452 5453 5454 5455 5456
	int ret, i;

	/* Reset rss size that was stored when reconfiguring rss for
	 * channel VSIs with non-power-of-2 queue count.
	 */
	vsi->current_rss_size = 0;

	/* perform cleanup for channels if they exist */
	if (list_empty(&vsi->ch_list))
		return;

	list_for_each_entry_safe(ch, ch_tmp, &vsi->ch_list, list) {
		struct i40e_vsi *p_vsi;

		list_del(&ch->list);
		p_vsi = ch->parent_vsi;
		if (!p_vsi || !ch->initialized) {
			kfree(ch);
N
Neerav Parikh 已提交
5457
			continue;
5458 5459 5460 5461 5462 5463 5464 5465 5466 5467 5468 5469 5470 5471
		}
		/* Reset queue contexts */
		for (i = 0; i < ch->num_queue_pairs; i++) {
			struct i40e_ring *tx_ring, *rx_ring;
			u16 pf_q;

			pf_q = ch->base_queue + i;
			tx_ring = vsi->tx_rings[pf_q];
			tx_ring->ch = NULL;

			rx_ring = vsi->rx_rings[pf_q];
			rx_ring->ch = NULL;
		}

5472 5473 5474 5475 5476 5477 5478
		/* Reset BW configured for this VSI via mqprio */
		ret = i40e_set_bw_limit(vsi, ch->seid, 0);
		if (ret)
			dev_info(&vsi->back->pdev->dev,
				 "Failed to reset tx rate for ch->seid %u\n",
				 ch->seid);

5479 5480 5481 5482 5483 5484 5485 5486 5487 5488 5489 5490 5491 5492 5493 5494 5495 5496 5497 5498 5499 5500 5501
		/* delete cloud filters associated with this channel */
		hlist_for_each_entry_safe(cfilter, node,
					  &pf->cloud_filter_list, cloud_node) {
			if (cfilter->seid != ch->seid)
				continue;

			hash_del(&cfilter->cloud_node);
			if (cfilter->dst_port)
				ret = i40e_add_del_cloud_filter_big_buf(vsi,
									cfilter,
									false);
			else
				ret = i40e_add_del_cloud_filter(vsi, cfilter,
								false);
			last_aq_status = pf->hw.aq.asq_last_status;
			if (ret)
				dev_info(&pf->pdev->dev,
					 "Failed to delete cloud filter, err %s aq_err %s\n",
					 i40e_stat_str(&pf->hw, ret),
					 i40e_aq_str(&pf->hw, last_aq_status));
			kfree(cfilter);
		}

5502 5503 5504 5505 5506 5507 5508 5509 5510 5511 5512 5513 5514 5515 5516 5517 5518 5519 5520 5521 5522 5523 5524 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 5551 5552 5553 5554 5555 5556 5557 5558 5559 5560 5561 5562 5563 5564 5565 5566 5567 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 5604 5605 5606 5607 5608 5609 5610 5611 5612 5613 5614 5615 5616 5617 5618 5619 5620 5621 5622 5623 5624 5625 5626 5627 5628 5629 5630 5631 5632 5633 5634 5635 5636 5637 5638 5639 5640 5641 5642 5643 5644 5645 5646 5647 5648 5649 5650 5651 5652 5653 5654 5655 5656 5657 5658 5659 5660 5661 5662 5663 5664 5665 5666 5667 5668 5669 5670 5671 5672 5673 5674 5675 5676 5677 5678 5679 5680 5681 5682 5683 5684 5685 5686 5687 5688 5689 5690 5691 5692 5693 5694 5695 5696 5697 5698 5699 5700 5701 5702 5703 5704 5705 5706 5707 5708 5709 5710 5711 5712 5713 5714 5715 5716 5717 5718 5719 5720 5721 5722 5723 5724 5725 5726 5727 5728 5729 5730 5731 5732 5733 5734 5735 5736 5737 5738 5739 5740 5741 5742 5743 5744 5745 5746 5747 5748 5749 5750 5751 5752 5753 5754 5755 5756 5757 5758 5759 5760 5761 5762 5763 5764 5765 5766 5767 5768 5769 5770 5771 5772 5773 5774 5775 5776 5777 5778 5779 5780 5781 5782 5783 5784 5785 5786 5787 5788 5789 5790 5791 5792 5793 5794 5795 5796 5797 5798 5799 5800 5801 5802 5803 5804 5805 5806 5807 5808 5809 5810 5811 5812 5813 5814 5815 5816 5817 5818 5819 5820 5821 5822 5823 5824 5825 5826 5827 5828 5829 5830 5831 5832 5833 5834 5835 5836 5837 5838 5839 5840 5841 5842 5843 5844 5845 5846 5847 5848 5849 5850 5851 5852 5853 5854 5855 5856 5857 5858 5859 5860 5861 5862 5863 5864 5865 5866 5867 5868 5869 5870 5871 5872 5873 5874 5875 5876 5877 5878 5879 5880 5881 5882 5883 5884 5885 5886 5887 5888 5889 5890 5891 5892 5893 5894 5895 5896 5897 5898 5899 5900 5901 5902 5903 5904 5905 5906 5907 5908 5909 5910 5911 5912 5913 5914 5915 5916 5917 5918 5919 5920 5921 5922 5923 5924 5925 5926 5927 5928 5929 5930 5931 5932 5933 5934 5935 5936 5937 5938 5939 5940 5941 5942 5943
		/* delete VSI from FW */
		ret = i40e_aq_delete_element(&vsi->back->hw, ch->seid,
					     NULL);
		if (ret)
			dev_err(&vsi->back->pdev->dev,
				"unable to remove channel (%d) for parent VSI(%d)\n",
				ch->seid, p_vsi->seid);
		kfree(ch);
	}
	INIT_LIST_HEAD(&vsi->ch_list);
}

/**
 * i40e_is_any_channel - channel exist or not
 * @vsi: ptr to VSI to which channels are associated with
 *
 * Returns true or false if channel(s) exist for associated VSI or not
 **/
static bool i40e_is_any_channel(struct i40e_vsi *vsi)
{
	struct i40e_channel *ch, *ch_tmp;

	list_for_each_entry_safe(ch, ch_tmp, &vsi->ch_list, list) {
		if (ch->initialized)
			return true;
	}

	return false;
}

/**
 * i40e_get_max_queues_for_channel
 * @vsi: ptr to VSI to which channels are associated with
 *
 * Helper function which returns max value among the queue counts set on the
 * channels/TCs created.
 **/
static int i40e_get_max_queues_for_channel(struct i40e_vsi *vsi)
{
	struct i40e_channel *ch, *ch_tmp;
	int max = 0;

	list_for_each_entry_safe(ch, ch_tmp, &vsi->ch_list, list) {
		if (!ch->initialized)
			continue;
		if (ch->num_queue_pairs > max)
			max = ch->num_queue_pairs;
	}

	return max;
}

/**
 * i40e_validate_num_queues - validate num_queues w.r.t channel
 * @pf: ptr to PF device
 * @num_queues: number of queues
 * @vsi: the parent VSI
 * @reconfig_rss: indicates should the RSS be reconfigured or not
 *
 * This function validates number of queues in the context of new channel
 * which is being established and determines if RSS should be reconfigured
 * or not for parent VSI.
 **/
static int i40e_validate_num_queues(struct i40e_pf *pf, int num_queues,
				    struct i40e_vsi *vsi, bool *reconfig_rss)
{
	int max_ch_queues;

	if (!reconfig_rss)
		return -EINVAL;

	*reconfig_rss = false;
	if (vsi->current_rss_size) {
		if (num_queues > vsi->current_rss_size) {
			dev_dbg(&pf->pdev->dev,
				"Error: num_queues (%d) > vsi's current_size(%d)\n",
				num_queues, vsi->current_rss_size);
			return -EINVAL;
		} else if ((num_queues < vsi->current_rss_size) &&
			   (!is_power_of_2(num_queues))) {
			dev_dbg(&pf->pdev->dev,
				"Error: num_queues (%d) < vsi's current_size(%d), but not power of 2\n",
				num_queues, vsi->current_rss_size);
			return -EINVAL;
		}
	}

	if (!is_power_of_2(num_queues)) {
		/* Find the max num_queues configured for channel if channel
		 * exist.
		 * if channel exist, then enforce 'num_queues' to be more than
		 * max ever queues configured for channel.
		 */
		max_ch_queues = i40e_get_max_queues_for_channel(vsi);
		if (num_queues < max_ch_queues) {
			dev_dbg(&pf->pdev->dev,
				"Error: num_queues (%d) < max queues configured for channel(%d)\n",
				num_queues, max_ch_queues);
			return -EINVAL;
		}
		*reconfig_rss = true;
	}

	return 0;
}

/**
 * i40e_vsi_reconfig_rss - reconfig RSS based on specified rss_size
 * @vsi: the VSI being setup
 * @rss_size: size of RSS, accordingly LUT gets reprogrammed
 *
 * This function reconfigures RSS by reprogramming LUTs using 'rss_size'
 **/
static int i40e_vsi_reconfig_rss(struct i40e_vsi *vsi, u16 rss_size)
{
	struct i40e_pf *pf = vsi->back;
	u8 seed[I40E_HKEY_ARRAY_SIZE];
	struct i40e_hw *hw = &pf->hw;
	int local_rss_size;
	u8 *lut;
	int ret;

	if (!vsi->rss_size)
		return -EINVAL;

	if (rss_size > vsi->rss_size)
		return -EINVAL;

	local_rss_size = min_t(int, vsi->rss_size, rss_size);
	lut = kzalloc(vsi->rss_table_size, GFP_KERNEL);
	if (!lut)
		return -ENOMEM;

	/* Ignoring user configured lut if there is one */
	i40e_fill_rss_lut(pf, lut, vsi->rss_table_size, local_rss_size);

	/* 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);

	ret = i40e_config_rss(vsi, seed, lut, vsi->rss_table_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));
		kfree(lut);
		return ret;
	}
	kfree(lut);

	/* Do the update w.r.t. storing rss_size */
	if (!vsi->orig_rss_size)
		vsi->orig_rss_size = vsi->rss_size;
	vsi->current_rss_size = local_rss_size;

	return ret;
}

/**
 * i40e_channel_setup_queue_map - Setup a channel queue map
 * @pf: ptr to PF device
 * @vsi: the VSI being setup
 * @ctxt: VSI context structure
 * @ch: ptr to channel structure
 *
 * Setup queue map for a specific channel
 **/
static void i40e_channel_setup_queue_map(struct i40e_pf *pf,
					 struct i40e_vsi_context *ctxt,
					 struct i40e_channel *ch)
{
	u16 qcount, qmap, sections = 0;
	u8 offset = 0;
	int pow;

	sections = I40E_AQ_VSI_PROP_QUEUE_MAP_VALID;
	sections |= I40E_AQ_VSI_PROP_SCHED_VALID;

	qcount = min_t(int, ch->num_queue_pairs, pf->num_lan_msix);
	ch->num_queue_pairs = qcount;

	/* find the next higher power-of-2 of num queue pairs */
	pow = ilog2(qcount);
	if (!is_power_of_2(qcount))
		pow++;

	qmap = (offset << I40E_AQ_VSI_TC_QUE_OFFSET_SHIFT) |
		(pow << I40E_AQ_VSI_TC_QUE_NUMBER_SHIFT);

	/* Setup queue TC[0].qmap for given VSI context */
	ctxt->info.tc_mapping[0] = cpu_to_le16(qmap);

	ctxt->info.up_enable_bits = 0x1; /* TC0 enabled */
	ctxt->info.mapping_flags |= cpu_to_le16(I40E_AQ_VSI_QUE_MAP_CONTIG);
	ctxt->info.queue_mapping[0] = cpu_to_le16(ch->base_queue);
	ctxt->info.valid_sections |= cpu_to_le16(sections);
}

/**
 * i40e_add_channel - add a channel by adding VSI
 * @pf: ptr to PF device
 * @uplink_seid: underlying HW switching element (VEB) ID
 * @ch: ptr to channel structure
 *
 * Add a channel (VSI) using add_vsi and queue_map
 **/
static int i40e_add_channel(struct i40e_pf *pf, u16 uplink_seid,
			    struct i40e_channel *ch)
{
	struct i40e_hw *hw = &pf->hw;
	struct i40e_vsi_context ctxt;
	u8 enabled_tc = 0x1; /* TC0 enabled */
	int ret;

	if (ch->type != I40E_VSI_VMDQ2) {
		dev_info(&pf->pdev->dev,
			 "add new vsi failed, ch->type %d\n", ch->type);
		return -EINVAL;
	}

	memset(&ctxt, 0, sizeof(ctxt));
	ctxt.pf_num = hw->pf_id;
	ctxt.vf_num = 0;
	ctxt.uplink_seid = uplink_seid;
	ctxt.connection_type = I40E_AQ_VSI_CONN_TYPE_NORMAL;
	if (ch->type == I40E_VSI_VMDQ2)
		ctxt.flags = I40E_AQ_VSI_TYPE_VMDQ2;

	if (pf->flags & I40E_FLAG_VEB_MODE_ENABLED) {
		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);
	}

	/* Set queue map for a given VSI context */
	i40e_channel_setup_queue_map(pf, &ctxt, ch);

	/* Now time to create VSI */
	ret = i40e_aq_add_vsi(hw, &ctxt, NULL);
	if (ret) {
		dev_info(&pf->pdev->dev,
			 "add new 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));
		return -ENOENT;
	}

	/* Success, update channel */
	ch->enabled_tc = enabled_tc;
	ch->seid = ctxt.seid;
	ch->vsi_number = ctxt.vsi_number;
	ch->stat_counter_idx = cpu_to_le16(ctxt.info.stat_counter_idx);

	/* copy just the sections touched not the entire info
	 * since not all sections are valid as returned by
	 * update vsi params
	 */
	ch->info.mapping_flags = ctxt.info.mapping_flags;
	memcpy(&ch->info.queue_mapping,
	       &ctxt.info.queue_mapping, sizeof(ctxt.info.queue_mapping));
	memcpy(&ch->info.tc_mapping, ctxt.info.tc_mapping,
	       sizeof(ctxt.info.tc_mapping));

	return 0;
}

static int i40e_channel_config_bw(struct i40e_vsi *vsi, struct i40e_channel *ch,
				  u8 *bw_share)
{
	struct i40e_aqc_configure_vsi_tc_bw_data bw_data;
	i40e_status ret;
	int i;

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

	ret = i40e_aq_config_vsi_tc_bw(&vsi->back->hw, ch->seid,
				       &bw_data, NULL);
	if (ret) {
		dev_info(&vsi->back->pdev->dev,
			 "Config VSI BW allocation per TC failed, aq_err: %d for new_vsi->seid %u\n",
			 vsi->back->hw.aq.asq_last_status, ch->seid);
		return -EINVAL;
	}

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

	return 0;
}

/**
 * i40e_channel_config_tx_ring - config TX ring associated with new channel
 * @pf: ptr to PF device
 * @vsi: the VSI being setup
 * @ch: ptr to channel structure
 *
 * Configure TX rings associated with channel (VSI) since queues are being
 * from parent VSI.
 **/
static int i40e_channel_config_tx_ring(struct i40e_pf *pf,
				       struct i40e_vsi *vsi,
				       struct i40e_channel *ch)
{
	i40e_status ret;
	int i;
	u8 bw_share[I40E_MAX_TRAFFIC_CLASS] = {0};

	/* Enable ETS TCs with equal BW Share for now across all VSIs */
	for (i = 0; i < I40E_MAX_TRAFFIC_CLASS; i++) {
		if (ch->enabled_tc & BIT(i))
			bw_share[i] = 1;
	}

	/* configure BW for new VSI */
	ret = i40e_channel_config_bw(vsi, ch, bw_share);
	if (ret) {
		dev_info(&vsi->back->pdev->dev,
			 "Failed configuring TC map %d for channel (seid %u)\n",
			 ch->enabled_tc, ch->seid);
		return ret;
	}

	for (i = 0; i < ch->num_queue_pairs; i++) {
		struct i40e_ring *tx_ring, *rx_ring;
		u16 pf_q;

		pf_q = ch->base_queue + i;

		/* Get to TX ring ptr of main VSI, for re-setup TX queue
		 * context
		 */
		tx_ring = vsi->tx_rings[pf_q];
		tx_ring->ch = ch;

		/* Get the RX ring ptr */
		rx_ring = vsi->rx_rings[pf_q];
		rx_ring->ch = ch;
	}

	return 0;
}

/**
 * i40e_setup_hw_channel - setup new channel
 * @pf: ptr to PF device
 * @vsi: the VSI being setup
 * @ch: ptr to channel structure
 * @uplink_seid: underlying HW switching element (VEB) ID
 * @type: type of channel to be created (VMDq2/VF)
 *
 * Setup new channel (VSI) based on specified type (VMDq2/VF)
 * and configures TX rings accordingly
 **/
static inline int i40e_setup_hw_channel(struct i40e_pf *pf,
					struct i40e_vsi *vsi,
					struct i40e_channel *ch,
					u16 uplink_seid, u8 type)
{
	int ret;

	ch->initialized = false;
	ch->base_queue = vsi->next_base_queue;
	ch->type = type;

	/* Proceed with creation of channel (VMDq2) VSI */
	ret = i40e_add_channel(pf, uplink_seid, ch);
	if (ret) {
		dev_info(&pf->pdev->dev,
			 "failed to add_channel using uplink_seid %u\n",
			 uplink_seid);
		return ret;
	}

	/* Mark the successful creation of channel */
	ch->initialized = true;

	/* Reconfigure TX queues using QTX_CTL register */
	ret = i40e_channel_config_tx_ring(pf, vsi, ch);
	if (ret) {
		dev_info(&pf->pdev->dev,
			 "failed to configure TX rings for channel %u\n",
			 ch->seid);
		return ret;
	}

	/* update 'next_base_queue' */
	vsi->next_base_queue = vsi->next_base_queue + ch->num_queue_pairs;
	dev_dbg(&pf->pdev->dev,
		"Added channel: vsi_seid %u, vsi_number %u, stat_counter_idx %u, num_queue_pairs %u, pf->next_base_queue %d\n",
		ch->seid, ch->vsi_number, ch->stat_counter_idx,
		ch->num_queue_pairs,
		vsi->next_base_queue);
	return ret;
}

/**
 * i40e_setup_channel - setup new channel using uplink element
 * @pf: ptr to PF device
 * @type: type of channel to be created (VMDq2/VF)
 * @uplink_seid: underlying HW switching element (VEB) ID
 * @ch: ptr to channel structure
 *
 * Setup new channel (VSI) based on specified type (VMDq2/VF)
 * and uplink switching element (uplink_seid)
 **/
static bool i40e_setup_channel(struct i40e_pf *pf, struct i40e_vsi *vsi,
			       struct i40e_channel *ch)
{
	u8 vsi_type;
	u16 seid;
	int ret;

	if (vsi->type == I40E_VSI_MAIN) {
		vsi_type = I40E_VSI_VMDQ2;
	} else {
		dev_err(&pf->pdev->dev, "unsupported parent vsi type(%d)\n",
			vsi->type);
		return false;
	}

	/* underlying switching element */
	seid = pf->vsi[pf->lan_vsi]->uplink_seid;

	/* create channel (VSI), configure TX rings */
	ret = i40e_setup_hw_channel(pf, vsi, ch, seid, vsi_type);
	if (ret) {
		dev_err(&pf->pdev->dev, "failed to setup hw_channel\n");
		return false;
	}

	return ch->initialized ? true : false;
}

5944 5945 5946 5947 5948 5949 5950 5951 5952 5953 5954 5955 5956 5957 5958 5959 5960 5961 5962 5963 5964 5965 5966 5967 5968 5969 5970 5971 5972 5973 5974 5975 5976 5977 5978 5979 5980
/**
 * i40e_validate_and_set_switch_mode - sets up switch mode correctly
 * @vsi: ptr to VSI which has PF backing
 *
 * Sets up switch mode correctly if it needs to be changed and perform
 * what are allowed modes.
 **/
static int i40e_validate_and_set_switch_mode(struct i40e_vsi *vsi)
{
	u8 mode;
	struct i40e_pf *pf = vsi->back;
	struct i40e_hw *hw = &pf->hw;
	int ret;

	ret = i40e_get_capabilities(pf, i40e_aqc_opc_list_dev_capabilities);
	if (ret)
		return -EINVAL;

	if (hw->dev_caps.switch_mode) {
		/* if switch mode is set, support mode2 (non-tunneled for
		 * cloud filter) for now
		 */
		u32 switch_mode = hw->dev_caps.switch_mode &
				  I40E_SWITCH_MODE_MASK;
		if (switch_mode >= I40E_CLOUD_FILTER_MODE1) {
			if (switch_mode == I40E_CLOUD_FILTER_MODE2)
				return 0;
			dev_err(&pf->pdev->dev,
				"Invalid switch_mode (%d), only non-tunneled mode for cloud filter is supported\n",
				hw->dev_caps.switch_mode);
			return -EINVAL;
		}
	}

	/* Set Bit 7 to be valid */
	mode = I40E_AQ_SET_SWITCH_BIT7_VALID;

5981 5982
	/* Set L4type for TCP support */
	mode |= I40E_AQ_SET_SWITCH_L4_TYPE_TCP;
5983 5984 5985 5986 5987 5988 5989 5990 5991 5992 5993 5994 5995 5996 5997 5998 5999 6000

	/* Set cloud filter mode */
	mode |= I40E_AQ_SET_SWITCH_MODE_NON_TUNNEL;

	/* Prep mode field for set_switch_config */
	ret = i40e_aq_set_switch_config(hw, pf->last_sw_conf_flags,
					pf->last_sw_conf_valid_flags,
					mode, NULL);
	if (ret && hw->aq.asq_last_status != I40E_AQ_RC_ESRCH)
		dev_err(&pf->pdev->dev,
			"couldn't set switch config bits, err %s aq_err %s\n",
			i40e_stat_str(hw, ret),
			i40e_aq_str(hw,
				    hw->aq.asq_last_status));

	return ret;
}

6001 6002 6003 6004 6005 6006 6007 6008 6009 6010 6011 6012 6013 6014 6015 6016 6017 6018 6019 6020 6021 6022 6023 6024 6025 6026 6027 6028 6029 6030 6031 6032 6033 6034 6035 6036 6037 6038 6039 6040 6041 6042 6043 6044 6045 6046 6047 6048 6049 6050 6051 6052 6053 6054 6055 6056 6057 6058 6059 6060 6061 6062 6063 6064 6065 6066 6067 6068 6069 6070 6071 6072 6073 6074 6075 6076 6077 6078 6079 6080 6081 6082 6083 6084 6085 6086 6087 6088 6089 6090 6091 6092
/**
 * i40e_create_queue_channel - function to create channel
 * @vsi: VSI to be configured
 * @ch: ptr to channel (it contains channel specific params)
 *
 * This function creates channel (VSI) using num_queues specified by user,
 * reconfigs RSS if needed.
 **/
int i40e_create_queue_channel(struct i40e_vsi *vsi,
			      struct i40e_channel *ch)
{
	struct i40e_pf *pf = vsi->back;
	bool reconfig_rss;
	int err;

	if (!ch)
		return -EINVAL;

	if (!ch->num_queue_pairs) {
		dev_err(&pf->pdev->dev, "Invalid num_queues requested: %d\n",
			ch->num_queue_pairs);
		return -EINVAL;
	}

	/* validate user requested num_queues for channel */
	err = i40e_validate_num_queues(pf, ch->num_queue_pairs, vsi,
				       &reconfig_rss);
	if (err) {
		dev_info(&pf->pdev->dev, "Failed to validate num_queues (%d)\n",
			 ch->num_queue_pairs);
		return -EINVAL;
	}

	/* By default we are in VEPA mode, if this is the first VF/VMDq
	 * VSI to be added switch to VEB mode.
	 */
	if ((!(pf->flags & I40E_FLAG_VEB_MODE_ENABLED)) ||
	    (!i40e_is_any_channel(vsi))) {
		if (!is_power_of_2(vsi->tc_config.tc_info[0].qcount)) {
			dev_dbg(&pf->pdev->dev,
				"Failed to create channel. Override queues (%u) not power of 2\n",
				vsi->tc_config.tc_info[0].qcount);
			return -EINVAL;
		}

		if (!(pf->flags & I40E_FLAG_VEB_MODE_ENABLED)) {
			pf->flags |= I40E_FLAG_VEB_MODE_ENABLED;

			if (vsi->type == I40E_VSI_MAIN) {
				if (pf->flags & I40E_FLAG_TC_MQPRIO)
					i40e_do_reset(pf, I40E_PF_RESET_FLAG,
						      true);
				else
					i40e_do_reset_safe(pf,
							   I40E_PF_RESET_FLAG);
			}
		}
		/* now onwards for main VSI, number of queues will be value
		 * of TC0's queue count
		 */
	}

	/* By this time, vsi->cnt_q_avail shall be set to non-zero and
	 * it should be more than num_queues
	 */
	if (!vsi->cnt_q_avail || vsi->cnt_q_avail < ch->num_queue_pairs) {
		dev_dbg(&pf->pdev->dev,
			"Error: cnt_q_avail (%u) less than num_queues %d\n",
			vsi->cnt_q_avail, ch->num_queue_pairs);
		return -EINVAL;
	}

	/* reconfig_rss only if vsi type is MAIN_VSI */
	if (reconfig_rss && (vsi->type == I40E_VSI_MAIN)) {
		err = i40e_vsi_reconfig_rss(vsi, ch->num_queue_pairs);
		if (err) {
			dev_info(&pf->pdev->dev,
				 "Error: unable to reconfig rss for num_queues (%u)\n",
				 ch->num_queue_pairs);
			return -EINVAL;
		}
	}

	if (!i40e_setup_channel(pf, vsi, ch)) {
		dev_info(&pf->pdev->dev, "Failed to setup channel\n");
		return -EINVAL;
	}

	dev_info(&pf->pdev->dev,
		 "Setup channel (id:%u) utilizing num_queues %d\n",
		 ch->seid, ch->num_queue_pairs);

6093 6094
	/* configure VSI for BW limit */
	if (ch->max_tx_rate) {
A
Alan Brady 已提交
6095 6096
		u64 credits = ch->max_tx_rate;

6097 6098 6099
		if (i40e_set_bw_limit(vsi, ch->seid, ch->max_tx_rate))
			return -EINVAL;

A
Alan Brady 已提交
6100
		do_div(credits, I40E_BW_CREDIT_DIVISOR);
6101 6102 6103
		dev_dbg(&pf->pdev->dev,
			"Set tx rate of %llu Mbps (count of 50Mbps %llu) for vsi->seid %u\n",
			ch->max_tx_rate,
A
Alan Brady 已提交
6104 6105
			credits,
			ch->seid);
6106 6107
	}

6108 6109 6110 6111 6112 6113 6114 6115 6116 6117 6118 6119 6120 6121 6122 6123 6124 6125
	/* in case of VF, this will be main SRIOV VSI */
	ch->parent_vsi = vsi;

	/* and update main_vsi's count for queue_available to use */
	vsi->cnt_q_avail -= ch->num_queue_pairs;

	return 0;
}

/**
 * i40e_configure_queue_channels - Add queue channel for the given TCs
 * @vsi: VSI to be configured
 *
 * Configures queue channel mapping to the given TCs
 **/
static int i40e_configure_queue_channels(struct i40e_vsi *vsi)
{
	struct i40e_channel *ch;
A
Alan Brady 已提交
6126
	u64 max_rate = 0;
6127 6128 6129
	int ret = 0, i;

	/* Create app vsi with the TCs. Main VSI with TC0 is already set up */
A
Amritha Nambiar 已提交
6130
	vsi->tc_seid_map[0] = vsi->seid;
6131 6132 6133 6134 6135 6136 6137 6138 6139 6140 6141 6142 6143 6144
	for (i = 1; i < I40E_MAX_TRAFFIC_CLASS; i++) {
		if (vsi->tc_config.enabled_tc & BIT(i)) {
			ch = kzalloc(sizeof(*ch), GFP_KERNEL);
			if (!ch) {
				ret = -ENOMEM;
				goto err_free;
			}

			INIT_LIST_HEAD(&ch->list);
			ch->num_queue_pairs =
				vsi->tc_config.tc_info[i].qcount;
			ch->base_queue =
				vsi->tc_config.tc_info[i].qoffset;

6145 6146 6147
			/* Bandwidth limit through tc interface is in bytes/s,
			 * change to Mbit/s
			 */
A
Alan Brady 已提交
6148 6149 6150
			max_rate = vsi->mqprio_qopt.max_rate[i];
			do_div(max_rate, I40E_BW_MBPS_DIVISOR);
			ch->max_tx_rate = max_rate;
6151

6152 6153 6154 6155 6156 6157 6158 6159 6160
			list_add_tail(&ch->list, &vsi->ch_list);

			ret = i40e_create_queue_channel(vsi, ch);
			if (ret) {
				dev_err(&vsi->back->pdev->dev,
					"Failed creating queue channel with TC%d: queues %d\n",
					i, ch->num_queue_pairs);
				goto err_free;
			}
A
Amritha Nambiar 已提交
6161
			vsi->tc_seid_map[i] = ch->seid;
6162 6163 6164 6165 6166 6167 6168 6169 6170
		}
	}
	return ret;

err_free:
	i40e_remove_queue_channels(vsi);
	return ret;
}

N
Neerav Parikh 已提交
6171 6172 6173 6174 6175 6176 6177 6178 6179 6180 6181 6182 6183 6184 6185 6186 6187 6188 6189 6190 6191 6192 6193
/**
 * 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
6194
		if (enabled_tc & BIT(i))
N
Neerav Parikh 已提交
6195 6196 6197 6198 6199 6200 6201
			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,
6202 6203 6204
			 "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
6205 6206 6207 6208 6209 6210 6211
		goto out;
	}

	/* Update the BW information */
	ret = i40e_veb_get_bw_info(veb);
	if (ret) {
		dev_info(&pf->pdev->dev,
6212 6213 6214
			 "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
6215 6216 6217 6218 6219 6220 6221 6222 6223 6224 6225 6226 6227 6228 6229 6230 6231 6232 6233 6234 6235 6236 6237 6238 6239 6240 6241 6242 6243 6244 6245 6246 6247 6248 6249 6250
	}

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
6251
	for (v = 0; v < pf->num_alloc_vsi; v++) {
N
Neerav Parikh 已提交
6252 6253 6254 6255 6256 6257 6258 6259 6260
		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
6261
			tc_map = I40E_DEFAULT_TRAFFIC_CLASS;
N
Neerav Parikh 已提交
6262 6263 6264 6265 6266 6267 6268 6269

		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 {
6270 6271
			/* Re-configure VSI vectors based on updated TC map */
			i40e_vsi_map_rings_to_vectors(pf->vsi[v]);
N
Neerav Parikh 已提交
6272 6273 6274 6275 6276 6277
			if (pf->vsi[v]->netdev)
				i40e_dcbnl_set_all(pf->vsi[v]);
		}
	}
}

6278 6279 6280 6281 6282 6283 6284 6285 6286 6287 6288 6289 6290 6291 6292
/**
 * 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,
6293 6294 6295
			 "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));
6296
		/* Schedule PF reset to recover */
6297
		set_bit(__I40E_PF_RESET_REQUESTED, pf->state);
6298 6299 6300 6301 6302 6303
		i40e_service_event_schedule(pf);
	}

	return ret;
}

N
Neerav Parikh 已提交
6304 6305 6306 6307 6308 6309 6310 6311 6312 6313 6314 6315
/**
 * 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;

6316 6317 6318 6319 6320
	/* Do not enable DCB for SW1 and SW2 images even if the FW is capable
	 * Also do not enable DCBx if FW LLDP agent is disabled
	 */
	if ((pf->hw_features & I40E_HW_NO_DCB_SUPPORT) ||
	    (pf->flags & I40E_FLAG_DISABLE_FW_LLDP))
6321 6322
		goto out;

N
Neerav Parikh 已提交
6323 6324 6325 6326 6327 6328 6329 6330 6331 6332 6333 6334
	/* 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;
6335 6336

			pf->flags |= I40E_FLAG_DCB_CAPABLE;
6337 6338 6339
			/* Enable DCB tagging only when more than one TC
			 * or explicitly disable if only one TC
			 */
6340 6341
			if (i40e_dcb_get_num_tc(&hw->local_dcbx_config) > 1)
				pf->flags |= I40E_FLAG_DCB_ENABLED;
6342 6343
			else
				pf->flags &= ~I40E_FLAG_DCB_ENABLED;
6344 6345
			dev_dbg(&pf->pdev->dev,
				"DCBX offload is supported for this PF.\n");
N
Neerav Parikh 已提交
6346
		}
A
Alan Brady 已提交
6347 6348 6349
	} else if (pf->hw.aq.asq_last_status == I40E_AQ_RC_EPERM) {
		dev_info(&pf->pdev->dev, "FW LLDP disabled for this PF.\n");
		pf->flags |= I40E_FLAG_DISABLE_FW_LLDP;
6350
	} else {
6351
		dev_info(&pf->pdev->dev,
6352 6353 6354
			 "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
6355 6356 6357 6358 6359 6360
	}

out:
	return err;
}
#endif /* CONFIG_I40E_DCB */
J
Jesse Brandeburg 已提交
6361 6362 6363 6364 6365 6366
#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
6367
void i40e_print_link_message(struct i40e_vsi *vsi, bool isup)
J
Jesse Brandeburg 已提交
6368
{
6369
	enum i40e_aq_link_speed new_speed;
6370
	struct i40e_pf *pf = vsi->back;
6371 6372
	char *speed = "Unknown";
	char *fc = "Unknown";
6373
	char *fec = "";
6374
	char *req_fec = "";
6375
	char *an = "";
J
Jesse Brandeburg 已提交
6376

6377
	new_speed = pf->hw.phy.link_info.link_speed;
6378 6379

	if ((vsi->current_isup == isup) && (vsi->current_speed == new_speed))
M
Matt Jared 已提交
6380 6381
		return;
	vsi->current_isup = isup;
6382
	vsi->current_speed = new_speed;
J
Jesse Brandeburg 已提交
6383 6384 6385 6386 6387
	if (!isup) {
		netdev_info(vsi->netdev, "NIC Link is Down\n");
		return;
	}

6388 6389 6390
	/* Warn user if link speed on NPAR enabled partition is not at
	 * least 10GB
	 */
6391 6392 6393
	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))
6394 6395 6396
		netdev_warn(vsi->netdev,
			    "The partition detected link speed that is less than 10Gbps\n");

6397
	switch (pf->hw.phy.link_info.link_speed) {
J
Jesse Brandeburg 已提交
6398
	case I40E_LINK_SPEED_40GB:
6399
		speed = "40 G";
J
Jesse Brandeburg 已提交
6400
		break;
6401
	case I40E_LINK_SPEED_20GB:
6402
		speed = "20 G";
6403
		break;
6404 6405 6406
	case I40E_LINK_SPEED_25GB:
		speed = "25 G";
		break;
J
Jesse Brandeburg 已提交
6407
	case I40E_LINK_SPEED_10GB:
6408
		speed = "10 G";
J
Jesse Brandeburg 已提交
6409 6410
		break;
	case I40E_LINK_SPEED_1GB:
6411
		speed = "1000 M";
J
Jesse Brandeburg 已提交
6412
		break;
M
Mitch Williams 已提交
6413
	case I40E_LINK_SPEED_100MB:
6414
		speed = "100 M";
M
Mitch Williams 已提交
6415
		break;
J
Jesse Brandeburg 已提交
6416 6417 6418 6419
	default:
		break;
	}

6420
	switch (pf->hw.fc.current_mode) {
J
Jesse Brandeburg 已提交
6421
	case I40E_FC_FULL:
6422
		fc = "RX/TX";
J
Jesse Brandeburg 已提交
6423 6424
		break;
	case I40E_FC_TX_PAUSE:
6425
		fc = "TX";
J
Jesse Brandeburg 已提交
6426 6427
		break;
	case I40E_FC_RX_PAUSE:
6428
		fc = "RX";
J
Jesse Brandeburg 已提交
6429 6430
		break;
	default:
6431
		fc = "None";
J
Jesse Brandeburg 已提交
6432 6433 6434
		break;
	}

6435
	if (pf->hw.phy.link_info.link_speed == I40E_LINK_SPEED_25GB) {
6436
		req_fec = ", Requested FEC: None";
6437 6438 6439
		fec = ", FEC: None";
		an = ", Autoneg: False";

6440
		if (pf->hw.phy.link_info.an_info & I40E_AQ_AN_COMPLETED)
6441 6442
			an = ", Autoneg: True";

6443
		if (pf->hw.phy.link_info.fec_info &
6444 6445
		    I40E_AQ_CONFIG_FEC_KR_ENA)
			fec = ", FEC: CL74 FC-FEC/BASE-R";
6446
		else if (pf->hw.phy.link_info.fec_info &
6447 6448
			 I40E_AQ_CONFIG_FEC_RS_ENA)
			fec = ", FEC: CL108 RS-FEC";
6449 6450 6451 6452 6453 6454 6455 6456 6457 6458 6459 6460

		/* '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";
		}
6461 6462
	}

6463 6464
	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
6465
}
N
Neerav Parikh 已提交
6466

J
Jesse Brandeburg 已提交
6467 6468 6469 6470 6471 6472 6473 6474 6475 6476 6477 6478 6479 6480 6481
/**
 * 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
6482
	err = i40e_vsi_start_rings(vsi);
J
Jesse Brandeburg 已提交
6483 6484 6485
	if (err)
		return err;

6486
	clear_bit(__I40E_VSI_DOWN, vsi->state);
J
Jesse Brandeburg 已提交
6487 6488 6489 6490 6491
	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
6492
		i40e_print_link_message(vsi, true);
J
Jesse Brandeburg 已提交
6493 6494 6495
		netif_tx_start_all_queues(vsi->netdev);
		netif_carrier_on(vsi->netdev);
	}
6496 6497

	/* replay FDIR SB filters */
6498 6499
	if (vsi->type == I40E_VSI_FDIR) {
		/* reset fd counters */
6500 6501
		pf->fd_add_err = 0;
		pf->fd_atr_cnt = 0;
6502
		i40e_fdir_filter_restore(vsi);
6503
	}
6504 6505 6506 6507 6508

	/* 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
6509 6510 6511 6512 6513 6514 6515 6516 6517 6518 6519 6520 6521 6522 6523 6524 6525
	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());
6526
	while (test_and_set_bit(__I40E_CONFIG_BUSY, pf->state))
J
Jesse Brandeburg 已提交
6527 6528 6529 6530
		usleep_range(1000, 2000);
	i40e_down(vsi);

	i40e_up(vsi);
6531
	clear_bit(__I40E_CONFIG_BUSY, pf->state);
J
Jesse Brandeburg 已提交
6532 6533 6534 6535 6536 6537 6538 6539 6540 6541 6542 6543 6544 6545 6546 6547 6548 6549 6550 6551 6552 6553 6554 6555 6556 6557
}

/**
 * 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
6558
	 * sets the vsi->state __I40E_VSI_DOWN bit.
J
Jesse Brandeburg 已提交
6559 6560 6561 6562 6563 6564
	 */
	if (vsi->netdev) {
		netif_carrier_off(vsi->netdev);
		netif_tx_disable(vsi->netdev);
	}
	i40e_vsi_disable_irq(vsi);
F
Filip Sadowski 已提交
6565
	i40e_vsi_stop_rings(vsi);
J
Jesse Brandeburg 已提交
6566 6567 6568
	i40e_napi_disable_all(vsi);

	for (i = 0; i < vsi->num_queue_pairs; i++) {
6569
		i40e_clean_tx_ring(vsi->tx_rings[i]);
6570 6571
		if (i40e_enabled_xdp_vsi(vsi))
			i40e_clean_tx_ring(vsi->xdp_rings[i]);
6572
		i40e_clean_rx_ring(vsi->rx_rings[i]);
J
Jesse Brandeburg 已提交
6573
	}
6574

J
Jesse Brandeburg 已提交
6575 6576
}

6577 6578 6579 6580 6581 6582 6583 6584
/**
 * i40e_validate_mqprio_qopt- validate queue mapping info
 * @vsi: the VSI being configured
 * @mqprio_qopt: queue parametrs
 **/
static int i40e_validate_mqprio_qopt(struct i40e_vsi *vsi,
				     struct tc_mqprio_qopt_offload *mqprio_qopt)
{
6585
	u64 sum_max_rate = 0;
A
Alan Brady 已提交
6586
	u64 max_rate = 0;
6587 6588 6589 6590 6591 6592 6593 6594 6595
	int i;

	if (mqprio_qopt->qopt.offset[0] != 0 ||
	    mqprio_qopt->qopt.num_tc < 1 ||
	    mqprio_qopt->qopt.num_tc > I40E_MAX_TRAFFIC_CLASS)
		return -EINVAL;
	for (i = 0; ; i++) {
		if (!mqprio_qopt->qopt.count[i])
			return -EINVAL;
6596 6597 6598
		if (mqprio_qopt->min_rate[i]) {
			dev_err(&vsi->back->pdev->dev,
				"Invalid min tx rate (greater than 0) specified\n");
6599
			return -EINVAL;
6600
		}
A
Alan Brady 已提交
6601 6602 6603
		max_rate = mqprio_qopt->max_rate[i];
		do_div(max_rate, I40E_BW_MBPS_DIVISOR);
		sum_max_rate += max_rate;
6604

6605 6606 6607 6608 6609 6610 6611 6612 6613 6614
		if (i >= mqprio_qopt->qopt.num_tc - 1)
			break;
		if (mqprio_qopt->qopt.offset[i + 1] !=
		    (mqprio_qopt->qopt.offset[i] + mqprio_qopt->qopt.count[i]))
			return -EINVAL;
	}
	if (vsi->num_queue_pairs <
	    (mqprio_qopt->qopt.offset[i] + mqprio_qopt->qopt.count[i])) {
		return -EINVAL;
	}
6615 6616 6617 6618 6619
	if (sum_max_rate > i40e_get_link_speed(vsi)) {
		dev_err(&vsi->back->pdev->dev,
			"Invalid max tx rate specified\n");
		return -EINVAL;
	}
6620 6621 6622 6623 6624 6625 6626 6627 6628 6629 6630 6631 6632 6633 6634 6635 6636 6637 6638 6639 6640 6641 6642 6643 6644 6645 6646 6647 6648 6649
	return 0;
}

/**
 * i40e_vsi_set_default_tc_config - set default values for tc configuration
 * @vsi: the VSI being configured
 **/
static void i40e_vsi_set_default_tc_config(struct i40e_vsi *vsi)
{
	u16 qcount;
	int i;

	/* Only TC0 is enabled */
	vsi->tc_config.numtc = 1;
	vsi->tc_config.enabled_tc = 1;
	qcount = min_t(int, vsi->alloc_queue_pairs,
		       i40e_pf_get_max_q_per_tc(vsi->back));
	for (i = 0; i < I40E_MAX_TRAFFIC_CLASS; i++) {
		/* For the TC that is not enabled set the offset to to default
		 * queue and allocate one queue for the given TC.
		 */
		vsi->tc_config.tc_info[i].qoffset = 0;
		if (i == 0)
			vsi->tc_config.tc_info[i].qcount = qcount;
		else
			vsi->tc_config.tc_info[i].qcount = 1;
		vsi->tc_config.tc_info[i].netdev_tc = 0;
	}
}

J
Jesse Brandeburg 已提交
6650 6651 6652
/**
 * i40e_setup_tc - configure multiple traffic classes
 * @netdev: net device to configure
6653
 * @type_data: tc offload data
J
Jesse Brandeburg 已提交
6654
 **/
6655
static int i40e_setup_tc(struct net_device *netdev, void *type_data)
J
Jesse Brandeburg 已提交
6656
{
6657
	struct tc_mqprio_qopt_offload *mqprio_qopt = type_data;
J
Jesse Brandeburg 已提交
6658 6659 6660
	struct i40e_netdev_priv *np = netdev_priv(netdev);
	struct i40e_vsi *vsi = np->vsi;
	struct i40e_pf *pf = vsi->back;
6661 6662
	u8 enabled_tc = 0, num_tc, hw;
	bool need_reset = false;
J
Jesse Brandeburg 已提交
6663
	int ret = -EINVAL;
6664
	u16 mode;
J
Jesse Brandeburg 已提交
6665 6666
	int i;

6667 6668 6669 6670 6671 6672 6673
	num_tc = mqprio_qopt->qopt.num_tc;
	hw = mqprio_qopt->qopt.hw;
	mode = mqprio_qopt->mode;
	if (!hw) {
		pf->flags &= ~I40E_FLAG_TC_MQPRIO;
		memcpy(&vsi->mqprio_qopt, mqprio_qopt, sizeof(*mqprio_qopt));
		goto config_tc;
J
Jesse Brandeburg 已提交
6674 6675 6676 6677
	}

	/* Check if MFP enabled */
	if (pf->flags & I40E_FLAG_MFP_ENABLED) {
6678 6679 6680
		netdev_info(netdev,
			    "Configuring TC not supported in MFP mode\n");
		return ret;
J
Jesse Brandeburg 已提交
6681
	}
6682 6683 6684
	switch (mode) {
	case TC_MQPRIO_MODE_DCB:
		pf->flags &= ~I40E_FLAG_TC_MQPRIO;
J
Jesse Brandeburg 已提交
6685

6686 6687 6688 6689 6690 6691 6692 6693 6694 6695 6696 6697 6698 6699 6700 6701 6702 6703 6704 6705 6706 6707 6708 6709 6710 6711 6712 6713 6714 6715 6716 6717
		/* Check if DCB enabled to continue */
		if (!(pf->flags & I40E_FLAG_DCB_ENABLED)) {
			netdev_info(netdev,
				    "DCB is not enabled for adapter\n");
			return ret;
		}

		/* Check whether tc count is within enabled limit */
		if (num_tc > i40e_pf_get_num_tc(pf)) {
			netdev_info(netdev,
				    "TC count greater than enabled on link for adapter\n");
			return ret;
		}
		break;
	case TC_MQPRIO_MODE_CHANNEL:
		if (pf->flags & I40E_FLAG_DCB_ENABLED) {
			netdev_info(netdev,
				    "Full offload of TC Mqprio options is not supported when DCB is enabled\n");
			return ret;
		}
		if (!(pf->flags & I40E_FLAG_MSIX_ENABLED))
			return ret;
		ret = i40e_validate_mqprio_qopt(vsi, mqprio_qopt);
		if (ret)
			return ret;
		memcpy(&vsi->mqprio_qopt, mqprio_qopt,
		       sizeof(*mqprio_qopt));
		pf->flags |= I40E_FLAG_TC_MQPRIO;
		pf->flags &= ~I40E_FLAG_DCB_ENABLED;
		break;
	default:
		return -EINVAL;
J
Jesse Brandeburg 已提交
6718 6719
	}

6720
config_tc:
J
Jesse Brandeburg 已提交
6721
	/* Generate TC map for number of tc requested */
6722
	for (i = 0; i < num_tc; i++)
J
Jesse Brandeburg 已提交
6723
		enabled_tc |= BIT(i);
J
Jesse Brandeburg 已提交
6724 6725

	/* Requesting same TC configuration as already enabled */
6726 6727
	if (enabled_tc == vsi->tc_config.enabled_tc &&
	    mode != TC_MQPRIO_MODE_CHANNEL)
J
Jesse Brandeburg 已提交
6728 6729 6730 6731 6732
		return 0;

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

6733 6734 6735
	if (!hw && !(pf->flags & I40E_FLAG_TC_MQPRIO))
		i40e_remove_queue_channels(vsi);

J
Jesse Brandeburg 已提交
6736 6737 6738 6739 6740
	/* 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);
6741
		need_reset = true;
J
Jesse Brandeburg 已提交
6742 6743 6744
		goto exit;
	}

6745
	if (pf->flags & I40E_FLAG_TC_MQPRIO) {
6746
		if (vsi->mqprio_qopt.max_rate[0]) {
A
Alan Brady 已提交
6747 6748 6749
			u64 max_tx_rate = vsi->mqprio_qopt.max_rate[0];

			do_div(max_tx_rate, I40E_BW_MBPS_DIVISOR);
6750 6751
			ret = i40e_set_bw_limit(vsi, vsi->seid, max_tx_rate);
			if (!ret) {
A
Alan Brady 已提交
6752 6753 6754
				u64 credits = max_tx_rate;

				do_div(credits, I40E_BW_CREDIT_DIVISOR);
6755 6756 6757
				dev_dbg(&vsi->back->pdev->dev,
					"Set tx rate of %llu Mbps (count of 50Mbps %llu) for vsi->seid %u\n",
					max_tx_rate,
A
Alan Brady 已提交
6758
					credits,
6759 6760 6761 6762 6763 6764
					vsi->seid);
			} else {
				need_reset = true;
				goto exit;
			}
		}
6765
		ret = i40e_configure_queue_channels(vsi);
N
Neerav Parikh 已提交
6766
		if (ret) {
6767 6768
			netdev_info(netdev,
				    "Failed configuring queue channels\n");
6769
			need_reset = true;
6770
			goto exit;
N
Neerav Parikh 已提交
6771 6772 6773
		}
	}

J
Jesse Brandeburg 已提交
6774
exit:
6775 6776 6777 6778 6779
	/* Reset the configuration data to defaults, only TC0 is enabled */
	if (need_reset) {
		i40e_vsi_set_default_tc_config(vsi);
		need_reset = false;
	}
N
Neerav Parikh 已提交
6780

6781 6782
	/* Unquiesce VSI */
	i40e_unquiesce_vsi(vsi);
J
Jesse Brandeburg 已提交
6783 6784
	return ret;
}
N
Neerav Parikh 已提交
6785

6786 6787 6788 6789 6790 6791 6792 6793 6794 6795 6796 6797 6798 6799 6800 6801 6802 6803 6804 6805 6806 6807 6808 6809 6810 6811 6812 6813
/**
 * i40e_set_cld_element - sets cloud filter element data
 * @filter: cloud filter rule
 * @cld: ptr to cloud filter element data
 *
 * This is helper function to copy data into cloud filter element
 **/
static inline void
i40e_set_cld_element(struct i40e_cloud_filter *filter,
		     struct i40e_aqc_cloud_filters_element_data *cld)
{
	int i, j;
	u32 ipa;

	memset(cld, 0, sizeof(*cld));
	ether_addr_copy(cld->outer_mac, filter->dst_mac);
	ether_addr_copy(cld->inner_mac, filter->src_mac);

	if (filter->n_proto != ETH_P_IP && filter->n_proto != ETH_P_IPV6)
		return;

	if (filter->n_proto == ETH_P_IPV6) {
#define IPV6_MAX_INDEX	(ARRAY_SIZE(filter->dst_ipv6) - 1)
		for (i = 0, j = 0; i < ARRAY_SIZE(filter->dst_ipv6);
		     i++, j += 2) {
			ipa = be32_to_cpu(filter->dst_ipv6[IPV6_MAX_INDEX - i]);
			ipa = cpu_to_le32(ipa);
			memcpy(&cld->ipaddr.raw_v6.data[j], &ipa, sizeof(ipa));
N
Neerav Parikh 已提交
6814
		}
6815 6816 6817
	} else {
		ipa = be32_to_cpu(filter->dst_ipv4);
		memcpy(&cld->ipaddr.v4.data, &ipa, sizeof(ipa));
N
Neerav Parikh 已提交
6818
	}
6819 6820 6821 6822 6823 6824 6825 6826

	cld->inner_vlan = cpu_to_le16(ntohs(filter->vlan_id));

	/* tenant_id is not supported by FW now, once the support is enabled
	 * fill the cld->tenant_id with cpu_to_le32(filter->tenant_id)
	 */
	if (filter->tenant_id)
		return;
N
Neerav Parikh 已提交
6827 6828
}

6829
/**
6830 6831 6832 6833
 * i40e_add_del_cloud_filter - Add/del cloud filter
 * @vsi: pointer to VSI
 * @filter: cloud filter rule
 * @add: if true, add, if false, delete
6834
 *
6835 6836
 * Add or delete a cloud filter for a specific flow spec.
 * Returns 0 if the filter were successfully added.
6837
 **/
6838 6839
int i40e_add_del_cloud_filter(struct i40e_vsi *vsi,
			      struct i40e_cloud_filter *filter, bool add)
6840
{
6841 6842
	struct i40e_aqc_cloud_filters_element_data cld_filter;
	struct i40e_pf *pf = vsi->back;
6843
	int ret;
6844 6845 6846 6847 6848 6849 6850 6851 6852 6853 6854 6855 6856 6857 6858 6859 6860 6861 6862 6863 6864 6865 6866 6867 6868 6869 6870 6871 6872 6873 6874 6875 6876
	static const u16 flag_table[128] = {
		[I40E_CLOUD_FILTER_FLAGS_OMAC]  =
			I40E_AQC_ADD_CLOUD_FILTER_OMAC,
		[I40E_CLOUD_FILTER_FLAGS_IMAC]  =
			I40E_AQC_ADD_CLOUD_FILTER_IMAC,
		[I40E_CLOUD_FILTER_FLAGS_IMAC_IVLAN]  =
			I40E_AQC_ADD_CLOUD_FILTER_IMAC_IVLAN,
		[I40E_CLOUD_FILTER_FLAGS_IMAC_TEN_ID] =
			I40E_AQC_ADD_CLOUD_FILTER_IMAC_TEN_ID,
		[I40E_CLOUD_FILTER_FLAGS_OMAC_TEN_ID_IMAC] =
			I40E_AQC_ADD_CLOUD_FILTER_OMAC_TEN_ID_IMAC,
		[I40E_CLOUD_FILTER_FLAGS_IMAC_IVLAN_TEN_ID] =
			I40E_AQC_ADD_CLOUD_FILTER_IMAC_IVLAN_TEN_ID,
		[I40E_CLOUD_FILTER_FLAGS_IIP] =
			I40E_AQC_ADD_CLOUD_FILTER_IIP,
	};

	if (filter->flags >= ARRAY_SIZE(flag_table))
		return I40E_ERR_CONFIG;

	/* copy element needed to add cloud filter from filter */
	i40e_set_cld_element(filter, &cld_filter);

	if (filter->tunnel_type != I40E_CLOUD_TNL_TYPE_NONE)
		cld_filter.flags = cpu_to_le16(filter->tunnel_type <<
					     I40E_AQC_ADD_CLOUD_TNL_TYPE_SHIFT);

	if (filter->n_proto == ETH_P_IPV6)
		cld_filter.flags |= cpu_to_le16(flag_table[filter->flags] |
						I40E_AQC_ADD_CLOUD_FLAGS_IPV6);
	else
		cld_filter.flags |= cpu_to_le16(flag_table[filter->flags] |
						I40E_AQC_ADD_CLOUD_FLAGS_IPV4);
6877

6878 6879 6880 6881 6882 6883 6884 6885 6886 6887 6888 6889
	if (add)
		ret = i40e_aq_add_cloud_filters(&pf->hw, filter->seid,
						&cld_filter, 1);
	else
		ret = i40e_aq_rem_cloud_filters(&pf->hw, filter->seid,
						&cld_filter, 1);
	if (ret)
		dev_dbg(&pf->pdev->dev,
			"Failed to %s cloud filter using l4 port %u, err %d aq_err %d\n",
			add ? "add" : "delete", filter->dst_port, ret,
			pf->hw.aq.asq_last_status);
	else
6890
		dev_info(&pf->pdev->dev,
6891 6892
			 "%s cloud filter for VSI: %d\n",
			 add ? "Added" : "Deleted", filter->seid);
6893 6894 6895
	return ret;
}

N
Neerav Parikh 已提交
6896
/**
6897 6898 6899 6900
 * i40e_add_del_cloud_filter_big_buf - Add/del cloud filter using big_buf
 * @vsi: pointer to VSI
 * @filter: cloud filter rule
 * @add: if true, add, if false, delete
N
Neerav Parikh 已提交
6901
 *
6902 6903
 * Add or delete a cloud filter for a specific flow spec using big buffer.
 * Returns 0 if the filter were successfully added.
N
Neerav Parikh 已提交
6904
 **/
6905 6906 6907
int i40e_add_del_cloud_filter_big_buf(struct i40e_vsi *vsi,
				      struct i40e_cloud_filter *filter,
				      bool add)
N
Neerav Parikh 已提交
6908
{
6909 6910 6911
	struct i40e_aqc_cloud_filters_element_bb cld_filter;
	struct i40e_pf *pf = vsi->back;
	int ret;
N
Neerav Parikh 已提交
6912

6913 6914 6915 6916 6917
	/* Both (src/dst) valid mac_addr are not supported */
	if ((is_valid_ether_addr(filter->dst_mac) &&
	     is_valid_ether_addr(filter->src_mac)) ||
	    (is_multicast_ether_addr(filter->dst_mac) &&
	     is_multicast_ether_addr(filter->src_mac)))
6918
		return -EOPNOTSUPP;
6919

6920 6921
	/* Big buffer cloud filter needs 'L4 port' to be non-zero. Also, UDP
	 * ports are not supported via big buffer now.
6922
	 */
6923 6924
	if (!filter->dst_port || filter->ip_proto == IPPROTO_UDP)
		return -EOPNOTSUPP;
6925

6926 6927 6928
	/* adding filter using src_port/src_ip is not supported at this stage */
	if (filter->src_port || filter->src_ipv4 ||
	    !ipv6_addr_any(&filter->ip.v6.src_ip6))
6929
		return -EOPNOTSUPP;
6930 6931 6932 6933 6934 6935 6936 6937 6938 6939

	/* copy element needed to add cloud filter from filter */
	i40e_set_cld_element(filter, &cld_filter.element);

	if (is_valid_ether_addr(filter->dst_mac) ||
	    is_valid_ether_addr(filter->src_mac) ||
	    is_multicast_ether_addr(filter->dst_mac) ||
	    is_multicast_ether_addr(filter->src_mac)) {
		/* MAC + IP : unsupported mode */
		if (filter->dst_ipv4)
6940
			return -EOPNOTSUPP;
6941 6942 6943 6944 6945 6946 6947 6948 6949 6950 6951

		/* since we validated that L4 port must be valid before
		 * we get here, start with respective "flags" value
		 * and update if vlan is present or not
		 */
		cld_filter.element.flags =
			cpu_to_le16(I40E_AQC_ADD_CLOUD_FILTER_MAC_PORT);

		if (filter->vlan_id) {
			cld_filter.element.flags =
			cpu_to_le16(I40E_AQC_ADD_CLOUD_FILTER_MAC_VLAN_PORT);
N
Neerav Parikh 已提交
6952
		}
6953 6954 6955 6956 6957 6958 6959 6960 6961 6962 6963

	} else if (filter->dst_ipv4 ||
		   !ipv6_addr_any(&filter->ip.v6.dst_ip6)) {
		cld_filter.element.flags =
				cpu_to_le16(I40E_AQC_ADD_CLOUD_FILTER_IP_PORT);
		if (filter->n_proto == ETH_P_IPV6)
			cld_filter.element.flags |=
				cpu_to_le16(I40E_AQC_ADD_CLOUD_FLAGS_IPV6);
		else
			cld_filter.element.flags |=
				cpu_to_le16(I40E_AQC_ADD_CLOUD_FLAGS_IPV4);
6964
	} else {
6965 6966 6967
		dev_err(&pf->pdev->dev,
			"either mac or ip has to be valid for cloud filter\n");
		return -EINVAL;
N
Neerav Parikh 已提交
6968 6969
	}

6970 6971 6972 6973 6974 6975 6976 6977 6978 6979 6980 6981 6982 6983 6984 6985 6986 6987 6988 6989 6990 6991 6992 6993 6994 6995 6996 6997 6998 6999 7000
	/* Now copy L4 port in Byte 6..7 in general fields */
	cld_filter.general_fields[I40E_AQC_ADD_CLOUD_FV_FLU_0X16_WORD0] =
						be16_to_cpu(filter->dst_port);

	if (add) {
		/* Validate current device switch mode, change if necessary */
		ret = i40e_validate_and_set_switch_mode(vsi);
		if (ret) {
			dev_err(&pf->pdev->dev,
				"failed to set switch mode, ret %d\n",
				ret);
			return ret;
		}

		ret = i40e_aq_add_cloud_filters_bb(&pf->hw, filter->seid,
						   &cld_filter, 1);
	} else {
		ret = i40e_aq_rem_cloud_filters_bb(&pf->hw, filter->seid,
						   &cld_filter, 1);
	}

	if (ret)
		dev_dbg(&pf->pdev->dev,
			"Failed to %s cloud filter(big buffer) err %d aq_err %d\n",
			add ? "add" : "delete", ret, pf->hw.aq.asq_last_status);
	else
		dev_info(&pf->pdev->dev,
			 "%s cloud filter for VSI: %d, L4 port: %d\n",
			 add ? "add" : "delete", filter->seid,
			 ntohs(filter->dst_port));
	return ret;
N
Neerav Parikh 已提交
7001
}
7002

J
Jesse Brandeburg 已提交
7003
/**
7004 7005 7006 7007 7008 7009 7010 7011 7012
 * i40e_parse_cls_flower - Parse tc flower filters provided by kernel
 * @vsi: Pointer to VSI
 * @cls_flower: Pointer to struct tc_cls_flower_offload
 * @filter: Pointer to cloud filter structure
 *
 **/
static int i40e_parse_cls_flower(struct i40e_vsi *vsi,
				 struct tc_cls_flower_offload *f,
				 struct i40e_cloud_filter *filter)
J
Jesse Brandeburg 已提交
7013
{
7014 7015 7016 7017 7018 7019 7020 7021 7022 7023 7024 7025 7026 7027 7028 7029 7030
	u16 n_proto_mask = 0, n_proto_key = 0, addr_type = 0;
	struct i40e_pf *pf = vsi->back;
	u8 field_flags = 0;

	if (f->dissector->used_keys &
	    ~(BIT(FLOW_DISSECTOR_KEY_CONTROL) |
	      BIT(FLOW_DISSECTOR_KEY_BASIC) |
	      BIT(FLOW_DISSECTOR_KEY_ETH_ADDRS) |
	      BIT(FLOW_DISSECTOR_KEY_VLAN) |
	      BIT(FLOW_DISSECTOR_KEY_IPV4_ADDRS) |
	      BIT(FLOW_DISSECTOR_KEY_IPV6_ADDRS) |
	      BIT(FLOW_DISSECTOR_KEY_PORTS) |
	      BIT(FLOW_DISSECTOR_KEY_ENC_KEYID))) {
		dev_err(&pf->pdev->dev, "Unsupported key used: 0x%x\n",
			f->dissector->used_keys);
		return -EOPNOTSUPP;
	}
J
Jesse Brandeburg 已提交
7031

7032 7033 7034 7035 7036
	if (dissector_uses_key(f->dissector, FLOW_DISSECTOR_KEY_ENC_KEYID)) {
		struct flow_dissector_key_keyid *key =
			skb_flow_dissector_target(f->dissector,
						  FLOW_DISSECTOR_KEY_ENC_KEYID,
						  f->key);
7037

7038 7039 7040 7041 7042 7043 7044 7045 7046
		struct flow_dissector_key_keyid *mask =
			skb_flow_dissector_target(f->dissector,
						  FLOW_DISSECTOR_KEY_ENC_KEYID,
						  f->mask);

		if (mask->keyid != 0)
			field_flags |= I40E_CLOUD_FIELD_TEN_ID;

		filter->tenant_id = be32_to_cpu(key->keyid);
J
Jesse Brandeburg 已提交
7047 7048
	}

7049 7050 7051 7052 7053
	if (dissector_uses_key(f->dissector, FLOW_DISSECTOR_KEY_BASIC)) {
		struct flow_dissector_key_basic *key =
			skb_flow_dissector_target(f->dissector,
						  FLOW_DISSECTOR_KEY_BASIC,
						  f->key);
7054

7055 7056 7057 7058 7059 7060 7061 7062 7063 7064 7065 7066 7067 7068
		struct flow_dissector_key_basic *mask =
			skb_flow_dissector_target(f->dissector,
						  FLOW_DISSECTOR_KEY_BASIC,
						  f->mask);

		n_proto_key = ntohs(key->n_proto);
		n_proto_mask = ntohs(mask->n_proto);

		if (n_proto_key == ETH_P_ALL) {
			n_proto_key = 0;
			n_proto_mask = 0;
		}
		filter->n_proto = n_proto_key & n_proto_mask;
		filter->ip_proto = key->ip_proto;
J
Jesse Brandeburg 已提交
7069 7070
	}

7071 7072 7073 7074 7075 7076 7077 7078 7079 7080 7081 7082 7083 7084 7085 7086 7087 7088 7089 7090 7091 7092 7093 7094 7095 7096 7097 7098 7099 7100 7101 7102 7103
	if (dissector_uses_key(f->dissector, FLOW_DISSECTOR_KEY_ETH_ADDRS)) {
		struct flow_dissector_key_eth_addrs *key =
			skb_flow_dissector_target(f->dissector,
						  FLOW_DISSECTOR_KEY_ETH_ADDRS,
						  f->key);

		struct flow_dissector_key_eth_addrs *mask =
			skb_flow_dissector_target(f->dissector,
						  FLOW_DISSECTOR_KEY_ETH_ADDRS,
						  f->mask);

		/* use is_broadcast and is_zero to check for all 0xf or 0 */
		if (!is_zero_ether_addr(mask->dst)) {
			if (is_broadcast_ether_addr(mask->dst)) {
				field_flags |= I40E_CLOUD_FIELD_OMAC;
			} else {
				dev_err(&pf->pdev->dev, "Bad ether dest mask %pM\n",
					mask->dst);
				return I40E_ERR_CONFIG;
			}
		}

		if (!is_zero_ether_addr(mask->src)) {
			if (is_broadcast_ether_addr(mask->src)) {
				field_flags |= I40E_CLOUD_FIELD_IMAC;
			} else {
				dev_err(&pf->pdev->dev, "Bad ether src mask %pM\n",
					mask->src);
				return I40E_ERR_CONFIG;
			}
		}
		ether_addr_copy(filter->dst_mac, key->dst);
		ether_addr_copy(filter->src_mac, key->src);
J
Jesse Brandeburg 已提交
7104 7105
	}

7106 7107 7108 7109 7110 7111 7112 7113 7114
	if (dissector_uses_key(f->dissector, FLOW_DISSECTOR_KEY_VLAN)) {
		struct flow_dissector_key_vlan *key =
			skb_flow_dissector_target(f->dissector,
						  FLOW_DISSECTOR_KEY_VLAN,
						  f->key);
		struct flow_dissector_key_vlan *mask =
			skb_flow_dissector_target(f->dissector,
						  FLOW_DISSECTOR_KEY_VLAN,
						  f->mask);
7115

7116 7117 7118
		if (mask->vlan_id) {
			if (mask->vlan_id == VLAN_VID_MASK) {
				field_flags |= I40E_CLOUD_FIELD_IVLAN;
7119

7120 7121 7122 7123 7124 7125
			} else {
				dev_err(&pf->pdev->dev, "Bad vlan mask 0x%04x\n",
					mask->vlan_id);
				return I40E_ERR_CONFIG;
			}
		}
7126

7127 7128 7129 7130 7131 7132 7133 7134 7135 7136 7137 7138 7139 7140 7141 7142 7143 7144 7145 7146 7147 7148 7149 7150 7151 7152 7153 7154 7155 7156 7157 7158 7159 7160 7161 7162 7163 7164 7165 7166 7167 7168 7169 7170 7171 7172 7173 7174 7175 7176 7177 7178 7179 7180 7181 7182 7183 7184 7185 7186 7187 7188 7189 7190
		filter->vlan_id = cpu_to_be16(key->vlan_id);
	}

	if (dissector_uses_key(f->dissector, FLOW_DISSECTOR_KEY_CONTROL)) {
		struct flow_dissector_key_control *key =
			skb_flow_dissector_target(f->dissector,
						  FLOW_DISSECTOR_KEY_CONTROL,
						  f->key);

		addr_type = key->addr_type;
	}

	if (addr_type == FLOW_DISSECTOR_KEY_IPV4_ADDRS) {
		struct flow_dissector_key_ipv4_addrs *key =
			skb_flow_dissector_target(f->dissector,
						  FLOW_DISSECTOR_KEY_IPV4_ADDRS,
						  f->key);
		struct flow_dissector_key_ipv4_addrs *mask =
			skb_flow_dissector_target(f->dissector,
						  FLOW_DISSECTOR_KEY_IPV4_ADDRS,
						  f->mask);

		if (mask->dst) {
			if (mask->dst == cpu_to_be32(0xffffffff)) {
				field_flags |= I40E_CLOUD_FIELD_IIP;
			} else {
				mask->dst = be32_to_cpu(mask->dst);
				dev_err(&pf->pdev->dev, "Bad ip dst mask %pI4\n",
					&mask->dst);
				return I40E_ERR_CONFIG;
			}
		}

		if (mask->src) {
			if (mask->src == cpu_to_be32(0xffffffff)) {
				field_flags |= I40E_CLOUD_FIELD_IIP;
			} else {
				mask->src = be32_to_cpu(mask->src);
				dev_err(&pf->pdev->dev, "Bad ip src mask %pI4\n",
					&mask->src);
				return I40E_ERR_CONFIG;
			}
		}

		if (field_flags & I40E_CLOUD_FIELD_TEN_ID) {
			dev_err(&pf->pdev->dev, "Tenant id not allowed for ip filter\n");
			return I40E_ERR_CONFIG;
		}
		filter->dst_ipv4 = key->dst;
		filter->src_ipv4 = key->src;
	}

	if (addr_type == FLOW_DISSECTOR_KEY_IPV6_ADDRS) {
		struct flow_dissector_key_ipv6_addrs *key =
			skb_flow_dissector_target(f->dissector,
						  FLOW_DISSECTOR_KEY_IPV6_ADDRS,
						  f->key);
		struct flow_dissector_key_ipv6_addrs *mask =
			skb_flow_dissector_target(f->dissector,
						  FLOW_DISSECTOR_KEY_IPV6_ADDRS,
						  f->mask);

		/* src and dest IPV6 address should not be LOOPBACK
		 * (0:0:0:0:0:0:0:1), which can be represented as ::1
7191
		 */
7192 7193 7194 7195 7196 7197 7198 7199 7200 7201 7202 7203 7204 7205 7206 7207 7208 7209 7210 7211 7212 7213 7214 7215 7216 7217 7218 7219 7220 7221 7222 7223 7224 7225 7226 7227 7228 7229 7230 7231 7232 7233 7234 7235 7236 7237 7238 7239 7240 7241 7242 7243 7244 7245 7246 7247
		if (ipv6_addr_loopback(&key->dst) ||
		    ipv6_addr_loopback(&key->src)) {
			dev_err(&pf->pdev->dev,
				"Bad ipv6, addr is LOOPBACK\n");
			return I40E_ERR_CONFIG;
		}
		if (!ipv6_addr_any(&mask->dst) || !ipv6_addr_any(&mask->src))
			field_flags |= I40E_CLOUD_FIELD_IIP;

		memcpy(&filter->src_ipv6, &key->src.s6_addr32,
		       sizeof(filter->src_ipv6));
		memcpy(&filter->dst_ipv6, &key->dst.s6_addr32,
		       sizeof(filter->dst_ipv6));
	}

	if (dissector_uses_key(f->dissector, FLOW_DISSECTOR_KEY_PORTS)) {
		struct flow_dissector_key_ports *key =
			skb_flow_dissector_target(f->dissector,
						  FLOW_DISSECTOR_KEY_PORTS,
						  f->key);
		struct flow_dissector_key_ports *mask =
			skb_flow_dissector_target(f->dissector,
						  FLOW_DISSECTOR_KEY_PORTS,
						  f->mask);

		if (mask->src) {
			if (mask->src == cpu_to_be16(0xffff)) {
				field_flags |= I40E_CLOUD_FIELD_IIP;
			} else {
				dev_err(&pf->pdev->dev, "Bad src port mask 0x%04x\n",
					be16_to_cpu(mask->src));
				return I40E_ERR_CONFIG;
			}
		}

		if (mask->dst) {
			if (mask->dst == cpu_to_be16(0xffff)) {
				field_flags |= I40E_CLOUD_FIELD_IIP;
			} else {
				dev_err(&pf->pdev->dev, "Bad dst port mask 0x%04x\n",
					be16_to_cpu(mask->dst));
				return I40E_ERR_CONFIG;
			}
		}

		filter->dst_port = key->dst;
		filter->src_port = key->src;

		switch (filter->ip_proto) {
		case IPPROTO_TCP:
		case IPPROTO_UDP:
			break;
		default:
			dev_err(&pf->pdev->dev,
				"Only UDP and TCP transport are supported\n");
			return -EINVAL;
7248
		}
7249
	}
7250 7251 7252
	filter->flags = field_flags;
	return 0;
}
7253

7254 7255 7256 7257 7258 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
/**
 * i40e_handle_tclass: Forward to a traffic class on the device
 * @vsi: Pointer to VSI
 * @tc: traffic class index on the device
 * @filter: Pointer to cloud filter structure
 *
 **/
static int i40e_handle_tclass(struct i40e_vsi *vsi, u32 tc,
			      struct i40e_cloud_filter *filter)
{
	struct i40e_channel *ch, *ch_tmp;

	/* direct to a traffic class on the same device */
	if (tc == 0) {
		filter->seid = vsi->seid;
		return 0;
	} else if (vsi->tc_config.enabled_tc & BIT(tc)) {
		if (!filter->dst_port) {
			dev_err(&vsi->back->pdev->dev,
				"Specify destination port to direct to traffic class that is not default\n");
			return -EINVAL;
		}
		if (list_empty(&vsi->ch_list))
			return -EINVAL;
		list_for_each_entry_safe(ch, ch_tmp, &vsi->ch_list,
					 list) {
			if (ch->seid == vsi->tc_seid_map[tc])
				filter->seid = ch->seid;
		}
		return 0;
	}
	dev_err(&vsi->back->pdev->dev, "TC is not enabled\n");
	return -EINVAL;
J
Jesse Brandeburg 已提交
7287
}
N
Neerav Parikh 已提交
7288

J
Jesse Brandeburg 已提交
7289
/**
7290 7291 7292 7293
 * i40e_configure_clsflower - Configure tc flower filters
 * @vsi: Pointer to VSI
 * @cls_flower: Pointer to struct tc_cls_flower_offload
 *
J
Jesse Brandeburg 已提交
7294
 **/
7295 7296
static int i40e_configure_clsflower(struct i40e_vsi *vsi,
				    struct tc_cls_flower_offload *cls_flower)
J
Jesse Brandeburg 已提交
7297
{
7298 7299
	int tc = tc_classid_to_hwtc(vsi->netdev, cls_flower->classid);
	struct i40e_cloud_filter *filter = NULL;
J
Jesse Brandeburg 已提交
7300
	struct i40e_pf *pf = vsi->back;
7301
	int err = 0;
J
Jesse Brandeburg 已提交
7302

7303 7304
	if (tc < 0) {
		dev_err(&vsi->back->pdev->dev, "Invalid traffic class\n");
7305
		return -EOPNOTSUPP;
7306
	}
J
Jesse Brandeburg 已提交
7307

7308 7309 7310
	if (test_bit(__I40E_RESET_RECOVERY_PENDING, pf->state) ||
	    test_bit(__I40E_RESET_INTR_RECEIVED, pf->state))
		return -EBUSY;
J
Jesse Brandeburg 已提交
7311

7312 7313 7314 7315 7316 7317
	if (pf->fdir_pf_active_filters ||
	    (!hlist_empty(&pf->fdir_filter_list))) {
		dev_err(&vsi->back->pdev->dev,
			"Flow Director Sideband filters exists, turn ntuple off to configure cloud filters\n");
		return -EINVAL;
	}
J
Jesse Brandeburg 已提交
7318

7319 7320 7321 7322 7323
	if (vsi->back->flags & I40E_FLAG_FD_SB_ENABLED) {
		dev_err(&vsi->back->pdev->dev,
			"Disable Flow Director Sideband, configuring Cloud filters via tc-flower\n");
		vsi->back->flags &= ~I40E_FLAG_FD_SB_ENABLED;
		vsi->back->flags |= I40E_FLAG_FD_SB_TO_CLOUD_FILTER;
J
Jesse Brandeburg 已提交
7324
	}
7325

7326 7327 7328 7329 7330 7331 7332 7333 7334 7335 7336 7337 7338 7339 7340 7341 7342 7343 7344 7345 7346 7347 7348 7349 7350
	filter = kzalloc(sizeof(*filter), GFP_KERNEL);
	if (!filter)
		return -ENOMEM;

	filter->cookie = cls_flower->cookie;

	err = i40e_parse_cls_flower(vsi, cls_flower, filter);
	if (err < 0)
		goto err;

	err = i40e_handle_tclass(vsi, tc, filter);
	if (err < 0)
		goto err;

	/* Add cloud filter */
	if (filter->dst_port)
		err = i40e_add_del_cloud_filter_big_buf(vsi, filter, true);
	else
		err = i40e_add_del_cloud_filter(vsi, filter, true);

	if (err) {
		dev_err(&pf->pdev->dev,
			"Failed to add cloud filter, err %s\n",
			i40e_stat_str(&pf->hw, err));
		goto err;
7351
	}
7352

7353 7354
	/* add filter to the ordered list */
	INIT_HLIST_NODE(&filter->cloud_node);
J
Jesse Brandeburg 已提交
7355

7356 7357 7358 7359 7360 7361 7362 7363
	hlist_add_head(&filter->cloud_node, &pf->cloud_filter_list);

	pf->num_cloud_filters++;

	return err;
err:
	kfree(filter);
	return err;
J
Jesse Brandeburg 已提交
7364 7365 7366
}

/**
7367 7368 7369
 * i40e_find_cloud_filter - Find the could filter in the list
 * @vsi: Pointer to VSI
 * @cookie: filter specific cookie
J
Jesse Brandeburg 已提交
7370 7371
 *
 **/
7372 7373
static struct i40e_cloud_filter *i40e_find_cloud_filter(struct i40e_vsi *vsi,
							unsigned long *cookie)
J
Jesse Brandeburg 已提交
7374
{
7375 7376
	struct i40e_cloud_filter *filter = NULL;
	struct hlist_node *node2;
J
Jesse Brandeburg 已提交
7377

7378 7379 7380 7381 7382
	hlist_for_each_entry_safe(filter, node2,
				  &vsi->back->cloud_filter_list, cloud_node)
		if (!memcmp(cookie, &filter->cookie, sizeof(filter->cookie)))
			return filter;
	return NULL;
J
Jesse Brandeburg 已提交
7383 7384 7385
}

/**
7386 7387 7388 7389
 * i40e_delete_clsflower - Remove tc flower filters
 * @vsi: Pointer to VSI
 * @cls_flower: Pointer to struct tc_cls_flower_offload
 *
J
Jesse Brandeburg 已提交
7390
 **/
7391 7392
static int i40e_delete_clsflower(struct i40e_vsi *vsi,
				 struct tc_cls_flower_offload *cls_flower)
J
Jesse Brandeburg 已提交
7393
{
7394 7395 7396
	struct i40e_cloud_filter *filter = NULL;
	struct i40e_pf *pf = vsi->back;
	int err = 0;
J
Jesse Brandeburg 已提交
7397

7398
	filter = i40e_find_cloud_filter(vsi, &cls_flower->cookie);
J
Jesse Brandeburg 已提交
7399

7400 7401
	if (!filter)
		return -EINVAL;
J
Jesse Brandeburg 已提交
7402

7403
	hash_del(&filter->cloud_node);
J
Jesse Brandeburg 已提交
7404

7405 7406 7407 7408
	if (filter->dst_port)
		err = i40e_add_del_cloud_filter_big_buf(vsi, filter, false);
	else
		err = i40e_add_del_cloud_filter(vsi, filter, false);
J
Jesse Brandeburg 已提交
7409

7410 7411 7412 7413 7414 7415
	kfree(filter);
	if (err) {
		dev_err(&pf->pdev->dev,
			"Failed to delete cloud filter, err %s\n",
			i40e_stat_str(&pf->hw, err));
		return i40e_aq_rc_to_posix(err, pf->hw.aq.asq_last_status);
J
Jesse Brandeburg 已提交
7416
	}
7417

7418 7419 7420 7421 7422 7423 7424 7425 7426
	pf->num_cloud_filters--;
	if (!pf->num_cloud_filters)
		if ((pf->flags & I40E_FLAG_FD_SB_TO_CLOUD_FILTER) &&
		    !(pf->flags & I40E_FLAG_FD_SB_INACTIVE)) {
			pf->flags |= I40E_FLAG_FD_SB_ENABLED;
			pf->flags &= ~I40E_FLAG_FD_SB_TO_CLOUD_FILTER;
			pf->flags &= ~I40E_FLAG_FD_SB_INACTIVE;
		}
	return 0;
J
Jesse Brandeburg 已提交
7427 7428 7429
}

/**
7430
 * i40e_setup_tc_cls_flower - flower classifier offloads
J
Jesse Brandeburg 已提交
7431
 * @netdev: net device to configure
7432
 * @type_data: offload data
J
Jesse Brandeburg 已提交
7433
 **/
7434 7435
static int i40e_setup_tc_cls_flower(struct i40e_netdev_priv *np,
				    struct tc_cls_flower_offload *cls_flower)
J
Jesse Brandeburg 已提交
7436 7437 7438
{
	struct i40e_vsi *vsi = np->vsi;

7439 7440 7441 7442 7443 7444 7445 7446 7447
	switch (cls_flower->command) {
	case TC_CLSFLOWER_REPLACE:
		return i40e_configure_clsflower(vsi, cls_flower);
	case TC_CLSFLOWER_DESTROY:
		return i40e_delete_clsflower(vsi, cls_flower);
	case TC_CLSFLOWER_STATS:
		return -EOPNOTSUPP;
	default:
		return -EINVAL;
J
Jesse Brandeburg 已提交
7448
	}
7449
}
J
Jesse Brandeburg 已提交
7450

7451 7452 7453 7454
static int i40e_setup_tc_block_cb(enum tc_setup_type type, void *type_data,
				  void *cb_priv)
{
	struct i40e_netdev_priv *np = cb_priv;
J
Jesse Brandeburg 已提交
7455

7456 7457 7458
	if (!tc_cls_can_offload_and_chain0(np->vsi->netdev, type_data))
		return -EOPNOTSUPP;

7459 7460 7461
	switch (type) {
	case TC_SETUP_CLSFLOWER:
		return i40e_setup_tc_cls_flower(np, type_data);
J
Jesse Brandeburg 已提交
7462

7463 7464
	default:
		return -EOPNOTSUPP;
J
Jesse Brandeburg 已提交
7465
	}
7466
}
J
Jesse Brandeburg 已提交
7467

7468 7469 7470 7471
static int i40e_setup_tc_block(struct net_device *dev,
			       struct tc_block_offload *f)
{
	struct i40e_netdev_priv *np = netdev_priv(dev);
J
Jesse Brandeburg 已提交
7472

7473 7474 7475 7476 7477 7478 7479 7480 7481 7482 7483 7484 7485
	if (f->binder_type != TCF_BLOCK_BINDER_TYPE_CLSACT_INGRESS)
		return -EOPNOTSUPP;

	switch (f->command) {
	case TC_BLOCK_BIND:
		return tcf_block_cb_register(f->block, i40e_setup_tc_block_cb,
					     np, np);
	case TC_BLOCK_UNBIND:
		tcf_block_cb_unregister(f->block, i40e_setup_tc_block_cb, np);
		return 0;
	default:
		return -EOPNOTSUPP;
	}
J
Jesse Brandeburg 已提交
7486 7487
}

7488
static int __i40e_setup_tc(struct net_device *netdev, enum tc_setup_type type,
7489
			   void *type_data)
7490
{
7491
	switch (type) {
7492
	case TC_SETUP_QDISC_MQPRIO:
7493 7494 7495 7496
		return i40e_setup_tc(netdev, type_data);
	case TC_SETUP_BLOCK:
		return i40e_setup_tc_block(netdev, type_data);
	default:
7497
		return -EOPNOTSUPP;
7498
	}
7499 7500
}

J
Jesse Brandeburg 已提交
7501 7502 7503 7504 7505 7506 7507 7508 7509 7510 7511 7512
/**
 * 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
 **/
7513
int i40e_open(struct net_device *netdev)
J
Jesse Brandeburg 已提交
7514 7515 7516 7517 7518 7519
{
	struct i40e_netdev_priv *np = netdev_priv(netdev);
	struct i40e_vsi *vsi = np->vsi;
	struct i40e_pf *pf = vsi->back;
	int err;

7520
	/* disallow open during test or if eeprom is broken */
7521 7522
	if (test_bit(__I40E_TESTING, pf->state) ||
	    test_bit(__I40E_BAD_EEPROM, pf->state))
J
Jesse Brandeburg 已提交
7523 7524 7525 7526
		return -EBUSY;

	netif_carrier_off(netdev);

7527 7528 7529 7530
	err = i40e_vsi_open(vsi);
	if (err)
		return err;

7531 7532 7533 7534 7535 7536 7537 7538
	/* 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);

7539
	udp_tunnel_get_rx_info(netdev);
7540

7541 7542 7543 7544 7545 7546 7547 7548 7549 7550
	return 0;
}

/**
 * i40e_vsi_open -
 * @vsi: the VSI to open
 *
 * Finish initialization of the VSI.
 *
 * Returns 0 on success, negative value on failure
7551 7552
 *
 * Note: expects to be called while under rtnl_lock()
7553 7554 7555 7556
 **/
int i40e_vsi_open(struct i40e_vsi *vsi)
{
	struct i40e_pf *pf = vsi->back;
7557
	char int_name[I40E_INT_NAME_STR_LEN];
7558 7559
	int err;

J
Jesse Brandeburg 已提交
7560 7561 7562 7563 7564 7565 7566 7567 7568 7569 7570 7571
	/* 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;

7572 7573 7574 7575 7576 7577
	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
7578

7579 7580 7581 7582 7583
		/* 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;
7584

7585 7586 7587 7588
		err = netif_set_real_num_rx_queues(vsi->netdev,
						   vsi->num_queue_pairs);
		if (err)
			goto err_set_queues;
7589 7590

	} else if (vsi->type == I40E_VSI_FDIR) {
7591
		snprintf(int_name, sizeof(int_name) - 1, "%s-%s:fdir",
7592 7593
			 dev_driver_string(&pf->pdev->dev),
			 dev_name(&pf->pdev->dev));
7594
		err = i40e_vsi_request_irq(vsi, int_name);
7595

7596
	} else {
7597
		err = -EINVAL;
7598 7599
		goto err_setup_rx;
	}
7600

J
Jesse Brandeburg 已提交
7601 7602 7603 7604 7605 7606 7607 7608
	err = i40e_up_complete(vsi);
	if (err)
		goto err_up_complete;

	return 0;

err_up_complete:
	i40e_down(vsi);
7609
err_set_queues:
J
Jesse Brandeburg 已提交
7610 7611 7612 7613 7614 7615
	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])
A
Amritha Nambiar 已提交
7616
		i40e_do_reset(pf, I40E_PF_RESET_FLAG, true);
J
Jesse Brandeburg 已提交
7617 7618 7619 7620

	return err;
}

7621 7622
/**
 * i40e_fdir_filter_exit - Cleans up the Flow Director accounting
7623
 * @pf: Pointer to PF
7624 7625 7626 7627 7628 7629 7630
 *
 * 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;
7631
	struct i40e_flex_pit *pit_entry, *tmp;
7632 7633 7634 7635 7636 7637 7638
	struct hlist_node *node2;

	hlist_for_each_entry_safe(filter, node2,
				  &pf->fdir_filter_list, fdir_node) {
		hlist_del(&filter->fdir_node);
		kfree(filter);
	}
7639

7640 7641 7642 7643 7644 7645 7646 7647 7648 7649 7650 7651
	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);

7652
	pf->fdir_pf_active_filters = 0;
7653 7654
	pf->fd_tcp4_filter_cnt = 0;
	pf->fd_udp4_filter_cnt = 0;
7655
	pf->fd_sctp4_filter_cnt = 0;
7656
	pf->fd_ip4_filter_cnt = 0;
7657 7658 7659 7660 7661 7662 7663 7664 7665 7666 7667 7668 7669 7670 7671 7672 7673 7674 7675

	/* 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);
7676 7677 7678

	i40e_write_fd_input_set(pf, I40E_FILTER_PCTYPE_FRAG_IPV4,
				I40E_L3_SRC_MASK | I40E_L3_DST_MASK);
7679 7680
}

A
Amritha Nambiar 已提交
7681 7682 7683 7684 7685 7686 7687 7688 7689 7690 7691 7692 7693 7694 7695 7696 7697 7698
/**
 * i40e_cloud_filter_exit - Cleans up the cloud filters
 * @pf: Pointer to PF
 *
 * This function destroys the hlist where all the cloud filters
 * were saved.
 **/
static void i40e_cloud_filter_exit(struct i40e_pf *pf)
{
	struct i40e_cloud_filter *cfilter;
	struct hlist_node *node;

	hlist_for_each_entry_safe(cfilter, node,
				  &pf->cloud_filter_list, cloud_node) {
		hlist_del(&cfilter->cloud_node);
		kfree(cfilter);
	}
	pf->num_cloud_filters = 0;
7699 7700 7701 7702 7703 7704 7705

	if ((pf->flags & I40E_FLAG_FD_SB_TO_CLOUD_FILTER) &&
	    !(pf->flags & I40E_FLAG_FD_SB_INACTIVE)) {
		pf->flags |= I40E_FLAG_FD_SB_ENABLED;
		pf->flags &= ~I40E_FLAG_FD_SB_TO_CLOUD_FILTER;
		pf->flags &= ~I40E_FLAG_FD_SB_INACTIVE;
	}
A
Amritha Nambiar 已提交
7706 7707
}

J
Jesse Brandeburg 已提交
7708 7709 7710 7711 7712 7713 7714 7715 7716 7717
/**
 * 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
 **/
7718
int i40e_close(struct net_device *netdev)
J
Jesse Brandeburg 已提交
7719 7720 7721 7722
{
	struct i40e_netdev_priv *np = netdev_priv(netdev);
	struct i40e_vsi *vsi = np->vsi;

7723
	i40e_vsi_close(vsi);
J
Jesse Brandeburg 已提交
7724 7725 7726 7727 7728 7729 7730 7731

	return 0;
}

/**
 * i40e_do_reset - Start a PF or Core Reset sequence
 * @pf: board private structure
 * @reset_flags: which reset is requested
7732 7733
 * @lock_acquired: indicates whether or not the lock has been acquired
 * before this function was called.
J
Jesse Brandeburg 已提交
7734 7735 7736 7737 7738
 *
 * 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.
 **/
7739
void i40e_do_reset(struct i40e_pf *pf, u32 reset_flags, bool lock_acquired)
J
Jesse Brandeburg 已提交
7740 7741 7742 7743 7744
{
	u32 val;

	WARN_ON(in_interrupt());

7745

J
Jesse Brandeburg 已提交
7746
	/* do the biggest reset indicated */
7747
	if (reset_flags & BIT_ULL(__I40E_GLOBAL_RESET_REQUESTED)) {
J
Jesse Brandeburg 已提交
7748 7749 7750 7751 7752 7753 7754 7755 7756

		/* 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
7757
		dev_dbg(&pf->pdev->dev, "GlobalR requested\n");
J
Jesse Brandeburg 已提交
7758 7759 7760 7761
		val = rd32(&pf->hw, I40E_GLGEN_RTRIG);
		val |= I40E_GLGEN_RTRIG_GLOBR_MASK;
		wr32(&pf->hw, I40E_GLGEN_RTRIG, val);

7762
	} else if (reset_flags & BIT_ULL(__I40E_CORE_RESET_REQUESTED)) {
J
Jesse Brandeburg 已提交
7763 7764 7765 7766 7767

		/* Request a Core Reset
		 *
		 * Same as Global Reset, except does *not* include the MAC/PHY
		 */
J
Jesse Brandeburg 已提交
7768
		dev_dbg(&pf->pdev->dev, "CoreR requested\n");
J
Jesse Brandeburg 已提交
7769 7770 7771 7772 7773
		val = rd32(&pf->hw, I40E_GLGEN_RTRIG);
		val |= I40E_GLGEN_RTRIG_CORER_MASK;
		wr32(&pf->hw, I40E_GLGEN_RTRIG, val);
		i40e_flush(&pf->hw);

A
Amritha Nambiar 已提交
7774
	} else if (reset_flags & I40E_PF_RESET_FLAG) {
J
Jesse Brandeburg 已提交
7775 7776 7777 7778 7779 7780 7781 7782 7783

		/* 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
7784
		dev_dbg(&pf->pdev->dev, "PFR requested\n");
7785
		i40e_handle_reset_warning(pf, lock_acquired);
J
Jesse Brandeburg 已提交
7786

7787
	} else if (reset_flags & BIT_ULL(__I40E_REINIT_REQUESTED)) {
J
Jesse Brandeburg 已提交
7788 7789 7790 7791 7792
		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
7793
		for (v = 0; v < pf->num_alloc_vsi; v++) {
J
Jesse Brandeburg 已提交
7794
			struct i40e_vsi *vsi = pf->vsi[v];
J
Jesse Brandeburg 已提交
7795

J
Jesse Brandeburg 已提交
7796
			if (vsi != NULL &&
7797
			    test_and_clear_bit(__I40E_VSI_REINIT_REQUESTED,
7798
					       vsi->state))
J
Jesse Brandeburg 已提交
7799 7800
				i40e_vsi_reinit_locked(pf->vsi[v]);
		}
7801
	} else if (reset_flags & BIT_ULL(__I40E_DOWN_REQUESTED)) {
7802 7803 7804 7805 7806 7807
		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
7808

7809
			if (vsi != NULL &&
7810
			    test_and_clear_bit(__I40E_VSI_DOWN_REQUESTED,
7811 7812
					       vsi->state)) {
				set_bit(__I40E_VSI_DOWN, vsi->state);
7813 7814 7815
				i40e_down(vsi);
			}
		}
J
Jesse Brandeburg 已提交
7816 7817 7818 7819 7820 7821
	} else {
		dev_info(&pf->pdev->dev,
			 "bad reset request 0x%08x\n", reset_flags);
	}
}

N
Neerav Parikh 已提交
7822 7823 7824 7825 7826 7827 7828 7829 7830 7831 7832 7833 7834 7835 7836 7837 7838 7839 7840 7841 7842 7843
#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
7844
			dev_dbg(&pf->pdev->dev, "ETS UP2TC changed.\n");
N
Neerav Parikh 已提交
7845 7846 7847 7848 7849
		}

		if (memcmp(&new_cfg->etscfg.tcbwtable,
			   &old_cfg->etscfg.tcbwtable,
			   sizeof(new_cfg->etscfg.tcbwtable)))
J
Jesse Brandeburg 已提交
7850
			dev_dbg(&pf->pdev->dev, "ETS TC BW Table changed.\n");
N
Neerav Parikh 已提交
7851 7852 7853 7854

		if (memcmp(&new_cfg->etscfg.tsatable,
			   &old_cfg->etscfg.tsatable,
			   sizeof(new_cfg->etscfg.tsatable)))
J
Jesse Brandeburg 已提交
7855
			dev_dbg(&pf->pdev->dev, "ETS TSA Table changed.\n");
N
Neerav Parikh 已提交
7856 7857 7858 7859 7860 7861 7862
	}

	/* 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
7863
		dev_dbg(&pf->pdev->dev, "PFC config change detected.\n");
N
Neerav Parikh 已提交
7864 7865 7866 7867 7868
	}

	/* Check if APP Table has changed */
	if (memcmp(&new_cfg->app,
		   &old_cfg->app,
7869
		   sizeof(new_cfg->app))) {
N
Neerav Parikh 已提交
7870
		need_reconfig = true;
J
Jesse Brandeburg 已提交
7871
		dev_dbg(&pf->pdev->dev, "APP Table change detected.\n");
7872
	}
N
Neerav Parikh 已提交
7873

S
Shannon Nelson 已提交
7874
	dev_dbg(&pf->pdev->dev, "dcb need_reconfig=%d\n", need_reconfig);
N
Neerav Parikh 已提交
7875 7876 7877 7878 7879 7880 7881 7882 7883 7884 7885 7886 7887 7888 7889 7890 7891 7892 7893
	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;

7894
	/* Not DCB capable or capability disabled */
7895
	if (!(pf->flags & I40E_FLAG_DCB_CAPABLE))
7896 7897
		return ret;

N
Neerav Parikh 已提交
7898 7899 7900
	/* 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
7901
	dev_dbg(&pf->pdev->dev, "LLDP event mib bridge type 0x%x\n", type);
N
Neerav Parikh 已提交
7902 7903 7904 7905 7906
	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;
7907
	dev_dbg(&pf->pdev->dev,
S
Shannon Nelson 已提交
7908
		"LLDP event mib type %s\n", type ? "remote" : "local");
N
Neerav Parikh 已提交
7909 7910 7911 7912 7913 7914 7915 7916
	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;
	}

7917
	/* Store the old configuration */
7918
	tmp_dcbx_cfg = hw->local_dcbx_config;
7919

7920 7921
	/* Reset the old DCBx configuration data */
	memset(&hw->local_dcbx_config, 0, sizeof(hw->local_dcbx_config));
7922 7923
	/* Get updated DCBX data from firmware */
	ret = i40e_get_dcb_config(&pf->hw);
N
Neerav Parikh 已提交
7924
	if (ret) {
7925 7926 7927 7928
		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
7929 7930 7931 7932
		goto exit;
	}

	/* No change detected in DCBX configs */
7933 7934
	if (!memcmp(&tmp_dcbx_cfg, &hw->local_dcbx_config,
		    sizeof(tmp_dcbx_cfg))) {
J
Jesse Brandeburg 已提交
7935
		dev_dbg(&pf->pdev->dev, "No change detected in DCBX configuration.\n");
N
Neerav Parikh 已提交
7936 7937 7938
		goto exit;
	}

7939 7940
	need_reconfig = i40e_dcb_need_reconfig(pf, &tmp_dcbx_cfg,
					       &hw->local_dcbx_config);
N
Neerav Parikh 已提交
7941

7942
	i40e_dcbnl_flush_apps(pf, &tmp_dcbx_cfg, &hw->local_dcbx_config);
N
Neerav Parikh 已提交
7943 7944 7945 7946

	if (!need_reconfig)
		goto exit;

7947
	/* Enable DCB tagging only when more than one TC */
7948
	if (i40e_dcb_get_num_tc(&hw->local_dcbx_config) > 1)
7949 7950 7951 7952
		pf->flags |= I40E_FLAG_DCB_ENABLED;
	else
		pf->flags &= ~I40E_FLAG_DCB_ENABLED;

7953
	set_bit(__I40E_PORT_SUSPENDED, pf->state);
N
Neerav Parikh 已提交
7954 7955 7956 7957 7958 7959
	/* Reconfiguration needed quiesce all VSIs */
	i40e_pf_quiesce_all_vsi(pf);

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

7960 7961
	ret = i40e_resume_port_tx(pf);

7962
	clear_bit(__I40E_PORT_SUSPENDED, pf->state);
7963
	/* In case of error no point in resuming VSIs */
7964 7965 7966
	if (ret)
		goto exit;

7967 7968
	/* Wait for the PF's queues to be disabled */
	ret = i40e_pf_wait_queues_disabled(pf);
7969 7970
	if (ret) {
		/* Schedule PF reset to recover */
7971
		set_bit(__I40E_PF_RESET_REQUESTED, pf->state);
7972 7973
		i40e_service_event_schedule(pf);
	} else {
7974
		i40e_pf_unquiesce_all_vsi(pf);
M
Mitch Williams 已提交
7975 7976
	pf->flags |= (I40E_FLAG_SERVICE_CLIENT_REQUESTED |
		      I40E_FLAG_CLIENT_L2_CHANGE);
7977 7978
	}

N
Neerav Parikh 已提交
7979 7980 7981 7982 7983
exit:
	return ret;
}
#endif /* CONFIG_I40E_DCB */

7984 7985 7986 7987 7988 7989 7990 7991 7992
/**
 * 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();
7993
	i40e_do_reset(pf, reset_flags, true);
7994 7995 7996
	rtnl_unlock();
}

J
Jesse Brandeburg 已提交
7997 7998 7999 8000 8001 8002 8003 8004 8005 8006 8007 8008 8009 8010 8011 8012 8013 8014 8015
/**
 * 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
8016 8017
	dev_dbg(&pf->pdev->dev, "overflow Rx Queue Number = %d QTX_CTL=0x%08x\n",
		queue, qtx_ctl);
J
Jesse Brandeburg 已提交
8018 8019 8020 8021 8022 8023 8024 8025 8026 8027 8028 8029 8030 8031 8032

	/* 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);
	}
}

8033
/**
8034 8035 8036
 * i40e_get_cur_guaranteed_fd_count - Get the consumed guaranteed FD filters
 * @pf: board private structure
 **/
8037
u32 i40e_get_cur_guaranteed_fd_count(struct i40e_pf *pf)
8038
{
8039
	u32 val, fcnt_prog;
8040 8041 8042 8043 8044 8045 8046

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

/**
8047
 * i40e_get_current_fd_count - Get total FD filters programmed for this PF
8048 8049
 * @pf: board private structure
 **/
8050
u32 i40e_get_current_fd_count(struct i40e_pf *pf)
8051
{
8052 8053
	u32 val, fcnt_prog;

8054 8055 8056 8057 8058 8059
	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;
}
8060

8061 8062 8063 8064 8065 8066 8067 8068 8069 8070 8071 8072 8073 8074 8075
/**
 * 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;
}

8076 8077 8078 8079 8080 8081
/**
 * 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)
{
8082
	struct i40e_fdir_filter *filter;
8083
	u32 fcnt_prog, fcnt_avail;
8084
	struct hlist_node *node;
8085

8086
	if (test_bit(__I40E_FD_FLUSH_REQUESTED, pf->state))
8087 8088
		return;

8089
	/* Check if we have enough room to re-enable FDir SB capability. */
8090
	fcnt_prog = i40e_get_global_fd_count(pf);
8091
	fcnt_avail = pf->fdir_pf_filter_count;
8092 8093 8094
	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)) {
8095 8096 8097 8098
		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))
8099
				dev_info(&pf->pdev->dev, "FD Sideband/ntuple is being enabled since we have space in the table now\n");
8100 8101
		}
	}
8102

8103 8104 8105
	/* 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.
8106
	 */
8107 8108 8109 8110 8111 8112
	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))
8113
				dev_info(&pf->pdev->dev, "ATR is being enabled since we have space in the table and there are no conflicting ntuple rules\n");
8114 8115
		}
	}
8116 8117 8118 8119 8120 8121 8122 8123 8124

	/* 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--;
8125
				pf->fd_inv = 0;
8126 8127 8128
			}
		}
	}
8129 8130
}

8131
#define I40E_MIN_FD_FLUSH_INTERVAL 10
8132
#define I40E_MIN_FD_FLUSH_SB_ATR_UNSTABLE 30
8133 8134 8135 8136 8137 8138
/**
 * 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)
{
8139
	unsigned long min_flush_time;
8140
	int flush_wait_retry = 50;
8141 8142
	bool disable_atr = false;
	int fd_room;
8143 8144
	int reg;

8145 8146 8147
	if (!time_after(jiffies, pf->fd_flush_timestamp +
				 (I40E_MIN_FD_FLUSH_INTERVAL * HZ)))
		return;
8148

8149 8150 8151 8152 8153 8154 8155 8156 8157 8158 8159 8160 8161 8162 8163
	/* 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;
8164
	pf->flags |= I40E_FLAG_FD_ATR_AUTO_DISABLED;
8165 8166 8167 8168 8169 8170 8171 8172 8173 8174 8175 8176 8177 8178 8179 8180 8181 8182
	/* 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]);
8183
		if (!disable_atr && !pf->fd_tcp4_filter_cnt)
8184
			pf->flags &= ~I40E_FLAG_FD_ATR_AUTO_DISABLED;
8185
		clear_bit(__I40E_FD_FLUSH_REQUESTED, pf->state);
8186 8187
		if (I40E_DEBUG_FD & pf->hw.debug_mask)
			dev_info(&pf->pdev->dev, "FD Filter table flushed and FD-SB replayed.\n");
8188 8189 8190 8191 8192 8193 8194
	}
}

/**
 * i40e_get_current_atr_count - Get the count of total FD ATR filters programmed
 * @pf: board private structure
 **/
8195
u32 i40e_get_current_atr_cnt(struct i40e_pf *pf)
8196 8197 8198 8199 8200 8201 8202 8203 8204 8205 8206
{
	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
8207 8208 8209 8210 8211 8212 8213 8214
/**
 * 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 */
8215
	if (test_bit(__I40E_DOWN, pf->state))
J
Jesse Brandeburg 已提交
8216
		return;
8217

8218
	if (test_bit(__I40E_FD_FLUSH_REQUESTED, pf->state))
8219 8220
		i40e_fdir_flush_and_replay(pf);

8221 8222
	i40e_fdir_check_and_reenable(pf);

J
Jesse Brandeburg 已提交
8223 8224 8225 8226 8227 8228 8229 8230 8231
}

/**
 * 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)
{
8232
	if (!vsi || test_bit(__I40E_VSI_DOWN, vsi->state))
J
Jesse Brandeburg 已提交
8233 8234 8235 8236 8237 8238 8239 8240 8241 8242 8243 8244 8245 8246 8247 8248 8249 8250 8251
		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:
8252
	case I40E_VSI_IWARP:
J
Jesse Brandeburg 已提交
8253 8254 8255 8256 8257 8258 8259 8260 8261 8262 8263 8264 8265 8266 8267 8268 8269 8270 8271 8272 8273 8274 8275 8276 8277 8278 8279
	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
8280
	for (i = 0; i < pf->num_alloc_vsi; i++)
J
Jesse Brandeburg 已提交
8281 8282 8283 8284 8285 8286 8287 8288 8289 8290
		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
8291
	struct i40e_vsi *vsi = pf->vsi[pf->lan_vsi];
8292
	u8 new_link_speed, old_link_speed;
8293 8294
	i40e_status status;
	bool new_link, old_link;
J
Jesse Brandeburg 已提交
8295

8296 8297 8298
	/* save off old link status information */
	pf->hw.phy.link_info_old = pf->hw.phy.link_info;

8299 8300 8301
	/* 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
8302
	old_link = (pf->hw.phy.link_info_old.link_info & I40E_AQ_LINK_UP);
8303 8304

	status = i40e_get_link_status(&pf->hw, &new_link);
8305 8306 8307 8308 8309 8310 8311 8312 8313 8314

	/* 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;
8315 8316 8317 8318 8319
		dev_dbg(&pf->pdev->dev, "couldn't get link state, status: %d\n",
			status);
		return;
	}

8320 8321
	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
8322

8323
	if (new_link == old_link &&
8324
	    new_link_speed == old_link_speed &&
8325
	    (test_bit(__I40E_VSI_DOWN, vsi->state) ||
M
Mitch Williams 已提交
8326
	     new_link == netif_carrier_ok(vsi->netdev)))
J
Jesse Brandeburg 已提交
8327
		return;
M
Mitch Williams 已提交
8328

8329
	i40e_print_link_message(vsi, new_link);
J
Jesse Brandeburg 已提交
8330 8331 8332 8333 8334 8335 8336

	/* 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
8337
		i40e_vsi_link_event(vsi, new_link);
J
Jesse Brandeburg 已提交
8338 8339 8340

	if (pf->vf)
		i40e_vc_notify_link_state(pf);
J
Jacob Keller 已提交
8341 8342 8343

	if (pf->flags & I40E_FLAG_PTP)
		i40e_ptp_set_increment(pf);
J
Jesse Brandeburg 已提交
8344 8345 8346
}

/**
8347
 * i40e_watchdog_subtask - periodic checks not using event driven response
J
Jesse Brandeburg 已提交
8348 8349 8350 8351 8352 8353 8354
 * @pf: board private structure
 **/
static void i40e_watchdog_subtask(struct i40e_pf *pf)
{
	int i;

	/* if interface is down do nothing */
8355
	if (test_bit(__I40E_DOWN, pf->state) ||
8356
	    test_bit(__I40E_CONFIG_BUSY, pf->state))
J
Jesse Brandeburg 已提交
8357 8358
		return;

8359 8360 8361 8362 8363 8364
	/* 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;

8365 8366
	if ((pf->flags & I40E_FLAG_LINK_POLLING_ENABLED) ||
	    (pf->flags & I40E_FLAG_TEMP_LINK_POLLING))
S
Shannon Nelson 已提交
8367
		i40e_link_event(pf);
8368

J
Jesse Brandeburg 已提交
8369 8370 8371
	/* 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
8372
	for (i = 0; i < pf->num_alloc_vsi; i++)
J
Jesse Brandeburg 已提交
8373 8374 8375
		if (pf->vsi[i] && pf->vsi[i]->netdev)
			i40e_update_stats(pf->vsi[i]);

8376 8377 8378 8379 8380 8381
	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
8382

8383
	i40e_ptp_rx_hang(pf);
8384
	i40e_ptp_tx_hang(pf);
J
Jesse Brandeburg 已提交
8385 8386 8387 8388 8389 8390 8391 8392 8393 8394
}

/**
 * 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;

8395
	if (test_bit(__I40E_REINIT_REQUESTED, pf->state)) {
J
Jesse Brandeburg 已提交
8396
		reset_flags |= BIT(__I40E_REINIT_REQUESTED);
8397
		clear_bit(__I40E_REINIT_REQUESTED, pf->state);
J
Jesse Brandeburg 已提交
8398
	}
8399
	if (test_bit(__I40E_PF_RESET_REQUESTED, pf->state)) {
J
Jesse Brandeburg 已提交
8400
		reset_flags |= BIT(__I40E_PF_RESET_REQUESTED);
8401
		clear_bit(__I40E_PF_RESET_REQUESTED, pf->state);
J
Jesse Brandeburg 已提交
8402
	}
8403
	if (test_bit(__I40E_CORE_RESET_REQUESTED, pf->state)) {
J
Jesse Brandeburg 已提交
8404
		reset_flags |= BIT(__I40E_CORE_RESET_REQUESTED);
8405
		clear_bit(__I40E_CORE_RESET_REQUESTED, pf->state);
J
Jesse Brandeburg 已提交
8406
	}
8407
	if (test_bit(__I40E_GLOBAL_RESET_REQUESTED, pf->state)) {
J
Jesse Brandeburg 已提交
8408
		reset_flags |= BIT(__I40E_GLOBAL_RESET_REQUESTED);
8409
		clear_bit(__I40E_GLOBAL_RESET_REQUESTED, pf->state);
J
Jesse Brandeburg 已提交
8410
	}
8411 8412 8413
	if (test_bit(__I40E_DOWN_REQUESTED, pf->state)) {
		reset_flags |= BIT(__I40E_DOWN_REQUESTED);
		clear_bit(__I40E_DOWN_REQUESTED, pf->state);
8414
	}
J
Jesse Brandeburg 已提交
8415 8416 8417 8418

	/* If there's a recovery already waiting, it takes
	 * precedence before starting a new reset sequence.
	 */
8419
	if (test_bit(__I40E_RESET_INTR_RECEIVED, pf->state)) {
8420 8421 8422
		i40e_prep_for_reset(pf, false);
		i40e_reset(pf);
		i40e_rebuild(pf, false, false);
J
Jesse Brandeburg 已提交
8423 8424 8425 8426
	}

	/* If we're already down or resetting, just bail */
	if (reset_flags &&
8427
	    !test_bit(__I40E_DOWN, pf->state) &&
8428
	    !test_bit(__I40E_CONFIG_BUSY, pf->state)) {
8429
		i40e_do_reset(pf, reset_flags, false);
8430
	}
J
Jesse Brandeburg 已提交
8431 8432 8433 8434 8435 8436 8437 8438 8439 8440 8441 8442 8443
}

/**
 * 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;

8444 8445 8446 8447 8448 8449 8450 8451
	/* 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);

8452 8453
	/* Check if module meets thermal requirements */
	if (status->phy_type == I40E_PHY_TYPE_NOT_SUPPORTED_HIGH_TEMP) {
8454
		dev_err(&pf->pdev->dev,
8455 8456 8457 8458 8459 8460 8461 8462 8463 8464 8465 8466 8467 8468 8469 8470 8471
			"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
8472 8473 8474 8475 8476 8477 8478 8479 8480 8481 8482 8483 8484
}

/**
 * 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;
8485
	u32 oldval;
J
Jesse Brandeburg 已提交
8486 8487
	u32 val;

8488
	/* Do not run clean AQ when PF reset fails */
8489
	if (test_bit(__I40E_RESET_FAILED, pf->state))
8490 8491
		return;

8492 8493 8494 8495
	/* 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
8496 8497
		if (hw->debug_mask & I40E_DEBUG_AQ)
			dev_info(&pf->pdev->dev, "ARQ VF Error detected\n");
8498 8499 8500
		val &= ~I40E_PF_ARQLEN_ARQVFE_MASK;
	}
	if (val & I40E_PF_ARQLEN_ARQOVFL_MASK) {
M
Mitch Williams 已提交
8501 8502
		if (hw->debug_mask & I40E_DEBUG_AQ)
			dev_info(&pf->pdev->dev, "ARQ Overflow Error detected\n");
8503
		val &= ~I40E_PF_ARQLEN_ARQOVFL_MASK;
8504
		pf->arq_overflows++;
8505 8506
	}
	if (val & I40E_PF_ARQLEN_ARQCRIT_MASK) {
M
Mitch Williams 已提交
8507 8508
		if (hw->debug_mask & I40E_DEBUG_AQ)
			dev_info(&pf->pdev->dev, "ARQ Critical Error detected\n");
8509 8510 8511 8512 8513 8514 8515 8516
		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
8517 8518
		if (pf->hw.debug_mask & I40E_DEBUG_AQ)
			dev_info(&pf->pdev->dev, "ASQ VF Error detected\n");
8519 8520 8521
		val &= ~I40E_PF_ATQLEN_ATQVFE_MASK;
	}
	if (val & I40E_PF_ATQLEN_ATQOVFL_MASK) {
M
Mitch Williams 已提交
8522 8523
		if (pf->hw.debug_mask & I40E_DEBUG_AQ)
			dev_info(&pf->pdev->dev, "ASQ Overflow Error detected\n");
8524 8525 8526
		val &= ~I40E_PF_ATQLEN_ATQOVFL_MASK;
	}
	if (val & I40E_PF_ATQLEN_ATQCRIT_MASK) {
M
Mitch Williams 已提交
8527 8528
		if (pf->hw.debug_mask & I40E_DEBUG_AQ)
			dev_info(&pf->pdev->dev, "ASQ Critical Error detected\n");
8529 8530 8531 8532 8533
		val &= ~I40E_PF_ATQLEN_ATQCRIT_MASK;
	}
	if (oldval != val)
		wr32(&pf->hw, pf->hw.aq.asq.len, val);

M
Mitch Williams 已提交
8534 8535
	event.buf_len = I40E_MAX_AQ_BUF_SIZE;
	event.msg_buf = kzalloc(event.buf_len, GFP_KERNEL);
J
Jesse Brandeburg 已提交
8536 8537 8538 8539 8540
	if (!event.msg_buf)
		return;

	do {
		ret = i40e_clean_arq_element(hw, &event, &pending);
M
Mitch Williams 已提交
8541
		if (ret == I40E_ERR_ADMIN_QUEUE_NO_WORK)
J
Jesse Brandeburg 已提交
8542
			break;
M
Mitch Williams 已提交
8543
		else if (ret) {
J
Jesse Brandeburg 已提交
8544 8545 8546 8547 8548 8549 8550 8551 8552 8553 8554 8555 8556 8557 8558 8559
			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
8560
					event.msg_len);
J
Jesse Brandeburg 已提交
8561 8562
			break;
		case i40e_aqc_opc_lldp_update_mib:
J
Jesse Brandeburg 已提交
8563
			dev_dbg(&pf->pdev->dev, "ARQ: Update LLDP MIB event received\n");
N
Neerav Parikh 已提交
8564 8565 8566 8567 8568
#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
8569 8570
			break;
		case i40e_aqc_opc_event_lan_overflow:
J
Jesse Brandeburg 已提交
8571
			dev_dbg(&pf->pdev->dev, "ARQ LAN queue overflow event received\n");
J
Jesse Brandeburg 已提交
8572 8573
			i40e_handle_lan_overflow_event(pf, &event);
			break;
8574 8575 8576
		case i40e_aqc_opc_send_msg_to_peer:
			dev_info(&pf->pdev->dev, "ARQ: Msg from other pf\n");
			break;
8577 8578
		case i40e_aqc_opc_nvm_erase:
		case i40e_aqc_opc_nvm_update:
8579
		case i40e_aqc_opc_oem_post_update:
8580 8581 8582
			i40e_debug(&pf->hw, I40E_DEBUG_NVM,
				   "ARQ NVM operation 0x%04x completed\n",
				   opcode);
8583
			break;
J
Jesse Brandeburg 已提交
8584 8585
		default:
			dev_info(&pf->pdev->dev,
8586
				 "ARQ: Unknown event 0x%04x ignored\n",
8587
				 opcode);
J
Jesse Brandeburg 已提交
8588 8589
			break;
		}
8590 8591 8592
	} while (i++ < pf->adminq_work_limit);

	if (i < pf->adminq_work_limit)
8593
		clear_bit(__I40E_ADMINQ_EVENT_PENDING, pf->state);
J
Jesse Brandeburg 已提交
8594 8595 8596 8597 8598 8599 8600 8601 8602 8603

	/* 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);
}

8604 8605 8606 8607 8608 8609 8610 8611 8612 8613 8614 8615 8616 8617 8618
/**
 * 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);
8619
			set_bit(__I40E_BAD_EEPROM, pf->state);
8620 8621 8622
		}
	}

8623
	if (!err && test_bit(__I40E_BAD_EEPROM, pf->state)) {
8624
		dev_info(&pf->pdev->dev, "eeprom check passed, Tx/Rx traffic enabled\n");
8625
		clear_bit(__I40E_BAD_EEPROM, pf->state);
8626 8627 8628
	}
}

8629 8630
/**
 * i40e_enable_pf_switch_lb
8631
 * @pf: pointer to the PF structure
8632 8633 8634 8635 8636 8637 8638
 *
 * 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;
8639
	int ret;
8640 8641 8642 8643

	ctxt.seid = pf->main_vsi_seid;
	ctxt.pf_num = pf->hw.pf_id;
	ctxt.vf_num = 0;
8644 8645
	ret = i40e_aq_get_vsi_params(&pf->hw, &ctxt, NULL);
	if (ret) {
8646
		dev_info(&pf->pdev->dev,
8647 8648 8649
			 "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));
8650 8651 8652 8653 8654 8655
		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);

8656 8657
	ret = i40e_aq_update_vsi_params(&vsi->back->hw, &ctxt, NULL);
	if (ret) {
8658
		dev_info(&pf->pdev->dev,
8659 8660 8661
			 "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));
8662 8663 8664 8665 8666
	}
}

/**
 * i40e_disable_pf_switch_lb
8667
 * @pf: pointer to the PF structure
8668 8669 8670 8671 8672 8673 8674
 *
 * 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;
8675
	int ret;
8676 8677 8678 8679

	ctxt.seid = pf->main_vsi_seid;
	ctxt.pf_num = pf->hw.pf_id;
	ctxt.vf_num = 0;
8680 8681
	ret = i40e_aq_get_vsi_params(&pf->hw, &ctxt, NULL);
	if (ret) {
8682
		dev_info(&pf->pdev->dev,
8683 8684 8685
			 "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));
8686 8687 8688 8689 8690 8691
		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);

8692 8693
	ret = i40e_aq_update_vsi_params(&vsi->back->hw, &ctxt, NULL);
	if (ret) {
8694
		dev_info(&pf->pdev->dev,
8695 8696 8697
			 "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));
8698 8699 8700
	}
}

8701 8702 8703 8704 8705 8706 8707 8708 8709 8710 8711 8712
/**
 * 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;

8713 8714 8715
	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");
8716 8717 8718 8719 8720 8721
	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
8722 8723 8724 8725 8726 8727 8728 8729 8730 8731 8732 8733 8734 8735 8736 8737 8738
/**
 * 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
8739
	for (v = 0; v < pf->num_alloc_vsi && !ctl_vsi; v++) {
J
Jesse Brandeburg 已提交
8740 8741 8742 8743 8744 8745 8746 8747 8748 8749 8750 8751 8752 8753 8754 8755 8756 8757
		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,
8758 8759
			 "rebuild of veb_idx %d owner VSI failed: %d\n",
			 veb->idx, ret);
J
Jesse Brandeburg 已提交
8760 8761 8762 8763 8764 8765 8766 8767 8768
		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;

8769 8770 8771 8772
	if (pf->flags & I40E_FLAG_VEB_MODE_ENABLED)
		veb->bridge_mode = BRIDGE_MODE_VEB;
	else
		veb->bridge_mode = BRIDGE_MODE_VEPA;
8773
	i40e_config_bridge_mode(veb);
8774

J
Jesse Brandeburg 已提交
8775
	/* create the remaining VSIs attached to this VEB */
M
Mitch Williams 已提交
8776
	for (v = 0; v < pf->num_alloc_vsi; v++) {
J
Jesse Brandeburg 已提交
8777 8778 8779 8780 8781
		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
8782

J
Jesse Brandeburg 已提交
8783 8784 8785 8786 8787 8788 8789 8790 8791 8792 8793 8794 8795 8796 8797 8798 8799 8800 8801 8802 8803 8804 8805 8806 8807 8808 8809 8810 8811 8812
			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
 **/
8813 8814
static int i40e_get_capabilities(struct i40e_pf *pf,
				 enum i40e_admin_queue_opc list_type)
J
Jesse Brandeburg 已提交
8815 8816 8817 8818 8819 8820 8821 8822 8823 8824 8825 8826 8827 8828
{
	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,
8829 8830
						    &data_size, list_type,
						    NULL);
J
Jesse Brandeburg 已提交
8831 8832 8833 8834 8835 8836 8837 8838
		/* 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,
8839 8840 8841 8842
				 "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
8843 8844 8845 8846
			return -ENODEV;
		}
	} while (err);

8847 8848 8849 8850 8851 8852 8853 8854 8855 8856 8857 8858 8859 8860 8861 8862 8863 8864 8865 8866 8867 8868 8869 8870 8871 8872 8873 8874
	if (pf->hw.debug_mask & I40E_DEBUG_USER) {
		if (list_type == i40e_aqc_opc_list_func_capabilities) {
			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);
		} else if (list_type == i40e_aqc_opc_list_dev_capabilities) {
			dev_info(&pf->pdev->dev,
				 "switch_mode=0x%04x, function_valid=0x%08x\n",
				 pf->hw.dev_caps.switch_mode,
				 pf->hw.dev_caps.valid_functions);
			dev_info(&pf->pdev->dev,
				 "SR-IOV=%d, num_vfs for all function=%u\n",
				 pf->hw.dev_caps.sr_iov_1_1,
				 pf->hw.dev_caps.num_vfs);
			dev_info(&pf->pdev->dev,
				 "num_vsis=%u, num_rx:%u, num_tx=%u\n",
				 pf->hw.dev_caps.num_vsis,
				 pf->hw.dev_caps.num_rx_qp,
				 pf->hw.dev_caps.num_tx_qp);
		}
	}
	if (list_type == i40e_aqc_opc_list_func_capabilities) {
8875 8876
#define DEF_NUM_VSI (1 + (pf->hw.func_caps.fcoe ? 1 : 0) \
		       + pf->hw.func_caps.num_vfs)
8877 8878 8879 8880 8881 8882 8883
		if (pf->hw.revision_id == 0 &&
		    pf->hw.func_caps.num_vsis < DEF_NUM_VSI) {
			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;
		}
8884
	}
J
Jesse Brandeburg 已提交
8885 8886 8887
	return 0;
}

8888 8889
static int i40e_vsi_clear(struct i40e_vsi *vsi);

J
Jesse Brandeburg 已提交
8890
/**
8891
 * i40e_fdir_sb_setup - initialize the Flow Director resources for Sideband
J
Jesse Brandeburg 已提交
8892 8893
 * @pf: board private structure
 **/
8894
static void i40e_fdir_sb_setup(struct i40e_pf *pf)
J
Jesse Brandeburg 已提交
8895 8896 8897
{
	struct i40e_vsi *vsi;

8898 8899 8900 8901 8902 8903 8904 8905 8906
	/* 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};
8907
		int i;
8908 8909 8910 8911 8912

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

8913
	if (!(pf->flags & I40E_FLAG_FD_SB_ENABLED))
J
Jesse Brandeburg 已提交
8914 8915
		return;

8916
	/* find existing VSI and see if it needs configuring */
8917
	vsi = i40e_find_vsi_by_type(pf, I40E_VSI_FDIR);
8918 8919

	/* create a new VSI if none exists */
J
Jesse Brandeburg 已提交
8920
	if (!vsi) {
8921 8922
		vsi = i40e_vsi_setup(pf, I40E_VSI_FDIR,
				     pf->vsi[pf->lan_vsi]->seid, 0);
J
Jesse Brandeburg 已提交
8923 8924
		if (!vsi) {
			dev_info(&pf->pdev->dev, "Couldn't create FDir VSI\n");
8925
			pf->flags &= ~I40E_FLAG_FD_SB_ENABLED;
8926
			pf->flags |= I40E_FLAG_FD_SB_INACTIVE;
8927
			return;
J
Jesse Brandeburg 已提交
8928
		}
8929
	}
J
Jesse Brandeburg 已提交
8930

8931
	i40e_vsi_setup_irqhandler(vsi, i40e_fdir_clean_ring);
J
Jesse Brandeburg 已提交
8932 8933 8934 8935 8936 8937 8938 8939
}

/**
 * i40e_fdir_teardown - release the Flow Director resources
 * @pf: board private structure
 **/
static void i40e_fdir_teardown(struct i40e_pf *pf)
{
8940
	struct i40e_vsi *vsi;
J
Jesse Brandeburg 已提交
8941

8942
	i40e_fdir_filter_exit(pf);
8943 8944 8945
	vsi = i40e_find_vsi_by_type(pf, I40E_VSI_FDIR);
	if (vsi)
		i40e_vsi_release(vsi);
J
Jesse Brandeburg 已提交
8946 8947
}

8948 8949 8950 8951 8952 8953 8954 8955 8956 8957 8958 8959 8960 8961 8962 8963 8964 8965 8966 8967 8968 8969 8970 8971 8972 8973
/**
 * i40e_rebuild_cloud_filters - Rebuilds cloud filters for VSIs
 * @vsi: PF main vsi
 * @seid: seid of main or channel VSIs
 *
 * Rebuilds cloud filters associated with main VSI and channel VSIs if they
 * existed before reset
 **/
static int i40e_rebuild_cloud_filters(struct i40e_vsi *vsi, u16 seid)
{
	struct i40e_cloud_filter *cfilter;
	struct i40e_pf *pf = vsi->back;
	struct hlist_node *node;
	i40e_status ret;

	/* Add cloud filters back if they exist */
	hlist_for_each_entry_safe(cfilter, node, &pf->cloud_filter_list,
				  cloud_node) {
		if (cfilter->seid != seid)
			continue;

		if (cfilter->dst_port)
			ret = i40e_add_del_cloud_filter_big_buf(vsi, cfilter,
								true);
		else
			ret = i40e_add_del_cloud_filter(vsi, cfilter, true);
8974

8975 8976 8977 8978 8979 8980 8981
		if (ret) {
			dev_dbg(&pf->pdev->dev,
				"Failed to rebuild cloud filter, 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;
J
Jesse Brandeburg 已提交
8982
		}
8983
	}
8984
	return 0;
J
Jesse Brandeburg 已提交
8985 8986 8987
}

/**
8988 8989 8990 8991
 * i40e_rebuild_channels - Rebuilds channel VSIs if they existed before reset
 * @vsi: PF main vsi
 *
 * Rebuilds channel VSIs if they existed before reset
J
Jesse Brandeburg 已提交
8992
 **/
8993
static int i40e_rebuild_channels(struct i40e_vsi *vsi)
J
Jesse Brandeburg 已提交
8994
{
8995 8996
	struct i40e_channel *ch, *ch_tmp;
	i40e_status ret;
J
Jesse Brandeburg 已提交
8997

8998 8999 9000 9001 9002 9003 9004 9005 9006 9007 9008 9009 9010 9011
	if (list_empty(&vsi->ch_list))
		return 0;

	list_for_each_entry_safe(ch, ch_tmp, &vsi->ch_list, list) {
		if (!ch->initialized)
			break;
		/* Proceed with creation of channel (VMDq2) VSI */
		ret = i40e_add_channel(vsi->back, vsi->uplink_seid, ch);
		if (ret) {
			dev_info(&vsi->back->pdev->dev,
				 "failed to rebuild channels using uplink_seid %u\n",
				 vsi->uplink_seid);
			return ret;
		}
9012 9013 9014 9015 9016 9017 9018 9019 9020 9021 9022
		/* Reconfigure TX queues using QTX_CTL register */
		ret = i40e_channel_config_tx_ring(vsi->back, vsi, ch);
		if (ret) {
			dev_info(&vsi->back->pdev->dev,
				 "failed to configure TX rings for channel %u\n",
				 ch->seid);
			return ret;
		}
		/* update 'next_base_queue' */
		vsi->next_base_queue = vsi->next_base_queue +
							ch->num_queue_pairs;
9023
		if (ch->max_tx_rate) {
A
Alan Brady 已提交
9024 9025
			u64 credits = ch->max_tx_rate;

9026 9027 9028 9029
			if (i40e_set_bw_limit(vsi, ch->seid,
					      ch->max_tx_rate))
				return -EINVAL;

A
Alan Brady 已提交
9030
			do_div(credits, I40E_BW_CREDIT_DIVISOR);
9031 9032 9033
			dev_dbg(&vsi->back->pdev->dev,
				"Set tx rate of %llu Mbps (count of 50Mbps %llu) for vsi->seid %u\n",
				ch->max_tx_rate,
A
Alan Brady 已提交
9034
				credits,
9035 9036
				ch->seid);
		}
9037 9038 9039 9040 9041 9042 9043
		ret = i40e_rebuild_cloud_filters(vsi, ch->seid);
		if (ret) {
			dev_dbg(&vsi->back->pdev->dev,
				"Failed to rebuild cloud filters for channel VSI %u\n",
				ch->seid);
			return ret;
		}
9044 9045
	}
	return 0;
J
Jesse Brandeburg 已提交
9046 9047 9048
}

/**
A
Anjali Singhai Jain 已提交
9049
 * i40e_prep_for_reset - prep for the core to reset
J
Jesse Brandeburg 已提交
9050
 * @pf: board private structure
9051 9052
 * @lock_acquired: indicates whether or not the lock has been acquired
 * before this function was called.
J
Jesse Brandeburg 已提交
9053
 *
9054
 * Close up the VFs and other things in prep for PF Reset.
A
Anjali Singhai Jain 已提交
9055
  **/
9056
static void i40e_prep_for_reset(struct i40e_pf *pf, bool lock_acquired)
J
Jesse Brandeburg 已提交
9057 9058
{
	struct i40e_hw *hw = &pf->hw;
9059
	i40e_status ret = 0;
J
Jesse Brandeburg 已提交
9060 9061
	u32 v;

9062 9063
	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
9064
		return;
M
Mitch Williams 已提交
9065 9066
	if (i40e_check_asq_alive(&pf->hw))
		i40e_vc_notify_reset(pf);
J
Jesse Brandeburg 已提交
9067

J
Jesse Brandeburg 已提交
9068
	dev_dbg(&pf->pdev->dev, "Tearing down internal switch for reset\n");
J
Jesse Brandeburg 已提交
9069 9070

	/* quiesce the VSIs and their queues that are not already DOWN */
9071 9072 9073
	/* pf_quiesce_all_vsi modifies netdev structures -rtnl_lock needed */
	if (!lock_acquired)
		rtnl_lock();
J
Jesse Brandeburg 已提交
9074
	i40e_pf_quiesce_all_vsi(pf);
9075 9076
	if (!lock_acquired)
		rtnl_unlock();
J
Jesse Brandeburg 已提交
9077

M
Mitch Williams 已提交
9078
	for (v = 0; v < pf->num_alloc_vsi; v++) {
J
Jesse Brandeburg 已提交
9079 9080 9081 9082 9083 9084
		if (pf->vsi[v])
			pf->vsi[v]->seid = 0;
	}

	i40e_shutdown_adminq(&pf->hw);

A
Anjali Singhai Jain 已提交
9085
	/* call shutdown HMC */
9086 9087
	if (hw->hmc.hmc_obj) {
		ret = i40e_shutdown_lan_hmc(hw);
S
Shannon Nelson 已提交
9088
		if (ret)
9089 9090
			dev_warn(&pf->pdev->dev,
				 "shutdown_lan_hmc failed: %d\n", ret);
A
Anjali Singhai Jain 已提交
9091 9092 9093
	}
}

J
Jesse Brandeburg 已提交
9094 9095 9096 9097 9098 9099 9100 9101 9102 9103 9104 9105
/**
 * 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;
9106
	strlcpy(dv.driver_string, DRV_VERSION, sizeof(dv.driver_string));
J
Jesse Brandeburg 已提交
9107 9108 9109
	i40e_aq_send_driver_version(&pf->hw, &dv, NULL);
}

9110 9111 9112 9113 9114 9115 9116 9117 9118 9119 9120 9121 9122 9123 9124 9125 9126 9127 9128 9129 9130 9131 9132 9133 9134 9135 9136 9137 9138 9139 9140 9141 9142 9143 9144 9145 9146 9147 9148 9149 9150 9151 9152 9153 9154
/**
 * 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
9155
/**
9156
 * i40e_reset - wait for core reset to finish reset, reset pf if corer not seen
A
Anjali Singhai Jain 已提交
9157 9158
 * @pf: board private structure
 **/
9159
static int i40e_reset(struct i40e_pf *pf)
A
Anjali Singhai Jain 已提交
9160 9161 9162 9163
{
	struct i40e_hw *hw = &pf->hw;
	i40e_status ret;

J
Jesse Brandeburg 已提交
9164
	ret = i40e_pf_reset(hw);
9165
	if (ret) {
J
Jesse Brandeburg 已提交
9166
		dev_info(&pf->pdev->dev, "PF reset failed, %d\n", ret);
9167 9168
		set_bit(__I40E_RESET_FAILED, pf->state);
		clear_bit(__I40E_RESET_RECOVERY_PENDING, pf->state);
9169 9170
	} else {
		pf->pfr_count++;
9171
	}
9172 9173 9174 9175 9176 9177 9178 9179 9180 9181 9182 9183
	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)
{
9184
	struct i40e_vsi *vsi = pf->vsi[pf->lan_vsi];
9185 9186 9187 9188 9189
	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
9190

9191
	if (test_bit(__I40E_DOWN, pf->state))
9192
		goto clear_recovery;
J
Jesse Brandeburg 已提交
9193
	dev_dbg(&pf->pdev->dev, "Rebuilding internal switch\n");
J
Jesse Brandeburg 已提交
9194 9195 9196 9197

	/* rebuild the basics for the AdminQ, HMC, and initial HW switch */
	ret = i40e_init_adminq(&pf->hw);
	if (ret) {
9198 9199 9200
		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));
9201
		goto clear_recovery;
J
Jesse Brandeburg 已提交
9202
	}
9203
	i40e_get_oem_version(&pf->hw);
J
Jesse Brandeburg 已提交
9204

9205 9206 9207 9208 9209 9210 9211 9212 9213 9214 9215
	if (test_bit(__I40E_EMP_RESET_INTR_RECEIVED, pf->state) &&
	    ((hw->aq.fw_maj_ver == 4 && hw->aq.fw_min_ver <= 33) ||
	     hw->aq.fw_maj_ver < 4) && hw->mac.type == I40E_MAC_XL710) {
		/* The following delay is necessary for 4.33 firmware and older
		 * to recover after EMP reset. 200 ms should suffice but we
		 * put here 300 ms to be sure that FW is ready to operate
		 * after reset.
		 */
		mdelay(300);
	}

9216
	/* re-verify the eeprom if we just had an EMP reset */
9217
	if (test_and_clear_bit(__I40E_EMP_RESET_INTR_RECEIVED, pf->state))
9218 9219
		i40e_verify_eeprom(pf);

9220
	i40e_clear_pxe_mode(hw);
9221
	ret = i40e_get_capabilities(pf, i40e_aqc_opc_list_func_capabilities);
9222
	if (ret)
J
Jesse Brandeburg 已提交
9223 9224 9225
		goto end_core_reset;

	ret = i40e_init_lan_hmc(hw, hw->func_caps.num_tx_qp,
9226
				hw->func_caps.num_rx_qp, 0, 0);
J
Jesse Brandeburg 已提交
9227 9228 9229 9230 9231 9232 9233 9234 9235 9236
	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;
	}

9237 9238 9239
	/* Enable FW to write a default DCB config on link-up */
	i40e_aq_set_dcb_parameters(hw, true, NULL);

N
Neerav Parikh 已提交
9240 9241 9242
#ifdef CONFIG_I40E_DCB
	ret = i40e_init_pf_dcb(pf);
	if (ret) {
9243 9244 9245
		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
9246 9247
	}
#endif /* CONFIG_I40E_DCB */
J
Jesse Brandeburg 已提交
9248
	/* do basic switch setup */
9249 9250
	if (!lock_acquired)
		rtnl_lock();
9251
	ret = i40e_setup_pf_switch(pf, reinit);
J
Jesse Brandeburg 已提交
9252
	if (ret)
9253
		goto end_unlock;
J
Jesse Brandeburg 已提交
9254

S
Shannon Nelson 已提交
9255 9256
	/* The driver only wants link up/down and module qualification
	 * reports from firmware.  Note the negative logic.
J
Jesse Brandeburg 已提交
9257 9258
	 */
	ret = i40e_aq_set_phy_int_mask(&pf->hw,
S
Shannon Nelson 已提交
9259
				       ~(I40E_AQ_EVENT_LINK_UPDOWN |
9260
					 I40E_AQ_EVENT_MEDIA_NA |
S
Shannon Nelson 已提交
9261
					 I40E_AQ_EVENT_MODULE_QUAL_FAIL), NULL);
J
Jesse Brandeburg 已提交
9262
	if (ret)
9263 9264 9265
		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
9266

9267 9268 9269
	/* make sure our flow control settings are restored */
	ret = i40e_set_fc(&pf->hw, &set_fc_aq_fail, true);
	if (ret)
9270 9271 9272
		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));
9273

J
Jesse Brandeburg 已提交
9274 9275 9276 9277 9278 9279 9280
	/* 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.
	 */
9281
	if (vsi->uplink_seid != pf->mac_seid) {
J
Jesse Brandeburg 已提交
9282
		dev_dbg(&pf->pdev->dev, "attempting to rebuild switch\n");
J
Jesse Brandeburg 已提交
9283 9284 9285 9286 9287 9288 9289 9290 9291 9292 9293 9294 9295 9296 9297 9298 9299 9300 9301 9302 9303 9304
		/* 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);
9305
					vsi->uplink_seid = pf->mac_seid;
J
Jesse Brandeburg 已提交
9306 9307 9308 9309 9310 9311 9312 9313 9314 9315
					break;
				} else if (pf->veb[v]->uplink_seid == 0) {
					dev_info(&pf->pdev->dev,
						 "rebuild of orphan VEB failed: %d\n",
						 ret);
				}
			}
		}
	}

9316
	if (vsi->uplink_seid == pf->mac_seid) {
S
Shannon Nelson 已提交
9317
		dev_dbg(&pf->pdev->dev, "attempting to rebuild PF VSI\n");
J
Jesse Brandeburg 已提交
9318
		/* no VEB, so rebuild only the Main VSI */
9319
		ret = i40e_add_vsi(vsi);
J
Jesse Brandeburg 已提交
9320 9321 9322
		if (ret) {
			dev_info(&pf->pdev->dev,
				 "rebuild of Main VSI failed: %d\n", ret);
9323
			goto end_unlock;
J
Jesse Brandeburg 已提交
9324 9325 9326
		}
	}

9327
	if (vsi->mqprio_qopt.max_rate[0]) {
A
Alan Brady 已提交
9328 9329
		u64 max_tx_rate = vsi->mqprio_qopt.max_rate[0];
		u64 credits = 0;
9330

A
Alan Brady 已提交
9331
		do_div(max_tx_rate, I40E_BW_MBPS_DIVISOR);
9332
		ret = i40e_set_bw_limit(vsi, vsi->seid, max_tx_rate);
A
Alan Brady 已提交
9333
		if (ret)
9334
			goto end_unlock;
A
Alan Brady 已提交
9335 9336 9337 9338 9339 9340 9341 9342

		credits = max_tx_rate;
		do_div(credits, I40E_BW_CREDIT_DIVISOR);
		dev_dbg(&vsi->back->pdev->dev,
			"Set tx rate of %llu Mbps (count of 50Mbps %llu) for vsi->seid %u\n",
			max_tx_rate,
			credits,
			vsi->seid);
9343 9344
	}

9345 9346 9347 9348
	ret = i40e_rebuild_cloud_filters(vsi, vsi->seid);
	if (ret)
		goto end_unlock;

9349 9350 9351
	/* PF Main VSI is rebuild by now, go ahead and rebuild channel VSIs
	 * for this main VSI if they exist
	 */
9352
	ret = i40e_rebuild_channels(vsi);
9353 9354 9355
	if (ret)
		goto end_unlock;

9356 9357 9358 9359 9360 9361 9362 9363 9364 9365 9366 9367 9368 9369
	/* 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);
	}

9370
	if (pf->hw_features & I40E_HW_RESTART_AUTONEG) {
9371 9372 9373
		msleep(75);
		ret = i40e_aq_set_link_restart_an(&pf->hw, true, NULL);
		if (ret)
9374 9375 9376 9377
			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));
9378
	}
J
Jesse Brandeburg 已提交
9379 9380 9381 9382
	/* reinit the misc interrupt */
	if (pf->flags & I40E_FLAG_MSIX_ENABLED)
		ret = i40e_setup_misc_vector(pf);

9383 9384 9385 9386 9387 9388 9389 9390 9391
	/* 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
9392 9393 9394
	/* restart the VSIs that were rebuilt and running before the reset */
	i40e_pf_unquiesce_all_vsi(pf);

9395 9396 9397 9398
	/* Release the RTNL lock before we start resetting VFs */
	if (!lock_acquired)
		rtnl_unlock();

A
Alan Brady 已提交
9399 9400 9401 9402 9403 9404 9405 9406 9407
	/* Restore promiscuous settings */
	ret = i40e_set_promiscuous(pf, pf->cur_promisc);
	if (ret)
		dev_warn(&pf->pdev->dev,
			 "Failed to restore promiscuous setting: %s, err %s aq_err %s\n",
			 pf->cur_promisc ? "on" : "off",
			 i40e_stat_str(&pf->hw, ret),
			 i40e_aq_str(&pf->hw, pf->hw.aq.asq_last_status));

9408
	i40e_reset_all_vfs(pf, true);
M
Mitch Williams 已提交
9409

J
Jesse Brandeburg 已提交
9410
	/* tell the firmware that we're starting */
J
Jesse Brandeburg 已提交
9411
	i40e_send_version(pf);
J
Jesse Brandeburg 已提交
9412

9413 9414 9415
	/* We've already released the lock, so don't do it again */
	goto end_core_reset;

9416
end_unlock:
9417 9418
	if (!lock_acquired)
		rtnl_unlock();
J
Jesse Brandeburg 已提交
9419
end_core_reset:
9420
	clear_bit(__I40E_RESET_FAILED, pf->state);
9421
clear_recovery:
9422
	clear_bit(__I40E_RESET_RECOVERY_PENDING, pf->state);
J
Jesse Brandeburg 已提交
9423 9424
}

9425 9426 9427 9428 9429 9430 9431 9432 9433 9434 9435 9436 9437 9438 9439 9440 9441 9442 9443 9444
/**
 * 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
9445
/**
9446
 * i40e_handle_reset_warning - prep for the PF to reset, reset and rebuild
A
Anjali Singhai Jain 已提交
9447 9448 9449 9450
 * @pf: board private structure
 *
 * Close up the VFs and other things in prep for a Core Reset,
 * then get ready to rebuild the world.
9451 9452
 * @lock_acquired: indicates whether or not the lock has been acquired
 * before this function was called.
A
Anjali Singhai Jain 已提交
9453
 **/
9454
static void i40e_handle_reset_warning(struct i40e_pf *pf, bool lock_acquired)
A
Anjali Singhai Jain 已提交
9455
{
9456 9457
	i40e_prep_for_reset(pf, lock_acquired);
	i40e_reset_and_rebuild(pf, false, lock_acquired);
A
Anjali Singhai Jain 已提交
9458 9459
}

J
Jesse Brandeburg 已提交
9460 9461
/**
 * i40e_handle_mdd_event
9462
 * @pf: pointer to the PF structure
J
Jesse Brandeburg 已提交
9463 9464 9465 9466 9467 9468 9469
 *
 * 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;
9470
	bool pf_mdd_detected = false;
J
Jesse Brandeburg 已提交
9471 9472 9473 9474
	struct i40e_vf *vf;
	u32 reg;
	int i;

9475
	if (!test_bit(__I40E_MDD_EVENT_PENDING, pf->state))
J
Jesse Brandeburg 已提交
9476 9477 9478 9479 9480
		return;

	/* find what triggered the MDD event */
	reg = rd32(hw, I40E_GL_MDET_TX);
	if (reg & I40E_GL_MDET_TX_VALID_MASK) {
9481 9482
		u8 pf_num = (reg & I40E_GL_MDET_TX_PF_NUM_MASK) >>
				I40E_GL_MDET_TX_PF_NUM_SHIFT;
9483
		u16 vf_num = (reg & I40E_GL_MDET_TX_VF_NUM_MASK) >>
9484
				I40E_GL_MDET_TX_VF_NUM_SHIFT;
9485
		u8 event = (reg & I40E_GL_MDET_TX_EVENT_MASK) >>
9486
				I40E_GL_MDET_TX_EVENT_SHIFT;
9487 9488 9489
		u16 queue = ((reg & I40E_GL_MDET_TX_QUEUE_MASK) >>
				I40E_GL_MDET_TX_QUEUE_SHIFT) -
				pf->hw.func_caps.base_queue;
9490
		if (netif_msg_tx_err(pf))
9491
			dev_info(&pf->pdev->dev, "Malicious Driver Detection event 0x%02x on TX queue %d PF number 0x%02x VF number 0x%02x\n",
9492
				 event, queue, pf_num, vf_num);
J
Jesse Brandeburg 已提交
9493 9494 9495 9496 9497
		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) {
9498 9499
		u8 func = (reg & I40E_GL_MDET_RX_FUNCTION_MASK) >>
				I40E_GL_MDET_RX_FUNCTION_SHIFT;
9500
		u8 event = (reg & I40E_GL_MDET_RX_EVENT_MASK) >>
9501
				I40E_GL_MDET_RX_EVENT_SHIFT;
9502 9503 9504
		u16 queue = ((reg & I40E_GL_MDET_RX_QUEUE_MASK) >>
				I40E_GL_MDET_RX_QUEUE_SHIFT) -
				pf->hw.func_caps.base_queue;
9505 9506 9507
		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
9508 9509 9510 9511
		wr32(hw, I40E_GL_MDET_RX, 0xffffffff);
		mdd_detected = true;
	}

9512 9513 9514 9515
	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);
9516
			dev_info(&pf->pdev->dev, "TX driver issue detected, PF reset issued\n");
9517 9518 9519 9520 9521
			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);
9522
			dev_info(&pf->pdev->dev, "RX driver issue detected, PF reset issued\n");
9523 9524 9525 9526
			pf_mdd_detected = true;
		}
		/* Queue belongs to the PF, initiate a reset */
		if (pf_mdd_detected) {
9527
			set_bit(__I40E_PF_RESET_REQUESTED, pf->state);
9528 9529 9530 9531
			i40e_service_event_schedule(pf);
		}
	}

J
Jesse Brandeburg 已提交
9532 9533 9534 9535 9536 9537 9538
	/* 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++;
9539 9540
			dev_info(&pf->pdev->dev, "TX driver issue detected on VF %d\n",
				 i);
J
Jesse Brandeburg 已提交
9541 9542 9543 9544 9545 9546
		}

		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++;
9547 9548
			dev_info(&pf->pdev->dev, "RX driver issue detected on VF %d\n",
				 i);
J
Jesse Brandeburg 已提交
9549 9550 9551 9552 9553 9554 9555
		}

		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");
9556
			set_bit(I40E_VF_STATE_DISABLED, &vf->vf_states);
J
Jesse Brandeburg 已提交
9557 9558 9559 9560
		}
	}

	/* re-enable mdd interrupt cause */
9561
	clear_bit(__I40E_MDD_EVENT_PENDING, pf->state);
J
Jesse Brandeburg 已提交
9562 9563 9564 9565 9566 9567
	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);
}

9568 9569 9570 9571 9572 9573 9574 9575 9576 9577 9578 9579
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";
	}
}

9580 9581 9582 9583 9584 9585 9586 9587 9588 9589 9590 9591 9592 9593 9594 9595 9596
/**
 * 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;
}

9597
/**
9598
 * i40e_sync_udp_filters_subtask - Sync the VSI filter list with HW
9599 9600
 * @pf: board private structure
 **/
9601
static void i40e_sync_udp_filters_subtask(struct i40e_pf *pf)
9602 9603 9604
{
	struct i40e_hw *hw = &pf->hw;
	i40e_status ret;
9605
	u16 port;
9606 9607
	int i;

9608
	if (!(pf->flags & I40E_FLAG_UDP_FILTER_SYNC))
9609 9610
		return;

9611
	pf->flags &= ~I40E_FLAG_UDP_FILTER_SYNC;
9612 9613

	for (i = 0; i < I40E_MAX_PF_UDP_OFFLOAD_PORTS; i++) {
9614 9615
		if (pf->pending_udp_bitmap & BIT_ULL(i)) {
			pf->pending_udp_bitmap &= ~BIT_ULL(i);
9616
			port = pf->udp_ports[i].port;
S
Shannon Nelson 已提交
9617
			if (port)
9618 9619 9620
				ret = i40e_aq_add_udp_tunnel(hw, port,
							pf->udp_ports[i].type,
							NULL, NULL);
S
Shannon Nelson 已提交
9621 9622
			else
				ret = i40e_aq_del_udp_tunnel(hw, i, NULL);
9623 9624

			if (ret) {
9625 9626 9627 9628 9629 9630 9631 9632
				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));
9633
				pf->udp_ports[i].port = 0;
9634 9635 9636 9637 9638
			}
		}
	}
}

J
Jesse Brandeburg 已提交
9639 9640 9641 9642 9643 9644 9645 9646 9647 9648 9649
/**
 * 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;

9650
	/* don't bother with service tasks if a reset is in progress */
9651
	if (test_bit(__I40E_RESET_RECOVERY_PENDING, pf->state))
9652 9653
		return;

9654
	if (test_and_set_bit(__I40E_SERVICE_SCHED, pf->state))
M
Mitch Williams 已提交
9655 9656
		return;

9657
	i40e_detect_recover_hung(pf->vsi[pf->lan_vsi]);
9658
	i40e_sync_filters_subtask(pf);
J
Jesse Brandeburg 已提交
9659 9660 9661 9662 9663
	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
9664 9665 9666 9667 9668 9669 9670 9671 9672 9673 9674 9675
	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
9676
	i40e_sync_filters_subtask(pf);
9677
	i40e_sync_udp_filters_subtask(pf);
J
Jesse Brandeburg 已提交
9678 9679
	i40e_clean_adminq_subtask(pf);

M
Mitch Williams 已提交
9680 9681
	/* flush memory to make sure state is correct before next watchdog */
	smp_mb__before_atomic();
9682
	clear_bit(__I40E_SERVICE_SCHED, pf->state);
J
Jesse Brandeburg 已提交
9683 9684 9685 9686 9687 9688

	/* 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)) ||
9689 9690 9691
	    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
9692 9693 9694 9695 9696 9697 9698
		i40e_service_event_schedule(pf);
}

/**
 * i40e_service_timer - timer callback
 * @data: pointer to PF struct
 **/
9699
static void i40e_service_timer(struct timer_list *t)
J
Jesse Brandeburg 已提交
9700
{
9701
	struct i40e_pf *pf = from_timer(pf, t, service_timer);
J
Jesse Brandeburg 已提交
9702 9703 9704 9705 9706 9707 9708 9709 9710 9711 9712 9713 9714 9715 9716 9717 9718 9719 9720 9721 9722 9723 9724 9725 9726 9727 9728 9729 9730 9731

	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);
9732
		vsi->num_q_vectors = pf->num_fdsb_msix;
J
Jesse Brandeburg 已提交
9733 9734 9735 9736 9737 9738 9739 9740 9741 9742 9743 9744 9745 9746 9747 9748 9749 9750 9751 9752 9753 9754 9755
		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
9756 9757
/**
 * i40e_vsi_alloc_arrays - Allocate queue and vector pointer arrays for the vsi
R
Rami Rosen 已提交
9758
 * @vsi: VSI pointer
9759
 * @alloc_qvectors: a bool to specify if q_vectors need to be allocated.
A
Anjali Singhai Jain 已提交
9760 9761 9762 9763
 *
 * On error: returns error code (negative)
 * On success: returns 0
 **/
9764
static int i40e_vsi_alloc_arrays(struct i40e_vsi *vsi, bool alloc_qvectors)
A
Anjali Singhai Jain 已提交
9765
{
9766
	struct i40e_ring **next_rings;
A
Anjali Singhai Jain 已提交
9767 9768 9769
	int size;
	int ret = 0;

9770 9771 9772
	/* 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
9773 9774 9775
	vsi->tx_rings = kzalloc(size, GFP_KERNEL);
	if (!vsi->tx_rings)
		return -ENOMEM;
9776 9777 9778 9779 9780 9781
	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
9782

9783 9784
	if (alloc_qvectors) {
		/* allocate memory for q_vector pointers */
9785
		size = sizeof(struct i40e_q_vector *) * vsi->num_q_vectors;
9786 9787 9788 9789 9790
		vsi->q_vectors = kzalloc(size, GFP_KERNEL);
		if (!vsi->q_vectors) {
			ret = -ENOMEM;
			goto err_vectors;
		}
A
Anjali Singhai Jain 已提交
9791 9792 9793 9794 9795 9796 9797 9798
	}
	return ret;

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

J
Jesse Brandeburg 已提交
9799 9800 9801 9802 9803 9804 9805 9806 9807 9808 9809 9810 9811 9812 9813 9814 9815 9816 9817 9818 9819 9820 9821 9822 9823
/**
 * 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
9824
	while (i < pf->num_alloc_vsi && pf->vsi[i])
J
Jesse Brandeburg 已提交
9825
		i++;
M
Mitch Williams 已提交
9826
	if (i >= pf->num_alloc_vsi) {
J
Jesse Brandeburg 已提交
9827 9828 9829 9830 9831
		i = 0;
		while (i < pf->next_vsi && pf->vsi[i])
			i++;
	}

M
Mitch Williams 已提交
9832
	if (i < pf->num_alloc_vsi && !pf->vsi[i]) {
J
Jesse Brandeburg 已提交
9833 9834 9835
		vsi_idx = i;             /* Found one! */
	} else {
		ret = -ENODEV;
9836
		goto unlock_pf;  /* out of VSI slots! */
J
Jesse Brandeburg 已提交
9837 9838 9839 9840 9841 9842
	}
	pf->next_vsi = ++i;

	vsi = kzalloc(sizeof(*vsi), GFP_KERNEL);
	if (!vsi) {
		ret = -ENOMEM;
9843
		goto unlock_pf;
J
Jesse Brandeburg 已提交
9844 9845 9846
	}
	vsi->type = type;
	vsi->back = pf;
9847
	set_bit(__I40E_VSI_DOWN, vsi->state);
J
Jesse Brandeburg 已提交
9848 9849
	vsi->flags = 0;
	vsi->idx = vsi_idx;
9850
	vsi->int_rate_limit = 0;
9851 9852
	vsi->rss_table_size = (vsi->type == I40E_VSI_MAIN) ?
				pf->rss_table_size : 64;
J
Jesse Brandeburg 已提交
9853 9854
	vsi->netdev_registered = false;
	vsi->work_limit = I40E_DEFAULT_IRQ_WORK;
9855
	hash_init(vsi->mac_filter_hash);
9856
	vsi->irqs_ready = false;
J
Jesse Brandeburg 已提交
9857

9858 9859 9860 9861
	ret = i40e_set_num_rings_in_vsi(vsi);
	if (ret)
		goto err_rings;

9862
	ret = i40e_vsi_alloc_arrays(vsi, true);
A
Anjali Singhai Jain 已提交
9863
	if (ret)
9864
		goto err_rings;
9865

J
Jesse Brandeburg 已提交
9866 9867 9868
	/* Setup default MSIX irq handler for VSI */
	i40e_vsi_setup_irqhandler(vsi, i40e_msix_clean_rings);

K
Kiran Patil 已提交
9869
	/* Initialize VSI lock */
9870
	spin_lock_init(&vsi->mac_filter_hash_lock);
J
Jesse Brandeburg 已提交
9871 9872
	pf->vsi[vsi_idx] = vsi;
	ret = vsi_idx;
9873 9874
	goto unlock_pf;

9875
err_rings:
9876 9877 9878
	pf->next_vsi = i - 1;
	kfree(vsi);
unlock_pf:
J
Jesse Brandeburg 已提交
9879 9880 9881 9882
	mutex_unlock(&pf->switch_mutex);
	return ret;
}

A
Anjali Singhai Jain 已提交
9883 9884 9885
/**
 * i40e_vsi_free_arrays - Free queue and vector pointer arrays for the VSI
 * @type: VSI pointer
9886
 * @free_qvectors: a bool to specify if q_vectors need to be freed.
A
Anjali Singhai Jain 已提交
9887 9888 9889 9890
 *
 * On error: returns error code (negative)
 * On success: returns 0
 **/
9891
static void i40e_vsi_free_arrays(struct i40e_vsi *vsi, bool free_qvectors)
A
Anjali Singhai Jain 已提交
9892 9893
{
	/* free the ring and vector containers */
9894 9895 9896 9897
	if (free_qvectors) {
		kfree(vsi->q_vectors);
		vsi->q_vectors = NULL;
	}
A
Anjali Singhai Jain 已提交
9898 9899 9900
	kfree(vsi->tx_rings);
	vsi->tx_rings = NULL;
	vsi->rx_rings = NULL;
9901
	vsi->xdp_rings = NULL;
A
Anjali Singhai Jain 已提交
9902 9903
}

9904 9905 9906 9907 9908 9909 9910 9911 9912 9913 9914 9915 9916 9917 9918 9919 9920
/**
 * 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
9921 9922 9923 9924 9925 9926 9927 9928 9929 9930 9931 9932 9933 9934 9935 9936 9937
/**
 * 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]) {
9938 9939
		dev_err(&pf->pdev->dev, "pf->vsi[%d] is NULL, just free vsi[%d](type %d)\n",
			vsi->idx, vsi->idx, vsi->type);
J
Jesse Brandeburg 已提交
9940 9941 9942 9943 9944
		goto unlock_vsi;
	}

	if (pf->vsi[vsi->idx] != vsi) {
		dev_err(&pf->pdev->dev,
9945
			"pf->vsi[%d](type %d) != vsi[%d](type %d): no free!\n",
J
Jesse Brandeburg 已提交
9946 9947
			pf->vsi[vsi->idx]->idx,
			pf->vsi[vsi->idx]->type,
9948
			vsi->idx, vsi->type);
J
Jesse Brandeburg 已提交
9949 9950 9951
		goto unlock_vsi;
	}

9952
	/* updates the PF for this cleared vsi */
J
Jesse Brandeburg 已提交
9953 9954 9955
	i40e_put_lump(pf->qp_pile, vsi->base_queue, vsi->idx);
	i40e_put_lump(pf->irq_pile, vsi->base_vector, vsi->idx);

9956
	i40e_vsi_free_arrays(vsi, true);
9957
	i40e_clear_rss_config_user(vsi);
9958

J
Jesse Brandeburg 已提交
9959 9960 9961 9962 9963 9964 9965 9966 9967 9968 9969 9970
	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;
}

9971 9972 9973 9974
/**
 * i40e_vsi_clear_rings - Deallocates the Rx and Tx rings for the provided VSI
 * @vsi: the VSI being cleaned
 **/
9975
static void i40e_vsi_clear_rings(struct i40e_vsi *vsi)
9976 9977 9978
{
	int i;

9979
	if (vsi->tx_rings && vsi->tx_rings[0]) {
N
Neerav Parikh 已提交
9980
		for (i = 0; i < vsi->alloc_queue_pairs; i++) {
9981 9982 9983
			kfree_rcu(vsi->tx_rings[i], rcu);
			vsi->tx_rings[i] = NULL;
			vsi->rx_rings[i] = NULL;
9984 9985
			if (vsi->xdp_rings)
				vsi->xdp_rings[i] = NULL;
9986
		}
9987
	}
9988 9989
}

J
Jesse Brandeburg 已提交
9990 9991 9992 9993 9994 9995
/**
 * 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)
{
9996
	int i, qpv = i40e_enabled_xdp_vsi(vsi) ? 3 : 2;
J
Jesse Brandeburg 已提交
9997
	struct i40e_pf *pf = vsi->back;
9998
	struct i40e_ring *ring;
J
Jesse Brandeburg 已提交
9999 10000

	/* Set basic values in the rings to be used later during open() */
N
Neerav Parikh 已提交
10001
	for (i = 0; i < vsi->alloc_queue_pairs; i++) {
10002
		/* allocate space for both Tx and Rx in one shot */
10003 10004
		ring = kcalloc(qpv, sizeof(struct i40e_ring), GFP_KERNEL);
		if (!ring)
10005
			goto err_out;
J
Jesse Brandeburg 已提交
10006

10007 10008 10009 10010 10011 10012 10013 10014 10015
		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;
10016
		if (vsi->back->hw_features & I40E_HW_WB_ON_ITR_CAPABLE)
10017
			ring->flags = I40E_TXR_FLAGS_WB_ON_ITR;
10018
		ring->itr_setting = pf->tx_itr_default;
10019 10020 10021 10022 10023 10024 10025 10026 10027 10028 10029 10030 10031 10032
		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;
10033
		if (vsi->back->hw_features & I40E_HW_WB_ON_ITR_CAPABLE)
10034 10035
			ring->flags = I40E_TXR_FLAGS_WB_ON_ITR;
		set_ring_xdp(ring);
10036
		ring->itr_setting = pf->tx_itr_default;
10037 10038 10039 10040 10041 10042 10043 10044 10045 10046 10047 10048
		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;
10049
		ring->itr_setting = pf->rx_itr_default;
10050
		vsi->rx_rings[i] = ring;
J
Jesse Brandeburg 已提交
10051 10052 10053
	}

	return 0;
10054 10055 10056 10057

err_out:
	i40e_vsi_clear_rings(vsi);
	return -ENOMEM;
J
Jesse Brandeburg 已提交
10058 10059 10060 10061 10062 10063 10064 10065 10066 10067 10068
}

/**
 * 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)
{
10069 10070 10071
	vectors = pci_enable_msix_range(pf->pdev, pf->msix_entries,
					I40E_MIN_MSIX, vectors);
	if (vectors < 0) {
J
Jesse Brandeburg 已提交
10072
		dev_info(&pf->pdev->dev,
10073
			 "MSI-X vector reservation failed: %d\n", vectors);
J
Jesse Brandeburg 已提交
10074 10075 10076 10077 10078 10079 10080 10081 10082 10083 10084 10085
		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.
 *
10086
 * Returns the number of vectors reserved or negative on failure
J
Jesse Brandeburg 已提交
10087 10088 10089 10090
 **/
static int i40e_init_msix(struct i40e_pf *pf)
{
	struct i40e_hw *hw = &pf->hw;
10091
	int cpus, extra_vectors;
S
Shannon Nelson 已提交
10092
	int vectors_left;
J
Jesse Brandeburg 已提交
10093
	int v_budget, i;
10094
	int v_actual;
10095
	int iwarp_requested = 0;
J
Jesse Brandeburg 已提交
10096 10097 10098 10099 10100 10101 10102

	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
10103 10104 10105 10106 10107
	 *	- 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
10108
	 *   - The number of VMDq pairs
10109
	 *   - The CPU count within the NUMA node if iWARP is enabled
J
Jesse Brandeburg 已提交
10110 10111 10112 10113 10114
	 * 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
10115 10116 10117 10118 10119 10120 10121 10122 10123
	vectors_left = hw->func_caps.num_msix_vectors;
	v_budget = 0;

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

10124 10125 10126 10127 10128 10129 10130 10131 10132
	/* 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
10133 10134 10135 10136 10137
	vectors_left -= pf->num_lan_msix;

	/* reserve one vector for sideband flow director */
	if (pf->flags & I40E_FLAG_FD_SB_ENABLED) {
		if (vectors_left) {
10138
			pf->num_fdsb_msix = 1;
S
Shannon Nelson 已提交
10139 10140 10141
			v_budget++;
			vectors_left--;
		} else {
10142
			pf->num_fdsb_msix = 0;
S
Shannon Nelson 已提交
10143 10144
		}
	}
10145

10146 10147
	/* can we reserve enough for iWARP? */
	if (pf->flags & I40E_FLAG_IWARP_ENABLED) {
10148 10149
		iwarp_requested = pf->num_iwarp_msix;

10150 10151 10152 10153 10154 10155 10156 10157
		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
10158 10159 10160 10161 10162
	/* 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);

10163 10164 10165 10166 10167 10168 10169 10170 10171 10172 10173 10174 10175
		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
10176

10177 10178 10179
			v_budget += vmdq_vecs;
			vectors_left -= vmdq_vecs;
		}
S
Shannon Nelson 已提交
10180
	}
J
Jesse Brandeburg 已提交
10181

10182 10183 10184 10185 10186 10187 10188 10189 10190 10191 10192 10193 10194 10195 10196 10197 10198
	/* 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
10199 10200 10201 10202 10203 10204 10205
	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;
10206
	v_actual = i40e_reserve_msix_vectors(pf, v_budget);
10207

10208
	if (v_actual < I40E_MIN_MSIX) {
J
Jesse Brandeburg 已提交
10209 10210 10211
		pf->flags &= ~I40E_FLAG_MSIX_ENABLED;
		kfree(pf->msix_entries);
		pf->msix_entries = NULL;
10212
		pci_disable_msix(pf->pdev);
J
Jesse Brandeburg 已提交
10213 10214
		return -ENODEV;

10215
	} else if (v_actual == I40E_MIN_MSIX) {
J
Jesse Brandeburg 已提交
10216 10217 10218 10219 10220 10221
		/* 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;

10222
	} else if (v_actual != v_budget) {
10223 10224 10225 10226 10227
		/* 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.
		 */
10228 10229
		int vec;

10230
		dev_info(&pf->pdev->dev,
10231 10232
			 "MSI-X vector limit reached with %d, wanted %d, attempting to redistribute vectors\n",
			 v_actual, v_budget);
10233
		/* reserve the misc vector */
10234
		vec = v_actual - 1;
10235

J
Jesse Brandeburg 已提交
10236 10237
		/* Scale vector usage down */
		pf->num_vmdq_msix = 1;    /* force VMDqs to only one vector */
10238
		pf->num_vmdq_vsis = 1;
S
Shannon Nelson 已提交
10239
		pf->num_vmdq_qps = 1;
J
Jesse Brandeburg 已提交
10240 10241 10242 10243 10244 10245 10246

		/* partition out the remaining vectors */
		switch (vec) {
		case 2:
			pf->num_lan_msix = 1;
			break;
		case 3:
10247 10248 10249 10250 10251 10252
			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
10253 10254
			break;
		default:
10255 10256 10257 10258 10259 10260 10261 10262 10263
			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);
			}
10264 10265 10266 10267
			if (pf->flags & I40E_FLAG_FD_SB_ENABLED) {
				pf->num_fdsb_msix = 1;
				vec--;
			}
10268 10269 10270
			pf->num_lan_msix = min_t(int,
			       (vec - (pf->num_iwarp_msix + pf->num_vmdq_vsis)),
							      pf->num_lan_msix);
10271
			pf->num_lan_qps = pf->num_lan_msix;
J
Jesse Brandeburg 已提交
10272 10273 10274 10275
			break;
		}
	}

10276 10277 10278 10279
	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;
10280
		pf->flags |= I40E_FLAG_FD_SB_INACTIVE;
10281
	}
10282 10283 10284 10285 10286
	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;
	}
10287 10288 10289 10290 10291 10292

	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;
	}
10293 10294 10295 10296 10297 10298 10299
	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);

10300
	return v_actual;
J
Jesse Brandeburg 已提交
10301 10302
}

10303
/**
10304
 * i40e_vsi_alloc_q_vector - Allocate memory for a single interrupt vector
10305 10306
 * @vsi: the VSI being configured
 * @v_idx: index of the vector in the vsi struct
10307
 * @cpu: cpu to be used on affinity_mask
10308 10309 10310
 *
 * We allocate one q_vector.  If allocation fails we return -ENOMEM.
 **/
10311
static int i40e_vsi_alloc_q_vector(struct i40e_vsi *vsi, int v_idx, int cpu)
10312 10313 10314 10315 10316 10317 10318 10319 10320 10321
{
	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;
10322
	cpumask_copy(&q_vector->affinity_mask, cpu_possible_mask);
10323

10324 10325
	if (vsi->netdev)
		netif_napi_add(vsi->netdev, &q_vector->napi,
J
Jesse Brandeburg 已提交
10326
			       i40e_napi_poll, NAPI_POLL_WEIGHT);
10327 10328 10329 10330 10331 10332 10333

	/* tie q_vector and vsi together */
	vsi->q_vectors[v_idx] = q_vector;

	return 0;
}

J
Jesse Brandeburg 已提交
10334
/**
10335
 * i40e_vsi_alloc_q_vectors - Allocate memory for interrupt vectors
J
Jesse Brandeburg 已提交
10336 10337 10338 10339 10340
 * @vsi: the VSI being configured
 *
 * We allocate one q_vector per queue interrupt.  If allocation fails we
 * return -ENOMEM.
 **/
10341
static int i40e_vsi_alloc_q_vectors(struct i40e_vsi *vsi)
J
Jesse Brandeburg 已提交
10342 10343
{
	struct i40e_pf *pf = vsi->back;
10344
	int err, v_idx, num_q_vectors, current_cpu;
J
Jesse Brandeburg 已提交
10345 10346 10347 10348 10349 10350 10351 10352 10353

	/* 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;

10354 10355
	current_cpu = cpumask_first(cpu_online_mask);

J
Jesse Brandeburg 已提交
10356
	for (v_idx = 0; v_idx < num_q_vectors; v_idx++) {
10357
		err = i40e_vsi_alloc_q_vector(vsi, v_idx, current_cpu);
10358 10359
		if (err)
			goto err_out;
10360 10361 10362
		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
10363 10364 10365
	}

	return 0;
10366 10367 10368 10369 10370 10371

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

	return err;
J
Jesse Brandeburg 已提交
10372 10373 10374 10375 10376 10377
}

/**
 * i40e_init_interrupt_scheme - Determine proper interrupt scheme
 * @pf: board private structure to initialize
 **/
10378
static int i40e_init_interrupt_scheme(struct i40e_pf *pf)
J
Jesse Brandeburg 已提交
10379
{
10380 10381
	int vectors = 0;
	ssize_t size;
J
Jesse Brandeburg 已提交
10382 10383

	if (pf->flags & I40E_FLAG_MSIX_ENABLED) {
10384 10385
		vectors = i40e_init_msix(pf);
		if (vectors < 0) {
J
Jesse Brandeburg 已提交
10386
			pf->flags &= ~(I40E_FLAG_MSIX_ENABLED	|
10387
				       I40E_FLAG_IWARP_ENABLED	|
J
Jesse Brandeburg 已提交
10388
				       I40E_FLAG_RSS_ENABLED	|
10389
				       I40E_FLAG_DCB_CAPABLE	|
10390
				       I40E_FLAG_DCB_ENABLED	|
J
Jesse Brandeburg 已提交
10391 10392 10393 10394
				       I40E_FLAG_SRIOV_ENABLED	|
				       I40E_FLAG_FD_SB_ENABLED	|
				       I40E_FLAG_FD_ATR_ENABLED	|
				       I40E_FLAG_VMDQ_ENABLED);
10395
			pf->flags |= I40E_FLAG_FD_SB_INACTIVE;
J
Jesse Brandeburg 已提交
10396 10397 10398 10399 10400 10401 10402 10403

			/* 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
10404
		dev_info(&pf->pdev->dev, "MSI-X not available, trying MSI\n");
10405 10406 10407 10408
		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
10409 10410
			pf->flags &= ~I40E_FLAG_MSI_ENABLED;
		}
10411
		vectors = 1;  /* one MSI or Legacy vector */
J
Jesse Brandeburg 已提交
10412 10413
	}

10414
	if (!(pf->flags & (I40E_FLAG_MSIX_ENABLED | I40E_FLAG_MSI_ENABLED)))
C
Catherine Sullivan 已提交
10415
		dev_info(&pf->pdev->dev, "MSI-X and MSI not available, falling back to Legacy IRQ\n");
10416

10417 10418 10419
	/* set up vector assignment tracking */
	size = sizeof(struct i40e_lump_tracking) + (sizeof(u16) * vectors);
	pf->irq_pile = kzalloc(size, GFP_KERNEL);
10420
	if (!pf->irq_pile)
10421
		return -ENOMEM;
10422

10423 10424 10425
	pf->irq_pile->num_entries = vectors;
	pf->irq_pile->search_hint = 0;

10426
	/* track first vector for misc interrupts, ignore return */
10427
	(void)i40e_get_lump(pf, pf->irq_pile, 1, I40E_PILE_VALID_BIT - 1);
10428 10429

	return 0;
J
Jesse Brandeburg 已提交
10430 10431
}

10432 10433 10434 10435 10436 10437 10438 10439 10440 10441 10442 10443 10444 10445 10446 10447 10448 10449 10450 10451 10452 10453 10454 10455 10456 10457 10458 10459 10460 10461 10462 10463 10464 10465 10466 10467 10468 10469 10470 10471 10472 10473 10474 10475 10476 10477 10478 10479 10480
/**
 * 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;
}

J
Jesse Brandeburg 已提交
10481 10482 10483 10484 10485 10486 10487 10488 10489 10490 10491 10492 10493
/**
 * 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;

10494 10495
	/* 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
10496
		err = request_irq(pf->msix_entries[0].vector,
10497
				  i40e_intr, 0, pf->int_name, pf);
J
Jesse Brandeburg 已提交
10498
		if (err) {
10499
			clear_bit(__I40E_MISC_IRQ_REQUESTED, pf->state);
J
Jesse Brandeburg 已提交
10500
			dev_info(&pf->pdev->dev,
C
Catherine Sullivan 已提交
10501
				 "request_irq for %s failed: %d\n",
10502
				 pf->int_name, err);
J
Jesse Brandeburg 已提交
10503 10504 10505 10506
			return -EFAULT;
		}
	}

10507
	i40e_enable_misc_int_causes(pf);
J
Jesse Brandeburg 已提交
10508 10509 10510 10511 10512 10513 10514

	/* 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);

10515
	i40e_irq_dynamic_enable_icr0(pf);
J
Jesse Brandeburg 已提交
10516 10517 10518 10519

	return err;
}

10520 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 10560 10561 10562 10563 10564 10565
/**
 * 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;
}

10566
/**
10567
 * i40e_config_rss_reg - Configure RSS keys and lut by writing registers
10568
 * @vsi: Pointer to vsi structure
10569
 * @seed: RSS hash seed
10570 10571 10572 10573
 * @lut: Lookup table
 * @lut_size: Lookup table size
 *
 * Returns 0 on success, negative on failure
J
Jesse Brandeburg 已提交
10574
 **/
10575 10576
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
10577
{
10578
	struct i40e_pf *pf = vsi->back;
10579
	struct i40e_hw *hw = &pf->hw;
10580
	u16 vf_id = vsi->vf_id;
10581
	u8 i;
J
Jesse Brandeburg 已提交
10582

10583
	/* Fill out hash function seed */
10584 10585 10586
	if (seed) {
		u32 *seed_dw = (u32 *)seed;

10587 10588
		if (vsi->type == I40E_VSI_MAIN) {
			for (i = 0; i <= I40E_PFQF_HKEY_MAX_INDEX; i++)
10589
				wr32(hw, I40E_PFQF_HKEY(i), seed_dw[i]);
10590 10591
		} else if (vsi->type == I40E_VSI_SRIOV) {
			for (i = 0; i <= I40E_VFQF_HKEY1_MAX_INDEX; i++)
10592
				wr32(hw, I40E_VFQF_HKEY1(i, vf_id), seed_dw[i]);
10593 10594 10595
		} else {
			dev_err(&pf->pdev->dev, "Cannot set RSS seed - invalid VSI type\n");
		}
10596 10597 10598 10599 10600
	}

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

10601 10602 10603 10604 10605 10606 10607 10608 10609
		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++)
10610
				wr32(hw, I40E_VFQF_HLUT1(i, vf_id), lut_dw[i]);
10611 10612 10613
		} else {
			dev_err(&pf->pdev->dev, "Cannot set RSS LUT - invalid VSI type\n");
		}
10614 10615 10616 10617 10618 10619
	}
	i40e_flush(hw);

	return 0;
}

10620 10621 10622 10623 10624 10625 10626 10627 10628 10629 10630 10631 10632 10633 10634 10635 10636 10637 10638 10639
/**
 * 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++)
10640
			seed_dw[i] = i40e_read_rx_ctl(hw, I40E_PFQF_HKEY(i));
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 10666
	}
	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;

10667
	if (pf->hw_features & I40E_HW_RSS_AQ_CAPABLE)
10668 10669 10670 10671 10672 10673 10674 10675 10676 10677 10678 10679 10680 10681 10682 10683
		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)
{
10684 10685
	struct i40e_pf *pf = vsi->back;

10686
	if (pf->hw_features & I40E_HW_RSS_AQ_CAPABLE)
10687 10688 10689
		return i40e_get_rss_aq(vsi, seed, lut, lut_size);
	else
		return i40e_get_rss_reg(vsi, seed, lut, lut_size);
10690 10691
}

10692 10693 10694 10695 10696 10697 10698
/**
 * 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
 */
10699 10700
void i40e_fill_rss_lut(struct i40e_pf *pf, u8 *lut,
		       u16 rss_table_size, u16 rss_size)
10701 10702 10703 10704 10705 10706 10707
{
	u16 i;

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

10708
/**
10709
 * i40e_pf_config_rss - Prepare for RSS if used
10710 10711
 * @pf: board private structure
 **/
10712
static int i40e_pf_config_rss(struct i40e_pf *pf)
10713 10714 10715
{
	struct i40e_vsi *vsi = pf->vsi[pf->lan_vsi];
	u8 seed[I40E_HKEY_ARRAY_SIZE];
10716
	u8 *lut;
10717 10718 10719
	struct i40e_hw *hw = &pf->hw;
	u32 reg_val;
	u64 hena;
10720
	int ret;
10721

J
Jesse Brandeburg 已提交
10722
	/* By default we enable TCP/UDP with IPv4/IPv6 ptypes */
10723 10724
	hena = (u64)i40e_read_rx_ctl(hw, I40E_PFQF_HENA(0)) |
		((u64)i40e_read_rx_ctl(hw, I40E_PFQF_HENA(1)) << 32);
10725 10726
	hena |= i40e_pf_get_default_rss_hena(pf);

10727 10728
	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
10729

10730
	/* Determine the RSS table size based on the hardware capabilities */
10731
	reg_val = i40e_read_rx_ctl(hw, I40E_PFQF_CTL_0);
10732 10733 10734
	reg_val = (pf->rss_table_size == 512) ?
			(reg_val | I40E_PFQF_CTL_0_HASHLUTSIZE_512) :
			(reg_val & ~I40E_PFQF_CTL_0_HASHLUTSIZE_512);
10735
	i40e_write_rx_ctl(hw, I40E_PFQF_CTL_0, reg_val);
10736

10737
	/* Determine the RSS size of the VSI */
10738 10739
	if (!vsi->rss_size) {
		u16 qcount;
M
Mitch Williams 已提交
10740 10741 10742 10743 10744 10745 10746
		/* If the firmware does something weird during VSI init, we
		 * could end up with zero TCs. Check for that to avoid
		 * divide-by-zero. It probably won't pass traffic, but it also
		 * won't panic.
		 */
		qcount = vsi->num_queue_pairs /
			 (vsi->tc_config.numtc ? vsi->tc_config.numtc : 1);
10747 10748
		vsi->rss_size = min_t(int, pf->alloc_rss_size, qcount);
	}
10749 10750
	if (!vsi->rss_size)
		return -EINVAL;
10751

10752 10753 10754 10755
	lut = kzalloc(vsi->rss_table_size, GFP_KERNEL);
	if (!lut)
		return -ENOMEM;

10756 10757 10758 10759 10760
	/* 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);
10761

10762 10763 10764 10765 10766 10767 10768
	/* 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);
10769
	ret = i40e_config_rss(vsi, seed, lut, vsi->rss_table_size);
10770 10771 10772
	kfree(lut);

	return ret;
J
Jesse Brandeburg 已提交
10773 10774
}

10775 10776 10777 10778 10779 10780 10781
/**
 * 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.
10782
 * Note: expects to be called while under rtnl_lock()
10783 10784 10785
 **/
int i40e_reconfig_rss_queues(struct i40e_pf *pf, int queue_count)
{
10786 10787 10788
	struct i40e_vsi *vsi = pf->vsi[pf->lan_vsi];
	int new_rss_size;

10789 10790 10791
	if (!(pf->flags & I40E_FLAG_RSS_ENABLED))
		return 0;

10792
	new_rss_size = min_t(int, queue_count, pf->rss_size_max);
10793

10794
	if (queue_count != vsi->num_queue_pairs) {
10795 10796
		u16 qcount;

10797
		vsi->req_queue_pairs = queue_count;
10798
		i40e_prep_for_reset(pf, true);
10799

10800
		pf->alloc_rss_size = new_rss_size;
10801

10802
		i40e_reset_and_rebuild(pf, true, true);
10803 10804 10805 10806 10807 10808 10809 10810 10811 10812 10813

		/* 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 */
10814 10815
		qcount = vsi->num_queue_pairs / vsi->tc_config.numtc;
		vsi->rss_size = min_t(int, pf->alloc_rss_size, qcount);
10816

10817
		i40e_pf_config_rss(pf);
10818
	}
10819 10820
	dev_info(&pf->pdev->dev, "User requested queue count/HW max RSS count:  %d/%d\n",
		 vsi->req_queue_pairs, pf->rss_size_max);
10821
	return pf->alloc_rss_size;
10822 10823
}

10824
/**
10825
 * i40e_get_partition_bw_setting - Retrieve BW settings for this PF partition
10826 10827
 * @pf: board private structure
 **/
10828
i40e_status i40e_get_partition_bw_setting(struct i40e_pf *pf)
10829 10830 10831 10832 10833 10834 10835 10836 10837 10838
{
	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)
10839
			pf->min_bw = min_bw;
10840
		if (max_valid)
10841
			pf->max_bw = max_bw;
10842 10843 10844 10845 10846 10847
	}

	return status;
}

/**
10848
 * i40e_set_partition_bw_setting - Set BW settings for this PF partition
10849 10850
 * @pf: board private structure
 **/
10851
i40e_status i40e_set_partition_bw_setting(struct i40e_pf *pf)
10852 10853 10854 10855
{
	struct i40e_aqc_configure_partition_bw_data bw_data;
	i40e_status status;

10856
	/* Set the valid bit for this PF */
10857
	bw_data.pf_valid_bits = cpu_to_le16(BIT(pf->hw.pf_id));
10858 10859
	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;
10860 10861 10862 10863 10864 10865 10866 10867

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

	return status;
}

/**
10868
 * i40e_commit_partition_bw_setting - Commit BW settings for this PF partition
10869 10870
 * @pf: board private structure
 **/
10871
i40e_status i40e_commit_partition_bw_setting(struct i40e_pf *pf)
10872 10873 10874 10875 10876 10877 10878 10879 10880 10881 10882 10883 10884 10885 10886 10887 10888 10889 10890
{
	/* 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,
10891 10892 10893
			 "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));
10894 10895 10896 10897 10898 10899 10900 10901 10902 10903 10904 10905 10906 10907
		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) {
10908 10909 10910
		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));
10911 10912 10913 10914 10915 10916 10917 10918 10919 10920 10921
		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,
10922 10923 10924
			 "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));
10925 10926 10927 10928 10929 10930 10931 10932 10933
		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),
10934
				 &nvm_word, true, 0, NULL);
10935 10936 10937 10938 10939 10940 10941
	/* 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,
10942 10943 10944
			 "BW settings NOT SAVED, err %s aq_err %s\n",
			 i40e_stat_str(&pf->hw, ret),
			 i40e_aq_str(&pf->hw, last_aq_status));
10945 10946 10947 10948 10949
bw_commit_out:

	return ret;
}

J
Jesse Brandeburg 已提交
10950 10951 10952 10953 10954 10955 10956 10957 10958 10959 10960 10961 10962 10963 10964 10965
/**
 * 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     |
10966 10967
		    I40E_FLAG_MSIX_ENABLED;

10968
	/* Set default ITR */
10969 10970
	pf->rx_itr_default = I40E_ITR_RX_DEF;
	pf->tx_itr_default = I40E_ITR_TX_DEF;
10971

10972 10973 10974
	/* Depending on PF configurations, it is possible that the RSS
	 * maximum might end up larger than the available queues
	 */
10975
	pf->rss_size_max = BIT(pf->hw.func_caps.rss_table_entry_width);
10976
	pf->alloc_rss_size = 1;
10977
	pf->rss_table_size = pf->hw.func_caps.rss_table_size;
10978 10979
	pf->rss_size_max = min_t(int, pf->rss_size_max,
				 pf->hw.func_caps.num_tx_qp);
J
Jesse Brandeburg 已提交
10980 10981
	if (pf->hw.func_caps.rss) {
		pf->flags |= I40E_FLAG_RSS_ENABLED;
10982 10983
		pf->alloc_rss_size = min_t(int, pf->rss_size_max,
					   num_online_cpus());
J
Jesse Brandeburg 已提交
10984 10985
	}

10986
	/* MFP mode enabled */
10987
	if (pf->hw.func_caps.npar_enable || pf->hw.func_caps.flex10_enable) {
10988 10989
		pf->flags |= I40E_FLAG_MFP_ENABLED;
		dev_info(&pf->pdev->dev, "MFP mode Enabled\n");
10990
		if (i40e_get_partition_bw_setting(pf)) {
10991
			dev_warn(&pf->pdev->dev,
10992 10993
				 "Could not get partition bw settings\n");
		} else {
10994
			dev_info(&pf->pdev->dev,
10995 10996 10997 10998 10999 11000
				 "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);
		}
11001 11002
	}

11003 11004 11005 11006
	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;
11007 11008
		if (pf->flags & I40E_FLAG_MFP_ENABLED &&
		    pf->hw.num_partitions > 1)
11009
			dev_info(&pf->pdev->dev,
A
Anjali Singhai Jain 已提交
11010
				 "Flow Director Sideband mode Disabled in MFP mode\n");
11011 11012
		else
			pf->flags |= I40E_FLAG_FD_SB_ENABLED;
11013 11014 11015 11016
		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
11017 11018
	}

11019
	if (pf->hw.mac.type == I40E_MAC_X722) {
11020 11021 11022 11023 11024 11025 11026 11027 11028 11029 11030
		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);
11031 11032 11033 11034 11035 11036 11037 11038

#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;
		}
11039 11040 11041 11042
	} 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 */
11043
		pf->hw_features |= I40E_HW_GENEVE_OFFLOAD_CAPABLE;
11044 11045 11046
	}

	/* Enable HW ATR eviction if possible */
11047
	if (pf->hw_features & I40E_HW_ATR_EVICT_CAPABLE)
11048 11049
		pf->flags |= I40E_FLAG_HW_ATR_EVICT_ENABLED;

B
Bimmy Pujari 已提交
11050
	if ((pf->hw.mac.type == I40E_MAC_XL710) &&
11051
	    (((pf->hw.aq.fw_maj_ver == 4) && (pf->hw.aq.fw_min_ver < 33)) ||
11052
	    (pf->hw.aq.fw_maj_ver < 4))) {
11053
		pf->hw_features |= I40E_HW_RESTART_AUTONEG;
11054
		/* No DCB support  for FW < v4.33 */
11055
		pf->hw_features |= I40E_HW_NO_DCB_SUPPORT;
11056 11057 11058
	}

	/* Disable FW LLDP if FW < v4.3 */
B
Bimmy Pujari 已提交
11059
	if ((pf->hw.mac.type == I40E_MAC_XL710) &&
11060 11061
	    (((pf->hw.aq.fw_maj_ver == 4) && (pf->hw.aq.fw_min_ver < 3)) ||
	    (pf->hw.aq.fw_maj_ver < 4)))
11062
		pf->hw_features |= I40E_HW_STOP_FW_LLDP;
11063 11064

	/* Use the FW Set LLDP MIB API if FW > v4.40 */
B
Bimmy Pujari 已提交
11065
	if ((pf->hw.mac.type == I40E_MAC_XL710) &&
11066 11067
	    (((pf->hw.aq.fw_maj_ver == 4) && (pf->hw.aq.fw_min_ver >= 40)) ||
	    (pf->hw.aq.fw_maj_ver >= 5)))
11068
		pf->hw_features |= I40E_HW_USE_SET_LLDP_MIB;
11069

11070 11071 11072 11073 11074
	/* 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;

A
Avinash Dayanand 已提交
11075
	if (pf->hw.func_caps.vmdq && num_online_cpus() != 1) {
J
Jesse Brandeburg 已提交
11076
		pf->num_vmdq_vsis = I40E_DEFAULT_NUM_VMDQ_VSI;
11077
		pf->flags |= I40E_FLAG_VMDQ_ENABLED;
J
Jesse Brandeburg 已提交
11078
		pf->num_vmdq_qps = i40e_default_queues_per_vmdq(pf);
J
Jesse Brandeburg 已提交
11079 11080
	}

A
Avinash Dayanand 已提交
11081
	if (pf->hw.func_caps.iwarp && num_online_cpus() != 1) {
11082 11083 11084 11085
		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;
	}
11086 11087 11088 11089 11090 11091 11092 11093 11094 11095
	/* Stopping the FW LLDP engine is only supported on the
	 * XL710 with a FW ver >= 1.7.  Also, stopping FW LLDP
	 * engine is not supported if NPAR is functioning on this
	 * part
	 */
	if (pf->hw.mac.type == I40E_MAC_XL710 &&
	    !pf->hw.func_caps.npar_enable &&
	    (pf->hw.aq.api_maj_ver > 1 ||
	     (pf->hw.aq.api_maj_ver == 1 && pf->hw.aq.api_min_ver > 6)))
		pf->hw_features |= I40E_HW_STOPPABLE_FW_LLDP;
11096

J
Jesse Brandeburg 已提交
11097
#ifdef CONFIG_PCI_IOV
11098
	if (pf->hw.func_caps.num_vfs && pf->hw.partition_id == 1) {
J
Jesse Brandeburg 已提交
11099 11100 11101 11102 11103 11104 11105 11106 11107 11108 11109
		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;

11110 11111 11112
	/* By default FW has this off for performance reasons */
	pf->flags &= ~I40E_FLAG_VEB_STATS_ENABLED;

J
Jesse Brandeburg 已提交
11113 11114 11115 11116 11117 11118 11119 11120 11121 11122 11123
	/* 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;

11124 11125
	pf->tx_timeout_recovery_level = 1;

J
Jesse Brandeburg 已提交
11126 11127 11128 11129 11130 11131
	mutex_init(&pf->switch_mutex);

sw_init_done:
	return err;
}

11132 11133 11134 11135 11136 11137 11138 11139 11140 11141 11142 11143 11144 11145 11146 11147 11148 11149
/**
 * 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;
11150 11151 11152 11153
		/* enable FD_SB only if there is MSI-X vector and no cloud
		 * filters exist
		 */
		if (pf->num_fdsb_msix > 0 && !pf->num_cloud_filters) {
11154
			pf->flags |= I40E_FLAG_FD_SB_ENABLED;
11155 11156
			pf->flags &= ~I40E_FLAG_FD_SB_INACTIVE;
		}
11157 11158 11159 11160 11161 11162
	} 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);
		}
11163 11164
		pf->flags &= ~(I40E_FLAG_FD_SB_ENABLED |
			       I40E_FLAG_FD_SB_AUTO_DISABLED);
11165 11166
		pf->flags |= I40E_FLAG_FD_SB_INACTIVE;

11167
		/* reset fd counters */
11168 11169
		pf->fd_add_err = 0;
		pf->fd_atr_cnt = 0;
11170
		/* if ATR was auto disabled it can be re-enabled. */
11171 11172 11173 11174
		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))
11175 11176
				dev_info(&pf->pdev->dev, "ATR re-enabled.\n");
		}
11177 11178 11179 11180
	}
	return need_reset;
}

11181 11182 11183 11184 11185 11186 11187 11188 11189 11190 11191 11192 11193 11194 11195 11196 11197 11198 11199 11200 11201 11202
/**
 * 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
11203 11204 11205 11206
/**
 * i40e_set_features - set the netdev feature flags
 * @netdev: ptr to the netdev being adjusted
 * @features: the feature set that the stack is suggesting
11207
 * Note: expects to be called while under rtnl_lock()
J
Jesse Brandeburg 已提交
11208 11209 11210 11211 11212 11213
 **/
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;
11214 11215
	struct i40e_pf *pf = vsi->back;
	bool need_reset;
J
Jesse Brandeburg 已提交
11216

11217 11218 11219 11220 11221 11222
	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
11223 11224 11225 11226 11227
	if (features & NETIF_F_HW_VLAN_CTAG_RX)
		i40e_vlan_stripping_enable(vsi);
	else
		i40e_vlan_stripping_disable(vsi);

11228 11229 11230 11231 11232 11233
	if (!(features & NETIF_F_HW_TC) && pf->num_cloud_filters) {
		dev_err(&pf->pdev->dev,
			"Offloaded tc filters active, can't turn hw_tc_offload off");
		return -EINVAL;
	}

11234 11235 11236
	need_reset = i40e_set_ntuple(pf, features);

	if (need_reset)
A
Amritha Nambiar 已提交
11237
		i40e_do_reset(pf, I40E_PF_RESET_FLAG, true);
11238

J
Jesse Brandeburg 已提交
11239 11240 11241
	return 0;
}

11242
/**
11243
 * i40e_get_udp_port_idx - Lookup a possibly offloaded for Rx UDP port
11244 11245 11246 11247 11248
 * @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
 **/
11249
static u8 i40e_get_udp_port_idx(struct i40e_pf *pf, u16 port)
11250 11251 11252 11253
{
	u8 i;

	for (i = 0; i < I40E_MAX_PF_UDP_OFFLOAD_PORTS; i++) {
11254
		if (pf->udp_ports[i].port == port)
11255 11256 11257 11258 11259 11260 11261
			return i;
	}

	return i;
}

/**
11262
 * i40e_udp_tunnel_add - Get notifications about UDP tunnel ports that come up
11263
 * @netdev: This physical port's netdev
11264
 * @ti: Tunnel endpoint information
11265
 **/
11266 11267
static void i40e_udp_tunnel_add(struct net_device *netdev,
				struct udp_tunnel_info *ti)
11268 11269 11270 11271
{
	struct i40e_netdev_priv *np = netdev_priv(netdev);
	struct i40e_vsi *vsi = np->vsi;
	struct i40e_pf *pf = vsi->back;
11272
	u16 port = ntohs(ti->port);
11273 11274 11275
	u8 next_idx;
	u8 idx;

11276
	idx = i40e_get_udp_port_idx(pf, port);
11277 11278 11279

	/* Check if port already exists */
	if (idx < I40E_MAX_PF_UDP_OFFLOAD_PORTS) {
11280
		netdev_info(netdev, "port %d already offloaded\n", port);
11281 11282 11283 11284
		return;
	}

	/* Now check if there is space to add the new port */
11285
	next_idx = i40e_get_udp_port_idx(pf, 0);
11286 11287

	if (next_idx == I40E_MAX_PF_UDP_OFFLOAD_PORTS) {
11288
		netdev_info(netdev, "maximum number of offloaded UDP ports reached, not adding port %d\n",
11289
			    port);
11290 11291 11292
		return;
	}

11293 11294 11295 11296 11297
	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:
11298
		if (!(pf->hw_features & I40E_HW_GENEVE_OFFLOAD_CAPABLE))
11299 11300 11301 11302
			return;
		pf->udp_ports[next_idx].type = I40E_AQC_TUNNEL_TYPE_NGE;
		break;
	default:
11303 11304 11305 11306
		return;
	}

	/* New port: add it and mark its index in the bitmap */
11307
	pf->udp_ports[next_idx].port = port;
11308 11309
	pf->pending_udp_bitmap |= BIT_ULL(next_idx);
	pf->flags |= I40E_FLAG_UDP_FILTER_SYNC;
11310 11311
}

11312
/**
11313
 * i40e_udp_tunnel_del - Get notifications about UDP tunnel ports that go away
11314
 * @netdev: This physical port's netdev
11315
 * @ti: Tunnel endpoint information
11316
 **/
11317 11318
static void i40e_udp_tunnel_del(struct net_device *netdev,
				struct udp_tunnel_info *ti)
11319 11320 11321 11322
{
	struct i40e_netdev_priv *np = netdev_priv(netdev);
	struct i40e_vsi *vsi = np->vsi;
	struct i40e_pf *pf = vsi->back;
11323
	u16 port = ntohs(ti->port);
11324 11325 11326 11327 11328
	u8 idx;

	idx = i40e_get_udp_port_idx(pf, port);

	/* Check if port already exists */
11329 11330
	if (idx >= I40E_MAX_PF_UDP_OFFLOAD_PORTS)
		goto not_found;
11331

11332 11333 11334 11335 11336 11337 11338 11339 11340 11341 11342
	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;
11343
	}
11344 11345 11346 11347

	/* if port exists, set it to 0 (mark for deletion)
	 * and make it pending
	 */
11348
	pf->udp_ports[idx].port = 0;
11349 11350 11351 11352 11353 11354
	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",
11355
		    port);
11356 11357
}

11358
static int i40e_get_phys_port_id(struct net_device *netdev,
11359
				 struct netdev_phys_item_id *ppid)
11360 11361 11362 11363 11364
{
	struct i40e_netdev_priv *np = netdev_priv(netdev);
	struct i40e_pf *pf = np->vsi->back;
	struct i40e_hw *hw = &pf->hw;

11365
	if (!(pf->hw_features & I40E_HW_PORT_ID_VALID))
11366 11367 11368 11369 11370 11371 11372 11373
		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;
}

11374 11375 11376 11377 11378 11379 11380 11381
/**
 * 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
 */
11382 11383
static int i40e_ndo_fdb_add(struct ndmsg *ndm, struct nlattr *tb[],
			    struct net_device *dev,
11384
			    const unsigned char *addr, u16 vid,
11385 11386 11387 11388 11389 11390 11391 11392 11393
			    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;

11394 11395 11396 11397 11398
	if (vid) {
		pr_info("%s: vlans aren't supported yet for dev_uc|mc_add()\n", dev->name);
		return -EINVAL;
	}

11399 11400 11401 11402 11403 11404 11405 11406 11407 11408 11409 11410 11411 11412 11413 11414 11415 11416 11417 11418 11419 11420
	/* 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;
}

11421 11422 11423 11424 11425 11426 11427 11428 11429 11430 11431
/**
 * 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.
11432 11433
 *
 * Note: expects to be called while under rtnl_lock()
11434 11435
 **/
static int i40e_ndo_bridge_setlink(struct net_device *dev,
11436 11437
				   struct nlmsghdr *nlh,
				   u16 flags)
11438 11439 11440 11441 11442 11443 11444 11445 11446 11447 11448 11449 11450 11451 11452 11453 11454 11455 11456 11457 11458 11459 11460 11461 11462 11463 11464 11465 11466 11467 11468 11469 11470 11471 11472 11473 11474 11475 11476 11477 11478 11479 11480 11481 11482 11483
{
	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;
11484 11485 11486 11487 11488
			/* 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;
A
Amritha Nambiar 已提交
11489
			i40e_do_reset(pf, I40E_PF_RESET_FLAG, true);
11490 11491 11492 11493 11494 11495 11496 11497 11498 11499 11500 11501 11502 11503
			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
11504
 * @nlflags: netlink flags passed in
11505 11506 11507 11508 11509 11510
 *
 * 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,
11511 11512
				   u32 __always_unused filter_mask,
				   int nlflags)
11513 11514 11515 11516 11517 11518 11519 11520 11521 11522 11523 11524 11525 11526 11527 11528 11529 11530 11531 11532
{
	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;

11533
	return ndo_dflt_bridge_getlink(skb, pid, seq, dev, veb->bridge_mode,
11534
				       0, 0, nlflags, filter_mask, NULL);
11535 11536
}

11537 11538 11539
/**
 * i40e_features_check - Validate encapsulated packet conforms to limits
 * @skb: skb buff
J
Jean Sacren 已提交
11540
 * @dev: This physical port's netdev
11541 11542 11543 11544 11545 11546
 * @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)
{
11547 11548 11549 11550 11551 11552 11553 11554 11555 11556 11557 11558 11559 11560 11561 11562 11563 11564 11565 11566 11567 11568 11569 11570 11571 11572 11573 11574 11575 11576 11577 11578 11579 11580 11581 11582 11583 11584 11585 11586 11587 11588
	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
	 */
11589 11590

	return features;
11591 11592
out_err:
	return features & ~(NETIF_F_CSUM_MASK | NETIF_F_GSO_MASK);
11593 11594
}

11595 11596 11597 11598 11599 11600 11601 11602 11603 11604 11605 11606 11607 11608 11609 11610 11611 11612 11613 11614 11615 11616 11617 11618 11619 11620 11621 11622 11623 11624 11625 11626 11627 11628 11629 11630 11631 11632 11633 11634 11635 11636
/**
 * 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;
}

/**
11637
 * i40e_xdp - implements ndo_bpf for i40e
11638 11639 11640 11641
 * @dev: netdevice
 * @xdp: XDP command
 **/
static int i40e_xdp(struct net_device *dev,
11642
		    struct netdev_bpf *xdp)
11643 11644 11645 11646 11647 11648 11649 11650 11651 11652 11653 11654
{
	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);
11655
		xdp->prog_id = vsi->xdp_prog ? vsi->xdp_prog->aux->id : 0;
11656 11657 11658 11659 11660 11661
		return 0;
	default:
		return -EINVAL;
	}
}

11662
static const struct net_device_ops i40e_netdev_ops = {
J
Jesse Brandeburg 已提交
11663 11664 11665 11666 11667 11668 11669 11670
	.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
11671
	.ndo_do_ioctl		= i40e_ioctl,
J
Jesse Brandeburg 已提交
11672 11673 11674 11675 11676 11677
	.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
11678
	.ndo_setup_tc		= __i40e_setup_tc,
J
Jesse Brandeburg 已提交
11679 11680 11681
	.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,
11682
	.ndo_set_vf_rate	= i40e_ndo_set_vf_bw,
J
Jesse Brandeburg 已提交
11683
	.ndo_get_vf_config	= i40e_ndo_get_vf_config,
M
Mitch Williams 已提交
11684
	.ndo_set_vf_link_state	= i40e_ndo_set_vf_link_state,
11685
	.ndo_set_vf_spoofchk	= i40e_ndo_set_vf_spoofchk,
11686
	.ndo_set_vf_trust	= i40e_ndo_set_vf_trust,
11687 11688
	.ndo_udp_tunnel_add	= i40e_udp_tunnel_add,
	.ndo_udp_tunnel_del	= i40e_udp_tunnel_del,
11689
	.ndo_get_phys_port_id	= i40e_get_phys_port_id,
11690
	.ndo_fdb_add		= i40e_ndo_fdb_add,
11691
	.ndo_features_check	= i40e_features_check,
11692 11693
	.ndo_bridge_getlink	= i40e_ndo_bridge_getlink,
	.ndo_bridge_setlink	= i40e_ndo_bridge_setlink,
11694
	.ndo_bpf		= i40e_xdp,
J
Jesse Brandeburg 已提交
11695 11696 11697 11698 11699 11700 11701 11702 11703 11704 11705 11706 11707 11708
};

/**
 * 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;
11709
	u8 broadcast[ETH_ALEN];
J
Jesse Brandeburg 已提交
11710 11711
	u8 mac_addr[ETH_ALEN];
	int etherdev_size;
11712 11713
	netdev_features_t hw_enc_features;
	netdev_features_t hw_features;
J
Jesse Brandeburg 已提交
11714 11715

	etherdev_size = sizeof(struct i40e_netdev_priv);
11716
	netdev = alloc_etherdev_mq(etherdev_size, vsi->alloc_queue_pairs);
J
Jesse Brandeburg 已提交
11717 11718 11719 11720 11721 11722 11723
	if (!netdev)
		return -ENOMEM;

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

11724 11725 11726 11727 11728 11729 11730 11731 11732 11733 11734 11735 11736 11737 11738 11739 11740
	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
11741

11742
	if (!(pf->hw_features & I40E_HW_OUTER_UDP_CSUM_CAPABLE))
11743 11744 11745
		netdev->gso_partial_features |= NETIF_F_GSO_UDP_TUNNEL_CSUM;

	netdev->gso_partial_features |= NETIF_F_GSO_GRE_CSUM;
11746

11747 11748
	netdev->hw_enc_features |= hw_enc_features;

11749
	/* record features VLANs can make use of */
11750
	netdev->vlan_features |= hw_enc_features | NETIF_F_TSO_MANGLEID;
J
Jesse Brandeburg 已提交
11751

11752
	if (!(pf->flags & I40E_FLAG_MFP_ENABLED))
11753 11754
		netdev->hw_features |= NETIF_F_NTUPLE | NETIF_F_HW_TC;

11755 11756 11757
	hw_features = hw_enc_features		|
		      NETIF_F_HW_VLAN_CTAG_TX	|
		      NETIF_F_HW_VLAN_CTAG_RX;
11758

11759
	netdev->hw_features |= hw_features;
11760

11761
	netdev->features |= hw_features | NETIF_F_HW_VLAN_CTAG_FILTER;
11762
	netdev->hw_enc_features |= NETIF_F_TSO_MANGLEID;
J
Jesse Brandeburg 已提交
11763 11764 11765

	if (vsi->type == I40E_VSI_MAIN) {
		SET_NETDEV_DEV(netdev, &pf->pdev->dev);
11766
		ether_addr_copy(mac_addr, hw->mac.perm_addr);
11767 11768 11769 11770 11771 11772 11773 11774 11775
		/* 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.
11776 11777
		 */
		i40e_rm_default_mac_filter(vsi, mac_addr);
11778
		spin_lock_bh(&vsi->mac_filter_hash_lock);
11779
		i40e_add_mac_filter(vsi, mac_addr);
11780
		spin_unlock_bh(&vsi->mac_filter_hash_lock);
J
Jesse Brandeburg 已提交
11781
	} else {
11782 11783 11784 11785 11786 11787 11788
		/* 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
11789 11790
			 pf->vsi[pf->lan_vsi]->netdev->name);
		random_ether_addr(mac_addr);
K
Kiran Patil 已提交
11791

11792
		spin_lock_bh(&vsi->mac_filter_hash_lock);
11793
		i40e_add_mac_filter(vsi, mac_addr);
11794
		spin_unlock_bh(&vsi->mac_filter_hash_lock);
J
Jesse Brandeburg 已提交
11795
	}
K
Kiran Patil 已提交
11796

11797 11798 11799 11800 11801 11802 11803 11804 11805 11806 11807 11808 11809 11810 11811
	/* 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);
11812
	i40e_add_mac_filter(vsi, broadcast);
11813 11814
	spin_unlock_bh(&vsi->mac_filter_hash_lock);

11815 11816
	ether_addr_copy(netdev->dev_addr, mac_addr);
	ether_addr_copy(netdev->perm_addr, mac_addr);
11817

J
Jesse Brandeburg 已提交
11818 11819 11820 11821 11822 11823 11824 11825 11826
	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);

11827 11828
	/* MTU range: 68 - 9706 */
	netdev->min_mtu = ETH_MIN_MTU;
11829
	netdev->max_mtu = I40E_MAX_RXBUFFER - I40E_PACKET_HDR_PAD;
11830

J
Jesse Brandeburg 已提交
11831 11832 11833 11834 11835 11836 11837 11838 11839 11840 11841 11842 11843 11844 11845 11846 11847 11848
	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);
}

11849 11850 11851 11852 11853 11854 11855 11856 11857 11858 11859 11860 11861 11862 11863 11864
/**
 * 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];
11865 11866 11867 11868 11869 11870
	if (!veb) {
		dev_info(&pf->pdev->dev,
			 "There is no veb associated with the bridge\n");
		return -ENOENT;
	}

11871
	/* Uplink is a bridge in VEPA mode */
11872
	if (veb->bridge_mode & BRIDGE_MODE_VEPA) {
11873
		return 0;
11874 11875 11876 11877
	} else {
		/* Uplink is a bridge in VEB mode */
		return 1;
	}
11878

11879 11880
	/* VEPA is now default bridge, so return 0 */
	return 0;
11881 11882
}

J
Jesse Brandeburg 已提交
11883 11884 11885 11886 11887 11888 11889 11890 11891 11892 11893 11894 11895
/**
 * 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;
11896 11897 11898
	struct i40e_mac_filter *f;
	struct hlist_node *h;
	int bkt;
K
Kiran Patil 已提交
11899

J
Jesse Brandeburg 已提交
11900 11901 11902 11903 11904 11905 11906 11907 11908 11909 11910 11911 11912 11913 11914 11915 11916 11917
	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,
11918 11919 11920 11921
				 "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
11922 11923
			return -ENOENT;
		}
11924
		vsi->info = ctxt.info;
J
Jesse Brandeburg 已提交
11925 11926 11927 11928 11929 11930 11931
		vsi->info.valid_sections = 0;

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

		enabled_tc = i40e_pf_get_tc_map(pf);

11932 11933 11934 11935 11936 11937 11938 11939 11940 11941 11942 11943 11944 11945 11946 11947 11948 11949 11950 11951 11952 11953 11954 11955 11956
		/* 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
11957
		/* MFP mode setup queue map and update VSI */
11958 11959
		if ((pf->flags & I40E_FLAG_MFP_ENABLED) &&
		    !(pf->hw.func_caps.iscsi)) { /* NIC type PF */
J
Jesse Brandeburg 已提交
11960 11961 11962 11963 11964 11965 11966 11967
			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,
11968 11969 11970 11971
					 "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
11972 11973 11974 11975 11976 11977 11978 11979 11980 11981
				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.
11982 11983
			 * For MFP case the iSCSI PF would use this
			 * flow to enable LAN+iSCSI TC.
J
Jesse Brandeburg 已提交
11984 11985 11986
			 */
			ret = i40e_vsi_config_tc(vsi, enabled_tc);
			if (ret) {
11987 11988 11989
				/* Single TC condition is not fatal,
				 * message and continue
				 */
J
Jesse Brandeburg 已提交
11990
				dev_info(&pf->pdev->dev,
11991 11992 11993 11994 11995
					 "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
11996 11997 11998 11999 12000
			}
		}
		break;

	case I40E_VSI_FDIR:
12001 12002 12003
		ctxt.pf_num = hw->pf_id;
		ctxt.vf_num = 0;
		ctxt.uplink_seid = vsi->uplink_seid;
12004
		ctxt.connection_type = I40E_AQ_VSI_CONN_TYPE_NORMAL;
12005
		ctxt.flags = I40E_AQ_VSI_TYPE_PF;
12006 12007
		if ((pf->flags & I40E_FLAG_VEB_MODE_ENABLED) &&
		    (i40e_is_vsi_uplink_mode_veb(vsi))) {
12008
			ctxt.info.valid_sections |=
12009
			     cpu_to_le16(I40E_AQ_VSI_PROP_SWITCH_VALID);
12010
			ctxt.info.switch_id =
12011
			   cpu_to_le16(I40E_AQ_VSI_SW_ID_FLAG_ALLOW_LB);
12012
		}
J
Jesse Brandeburg 已提交
12013 12014 12015 12016 12017 12018 12019
		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;
12020
		ctxt.connection_type = I40E_AQ_VSI_CONN_TYPE_NORMAL;
J
Jesse Brandeburg 已提交
12021 12022 12023 12024 12025
		ctxt.flags = I40E_AQ_VSI_TYPE_VMDQ2;

		/* This VSI is connected to VEB so the switch_id
		 * should be set to zero by default.
		 */
12026 12027 12028 12029 12030 12031
		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
12032 12033 12034 12035 12036 12037 12038 12039 12040

		/* 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;
12041
		ctxt.connection_type = I40E_AQ_VSI_CONN_TYPE_NORMAL;
J
Jesse Brandeburg 已提交
12042 12043 12044 12045 12046
		ctxt.flags = I40E_AQ_VSI_TYPE_VF;

		/* This VSI is connected to VEB so the switch_id
		 * should be set to zero by default.
		 */
12047 12048 12049 12050 12051 12052
		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
12053

12054 12055 12056 12057
		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 |=
12058 12059
				(I40E_AQ_VSI_QUE_OPT_TCP_ENA |
				 I40E_AQ_VSI_QUE_OPT_RSS_LUT_VSI);
12060 12061
		}

J
Jesse Brandeburg 已提交
12062 12063
		ctxt.info.valid_sections |= cpu_to_le16(I40E_AQ_VSI_PROP_VLAN_VALID);
		ctxt.info.port_vlan_flags |= I40E_AQ_VSI_PVLAN_MODE_ALL;
12064 12065 12066 12067 12068 12069 12070
		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
12071 12072 12073 12074
		/* Setup the VSI tx/rx queue map for TC0 only for now */
		i40e_vsi_setup_queue_map(vsi, &ctxt, enabled_tc, true);
		break;

12075 12076 12077 12078
	case I40E_VSI_IWARP:
		/* send down message to iWARP */
		break;

J
Jesse Brandeburg 已提交
12079 12080 12081 12082 12083 12084 12085 12086
	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,
12087 12088 12089 12090
				 "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
12091 12092 12093
			ret = -ENOENT;
			goto err;
		}
12094
		vsi->info = ctxt.info;
J
Jesse Brandeburg 已提交
12095 12096 12097 12098 12099
		vsi->info.valid_sections = 0;
		vsi->seid = ctxt.seid;
		vsi->id = ctxt.vsi_number;
	}

12100
	vsi->active_filters = 0;
12101
	clear_bit(__I40E_VSI_OVERFLOW_PROMISC, vsi->state);
12102
	spin_lock_bh(&vsi->mac_filter_hash_lock);
J
Jesse Brandeburg 已提交
12103
	/* If macvlan filters already exist, force them to get loaded */
12104
	hash_for_each_safe(vsi->mac_filter_hash, bkt, h, f, hlist) {
12105
		f->state = I40E_FILTER_NEW;
J
Jesse Brandeburg 已提交
12106
		f_count++;
K
Kiran Patil 已提交
12107
	}
12108
	spin_unlock_bh(&vsi->mac_filter_hash_lock);
12109

J
Jesse Brandeburg 已提交
12110 12111 12112 12113 12114 12115 12116 12117 12118
	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,
12119 12120 12121
			 "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
12122 12123 12124 12125 12126 12127 12128 12129 12130 12131 12132 12133 12134 12135 12136 12137
		/* 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)
{
12138 12139
	struct i40e_mac_filter *f;
	struct hlist_node *h;
J
Jesse Brandeburg 已提交
12140 12141 12142
	struct i40e_veb *veb = NULL;
	struct i40e_pf *pf;
	u16 uplink_seid;
12143
	int i, n, bkt;
J
Jesse Brandeburg 已提交
12144 12145 12146 12147 12148 12149 12150 12151 12152 12153

	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] &&
12154
	    !test_bit(__I40E_DOWN, pf->state)) {
J
Jesse Brandeburg 已提交
12155 12156 12157 12158 12159 12160 12161 12162 12163 12164 12165 12166 12167
		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 {
12168
			i40e_vsi_close(vsi);
J
Jesse Brandeburg 已提交
12169 12170 12171 12172
		}
		i40e_vsi_disable_irq(vsi);
	}

12173
	spin_lock_bh(&vsi->mac_filter_hash_lock);
12174 12175 12176 12177 12178 12179 12180 12181

	/* 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 */
12182
	hash_for_each_safe(vsi->mac_filter_hash, bkt, h, f, hlist)
12183
		__i40e_del_filter(vsi, f);
12184

12185
	spin_unlock_bh(&vsi->mac_filter_hash_lock);
K
Kiran Patil 已提交
12186

J
Jesse Brandeburg 已提交
12187
	i40e_sync_vsi_filters(vsi);
J
Jesse Brandeburg 已提交
12188 12189 12190

	i40e_vsi_delete(vsi);
	i40e_vsi_free_q_vectors(vsi);
12191 12192 12193 12194
	if (vsi->netdev) {
		free_netdev(vsi->netdev);
		vsi->netdev = NULL;
	}
J
Jesse Brandeburg 已提交
12195 12196 12197 12198 12199 12200 12201 12202 12203 12204 12205
	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
12206
	for (n = 0, i = 0; i < pf->num_alloc_vsi; i++) {
J
Jesse Brandeburg 已提交
12207 12208 12209 12210 12211 12212 12213 12214 12215 12216 12217 12218 12219 12220 12221 12222 12223 12224 12225 12226 12227 12228 12229 12230 12231 12232 12233 12234 12235 12236 12237 12238 12239 12240 12241
		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;

12242
	if (vsi->q_vectors[0]) {
J
Jesse Brandeburg 已提交
12243 12244 12245 12246 12247 12248
		dev_info(&pf->pdev->dev, "VSI %d has existing q_vectors\n",
			 vsi->seid);
		return -EEXIST;
	}

	if (vsi->base_vector) {
12249
		dev_info(&pf->pdev->dev, "VSI %d has non-zero base vector %d\n",
J
Jesse Brandeburg 已提交
12250 12251 12252 12253
			 vsi->seid, vsi->base_vector);
		return -EEXIST;
	}

12254
	ret = i40e_vsi_alloc_q_vectors(vsi);
J
Jesse Brandeburg 已提交
12255 12256 12257 12258 12259 12260 12261 12262
	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;
	}

12263 12264 12265 12266 12267
	/* 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;
12268 12269 12270
	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
12271 12272
	if (vsi->base_vector < 0) {
		dev_info(&pf->pdev->dev,
12273 12274
			 "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
12275 12276 12277 12278 12279 12280 12281 12282 12283
		i40e_vsi_free_q_vectors(vsi);
		ret = -ENOENT;
		goto vector_setup_out;
	}

vector_setup_out:
	return ret;
}

12284 12285 12286 12287 12288 12289 12290 12291 12292 12293 12294
/**
 * 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)
{
12295
	u16 alloc_queue_pairs;
J
John Underwood 已提交
12296
	struct i40e_pf *pf;
12297 12298 12299
	u8 enabled_tc;
	int ret;

J
John Underwood 已提交
12300 12301 12302 12303 12304
	if (!vsi)
		return NULL;

	pf = vsi->back;

12305 12306 12307 12308 12309 12310 12311 12312 12313
	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;

12314 12315 12316 12317
	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);
12318
	if (ret < 0) {
12319
		dev_info(&pf->pdev->dev,
12320
			 "failed to get tracking for %d queues for VSI %d err %d\n",
12321
			 alloc_queue_pairs, vsi->seid, ret);
12322 12323 12324 12325 12326 12327 12328 12329 12330 12331 12332
		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);
12333 12334
	if (vsi->type == I40E_VSI_MAIN)
		i40e_rm_default_mac_filter(vsi, pf->hw.mac.perm_addr);
12335 12336 12337 12338 12339 12340 12341 12342 12343 12344 12345 12346 12347 12348 12349 12350 12351 12352 12353 12354 12355 12356 12357 12358

	/* 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
12359 12360 12361 12362 12363 12364 12365 12366 12367 12368 12369 12370 12371 12372 12373 12374 12375 12376
/**
 * 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;
12377
	u16 alloc_queue_pairs;
J
Jesse Brandeburg 已提交
12378 12379 12380 12381 12382 12383 12384 12385 12386 12387 12388 12389 12390 12391 12392 12393 12394 12395 12396 12397 12398 12399 12400 12401 12402
	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
12403
		for (i = 0; i < pf->num_alloc_vsi; i++) {
J
Jesse Brandeburg 已提交
12404 12405 12406 12407 12408 12409 12410 12411 12412 12413 12414 12415 12416 12417 12418 12419 12420
			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);
12421 12422 12423
		if (veb) {
			if (vsi->seid != pf->vsi[pf->lan_vsi]->seid) {
				dev_info(&vsi->back->pdev->dev,
S
Shannon Nelson 已提交
12424
					 "New VSI creation error, uplink seid of LAN VSI expected.\n");
12425 12426
				return NULL;
			}
12427 12428 12429 12430 12431 12432 12433 12434
			/* 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;
			}
12435
			i40e_config_bridge_mode(veb);
12436
		}
J
Jesse Brandeburg 已提交
12437 12438 12439 12440 12441 12442 12443 12444 12445 12446 12447 12448 12449 12450 12451 12452 12453 12454
		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];
12455 12456
	if (!vsi)
		goto err_alloc;
J
Jesse Brandeburg 已提交
12457 12458 12459 12460 12461 12462 12463 12464
	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 */
12465 12466 12467 12468
	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
12469
	if (ret < 0) {
12470 12471
		dev_info(&pf->pdev->dev,
			 "failed to get tracking for %d queues for VSI %d err=%d\n",
12472
			 alloc_queue_pairs, vsi->seid, ret);
J
Jesse Brandeburg 已提交
12473 12474 12475 12476 12477 12478 12479 12480 12481 12482 12483 12484 12485 12486 12487 12488 12489 12490 12491 12492 12493 12494
		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
12495 12496 12497 12498
#ifdef CONFIG_I40E_DCB
		/* Setup DCB netlink interface */
		i40e_dcbnl_setup(vsi);
#endif /* CONFIG_I40E_DCB */
J
Jesse Brandeburg 已提交
12499 12500 12501 12502 12503 12504 12505 12506 12507 12508 12509 12510 12511 12512 12513 12514 12515 12516 12517 12518 12519 12520 12521
		/* 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;
	}

12522
	if ((pf->hw_features & I40E_HW_RSS_AQ_CAPABLE) &&
12523 12524 12525
	    (vsi->type == I40E_VSI_VMDQ2)) {
		ret = i40e_vsi_config_rss(vsi);
	}
J
Jesse Brandeburg 已提交
12526 12527 12528 12529 12530 12531 12532 12533 12534 12535 12536 12537 12538 12539 12540 12541 12542 12543 12544 12545 12546 12547 12548 12549 12550 12551 12552 12553 12554 12555 12556 12557 12558 12559 12560 12561 12562 12563 12564
	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,
12565 12566 12567
			 "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
12568 12569 12570 12571 12572 12573 12574
		goto out;
	}

	ret = i40e_aq_query_switch_comp_ets_config(hw, veb->seid,
						   &ets_data, NULL);
	if (ret) {
		dev_info(&pf->pdev->dev,
12575 12576 12577
			 "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
12578 12579 12580 12581 12582 12583
		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;
12584
	veb->enabled_tc = ets_data.tc_valid_bits;
J
Jesse Brandeburg 已提交
12585 12586 12587 12588 12589 12590 12591 12592 12593 12594 12595 12596 12597 12598 12599 12600 12601 12602 12603 12604 12605 12606 12607 12608 12609 12610 12611 12612 12613 12614 12615 12616 12617 12618 12619 12620 12621 12622 12623 12624 12625 12626 12627 12628 12629 12630 12631 12632 12633 12634 12635 12636 12637 12638 12639 12640 12641 12642 12643 12644 12645 12646 12647 12648 12649 12650 12651 12652 12653 12654 12655 12656 12657 12658 12659 12660 12661 12662 12663 12664 12665 12666 12667 12668 12669 12670
	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
12671
	for (i = 0; i < pf->num_alloc_vsi; i++) {
J
Jesse Brandeburg 已提交
12672 12673 12674 12675 12676 12677 12678 12679 12680 12681 12682 12683 12684 12685 12686 12687 12688 12689 12690 12691 12692 12693 12694 12695 12696 12697 12698 12699 12700 12701 12702 12703 12704 12705 12706 12707 12708 12709 12710 12711 12712 12713 12714 12715 12716 12717 12718 12719 12720 12721 12722
		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
12723
	for (i = 0; i < pf->num_alloc_vsi; i++) {
J
Jesse Brandeburg 已提交
12724 12725 12726 12727 12728 12729 12730 12731 12732 12733 12734 12735 12736 12737 12738 12739 12740 12741 12742 12743 12744 12745 12746 12747 12748 12749 12750 12751 12752 12753 12754 12755 12756 12757 12758 12759 12760
		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)
{
12761
	struct i40e_pf *pf = veb->pf;
12762
	bool enable_stats = !!(pf->flags & I40E_FLAG_VEB_STATS_ENABLED);
J
Jesse Brandeburg 已提交
12763 12764
	int ret;

12765
	ret = i40e_aq_add_veb(&pf->hw, veb->uplink_seid, vsi->seid,
12766
			      veb->enabled_tc, false,
12767
			      &veb->seid, enable_stats, NULL);
12768 12769

	/* get a VEB from the hardware */
J
Jesse Brandeburg 已提交
12770
	if (ret) {
12771 12772 12773 12774
		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
12775 12776 12777 12778
		return -EPERM;
	}

	/* get statistics counter */
12779
	ret = i40e_aq_get_veb_parameters(&pf->hw, veb->seid, NULL, NULL,
J
Jesse Brandeburg 已提交
12780 12781
					 &veb->stats_idx, NULL, NULL, NULL);
	if (ret) {
12782 12783 12784 12785
		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
12786 12787 12788 12789
		return -EPERM;
	}
	ret = i40e_veb_get_bw_info(veb);
	if (ret) {
12790 12791 12792 12793 12794
		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
12795 12796 12797 12798 12799 12800 12801 12802 12803 12804 12805 12806 12807 12808 12809 12810 12811 12812 12813 12814 12815 12816 12817 12818 12819 12820 12821 12822 12823 12824 12825 12826 12827 12828 12829 12830 12831 12832 12833 12834 12835 12836 12837 12838
		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
12839
	for (vsi_idx = 0; vsi_idx < pf->num_alloc_vsi; vsi_idx++)
J
Jesse Brandeburg 已提交
12840 12841
		if (pf->vsi[vsi_idx] && pf->vsi[vsi_idx]->seid == vsi_seid)
			break;
M
Mitch Williams 已提交
12842
	if (vsi_idx >= pf->num_alloc_vsi && vsi_seid != 0) {
J
Jesse Brandeburg 已提交
12843 12844 12845 12846 12847 12848 12849 12850 12851 12852 12853 12854 12855 12856 12857 12858 12859 12860 12861 12862 12863 12864 12865 12866 12867 12868 12869 12870 12871 12872 12873 12874 12875 12876
		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
12877 12878
	if (vsi_idx == pf->lan_vsi)
		pf->lan_veb = veb->idx;
J
Jesse Brandeburg 已提交
12879 12880 12881 12882 12883 12884 12885 12886 12887 12888

	return veb;

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

/**
12889
 * i40e_setup_pf_switch_element - set PF vars based on switch type
J
Jesse Brandeburg 已提交
12890 12891 12892 12893 12894 12895 12896 12897 12898 12899 12900 12901 12902 12903 12904 12905 12906 12907 12908 12909 12910 12911 12912 12913 12914 12915 12916 12917 12918 12919 12920 12921 12922 12923 12924 12925 12926 12927 12928 12929 12930 12931 12932 12933 12934 12935 12936 12937 12938 12939 12940 12941 12942 12943 12944 12945 12946 12947 12948 12949 12950 12951 12952 12953 12954 12955 12956 12957 12958 12959 12960 12961 12962 12963 12964 12965 12966 12967 12968 12969 12970 12971 12972 12973 12974 12975 12976 12977 12978 12979 12980 12981 12982 12983 12984 12985 12986 12987 12988 12989 12990 12991 12992 12993 12994 12995 12996 12997 12998 12999
 * @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,
13000 13001 13002 13003
				 "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
13004 13005 13006 13007 13008 13009 13010 13011 13012 13013 13014 13015 13016 13017 13018 13019 13020 13021 13022 13023 13024 13025 13026 13027 13028 13029 13030 13031
			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
13032
 * @reinit: if the Main VSI needs to re-initialized.
J
Jesse Brandeburg 已提交
13033 13034 13035
 *
 * Returns 0 on success, negative value on failure
 **/
13036
static int i40e_setup_pf_switch(struct i40e_pf *pf, bool reinit)
J
Jesse Brandeburg 已提交
13037
{
13038
	u16 flags = 0;
J
Jesse Brandeburg 已提交
13039 13040 13041 13042 13043 13044
	int ret;

	/* find out what's out there already */
	ret = i40e_fetch_switch_configuration(pf, false);
	if (ret) {
		dev_info(&pf->pdev->dev,
13045 13046 13047
			 "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
13048 13049 13050 13051
		return ret;
	}
	i40e_pf_reset_stats(pf);

13052 13053 13054 13055 13056 13057 13058
	/* 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) &&
13059
	    !(pf->flags & I40E_FLAG_TRUE_PROMISC_SUPPORT)) {
13060
		flags = I40E_AQ_SET_SWITCH_CFG_PROMISC;
13061 13062
		pf->last_sw_conf_flags = flags;
	}
13063 13064 13065 13066 13067

	if (pf->hw.pf_id == 0) {
		u16 valid_flags;

		valid_flags = I40E_AQ_SET_SWITCH_CFG_PROMISC;
13068
		ret = i40e_aq_set_switch_config(&pf->hw, flags, valid_flags, 0,
13069 13070 13071 13072 13073 13074 13075 13076 13077
						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 */
		}
13078
		pf->last_sw_conf_valid_flags = valid_flags;
13079 13080
	}

J
Jesse Brandeburg 已提交
13081
	/* first time setup */
13082
	if (pf->lan_vsi == I40E_NO_VSI || reinit) {
J
Jesse Brandeburg 已提交
13083 13084 13085 13086 13087 13088 13089 13090 13091 13092
		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;
13093 13094 13095 13096
		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
13097 13098
		if (!vsi) {
			dev_info(&pf->pdev->dev, "setup of MAIN VSI failed\n");
A
Amritha Nambiar 已提交
13099
			i40e_cloud_filter_exit(pf);
J
Jesse Brandeburg 已提交
13100 13101 13102 13103 13104 13105
			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
13106

J
Jesse Brandeburg 已提交
13107 13108 13109 13110 13111 13112
		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]);

13113 13114
	i40e_fdir_sb_setup(pf);

J
Jesse Brandeburg 已提交
13115 13116 13117 13118 13119 13120 13121 13122 13123 13124 13125 13126
	/* 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))
13127
		i40e_pf_config_rss(pf);
J
Jesse Brandeburg 已提交
13128 13129

	/* fill in link information and enable LSE reporting */
13130 13131
	i40e_link_event(pf);

13132
	/* Initialize user-specific link properties */
J
Jesse Brandeburg 已提交
13133 13134
	pf->fc_autoneg_status = ((pf->hw.phy.link_info.an_info &
				  I40E_AQ_AN_COMPLETED) ? true : false);
13135

J
Jacob Keller 已提交
13136 13137
	i40e_ptp_init(pf);

13138 13139 13140
	/* repopulate tunnel port filters */
	i40e_sync_udp_filters(pf);

J
Jesse Brandeburg 已提交
13141 13142 13143 13144 13145 13146 13147 13148 13149 13150
	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;
13151
	int q_max;
J
Jesse Brandeburg 已提交
13152 13153 13154 13155 13156 13157 13158 13159 13160

	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;

13161
	if ((queues_left == 1) ||
13162
	    !(pf->flags & I40E_FLAG_MSIX_ENABLED)) {
J
Jesse Brandeburg 已提交
13163 13164
		/* one qp for PF, no queues for anything else */
		queues_left = 0;
13165
		pf->alloc_rss_size = pf->num_lan_qps = 1;
J
Jesse Brandeburg 已提交
13166 13167

		/* make sure all the fancies are disabled */
J
Jesse Brandeburg 已提交
13168
		pf->flags &= ~(I40E_FLAG_RSS_ENABLED	|
13169
			       I40E_FLAG_IWARP_ENABLED	|
J
Jesse Brandeburg 已提交
13170 13171
			       I40E_FLAG_FD_SB_ENABLED	|
			       I40E_FLAG_FD_ATR_ENABLED	|
13172
			       I40E_FLAG_DCB_CAPABLE	|
13173
			       I40E_FLAG_DCB_ENABLED	|
J
Jesse Brandeburg 已提交
13174 13175
			       I40E_FLAG_SRIOV_ENABLED	|
			       I40E_FLAG_VMDQ_ENABLED);
13176
		pf->flags |= I40E_FLAG_FD_SB_INACTIVE;
13177 13178
	} else if (!(pf->flags & (I40E_FLAG_RSS_ENABLED |
				  I40E_FLAG_FD_SB_ENABLED |
13179
				  I40E_FLAG_FD_ATR_ENABLED |
13180
				  I40E_FLAG_DCB_CAPABLE))) {
13181
		/* one qp for PF */
13182
		pf->alloc_rss_size = pf->num_lan_qps = 1;
13183 13184 13185
		queues_left -= pf->num_lan_qps;

		pf->flags &= ~(I40E_FLAG_RSS_ENABLED	|
13186
			       I40E_FLAG_IWARP_ENABLED	|
13187 13188 13189 13190
			       I40E_FLAG_FD_SB_ENABLED	|
			       I40E_FLAG_FD_ATR_ENABLED	|
			       I40E_FLAG_DCB_ENABLED	|
			       I40E_FLAG_VMDQ_ENABLED);
13191
		pf->flags |= I40E_FLAG_FD_SB_INACTIVE;
J
Jesse Brandeburg 已提交
13192
	} else {
13193
		/* Not enough queues for all TCs */
13194
		if ((pf->flags & I40E_FLAG_DCB_CAPABLE) &&
13195
		    (queues_left < I40E_MAX_TRAFFIC_CLASS)) {
13196 13197
			pf->flags &= ~(I40E_FLAG_DCB_CAPABLE |
					I40E_FLAG_DCB_ENABLED);
13198 13199
			dev_info(&pf->pdev->dev, "not enough queues for DCB. DCB is disabled.\n");
		}
13200 13201 13202 13203 13204 13205

		/* 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;
13206

13207 13208 13209 13210 13211 13212 13213 13214
		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;
13215
			pf->flags |= I40E_FLAG_FD_SB_INACTIVE;
13216 13217
			dev_info(&pf->pdev->dev, "not enough queues for Flow Director. Flow Director feature is disabled\n");
		}
J
Jesse Brandeburg 已提交
13218 13219 13220 13221
	}

	if ((pf->flags & I40E_FLAG_SRIOV_ENABLED) &&
	    pf->num_vf_qps && pf->num_req_vfs && queues_left) {
13222 13223
		pf->num_req_vfs = min_t(int, pf->num_req_vfs,
					(queues_left / pf->num_vf_qps));
J
Jesse Brandeburg 已提交
13224 13225 13226 13227 13228 13229 13230 13231 13232 13233
		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);
	}

13234
	pf->queues_left = queues_left;
13235 13236 13237 13238
	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),
13239 13240 13241
		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
13242 13243 13244 13245 13246 13247
}

/**
 * i40e_setup_pf_filter_control - Setup PF static filter control
 * @pf: PF to be setup
 *
13248
 * i40e_setup_pf_filter_control sets up a PF's initial filter control
J
Jesse Brandeburg 已提交
13249 13250 13251 13252 13253 13254 13255 13256 13257 13258 13259 13260 13261
 * 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
13262
	if (pf->flags & (I40E_FLAG_FD_SB_ENABLED | I40E_FLAG_FD_ATR_ENABLED))
J
Jesse Brandeburg 已提交
13263 13264 13265 13266 13267 13268 13269 13270 13271 13272 13273 13274
		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
13275
#define INFO_STRING_LEN 255
13276
#define REMAIN(__x) (INFO_STRING_LEN - (__x))
J
Jesse Brandeburg 已提交
13277 13278 13279
static void i40e_print_features(struct i40e_pf *pf)
{
	struct i40e_hw *hw = &pf->hw;
13280 13281
	char *buf;
	int i;
J
Jesse Brandeburg 已提交
13282

13283 13284
	buf = kmalloc(INFO_STRING_LEN, GFP_KERNEL);
	if (!buf)
J
Jesse Brandeburg 已提交
13285 13286
		return;

13287
	i = snprintf(buf, INFO_STRING_LEN, "Features: PF-id[%d]", hw->pf_id);
J
Jesse Brandeburg 已提交
13288
#ifdef CONFIG_PCI_IOV
13289
	i += snprintf(&buf[i], REMAIN(i), " VFs: %d", pf->num_req_vfs);
J
Jesse Brandeburg 已提交
13290
#endif
13291
	i += snprintf(&buf[i], REMAIN(i), " VSIs: %d QP: %d",
13292
		      pf->hw.func_caps.num_vsis,
13293
		      pf->vsi[pf->lan_vsi]->num_queue_pairs);
J
Jesse Brandeburg 已提交
13294
	if (pf->flags & I40E_FLAG_RSS_ENABLED)
13295
		i += snprintf(&buf[i], REMAIN(i), " RSS");
J
Jesse Brandeburg 已提交
13296
	if (pf->flags & I40E_FLAG_FD_ATR_ENABLED)
13297
		i += snprintf(&buf[i], REMAIN(i), " FD_ATR");
13298
	if (pf->flags & I40E_FLAG_FD_SB_ENABLED) {
13299 13300
		i += snprintf(&buf[i], REMAIN(i), " FD_SB");
		i += snprintf(&buf[i], REMAIN(i), " NTUPLE");
13301
	}
13302
	if (pf->flags & I40E_FLAG_DCB_CAPABLE)
13303 13304
		i += snprintf(&buf[i], REMAIN(i), " DCB");
	i += snprintf(&buf[i], REMAIN(i), " VxLAN");
13305
	i += snprintf(&buf[i], REMAIN(i), " Geneve");
J
Jesse Brandeburg 已提交
13306
	if (pf->flags & I40E_FLAG_PTP)
13307
		i += snprintf(&buf[i], REMAIN(i), " PTP");
13308
	if (pf->flags & I40E_FLAG_VEB_MODE_ENABLED)
13309
		i += snprintf(&buf[i], REMAIN(i), " VEB");
13310
	else
13311
		i += snprintf(&buf[i], REMAIN(i), " VEPA");
J
Jesse Brandeburg 已提交
13312

13313 13314
	dev_info(&pf->pdev->dev, "%s\n", buf);
	kfree(buf);
13315
	WARN_ON(i > INFO_STRING_LEN);
J
Jesse Brandeburg 已提交
13316 13317
}

13318 13319 13320 13321 13322
/**
 * i40e_get_platform_mac_addr - get platform-specific MAC address
 * @pdev: PCI device information struct
 * @pf: board private structure
 *
13323 13324 13325 13326
 * 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.
13327 13328 13329
 **/
static void i40e_get_platform_mac_addr(struct pci_dev *pdev, struct i40e_pf *pf)
{
13330 13331
	if (eth_platform_get_mac_address(&pdev->dev, pf->hw.mac.addr))
		i40e_get_mac_addr(&pf->hw, pf->hw.mac.addr);
13332 13333
}

J
Jesse Brandeburg 已提交
13334 13335 13336 13337 13338
/**
 * i40e_probe - Device initialization routine
 * @pdev: PCI device information struct
 * @ent: entry in i40e_pci_tbl
 *
13339 13340
 * 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
13341 13342 13343 13344 13345 13346
 * 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)
{
13347
	struct i40e_aq_get_phy_abilities_resp abilities;
J
Jesse Brandeburg 已提交
13348 13349
	struct i40e_pf *pf;
	struct i40e_hw *hw;
13350
	static u16 pfs_found;
13351
	u16 wol_nvm_bits;
13352
	u16 link_status;
13353
	int err;
13354
	u32 val;
13355
	u32 i;
13356
	u8 set_fc_aq_fail;
J
Jesse Brandeburg 已提交
13357 13358 13359 13360 13361 13362

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

	/* set up for high or low dma */
13363 13364
	err = dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(64));
	if (err) {
13365 13366 13367 13368 13369 13370
		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
13371 13372 13373
	}

	/* set up pci connections */
13374
	err = pci_request_mem_regions(pdev, i40e_driver_name);
J
Jesse Brandeburg 已提交
13375 13376 13377 13378 13379 13380 13381 13382 13383 13384 13385 13386 13387 13388 13389 13390 13391 13392 13393 13394 13395
	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;
13396
	set_bit(__I40E_DOWN, pf->state);
J
Jesse Brandeburg 已提交
13397 13398 13399

	hw = &pf->hw;
	hw->back = pf;
A
Anjali Singhai 已提交
13400

S
Shannon Nelson 已提交
13401 13402
	pf->ioremap_len = min_t(int, pci_resource_len(pdev, 0),
				I40E_MAX_CSR_SPACE);
A
Anjali Singhai 已提交
13403

S
Shannon Nelson 已提交
13404
	hw->hw_addr = ioremap(pci_resource_start(pdev, 0), pf->ioremap_len);
J
Jesse Brandeburg 已提交
13405 13406 13407 13408
	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
13409
			 pf->ioremap_len, err);
J
Jesse Brandeburg 已提交
13410 13411 13412 13413 13414 13415 13416 13417 13418
		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);
13419
	hw->bus.bus_id = pdev->bus->number;
13420
	pf->instance = pfs_found;
J
Jesse Brandeburg 已提交
13421

13422 13423 13424 13425 13426 13427 13428
	/* 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;

13429 13430 13431
	INIT_LIST_HEAD(&pf->l3_flex_pit_list);
	INIT_LIST_HEAD(&pf->l4_flex_pit_list);

13432 13433 13434 13435 13436 13437
	/* 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);

13438 13439 13440 13441 13442 13443
	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
13444

13445 13446 13447 13448 13449 13450 13451 13452 13453 13454 13455
	/* 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
13456
	/* Reset here to make sure all is clean and to define PF 'n' */
13457
	i40e_clear_hw(hw);
J
Jesse Brandeburg 已提交
13458 13459 13460 13461 13462 13463 13464 13465 13466 13467 13468 13469
	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;
13470

13471
	snprintf(pf->int_name, sizeof(pf->int_name) - 1,
13472 13473
		 "%s-%s:misc",
		 dev_driver_string(&pf->pdev->dev), dev_name(&pdev->dev));
J
Jesse Brandeburg 已提交
13474 13475 13476

	err = i40e_init_shared_code(hw);
	if (err) {
13477 13478
		dev_warn(&pdev->dev, "unidentified MAC or BLANK NVM: %d\n",
			 err);
J
Jesse Brandeburg 已提交
13479 13480 13481
		goto err_pf_reset;
	}

13482 13483 13484
	/* set up a default setting for link flow control */
	pf->hw.fc.requested_mode = I40E_FC_NONE;

J
Jesse Brandeburg 已提交
13485
	err = i40e_init_adminq(hw);
C
Carolyn Wyborny 已提交
13486 13487 13488 13489 13490 13491 13492 13493 13494 13495
	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;
	}
13496
	i40e_get_oem_version(hw);
13497

13498 13499 13500 13501 13502
	/* 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));
13503

13504
	if (hw->aq.api_maj_ver == I40E_FW_API_VERSION_MAJOR &&
13505
	    hw->aq.api_min_ver > I40E_FW_MINOR_VERSION(hw))
13506
		dev_info(&pdev->dev,
13507
			 "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");
13508
	else if (hw->aq.api_maj_ver == 1 && hw->aq.api_min_ver < 4)
13509
		dev_info(&pdev->dev,
13510
			 "The driver for the device detected an older version of the NVM image than expected. Please update the NVM image.\n");
13511

13512 13513
	i40e_verify_eeprom(pf);

13514 13515 13516 13517
	/* 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");

13518
	i40e_clear_pxe_mode(hw);
13519
	err = i40e_get_capabilities(pf, i40e_aqc_opc_list_func_capabilities);
J
Jesse Brandeburg 已提交
13520 13521 13522 13523 13524 13525 13526 13527 13528 13529
	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,
13530
				hw->func_caps.num_rx_qp, 0, 0);
J
Jesse Brandeburg 已提交
13531 13532 13533 13534 13535 13536 13537 13538 13539 13540 13541 13542
	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;
	}

13543 13544 13545 13546
	/* 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
	 */
13547
	if (pf->hw_features & I40E_HW_STOP_FW_LLDP) {
13548 13549 13550 13551
		dev_info(&pdev->dev, "Stopping firmware LLDP agent.\n");
		i40e_aq_stop_lldp(hw, true, NULL);
	}

13552 13553
	/* allow a platform config to override the HW addr */
	i40e_get_platform_mac_addr(pdev, pf);
13554

13555
	if (!is_valid_ether_addr(hw->mac.addr)) {
J
Jesse Brandeburg 已提交
13556 13557 13558 13559 13560
		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);
13561
	ether_addr_copy(hw->mac.perm_addr, hw->mac.addr);
13562 13563
	i40e_get_port_mac_addr(hw, hw->mac.port_addr);
	if (is_valid_ether_addr(hw->mac.port_addr))
13564
		pf->hw_features |= I40E_HW_PORT_ID_VALID;
J
Jesse Brandeburg 已提交
13565 13566 13567

	pci_set_drvdata(pdev, pf);
	pci_save_state(pdev);
13568 13569 13570 13571

	/* Enable FW to write default DCB config on link-up */
	i40e_aq_set_dcb_parameters(hw, true, NULL);

N
Neerav Parikh 已提交
13572 13573 13574
#ifdef CONFIG_I40E_DCB
	err = i40e_init_pf_dcb(pf);
	if (err) {
13575
		dev_info(&pdev->dev, "DCB init failed %d, disabled\n", err);
13576
		pf->flags &= ~(I40E_FLAG_DCB_CAPABLE | I40E_FLAG_DCB_ENABLED);
13577
		/* Continue without DCB enabled */
N
Neerav Parikh 已提交
13578 13579
	}
#endif /* CONFIG_I40E_DCB */
J
Jesse Brandeburg 已提交
13580 13581

	/* set up periodic task facility */
13582
	timer_setup(&pf->service_timer, i40e_service_timer, 0);
J
Jesse Brandeburg 已提交
13583 13584 13585
	pf->service_timer_period = HZ;

	INIT_WORK(&pf->service_task, i40e_service_task);
13586
	clear_bit(__I40E_SERVICE_SCHED, pf->state);
J
Jesse Brandeburg 已提交
13587

13588 13589
	/* 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
13590
	if (BIT (hw->port) & wol_nvm_bits || hw->partition_id != 1)
13591 13592 13593
		pf->wol_en = false;
	else
		pf->wol_en = true;
S
Shannon Nelson 已提交
13594 13595
	device_set_wakeup_enable(&pf->pdev->dev, pf->wol_en);

J
Jesse Brandeburg 已提交
13596 13597
	/* set up the main switch operations */
	i40e_determine_queue_usage(pf);
13598 13599 13600
	err = i40e_init_interrupt_scheme(pf);
	if (err)
		goto err_switch_setup;
J
Jesse Brandeburg 已提交
13601

M
Mitch Williams 已提交
13602 13603 13604 13605
	/* 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
13606
	 */
M
Mitch Williams 已提交
13607 13608 13609 13610 13611 13612
	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. */
13613 13614
	pf->vsi = kcalloc(pf->num_alloc_vsi, sizeof(struct i40e_vsi *),
			  GFP_KERNEL);
13615 13616
	if (!pf->vsi) {
		err = -ENOMEM;
J
Jesse Brandeburg 已提交
13617
		goto err_switch_setup;
13618
	}
J
Jesse Brandeburg 已提交
13619

13620 13621 13622 13623
#ifdef CONFIG_PCI_IOV
	/* prep for VF support */
	if ((pf->flags & I40E_FLAG_SRIOV_ENABLED) &&
	    (pf->flags & I40E_FLAG_MSIX_ENABLED) &&
13624
	    !test_bit(__I40E_BAD_EEPROM, pf->state)) {
13625 13626 13627 13628
		if (pci_num_vf(pdev))
			pf->flags |= I40E_FLAG_VEB_MODE_ENABLED;
	}
#endif
13629
	err = i40e_setup_pf_switch(pf, false);
J
Jesse Brandeburg 已提交
13630 13631 13632 13633
	if (err) {
		dev_info(&pdev->dev, "setup_pf_switch failed: %d\n", err);
		goto err_vsis;
	}
13634
	INIT_LIST_HEAD(&pf->vsi[pf->lan_vsi]->ch_list);
13635 13636 13637 13638 13639 13640 13641 13642 13643 13644 13645 13646 13647 13648 13649 13650 13651 13652 13653

	/* 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));

13654
	/* if FDIR VSI was set up, start it now */
M
Mitch Williams 已提交
13655
	for (i = 0; i < pf->num_alloc_vsi; i++) {
13656 13657 13658 13659 13660
		if (pf->vsi[i] && pf->vsi[i]->type == I40E_VSI_FDIR) {
			i40e_vsi_open(pf->vsi[i]);
			break;
		}
	}
J
Jesse Brandeburg 已提交
13661

S
Shannon Nelson 已提交
13662 13663
	/* The driver only wants link up/down and module qualification
	 * reports from firmware.  Note the negative logic.
J
Jesse Brandeburg 已提交
13664 13665
	 */
	err = i40e_aq_set_phy_int_mask(&pf->hw,
S
Shannon Nelson 已提交
13666
				       ~(I40E_AQ_EVENT_LINK_UPDOWN |
13667
					 I40E_AQ_EVENT_MEDIA_NA |
S
Shannon Nelson 已提交
13668
					 I40E_AQ_EVENT_MODULE_QUAL_FAIL), NULL);
J
Jesse Brandeburg 已提交
13669
	if (err)
13670 13671 13672
		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
13673

13674 13675 13676 13677 13678 13679 13680 13681 13682 13683 13684
	/* 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);
	}

13685
	if (pf->hw_features & I40E_HW_RESTART_AUTONEG) {
13686 13687 13688
		msleep(75);
		err = i40e_aq_set_link_restart_an(&pf->hw, true, NULL);
		if (err)
13689 13690 13691 13692
			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));
13693
	}
J
Jesse Brandeburg 已提交
13694 13695 13696 13697
	/* 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.
	 */
13698
	clear_bit(__I40E_DOWN, pf->state);
J
Jesse Brandeburg 已提交
13699 13700 13701 13702 13703 13704 13705 13706 13707 13708 13709 13710 13711 13712 13713

	/* 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;
		}
	}

13714
#ifdef CONFIG_PCI_IOV
J
Jesse Brandeburg 已提交
13715 13716
	/* prep for VF support */
	if ((pf->flags & I40E_FLAG_SRIOV_ENABLED) &&
13717
	    (pf->flags & I40E_FLAG_MSIX_ENABLED) &&
13718
	    !test_bit(__I40E_BAD_EEPROM, pf->state)) {
J
Jesse Brandeburg 已提交
13719 13720 13721 13722 13723
		/* 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
13724 13725 13726 13727 13728 13729 13730 13731 13732 13733

		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
13734
	}
13735
#endif /* CONFIG_PCI_IOV */
J
Jesse Brandeburg 已提交
13736

13737 13738 13739 13740 13741 13742 13743 13744 13745 13746 13747
	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;
		}
	}
13748

J
Jesse Brandeburg 已提交
13749 13750 13751
	i40e_dbg_pf_init(pf);

	/* tell the firmware that we're starting */
J
Jesse Brandeburg 已提交
13752
	i40e_send_version(pf);
J
Jesse Brandeburg 已提交
13753 13754 13755 13756 13757

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

13758
	/* add this PF to client device list and launch a client service task */
13759 13760 13761 13762 13763 13764
	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);
	}
13765

13766 13767 13768 13769 13770 13771
#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.
	 */
13772
	if (!(pf->hw_features & I40E_HW_NO_PCI_LINK_CHECK)) {
13773 13774 13775 13776 13777 13778 13779 13780 13781 13782 13783 13784 13785 13786 13787 13788 13789 13790 13791 13792 13793 13794 13795 13796 13797 13798 13799 13800 13801 13802 13803 13804 13805 13806 13807 13808 13809 13810 13811 13812 13813 13814
		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");
		}
13815 13816
	}

13817 13818 13819
	/* get the requested speeds from the fw */
	err = i40e_aq_get_phy_capabilities(hw, false, false, &abilities, NULL);
	if (err)
13820 13821 13822
		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));
13823 13824
	pf->hw.phy.link_info.requested_speeds = abilities.link_speed;

13825 13826 13827 13828 13829 13830 13831
	/* 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));

13832 13833 13834 13835 13836 13837 13838 13839 13840
	/* 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);

13841
	if ((pf->hw.device_id == I40E_DEV_ID_10G_BASE_T) ||
H
Henry Tieman 已提交
13842
		(pf->hw.device_id == I40E_DEV_ID_10G_BASE_T4))
13843
		pf->hw_features |= I40E_HW_PHY_CONTROLS_LEDS;
13844
	if (pf->hw.device_id == I40E_DEV_ID_SFP_I_X722)
13845
		pf->hw_features |= I40E_HW_HAVE_CRT_RETIMER;
J
Jesse Brandeburg 已提交
13846 13847 13848
	/* print a string summarizing features */
	i40e_print_features(pf);

J
Jesse Brandeburg 已提交
13849 13850 13851 13852
	return 0;

	/* Unwind what we've done if something failed in the setup */
err_vsis:
13853
	set_bit(__I40E_DOWN, pf->state);
J
Jesse Brandeburg 已提交
13854 13855
	i40e_clear_interrupt_scheme(pf);
	kfree(pf->vsi);
13856 13857
err_switch_setup:
	i40e_reset_interrupt_capability(pf);
J
Jesse Brandeburg 已提交
13858 13859 13860 13861 13862 13863 13864 13865 13866 13867 13868 13869 13870 13871
	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);
13872
	pci_release_mem_regions(pdev);
J
Jesse Brandeburg 已提交
13873 13874 13875 13876 13877 13878 13879 13880 13881 13882 13883 13884 13885 13886 13887 13888 13889 13890
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);
13891
	struct i40e_hw *hw = &pf->hw;
J
Jesse Brandeburg 已提交
13892 13893 13894 13895 13896
	i40e_status ret_code;
	int i;

	i40e_dbg_pf_exit(pf);

J
Jacob Keller 已提交
13897 13898
	i40e_ptp_stop(pf);

13899
	/* Disable RSS in hw */
13900 13901
	i40e_write_rx_ctl(hw, I40E_PFQF_HENA(0), 0);
	i40e_write_rx_ctl(hw, I40E_PFQF_HENA(1), 0);
13902

J
Jesse Brandeburg 已提交
13903
	/* no more scheduling of any task */
13904
	set_bit(__I40E_SUSPENDED, pf->state);
13905
	set_bit(__I40E_DOWN, pf->state);
13906
	if (pf->service_timer.function)
13907 13908 13909
		del_timer_sync(&pf->service_timer);
	if (pf->service_task.func)
		cancel_work_sync(&pf->service_task);
J
Jesse Brandeburg 已提交
13910

13911 13912 13913 13914 13915
	/* 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);

13916 13917 13918 13919 13920
	if (pf->flags & I40E_FLAG_SRIOV_ENABLED) {
		i40e_free_vfs(pf);
		pf->flags &= ~I40E_FLAG_SRIOV_ENABLED;
	}

J
Jesse Brandeburg 已提交
13921 13922 13923 13924 13925 13926 13927 13928 13929 13930 13931 13932 13933 13934 13935 13936 13937 13938 13939 13940
	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]);

A
Amritha Nambiar 已提交
13941 13942
	i40e_cloud_filter_exit(pf);

13943
	/* remove attached clients */
13944 13945 13946 13947 13948
	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);
13949 13950
	}

J
Jesse Brandeburg 已提交
13951
	/* shutdown and destroy the HMC */
13952 13953
	if (hw->hmc.hmc_obj) {
		ret_code = i40e_shutdown_lan_hmc(hw);
13954 13955 13956 13957 13958
		if (ret_code)
			dev_warn(&pdev->dev,
				 "Failed to destroy the HMC resources: %d\n",
				 ret_code);
	}
J
Jesse Brandeburg 已提交
13959 13960

	/* shutdown the adminq */
H
Henry Tieman 已提交
13961
	i40e_shutdown_adminq(hw);
J
Jesse Brandeburg 已提交
13962

13963 13964 13965 13966
	/* destroy the locks only once, here */
	mutex_destroy(&hw->aq.arq_mutex);
	mutex_destroy(&hw->aq.asq_mutex);

J
Jesse Brandeburg 已提交
13967 13968
	/* Clear all dynamic memory lists of rings, q_vectors, and VSIs */
	i40e_clear_interrupt_scheme(pf);
M
Mitch Williams 已提交
13969
	for (i = 0; i < pf->num_alloc_vsi; i++) {
J
Jesse Brandeburg 已提交
13970 13971 13972 13973 13974 13975 13976 13977 13978 13979 13980 13981 13982 13983 13984
		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);

13985
	iounmap(hw->hw_addr);
J
Jesse Brandeburg 已提交
13986
	kfree(pf);
13987
	pci_release_mem_regions(pdev);
J
Jesse Brandeburg 已提交
13988 13989 13990 13991 13992 13993 13994 13995 13996 13997 13998 13999 14000 14001 14002 14003 14004 14005 14006 14007

	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);

14008 14009 14010 14011 14012 14013
	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
14014
	/* shutdown all operations */
14015 14016
	if (!test_bit(__I40E_SUSPENDED, pf->state))
		i40e_prep_for_reset(pf, false);
J
Jesse Brandeburg 已提交
14017 14018 14019 14020 14021 14022 14023 14024 14025 14026 14027 14028 14029 14030 14031 14032 14033 14034 14035 14036 14037

	/* 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
14038
	dev_dbg(&pdev->dev, "%s\n", __func__);
J
Jesse Brandeburg 已提交
14039 14040 14041 14042 14043 14044 14045 14046 14047 14048 14049 14050 14051 14052 14053 14054 14055 14056 14057 14058 14059 14060 14061 14062 14063 14064 14065 14066
	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;
}

14067 14068 14069 14070 14071 14072 14073 14074 14075 14076 14077 14078 14079 14080 14081 14082 14083 14084 14085 14086 14087 14088
/**
 * 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
14089 14090 14091 14092 14093 14094 14095 14096 14097 14098 14099
/**
 * 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
14100
	dev_dbg(&pdev->dev, "%s\n", __func__);
14101
	if (test_bit(__I40E_SUSPENDED, pf->state))
14102 14103
		return;

14104
	i40e_handle_reset_warning(pf, false);
14105 14106
}

14107 14108 14109 14110 14111 14112 14113 14114 14115 14116 14117 14118 14119 14120 14121 14122 14123 14124 14125 14126 14127 14128 14129 14130 14131 14132 14133 14134 14135 14136 14137 14138 14139 14140 14141 14142 14143 14144 14145 14146 14147 14148 14149 14150 14151 14152 14153
/**
 * 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");
}

14154 14155 14156 14157 14158 14159 14160
/**
 * 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
14161
	struct i40e_hw *hw = &pf->hw;
14162

14163
	set_bit(__I40E_SUSPENDED, pf->state);
14164
	set_bit(__I40E_DOWN, pf->state);
14165
	rtnl_lock();
14166
	i40e_prep_for_reset(pf, true);
14167 14168
	rtnl_unlock();

S
Shannon Nelson 已提交
14169 14170 14171
	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));

14172 14173
	del_timer_sync(&pf->service_timer);
	cancel_work_sync(&pf->service_task);
A
Amritha Nambiar 已提交
14174
	i40e_cloud_filter_exit(pf);
14175 14176
	i40e_fdir_teardown(pf);

14177 14178 14179 14180 14181
	/* 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);

14182
	if (pf->wol_en && (pf->hw_features & I40E_HW_WOL_MC_MAGIC_PKT_WAKE))
14183 14184
		i40e_enable_mc_magic_wake(pf);

14185
	i40e_prep_for_reset(pf, false);
14186 14187 14188 14189 14190 14191

	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));

14192 14193
	i40e_clear_interrupt_scheme(pf);

14194
	if (system_state == SYSTEM_POWER_OFF) {
S
Shannon Nelson 已提交
14195
		pci_wake_from_d3(pdev, pf->wol_en);
14196 14197 14198 14199 14200
		pci_set_power_state(pdev, PCI_D3hot);
	}
}

/**
14201 14202
 * i40e_suspend - PM callback for moving to D3
 * @dev: generic device information structure
14203
 **/
14204
static int __maybe_unused i40e_suspend(struct device *dev)
14205
{
14206
	struct pci_dev *pdev = to_pci_dev(dev);
14207
	struct i40e_pf *pf = pci_get_drvdata(pdev);
S
Shannon Nelson 已提交
14208
	struct i40e_hw *hw = &pf->hw;
14209

14210 14211 14212 14213
	/* If we're already suspended, then there is nothing to do */
	if (test_and_set_bit(__I40E_SUSPENDED, pf->state))
		return 0;

14214
	set_bit(__I40E_DOWN, pf->state);
14215

14216 14217 14218 14219
	/* Ensure service task will not be running */
	del_timer_sync(&pf->service_timer);
	cancel_work_sync(&pf->service_task);

14220
	if (pf->wol_en && (pf->hw_features & I40E_HW_WOL_MC_MAGIC_PKT_WAKE))
14221 14222
		i40e_enable_mc_magic_wake(pf);

14223
	i40e_prep_for_reset(pf, false);
14224

S
Shannon Nelson 已提交
14225 14226 14227
	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));

14228 14229 14230 14231 14232 14233
	/* 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);
14234

14235
	return 0;
J
Jesse Brandeburg 已提交
14236 14237
}

14238
/**
14239 14240
 * i40e_resume - PM callback for waking up from D3
 * @dev: generic device information structure
14241
 **/
14242
static int __maybe_unused i40e_resume(struct device *dev)
14243
{
14244
	struct pci_dev *pdev = to_pci_dev(dev);
14245
	struct i40e_pf *pf = pci_get_drvdata(pdev);
14246
	int err;
14247

14248 14249 14250
	/* If we're not suspended, then there is nothing to do */
	if (!test_bit(__I40E_SUSPENDED, pf->state))
		return 0;
14251

14252 14253 14254 14255
	/* 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);
14256
	if (err) {
14257 14258
		dev_err(&pdev->dev, "Cannot restore interrupt scheme: %d\n",
			err);
14259 14260
	}

14261 14262 14263 14264 14265
	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);
14266

14267 14268 14269
	/* Restart the service task */
	mod_timer(&pf->service_timer,
		  round_jiffies(jiffies + pf->service_timer_period));
14270 14271 14272 14273

	return 0;
}

J
Jesse Brandeburg 已提交
14274 14275 14276
static const struct pci_error_handlers i40e_err_handler = {
	.error_detected = i40e_pci_error_detected,
	.slot_reset = i40e_pci_error_slot_reset,
14277 14278
	.reset_prepare = i40e_pci_error_reset_prepare,
	.reset_done = i40e_pci_error_reset_done,
J
Jesse Brandeburg 已提交
14279 14280 14281
	.resume = i40e_pci_error_resume,
};

14282 14283
static SIMPLE_DEV_PM_OPS(i40e_pm_ops, i40e_suspend, i40e_resume);

J
Jesse Brandeburg 已提交
14284 14285 14286 14287 14288
static struct pci_driver i40e_driver = {
	.name     = i40e_driver_name,
	.id_table = i40e_pci_tbl,
	.probe    = i40e_probe,
	.remove   = i40e_remove,
14289 14290 14291
	.driver   = {
		.pm = &i40e_pm_ops,
	},
14292
	.shutdown = i40e_shutdown,
J
Jesse Brandeburg 已提交
14293 14294 14295 14296 14297 14298 14299 14300 14301 14302 14303 14304 14305 14306 14307
	.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);
14308

14309 14310 14311 14312 14313 14314
	/* 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.
14315
	 */
14316
	i40e_wq = alloc_workqueue("%s", WQ_MEM_RECLAIM, 0, i40e_driver_name);
14317 14318 14319 14320 14321
	if (!i40e_wq) {
		pr_err("%s: Failed to create workqueue\n", i40e_driver_name);
		return -ENOMEM;
	}

J
Jesse Brandeburg 已提交
14322 14323 14324 14325 14326 14327 14328 14329 14330 14331 14332 14333 14334 14335
	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);
14336
	destroy_workqueue(i40e_wq);
J
Jesse Brandeburg 已提交
14337 14338 14339
	i40e_dbg_exit();
}
module_exit(i40e_exit_module);