ixgbe_main.c 144.4 KB
Newer Older
1 2 3
/*******************************************************************************

  Intel 10 Gigabit PCI Express Linux driver
4
  Copyright(c) 1999 - 2009 Intel Corporation.
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46

  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.

  You should have received a copy of the GNU General Public License along with
  this program; if not, write to the Free Software Foundation, Inc.,
  51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA.

  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

*******************************************************************************/

#include <linux/types.h>
#include <linux/module.h>
#include <linux/pci.h>
#include <linux/netdevice.h>
#include <linux/vmalloc.h>
#include <linux/string.h>
#include <linux/in.h>
#include <linux/ip.h>
#include <linux/tcp.h>
#include <linux/ipv6.h>
#include <net/checksum.h>
#include <net/ip6_checksum.h>
#include <linux/ethtool.h>
#include <linux/if_vlan.h>

#include "ixgbe.h"
#include "ixgbe_common.h"

char ixgbe_driver_name[] = "ixgbe";
S
Stephen Hemminger 已提交
47
static const char ixgbe_driver_string[] =
48
                              "Intel(R) 10 Gigabit PCI Express Network Driver";
49

50
#define DRV_VERSION "2.0.16-k2"
S
Stephen Hemminger 已提交
51
const char ixgbe_driver_version[] = DRV_VERSION;
52
static char ixgbe_copyright[] = "Copyright (c) 1999-2009 Intel Corporation.";
53 54

static const struct ixgbe_info *ixgbe_info_tbl[] = {
55
	[board_82598] = &ixgbe_82598_info,
56
	[board_82599] = &ixgbe_82599_info,
57 58 59 60 61 62 63 64 65 66 67
};

/* ixgbe_pci_tbl - PCI Device ID Table
 *
 * Wildcard entries (PCI_ANY_ID) should come last
 * Last entry must be all 0s
 *
 * { Vendor ID, Device ID, SubVendor ID, SubDevice ID,
 *   Class, Class Mask, private data (not used) }
 */
static struct pci_device_id ixgbe_pci_tbl[] = {
D
Don Skidmore 已提交
68 69
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_82598),
	 board_82598 },
70
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_82598AF_DUAL_PORT),
71
	 board_82598 },
72
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_82598AF_SINGLE_PORT),
73
	 board_82598 },
74 75
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_82598AT),
	 board_82598 },
76
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_82598EB_CX4),
77
	 board_82598 },
J
Jesse Brandeburg 已提交
78 79
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_82598_CX4_DUAL_PORT),
	 board_82598 },
D
Donald Skidmore 已提交
80 81 82 83
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_82598_DA_DUAL_PORT),
	 board_82598 },
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_82598_SR_DUAL_PORT_EM),
	 board_82598 },
84 85
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_82598EB_XF_LR),
	 board_82598 },
D
Donald Skidmore 已提交
86 87
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_82598EB_SFP_LOM),
	 board_82598 },
88 89
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_82598_BX),
	 board_82598 },
90 91 92 93
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_82599_KX4),
	 board_82599 },
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_82599_SFP),
	 board_82599 },
94 95 96 97 98 99

	/* required last entry */
	{0, }
};
MODULE_DEVICE_TABLE(pci, ixgbe_pci_tbl);

100
#ifdef CONFIG_IXGBE_DCA
101
static int ixgbe_notify_dca(struct notifier_block *, unsigned long event,
102
                            void *p);
103 104 105 106 107 108 109
static struct notifier_block dca_notifier = {
	.notifier_call = ixgbe_notify_dca,
	.next          = NULL,
	.priority      = 0
};
#endif

110 111 112 113 114 115 116
MODULE_AUTHOR("Intel Corporation, <linux.nics@intel.com>");
MODULE_DESCRIPTION("Intel(R) 10 Gigabit PCI Express Network Driver");
MODULE_LICENSE("GPL");
MODULE_VERSION(DRV_VERSION);

#define DEFAULT_DEBUG_LEVEL_SHIFT 3

117 118 119 120 121 122 123
static void ixgbe_release_hw_control(struct ixgbe_adapter *adapter)
{
	u32 ctrl_ext;

	/* Let firmware take over control of h/w */
	ctrl_ext = IXGBE_READ_REG(&adapter->hw, IXGBE_CTRL_EXT);
	IXGBE_WRITE_REG(&adapter->hw, IXGBE_CTRL_EXT,
124
	                ctrl_ext & ~IXGBE_CTRL_EXT_DRV_LOAD);
125 126 127 128 129 130 131 132 133
}

static void ixgbe_get_hw_control(struct ixgbe_adapter *adapter)
{
	u32 ctrl_ext;

	/* Let firmware know the driver has taken over */
	ctrl_ext = IXGBE_READ_REG(&adapter->hw, IXGBE_CTRL_EXT);
	IXGBE_WRITE_REG(&adapter->hw, IXGBE_CTRL_EXT,
134
	                ctrl_ext | IXGBE_CTRL_EXT_DRV_LOAD);
135
}
136

137 138 139 140 141 142 143 144 145 146
/*
 * ixgbe_set_ivar - set the IVAR registers, mapping interrupt causes to vectors
 * @adapter: pointer to adapter struct
 * @direction: 0 for Rx, 1 for Tx, -1 for other causes
 * @queue: queue to map the corresponding interrupt to
 * @msix_vector: the vector to map to the corresponding queue
 *
 */
static void ixgbe_set_ivar(struct ixgbe_adapter *adapter, s8 direction,
	                   u8 queue, u8 msix_vector)
147 148
{
	u32 ivar, index;
149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183
	struct ixgbe_hw *hw = &adapter->hw;
	switch (hw->mac.type) {
	case ixgbe_mac_82598EB:
		msix_vector |= IXGBE_IVAR_ALLOC_VAL;
		if (direction == -1)
			direction = 0;
		index = (((direction * 64) + queue) >> 2) & 0x1F;
		ivar = IXGBE_READ_REG(hw, IXGBE_IVAR(index));
		ivar &= ~(0xFF << (8 * (queue & 0x3)));
		ivar |= (msix_vector << (8 * (queue & 0x3)));
		IXGBE_WRITE_REG(hw, IXGBE_IVAR(index), ivar);
		break;
	case ixgbe_mac_82599EB:
		if (direction == -1) {
			/* other causes */
			msix_vector |= IXGBE_IVAR_ALLOC_VAL;
			index = ((queue & 1) * 8);
			ivar = IXGBE_READ_REG(&adapter->hw, IXGBE_IVAR_MISC);
			ivar &= ~(0xFF << index);
			ivar |= (msix_vector << index);
			IXGBE_WRITE_REG(&adapter->hw, IXGBE_IVAR_MISC, ivar);
			break;
		} else {
			/* tx or rx causes */
			msix_vector |= IXGBE_IVAR_ALLOC_VAL;
			index = ((16 * (queue & 1)) + (8 * direction));
			ivar = IXGBE_READ_REG(hw, IXGBE_IVAR(queue >> 1));
			ivar &= ~(0xFF << index);
			ivar |= (msix_vector << index);
			IXGBE_WRITE_REG(hw, IXGBE_IVAR(queue >> 1), ivar);
			break;
		}
	default:
		break;
	}
184 185 186
}

static void ixgbe_unmap_and_free_tx_resource(struct ixgbe_adapter *adapter,
187 188
                                             struct ixgbe_tx_buffer
                                             *tx_buffer_info)
189
{
190
	tx_buffer_info->dma = 0;
191
	if (tx_buffer_info->skb) {
192 193
		skb_dma_unmap(&adapter->pdev->dev, tx_buffer_info->skb,
		              DMA_TO_DEVICE);
194 195 196
		dev_kfree_skb_any(tx_buffer_info->skb);
		tx_buffer_info->skb = NULL;
	}
197
	tx_buffer_info->time_stamp = 0;
198 199 200 201
	/* tx_buffer_info must be completely set up in the transmit path */
}

static inline bool ixgbe_check_tx_hang(struct ixgbe_adapter *adapter,
202 203
                                       struct ixgbe_ring *tx_ring,
                                       unsigned int eop)
204
{
205 206
	struct ixgbe_hw *hw = &adapter->hw;

207
	/* Detect a transmit hang in hardware, this serializes the
208
	 * check with the clearing of time_stamp and movement of eop */
209
	adapter->detect_tx_hung = false;
210
	if (tx_ring->tx_buffer_info[eop].time_stamp &&
211 212 213
	    time_after(jiffies, tx_ring->tx_buffer_info[eop].time_stamp + HZ) &&
	    !(IXGBE_READ_REG(&adapter->hw, IXGBE_TFCS) & IXGBE_TFCS_TXOFF)) {
		/* detected Tx unit hang */
214 215
		union ixgbe_adv_tx_desc *tx_desc;
		tx_desc = IXGBE_TX_DESC_ADV(*tx_ring, eop);
216
		DPRINTK(DRV, ERR, "Detected Tx Unit Hang\n"
217 218
			"  Tx Queue             <%d>\n"
			"  TDH, TDT             <%x>, <%x>\n"
219 220 221 222
			"  next_to_use          <%x>\n"
			"  next_to_clean        <%x>\n"
			"tx_buffer_info[next_to_clean]\n"
			"  time_stamp           <%lx>\n"
223 224
			"  jiffies              <%lx>\n",
			tx_ring->queue_index,
225 226
			IXGBE_READ_REG(hw, tx_ring->head),
			IXGBE_READ_REG(hw, tx_ring->tail),
227 228
			tx_ring->next_to_use, eop,
			tx_ring->tx_buffer_info[eop].time_stamp, jiffies);
229 230 231 232 233 234
		return true;
	}

	return false;
}

235 236
#define IXGBE_MAX_TXD_PWR       14
#define IXGBE_MAX_DATA_PER_TXD  (1 << IXGBE_MAX_TXD_PWR)
237 238 239 240 241

/* Tx Descriptors needed, worst case */
#define TXD_USE_COUNT(S) (((S) >> IXGBE_MAX_TXD_PWR) + \
			 (((S) & (IXGBE_MAX_DATA_PER_TXD - 1)) ? 1 : 0))
#define DESC_NEEDED (TXD_USE_COUNT(IXGBE_MAX_DATA_PER_TXD) /* skb->data */ + \
242
	MAX_SKB_FRAGS * TXD_USE_COUNT(PAGE_SIZE) + 1) /* for context */
243

244 245
static void ixgbe_tx_timeout(struct net_device *netdev);

246 247 248
/**
 * ixgbe_clean_tx_irq - Reclaim resources after transmit completes
 * @adapter: board private structure
249
 * @tx_ring: tx ring to clean
250 251
 *
 * returns true if transmit work is done
252 253
 **/
static bool ixgbe_clean_tx_irq(struct ixgbe_adapter *adapter,
254
                               struct ixgbe_ring *tx_ring)
255
{
256
	struct net_device *netdev = adapter->netdev;
257 258 259
	union ixgbe_adv_tx_desc *tx_desc, *eop_desc;
	struct ixgbe_tx_buffer *tx_buffer_info;
	unsigned int i, eop, count = 0;
260
	unsigned int total_bytes = 0, total_packets = 0;
261 262

	i = tx_ring->next_to_clean;
263 264 265 266
	eop = tx_ring->tx_buffer_info[i].next_to_watch;
	eop_desc = IXGBE_TX_DESC_ADV(*tx_ring, eop);

	while ((eop_desc->wb.status & cpu_to_le32(IXGBE_TXD_STAT_DD)) &&
267
	       (count < tx_ring->work_limit)) {
268 269 270
		bool cleaned = false;
		for ( ; !cleaned; count++) {
			struct sk_buff *skb;
271 272
			tx_desc = IXGBE_TX_DESC_ADV(*tx_ring, i);
			tx_buffer_info = &tx_ring->tx_buffer_info[i];
273
			cleaned = (i == eop);
274
			skb = tx_buffer_info->skb;
275

276
			if (cleaned && skb) {
277
				unsigned int segs, bytecount;
278 279

				/* gso_segs is currently only valid for tcp */
280 281 282
				segs = skb_shinfo(skb)->gso_segs ?: 1;
				/* multiply data chunks by size of headers */
				bytecount = ((segs - 1) * skb_headlen(skb)) +
283 284 285
				            skb->len;
				total_packets += segs;
				total_bytes += bytecount;
286
			}
287

288
			ixgbe_unmap_and_free_tx_resource(adapter,
289
			                                 tx_buffer_info);
290

291 292
			tx_desc->wb.status = 0;

293 294 295
			i++;
			if (i == tx_ring->count)
				i = 0;
296
		}
297 298 299 300 301

		eop = tx_ring->tx_buffer_info[i].next_to_watch;
		eop_desc = IXGBE_TX_DESC_ADV(*tx_ring, eop);
	}

302 303
	tx_ring->next_to_clean = i;

304
#define TX_WAKE_THRESHOLD (DESC_NEEDED * 2)
305 306
	if (unlikely(count && netif_carrier_ok(netdev) &&
	             (IXGBE_DESC_UNUSED(tx_ring) >= TX_WAKE_THRESHOLD))) {
307 308 309 310
		/* Make sure that anybody stopping the queue after this
		 * sees the new next_to_clean.
		 */
		smp_mb();
311 312 313
		if (__netif_subqueue_stopped(netdev, tx_ring->queue_index) &&
		    !test_bit(__IXGBE_DOWN, &adapter->state)) {
			netif_wake_subqueue(netdev, tx_ring->queue_index);
314
			++adapter->restart_queue;
315
		}
316
	}
317

318 319 320 321 322 323 324 325 326
	if (adapter->detect_tx_hung) {
		if (ixgbe_check_tx_hang(adapter, tx_ring, i)) {
			/* schedule immediate reset if we believe we hung */
			DPRINTK(PROBE, INFO,
			        "tx hang %d detected, resetting adapter\n",
			        adapter->tx_timeout_count + 1);
			ixgbe_tx_timeout(adapter->netdev);
		}
	}
327

328
	/* re-arm the interrupt */
329
	if (count >= tx_ring->work_limit)
330
		IXGBE_WRITE_REG(&adapter->hw, IXGBE_EICS, tx_ring->v_idx);
331

332 333 334
	tx_ring->total_bytes += total_bytes;
	tx_ring->total_packets += total_packets;
	tx_ring->stats.packets += total_packets;
335
	tx_ring->stats.bytes += total_bytes;
336 337
	adapter->net_stats.tx_bytes += total_bytes;
	adapter->net_stats.tx_packets += total_packets;
338
	return (count < tx_ring->work_limit);
339 340
}

341
#ifdef CONFIG_IXGBE_DCA
342
static void ixgbe_update_rx_dca(struct ixgbe_adapter *adapter,
343
                                struct ixgbe_ring *rx_ring)
344 345 346
{
	u32 rxctrl;
	int cpu = get_cpu();
347
	int q = rx_ring - adapter->rx_ring;
348

349
	if (rx_ring->cpu != cpu) {
350
		rxctrl = IXGBE_READ_REG(&adapter->hw, IXGBE_DCA_RXCTRL(q));
351 352 353 354 355 356 357 358
		if (adapter->hw.mac.type == ixgbe_mac_82598EB) {
			rxctrl &= ~IXGBE_DCA_RXCTRL_CPUID_MASK;
			rxctrl |= dca3_get_tag(&adapter->pdev->dev, cpu);
		} else if (adapter->hw.mac.type == ixgbe_mac_82599EB) {
			rxctrl &= ~IXGBE_DCA_RXCTRL_CPUID_MASK_82599;
			rxctrl |= (dca3_get_tag(&adapter->pdev->dev, cpu) <<
			           IXGBE_DCA_RXCTRL_CPUID_SHIFT_82599);
		}
359 360
		rxctrl |= IXGBE_DCA_RXCTRL_DESC_DCA_EN;
		rxctrl |= IXGBE_DCA_RXCTRL_HEAD_DCA_EN;
361 362
		rxctrl &= ~(IXGBE_DCA_RXCTRL_DESC_RRO_EN);
		rxctrl &= ~(IXGBE_DCA_RXCTRL_DESC_WRO_EN |
363
		            IXGBE_DCA_RXCTRL_DESC_HSRO_EN);
364
		IXGBE_WRITE_REG(&adapter->hw, IXGBE_DCA_RXCTRL(q), rxctrl);
365
		rx_ring->cpu = cpu;
366 367 368 369 370
	}
	put_cpu();
}

static void ixgbe_update_tx_dca(struct ixgbe_adapter *adapter,
371
                                struct ixgbe_ring *tx_ring)
372 373 374
{
	u32 txctrl;
	int cpu = get_cpu();
375
	int q = tx_ring - adapter->tx_ring;
376

377
	if (tx_ring->cpu != cpu) {
378
		txctrl = IXGBE_READ_REG(&adapter->hw, IXGBE_DCA_TXCTRL(q));
379 380 381 382 383 384 385 386
		if (adapter->hw.mac.type == ixgbe_mac_82598EB) {
			txctrl &= ~IXGBE_DCA_TXCTRL_CPUID_MASK;
			txctrl |= dca3_get_tag(&adapter->pdev->dev, cpu);
		} else if (adapter->hw.mac.type == ixgbe_mac_82599EB) {
			txctrl &= ~IXGBE_DCA_TXCTRL_CPUID_MASK_82599;
			txctrl |= (dca3_get_tag(&adapter->pdev->dev, cpu) <<
			           IXGBE_DCA_TXCTRL_CPUID_SHIFT_82599);
		}
387 388
		txctrl |= IXGBE_DCA_TXCTRL_DESC_DCA_EN;
		IXGBE_WRITE_REG(&adapter->hw, IXGBE_DCA_TXCTRL(q), txctrl);
389
		tx_ring->cpu = cpu;
390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418
	}
	put_cpu();
}

static void ixgbe_setup_dca(struct ixgbe_adapter *adapter)
{
	int i;

	if (!(adapter->flags & IXGBE_FLAG_DCA_ENABLED))
		return;

	for (i = 0; i < adapter->num_tx_queues; i++) {
		adapter->tx_ring[i].cpu = -1;
		ixgbe_update_tx_dca(adapter, &adapter->tx_ring[i]);
	}
	for (i = 0; i < adapter->num_rx_queues; i++) {
		adapter->rx_ring[i].cpu = -1;
		ixgbe_update_rx_dca(adapter, &adapter->rx_ring[i]);
	}
}

static int __ixgbe_notify_dca(struct device *dev, void *data)
{
	struct net_device *netdev = dev_get_drvdata(dev);
	struct ixgbe_adapter *adapter = netdev_priv(netdev);
	unsigned long event = *(unsigned long *)data;

	switch (event) {
	case DCA_PROVIDER_ADD:
419 420 421
		/* if we're already enabled, don't do it again */
		if (adapter->flags & IXGBE_FLAG_DCA_ENABLED)
			break;
422 423 424
		/* Always use CB2 mode, difference is masked
		 * in the CB driver. */
		IXGBE_WRITE_REG(&adapter->hw, IXGBE_DCA_CTRL, 2);
425
		if (dca_add_requester(dev) == 0) {
426
			adapter->flags |= IXGBE_FLAG_DCA_ENABLED;
427 428 429 430 431 432 433 434 435 436 437 438 439
			ixgbe_setup_dca(adapter);
			break;
		}
		/* Fall Through since DCA is disabled. */
	case DCA_PROVIDER_REMOVE:
		if (adapter->flags & IXGBE_FLAG_DCA_ENABLED) {
			dca_remove_requester(dev);
			adapter->flags &= ~IXGBE_FLAG_DCA_ENABLED;
			IXGBE_WRITE_REG(&adapter->hw, IXGBE_DCA_CTRL, 1);
		}
		break;
	}

440
	return 0;
441 442
}

443
#endif /* CONFIG_IXGBE_DCA */
444 445 446 447
/**
 * ixgbe_receive_skb - Send a completed packet up the stack
 * @adapter: board private structure
 * @skb: packet to send up
448 449 450
 * @status: hardware indication of status of receive
 * @rx_ring: rx descriptor ring (for a specific queue) to setup
 * @rx_desc: rx descriptor
451
 **/
H
Herbert Xu 已提交
452
static void ixgbe_receive_skb(struct ixgbe_q_vector *q_vector,
453
                              struct sk_buff *skb, u8 status,
454
                              union ixgbe_adv_rx_desc *rx_desc)
455
{
H
Herbert Xu 已提交
456 457
	struct ixgbe_adapter *adapter = q_vector->adapter;
	struct napi_struct *napi = &q_vector->napi;
458 459
	bool is_vlan = (status & IXGBE_RXD_STAT_VP);
	u16 tag = le16_to_cpu(rx_desc->wb.upper.vlan);
460

461
	skb_record_rx_queue(skb, q_vector - &adapter->q_vector[0]);
H
Herbert Xu 已提交
462
	if (skb->ip_summed == CHECKSUM_UNNECESSARY) {
463
		if (adapter->vlgrp && is_vlan && (tag != 0))
H
Herbert Xu 已提交
464
			vlan_gro_receive(napi, adapter->vlgrp, tag, skb);
465
		else
H
Herbert Xu 已提交
466
			napi_gro_receive(napi, skb);
467 468
	} else {
		if (!(adapter->flags & IXGBE_FLAG_IN_NETPOLL)) {
469
			if (adapter->vlgrp && is_vlan && (tag != 0))
470 471 472 473
				vlan_hwaccel_receive_skb(skb, adapter->vlgrp, tag);
			else
				netif_receive_skb(skb);
		} else {
474
			if (adapter->vlgrp && is_vlan && (tag != 0))
475 476 477 478
				vlan_hwaccel_rx(skb, adapter->vlgrp, tag);
			else
				netif_rx(skb);
		}
479 480 481
	}
}

482 483 484 485 486 487
/**
 * ixgbe_rx_checksum - indicate in skb if hw indicated a good cksum
 * @adapter: address of board private structure
 * @status_err: hardware indication of status of receive
 * @skb: skb currently being received and modified
 **/
488
static inline void ixgbe_rx_checksum(struct ixgbe_adapter *adapter,
489
                                     u32 status_err, struct sk_buff *skb)
490 491 492
{
	skb->ip_summed = CHECKSUM_NONE;

493 494
	/* Rx csum disabled */
	if (!(adapter->flags & IXGBE_FLAG_RX_CSUM_ENABLED))
495
		return;
496 497 498 499

	/* if IP and error */
	if ((status_err & IXGBE_RXD_STAT_IPCS) &&
	    (status_err & IXGBE_RXDADV_ERR_IPE)) {
500 501 502
		adapter->hw_csum_rx_error++;
		return;
	}
503 504 505 506 507 508 509 510 511

	if (!(status_err & IXGBE_RXD_STAT_L4CS))
		return;

	if (status_err & IXGBE_RXDADV_ERR_TCPE) {
		adapter->hw_csum_rx_error++;
		return;
	}

512
	/* It must be a TCP or UDP packet with a valid checksum */
513
	skb->ip_summed = CHECKSUM_UNNECESSARY;
514 515 516
	adapter->hw_csum_rx_good++;
}

517 518 519 520 521 522 523 524 525 526 527 528 529
static inline void ixgbe_release_rx_desc(struct ixgbe_hw *hw,
                                         struct ixgbe_ring *rx_ring, u32 val)
{
	/*
	 * Force memory writes to complete before letting h/w
	 * know there are new descriptors to fetch.  (Only
	 * applicable for weak-ordered memory model archs,
	 * such as IA-64).
	 */
	wmb();
	IXGBE_WRITE_REG(hw, IXGBE_RDT(rx_ring->reg_idx), val);
}

530 531 532 533 534
/**
 * ixgbe_alloc_rx_buffers - Replace used receive buffers; packet split
 * @adapter: address of board private structure
 **/
static void ixgbe_alloc_rx_buffers(struct ixgbe_adapter *adapter,
535 536
                                   struct ixgbe_ring *rx_ring,
                                   int cleaned_count)
537 538 539
{
	struct pci_dev *pdev = adapter->pdev;
	union ixgbe_adv_rx_desc *rx_desc;
540
	struct ixgbe_rx_buffer *bi;
541
	unsigned int i;
542
	unsigned int bufsz = rx_ring->rx_buf_len + NET_IP_ALIGN;
543 544

	i = rx_ring->next_to_use;
545
	bi = &rx_ring->rx_buffer_info[i];
546 547 548 549

	while (cleaned_count--) {
		rx_desc = IXGBE_RX_DESC_ADV(*rx_ring, i);

550
		if (!bi->page_dma &&
551 552
		    (adapter->flags & IXGBE_FLAG_RX_PS_ENABLED)) {
			if (!bi->page) {
553 554 555 556 557 558 559 560 561
				bi->page = alloc_page(GFP_ATOMIC);
				if (!bi->page) {
					adapter->alloc_rx_page_failed++;
					goto no_buffers;
				}
				bi->page_offset = 0;
			} else {
				/* use a half page if we're re-using */
				bi->page_offset ^= (PAGE_SIZE / 2);
562
			}
563 564 565 566 567

			bi->page_dma = pci_map_page(pdev, bi->page,
			                            bi->page_offset,
			                            (PAGE_SIZE / 2),
			                            PCI_DMA_FROMDEVICE);
568 569
		}

570
		if (!bi->skb) {
571
			struct sk_buff *skb;
572
			skb = netdev_alloc_skb(adapter->netdev, bufsz);
573 574 575 576 577 578 579 580 581 582 583 584 585

			if (!skb) {
				adapter->alloc_rx_buff_failed++;
				goto no_buffers;
			}

			/*
			 * Make buffer alignment 2 beyond a 16 byte boundary
			 * this will result in a 16 byte aligned IP header after
			 * the 14 byte MAC header is removed
			 */
			skb_reserve(skb, NET_IP_ALIGN);

586
			bi->skb = skb;
587
			bi->dma = pci_map_single(pdev, skb->data, bufsz,
588
			                         PCI_DMA_FROMDEVICE);
589 590 591 592
		}
		/* Refresh the desc even if buffer_addrs didn't change because
		 * each write-back erases this info. */
		if (adapter->flags & IXGBE_FLAG_RX_PS_ENABLED) {
593 594
			rx_desc->read.pkt_addr = cpu_to_le64(bi->page_dma);
			rx_desc->read.hdr_addr = cpu_to_le64(bi->dma);
595
		} else {
596
			rx_desc->read.pkt_addr = cpu_to_le64(bi->dma);
597 598 599 600 601
		}

		i++;
		if (i == rx_ring->count)
			i = 0;
602
		bi = &rx_ring->rx_buffer_info[i];
603
	}
604

605 606 607 608 609 610
no_buffers:
	if (rx_ring->next_to_use != i) {
		rx_ring->next_to_use = i;
		if (i-- == 0)
			i = (rx_ring->count - 1);

611
		ixgbe_release_rx_desc(&adapter->hw, rx_ring, i);
612 613 614
	}
}

615 616 617 618 619 620 621 622 623 624
static inline u16 ixgbe_get_hdr_info(union ixgbe_adv_rx_desc *rx_desc)
{
	return rx_desc->wb.lower.lo_dword.hs_rss.hdr_info;
}

static inline u16 ixgbe_get_pkt_info(union ixgbe_adv_rx_desc *rx_desc)
{
	return rx_desc->wb.lower.lo_dword.hs_rss.pkt_info;
}

H
Herbert Xu 已提交
625
static bool ixgbe_clean_rx_irq(struct ixgbe_q_vector *q_vector,
626 627
                               struct ixgbe_ring *rx_ring,
                               int *work_done, int work_to_do)
628
{
H
Herbert Xu 已提交
629
	struct ixgbe_adapter *adapter = q_vector->adapter;
630 631 632 633 634
	struct pci_dev *pdev = adapter->pdev;
	union ixgbe_adv_rx_desc *rx_desc, *next_rxd;
	struct ixgbe_rx_buffer *rx_buffer_info, *next_buffer;
	struct sk_buff *skb;
	unsigned int i;
635
	u32 len, staterr;
636 637
	u16 hdr_info;
	bool cleaned = false;
638
	int cleaned_count = 0;
639
	unsigned int total_rx_bytes = 0, total_rx_packets = 0;
640 641 642 643 644 645 646

	i = rx_ring->next_to_clean;
	rx_desc = IXGBE_RX_DESC_ADV(*rx_ring, i);
	staterr = le32_to_cpu(rx_desc->wb.upper.status_error);
	rx_buffer_info = &rx_ring->rx_buffer_info[i];

	while (staterr & IXGBE_RXD_STAT_DD) {
647
		u32 upper_len = 0;
648 649 650 651 652
		if (*work_done >= work_to_do)
			break;
		(*work_done)++;

		if (adapter->flags & IXGBE_FLAG_RX_PS_ENABLED) {
653 654
			hdr_info = le16_to_cpu(ixgbe_get_hdr_info(rx_desc));
			len = (hdr_info & IXGBE_RXDADV_HDRBUFLEN_MASK) >>
655
			       IXGBE_RXDADV_HDRBUFLEN_SHIFT;
656 657 658 659 660
			if (hdr_info & IXGBE_RXDADV_SPH)
				adapter->rx_hdr_split++;
			if (len > IXGBE_RX_HDR_SIZE)
				len = IXGBE_RX_HDR_SIZE;
			upper_len = le16_to_cpu(rx_desc->wb.upper.length);
661
		} else {
662
			len = le16_to_cpu(rx_desc->wb.upper.length);
663
		}
664 665 666 667 668 669 670 671

		cleaned = true;
		skb = rx_buffer_info->skb;
		prefetch(skb->data - NET_IP_ALIGN);
		rx_buffer_info->skb = NULL;

		if (len && !skb_shinfo(skb)->nr_frags) {
			pci_unmap_single(pdev, rx_buffer_info->dma,
672
			                 rx_ring->rx_buf_len,
673
			                 PCI_DMA_FROMDEVICE);
674 675 676 677 678
			skb_put(skb, len);
		}

		if (upper_len) {
			pci_unmap_page(pdev, rx_buffer_info->page_dma,
679
			               PAGE_SIZE / 2, PCI_DMA_FROMDEVICE);
680 681
			rx_buffer_info->page_dma = 0;
			skb_fill_page_desc(skb, skb_shinfo(skb)->nr_frags,
682 683 684 685 686 687 688 689 690
			                   rx_buffer_info->page,
			                   rx_buffer_info->page_offset,
			                   upper_len);

			if ((rx_ring->rx_buf_len > (PAGE_SIZE / 2)) ||
			    (page_count(rx_buffer_info->page) != 1))
				rx_buffer_info->page = NULL;
			else
				get_page(rx_buffer_info->page);
691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712

			skb->len += upper_len;
			skb->data_len += upper_len;
			skb->truesize += upper_len;
		}

		i++;
		if (i == rx_ring->count)
			i = 0;
		next_buffer = &rx_ring->rx_buffer_info[i];

		next_rxd = IXGBE_RX_DESC_ADV(*rx_ring, i);
		prefetch(next_rxd);

		cleaned_count++;
		if (staterr & IXGBE_RXD_STAT_EOP) {
			rx_ring->stats.packets++;
			rx_ring->stats.bytes += skb->len;
		} else {
			rx_buffer_info->skb = next_buffer->skb;
			rx_buffer_info->dma = next_buffer->dma;
			next_buffer->skb = skb;
713
			next_buffer->dma = 0;
714 715 716 717 718 719 720 721 722 723
			adapter->non_eop_descs++;
			goto next_desc;
		}

		if (staterr & IXGBE_RXDADV_ERR_FRAME_ERR_MASK) {
			dev_kfree_skb_irq(skb);
			goto next_desc;
		}

		ixgbe_rx_checksum(adapter, staterr, skb);
724 725 726 727 728

		/* probably a little skewed due to removing CRC */
		total_rx_bytes += skb->len;
		total_rx_packets++;

729
		skb->protocol = eth_type_trans(skb, adapter->netdev);
H
Herbert Xu 已提交
730
		ixgbe_receive_skb(q_vector, skb, staterr, rx_desc);
731 732 733 734 735 736 737 738 739 740 741 742 743 744 745

next_desc:
		rx_desc->wb.upper.status_error = 0;

		/* return some buffers to hardware, one at a time is too slow */
		if (cleaned_count >= IXGBE_RX_BUFFER_WRITE) {
			ixgbe_alloc_rx_buffers(adapter, rx_ring, cleaned_count);
			cleaned_count = 0;
		}

		/* use prefetched values */
		rx_desc = next_rxd;
		rx_buffer_info = next_buffer;

		staterr = le32_to_cpu(rx_desc->wb.upper.status_error);
746 747
	}

748 749 750 751 752 753
	rx_ring->next_to_clean = i;
	cleaned_count = IXGBE_DESC_UNUSED(rx_ring);

	if (cleaned_count)
		ixgbe_alloc_rx_buffers(adapter, rx_ring, cleaned_count);

754 755 756 757 758
	rx_ring->total_packets += total_rx_packets;
	rx_ring->total_bytes += total_rx_bytes;
	adapter->net_stats.rx_bytes += total_rx_bytes;
	adapter->net_stats.rx_packets += total_rx_packets;

759 760 761
	return cleaned;
}

762
static int ixgbe_clean_rxonly(struct napi_struct *, int);
763 764 765 766 767 768 769 770 771
/**
 * ixgbe_configure_msix - Configure MSI-X hardware
 * @adapter: board private structure
 *
 * ixgbe_configure_msix sets up the hardware to properly generate MSI-X
 * interrupts.
 **/
static void ixgbe_configure_msix(struct ixgbe_adapter *adapter)
{
772 773 774
	struct ixgbe_q_vector *q_vector;
	int i, j, q_vectors, v_idx, r_idx;
	u32 mask;
775

776
	q_vectors = adapter->num_msix_vectors - NON_Q_VECTORS;
777

778 779
	/*
	 * Populate the IVAR table and set the ITR values to the
780 781 782 783 784 785
	 * corresponding register.
	 */
	for (v_idx = 0; v_idx < q_vectors; v_idx++) {
		q_vector = &adapter->q_vector[v_idx];
		/* XXX for_each_bit(...) */
		r_idx = find_first_bit(q_vector->rxr_idx,
786
		                       adapter->num_rx_queues);
787 788 789

		for (i = 0; i < q_vector->rxr_count; i++) {
			j = adapter->rx_ring[r_idx].reg_idx;
790
			ixgbe_set_ivar(adapter, 0, j, v_idx);
791
			r_idx = find_next_bit(q_vector->rxr_idx,
792 793
			                      adapter->num_rx_queues,
			                      r_idx + 1);
794 795
		}
		r_idx = find_first_bit(q_vector->txr_idx,
796
		                       adapter->num_tx_queues);
797 798 799

		for (i = 0; i < q_vector->txr_count; i++) {
			j = adapter->tx_ring[r_idx].reg_idx;
800
			ixgbe_set_ivar(adapter, 1, j, v_idx);
801
			r_idx = find_next_bit(q_vector->txr_idx,
802 803
			                      adapter->num_tx_queues,
			                      r_idx + 1);
804 805
		}

806
		/* if this is a tx only vector halve the interrupt rate */
807
		if (q_vector->txr_count && !q_vector->rxr_count)
808
			q_vector->eitr = (adapter->eitr_param >> 1);
809
		else if (q_vector->rxr_count)
810 811
			/* rx only */
			q_vector->eitr = adapter->eitr_param;
812

813
		/*
814
		 * since this is initial set up don't need to call
815 816
		 * ixgbe_write_eitr helper
		 */
817
		IXGBE_WRITE_REG(&adapter->hw, IXGBE_EITR(v_idx),
818
		                EITR_INTS_PER_SEC_TO_REG(q_vector->eitr));
819 820
	}

821 822 823 824 825
	if (adapter->hw.mac.type == ixgbe_mac_82598EB)
		ixgbe_set_ivar(adapter, -1, IXGBE_IVAR_OTHER_CAUSES_INDEX,
		               v_idx);
	else if (adapter->hw.mac.type == ixgbe_mac_82599EB)
		ixgbe_set_ivar(adapter, -1, 1, v_idx);
826 827
	IXGBE_WRITE_REG(&adapter->hw, IXGBE_EITR(v_idx), 1950);

828
	/* set up to autoclear timer, and the vectors */
829
	mask = IXGBE_EIMS_ENABLE_MASK;
830
	mask &= ~(IXGBE_EIMS_OTHER | IXGBE_EIMS_LSC);
831
	IXGBE_WRITE_REG(&adapter->hw, IXGBE_EIAC, mask);
832 833
}

834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859
enum latency_range {
	lowest_latency = 0,
	low_latency = 1,
	bulk_latency = 2,
	latency_invalid = 255
};

/**
 * ixgbe_update_itr - update the dynamic ITR value based on statistics
 * @adapter: pointer to adapter
 * @eitr: eitr setting (ints per sec) to give last timeslice
 * @itr_setting: current throttle rate in ints/second
 * @packets: the number of packets during this measurement interval
 * @bytes: the number of bytes during this measurement interval
 *
 *      Stores a new ITR value based on packets and byte
 *      counts during the last interrupt.  The advantage of per interrupt
 *      computation is faster updates and more accurate ITR for the current
 *      traffic pattern.  Constants in this function were computed
 *      based on theoretical maximum wire speed and thresholds were set based
 *      on testing data as well as attempting to minimize response time
 *      while increasing bulk throughput.
 *      this functionality is controlled by the InterruptThrottleRate module
 *      parameter (see ixgbe_param.c)
 **/
static u8 ixgbe_update_itr(struct ixgbe_adapter *adapter,
860 861
                           u32 eitr, u8 itr_setting,
                           int packets, int bytes)
862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900
{
	unsigned int retval = itr_setting;
	u32 timepassed_us;
	u64 bytes_perint;

	if (packets == 0)
		goto update_itr_done;


	/* simple throttlerate management
	 *    0-20MB/s lowest (100000 ints/s)
	 *   20-100MB/s low   (20000 ints/s)
	 *  100-1249MB/s bulk (8000 ints/s)
	 */
	/* what was last interrupt timeslice? */
	timepassed_us = 1000000/eitr;
	bytes_perint = bytes / timepassed_us; /* bytes/usec */

	switch (itr_setting) {
	case lowest_latency:
		if (bytes_perint > adapter->eitr_low)
			retval = low_latency;
		break;
	case low_latency:
		if (bytes_perint > adapter->eitr_high)
			retval = bulk_latency;
		else if (bytes_perint <= adapter->eitr_low)
			retval = lowest_latency;
		break;
	case bulk_latency:
		if (bytes_perint <= adapter->eitr_high)
			retval = low_latency;
		break;
	}

update_itr_done:
	return retval;
}

901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926
/**
 * ixgbe_write_eitr - write EITR register in hardware specific way
 * @adapter: pointer to adapter struct
 * @v_idx: vector index into q_vector array
 * @itr_reg: new value to be written in *register* format, not ints/s
 *
 * This function is made to be called by ethtool and by the driver
 * when it needs to update EITR registers at runtime.  Hardware
 * specific quirks/differences are taken care of here.
 */
void ixgbe_write_eitr(struct ixgbe_adapter *adapter, int v_idx, u32 itr_reg)
{
	struct ixgbe_hw *hw = &adapter->hw;
	if (adapter->hw.mac.type == ixgbe_mac_82598EB) {
		/* must write high and low 16 bits to reset counter */
		itr_reg |= (itr_reg << 16);
	} else if (adapter->hw.mac.type == ixgbe_mac_82599EB) {
		/*
		 * set the WDIS bit to not clear the timer bits and cause an
		 * immediate assertion of the interrupt
		 */
		itr_reg |= IXGBE_EITR_CNT_WDIS;
	}
	IXGBE_WRITE_REG(hw, IXGBE_EITR(v_idx), itr_reg);
}

927 928 929 930 931 932
static void ixgbe_set_itr_msix(struct ixgbe_q_vector *q_vector)
{
	struct ixgbe_adapter *adapter = q_vector->adapter;
	u32 new_itr;
	u8 current_itr, ret_itr;
	int i, r_idx, v_idx = ((void *)q_vector - (void *)(adapter->q_vector)) /
933
	                       sizeof(struct ixgbe_q_vector);
934 935 936 937 938 939
	struct ixgbe_ring *rx_ring, *tx_ring;

	r_idx = find_first_bit(q_vector->txr_idx, adapter->num_tx_queues);
	for (i = 0; i < q_vector->txr_count; i++) {
		tx_ring = &(adapter->tx_ring[r_idx]);
		ret_itr = ixgbe_update_itr(adapter, q_vector->eitr,
940 941 942
		                           q_vector->tx_itr,
		                           tx_ring->total_packets,
		                           tx_ring->total_bytes);
943 944
		/* if the result for this queue would decrease interrupt
		 * rate for this vector then use that result */
945
		q_vector->tx_itr = ((q_vector->tx_itr > ret_itr) ?
946
		                    q_vector->tx_itr - 1 : ret_itr);
947
		r_idx = find_next_bit(q_vector->txr_idx, adapter->num_tx_queues,
948
		                      r_idx + 1);
949 950 951 952 953 954
	}

	r_idx = find_first_bit(q_vector->rxr_idx, adapter->num_rx_queues);
	for (i = 0; i < q_vector->rxr_count; i++) {
		rx_ring = &(adapter->rx_ring[r_idx]);
		ret_itr = ixgbe_update_itr(adapter, q_vector->eitr,
955 956 957
		                           q_vector->rx_itr,
		                           rx_ring->total_packets,
		                           rx_ring->total_bytes);
958 959
		/* if the result for this queue would decrease interrupt
		 * rate for this vector then use that result */
960
		q_vector->rx_itr = ((q_vector->rx_itr > ret_itr) ?
961
		                    q_vector->rx_itr - 1 : ret_itr);
962
		r_idx = find_next_bit(q_vector->rxr_idx, adapter->num_rx_queues,
963
		                      r_idx + 1);
964 965
	}

966
	current_itr = max(q_vector->rx_itr, q_vector->tx_itr);
967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983

	switch (current_itr) {
	/* counts and packets in update_itr are dependent on these numbers */
	case lowest_latency:
		new_itr = 100000;
		break;
	case low_latency:
		new_itr = 20000; /* aka hwitr = ~200 */
		break;
	case bulk_latency:
	default:
		new_itr = 8000;
		break;
	}

	if (new_itr != q_vector->eitr) {
		u32 itr_reg;
984 985 986

		/* save the algorithm value here, not the smoothed one */
		q_vector->eitr = new_itr;
987 988 989
		/* do an exponential smoothing */
		new_itr = ((q_vector->eitr * 90)/100) + ((new_itr * 10)/100);
		itr_reg = EITR_INTS_PER_SEC_TO_REG(new_itr);
990
		ixgbe_write_eitr(adapter, v_idx, itr_reg);
991 992 993 994 995
	}

	return;
}

996 997 998 999 1000 1001 1002 1003 1004 1005 1006
static void ixgbe_check_fan_failure(struct ixgbe_adapter *adapter, u32 eicr)
{
	struct ixgbe_hw *hw = &adapter->hw;

	if ((adapter->flags & IXGBE_FLAG_FAN_FAIL_CAPABLE) &&
	    (eicr & IXGBE_EICR_GPI_SDP1)) {
		DPRINTK(PROBE, CRIT, "Fan has stopped, replace the adapter\n");
		/* write to clear the interrupt */
		IXGBE_WRITE_REG(hw, IXGBE_EICR, IXGBE_EICR_GPI_SDP1);
	}
}
1007

1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025
static void ixgbe_check_sfp_event(struct ixgbe_adapter *adapter, u32 eicr)
{
	struct ixgbe_hw *hw = &adapter->hw;

	if (eicr & IXGBE_EICR_GPI_SDP1) {
		/* Clear the interrupt */
		IXGBE_WRITE_REG(hw, IXGBE_EICR, IXGBE_EICR_GPI_SDP1);
		schedule_work(&adapter->multispeed_fiber_task);
	} else if (eicr & IXGBE_EICR_GPI_SDP2) {
		/* Clear the interrupt */
		IXGBE_WRITE_REG(hw, IXGBE_EICR, IXGBE_EICR_GPI_SDP2);
		schedule_work(&adapter->sfp_config_module_task);
	} else {
		/* Interrupt isn't for us... */
		return;
	}
}

1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038
static void ixgbe_check_lsc(struct ixgbe_adapter *adapter)
{
	struct ixgbe_hw *hw = &adapter->hw;

	adapter->lsc_int++;
	adapter->flags |= IXGBE_FLAG_NEED_LINK_UPDATE;
	adapter->link_check_timeout = jiffies;
	if (!test_bit(__IXGBE_DOWN, &adapter->state)) {
		IXGBE_WRITE_REG(hw, IXGBE_EIMC, IXGBE_EIMC_LSC);
		schedule_work(&adapter->watchdog_task);
	}
}

1039 1040 1041 1042 1043
static irqreturn_t ixgbe_msix_lsc(int irq, void *data)
{
	struct net_device *netdev = data;
	struct ixgbe_adapter *adapter = netdev_priv(netdev);
	struct ixgbe_hw *hw = &adapter->hw;
1044 1045 1046 1047 1048 1049 1050 1051 1052 1053
	u32 eicr;

	/*
	 * Workaround for Silicon errata.  Use clear-by-write instead
	 * of clear-by-read.  Reading with EICS will return the
	 * interrupt causes without clearing, which later be done
	 * with the write to EICR.
	 */
	eicr = IXGBE_READ_REG(hw, IXGBE_EICS);
	IXGBE_WRITE_REG(hw, IXGBE_EICR, eicr);
1054

1055 1056
	if (eicr & IXGBE_EICR_LSC)
		ixgbe_check_lsc(adapter);
1057

1058 1059
	if (hw->mac.type == ixgbe_mac_82598EB)
		ixgbe_check_fan_failure(adapter, eicr);
1060

1061 1062
	if (hw->mac.type == ixgbe_mac_82599EB)
		ixgbe_check_sfp_event(adapter, eicr);
1063 1064
	if (!test_bit(__IXGBE_DOWN, &adapter->state))
		IXGBE_WRITE_REG(hw, IXGBE_EIMS, IXGBE_EIMS_OTHER);
1065 1066 1067 1068 1069 1070

	return IRQ_HANDLED;
}

static irqreturn_t ixgbe_msix_clean_tx(int irq, void *data)
{
1071 1072
	struct ixgbe_q_vector *q_vector = data;
	struct ixgbe_adapter  *adapter = q_vector->adapter;
1073
	struct ixgbe_ring     *tx_ring;
1074 1075 1076 1077 1078 1079 1080
	int i, r_idx;

	if (!q_vector->txr_count)
		return IRQ_HANDLED;

	r_idx = find_first_bit(q_vector->txr_idx, adapter->num_tx_queues);
	for (i = 0; i < q_vector->txr_count; i++) {
1081
		tx_ring = &(adapter->tx_ring[r_idx]);
1082
#ifdef CONFIG_IXGBE_DCA
1083
		if (adapter->flags & IXGBE_FLAG_DCA_ENABLED)
1084
			ixgbe_update_tx_dca(adapter, tx_ring);
1085
#endif
1086 1087 1088
		tx_ring->total_bytes = 0;
		tx_ring->total_packets = 0;
		ixgbe_clean_tx_irq(adapter, tx_ring);
1089
		r_idx = find_next_bit(q_vector->txr_idx, adapter->num_tx_queues,
1090
		                      r_idx + 1);
1091
	}
1092 1093 1094 1095

	return IRQ_HANDLED;
}

1096 1097 1098 1099 1100
/**
 * ixgbe_msix_clean_rx - single unshared vector rx clean (all queues)
 * @irq: unused
 * @data: pointer to our q_vector struct for this interrupt vector
 **/
1101 1102
static irqreturn_t ixgbe_msix_clean_rx(int irq, void *data)
{
1103 1104
	struct ixgbe_q_vector *q_vector = data;
	struct ixgbe_adapter  *adapter = q_vector->adapter;
1105
	struct ixgbe_ring  *rx_ring;
1106
	int r_idx;
1107
	int i;
1108 1109

	r_idx = find_first_bit(q_vector->rxr_idx, adapter->num_rx_queues);
1110 1111 1112 1113 1114 1115 1116 1117
	for (i = 0;  i < q_vector->rxr_count; i++) {
		rx_ring = &(adapter->rx_ring[r_idx]);
		rx_ring->total_bytes = 0;
		rx_ring->total_packets = 0;
		r_idx = find_next_bit(q_vector->rxr_idx, adapter->num_rx_queues,
		                      r_idx + 1);
	}

1118 1119 1120
	if (!q_vector->rxr_count)
		return IRQ_HANDLED;

1121
	r_idx = find_first_bit(q_vector->rxr_idx, adapter->num_rx_queues);
1122
	rx_ring = &(adapter->rx_ring[r_idx]);
1123
	/* disable interrupts on this vector only */
1124
	IXGBE_WRITE_REG(&adapter->hw, IXGBE_EIMC, rx_ring->v_idx);
1125
	napi_schedule(&q_vector->napi);
1126 1127 1128 1129 1130 1131 1132 1133

	return IRQ_HANDLED;
}

static irqreturn_t ixgbe_msix_clean_many(int irq, void *data)
{
	ixgbe_msix_clean_rx(irq, data);
	ixgbe_msix_clean_tx(irq, data);
1134 1135 1136 1137

	return IRQ_HANDLED;
}

1138 1139 1140 1141 1142
/**
 * ixgbe_clean_rxonly - msix (aka one shot) rx clean routine
 * @napi: napi struct with our devices info in it
 * @budget: amount of work driver is allowed to do this pass, in packets
 *
1143 1144
 * This function is optimized for cleaning one queue only on a single
 * q_vector!!!
1145
 **/
1146 1147
static int ixgbe_clean_rxonly(struct napi_struct *napi, int budget)
{
1148
	struct ixgbe_q_vector *q_vector =
1149
	                       container_of(napi, struct ixgbe_q_vector, napi);
1150
	struct ixgbe_adapter *adapter = q_vector->adapter;
1151
	struct ixgbe_ring *rx_ring = NULL;
1152
	int work_done = 0;
1153
	long r_idx;
1154

1155
	r_idx = find_first_bit(q_vector->rxr_idx, adapter->num_rx_queues);
1156
	rx_ring = &(adapter->rx_ring[r_idx]);
1157
#ifdef CONFIG_IXGBE_DCA
1158
	if (adapter->flags & IXGBE_FLAG_DCA_ENABLED)
1159
		ixgbe_update_rx_dca(adapter, rx_ring);
1160
#endif
1161

H
Herbert Xu 已提交
1162
	ixgbe_clean_rx_irq(q_vector, rx_ring, &work_done, budget);
1163

1164 1165
	/* If all Rx work done, exit the polling mode */
	if (work_done < budget) {
1166
		napi_complete(napi);
1167
		if (adapter->itr_setting & 1)
1168
			ixgbe_set_itr_msix(q_vector);
1169
		if (!test_bit(__IXGBE_DOWN, &adapter->state))
1170
			IXGBE_WRITE_REG(&adapter->hw, IXGBE_EIMS, rx_ring->v_idx);
1171 1172 1173 1174 1175
	}

	return work_done;
}

1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200
/**
 * ixgbe_clean_rxonly_many - msix (aka one shot) rx clean routine
 * @napi: napi struct with our devices info in it
 * @budget: amount of work driver is allowed to do this pass, in packets
 *
 * This function will clean more than one rx queue associated with a
 * q_vector.
 **/
static int ixgbe_clean_rxonly_many(struct napi_struct *napi, int budget)
{
	struct ixgbe_q_vector *q_vector =
	                       container_of(napi, struct ixgbe_q_vector, napi);
	struct ixgbe_adapter *adapter = q_vector->adapter;
	struct ixgbe_ring *rx_ring = NULL;
	int work_done = 0, i;
	long r_idx;
	u16 enable_mask = 0;

	/* attempt to distribute budget to each queue fairly, but don't allow
	 * the budget to go below 1 because we'll exit polling */
	budget /= (q_vector->rxr_count ?: 1);
	budget = max(budget, 1);
	r_idx = find_first_bit(q_vector->rxr_idx, adapter->num_rx_queues);
	for (i = 0; i < q_vector->rxr_count; i++) {
		rx_ring = &(adapter->rx_ring[r_idx]);
1201
#ifdef CONFIG_IXGBE_DCA
1202 1203 1204
		if (adapter->flags & IXGBE_FLAG_DCA_ENABLED)
			ixgbe_update_rx_dca(adapter, rx_ring);
#endif
H
Herbert Xu 已提交
1205
		ixgbe_clean_rx_irq(q_vector, rx_ring, &work_done, budget);
1206 1207 1208 1209 1210 1211 1212 1213
		enable_mask |= rx_ring->v_idx;
		r_idx = find_next_bit(q_vector->rxr_idx, adapter->num_rx_queues,
		                      r_idx + 1);
	}

	r_idx = find_first_bit(q_vector->rxr_idx, adapter->num_rx_queues);
	rx_ring = &(adapter->rx_ring[r_idx]);
	/* If all Rx work done, exit the polling mode */
1214
	if (work_done < budget) {
1215
		napi_complete(napi);
1216
		if (adapter->itr_setting & 1)
1217 1218 1219 1220 1221 1222 1223 1224
			ixgbe_set_itr_msix(q_vector);
		if (!test_bit(__IXGBE_DOWN, &adapter->state))
			IXGBE_WRITE_REG(&adapter->hw, IXGBE_EIMS, enable_mask);
		return 0;
	}

	return work_done;
}
1225
static inline void map_vector_to_rxq(struct ixgbe_adapter *a, int v_idx,
1226
                                     int r_idx)
1227 1228 1229 1230 1231 1232 1233 1234
{
	a->q_vector[v_idx].adapter = a;
	set_bit(r_idx, a->q_vector[v_idx].rxr_idx);
	a->q_vector[v_idx].rxr_count++;
	a->rx_ring[r_idx].v_idx = 1 << v_idx;
}

static inline void map_vector_to_txq(struct ixgbe_adapter *a, int v_idx,
1235
                                     int r_idx)
1236 1237 1238 1239 1240 1241 1242
{
	a->q_vector[v_idx].adapter = a;
	set_bit(r_idx, a->q_vector[v_idx].txr_idx);
	a->q_vector[v_idx].txr_count++;
	a->tx_ring[r_idx].v_idx = 1 << v_idx;
}

1243
/**
1244 1245 1246
 * ixgbe_map_rings_to_vectors - Maps descriptor rings to vectors
 * @adapter: board private structure to initialize
 * @vectors: allotted vector count for descriptor rings
1247
 *
1248 1249 1250 1251 1252
 * 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 ring/queue, but on a constrained vector budget, we
 * group the rings as "efficiently" as possible.  You would add new
 * mapping configurations in here.
1253
 **/
1254
static int ixgbe_map_rings_to_vectors(struct ixgbe_adapter *adapter,
1255
                                      int vectors)
1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267
{
	int v_start = 0;
	int rxr_idx = 0, txr_idx = 0;
	int rxr_remaining = adapter->num_rx_queues;
	int txr_remaining = adapter->num_tx_queues;
	int i, j;
	int rqpv, tqpv;
	int err = 0;

	/* No mapping required if MSI-X is disabled. */
	if (!(adapter->flags & IXGBE_FLAG_MSIX_ENABLED))
		goto out;
1268

1269 1270 1271 1272 1273 1274 1275
	/*
	 * The ideal configuration...
	 * We have enough vectors to map one per queue.
	 */
	if (vectors == adapter->num_rx_queues + adapter->num_tx_queues) {
		for (; rxr_idx < rxr_remaining; v_start++, rxr_idx++)
			map_vector_to_rxq(adapter, v_start, rxr_idx);
1276

1277 1278
		for (; txr_idx < txr_remaining; v_start++, txr_idx++)
			map_vector_to_txq(adapter, v_start, txr_idx);
1279 1280

		goto out;
1281
	}
1282

1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302
	/*
	 * 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.
	 */
	/* Re-adjusting *qpv takes care of the remainder. */
	for (i = v_start; i < vectors; i++) {
		rqpv = DIV_ROUND_UP(rxr_remaining, vectors - i);
		for (j = 0; j < rqpv; j++) {
			map_vector_to_rxq(adapter, i, rxr_idx);
			rxr_idx++;
			rxr_remaining--;
		}
	}
	for (i = v_start; i < vectors; i++) {
		tqpv = DIV_ROUND_UP(txr_remaining, vectors - i);
		for (j = 0; j < tqpv; j++) {
			map_vector_to_txq(adapter, i, txr_idx);
			txr_idx++;
			txr_remaining--;
1303 1304 1305
		}
	}

1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321
out:
	return err;
}

/**
 * ixgbe_request_msix_irqs - Initialize MSI-X interrupts
 * @adapter: board private structure
 *
 * ixgbe_request_msix_irqs allocates MSI-X vectors and requests
 * interrupts from the kernel.
 **/
static int ixgbe_request_msix_irqs(struct ixgbe_adapter *adapter)
{
	struct net_device *netdev = adapter->netdev;
	irqreturn_t (*handler)(int, void *);
	int i, vector, q_vectors, err;
R
Robert Olsson 已提交
1322
	int ri=0, ti=0;
1323 1324 1325 1326 1327 1328 1329 1330 1331 1332

	/* Decrement for Other and TCP Timer vectors */
	q_vectors = adapter->num_msix_vectors - NON_Q_VECTORS;

	/* Map the Tx/Rx rings to the vectors we were allotted. */
	err = ixgbe_map_rings_to_vectors(adapter, q_vectors);
	if (err)
		goto out;

#define SET_HANDLER(_v) ((!(_v)->rxr_count) ? &ixgbe_msix_clean_tx : \
1333 1334
                         (!(_v)->txr_count) ? &ixgbe_msix_clean_rx : \
                         &ixgbe_msix_clean_many)
1335 1336
	for (vector = 0; vector < q_vectors; vector++) {
		handler = SET_HANDLER(&adapter->q_vector[vector]);
R
Robert Olsson 已提交
1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349

		if(handler == &ixgbe_msix_clean_rx) {
			sprintf(adapter->name[vector], "%s-%s-%d",
				netdev->name, "rx", ri++);
		}
		else if(handler == &ixgbe_msix_clean_tx) {
			sprintf(adapter->name[vector], "%s-%s-%d",
				netdev->name, "tx", ti++);
		}
		else
			sprintf(adapter->name[vector], "%s-%s-%d",
				netdev->name, "TxRx", vector);

1350
		err = request_irq(adapter->msix_entries[vector].vector,
1351 1352
		                  handler, 0, adapter->name[vector],
		                  &(adapter->q_vector[vector]));
1353 1354
		if (err) {
			DPRINTK(PROBE, ERR,
1355 1356
			        "request_irq failed for MSIX interrupt "
			        "Error: %d\n", err);
1357
			goto free_queue_irqs;
1358 1359 1360
		}
	}

1361 1362
	sprintf(adapter->name[vector], "%s:lsc", netdev->name);
	err = request_irq(adapter->msix_entries[vector].vector,
1363
	                  &ixgbe_msix_lsc, 0, adapter->name[vector], netdev);
1364 1365 1366
	if (err) {
		DPRINTK(PROBE, ERR,
			"request_irq for msix_lsc failed: %d\n", err);
1367
		goto free_queue_irqs;
1368 1369 1370 1371
	}

	return 0;

1372 1373 1374
free_queue_irqs:
	for (i = vector - 1; i >= 0; i--)
		free_irq(adapter->msix_entries[--vector].vector,
1375
		         &(adapter->q_vector[i]));
1376 1377
	adapter->flags &= ~IXGBE_FLAG_MSIX_ENABLED;
	pci_disable_msix(adapter->pdev);
1378 1379
	kfree(adapter->msix_entries);
	adapter->msix_entries = NULL;
1380
out:
1381 1382 1383
	return err;
}

1384 1385 1386 1387 1388 1389 1390 1391
static void ixgbe_set_itr(struct ixgbe_adapter *adapter)
{
	struct ixgbe_q_vector *q_vector = adapter->q_vector;
	u8 current_itr;
	u32 new_itr = q_vector->eitr;
	struct ixgbe_ring *rx_ring = &adapter->rx_ring[0];
	struct ixgbe_ring *tx_ring = &adapter->tx_ring[0];

1392
	q_vector->tx_itr = ixgbe_update_itr(adapter, new_itr,
1393 1394 1395
	                                    q_vector->tx_itr,
	                                    tx_ring->total_packets,
	                                    tx_ring->total_bytes);
1396
	q_vector->rx_itr = ixgbe_update_itr(adapter, new_itr,
1397 1398 1399
	                                    q_vector->rx_itr,
	                                    rx_ring->total_packets,
	                                    rx_ring->total_bytes);
1400

1401
	current_itr = max(q_vector->rx_itr, q_vector->tx_itr);
1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419

	switch (current_itr) {
	/* counts and packets in update_itr are dependent on these numbers */
	case lowest_latency:
		new_itr = 100000;
		break;
	case low_latency:
		new_itr = 20000; /* aka hwitr = ~200 */
		break;
	case bulk_latency:
		new_itr = 8000;
		break;
	default:
		break;
	}

	if (new_itr != q_vector->eitr) {
		u32 itr_reg;
1420 1421 1422

		/* save the algorithm value here, not the smoothed one */
		q_vector->eitr = new_itr;
1423 1424 1425
		/* do an exponential smoothing */
		new_itr = ((q_vector->eitr * 90)/100) + ((new_itr * 10)/100);
		itr_reg = EITR_INTS_PER_SEC_TO_REG(new_itr);
1426
		ixgbe_write_eitr(adapter, 0, itr_reg);
1427 1428 1429 1430 1431
	}

	return;
}

1432 1433 1434 1435 1436 1437 1438 1439
/**
 * ixgbe_irq_enable - Enable default interrupt generation settings
 * @adapter: board private structure
 **/
static inline void ixgbe_irq_enable(struct ixgbe_adapter *adapter)
{
	u32 mask;
	mask = IXGBE_EIMS_ENABLE_MASK;
1440 1441
	if (adapter->flags & IXGBE_FLAG_FAN_FAIL_CAPABLE)
		mask |= IXGBE_EIMS_GPI_SDP1;
1442
	if (adapter->hw.mac.type == ixgbe_mac_82599EB) {
1443
		mask |= IXGBE_EIMS_ECC;
1444 1445 1446 1447
		mask |= IXGBE_EIMS_GPI_SDP1;
		mask |= IXGBE_EIMS_GPI_SDP2;
	}

1448
	IXGBE_WRITE_REG(&adapter->hw, IXGBE_EIMS, mask);
1449 1450 1451 1452 1453 1454 1455 1456
	if (adapter->hw.mac.type == ixgbe_mac_82599EB) {
		/* enable the rest of the queue vectors */
		IXGBE_WRITE_REG(&adapter->hw, IXGBE_EIMS_EX(1),
		                (IXGBE_EIMS_RTX_QUEUE << 16));
		IXGBE_WRITE_REG(&adapter->hw, IXGBE_EIMS_EX(2),
		                ((IXGBE_EIMS_RTX_QUEUE << 16) |
		                  IXGBE_EIMS_RTX_QUEUE));
	}
1457 1458
	IXGBE_WRITE_FLUSH(&adapter->hw);
}
1459

1460
/**
1461
 * ixgbe_intr - legacy mode Interrupt Handler
1462 1463 1464 1465 1466 1467 1468 1469 1470 1471
 * @irq: interrupt number
 * @data: pointer to a network interface device structure
 **/
static irqreturn_t ixgbe_intr(int irq, void *data)
{
	struct net_device *netdev = data;
	struct ixgbe_adapter *adapter = netdev_priv(netdev);
	struct ixgbe_hw *hw = &adapter->hw;
	u32 eicr;

1472 1473 1474 1475 1476 1477
	/*
	 * Workaround for silicon errata.  Mask the interrupts
	 * before the read of EICR.
	 */
	IXGBE_WRITE_REG(hw, IXGBE_EIMC, IXGBE_IRQ_CLEAR_MASK);

1478 1479 1480
	/* for NAPI, using EIAM to auto-mask tx/rx interrupt bits on read
	 * therefore no explict interrupt disable is necessary */
	eicr = IXGBE_READ_REG(hw, IXGBE_EICR);
1481 1482 1483 1484 1485
	if (!eicr) {
		/* shared interrupt alert!
		 * make sure interrupts are enabled because the read will
		 * have disabled interrupts due to EIAM */
		ixgbe_irq_enable(adapter);
1486
		return IRQ_NONE;	/* Not our interrupt */
1487
	}
1488

1489 1490
	if (eicr & IXGBE_EICR_LSC)
		ixgbe_check_lsc(adapter);
1491

1492 1493 1494
	if (hw->mac.type == ixgbe_mac_82599EB)
		ixgbe_check_sfp_event(adapter, eicr);

1495 1496
	ixgbe_check_fan_failure(adapter, eicr);

1497
	if (napi_schedule_prep(&adapter->q_vector[0].napi)) {
1498 1499 1500 1501
		adapter->tx_ring[0].total_packets = 0;
		adapter->tx_ring[0].total_bytes = 0;
		adapter->rx_ring[0].total_packets = 0;
		adapter->rx_ring[0].total_bytes = 0;
1502
		/* would disable interrupts here but EIAM disabled it */
1503
		__napi_schedule(&adapter->q_vector[0].napi);
1504 1505 1506 1507 1508
	}

	return IRQ_HANDLED;
}

1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521
static inline void ixgbe_reset_q_vectors(struct ixgbe_adapter *adapter)
{
	int i, q_vectors = adapter->num_msix_vectors - NON_Q_VECTORS;

	for (i = 0; i < q_vectors; i++) {
		struct ixgbe_q_vector *q_vector = &adapter->q_vector[i];
		bitmap_zero(q_vector->rxr_idx, MAX_RX_QUEUES);
		bitmap_zero(q_vector->txr_idx, MAX_TX_QUEUES);
		q_vector->rxr_count = 0;
		q_vector->txr_count = 0;
	}
}

1522 1523 1524 1525 1526 1527 1528
/**
 * ixgbe_request_irq - initialize interrupts
 * @adapter: board private structure
 *
 * Attempts to configure interrupts using the best available
 * capabilities of the hardware and kernel.
 **/
1529
static int ixgbe_request_irq(struct ixgbe_adapter *adapter)
1530 1531
{
	struct net_device *netdev = adapter->netdev;
1532
	int err;
1533

1534 1535 1536 1537
	if (adapter->flags & IXGBE_FLAG_MSIX_ENABLED) {
		err = ixgbe_request_msix_irqs(adapter);
	} else if (adapter->flags & IXGBE_FLAG_MSI_ENABLED) {
		err = request_irq(adapter->pdev->irq, &ixgbe_intr, 0,
1538
		                  netdev->name, netdev);
1539 1540
	} else {
		err = request_irq(adapter->pdev->irq, &ixgbe_intr, IRQF_SHARED,
1541
		                  netdev->name, netdev);
1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554
	}

	if (err)
		DPRINTK(PROBE, ERR, "request_irq failed, Error %d\n", err);

	return err;
}

static void ixgbe_free_irq(struct ixgbe_adapter *adapter)
{
	struct net_device *netdev = adapter->netdev;

	if (adapter->flags & IXGBE_FLAG_MSIX_ENABLED) {
1555
		int i, q_vectors;
1556

1557 1558 1559
		q_vectors = adapter->num_msix_vectors;

		i = q_vectors - 1;
1560 1561
		free_irq(adapter->msix_entries[i].vector, netdev);

1562 1563 1564
		i--;
		for (; i >= 0; i--) {
			free_irq(adapter->msix_entries[i].vector,
1565
			         &(adapter->q_vector[i]));
1566 1567 1568 1569 1570
		}

		ixgbe_reset_q_vectors(adapter);
	} else {
		free_irq(adapter->pdev->irq, netdev);
1571 1572 1573
	}
}

1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606
/**
 * ixgbe_irq_disable - Mask off interrupt generation on the NIC
 * @adapter: board private structure
 **/
static inline void ixgbe_irq_disable(struct ixgbe_adapter *adapter)
{
	IXGBE_WRITE_REG(&adapter->hw, IXGBE_EIMC, ~0);
	if (adapter->hw.mac.type == ixgbe_mac_82599EB) {
		IXGBE_WRITE_REG(&adapter->hw, IXGBE_EIMC_EX(1), ~0);
		IXGBE_WRITE_REG(&adapter->hw, IXGBE_EIMC_EX(2), ~0);
	}
	IXGBE_WRITE_FLUSH(&adapter->hw);
	if (adapter->flags & IXGBE_FLAG_MSIX_ENABLED) {
		int i;
		for (i = 0; i < adapter->num_msix_vectors; i++)
			synchronize_irq(adapter->msix_entries[i].vector);
	} else {
		synchronize_irq(adapter->pdev->irq);
	}
}

static inline void ixgbe_irq_enable_queues(struct ixgbe_adapter *adapter)
{
	u32 mask = IXGBE_EIMS_RTX_QUEUE;
	IXGBE_WRITE_REG(&adapter->hw, IXGBE_EIMS, mask);
	if (adapter->hw.mac.type == ixgbe_mac_82599EB) {
		IXGBE_WRITE_REG(&adapter->hw, IXGBE_EIMS_EX(1), mask << 16);
		IXGBE_WRITE_REG(&adapter->hw, IXGBE_EIMS_EX(2),
		                (mask << 16 | mask));
	}
	/* skip the flush */
}

1607 1608 1609 1610 1611 1612 1613 1614
/**
 * ixgbe_configure_msi_and_legacy - Initialize PIN (INTA...) and MSI interrupts
 *
 **/
static void ixgbe_configure_msi_and_legacy(struct ixgbe_adapter *adapter)
{
	struct ixgbe_hw *hw = &adapter->hw;

1615
	IXGBE_WRITE_REG(hw, IXGBE_EITR(0),
1616
	                EITR_INTS_PER_SEC_TO_REG(adapter->eitr_param));
1617

1618 1619
	ixgbe_set_ivar(adapter, 0, 0, 0);
	ixgbe_set_ivar(adapter, 1, 0, 0);
1620 1621 1622 1623 1624

	map_vector_to_rxq(adapter, 0, 0);
	map_vector_to_txq(adapter, 0, 0);

	DPRINTK(HW, INFO, "Legacy interrupt IVAR setup done\n");
1625 1626 1627
}

/**
1628
 * ixgbe_configure_tx - Configure 8259x Transmit Unit after Reset
1629 1630 1631 1632 1633 1634
 * @adapter: board private structure
 *
 * Configure the Tx unit of the MAC after a reset.
 **/
static void ixgbe_configure_tx(struct ixgbe_adapter *adapter)
{
1635
	u64 tdba;
1636
	struct ixgbe_hw *hw = &adapter->hw;
1637
	u32 i, j, tdlen, txctrl;
1638 1639 1640

	/* Setup the HW Tx Head and Tail descriptor pointers */
	for (i = 0; i < adapter->num_tx_queues; i++) {
1641 1642 1643 1644
		struct ixgbe_ring *ring = &adapter->tx_ring[i];
		j = ring->reg_idx;
		tdba = ring->dma;
		tdlen = ring->count * sizeof(union ixgbe_adv_tx_desc);
1645
		IXGBE_WRITE_REG(hw, IXGBE_TDBAL(j),
1646
		                (tdba & DMA_BIT_MASK(32)));
1647 1648 1649 1650 1651 1652 1653 1654 1655
		IXGBE_WRITE_REG(hw, IXGBE_TDBAH(j), (tdba >> 32));
		IXGBE_WRITE_REG(hw, IXGBE_TDLEN(j), tdlen);
		IXGBE_WRITE_REG(hw, IXGBE_TDH(j), 0);
		IXGBE_WRITE_REG(hw, IXGBE_TDT(j), 0);
		adapter->tx_ring[i].head = IXGBE_TDH(j);
		adapter->tx_ring[i].tail = IXGBE_TDT(j);
		/* Disable Tx Head Writeback RO bit, since this hoses
		 * bookkeeping if things aren't delivered in order.
		 */
1656
		txctrl = IXGBE_READ_REG(hw, IXGBE_DCA_TXCTRL(j));
1657
		txctrl &= ~IXGBE_DCA_TXCTRL_TX_WB_RO_EN;
1658
		IXGBE_WRITE_REG(hw, IXGBE_DCA_TXCTRL(j), txctrl);
1659
	}
1660 1661 1662 1663 1664 1665
	if (hw->mac.type == ixgbe_mac_82599EB) {
		/* We enable 8 traffic classes, DCB only */
		if (adapter->flags & IXGBE_FLAG_DCB_ENABLED)
			IXGBE_WRITE_REG(hw, IXGBE_MTQC, (IXGBE_MTQC_RT_ENA |
			                IXGBE_MTQC_8TC_8TQ));
	}
1666 1667
}

1668
#define IXGBE_SRRCTL_BSIZEHDRSIZE_SHIFT 2
1669 1670 1671 1672 1673

static void ixgbe_configure_srrctl(struct ixgbe_adapter *adapter, int index)
{
	struct ixgbe_ring *rx_ring;
	u32 srrctl;
1674
	int queue0 = 0;
1675 1676
	unsigned long mask;

1677 1678
	if (adapter->hw.mac.type == ixgbe_mac_82599EB) {
		queue0 = index;
1679
	} else {
1680 1681 1682
		mask = (unsigned long) adapter->ring_feature[RING_F_RSS].mask;
		queue0 = index & mask;
		index = index & mask;
1683
	}
1684

1685 1686 1687 1688 1689 1690 1691 1692
	rx_ring = &adapter->rx_ring[queue0];

	srrctl = IXGBE_READ_REG(&adapter->hw, IXGBE_SRRCTL(index));

	srrctl &= ~IXGBE_SRRCTL_BSIZEHDR_MASK;
	srrctl &= ~IXGBE_SRRCTL_BSIZEPKT_MASK;

	if (adapter->flags & IXGBE_FLAG_RX_PS_ENABLED) {
1693 1694 1695 1696 1697 1698 1699 1700
		u16 bufsz = IXGBE_RXBUFFER_2048;
		/* grow the amount we can receive on large page machines */
		if (bufsz < (PAGE_SIZE / 2))
			bufsz = (PAGE_SIZE / 2);
		/* cap the bufsz at our largest descriptor size */
		bufsz = min((u16)IXGBE_MAX_RXBUFFER, bufsz);

		srrctl |= bufsz >> IXGBE_SRRCTL_BSIZEPKT_SHIFT;
1701 1702
		srrctl |= IXGBE_SRRCTL_DESCTYPE_HDR_SPLIT_ALWAYS;
		srrctl |= ((IXGBE_RX_HDR_SIZE <<
1703 1704
		            IXGBE_SRRCTL_BSIZEHDRSIZE_SHIFT) &
		           IXGBE_SRRCTL_BSIZEHDR_MASK);
1705 1706 1707 1708 1709 1710 1711 1712 1713 1714
	} else {
		srrctl |= IXGBE_SRRCTL_DESCTYPE_ADV_ONEBUF;

		if (rx_ring->rx_buf_len == MAXIMUM_ETHERNET_VLAN_SIZE)
			srrctl |= IXGBE_RXBUFFER_2048 >>
			          IXGBE_SRRCTL_BSIZEPKT_SHIFT;
		else
			srrctl |= rx_ring->rx_buf_len >>
			          IXGBE_SRRCTL_BSIZEPKT_SHIFT;
	}
1715

1716 1717
	IXGBE_WRITE_REG(&adapter->hw, IXGBE_SRRCTL(index), srrctl);
}
1718 1719

/**
1720
 * ixgbe_configure_rx - Configure 8259x Receive Unit after Reset
1721 1722 1723 1724 1725 1726 1727 1728 1729 1730
 * @adapter: board private structure
 *
 * Configure the Rx unit of the MAC after a reset.
 **/
static void ixgbe_configure_rx(struct ixgbe_adapter *adapter)
{
	u64 rdba;
	struct ixgbe_hw *hw = &adapter->hw;
	struct net_device *netdev = adapter->netdev;
	int max_frame = netdev->mtu + ETH_HLEN + ETH_FCS_LEN;
1731
	int i, j;
1732
	u32 rdlen, rxctrl, rxcsum;
1733 1734 1735
	static const u32 seed[10] = { 0xE291D73D, 0x1805EC6C, 0x2A94B30D,
	                  0xA54F2BEC, 0xEA49AF7C, 0xE214AD3D, 0xB855AABE,
	                  0x6A3E67EA, 0x14364D17, 0x3BED200D};
1736
	u32 fctrl, hlreg0;
1737
	u32 reta = 0, mrqc = 0;
1738
	u32 rdrxctl;
1739
	int rx_buf_len;
1740 1741

	/* Decide whether to use packet split mode or not */
1742
	adapter->flags |= IXGBE_FLAG_RX_PS_ENABLED;
1743 1744 1745

	/* Set the RX buffer length according to the mode */
	if (adapter->flags & IXGBE_FLAG_RX_PS_ENABLED) {
1746
		rx_buf_len = IXGBE_RX_HDR_SIZE;
1747 1748 1749 1750 1751 1752 1753 1754
		if (hw->mac.type == ixgbe_mac_82599EB) {
			/* PSRTYPE must be initialized in 82599 */
			u32 psrtype = IXGBE_PSRTYPE_TCPHDR |
			              IXGBE_PSRTYPE_UDPHDR |
			              IXGBE_PSRTYPE_IPV4HDR |
			              IXGBE_PSRTYPE_IPV6HDR;
			IXGBE_WRITE_REG(hw, IXGBE_PSRTYPE(0), psrtype);
		}
1755 1756
	} else {
		if (netdev->mtu <= ETH_DATA_LEN)
1757
			rx_buf_len = MAXIMUM_ETHERNET_VLAN_SIZE;
1758
		else
1759
			rx_buf_len = ALIGN(max_frame, 1024);
1760 1761 1762 1763
	}

	fctrl = IXGBE_READ_REG(&adapter->hw, IXGBE_FCTRL);
	fctrl |= IXGBE_FCTRL_BAM;
1764
	fctrl |= IXGBE_FCTRL_DPF; /* discard pause frames when FC enabled */
1765
	fctrl |= IXGBE_FCTRL_PMCF;
1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783
	IXGBE_WRITE_REG(&adapter->hw, IXGBE_FCTRL, fctrl);

	hlreg0 = IXGBE_READ_REG(hw, IXGBE_HLREG0);
	if (adapter->netdev->mtu <= ETH_DATA_LEN)
		hlreg0 &= ~IXGBE_HLREG0_JUMBOEN;
	else
		hlreg0 |= IXGBE_HLREG0_JUMBOEN;
	IXGBE_WRITE_REG(hw, IXGBE_HLREG0, hlreg0);

	rdlen = adapter->rx_ring[0].count * sizeof(union ixgbe_adv_rx_desc);
	/* disable receives while setting up the descriptors */
	rxctrl = IXGBE_READ_REG(hw, IXGBE_RXCTRL);
	IXGBE_WRITE_REG(hw, IXGBE_RXCTRL, rxctrl & ~IXGBE_RXCTRL_RXEN);

	/* Setup the HW Rx Head and Tail Descriptor Pointers and
	 * the Base and Length of the Rx Descriptor Ring */
	for (i = 0; i < adapter->num_rx_queues; i++) {
		rdba = adapter->rx_ring[i].dma;
1784
		j = adapter->rx_ring[i].reg_idx;
1785
		IXGBE_WRITE_REG(hw, IXGBE_RDBAL(j), (rdba & DMA_BIT_MASK(32)));
1786 1787 1788 1789 1790 1791 1792
		IXGBE_WRITE_REG(hw, IXGBE_RDBAH(j), (rdba >> 32));
		IXGBE_WRITE_REG(hw, IXGBE_RDLEN(j), rdlen);
		IXGBE_WRITE_REG(hw, IXGBE_RDH(j), 0);
		IXGBE_WRITE_REG(hw, IXGBE_RDT(j), 0);
		adapter->rx_ring[i].head = IXGBE_RDH(j);
		adapter->rx_ring[i].tail = IXGBE_RDT(j);
		adapter->rx_ring[i].rx_buf_len = rx_buf_len;
1793 1794

		ixgbe_configure_srrctl(adapter, j);
1795 1796
	}

1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807
	if (hw->mac.type == ixgbe_mac_82598EB) {
		/*
		 * For VMDq support of different descriptor types or
		 * buffer sizes through the use of multiple SRRCTL
		 * registers, RDRXCTL.MVMEN must be set to 1
		 *
		 * also, the manual doesn't mention it clearly but DCA hints
		 * will only use queue 0's tags unless this bit is set.  Side
		 * effects of setting this bit are only that SRRCTL must be
		 * fully programmed [0..15]
		 */
1808 1809 1810
		rdrxctl = IXGBE_READ_REG(hw, IXGBE_RDRXCTL);
		rdrxctl |= IXGBE_RDRXCTL_MVMEN;
		IXGBE_WRITE_REG(hw, IXGBE_RDRXCTL, rdrxctl);
1811
	}
1812

1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830
	/* Program MRQC for the distribution of queues */
	if (hw->mac.type == ixgbe_mac_82599EB) {
		int mask = adapter->flags & (
				IXGBE_FLAG_RSS_ENABLED
				| IXGBE_FLAG_DCB_ENABLED
				);

		switch (mask) {
		case (IXGBE_FLAG_RSS_ENABLED):
			mrqc = IXGBE_MRQC_RSSEN;
			break;
		case (IXGBE_FLAG_DCB_ENABLED):
			mrqc = IXGBE_MRQC_RT8TCEN;
			break;
		default:
			break;
		}
	}
1831
	if (adapter->flags & IXGBE_FLAG_RSS_ENABLED) {
1832
		/* Fill out redirection table */
1833 1834 1835 1836 1837 1838 1839 1840
		for (i = 0, j = 0; i < 128; i++, j++) {
			if (j == adapter->ring_feature[RING_F_RSS].indices)
				j = 0;
			/* reta = 4-byte sliding window of
			 * 0x00..(indices-1)(indices-1)00..etc. */
			reta = (reta << 8) | (j * 0x11);
			if ((i & 3) == 3)
				IXGBE_WRITE_REG(hw, IXGBE_RETA(i >> 2), reta);
1841 1842 1843 1844
		}

		/* Fill out hash function seeds */
		for (i = 0; i < 10; i++)
1845
			IXGBE_WRITE_REG(hw, IXGBE_RSSRK(i), seed[i]);
1846

1847 1848
		if (hw->mac.type == ixgbe_mac_82598EB)
			mrqc |= IXGBE_MRQC_RSSEN;
1849
		    /* Perform hash on these packet types */
1850 1851 1852 1853 1854 1855
		mrqc |= IXGBE_MRQC_RSS_FIELD_IPV4
		      | IXGBE_MRQC_RSS_FIELD_IPV4_TCP
		      | IXGBE_MRQC_RSS_FIELD_IPV4_UDP
		      | IXGBE_MRQC_RSS_FIELD_IPV6
		      | IXGBE_MRQC_RSS_FIELD_IPV6_TCP
		      | IXGBE_MRQC_RSS_FIELD_IPV6_UDP;
1856
	}
1857
	IXGBE_WRITE_REG(hw, IXGBE_MRQC, mrqc);
1858

1859 1860 1861 1862 1863 1864
	rxcsum = IXGBE_READ_REG(hw, IXGBE_RXCSUM);

	if (adapter->flags & IXGBE_FLAG_RSS_ENABLED ||
	    adapter->flags & IXGBE_FLAG_RX_CSUM_ENABLED) {
		/* Disable indicating checksum in descriptor, enables
		 * RSS hash */
1865 1866
		rxcsum |= IXGBE_RXCSUM_PCSD;
	}
1867 1868 1869 1870 1871 1872 1873
	if (!(rxcsum & IXGBE_RXCSUM_PCSD)) {
		/* Enable IPv4 payload checksum for UDP fragments
		 * if PCSD is not set */
		rxcsum |= IXGBE_RXCSUM_IPPCSE;
	}

	IXGBE_WRITE_REG(hw, IXGBE_RXCSUM, rxcsum);
1874 1875 1876 1877 1878 1879

	if (hw->mac.type == ixgbe_mac_82599EB) {
		rdrxctl = IXGBE_READ_REG(hw, IXGBE_RDRXCTL);
		rdrxctl |= IXGBE_RDRXCTL_CRCSTRIP;
		IXGBE_WRITE_REG(hw, IXGBE_RDRXCTL, rdrxctl);
	}
1880 1881
}

1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907
static void ixgbe_vlan_rx_add_vid(struct net_device *netdev, u16 vid)
{
	struct ixgbe_adapter *adapter = netdev_priv(netdev);
	struct ixgbe_hw *hw = &adapter->hw;

	/* add VID to filter table */
	hw->mac.ops.set_vfta(&adapter->hw, vid, 0, true);
}

static void ixgbe_vlan_rx_kill_vid(struct net_device *netdev, u16 vid)
{
	struct ixgbe_adapter *adapter = netdev_priv(netdev);
	struct ixgbe_hw *hw = &adapter->hw;

	if (!test_bit(__IXGBE_DOWN, &adapter->state))
		ixgbe_irq_disable(adapter);

	vlan_group_set_device(adapter->vlgrp, vid, NULL);

	if (!test_bit(__IXGBE_DOWN, &adapter->state))
		ixgbe_irq_enable(adapter);

	/* remove VID from filter table */
	hw->mac.ops.set_vfta(&adapter->hw, vid, 0, false);
}

1908
static void ixgbe_vlan_rx_register(struct net_device *netdev,
1909
                                   struct vlan_group *grp)
1910 1911 1912
{
	struct ixgbe_adapter *adapter = netdev_priv(netdev);
	u32 ctrl;
1913
	int i, j;
1914

1915 1916
	if (!test_bit(__IXGBE_DOWN, &adapter->state))
		ixgbe_irq_disable(adapter);
1917 1918
	adapter->vlgrp = grp;

1919 1920 1921 1922 1923 1924
	/*
	 * For a DCB driver, always enable VLAN tag stripping so we can
	 * still receive traffic from a DCB-enabled host even if we're
	 * not in DCB mode.
	 */
	ctrl = IXGBE_READ_REG(&adapter->hw, IXGBE_VLNCTRL);
1925 1926 1927 1928 1929 1930
	if (adapter->hw.mac.type == ixgbe_mac_82598EB) {
		ctrl |= IXGBE_VLNCTRL_VME | IXGBE_VLNCTRL_VFE;
		ctrl &= ~IXGBE_VLNCTRL_CFIEN;
		IXGBE_WRITE_REG(&adapter->hw, IXGBE_VLNCTRL, ctrl);
	} else if (adapter->hw.mac.type == ixgbe_mac_82599EB) {
		ctrl |= IXGBE_VLNCTRL_VFE;
1931 1932 1933 1934
		/* enable VLAN tag insert/strip */
		ctrl = IXGBE_READ_REG(&adapter->hw, IXGBE_VLNCTRL);
		ctrl &= ~IXGBE_VLNCTRL_CFIEN;
		IXGBE_WRITE_REG(&adapter->hw, IXGBE_VLNCTRL, ctrl);
1935 1936 1937 1938 1939 1940
		for (i = 0; i < adapter->num_rx_queues; i++) {
			j = adapter->rx_ring[i].reg_idx;
			ctrl = IXGBE_READ_REG(&adapter->hw, IXGBE_RXDCTL(j));
			ctrl |= IXGBE_RXDCTL_VME;
			IXGBE_WRITE_REG(&adapter->hw, IXGBE_RXDCTL(j), ctrl);
		}
1941
	}
1942
	ixgbe_vlan_rx_add_vid(netdev, 0);
1943

1944 1945
	if (!test_bit(__IXGBE_DOWN, &adapter->state))
		ixgbe_irq_enable(adapter);
1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961
}

static void ixgbe_restore_vlan(struct ixgbe_adapter *adapter)
{
	ixgbe_vlan_rx_register(adapter->netdev, adapter->vlgrp);

	if (adapter->vlgrp) {
		u16 vid;
		for (vid = 0; vid < VLAN_GROUP_ARRAY_LEN; vid++) {
			if (!vlan_group_get_device(adapter->vlgrp, vid))
				continue;
			ixgbe_vlan_rx_add_vid(adapter->netdev, vid);
		}
	}
}

1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976
static u8 *ixgbe_addr_list_itr(struct ixgbe_hw *hw, u8 **mc_addr_ptr, u32 *vmdq)
{
	struct dev_mc_list *mc_ptr;
	u8 *addr = *mc_addr_ptr;
	*vmdq = 0;

	mc_ptr = container_of(addr, struct dev_mc_list, dmi_addr[0]);
	if (mc_ptr->next)
		*mc_addr_ptr = mc_ptr->next->dmi_addr;
	else
		*mc_addr_ptr = NULL;

	return addr;
}

1977
/**
1978
 * ixgbe_set_rx_mode - Unicast, Multicast and Promiscuous mode set
1979 1980
 * @netdev: network interface device structure
 *
1981 1982 1983 1984
 * The set_rx_method entry point is called whenever the unicast/multicast
 * address list or the network interface flags are updated.  This routine is
 * responsible for configuring the hardware for proper unicast, multicast and
 * promiscuous mode.
1985
 **/
1986
static void ixgbe_set_rx_mode(struct net_device *netdev)
1987 1988 1989
{
	struct ixgbe_adapter *adapter = netdev_priv(netdev);
	struct ixgbe_hw *hw = &adapter->hw;
A
Alexander Duyck 已提交
1990
	u32 fctrl, vlnctrl;
1991 1992
	u8 *addr_list = NULL;
	int addr_count = 0;
1993 1994 1995 1996

	/* Check for Promiscuous and All Multicast modes */

	fctrl = IXGBE_READ_REG(hw, IXGBE_FCTRL);
A
Alexander Duyck 已提交
1997
	vlnctrl = IXGBE_READ_REG(hw, IXGBE_VLNCTRL);
1998 1999

	if (netdev->flags & IFF_PROMISC) {
2000
		hw->addr_ctrl.user_set_promisc = 1;
2001
		fctrl |= (IXGBE_FCTRL_UPE | IXGBE_FCTRL_MPE);
A
Alexander Duyck 已提交
2002
		vlnctrl &= ~IXGBE_VLNCTRL_VFE;
2003
	} else {
2004 2005 2006 2007 2008 2009
		if (netdev->flags & IFF_ALLMULTI) {
			fctrl |= IXGBE_FCTRL_MPE;
			fctrl &= ~IXGBE_FCTRL_UPE;
		} else {
			fctrl &= ~(IXGBE_FCTRL_UPE | IXGBE_FCTRL_MPE);
		}
A
Alexander Duyck 已提交
2010
		vlnctrl |= IXGBE_VLNCTRL_VFE;
2011
		hw->addr_ctrl.user_set_promisc = 0;
2012 2013 2014
	}

	IXGBE_WRITE_REG(hw, IXGBE_FCTRL, fctrl);
A
Alexander Duyck 已提交
2015
	IXGBE_WRITE_REG(hw, IXGBE_VLNCTRL, vlnctrl);
2016

2017 2018 2019 2020
	/* reprogram secondary unicast list */
	addr_count = netdev->uc_count;
	if (addr_count)
		addr_list = netdev->uc_list->dmi_addr;
2021 2022
	hw->mac.ops.update_uc_addr_list(hw, addr_list, addr_count,
	                                  ixgbe_addr_list_itr);
2023

2024 2025 2026 2027
	/* reprogram multicast list */
	addr_count = netdev->mc_count;
	if (addr_count)
		addr_list = netdev->mc_list->dmi_addr;
2028 2029
	hw->mac.ops.update_mc_addr_list(hw, addr_list, addr_count,
	                                ixgbe_addr_list_itr);
2030 2031
}

2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042
static void ixgbe_napi_enable_all(struct ixgbe_adapter *adapter)
{
	int q_idx;
	struct ixgbe_q_vector *q_vector;
	int q_vectors = adapter->num_msix_vectors - NON_Q_VECTORS;

	/* legacy and MSI only use one vector */
	if (!(adapter->flags & IXGBE_FLAG_MSIX_ENABLED))
		q_vectors = 1;

	for (q_idx = 0; q_idx < q_vectors; q_idx++) {
2043
		struct napi_struct *napi;
2044 2045 2046
		q_vector = &adapter->q_vector[q_idx];
		if (!q_vector->rxr_count)
			continue;
2047 2048 2049 2050 2051 2052
		napi = &q_vector->napi;
		if ((adapter->flags & IXGBE_FLAG_MSIX_ENABLED) &&
		    (q_vector->rxr_count > 1))
			napi->poll = &ixgbe_clean_rxonly_many;

		napi_enable(napi);
2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073
	}
}

static void ixgbe_napi_disable_all(struct ixgbe_adapter *adapter)
{
	int q_idx;
	struct ixgbe_q_vector *q_vector;
	int q_vectors = adapter->num_msix_vectors - NON_Q_VECTORS;

	/* legacy and MSI only use one vector */
	if (!(adapter->flags & IXGBE_FLAG_MSIX_ENABLED))
		q_vectors = 1;

	for (q_idx = 0; q_idx < q_vectors; q_idx++) {
		q_vector = &adapter->q_vector[q_idx];
		if (!q_vector->rxr_count)
			continue;
		napi_disable(&q_vector->napi);
	}
}

J
Jeff Kirsher 已提交
2074
#ifdef CONFIG_IXGBE_DCB
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
/*
 * ixgbe_configure_dcb - Configure DCB hardware
 * @adapter: ixgbe adapter struct
 *
 * This is called by the driver on open to configure the DCB hardware.
 * This is also called by the gennetlink interface when reconfiguring
 * the DCB state.
 */
static void ixgbe_configure_dcb(struct ixgbe_adapter *adapter)
{
	struct ixgbe_hw *hw = &adapter->hw;
	u32 txdctl, vlnctrl;
	int i, j;

	ixgbe_dcb_check_config(&adapter->dcb_cfg);
	ixgbe_dcb_calculate_tc_credits(&adapter->dcb_cfg, DCB_TX_CONFIG);
	ixgbe_dcb_calculate_tc_credits(&adapter->dcb_cfg, DCB_RX_CONFIG);

	/* reconfigure the hardware */
	ixgbe_dcb_hw_config(&adapter->hw, &adapter->dcb_cfg);

	for (i = 0; i < adapter->num_tx_queues; i++) {
		j = adapter->tx_ring[i].reg_idx;
		txdctl = IXGBE_READ_REG(hw, IXGBE_TXDCTL(j));
		/* PThresh workaround for Tx hang with DFP enabled. */
		txdctl |= 32;
		IXGBE_WRITE_REG(hw, IXGBE_TXDCTL(j), txdctl);
	}
	/* Enable VLAN tag insert/strip */
	vlnctrl = IXGBE_READ_REG(hw, IXGBE_VLNCTRL);
2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119
	if (hw->mac.type == ixgbe_mac_82598EB) {
		vlnctrl |= IXGBE_VLNCTRL_VME | IXGBE_VLNCTRL_VFE;
		vlnctrl &= ~IXGBE_VLNCTRL_CFIEN;
		IXGBE_WRITE_REG(hw, IXGBE_VLNCTRL, vlnctrl);
	} else if (hw->mac.type == ixgbe_mac_82599EB) {
		vlnctrl |= IXGBE_VLNCTRL_VFE;
		vlnctrl &= ~IXGBE_VLNCTRL_CFIEN;
		IXGBE_WRITE_REG(hw, IXGBE_VLNCTRL, vlnctrl);
		for (i = 0; i < adapter->num_rx_queues; i++) {
			j = adapter->rx_ring[i].reg_idx;
			vlnctrl = IXGBE_READ_REG(hw, IXGBE_RXDCTL(j));
			vlnctrl |= IXGBE_RXDCTL_VME;
			IXGBE_WRITE_REG(hw, IXGBE_RXDCTL(j), vlnctrl);
		}
	}
2120 2121 2122 2123
	hw->mac.ops.set_vfta(&adapter->hw, 0, 0, true);
}

#endif
2124 2125 2126 2127 2128
static void ixgbe_configure(struct ixgbe_adapter *adapter)
{
	struct net_device *netdev = adapter->netdev;
	int i;

2129
	ixgbe_set_rx_mode(netdev);
2130 2131

	ixgbe_restore_vlan(adapter);
J
Jeff Kirsher 已提交
2132
#ifdef CONFIG_IXGBE_DCB
2133 2134 2135 2136 2137 2138 2139 2140 2141
	if (adapter->flags & IXGBE_FLAG_DCB_ENABLED) {
		netif_set_gso_max_size(netdev, 32768);
		ixgbe_configure_dcb(adapter);
	} else {
		netif_set_gso_max_size(netdev, 65536);
	}
#else
	netif_set_gso_max_size(netdev, 65536);
#endif
2142 2143 2144 2145 2146

	ixgbe_configure_tx(adapter);
	ixgbe_configure_rx(adapter);
	for (i = 0; i < adapter->num_rx_queues; i++)
		ixgbe_alloc_rx_buffers(adapter, &adapter->rx_ring[i],
2147
		                       (adapter->rx_ring[i].count - 1));
2148 2149
}

2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164
static inline bool ixgbe_is_sfp(struct ixgbe_hw *hw)
{
	switch (hw->phy.type) {
	case ixgbe_phy_sfp_avago:
	case ixgbe_phy_sfp_ftl:
	case ixgbe_phy_sfp_intel:
	case ixgbe_phy_sfp_unknown:
	case ixgbe_phy_tw_tyco:
	case ixgbe_phy_tw_unknown:
		return true;
	default:
		return false;
	}
}

2165
/**
2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198
 * ixgbe_sfp_link_config - set up SFP+ link
 * @adapter: pointer to private adapter struct
 **/
static void ixgbe_sfp_link_config(struct ixgbe_adapter *adapter)
{
	struct ixgbe_hw *hw = &adapter->hw;

		if (hw->phy.multispeed_fiber) {
			/*
			 * In multispeed fiber setups, the device may not have
			 * had a physical connection when the driver loaded.
			 * If that's the case, the initial link configuration
			 * couldn't get the MAC into 10G or 1G mode, so we'll
			 * never have a link status change interrupt fire.
			 * We need to try and force an autonegotiation
			 * session, then bring up link.
			 */
			hw->mac.ops.setup_sfp(hw);
			if (!(adapter->flags & IXGBE_FLAG_IN_SFP_LINK_TASK))
				schedule_work(&adapter->multispeed_fiber_task);
		} else {
			/*
			 * Direct Attach Cu and non-multispeed fiber modules
			 * still need to be configured properly prior to
			 * attempting link.
			 */
			if (!(adapter->flags & IXGBE_FLAG_IN_SFP_MOD_TASK))
				schedule_work(&adapter->sfp_config_module_task);
		}
}

/**
 * ixgbe_non_sfp_link_config - set up non-SFP+ link
2199 2200 2201 2202
 * @hw: pointer to private hardware struct
 *
 * Returns 0 on success, negative on failure
 **/
2203
static int ixgbe_non_sfp_link_config(struct ixgbe_hw *hw)
2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 2224 2225 2226
{
	u32 autoneg;
	bool link_up = false;
	u32 ret = IXGBE_ERR_LINK_SETUP;

	if (hw->mac.ops.check_link)
		ret = hw->mac.ops.check_link(hw, &autoneg, &link_up, false);

	if (ret)
		goto link_cfg_out;

	if (hw->mac.ops.get_link_capabilities)
		ret = hw->mac.ops.get_link_capabilities(hw, &autoneg,
		                                        &hw->mac.autoneg);
	if (ret)
		goto link_cfg_out;

	if (hw->mac.ops.setup_link_speed)
		ret = hw->mac.ops.setup_link_speed(hw, autoneg, true, link_up);
link_cfg_out:
	return ret;
}

2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240 2241 2242 2243 2244 2245 2246 2247 2248
#define IXGBE_MAX_RX_DESC_POLL 10
static inline void ixgbe_rx_desc_queue_enable(struct ixgbe_adapter *adapter,
	                                      int rxr)
{
	int j = adapter->rx_ring[rxr].reg_idx;
	int k;

	for (k = 0; k < IXGBE_MAX_RX_DESC_POLL; k++) {
		if (IXGBE_READ_REG(&adapter->hw,
		                   IXGBE_RXDCTL(j)) & IXGBE_RXDCTL_ENABLE)
			break;
		else
			msleep(1);
	}
	if (k >= IXGBE_MAX_RX_DESC_POLL) {
		DPRINTK(DRV, ERR, "RXDCTL.ENABLE on Rx queue %d "
		        "not set within the polling period\n", rxr);
	}
	ixgbe_release_rx_desc(&adapter->hw, &adapter->rx_ring[rxr],
	                      (adapter->rx_ring[rxr].count - 1));
}

2249 2250 2251 2252
static int ixgbe_up_complete(struct ixgbe_adapter *adapter)
{
	struct net_device *netdev = adapter->netdev;
	struct ixgbe_hw *hw = &adapter->hw;
2253
	int i, j = 0;
2254
	int num_rx_rings = adapter->num_rx_queues;
2255
	int err;
2256
	int max_frame = netdev->mtu + ETH_HLEN + ETH_FCS_LEN;
2257
	u32 txdctl, rxdctl, mhadd;
2258
	u32 dmatxctl;
2259
	u32 gpie;
2260

2261 2262
	ixgbe_get_hw_control(adapter);

2263 2264
	if ((adapter->flags & IXGBE_FLAG_MSIX_ENABLED) ||
	    (adapter->flags & IXGBE_FLAG_MSI_ENABLED)) {
2265 2266
		if (adapter->flags & IXGBE_FLAG_MSIX_ENABLED) {
			gpie = (IXGBE_GPIE_MSIX_MODE | IXGBE_GPIE_EIAME |
2267
			        IXGBE_GPIE_PBA_SUPPORT | IXGBE_GPIE_OCD);
2268 2269
		} else {
			/* MSI only */
2270
			gpie = 0;
2271
		}
2272 2273 2274
		/* XXX: to interrupt immediately for EICS writes, enable this */
		/* gpie |= IXGBE_GPIE_EIMEN; */
		IXGBE_WRITE_REG(hw, IXGBE_GPIE, gpie);
2275 2276
	}

2277 2278 2279 2280 2281
	if (!(adapter->flags & IXGBE_FLAG_MSIX_ENABLED)) {
		/* legacy interrupts, use EIAM to auto-mask when reading EICR,
		 * specifically only auto mask tx and rx interrupts */
		IXGBE_WRITE_REG(hw, IXGBE_EIAM, IXGBE_EICS_RTX_QUEUE);
	}
2282

2283 2284 2285 2286 2287 2288 2289
	/* Enable fan failure interrupt if media type is copper */
	if (adapter->flags & IXGBE_FLAG_FAN_FAIL_CAPABLE) {
		gpie = IXGBE_READ_REG(hw, IXGBE_GPIE);
		gpie |= IXGBE_SDP1_GPIEN;
		IXGBE_WRITE_REG(hw, IXGBE_GPIE, gpie);
	}

2290 2291 2292 2293 2294 2295 2296
	if (hw->mac.type == ixgbe_mac_82599EB) {
		gpie = IXGBE_READ_REG(hw, IXGBE_GPIE);
		gpie |= IXGBE_SDP1_GPIEN;
		gpie |= IXGBE_SDP2_GPIEN;
		IXGBE_WRITE_REG(hw, IXGBE_GPIE, gpie);
	}

2297
	mhadd = IXGBE_READ_REG(hw, IXGBE_MHADD);
2298 2299 2300 2301 2302 2303 2304 2305
	if (max_frame != (mhadd >> IXGBE_MHADD_MFS_SHIFT)) {
		mhadd &= ~IXGBE_MHADD_MFS_MASK;
		mhadd |= max_frame << IXGBE_MHADD_MFS_SHIFT;

		IXGBE_WRITE_REG(hw, IXGBE_MHADD, mhadd);
	}

	for (i = 0; i < adapter->num_tx_queues; i++) {
2306 2307
		j = adapter->tx_ring[i].reg_idx;
		txdctl = IXGBE_READ_REG(hw, IXGBE_TXDCTL(j));
2308 2309
		/* enable WTHRESH=8 descriptors, to encourage burst writeback */
		txdctl |= (8 << 16);
2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321
		IXGBE_WRITE_REG(hw, IXGBE_TXDCTL(j), txdctl);
	}

	if (hw->mac.type == ixgbe_mac_82599EB) {
		/* DMATXCTL.EN must be set after all Tx queue config is done */
		dmatxctl = IXGBE_READ_REG(hw, IXGBE_DMATXCTL);
		dmatxctl |= IXGBE_DMATXCTL_TE;
		IXGBE_WRITE_REG(hw, IXGBE_DMATXCTL, dmatxctl);
	}
	for (i = 0; i < adapter->num_tx_queues; i++) {
		j = adapter->tx_ring[i].reg_idx;
		txdctl = IXGBE_READ_REG(hw, IXGBE_TXDCTL(j));
2322
		txdctl |= IXGBE_TXDCTL_ENABLE;
2323
		IXGBE_WRITE_REG(hw, IXGBE_TXDCTL(j), txdctl);
2324 2325
	}

2326
	for (i = 0; i < num_rx_rings; i++) {
2327 2328 2329 2330 2331 2332
		j = adapter->rx_ring[i].reg_idx;
		rxdctl = IXGBE_READ_REG(hw, IXGBE_RXDCTL(j));
		/* enable PTHRESH=32 descriptors (half the internal cache)
		 * and HTHRESH=0 descriptors (to minimize latency on fetch),
		 * this also removes a pesky rx_no_buffer_count increment */
		rxdctl |= 0x0020;
2333
		rxdctl |= IXGBE_RXDCTL_ENABLE;
2334
		IXGBE_WRITE_REG(hw, IXGBE_RXDCTL(j), rxdctl);
2335 2336
		if (hw->mac.type == ixgbe_mac_82599EB)
			ixgbe_rx_desc_queue_enable(adapter, i);
2337 2338 2339
	}
	/* enable all receives */
	rxdctl = IXGBE_READ_REG(hw, IXGBE_RXCTRL);
2340 2341 2342 2343 2344
	if (hw->mac.type == ixgbe_mac_82598EB)
		rxdctl |= (IXGBE_RXCTRL_DMBYPS | IXGBE_RXCTRL_RXEN);
	else
		rxdctl |= IXGBE_RXCTRL_RXEN;
	hw->mac.ops.enable_rx_dma(hw, rxdctl);
2345 2346 2347 2348 2349 2350 2351

	if (adapter->flags & IXGBE_FLAG_MSIX_ENABLED)
		ixgbe_configure_msix(adapter);
	else
		ixgbe_configure_msi_and_legacy(adapter);

	clear_bit(__IXGBE_DOWN, &adapter->state);
2352 2353 2354 2355 2356
	ixgbe_napi_enable_all(adapter);

	/* clear any pending interrupts, may auto mask */
	IXGBE_READ_REG(hw, IXGBE_EICR);

2357 2358
	ixgbe_irq_enable(adapter);

2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379
	/*
	 * For hot-pluggable SFP+ devices, a new SFP+ module may have
	 * arrived before interrupts were enabled.  We need to kick off
	 * the SFP+ module setup first, then try to bring up link.
	 * If we're not hot-pluggable SFP+, we just need to configure link
	 * and bring it up.
	 */
	err = hw->phy.ops.identify(hw);
	if (err == IXGBE_ERR_SFP_NOT_SUPPORTED) {
		DPRINTK(PROBE, ERR, "PHY not supported on this NIC %d\n", err);
		ixgbe_down(adapter);
		return err;
	}

	if (ixgbe_is_sfp(hw)) {
		ixgbe_sfp_link_config(adapter);
	} else {
		err = ixgbe_non_sfp_link_config(hw);
		if (err)
			DPRINTK(PROBE, ERR, "link_config FAILED %d\n", err);
	}
2380

2381 2382 2383
	/* enable transmits */
	netif_tx_start_all_queues(netdev);

2384 2385
	/* bring the link up in the watchdog, this could race with our first
	 * link up interrupt but shouldn't be a problem */
2386 2387
	adapter->flags |= IXGBE_FLAG_NEED_LINK_UPDATE;
	adapter->link_check_timeout = jiffies;
2388 2389 2390 2391
	mod_timer(&adapter->watchdog_timer, jiffies);
	return 0;
}

2392 2393 2394 2395 2396 2397 2398 2399 2400 2401
void ixgbe_reinit_locked(struct ixgbe_adapter *adapter)
{
	WARN_ON(in_interrupt());
	while (test_and_set_bit(__IXGBE_RESETTING, &adapter->state))
		msleep(1);
	ixgbe_down(adapter);
	ixgbe_up(adapter);
	clear_bit(__IXGBE_RESETTING, &adapter->state);
}

2402 2403 2404 2405 2406
int ixgbe_up(struct ixgbe_adapter *adapter)
{
	/* hardware has been reset, we need to reload some things */
	ixgbe_configure(adapter);

2407 2408
	ixgbe_napi_add_all(adapter);

2409 2410 2411 2412 2413
	return ixgbe_up_complete(adapter);
}

void ixgbe_reset(struct ixgbe_adapter *adapter)
{
2414 2415 2416
	struct ixgbe_hw *hw = &adapter->hw;
	if (hw->mac.ops.init_hw(hw))
		dev_err(&adapter->pdev->dev, "Hardware Error\n");
2417 2418

	/* reprogram the RAR[0] in case user changed it. */
2419
	hw->mac.ops.set_rar(hw, 0, hw->mac.addr, 0, IXGBE_RAH_AV);
2420 2421 2422 2423 2424 2425 2426 2427 2428

}

/**
 * ixgbe_clean_rx_ring - Free Rx Buffers per Queue
 * @adapter: board private structure
 * @rx_ring: ring to free buffers from
 **/
static void ixgbe_clean_rx_ring(struct ixgbe_adapter *adapter,
2429
                                struct ixgbe_ring *rx_ring)
2430 2431 2432 2433 2434 2435 2436 2437 2438 2439 2440 2441 2442
{
	struct pci_dev *pdev = adapter->pdev;
	unsigned long size;
	unsigned int i;

	/* Free all the Rx ring sk_buffs */

	for (i = 0; i < rx_ring->count; i++) {
		struct ixgbe_rx_buffer *rx_buffer_info;

		rx_buffer_info = &rx_ring->rx_buffer_info[i];
		if (rx_buffer_info->dma) {
			pci_unmap_single(pdev, rx_buffer_info->dma,
2443 2444
			                 rx_ring->rx_buf_len,
			                 PCI_DMA_FROMDEVICE);
2445 2446 2447 2448 2449 2450 2451 2452
			rx_buffer_info->dma = 0;
		}
		if (rx_buffer_info->skb) {
			dev_kfree_skb(rx_buffer_info->skb);
			rx_buffer_info->skb = NULL;
		}
		if (!rx_buffer_info->page)
			continue;
2453 2454
		pci_unmap_page(pdev, rx_buffer_info->page_dma, PAGE_SIZE / 2,
		               PCI_DMA_FROMDEVICE);
2455 2456 2457
		rx_buffer_info->page_dma = 0;
		put_page(rx_buffer_info->page);
		rx_buffer_info->page = NULL;
2458
		rx_buffer_info->page_offset = 0;
2459 2460 2461 2462 2463 2464 2465 2466 2467 2468 2469
	}

	size = sizeof(struct ixgbe_rx_buffer) * rx_ring->count;
	memset(rx_ring->rx_buffer_info, 0, size);

	/* Zero out the descriptor ring */
	memset(rx_ring->desc, 0, rx_ring->size);

	rx_ring->next_to_clean = 0;
	rx_ring->next_to_use = 0;

2470 2471 2472 2473
	if (rx_ring->head)
		writel(0, adapter->hw.hw_addr + rx_ring->head);
	if (rx_ring->tail)
		writel(0, adapter->hw.hw_addr + rx_ring->tail);
2474 2475 2476 2477 2478 2479 2480 2481
}

/**
 * ixgbe_clean_tx_ring - Free Tx Buffers
 * @adapter: board private structure
 * @tx_ring: ring to be cleaned
 **/
static void ixgbe_clean_tx_ring(struct ixgbe_adapter *adapter,
2482
                                struct ixgbe_ring *tx_ring)
2483 2484 2485 2486 2487 2488 2489 2490 2491 2492 2493 2494 2495 2496 2497 2498 2499 2500 2501 2502 2503
{
	struct ixgbe_tx_buffer *tx_buffer_info;
	unsigned long size;
	unsigned int i;

	/* Free all the Tx ring sk_buffs */

	for (i = 0; i < tx_ring->count; i++) {
		tx_buffer_info = &tx_ring->tx_buffer_info[i];
		ixgbe_unmap_and_free_tx_resource(adapter, tx_buffer_info);
	}

	size = sizeof(struct ixgbe_tx_buffer) * tx_ring->count;
	memset(tx_ring->tx_buffer_info, 0, size);

	/* Zero out the descriptor ring */
	memset(tx_ring->desc, 0, tx_ring->size);

	tx_ring->next_to_use = 0;
	tx_ring->next_to_clean = 0;

2504 2505 2506 2507
	if (tx_ring->head)
		writel(0, adapter->hw.hw_addr + tx_ring->head);
	if (tx_ring->tail)
		writel(0, adapter->hw.hw_addr + tx_ring->tail);
2508 2509 2510
}

/**
2511
 * ixgbe_clean_all_rx_rings - Free Rx Buffers for all queues
2512 2513
 * @adapter: board private structure
 **/
2514
static void ixgbe_clean_all_rx_rings(struct ixgbe_adapter *adapter)
2515 2516 2517
{
	int i;

2518 2519
	for (i = 0; i < adapter->num_rx_queues; i++)
		ixgbe_clean_rx_ring(adapter, &adapter->rx_ring[i]);
2520 2521 2522
}

/**
2523
 * ixgbe_clean_all_tx_rings - Free Tx Buffers for all queues
2524 2525
 * @adapter: board private structure
 **/
2526
static void ixgbe_clean_all_tx_rings(struct ixgbe_adapter *adapter)
2527 2528 2529
{
	int i;

2530 2531
	for (i = 0; i < adapter->num_tx_queues; i++)
		ixgbe_clean_tx_ring(adapter, &adapter->tx_ring[i]);
2532 2533 2534 2535 2536
}

void ixgbe_down(struct ixgbe_adapter *adapter)
{
	struct net_device *netdev = adapter->netdev;
2537
	struct ixgbe_hw *hw = &adapter->hw;
2538
	u32 rxctrl;
2539 2540
	u32 txdctl;
	int i, j;
2541 2542 2543 2544 2545

	/* signal that we are down to the interrupt handler */
	set_bit(__IXGBE_DOWN, &adapter->state);

	/* disable receives */
2546 2547
	rxctrl = IXGBE_READ_REG(hw, IXGBE_RXCTRL);
	IXGBE_WRITE_REG(hw, IXGBE_RXCTRL, rxctrl & ~IXGBE_RXCTRL_RXEN);
2548 2549 2550

	netif_tx_disable(netdev);

2551
	IXGBE_WRITE_FLUSH(hw);
2552 2553
	msleep(10);

2554 2555
	netif_tx_stop_all_queues(netdev);

2556 2557
	ixgbe_irq_disable(adapter);

2558
	ixgbe_napi_disable_all(adapter);
2559

2560
	del_timer_sync(&adapter->watchdog_timer);
2561
	cancel_work_sync(&adapter->watchdog_task);
2562

2563 2564 2565 2566 2567 2568 2569
	/* disable transmits in the hardware now that interrupts are off */
	for (i = 0; i < adapter->num_tx_queues; i++) {
		j = adapter->tx_ring[i].reg_idx;
		txdctl = IXGBE_READ_REG(hw, IXGBE_TXDCTL(j));
		IXGBE_WRITE_REG(hw, IXGBE_TXDCTL(j),
		                (txdctl & ~IXGBE_TXDCTL_ENABLE));
	}
2570 2571 2572 2573 2574
	/* Disable the Tx DMA engine on 82599 */
	if (hw->mac.type == ixgbe_mac_82599EB)
		IXGBE_WRITE_REG(hw, IXGBE_DMATXCTL,
		                (IXGBE_READ_REG(hw, IXGBE_DMATXCTL) &
		                 ~IXGBE_DMATXCTL_TE));
2575

2576 2577
	netif_carrier_off(netdev);

2578
#ifdef CONFIG_IXGBE_DCA
2579 2580 2581 2582 2583 2584
	if (adapter->flags & IXGBE_FLAG_DCA_ENABLED) {
		adapter->flags &= ~IXGBE_FLAG_DCA_ENABLED;
		dca_remove_requester(&adapter->pdev->dev);
	}

#endif
2585 2586
	if (!pci_channel_offline(adapter->pdev))
		ixgbe_reset(adapter);
2587 2588 2589
	ixgbe_clean_all_tx_rings(adapter);
	ixgbe_clean_all_rx_rings(adapter);

2590
#ifdef CONFIG_IXGBE_DCA
2591 2592 2593 2594 2595
	/* since we reset the hardware DCA settings were cleared */
	if (dca_add_requester(&adapter->pdev->dev) == 0) {
		adapter->flags |= IXGBE_FLAG_DCA_ENABLED;
		/* always use CB2 mode, difference is masked
		 * in the CB driver */
2596
		IXGBE_WRITE_REG(hw, IXGBE_DCA_CTRL, 2);
2597 2598 2599
		ixgbe_setup_dca(adapter);
	}
#endif
2600 2601 2602
}

/**
2603 2604 2605 2606 2607
 * ixgbe_poll - NAPI Rx polling callback
 * @napi: structure for representing this polling device
 * @budget: how many packets driver is allowed to clean
 *
 * This function is used for legacy and MSI, NAPI mode
2608
 **/
2609
static int ixgbe_poll(struct napi_struct *napi, int budget)
2610
{
2611 2612
	struct ixgbe_q_vector *q_vector =
	                        container_of(napi, struct ixgbe_q_vector, napi);
2613
	struct ixgbe_adapter *adapter = q_vector->adapter;
2614
	int tx_clean_complete, work_done = 0;
2615

2616
#ifdef CONFIG_IXGBE_DCA
2617 2618 2619 2620 2621 2622
	if (adapter->flags & IXGBE_FLAG_DCA_ENABLED) {
		ixgbe_update_tx_dca(adapter, adapter->tx_ring);
		ixgbe_update_rx_dca(adapter, adapter->rx_ring);
	}
#endif

2623
	tx_clean_complete = ixgbe_clean_tx_irq(adapter, adapter->tx_ring);
H
Herbert Xu 已提交
2624
	ixgbe_clean_rx_irq(q_vector, adapter->rx_ring, &work_done, budget);
2625

2626
	if (!tx_clean_complete)
2627 2628
		work_done = budget;

2629 2630
	/* If budget not fully consumed, exit the polling mode */
	if (work_done < budget) {
2631
		napi_complete(napi);
2632
		if (adapter->itr_setting & 1)
2633
			ixgbe_set_itr(adapter);
2634
		if (!test_bit(__IXGBE_DOWN, &adapter->state))
2635
			ixgbe_irq_enable_queues(adapter);
2636 2637 2638 2639 2640 2641 2642 2643 2644 2645 2646 2647 2648 2649 2650 2651 2652 2653 2654 2655 2656
	}
	return work_done;
}

/**
 * ixgbe_tx_timeout - Respond to a Tx Hang
 * @netdev: network interface device structure
 **/
static void ixgbe_tx_timeout(struct net_device *netdev)
{
	struct ixgbe_adapter *adapter = netdev_priv(netdev);

	/* Do the reset outside of interrupt context */
	schedule_work(&adapter->reset_task);
}

static void ixgbe_reset_task(struct work_struct *work)
{
	struct ixgbe_adapter *adapter;
	adapter = container_of(work, struct ixgbe_adapter, reset_task);

2657 2658 2659 2660 2661
	/* If we're already down or resetting, just bail */
	if (test_bit(__IXGBE_DOWN, &adapter->state) ||
	    test_bit(__IXGBE_RESETTING, &adapter->state))
		return;

2662 2663
	adapter->tx_timeout_count++;

2664
	ixgbe_reinit_locked(adapter);
2665 2666
}

2667 2668
#ifdef CONFIG_IXGBE_DCB
static inline bool ixgbe_set_dcb_queues(struct ixgbe_adapter *adapter)
2669
{
2670
	bool ret = false;
2671

2672 2673 2674 2675 2676 2677 2678 2679 2680 2681
	if (adapter->flags & IXGBE_FLAG_DCB_ENABLED) {
		adapter->ring_feature[RING_F_DCB].mask = 0x7 << 3;
		adapter->num_rx_queues =
		                      adapter->ring_feature[RING_F_DCB].indices;
		adapter->num_tx_queues =
		                      adapter->ring_feature[RING_F_DCB].indices;
		ret = true;
	} else {
		ret = false;
	}
2682

2683 2684 2685 2686
	return ret;
}
#endif

2687 2688 2689 2690 2691 2692 2693 2694
/**
 * ixgbe_set_rss_queues: Allocate queues for RSS
 * @adapter: board private structure to initialize
 *
 * This is our "base" multiqueue mode.  RSS (Receive Side Scaling) will try
 * to allocate one Rx queue per CPU, and if available, one Tx queue per CPU.
 *
 **/
2695 2696 2697 2698 2699 2700 2701 2702 2703 2704 2705 2706 2707
static inline bool ixgbe_set_rss_queues(struct ixgbe_adapter *adapter)
{
	bool ret = false;

	if (adapter->flags & IXGBE_FLAG_RSS_ENABLED) {
		adapter->ring_feature[RING_F_RSS].mask = 0xF;
		adapter->num_rx_queues =
		                      adapter->ring_feature[RING_F_RSS].indices;
		adapter->num_tx_queues =
		                      adapter->ring_feature[RING_F_RSS].indices;
		ret = true;
	} else {
		ret = false;
2708 2709
	}

2710 2711 2712
	return ret;
}

2713 2714 2715 2716 2717 2718 2719 2720 2721 2722 2723
/*
 * ixgbe_set_num_queues: Allocate queues for device, feature dependant
 * @adapter: board private structure to initialize
 *
 * This is the top level queue allocation routine.  The order here is very
 * important, starting with the "most" number of features turned on at once,
 * and ending with the smallest set of features.  This way large combinations
 * can be allocated if they're turned on, and smaller combinations are the
 * fallthrough conditions.
 *
 **/
2724 2725 2726 2727
static void ixgbe_set_num_queues(struct ixgbe_adapter *adapter)
{
#ifdef CONFIG_IXGBE_DCB
	if (ixgbe_set_dcb_queues(adapter))
2728
		goto done;
2729 2730 2731

#endif
	if (ixgbe_set_rss_queues(adapter))
2732 2733 2734 2735 2736 2737 2738 2739 2740
		goto done;

	/* fallback to base case */
	adapter->num_rx_queues = 1;
	adapter->num_tx_queues = 1;

done:
	/* Notify the stack of the (possibly) reduced Tx Queue count. */
	adapter->netdev->real_num_tx_queues = adapter->num_tx_queues;
2741 2742
}

2743
static void ixgbe_acquire_msix_vectors(struct ixgbe_adapter *adapter,
2744
                                       int vectors)
2745 2746 2747 2748 2749 2750 2751 2752 2753 2754 2755 2756 2757 2758 2759 2760 2761 2762
{
	int err, vector_threshold;

	/* We'll want at least 3 (vector_threshold):
	 * 1) TxQ[0] Cleanup
	 * 2) RxQ[0] Cleanup
	 * 3) Other (Link Status Change, etc.)
	 * 4) TCP Timer (optional)
	 */
	vector_threshold = MIN_MSIX_COUNT;

	/* The more we get, the more we will assign to Tx/Rx Cleanup
	 * for the separate queues...where Rx Cleanup >= Tx Cleanup.
	 * Right now, we simply care about how many we'll get; we'll
	 * set them up later while requesting irq's.
	 */
	while (vectors >= vector_threshold) {
		err = pci_enable_msix(adapter->pdev, adapter->msix_entries,
2763
		                      vectors);
2764 2765 2766 2767 2768 2769 2770 2771 2772 2773 2774 2775 2776 2777 2778 2779 2780
		if (!err) /* Success in acquiring all requested vectors. */
			break;
		else if (err < 0)
			vectors = 0; /* Nasty failure, quit now */
		else /* err == number of vectors we should try again with */
			vectors = err;
	}

	if (vectors < vector_threshold) {
		/* Can't allocate enough MSI-X interrupts?  Oh well.
		 * This just means we'll go with either a single MSI
		 * vector or fall back to legacy interrupts.
		 */
		DPRINTK(HW, DEBUG, "Unable to allocate MSI-X interrupts\n");
		adapter->flags &= ~IXGBE_FLAG_MSIX_ENABLED;
		kfree(adapter->msix_entries);
		adapter->msix_entries = NULL;
2781
		adapter->flags &= ~IXGBE_FLAG_DCB_ENABLED;
2782
		adapter->flags &= ~IXGBE_FLAG_RSS_ENABLED;
2783
		ixgbe_set_num_queues(adapter);
2784 2785
	} else {
		adapter->flags |= IXGBE_FLAG_MSIX_ENABLED; /* Woot! */
2786 2787 2788 2789 2790 2791 2792
		/*
		 * Adjust for only the vectors we'll use, which is minimum
		 * of max_msix_q_vectors + NON_Q_VECTORS, or the number of
		 * vectors we were allocated.
		 */
		adapter->num_msix_vectors = min(vectors,
		                   adapter->max_msix_q_vectors + NON_Q_VECTORS);
2793 2794 2795 2796
	}
}

/**
2797
 * ixgbe_cache_ring_rss - Descriptor ring to register mapping for RSS
2798 2799
 * @adapter: board private structure to initialize
 *
2800 2801
 * Cache the descriptor ring offsets for RSS to the assigned rings.
 *
2802
 **/
2803
static inline bool ixgbe_cache_ring_rss(struct ixgbe_adapter *adapter)
2804
{
2805 2806 2807 2808 2809 2810 2811 2812 2813 2814 2815 2816 2817 2818 2819 2820 2821 2822 2823 2824 2825 2826 2827 2828 2829 2830 2831 2832 2833 2834 2835 2836
	int i;
	bool ret = false;

	if (adapter->flags & IXGBE_FLAG_RSS_ENABLED) {
		for (i = 0; i < adapter->num_rx_queues; i++)
			adapter->rx_ring[i].reg_idx = i;
		for (i = 0; i < adapter->num_tx_queues; i++)
			adapter->tx_ring[i].reg_idx = i;
		ret = true;
	} else {
		ret = false;
	}

	return ret;
}

#ifdef CONFIG_IXGBE_DCB
/**
 * ixgbe_cache_ring_dcb - Descriptor ring to register mapping for DCB
 * @adapter: board private structure to initialize
 *
 * Cache the descriptor ring offsets for DCB to the assigned rings.
 *
 **/
static inline bool ixgbe_cache_ring_dcb(struct ixgbe_adapter *adapter)
{
	int i;
	bool ret = false;
	int dcb_i = adapter->ring_feature[RING_F_DCB].indices;

	if (adapter->flags & IXGBE_FLAG_DCB_ENABLED) {
		if (adapter->hw.mac.type == ixgbe_mac_82598EB) {
2837 2838 2839 2840 2841
			/* the number of queues is assumed to be symmetric */
			for (i = 0; i < dcb_i; i++) {
				adapter->rx_ring[i].reg_idx = i << 3;
				adapter->tx_ring[i].reg_idx = i << 2;
			}
2842
			ret = true;
2843 2844 2845 2846 2847 2848
		} else if (adapter->hw.mac.type == ixgbe_mac_82599EB) {
			for (i = 0; i < dcb_i; i++) {
				adapter->rx_ring[i].reg_idx = i << 4;
				adapter->tx_ring[i].reg_idx = i << 4;
			}
			ret = true;
2849 2850
		} else {
			ret = false;
2851
		}
2852 2853
	} else {
		ret = false;
2854
	}
2855 2856 2857 2858 2859 2860 2861 2862 2863 2864 2865 2866 2867 2868 2869 2870 2871 2872 2873 2874 2875 2876 2877 2878 2879 2880 2881 2882 2883

	return ret;
}
#endif

/**
 * ixgbe_cache_ring_register - Descriptor ring to register mapping
 * @adapter: board private structure to initialize
 *
 * Once we know the feature-set enabled for the device, we'll cache
 * the register offset the descriptor ring is assigned to.
 *
 * Note, the order the various feature calls is important.  It must start with
 * the "most" features enabled at the same time, then trickle down to the
 * least amount of features turned on at once.
 **/
static void ixgbe_cache_ring_register(struct ixgbe_adapter *adapter)
{
	/* start with default case */
	adapter->rx_ring[0].reg_idx = 0;
	adapter->tx_ring[0].reg_idx = 0;

#ifdef CONFIG_IXGBE_DCB
	if (ixgbe_cache_ring_dcb(adapter))
		return;

#endif
	if (ixgbe_cache_ring_rss(adapter))
		return;
2884 2885
}

2886 2887 2888 2889 2890
/**
 * ixgbe_alloc_queues - Allocate memory for all rings
 * @adapter: board private structure to initialize
 *
 * We allocate one ring per queue at run-time since we don't know the
2891 2892
 * number of queues at compile-time.  The polling_netdev array is
 * intended for Multiqueue, but should work fine with a single queue.
2893
 **/
2894
static int ixgbe_alloc_queues(struct ixgbe_adapter *adapter)
2895 2896 2897 2898
{
	int i;

	adapter->tx_ring = kcalloc(adapter->num_tx_queues,
2899
	                           sizeof(struct ixgbe_ring), GFP_KERNEL);
2900
	if (!adapter->tx_ring)
2901
		goto err_tx_ring_allocation;
2902 2903

	adapter->rx_ring = kcalloc(adapter->num_rx_queues,
2904
	                           sizeof(struct ixgbe_ring), GFP_KERNEL);
2905 2906
	if (!adapter->rx_ring)
		goto err_rx_ring_allocation;
2907

2908
	for (i = 0; i < adapter->num_tx_queues; i++) {
2909
		adapter->tx_ring[i].count = adapter->tx_ring_count;
2910 2911
		adapter->tx_ring[i].queue_index = i;
	}
2912

2913
	for (i = 0; i < adapter->num_rx_queues; i++) {
2914
		adapter->rx_ring[i].count = adapter->rx_ring_count;
2915 2916 2917 2918 2919 2920 2921 2922 2923 2924 2925 2926 2927 2928 2929 2930 2931 2932 2933 2934
		adapter->rx_ring[i].queue_index = i;
	}

	ixgbe_cache_ring_register(adapter);

	return 0;

err_rx_ring_allocation:
	kfree(adapter->tx_ring);
err_tx_ring_allocation:
	return -ENOMEM;
}

/**
 * ixgbe_set_interrupt_capability - set MSI-X or MSI if supported
 * @adapter: board private structure to initialize
 *
 * Attempt to configure the interrupts using the best available
 * capabilities of the hardware and the kernel.
 **/
A
Al Viro 已提交
2935
static int ixgbe_set_interrupt_capability(struct ixgbe_adapter *adapter)
2936
{
2937
	struct ixgbe_hw *hw = &adapter->hw;
2938 2939 2940 2941 2942 2943 2944 2945 2946 2947
	int err = 0;
	int vector, v_budget;

	/*
	 * It's easy to be greedy for MSI-X vectors, but it really
	 * doesn't do us much good if we have a lot more vectors
	 * than CPU's.  So let's be conservative and only ask for
	 * (roughly) twice the number of vectors as there are CPU's.
	 */
	v_budget = min(adapter->num_rx_queues + adapter->num_tx_queues,
2948
	               (int)(num_online_cpus() * 2)) + NON_Q_VECTORS;
2949 2950 2951

	/*
	 * At the same time, hardware can only support a maximum of
2952 2953 2954 2955
	 * hw.mac->max_msix_vectors vectors.  With features
	 * such as RSS and VMDq, we can easily surpass the number of Rx and Tx
	 * descriptor queues supported by our device.  Thus, we cap it off in
	 * those rare cases where the cpu count also exceeds our vector limit.
2956
	 */
2957
	v_budget = min(v_budget, (int)hw->mac.max_msix_vectors);
2958 2959 2960 2961

	/* A failure in MSI-X entry allocation isn't fatal, but it does
	 * mean we disable MSI-X capabilities of the adapter. */
	adapter->msix_entries = kcalloc(v_budget,
2962
	                                sizeof(struct msix_entry), GFP_KERNEL);
2963
	if (!adapter->msix_entries) {
2964
		adapter->flags &= ~IXGBE_FLAG_DCB_ENABLED;
2965 2966 2967 2968 2969 2970 2971
		adapter->flags &= ~IXGBE_FLAG_RSS_ENABLED;
		ixgbe_set_num_queues(adapter);
		kfree(adapter->tx_ring);
		kfree(adapter->rx_ring);
		err = ixgbe_alloc_queues(adapter);
		if (err) {
			DPRINTK(PROBE, ERR, "Unable to allocate memory "
2972
			        "for queues\n");
2973 2974 2975 2976 2977 2978 2979 2980 2981 2982 2983 2984 2985 2986 2987 2988 2989 2990 2991 2992
			goto out;
		}

		goto try_msi;
	}

	for (vector = 0; vector < v_budget; vector++)
		adapter->msix_entries[vector].entry = vector;

	ixgbe_acquire_msix_vectors(adapter, v_budget);

	if (adapter->flags & IXGBE_FLAG_MSIX_ENABLED)
		goto out;

try_msi:
	err = pci_enable_msi(adapter->pdev);
	if (!err) {
		adapter->flags |= IXGBE_FLAG_MSI_ENABLED;
	} else {
		DPRINTK(HW, DEBUG, "Unable to allocate MSI interrupt, "
2993
		        "falling back to legacy.  Error: %d\n", err);
2994 2995 2996 2997 2998 2999 3000 3001
		/* reset err */
		err = 0;
	}

out:
	return err;
}

3002
void ixgbe_reset_interrupt_capability(struct ixgbe_adapter *adapter)
3003 3004 3005 3006 3007 3008 3009 3010 3011 3012 3013 3014 3015 3016 3017 3018 3019 3020 3021 3022 3023 3024 3025
{
	if (adapter->flags & IXGBE_FLAG_MSIX_ENABLED) {
		adapter->flags &= ~IXGBE_FLAG_MSIX_ENABLED;
		pci_disable_msix(adapter->pdev);
		kfree(adapter->msix_entries);
		adapter->msix_entries = NULL;
	} else if (adapter->flags & IXGBE_FLAG_MSI_ENABLED) {
		adapter->flags &= ~IXGBE_FLAG_MSI_ENABLED;
		pci_disable_msi(adapter->pdev);
	}
	return;
}

/**
 * ixgbe_init_interrupt_scheme - Determine proper interrupt scheme
 * @adapter: board private structure to initialize
 *
 * We determine which interrupt scheme to use based on...
 * - Kernel support (MSI, MSI-X)
 *   - which can be user-defined (via MODULE_PARAM)
 * - Hardware queue count (num_*_queues)
 *   - defined by miscellaneous hardware support/features (RSS, etc.)
 **/
3026
int ixgbe_init_interrupt_scheme(struct ixgbe_adapter *adapter)
3027 3028 3029 3030 3031 3032 3033 3034 3035 3036 3037 3038 3039 3040 3041 3042
{
	int err;

	/* Number of supported queues */
	ixgbe_set_num_queues(adapter);

	err = ixgbe_alloc_queues(adapter);
	if (err) {
		DPRINTK(PROBE, ERR, "Unable to allocate memory for queues\n");
		goto err_alloc_queues;
	}

	err = ixgbe_set_interrupt_capability(adapter);
	if (err) {
		DPRINTK(PROBE, ERR, "Unable to setup interrupt capabilities\n");
		goto err_set_interrupt;
3043 3044
	}

3045
	DPRINTK(DRV, INFO, "Multiqueue %s: Rx Queue count = %u, "
3046 3047 3048
	        "Tx Queue count = %u\n",
	        (adapter->num_rx_queues > 1) ? "Enabled" :
	        "Disabled", adapter->num_rx_queues, adapter->num_tx_queues);
3049 3050 3051

	set_bit(__IXGBE_DOWN, &adapter->state);

3052
	return 0;
3053 3054 3055 3056 3057 3058

err_set_interrupt:
	kfree(adapter->tx_ring);
	kfree(adapter->rx_ring);
err_alloc_queues:
	return err;
3059 3060
}

D
Donald Skidmore 已提交
3061 3062 3063 3064 3065 3066 3067 3068
/**
 * ixgbe_sfp_timer - worker thread to find a missing module
 * @data: pointer to our adapter struct
 **/
static void ixgbe_sfp_timer(unsigned long data)
{
	struct ixgbe_adapter *adapter = (struct ixgbe_adapter *)data;

3069 3070
	/*
	 * Do the sfp_timer outside of interrupt context due to the
D
Donald Skidmore 已提交
3071 3072 3073 3074 3075 3076 3077 3078 3079 3080 3081 3082 3083 3084 3085 3086 3087 3088 3089 3090 3091 3092 3093 3094 3095 3096 3097 3098 3099 3100 3101 3102 3103 3104 3105 3106 3107 3108 3109 3110 3111 3112
	 * delays that sfp+ detection requires
	 */
	schedule_work(&adapter->sfp_task);
}

/**
 * ixgbe_sfp_task - worker thread to find a missing module
 * @work: pointer to work_struct containing our data
 **/
static void ixgbe_sfp_task(struct work_struct *work)
{
	struct ixgbe_adapter *adapter = container_of(work,
	                                             struct ixgbe_adapter,
	                                             sfp_task);
	struct ixgbe_hw *hw = &adapter->hw;

	if ((hw->phy.type == ixgbe_phy_nl) &&
	    (hw->phy.sfp_type == ixgbe_sfp_type_not_present)) {
		s32 ret = hw->phy.ops.identify_sfp(hw);
		if (ret)
			goto reschedule;
		ret = hw->phy.ops.reset(hw);
		if (ret == IXGBE_ERR_SFP_NOT_SUPPORTED) {
			DPRINTK(PROBE, ERR, "failed to initialize because an "
			        "unsupported SFP+ module type was detected.\n"
			        "Reload the driver after installing a "
			        "supported module.\n");
			unregister_netdev(adapter->netdev);
		} else {
			DPRINTK(PROBE, INFO, "detected SFP+: %d\n",
			        hw->phy.sfp_type);
		}
		/* don't need this routine any more */
		clear_bit(__IXGBE_SFP_MODULE_NOT_FOUND, &adapter->state);
	}
	return;
reschedule:
	if (test_bit(__IXGBE_SFP_MODULE_NOT_FOUND, &adapter->state))
		mod_timer(&adapter->sfp_timer,
		          round_jiffies(jiffies + (2 * HZ)));
}

3113 3114 3115 3116 3117 3118 3119 3120 3121 3122 3123 3124
/**
 * ixgbe_sw_init - Initialize general software structures (struct ixgbe_adapter)
 * @adapter: board private structure to initialize
 *
 * ixgbe_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 __devinit ixgbe_sw_init(struct ixgbe_adapter *adapter)
{
	struct ixgbe_hw *hw = &adapter->hw;
	struct pci_dev *pdev = adapter->pdev;
3125
	unsigned int rss;
J
Jeff Kirsher 已提交
3126
#ifdef CONFIG_IXGBE_DCB
3127 3128 3129
	int j;
	struct tc_configuration *tc;
#endif
3130

3131 3132 3133 3134 3135 3136 3137 3138
	/* PCI config space info */

	hw->vendor_id = pdev->vendor;
	hw->device_id = pdev->device;
	hw->revision_id = pdev->revision;
	hw->subsystem_vendor_id = pdev->subsystem_vendor;
	hw->subsystem_device_id = pdev->subsystem_device;

3139 3140 3141 3142
	/* Set capability flags */
	rss = min(IXGBE_MAX_RSS_INDICES, (int)num_online_cpus());
	adapter->ring_feature[RING_F_RSS].indices = rss;
	adapter->flags |= IXGBE_FLAG_RSS_ENABLED;
3143
	adapter->ring_feature[RING_F_DCB].indices = IXGBE_MAX_DCB_INDICES;
3144 3145 3146 3147
	if (hw->mac.type == ixgbe_mac_82598EB)
		adapter->max_msix_q_vectors = MAX_MSIX_Q_VECTORS_82598;
	else if (hw->mac.type == ixgbe_mac_82599EB)
		adapter->max_msix_q_vectors = MAX_MSIX_Q_VECTORS_82599;
3148

J
Jeff Kirsher 已提交
3149
#ifdef CONFIG_IXGBE_DCB
3150 3151 3152 3153 3154 3155 3156 3157 3158 3159 3160 3161 3162 3163 3164 3165 3166 3167
	/* Configure DCB traffic classes */
	for (j = 0; j < MAX_TRAFFIC_CLASS; j++) {
		tc = &adapter->dcb_cfg.tc_config[j];
		tc->path[DCB_TX_CONFIG].bwg_id = 0;
		tc->path[DCB_TX_CONFIG].bwg_percent = 12 + (j & 1);
		tc->path[DCB_RX_CONFIG].bwg_id = 0;
		tc->path[DCB_RX_CONFIG].bwg_percent = 12 + (j & 1);
		tc->dcb_pfc = pfc_disabled;
	}
	adapter->dcb_cfg.bw_percentage[DCB_TX_CONFIG][0] = 100;
	adapter->dcb_cfg.bw_percentage[DCB_RX_CONFIG][0] = 100;
	adapter->dcb_cfg.rx_pba_cfg = pba_equal;
	adapter->dcb_cfg.round_robin_enable = false;
	adapter->dcb_set_bitmap = 0x00;
	ixgbe_copy_dcb_cfg(&adapter->dcb_cfg, &adapter->temp_dcb_cfg,
	                   adapter->ring_feature[RING_F_DCB].indices);

#endif
3168 3169

	/* default flow control settings */
3170
	hw->fc.requested_mode = ixgbe_fc_full;
D
Don Skidmore 已提交
3171
	hw->fc.current_mode = ixgbe_fc_full;	/* init for ethtool output */
3172 3173 3174 3175
	hw->fc.high_water = IXGBE_DEFAULT_FCRTH;
	hw->fc.low_water = IXGBE_DEFAULT_FCRTL;
	hw->fc.pause_time = IXGBE_DEFAULT_FCPAUSE;
	hw->fc.send_xon = true;
D
Don Skidmore 已提交
3176
	hw->fc.disable_fc_autoneg = false;
3177

3178 3179 3180 3181 3182 3183 3184 3185 3186 3187 3188 3189
	/* enable itr by default in dynamic mode */
	adapter->itr_setting = 1;
	adapter->eitr_param = 20000;

	/* set defaults for eitr in MegaBytes */
	adapter->eitr_low = 10;
	adapter->eitr_high = 20;

	/* set default ring sizes */
	adapter->tx_ring_count = IXGBE_DEFAULT_TXD;
	adapter->rx_ring_count = IXGBE_DEFAULT_RXD;

3190
	/* initialize eeprom parameters */
3191
	if (ixgbe_init_eeprom_params_generic(hw)) {
3192 3193 3194 3195
		dev_err(&pdev->dev, "EEPROM initialization failed\n");
		return -EIO;
	}

3196
	/* enable rx csum by default */
3197 3198 3199 3200 3201 3202 3203 3204 3205 3206
	adapter->flags |= IXGBE_FLAG_RX_CSUM_ENABLED;

	set_bit(__IXGBE_DOWN, &adapter->state);

	return 0;
}

/**
 * ixgbe_setup_tx_resources - allocate Tx resources (Descriptors)
 * @adapter: board private structure
3207
 * @tx_ring:    tx descriptor ring (for a specific queue) to setup
3208 3209 3210 3211
 *
 * Return 0 on success, negative on failure
 **/
int ixgbe_setup_tx_resources(struct ixgbe_adapter *adapter,
3212
                             struct ixgbe_ring *tx_ring)
3213 3214 3215 3216
{
	struct pci_dev *pdev = adapter->pdev;
	int size;

3217 3218
	size = sizeof(struct ixgbe_tx_buffer) * tx_ring->count;
	tx_ring->tx_buffer_info = vmalloc(size);
3219 3220
	if (!tx_ring->tx_buffer_info)
		goto err;
3221
	memset(tx_ring->tx_buffer_info, 0, size);
3222 3223

	/* round up to nearest 4K */
3224
	tx_ring->size = tx_ring->count * sizeof(union ixgbe_adv_tx_desc);
3225
	tx_ring->size = ALIGN(tx_ring->size, 4096);
3226

3227 3228
	tx_ring->desc = pci_alloc_consistent(pdev, tx_ring->size,
	                                     &tx_ring->dma);
3229 3230
	if (!tx_ring->desc)
		goto err;
3231

3232 3233 3234
	tx_ring->next_to_use = 0;
	tx_ring->next_to_clean = 0;
	tx_ring->work_limit = tx_ring->count;
3235
	return 0;
3236 3237 3238 3239 3240 3241 3242

err:
	vfree(tx_ring->tx_buffer_info);
	tx_ring->tx_buffer_info = NULL;
	DPRINTK(PROBE, ERR, "Unable to allocate memory for the transmit "
	                    "descriptor ring\n");
	return -ENOMEM;
3243 3244
}

3245 3246 3247 3248 3249 3250 3251 3252 3253 3254 3255 3256 3257 3258 3259 3260 3261 3262 3263 3264 3265 3266 3267 3268 3269
/**
 * ixgbe_setup_all_tx_resources - allocate all queues Tx resources
 * @adapter: board private structure
 *
 * 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 ixgbe_setup_all_tx_resources(struct ixgbe_adapter *adapter)
{
	int i, err = 0;

	for (i = 0; i < adapter->num_tx_queues; i++) {
		err = ixgbe_setup_tx_resources(adapter, &adapter->tx_ring[i]);
		if (!err)
			continue;
		DPRINTK(PROBE, ERR, "Allocation for Tx Queue %u failed\n", i);
		break;
	}

	return err;
}

3270 3271 3272
/**
 * ixgbe_setup_rx_resources - allocate Rx resources (Descriptors)
 * @adapter: board private structure
3273
 * @rx_ring:    rx descriptor ring (for a specific queue) to setup
3274 3275 3276 3277
 *
 * Returns 0 on success, negative on failure
 **/
int ixgbe_setup_rx_resources(struct ixgbe_adapter *adapter,
3278
                             struct ixgbe_ring *rx_ring)
3279 3280
{
	struct pci_dev *pdev = adapter->pdev;
3281
	int size;
3282

3283 3284 3285
	size = sizeof(struct ixgbe_rx_buffer) * rx_ring->count;
	rx_ring->rx_buffer_info = vmalloc(size);
	if (!rx_ring->rx_buffer_info) {
3286
		DPRINTK(PROBE, ERR,
3287
		        "vmalloc allocation failed for the rx desc ring\n");
3288
		goto alloc_failed;
3289
	}
3290
	memset(rx_ring->rx_buffer_info, 0, size);
3291 3292

	/* Round up to nearest 4K */
3293 3294
	rx_ring->size = rx_ring->count * sizeof(union ixgbe_adv_rx_desc);
	rx_ring->size = ALIGN(rx_ring->size, 4096);
3295

3296
	rx_ring->desc = pci_alloc_consistent(pdev, rx_ring->size, &rx_ring->dma);
3297

3298
	if (!rx_ring->desc) {
3299
		DPRINTK(PROBE, ERR,
3300
		        "Memory allocation failed for the rx desc ring\n");
3301
		vfree(rx_ring->rx_buffer_info);
3302
		goto alloc_failed;
3303 3304
	}

3305 3306
	rx_ring->next_to_clean = 0;
	rx_ring->next_to_use = 0;
3307 3308

	return 0;
3309 3310 3311

alloc_failed:
	return -ENOMEM;
3312 3313
}

3314 3315 3316 3317 3318 3319 3320 3321 3322 3323 3324 3325 3326 3327 3328 3329 3330 3331 3332 3333 3334 3335 3336 3337 3338 3339
/**
 * ixgbe_setup_all_rx_resources - allocate all queues Rx resources
 * @adapter: board private structure
 *
 * 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 ixgbe_setup_all_rx_resources(struct ixgbe_adapter *adapter)
{
	int i, err = 0;

	for (i = 0; i < adapter->num_rx_queues; i++) {
		err = ixgbe_setup_rx_resources(adapter, &adapter->rx_ring[i]);
		if (!err)
			continue;
		DPRINTK(PROBE, ERR, "Allocation for Rx Queue %u failed\n", i);
		break;
	}

	return err;
}

3340 3341 3342 3343 3344 3345 3346
/**
 * ixgbe_free_tx_resources - Free Tx Resources per Queue
 * @adapter: board private structure
 * @tx_ring: Tx descriptor ring for a specific queue
 *
 * Free all transmit software resources
 **/
3347 3348
void ixgbe_free_tx_resources(struct ixgbe_adapter *adapter,
                             struct ixgbe_ring *tx_ring)
3349 3350 3351 3352 3353 3354 3355 3356 3357 3358 3359 3360 3361 3362 3363 3364 3365 3366 3367 3368 3369 3370 3371 3372
{
	struct pci_dev *pdev = adapter->pdev;

	ixgbe_clean_tx_ring(adapter, tx_ring);

	vfree(tx_ring->tx_buffer_info);
	tx_ring->tx_buffer_info = NULL;

	pci_free_consistent(pdev, tx_ring->size, tx_ring->desc, tx_ring->dma);

	tx_ring->desc = NULL;
}

/**
 * ixgbe_free_all_tx_resources - Free Tx Resources for All Queues
 * @adapter: board private structure
 *
 * Free all transmit software resources
 **/
static void ixgbe_free_all_tx_resources(struct ixgbe_adapter *adapter)
{
	int i;

	for (i = 0; i < adapter->num_tx_queues; i++)
3373 3374
		if (adapter->tx_ring[i].desc)
			ixgbe_free_tx_resources(adapter, &adapter->tx_ring[i]);
3375 3376 3377
}

/**
3378
 * ixgbe_free_rx_resources - Free Rx Resources
3379 3380 3381 3382 3383
 * @adapter: board private structure
 * @rx_ring: ring to clean the resources from
 *
 * Free all receive software resources
 **/
3384 3385
void ixgbe_free_rx_resources(struct ixgbe_adapter *adapter,
                             struct ixgbe_ring *rx_ring)
3386 3387 3388 3389 3390 3391 3392 3393 3394 3395 3396 3397 3398 3399 3400 3401 3402 3403 3404 3405 3406 3407 3408 3409
{
	struct pci_dev *pdev = adapter->pdev;

	ixgbe_clean_rx_ring(adapter, rx_ring);

	vfree(rx_ring->rx_buffer_info);
	rx_ring->rx_buffer_info = NULL;

	pci_free_consistent(pdev, rx_ring->size, rx_ring->desc, rx_ring->dma);

	rx_ring->desc = NULL;
}

/**
 * ixgbe_free_all_rx_resources - Free Rx Resources for All Queues
 * @adapter: board private structure
 *
 * Free all receive software resources
 **/
static void ixgbe_free_all_rx_resources(struct ixgbe_adapter *adapter)
{
	int i;

	for (i = 0; i < adapter->num_rx_queues; i++)
3410 3411
		if (adapter->rx_ring[i].desc)
			ixgbe_free_rx_resources(adapter, &adapter->rx_ring[i]);
3412 3413 3414 3415 3416 3417 3418 3419 3420 3421 3422 3423 3424 3425
}

/**
 * ixgbe_change_mtu - 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 ixgbe_change_mtu(struct net_device *netdev, int new_mtu)
{
	struct ixgbe_adapter *adapter = netdev_priv(netdev);
	int max_frame = new_mtu + ETH_HLEN + ETH_FCS_LEN;

3426 3427
	/* MTU < 68 is an error and causes problems on some kernels */
	if ((new_mtu < 68) || (max_frame > IXGBE_MAX_JUMBO_FRAME_SIZE))
3428 3429
		return -EINVAL;

3430
	DPRINTK(PROBE, INFO, "changing MTU from %d to %d\n",
3431
	        netdev->mtu, new_mtu);
3432
	/* must set new MTU before calling down or up */
3433 3434
	netdev->mtu = new_mtu;

3435 3436
	if (netif_running(netdev))
		ixgbe_reinit_locked(adapter);
3437 3438 3439 3440 3441 3442 3443 3444 3445 3446 3447 3448 3449 3450 3451 3452 3453 3454 3455 3456

	return 0;
}

/**
 * ixgbe_open - Called when a network interface is made active
 * @netdev: network interface device structure
 *
 * Returns 0 on success, negative value on failure
 *
 * 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 watchdog timer is started,
 * and the stack is notified that the interface is ready.
 **/
static int ixgbe_open(struct net_device *netdev)
{
	struct ixgbe_adapter *adapter = netdev_priv(netdev);
	int err;
3457 3458 3459 3460

	/* disallow open during test */
	if (test_bit(__IXGBE_TESTING, &adapter->state))
		return -EBUSY;
3461 3462 3463 3464 3465 3466 3467 3468 3469 3470 3471 3472 3473

	/* allocate transmit descriptors */
	err = ixgbe_setup_all_tx_resources(adapter);
	if (err)
		goto err_setup_tx;

	/* allocate receive descriptors */
	err = ixgbe_setup_all_rx_resources(adapter);
	if (err)
		goto err_setup_rx;

	ixgbe_configure(adapter);

3474 3475
	ixgbe_napi_add_all(adapter);

3476
	err = ixgbe_request_irq(adapter);
3477 3478 3479 3480 3481 3482 3483
	if (err)
		goto err_req_irq;

	err = ixgbe_up_complete(adapter);
	if (err)
		goto err_up;

3484 3485
	netif_tx_start_all_queues(netdev);

3486 3487 3488
	return 0;

err_up:
3489
	ixgbe_release_hw_control(adapter);
3490 3491 3492
	ixgbe_free_irq(adapter);
err_req_irq:
err_setup_rx:
3493
	ixgbe_free_all_rx_resources(adapter);
3494
err_setup_tx:
3495
	ixgbe_free_all_tx_resources(adapter);
3496 3497 3498 3499 3500 3501 3502 3503 3504 3505 3506 3507 3508 3509 3510 3511 3512 3513 3514 3515 3516 3517 3518 3519 3520 3521
	ixgbe_reset(adapter);

	return err;
}

/**
 * ixgbe_close - Disables a network interface
 * @netdev: network interface device structure
 *
 * Returns 0, this is not allowed to fail
 *
 * The close entry point is called when an interface is de-activated
 * by the OS.  The hardware is still under the drivers control, but
 * needs to be disabled.  A global MAC reset is issued to stop the
 * hardware, and all transmit and receive resources are freed.
 **/
static int ixgbe_close(struct net_device *netdev)
{
	struct ixgbe_adapter *adapter = netdev_priv(netdev);

	ixgbe_down(adapter);
	ixgbe_free_irq(adapter);

	ixgbe_free_all_tx_resources(adapter);
	ixgbe_free_all_rx_resources(adapter);

3522
	ixgbe_release_hw_control(adapter);
3523 3524 3525 3526

	return 0;
}

3527 3528 3529
/**
 * ixgbe_napi_add_all - prep napi structs for use
 * @adapter: private struct
3530
 *
3531 3532
 * helper function to napi_add each possible q_vector->napi
 */
3533
void ixgbe_napi_add_all(struct ixgbe_adapter *adapter)
3534 3535
{
	int q_idx, q_vectors;
3536
	struct net_device *netdev = adapter->netdev;
3537 3538
	int (*poll)(struct napi_struct *, int);

3539 3540 3541 3542
	/* check if we already have our netdev->napi_list populated */
	if (&netdev->napi_list != netdev->napi_list.next)
		return;

3543 3544 3545 3546 3547 3548 3549 3550 3551 3552 3553 3554 3555 3556 3557 3558
	if (adapter->flags & IXGBE_FLAG_MSIX_ENABLED) {
		poll = &ixgbe_clean_rxonly;
		/* Only enable as many vectors as we have rx queues. */
		q_vectors = adapter->num_rx_queues;
	} else {
		poll = &ixgbe_poll;
		/* only one q_vector for legacy modes */
		q_vectors = 1;
	}

	for (q_idx = 0; q_idx < q_vectors; q_idx++) {
		struct ixgbe_q_vector *q_vector = &adapter->q_vector[q_idx];
		netif_napi_add(adapter->netdev, &q_vector->napi, (*poll), 64);
	}
}

3559
void ixgbe_napi_del_all(struct ixgbe_adapter *adapter)
3560 3561 3562 3563 3564 3565 3566 3567 3568 3569 3570 3571 3572 3573 3574 3575 3576 3577 3578 3579 3580 3581 3582 3583 3584 3585 3586
{
	int q_idx;
	int q_vectors = adapter->num_msix_vectors - NON_Q_VECTORS;

	/* legacy and MSI only use one vector */
	if (!(adapter->flags & IXGBE_FLAG_MSIX_ENABLED))
		q_vectors = 1;

	for (q_idx = 0; q_idx < q_vectors; q_idx++) {
		struct ixgbe_q_vector *q_vector = &adapter->q_vector[q_idx];
		if (!q_vector->rxr_count)
			continue;
		netif_napi_del(&q_vector->napi);
	}
}

#ifdef CONFIG_PM
static int ixgbe_resume(struct pci_dev *pdev)
{
	struct net_device *netdev = pci_get_drvdata(pdev);
	struct ixgbe_adapter *adapter = netdev_priv(netdev);
	u32 err;

	pci_set_power_state(pdev, PCI_D0);
	pci_restore_state(pdev);
	err = pci_enable_device(pdev);
	if (err) {
3587
		printk(KERN_ERR "ixgbe: Cannot enable PCI device from "
3588 3589 3590 3591 3592 3593 3594 3595 3596 3597 3598 3599 3600 3601 3602 3603 3604 3605 3606 3607 3608 3609 3610 3611 3612 3613 3614 3615
				"suspend\n");
		return err;
	}
	pci_set_master(pdev);

	pci_enable_wake(pdev, PCI_D3hot, 0);
	pci_enable_wake(pdev, PCI_D3cold, 0);

	err = ixgbe_init_interrupt_scheme(adapter);
	if (err) {
		printk(KERN_ERR "ixgbe: Cannot initialize interrupts for "
		                "device\n");
		return err;
	}

	ixgbe_reset(adapter);

	if (netif_running(netdev)) {
		err = ixgbe_open(adapter->netdev);
		if (err)
			return err;
	}

	netif_device_attach(netdev);

	return 0;
}
#endif /* CONFIG_PM */
3616 3617

static int __ixgbe_shutdown(struct pci_dev *pdev, bool *enable_wake)
3618 3619 3620
{
	struct net_device *netdev = pci_get_drvdata(pdev);
	struct ixgbe_adapter *adapter = netdev_priv(netdev);
3621 3622 3623
	struct ixgbe_hw *hw = &adapter->hw;
	u32 ctrl, fctrl;
	u32 wufc = adapter->wol;
3624 3625 3626 3627 3628 3629 3630 3631 3632 3633 3634 3635 3636 3637
#ifdef CONFIG_PM
	int retval = 0;
#endif

	netif_device_detach(netdev);

	if (netif_running(netdev)) {
		ixgbe_down(adapter);
		ixgbe_free_irq(adapter);
		ixgbe_free_all_tx_resources(adapter);
		ixgbe_free_all_rx_resources(adapter);
	}
	ixgbe_reset_interrupt_capability(adapter);
	ixgbe_napi_del_all(adapter);
3638
	INIT_LIST_HEAD(&netdev->napi_list);
3639 3640 3641 3642 3643 3644 3645
	kfree(adapter->tx_ring);
	kfree(adapter->rx_ring);

#ifdef CONFIG_PM
	retval = pci_save_state(pdev);
	if (retval)
		return retval;
3646

3647
#endif
3648 3649
	if (wufc) {
		ixgbe_set_rx_mode(netdev);
3650

3651 3652 3653 3654 3655 3656 3657 3658 3659 3660 3661 3662 3663 3664 3665 3666 3667 3668 3669 3670 3671 3672 3673 3674
		/* turn on all-multi mode if wake on multicast is enabled */
		if (wufc & IXGBE_WUFC_MC) {
			fctrl = IXGBE_READ_REG(hw, IXGBE_FCTRL);
			fctrl |= IXGBE_FCTRL_MPE;
			IXGBE_WRITE_REG(hw, IXGBE_FCTRL, fctrl);
		}

		ctrl = IXGBE_READ_REG(hw, IXGBE_CTRL);
		ctrl |= IXGBE_CTRL_GIO_DIS;
		IXGBE_WRITE_REG(hw, IXGBE_CTRL, ctrl);

		IXGBE_WRITE_REG(hw, IXGBE_WUFC, wufc);
	} else {
		IXGBE_WRITE_REG(hw, IXGBE_WUC, 0);
		IXGBE_WRITE_REG(hw, IXGBE_WUFC, 0);
	}

	if (wufc && hw->mac.type == ixgbe_mac_82599EB) {
		pci_enable_wake(pdev, PCI_D3hot, 1);
		pci_enable_wake(pdev, PCI_D3cold, 1);
	} else {
		pci_enable_wake(pdev, PCI_D3hot, 0);
		pci_enable_wake(pdev, PCI_D3cold, 0);
	}
3675

3676 3677
	*enable_wake = !!wufc;

3678 3679 3680 3681
	ixgbe_release_hw_control(adapter);

	pci_disable_device(pdev);

3682 3683 3684 3685 3686 3687 3688 3689 3690 3691 3692 3693 3694 3695 3696 3697 3698 3699 3700
	return 0;
}

#ifdef CONFIG_PM
static int ixgbe_suspend(struct pci_dev *pdev, pm_message_t state)
{
	int retval;
	bool wake;

	retval = __ixgbe_shutdown(pdev, &wake);
	if (retval)
		return retval;

	if (wake) {
		pci_prepare_to_sleep(pdev);
	} else {
		pci_wake_from_d3(pdev, false);
		pci_set_power_state(pdev, PCI_D3hot);
	}
3701 3702 3703

	return 0;
}
3704
#endif /* CONFIG_PM */
3705 3706 3707

static void ixgbe_shutdown(struct pci_dev *pdev)
{
3708 3709 3710 3711 3712 3713 3714 3715
	bool wake;

	__ixgbe_shutdown(pdev, &wake);

	if (system_state == SYSTEM_POWER_OFF) {
		pci_wake_from_d3(pdev, wake);
		pci_set_power_state(pdev, PCI_D3hot);
	}
3716 3717
}

3718 3719 3720 3721 3722 3723 3724
/**
 * ixgbe_update_stats - Update the board statistics counters.
 * @adapter: board private structure
 **/
void ixgbe_update_stats(struct ixgbe_adapter *adapter)
{
	struct ixgbe_hw *hw = &adapter->hw;
3725 3726
	u64 total_mpc = 0;
	u32 i, missed_rx = 0, mpc, bprc, lxon, lxoff, xon_off_tot;
3727

3728 3729 3730 3731 3732 3733
	if (hw->mac.type == ixgbe_mac_82599EB) {
		for (i = 0; i < 16; i++)
			adapter->hw_rx_no_dma_resources +=
			                     IXGBE_READ_REG(hw, IXGBE_QPRDC(i));
	}

3734
	adapter->stats.crcerrs += IXGBE_READ_REG(hw, IXGBE_CRCERRS);
3735 3736 3737 3738 3739 3740
	for (i = 0; i < 8; i++) {
		/* for packet buffers not used, the register should read 0 */
		mpc = IXGBE_READ_REG(hw, IXGBE_MPC(i));
		missed_rx += mpc;
		adapter->stats.mpc[i] += mpc;
		total_mpc += adapter->stats.mpc[i];
3741 3742
		if (hw->mac.type == ixgbe_mac_82598EB)
			adapter->stats.rnbc[i] += IXGBE_READ_REG(hw, IXGBE_RNBC(i));
3743 3744 3745 3746
		adapter->stats.qptc[i] += IXGBE_READ_REG(hw, IXGBE_QPTC(i));
		adapter->stats.qbtc[i] += IXGBE_READ_REG(hw, IXGBE_QBTC(i));
		adapter->stats.qprc[i] += IXGBE_READ_REG(hw, IXGBE_QPRC(i));
		adapter->stats.qbrc[i] += IXGBE_READ_REG(hw, IXGBE_QBRC(i));
3747 3748 3749 3750 3751 3752 3753 3754 3755 3756 3757 3758
		if (hw->mac.type == ixgbe_mac_82599EB) {
			adapter->stats.pxonrxc[i] += IXGBE_READ_REG(hw,
			                                    IXGBE_PXONRXCNT(i));
			adapter->stats.pxoffrxc[i] += IXGBE_READ_REG(hw,
			                                   IXGBE_PXOFFRXCNT(i));
			adapter->stats.qprdc[i] += IXGBE_READ_REG(hw, IXGBE_QPRDC(i));
		} else {
			adapter->stats.pxonrxc[i] += IXGBE_READ_REG(hw,
			                                      IXGBE_PXONRXC(i));
			adapter->stats.pxoffrxc[i] += IXGBE_READ_REG(hw,
			                                     IXGBE_PXOFFRXC(i));
		}
3759 3760 3761
		adapter->stats.pxontxc[i] += IXGBE_READ_REG(hw,
		                                            IXGBE_PXONTXC(i));
		adapter->stats.pxofftxc[i] += IXGBE_READ_REG(hw,
3762
		                                             IXGBE_PXOFFTXC(i));
3763 3764 3765 3766 3767 3768
	}
	adapter->stats.gprc += IXGBE_READ_REG(hw, IXGBE_GPRC);
	/* work around hardware counting issue */
	adapter->stats.gprc -= missed_rx;

	/* 82598 hardware only has a 32 bit counter in the high register */
3769 3770 3771 3772 3773 3774 3775 3776 3777 3778 3779 3780 3781 3782 3783 3784
	if (hw->mac.type == ixgbe_mac_82599EB) {
		adapter->stats.gorc += IXGBE_READ_REG(hw, IXGBE_GORCL);
		IXGBE_READ_REG(hw, IXGBE_GORCH); /* to clear */
		adapter->stats.gotc += IXGBE_READ_REG(hw, IXGBE_GOTCL);
		IXGBE_READ_REG(hw, IXGBE_GOTCH); /* to clear */
		adapter->stats.tor += IXGBE_READ_REG(hw, IXGBE_TORL);
		IXGBE_READ_REG(hw, IXGBE_TORH); /* to clear */
		adapter->stats.lxonrxc += IXGBE_READ_REG(hw, IXGBE_LXONRXCNT);
		adapter->stats.lxoffrxc += IXGBE_READ_REG(hw, IXGBE_LXOFFRXCNT);
	} else {
		adapter->stats.lxonrxc += IXGBE_READ_REG(hw, IXGBE_LXONRXC);
		adapter->stats.lxoffrxc += IXGBE_READ_REG(hw, IXGBE_LXOFFRXC);
		adapter->stats.gorc += IXGBE_READ_REG(hw, IXGBE_GORCH);
		adapter->stats.gotc += IXGBE_READ_REG(hw, IXGBE_GOTCH);
		adapter->stats.tor += IXGBE_READ_REG(hw, IXGBE_TORH);
	}
3785 3786 3787
	bprc = IXGBE_READ_REG(hw, IXGBE_BPRC);
	adapter->stats.bprc += bprc;
	adapter->stats.mprc += IXGBE_READ_REG(hw, IXGBE_MPRC);
3788 3789
	if (hw->mac.type == ixgbe_mac_82598EB)
		adapter->stats.mprc -= bprc;
3790 3791 3792 3793 3794 3795 3796 3797
	adapter->stats.roc += IXGBE_READ_REG(hw, IXGBE_ROC);
	adapter->stats.prc64 += IXGBE_READ_REG(hw, IXGBE_PRC64);
	adapter->stats.prc127 += IXGBE_READ_REG(hw, IXGBE_PRC127);
	adapter->stats.prc255 += IXGBE_READ_REG(hw, IXGBE_PRC255);
	adapter->stats.prc511 += IXGBE_READ_REG(hw, IXGBE_PRC511);
	adapter->stats.prc1023 += IXGBE_READ_REG(hw, IXGBE_PRC1023);
	adapter->stats.prc1522 += IXGBE_READ_REG(hw, IXGBE_PRC1522);
	adapter->stats.rlec += IXGBE_READ_REG(hw, IXGBE_RLEC);
3798 3799 3800 3801
	lxon = IXGBE_READ_REG(hw, IXGBE_LXONTXC);
	adapter->stats.lxontxc += lxon;
	lxoff = IXGBE_READ_REG(hw, IXGBE_LXOFFTXC);
	adapter->stats.lxofftxc += lxoff;
3802 3803
	adapter->stats.ruc += IXGBE_READ_REG(hw, IXGBE_RUC);
	adapter->stats.gptc += IXGBE_READ_REG(hw, IXGBE_GPTC);
3804 3805 3806 3807 3808 3809 3810 3811
	adapter->stats.mptc += IXGBE_READ_REG(hw, IXGBE_MPTC);
	/*
	 * 82598 errata - tx of flow control packets is included in tx counters
	 */
	xon_off_tot = lxon + lxoff;
	adapter->stats.gptc -= xon_off_tot;
	adapter->stats.mptc -= xon_off_tot;
	adapter->stats.gotc -= (xon_off_tot * (ETH_ZLEN + ETH_FCS_LEN));
3812 3813 3814 3815 3816
	adapter->stats.ruc += IXGBE_READ_REG(hw, IXGBE_RUC);
	adapter->stats.rfc += IXGBE_READ_REG(hw, IXGBE_RFC);
	adapter->stats.rjc += IXGBE_READ_REG(hw, IXGBE_RJC);
	adapter->stats.tpr += IXGBE_READ_REG(hw, IXGBE_TPR);
	adapter->stats.ptc64 += IXGBE_READ_REG(hw, IXGBE_PTC64);
3817
	adapter->stats.ptc64 -= xon_off_tot;
3818 3819 3820 3821 3822 3823 3824 3825 3826 3827 3828 3829
	adapter->stats.ptc127 += IXGBE_READ_REG(hw, IXGBE_PTC127);
	adapter->stats.ptc255 += IXGBE_READ_REG(hw, IXGBE_PTC255);
	adapter->stats.ptc511 += IXGBE_READ_REG(hw, IXGBE_PTC511);
	adapter->stats.ptc1023 += IXGBE_READ_REG(hw, IXGBE_PTC1023);
	adapter->stats.ptc1522 += IXGBE_READ_REG(hw, IXGBE_PTC1522);
	adapter->stats.bptc += IXGBE_READ_REG(hw, IXGBE_BPTC);

	/* Fill out the OS statistics structure */
	adapter->net_stats.multicast = adapter->stats.mprc;

	/* Rx Errors */
	adapter->net_stats.rx_errors = adapter->stats.crcerrs +
3830
	                               adapter->stats.rlec;
3831 3832 3833
	adapter->net_stats.rx_dropped = 0;
	adapter->net_stats.rx_length_errors = adapter->stats.rlec;
	adapter->net_stats.rx_crc_errors = adapter->stats.crcerrs;
3834
	adapter->net_stats.rx_missed_errors = total_mpc;
3835 3836 3837 3838 3839 3840 3841 3842 3843
}

/**
 * ixgbe_watchdog - Timer Call-back
 * @data: pointer to adapter cast into an unsigned long
 **/
static void ixgbe_watchdog(unsigned long data)
{
	struct ixgbe_adapter *adapter = (struct ixgbe_adapter *)data;
3844 3845 3846 3847 3848
	struct ixgbe_hw *hw = &adapter->hw;

	/* Do the watchdog outside of interrupt context due to the lovely
	 * delays that some of the newer hardware requires */
	if (!test_bit(__IXGBE_DOWN, &adapter->state)) {
3849 3850 3851 3852 3853 3854
		u64 eics = 0;
		int i;

		for (i = 0; i < adapter->num_msix_vectors - NON_Q_VECTORS; i++)
			eics |= (1 << i);

3855
		/* Cause software interrupt to ensure rx rings are cleaned */
3856 3857 3858 3859 3860 3861 3862 3863 3864 3865 3866 3867 3868 3869 3870 3871 3872 3873 3874 3875 3876 3877 3878 3879 3880 3881 3882 3883 3884 3885 3886 3887 3888 3889 3890 3891 3892 3893 3894 3895 3896
		switch (hw->mac.type) {
		case ixgbe_mac_82598EB:
			if (adapter->flags & IXGBE_FLAG_MSIX_ENABLED) {
				IXGBE_WRITE_REG(hw, IXGBE_EICS, (u32)eics);
			} else {
				/*
				 * for legacy and MSI interrupts don't set any
				 * bits that are enabled for EIAM, because this
				 * operation would set *both* EIMS and EICS for
				 * any bit in EIAM
				 */
				IXGBE_WRITE_REG(hw, IXGBE_EICS,
				     (IXGBE_EICS_TCP_TIMER | IXGBE_EICS_OTHER));
			}
			break;
		case ixgbe_mac_82599EB:
			if (adapter->flags & IXGBE_FLAG_MSIX_ENABLED) {
				/*
				 * EICS(0..15) first 0-15 q vectors
				 * EICS[1] (16..31) q vectors 16-31
				 * EICS[2] (0..31) q vectors 32-63
				 */
				IXGBE_WRITE_REG(hw, IXGBE_EICS,
				                (u32)(eics & 0xFFFF));
				IXGBE_WRITE_REG(hw, IXGBE_EICS_EX(1),
				                (u32)(eics & 0xFFFF0000));
				IXGBE_WRITE_REG(hw, IXGBE_EICS_EX(2),
				                (u32)(eics >> 32));
			} else {
				/*
				 * for legacy and MSI interrupts don't set any
				 * bits that are enabled for EIAM, because this
				 * operation would set *both* EIMS and EICS for
				 * any bit in EIAM
				 */
				IXGBE_WRITE_REG(hw, IXGBE_EICS,
				     (IXGBE_EICS_TCP_TIMER | IXGBE_EICS_OTHER));
			}
			break;
		default:
			break;
3897 3898 3899 3900 3901
		}
		/* Reset the timer */
		mod_timer(&adapter->watchdog_timer,
		          round_jiffies(jiffies + 2 * HZ));
	}
3902

3903 3904 3905
	schedule_work(&adapter->watchdog_task);
}

3906 3907 3908 3909 3910 3911 3912 3913 3914 3915 3916 3917 3918 3919 3920 3921 3922 3923 3924 3925 3926 3927 3928 3929 3930 3931 3932 3933 3934 3935 3936 3937 3938 3939 3940 3941 3942 3943 3944 3945 3946 3947 3948
/**
 * ixgbe_multispeed_fiber_task - worker thread to configure multispeed fiber
 * @work: pointer to work_struct containing our data
 **/
static void ixgbe_multispeed_fiber_task(struct work_struct *work)
{
	struct ixgbe_adapter *adapter = container_of(work,
	                                             struct ixgbe_adapter,
	                                             multispeed_fiber_task);
	struct ixgbe_hw *hw = &adapter->hw;
	u32 autoneg;

	adapter->flags |= IXGBE_FLAG_IN_SFP_LINK_TASK;
	if (hw->mac.ops.get_link_capabilities)
		hw->mac.ops.get_link_capabilities(hw, &autoneg,
		                                  &hw->mac.autoneg);
	if (hw->mac.ops.setup_link_speed)
		hw->mac.ops.setup_link_speed(hw, autoneg, true, true);
	adapter->flags |= IXGBE_FLAG_NEED_LINK_UPDATE;
	adapter->flags &= ~IXGBE_FLAG_IN_SFP_LINK_TASK;
}

/**
 * ixgbe_sfp_config_module_task - worker thread to configure a new SFP+ module
 * @work: pointer to work_struct containing our data
 **/
static void ixgbe_sfp_config_module_task(struct work_struct *work)
{
	struct ixgbe_adapter *adapter = container_of(work,
	                                             struct ixgbe_adapter,
	                                             sfp_config_module_task);
	struct ixgbe_hw *hw = &adapter->hw;
	u32 err;

	adapter->flags |= IXGBE_FLAG_IN_SFP_MOD_TASK;
	err = hw->phy.ops.identify_sfp(hw);
	if (err == IXGBE_ERR_SFP_NOT_SUPPORTED) {
		DPRINTK(PROBE, ERR, "PHY not supported on this NIC %d\n", err);
		ixgbe_down(adapter);
		return;
	}
	hw->mac.ops.setup_sfp(hw);

3949
	if (!(adapter->flags & IXGBE_FLAG_IN_SFP_LINK_TASK))
3950 3951 3952 3953 3954
		/* This will also work for DA Twinax connections */
		schedule_work(&adapter->multispeed_fiber_task);
	adapter->flags &= ~IXGBE_FLAG_IN_SFP_MOD_TASK;
}

3955
/**
3956 3957
 * ixgbe_watchdog_task - worker thread to bring link up
 * @work: pointer to work_struct containing our data
3958 3959 3960 3961 3962 3963 3964 3965 3966 3967 3968 3969 3970 3971 3972 3973 3974 3975 3976 3977 3978 3979 3980 3981
 **/
static void ixgbe_watchdog_task(struct work_struct *work)
{
	struct ixgbe_adapter *adapter = container_of(work,
	                                             struct ixgbe_adapter,
	                                             watchdog_task);
	struct net_device *netdev = adapter->netdev;
	struct ixgbe_hw *hw = &adapter->hw;
	u32 link_speed = adapter->link_speed;
	bool link_up = adapter->link_up;

	adapter->flags |= IXGBE_FLAG_IN_WATCHDOG_TASK;

	if (adapter->flags & IXGBE_FLAG_NEED_LINK_UPDATE) {
		hw->mac.ops.check_link(hw, &link_speed, &link_up, false);
		if (link_up ||
		    time_after(jiffies, (adapter->link_check_timeout +
		                         IXGBE_TRY_LINK_TIMEOUT))) {
			IXGBE_WRITE_REG(hw, IXGBE_EIMS, IXGBE_EIMC_LSC);
			adapter->flags &= ~IXGBE_FLAG_NEED_LINK_UPDATE;
		}
		adapter->link_up = link_up;
		adapter->link_speed = link_speed;
	}
3982 3983 3984

	if (link_up) {
		if (!netif_carrier_ok(netdev)) {
3985 3986 3987 3988 3989 3990 3991 3992 3993 3994 3995 3996 3997 3998
			bool flow_rx, flow_tx;

			if (hw->mac.type == ixgbe_mac_82599EB) {
				u32 mflcn = IXGBE_READ_REG(hw, IXGBE_MFLCN);
				u32 fccfg = IXGBE_READ_REG(hw, IXGBE_FCCFG);
				flow_rx = (mflcn & IXGBE_MFLCN_RFCE);
				flow_tx = (fccfg & IXGBE_FCCFG_TFCE_802_3X);
			} else {
				u32 frctl = IXGBE_READ_REG(hw, IXGBE_FCTRL);
				u32 rmcs = IXGBE_READ_REG(hw, IXGBE_RMCS);
				flow_rx = (frctl & IXGBE_FCTRL_RFCE);
				flow_tx = (rmcs & IXGBE_RMCS_TFCE_802_3X);
			}

3999 4000 4001 4002 4003 4004 4005
			printk(KERN_INFO "ixgbe: %s NIC Link is Up %s, "
			       "Flow Control: %s\n",
			       netdev->name,
			       (link_speed == IXGBE_LINK_SPEED_10GB_FULL ?
			        "10 Gbps" :
			        (link_speed == IXGBE_LINK_SPEED_1GB_FULL ?
			         "1 Gbps" : "unknown speed")),
4006 4007 4008
			       ((flow_rx && flow_tx) ? "RX/TX" :
			        (flow_rx ? "RX" :
			        (flow_tx ? "TX" : "None"))));
4009 4010 4011 4012 4013 4014 4015

			netif_carrier_on(netdev);
		} else {
			/* Force detection of hung controller */
			adapter->detect_tx_hung = true;
		}
	} else {
4016 4017
		adapter->link_up = false;
		adapter->link_speed = 0;
4018
		if (netif_carrier_ok(netdev)) {
4019 4020
			printk(KERN_INFO "ixgbe: %s NIC Link is Down\n",
			       netdev->name);
4021 4022 4023 4024 4025
			netif_carrier_off(netdev);
		}
	}

	ixgbe_update_stats(adapter);
4026
	adapter->flags &= ~IXGBE_FLAG_IN_WATCHDOG_TASK;
4027 4028 4029
}

static int ixgbe_tso(struct ixgbe_adapter *adapter,
4030 4031
                     struct ixgbe_ring *tx_ring, struct sk_buff *skb,
                     u32 tx_flags, u8 *hdr_len)
4032 4033 4034 4035 4036
{
	struct ixgbe_adv_tx_context_desc *context_desc;
	unsigned int i;
	int err;
	struct ixgbe_tx_buffer *tx_buffer_info;
J
Jesse Brandeburg 已提交
4037 4038
	u32 vlan_macip_lens = 0, type_tucmd_mlhl;
	u32 mss_l4len_idx, l4len;
4039 4040 4041 4042 4043 4044 4045 4046 4047 4048

	if (skb_is_gso(skb)) {
		if (skb_header_cloned(skb)) {
			err = pskb_expand_head(skb, 0, 0, GFP_ATOMIC);
			if (err)
				return err;
		}
		l4len = tcp_hdrlen(skb);
		*hdr_len += l4len;

A
Al Viro 已提交
4049
		if (skb->protocol == htons(ETH_P_IP)) {
4050 4051 4052 4053
			struct iphdr *iph = ip_hdr(skb);
			iph->tot_len = 0;
			iph->check = 0;
			tcp_hdr(skb)->check = ~csum_tcpudp_magic(iph->saddr,
4054 4055 4056
			                                         iph->daddr, 0,
			                                         IPPROTO_TCP,
			                                         0);
4057 4058 4059 4060 4061
			adapter->hw_tso_ctxt++;
		} else if (skb_shinfo(skb)->gso_type == SKB_GSO_TCPV6) {
			ipv6_hdr(skb)->payload_len = 0;
			tcp_hdr(skb)->check =
			    ~csum_ipv6_magic(&ipv6_hdr(skb)->saddr,
4062 4063
			                     &ipv6_hdr(skb)->daddr,
			                     0, IPPROTO_TCP, 0);
4064 4065 4066 4067 4068 4069 4070 4071 4072 4073 4074 4075 4076
			adapter->hw_tso6_ctxt++;
		}

		i = tx_ring->next_to_use;

		tx_buffer_info = &tx_ring->tx_buffer_info[i];
		context_desc = IXGBE_TX_CTXTDESC_ADV(*tx_ring, i);

		/* VLAN MACLEN IPLEN */
		if (tx_flags & IXGBE_TX_FLAGS_VLAN)
			vlan_macip_lens |=
			    (tx_flags & IXGBE_TX_FLAGS_VLAN_MASK);
		vlan_macip_lens |= ((skb_network_offset(skb)) <<
4077
		                    IXGBE_ADVTXD_MACLEN_SHIFT);
4078 4079 4080 4081 4082 4083 4084 4085 4086
		*hdr_len += skb_network_offset(skb);
		vlan_macip_lens |=
		    (skb_transport_header(skb) - skb_network_header(skb));
		*hdr_len +=
		    (skb_transport_header(skb) - skb_network_header(skb));
		context_desc->vlan_macip_lens = cpu_to_le32(vlan_macip_lens);
		context_desc->seqnum_seed = 0;

		/* ADV DTYP TUCMD MKRLOC/ISCSIHEDLEN */
J
Jesse Brandeburg 已提交
4087
		type_tucmd_mlhl = (IXGBE_TXD_CMD_DEXT |
4088
		                   IXGBE_ADVTXD_DTYP_CTXT);
4089

A
Al Viro 已提交
4090
		if (skb->protocol == htons(ETH_P_IP))
4091 4092 4093 4094 4095
			type_tucmd_mlhl |= IXGBE_ADVTXD_TUCMD_IPV4;
		type_tucmd_mlhl |= IXGBE_ADVTXD_TUCMD_L4T_TCP;
		context_desc->type_tucmd_mlhl = cpu_to_le32(type_tucmd_mlhl);

		/* MSS L4LEN IDX */
J
Jesse Brandeburg 已提交
4096
		mss_l4len_idx =
4097 4098
		    (skb_shinfo(skb)->gso_size << IXGBE_ADVTXD_MSS_SHIFT);
		mss_l4len_idx |= (l4len << IXGBE_ADVTXD_L4LEN_SHIFT);
4099 4100
		/* use index 1 for TSO */
		mss_l4len_idx |= (1 << IXGBE_ADVTXD_IDX_SHIFT);
4101 4102 4103 4104 4105 4106 4107 4108 4109 4110 4111 4112 4113 4114 4115 4116
		context_desc->mss_l4len_idx = cpu_to_le32(mss_l4len_idx);

		tx_buffer_info->time_stamp = jiffies;
		tx_buffer_info->next_to_watch = i;

		i++;
		if (i == tx_ring->count)
			i = 0;
		tx_ring->next_to_use = i;

		return true;
	}
	return false;
}

static bool ixgbe_tx_csum(struct ixgbe_adapter *adapter,
4117 4118
                          struct ixgbe_ring *tx_ring,
                          struct sk_buff *skb, u32 tx_flags)
4119 4120 4121 4122 4123 4124 4125 4126 4127 4128 4129 4130 4131 4132 4133 4134
{
	struct ixgbe_adv_tx_context_desc *context_desc;
	unsigned int i;
	struct ixgbe_tx_buffer *tx_buffer_info;
	u32 vlan_macip_lens = 0, type_tucmd_mlhl = 0;

	if (skb->ip_summed == CHECKSUM_PARTIAL ||
	    (tx_flags & IXGBE_TX_FLAGS_VLAN)) {
		i = tx_ring->next_to_use;
		tx_buffer_info = &tx_ring->tx_buffer_info[i];
		context_desc = IXGBE_TX_CTXTDESC_ADV(*tx_ring, i);

		if (tx_flags & IXGBE_TX_FLAGS_VLAN)
			vlan_macip_lens |=
			    (tx_flags & IXGBE_TX_FLAGS_VLAN_MASK);
		vlan_macip_lens |= (skb_network_offset(skb) <<
4135
		                    IXGBE_ADVTXD_MACLEN_SHIFT);
4136 4137
		if (skb->ip_summed == CHECKSUM_PARTIAL)
			vlan_macip_lens |= (skb_transport_header(skb) -
4138
			                    skb_network_header(skb));
4139 4140 4141 4142 4143

		context_desc->vlan_macip_lens = cpu_to_le32(vlan_macip_lens);
		context_desc->seqnum_seed = 0;

		type_tucmd_mlhl |= (IXGBE_TXD_CMD_DEXT |
4144
		                    IXGBE_ADVTXD_DTYP_CTXT);
4145 4146

		if (skb->ip_summed == CHECKSUM_PARTIAL) {
4147
			switch (skb->protocol) {
4148
			case cpu_to_be16(ETH_P_IP):
4149
				type_tucmd_mlhl |= IXGBE_ADVTXD_TUCMD_IPV4;
4150 4151
				if (ip_hdr(skb)->protocol == IPPROTO_TCP)
					type_tucmd_mlhl |=
4152
					        IXGBE_ADVTXD_TUCMD_L4T_TCP;
4153
				break;
4154
			case cpu_to_be16(ETH_P_IPV6):
4155 4156 4157
				/* XXX what about other V6 headers?? */
				if (ipv6_hdr(skb)->nexthdr == IPPROTO_TCP)
					type_tucmd_mlhl |=
4158
					        IXGBE_ADVTXD_TUCMD_L4T_TCP;
4159 4160 4161 4162 4163 4164 4165 4166 4167
				break;
			default:
				if (unlikely(net_ratelimit())) {
					DPRINTK(PROBE, WARNING,
					 "partial checksum but proto=%x!\n",
					 skb->protocol);
				}
				break;
			}
4168 4169 4170
		}

		context_desc->type_tucmd_mlhl = cpu_to_le32(type_tucmd_mlhl);
4171
		/* use index zero for tx checksum offload */
4172 4173 4174 4175
		context_desc->mss_l4len_idx = 0;

		tx_buffer_info->time_stamp = jiffies;
		tx_buffer_info->next_to_watch = i;
J
Jesse Brandeburg 已提交
4176

4177 4178 4179 4180 4181 4182 4183 4184
		adapter->hw_csum_tx_good++;
		i++;
		if (i == tx_ring->count)
			i = 0;
		tx_ring->next_to_use = i;

		return true;
	}
J
Jesse Brandeburg 已提交
4185

4186 4187 4188 4189
	return false;
}

static int ixgbe_tx_map(struct ixgbe_adapter *adapter,
4190 4191
                        struct ixgbe_ring *tx_ring,
                        struct sk_buff *skb, unsigned int first)
4192 4193
{
	struct ixgbe_tx_buffer *tx_buffer_info;
4194
	unsigned int len = skb_headlen(skb);
4195 4196 4197
	unsigned int offset = 0, size, count = 0, i;
	unsigned int nr_frags = skb_shinfo(skb)->nr_frags;
	unsigned int f;
4198
	dma_addr_t *map;
4199 4200 4201

	i = tx_ring->next_to_use;

4202 4203 4204 4205 4206 4207 4208
	if (skb_dma_map(&adapter->pdev->dev, skb, DMA_TO_DEVICE)) {
		dev_err(&adapter->pdev->dev, "TX DMA map failed\n");
		return 0;
	}

	map = skb_shinfo(skb)->dma_maps;

4209 4210 4211 4212 4213
	while (len) {
		tx_buffer_info = &tx_ring->tx_buffer_info[i];
		size = min(len, (uint)IXGBE_MAX_DATA_PER_TXD);

		tx_buffer_info->length = size;
4214
		tx_buffer_info->dma = map[0] + offset;
4215 4216 4217 4218 4219 4220
		tx_buffer_info->time_stamp = jiffies;
		tx_buffer_info->next_to_watch = i;

		len -= size;
		offset += size;
		count++;
4221 4222 4223 4224 4225 4226

		if (len) {
			i++;
			if (i == tx_ring->count)
				i = 0;
		}
4227 4228 4229 4230 4231 4232 4233
	}

	for (f = 0; f < nr_frags; f++) {
		struct skb_frag_struct *frag;

		frag = &skb_shinfo(skb)->frags[f];
		len = frag->size;
4234
		offset = 0;
4235 4236

		while (len) {
4237 4238 4239 4240
			i++;
			if (i == tx_ring->count)
				i = 0;

4241 4242 4243 4244
			tx_buffer_info = &tx_ring->tx_buffer_info[i];
			size = min(len, (uint)IXGBE_MAX_DATA_PER_TXD);

			tx_buffer_info->length = size;
4245
			tx_buffer_info->dma = map[f + 1] + offset;
4246 4247 4248 4249 4250 4251 4252 4253
			tx_buffer_info->time_stamp = jiffies;
			tx_buffer_info->next_to_watch = i;

			len -= size;
			offset += size;
			count++;
		}
	}
4254

4255 4256 4257 4258 4259 4260 4261
	tx_ring->tx_buffer_info[i].skb = skb;
	tx_ring->tx_buffer_info[first].next_to_watch = i;

	return count;
}

static void ixgbe_tx_queue(struct ixgbe_adapter *adapter,
4262 4263
                           struct ixgbe_ring *tx_ring,
                           int tx_flags, int count, u32 paylen, u8 hdr_len)
4264 4265 4266 4267 4268 4269 4270 4271 4272 4273 4274 4275 4276 4277 4278 4279 4280 4281
{
	union ixgbe_adv_tx_desc *tx_desc = NULL;
	struct ixgbe_tx_buffer *tx_buffer_info;
	u32 olinfo_status = 0, cmd_type_len = 0;
	unsigned int i;
	u32 txd_cmd = IXGBE_TXD_CMD_EOP | IXGBE_TXD_CMD_RS | IXGBE_TXD_CMD_IFCS;

	cmd_type_len |= IXGBE_ADVTXD_DTYP_DATA;

	cmd_type_len |= IXGBE_ADVTXD_DCMD_IFCS | IXGBE_ADVTXD_DCMD_DEXT;

	if (tx_flags & IXGBE_TX_FLAGS_VLAN)
		cmd_type_len |= IXGBE_ADVTXD_DCMD_VLE;

	if (tx_flags & IXGBE_TX_FLAGS_TSO) {
		cmd_type_len |= IXGBE_ADVTXD_DCMD_TSE;

		olinfo_status |= IXGBE_TXD_POPTS_TXSM <<
4282
		                 IXGBE_ADVTXD_POPTS_SHIFT;
4283

4284 4285
		/* use index 1 context for tso */
		olinfo_status |= (1 << IXGBE_ADVTXD_IDX_SHIFT);
4286 4287
		if (tx_flags & IXGBE_TX_FLAGS_IPV4)
			olinfo_status |= IXGBE_TXD_POPTS_IXSM <<
4288
			                 IXGBE_ADVTXD_POPTS_SHIFT;
4289 4290 4291

	} else if (tx_flags & IXGBE_TX_FLAGS_CSUM)
		olinfo_status |= IXGBE_TXD_POPTS_TXSM <<
4292
		                 IXGBE_ADVTXD_POPTS_SHIFT;
4293 4294 4295 4296 4297 4298 4299 4300 4301

	olinfo_status |= ((paylen - hdr_len) << IXGBE_ADVTXD_PAYLEN_SHIFT);

	i = tx_ring->next_to_use;
	while (count--) {
		tx_buffer_info = &tx_ring->tx_buffer_info[i];
		tx_desc = IXGBE_TX_DESC_ADV(*tx_ring, i);
		tx_desc->read.buffer_addr = cpu_to_le64(tx_buffer_info->dma);
		tx_desc->read.cmd_type_len =
4302
		        cpu_to_le32(cmd_type_len | tx_buffer_info->length);
4303 4304 4305 4306 4307 4308 4309 4310 4311 4312 4313 4314 4315 4316 4317 4318 4319 4320 4321 4322
		tx_desc->read.olinfo_status = cpu_to_le32(olinfo_status);
		i++;
		if (i == tx_ring->count)
			i = 0;
	}

	tx_desc->read.cmd_type_len |= cpu_to_le32(txd_cmd);

	/*
	 * Force memory writes to complete before letting h/w
	 * know there are new descriptors to fetch.  (Only
	 * applicable for weak-ordered memory model archs,
	 * such as IA-64).
	 */
	wmb();

	tx_ring->next_to_use = i;
	writel(i, adapter->hw.hw_addr + tx_ring->tail);
}

4323
static int __ixgbe_maybe_stop_tx(struct net_device *netdev,
4324
                                 struct ixgbe_ring *tx_ring, int size)
4325 4326 4327
{
	struct ixgbe_adapter *adapter = netdev_priv(netdev);

4328
	netif_stop_subqueue(netdev, tx_ring->queue_index);
4329 4330 4331 4332 4333 4334 4335 4336 4337 4338 4339
	/* Herbert's original patch had:
	 *  smp_mb__after_netif_stop_queue();
	 * but since that doesn't exist yet, just open code it. */
	smp_mb();

	/* We need to check again in a case another CPU has just
	 * made room available. */
	if (likely(IXGBE_DESC_UNUSED(tx_ring) < size))
		return -EBUSY;

	/* A reprieve! - use start_queue because it doesn't call schedule */
4340
	netif_start_subqueue(netdev, tx_ring->queue_index);
4341 4342 4343 4344 4345
	++adapter->restart_queue;
	return 0;
}

static int ixgbe_maybe_stop_tx(struct net_device *netdev,
4346
                              struct ixgbe_ring *tx_ring, int size)
4347 4348 4349 4350 4351 4352
{
	if (likely(IXGBE_DESC_UNUSED(tx_ring) >= size))
		return 0;
	return __ixgbe_maybe_stop_tx(netdev, tx_ring, size);
}

4353 4354 4355 4356 4357 4358 4359 4360 4361 4362
static u16 ixgbe_select_queue(struct net_device *dev, struct sk_buff *skb)
{
	struct ixgbe_adapter *adapter = netdev_priv(dev);

	if (adapter->flags & IXGBE_FLAG_DCB_ENABLED)
		return 0;  /* All traffic should default to class 0 */

	return skb_tx_hash(dev, skb);
}

4363 4364 4365 4366 4367 4368
static int ixgbe_xmit_frame(struct sk_buff *skb, struct net_device *netdev)
{
	struct ixgbe_adapter *adapter = netdev_priv(netdev);
	struct ixgbe_ring *tx_ring;
	unsigned int first;
	unsigned int tx_flags = 0;
4369 4370
	u8 hdr_len = 0;
	int r_idx = 0, tso;
4371 4372
	int count = 0;
	unsigned int f;
J
Jesse Brandeburg 已提交
4373

W
Wu Fengguang 已提交
4374
	r_idx = skb->queue_mapping;
4375
	tx_ring = &adapter->tx_ring[r_idx];
4376

J
Jesse Brandeburg 已提交
4377 4378
	if (adapter->vlgrp && vlan_tx_tag_present(skb)) {
		tx_flags |= vlan_tx_tag_get(skb);
4379 4380 4381 4382 4383 4384 4385 4386
		if (adapter->flags & IXGBE_FLAG_DCB_ENABLED) {
			tx_flags &= ~IXGBE_TX_FLAGS_VLAN_PRIO_MASK;
			tx_flags |= (skb->queue_mapping << 13);
		}
		tx_flags <<= IXGBE_TX_FLAGS_VLAN_SHIFT;
		tx_flags |= IXGBE_TX_FLAGS_VLAN;
	} else if (adapter->flags & IXGBE_FLAG_DCB_ENABLED) {
		tx_flags |= (skb->queue_mapping << 13);
J
Jesse Brandeburg 已提交
4387 4388
		tx_flags <<= IXGBE_TX_FLAGS_VLAN_SHIFT;
		tx_flags |= IXGBE_TX_FLAGS_VLAN;
4389
	}
J
Jesse Brandeburg 已提交
4390 4391 4392 4393
	/* three things can cause us to need a context descriptor */
	if (skb_is_gso(skb) ||
	    (skb->ip_summed == CHECKSUM_PARTIAL) ||
	    (tx_flags & IXGBE_TX_FLAGS_VLAN))
4394 4395
		count++;

J
Jesse Brandeburg 已提交
4396 4397
	count += TXD_USE_COUNT(skb_headlen(skb));
	for (f = 0; f < skb_shinfo(skb)->nr_frags; f++)
4398 4399
		count += TXD_USE_COUNT(skb_shinfo(skb)->frags[f].size);

4400
	if (ixgbe_maybe_stop_tx(netdev, tx_ring, count)) {
4401 4402 4403 4404
		adapter->tx_busy++;
		return NETDEV_TX_BUSY;
	}

A
Al Viro 已提交
4405
	if (skb->protocol == htons(ETH_P_IP))
4406 4407 4408 4409 4410 4411 4412 4413 4414 4415 4416
		tx_flags |= IXGBE_TX_FLAGS_IPV4;
	first = tx_ring->next_to_use;
	tso = ixgbe_tso(adapter, tx_ring, skb, tx_flags, &hdr_len);
	if (tso < 0) {
		dev_kfree_skb_any(skb);
		return NETDEV_TX_OK;
	}

	if (tso)
		tx_flags |= IXGBE_TX_FLAGS_TSO;
	else if (ixgbe_tx_csum(adapter, tx_ring, skb, tx_flags) &&
4417
	         (skb->ip_summed == CHECKSUM_PARTIAL))
4418 4419
		tx_flags |= IXGBE_TX_FLAGS_CSUM;

4420
	count = ixgbe_tx_map(adapter, tx_ring, skb, first);
4421

4422 4423 4424 4425 4426
	if (count) {
		ixgbe_tx_queue(adapter, tx_ring, tx_flags, count, skb->len,
		               hdr_len);
		netdev->trans_start = jiffies;
		ixgbe_maybe_stop_tx(netdev, tx_ring, DESC_NEEDED);
4427

4428 4429 4430 4431 4432
	} else {
		dev_kfree_skb_any(skb);
		tx_ring->tx_buffer_info[first].time_stamp = 0;
		tx_ring->next_to_use = first;
	}
4433 4434 4435 4436 4437 4438 4439 4440 4441 4442 4443 4444 4445 4446 4447 4448 4449 4450 4451 4452 4453 4454 4455 4456 4457 4458 4459 4460 4461

	return NETDEV_TX_OK;
}

/**
 * ixgbe_get_stats - Get System Network Statistics
 * @netdev: network interface device structure
 *
 * Returns the address of the device statistics structure.
 * The statistics are actually updated from the timer callback.
 **/
static struct net_device_stats *ixgbe_get_stats(struct net_device *netdev)
{
	struct ixgbe_adapter *adapter = netdev_priv(netdev);

	/* only return the current stats */
	return &adapter->net_stats;
}

/**
 * ixgbe_set_mac - Change the Ethernet Address of the NIC
 * @netdev: network interface device structure
 * @p: pointer to an address structure
 *
 * Returns 0 on success, negative on failure
 **/
static int ixgbe_set_mac(struct net_device *netdev, void *p)
{
	struct ixgbe_adapter *adapter = netdev_priv(netdev);
4462
	struct ixgbe_hw *hw = &adapter->hw;
4463 4464 4465 4466 4467 4468
	struct sockaddr *addr = p;

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

	memcpy(netdev->dev_addr, addr->sa_data, netdev->addr_len);
4469
	memcpy(hw->mac.addr, addr->sa_data, netdev->addr_len);
4470

4471
	hw->mac.ops.set_rar(hw, 0, hw->mac.addr, 0, IXGBE_RAH_AV);
4472 4473 4474 4475 4476 4477 4478 4479 4480 4481 4482 4483 4484 4485 4486 4487 4488 4489 4490 4491 4492 4493

	return 0;
}

#ifdef CONFIG_NET_POLL_CONTROLLER
/*
 * Polling 'interrupt' - used by things like netconsole to send skbs
 * without having to re-enable interrupts. It's not called while
 * the interrupt routine is executing.
 */
static void ixgbe_netpoll(struct net_device *netdev)
{
	struct ixgbe_adapter *adapter = netdev_priv(netdev);

	disable_irq(adapter->pdev->irq);
	adapter->flags |= IXGBE_FLAG_IN_NETPOLL;
	ixgbe_intr(adapter->pdev->irq, netdev);
	adapter->flags &= ~IXGBE_FLAG_IN_NETPOLL;
	enable_irq(adapter->pdev->irq);
}
#endif

4494 4495 4496
static const struct net_device_ops ixgbe_netdev_ops = {
	.ndo_open 		= ixgbe_open,
	.ndo_stop		= ixgbe_close,
4497
	.ndo_start_xmit		= ixgbe_xmit_frame,
4498
	.ndo_select_queue	= ixgbe_select_queue,
4499
	.ndo_get_stats		= ixgbe_get_stats,
4500
	.ndo_set_rx_mode        = ixgbe_set_rx_mode,
4501 4502 4503 4504 4505 4506 4507 4508 4509 4510 4511 4512 4513
	.ndo_set_multicast_list	= ixgbe_set_rx_mode,
	.ndo_validate_addr	= eth_validate_addr,
	.ndo_set_mac_address	= ixgbe_set_mac,
	.ndo_change_mtu		= ixgbe_change_mtu,
	.ndo_tx_timeout		= ixgbe_tx_timeout,
	.ndo_vlan_rx_register	= ixgbe_vlan_rx_register,
	.ndo_vlan_rx_add_vid	= ixgbe_vlan_rx_add_vid,
	.ndo_vlan_rx_kill_vid	= ixgbe_vlan_rx_kill_vid,
#ifdef CONFIG_NET_POLL_CONTROLLER
	.ndo_poll_controller	= ixgbe_netpoll,
#endif
};

4514 4515 4516 4517 4518 4519 4520 4521 4522 4523 4524 4525
/**
 * ixgbe_probe - Device Initialization Routine
 * @pdev: PCI device information struct
 * @ent: entry in ixgbe_pci_tbl
 *
 * Returns 0 on success, negative on failure
 *
 * ixgbe_probe initializes an adapter identified by a pci_dev structure.
 * The OS initialization, configuring of the adapter private structure,
 * and a hardware reset occur.
 **/
static int __devinit ixgbe_probe(struct pci_dev *pdev,
4526
                                 const struct pci_device_id *ent)
4527 4528 4529 4530 4531 4532 4533
{
	struct net_device *netdev;
	struct ixgbe_adapter *adapter = NULL;
	struct ixgbe_hw *hw;
	const struct ixgbe_info *ii = ixgbe_info_tbl[ent->driver_data];
	static int cards_found;
	int i, err, pci_using_dac;
4534
	u16 pm_value = 0;
4535
	u32 part_num, eec;
4536 4537 4538 4539 4540

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

4541 4542
	if (!pci_set_dma_mask(pdev, DMA_BIT_MASK(64)) &&
	    !pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(64))) {
4543 4544
		pci_using_dac = 1;
	} else {
4545
		err = pci_set_dma_mask(pdev, DMA_BIT_MASK(32));
4546
		if (err) {
4547
			err = pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(32));
4548
			if (err) {
4549 4550
				dev_err(&pdev->dev, "No usable DMA "
				        "configuration, aborting\n");
4551 4552 4553 4554 4555 4556 4557 4558 4559 4560 4561 4562
				goto err_dma;
			}
		}
		pci_using_dac = 0;
	}

	err = pci_request_regions(pdev, ixgbe_driver_name);
	if (err) {
		dev_err(&pdev->dev, "pci_request_regions failed 0x%x\n", err);
		goto err_pci_reg;
	}

4563 4564 4565 4566 4567 4568 4569
	err = pci_enable_pcie_error_reporting(pdev);
	if (err) {
		dev_err(&pdev->dev, "pci_enable_pcie_error_reporting failed "
		                    "0x%x\n", err);
		/* non-fatal, continue */
	}

4570
	pci_set_master(pdev);
4571
	pci_save_state(pdev);
4572

4573
	netdev = alloc_etherdev_mq(sizeof(struct ixgbe_adapter), MAX_TX_QUEUES);
4574 4575 4576 4577 4578 4579 4580 4581 4582 4583 4584 4585 4586 4587 4588 4589
	if (!netdev) {
		err = -ENOMEM;
		goto err_alloc_etherdev;
	}

	SET_NETDEV_DEV(netdev, &pdev->dev);

	pci_set_drvdata(pdev, netdev);
	adapter = netdev_priv(netdev);

	adapter->netdev = netdev;
	adapter->pdev = pdev;
	hw = &adapter->hw;
	hw->back = adapter;
	adapter->msg_enable = (1 << DEFAULT_DEBUG_LEVEL_SHIFT) - 1;

4590 4591
	hw->hw_addr = ioremap(pci_resource_start(pdev, 0),
	                      pci_resource_len(pdev, 0));
4592 4593 4594 4595 4596 4597 4598 4599 4600 4601
	if (!hw->hw_addr) {
		err = -EIO;
		goto err_ioremap;
	}

	for (i = 1; i <= 5; i++) {
		if (pci_resource_len(pdev, i) == 0)
			continue;
	}

4602
	netdev->netdev_ops = &ixgbe_netdev_ops;
4603 4604 4605 4606 4607 4608 4609 4610
	ixgbe_set_ethtool_ops(netdev);
	netdev->watchdog_timeo = 5 * HZ;
	strcpy(netdev->name, pci_name(pdev));

	adapter->bd_number = cards_found;

	/* Setup hw api */
	memcpy(&hw->mac.ops, ii->mac_ops, sizeof(hw->mac.ops));
4611
	hw->mac.type  = ii->mac;
4612

4613 4614 4615 4616 4617 4618 4619 4620 4621
	/* EEPROM */
	memcpy(&hw->eeprom.ops, ii->eeprom_ops, sizeof(hw->eeprom.ops));
	eec = IXGBE_READ_REG(hw, IXGBE_EEC);
	/* If EEPROM is valid (bit 8 = 1), use default otherwise use bit bang */
	if (!(eec & (1 << 8)))
		hw->eeprom.ops.read = &ixgbe_read_eeprom_bit_bang_generic;

	/* PHY */
	memcpy(&hw->phy.ops, ii->phy_ops, sizeof(hw->phy.ops));
D
Donald Skidmore 已提交
4622 4623 4624 4625 4626 4627 4628 4629 4630 4631
	hw->phy.sfp_type = ixgbe_sfp_type_unknown;

	/* set up this timer and work struct before calling get_invariants
	 * which might start the timer
	 */
	init_timer(&adapter->sfp_timer);
	adapter->sfp_timer.function = &ixgbe_sfp_timer;
	adapter->sfp_timer.data = (unsigned long) adapter;

	INIT_WORK(&adapter->sfp_task, ixgbe_sfp_task);
4632

4633 4634 4635 4636 4637 4638 4639
	/* multispeed fiber has its own tasklet, called from GPI SDP1 context */
	INIT_WORK(&adapter->multispeed_fiber_task, ixgbe_multispeed_fiber_task);

	/* a new SFP+ module arrival, called from GPI SDP2 context */
	INIT_WORK(&adapter->sfp_config_module_task,
	          ixgbe_sfp_config_module_task);

4640
	err = ii->get_invariants(hw);
D
Donald Skidmore 已提交
4641 4642 4643 4644 4645 4646 4647 4648 4649
	if (err == IXGBE_ERR_SFP_NOT_PRESENT) {
		/* start a kernel thread to watch for a module to arrive */
		set_bit(__IXGBE_SFP_MODULE_NOT_FOUND, &adapter->state);
		mod_timer(&adapter->sfp_timer,
		          round_jiffies(jiffies + (2 * HZ)));
		err = 0;
	} else if (err == IXGBE_ERR_SFP_NOT_SUPPORTED) {
		DPRINTK(PROBE, ERR, "failed to load because an "
		        "unsupported SFP+ module type was detected.\n");
4650
		goto err_hw_init;
D
Donald Skidmore 已提交
4651 4652 4653
	} else if (err) {
		goto err_hw_init;
	}
4654 4655 4656 4657 4658 4659

	/* setup the private structure */
	err = ixgbe_sw_init(adapter);
	if (err)
		goto err_sw_init;

4660 4661
	/* reset_hw fills in the perm_addr as well */
	err = hw->mac.ops.reset_hw(hw);
4662 4663 4664 4665 4666
	if (err == IXGBE_ERR_SFP_NOT_SUPPORTED) {
		dev_err(&adapter->pdev->dev, "failed to load because an "
		        "unsupported SFP+ module type was detected.\n");
		goto err_sw_init;
	} else if (err) {
4667 4668 4669 4670
		dev_err(&adapter->pdev->dev, "HW Init failed: %d\n", err);
		goto err_sw_init;
	}

4671
	netdev->features = NETIF_F_SG |
4672 4673 4674 4675
	                   NETIF_F_IP_CSUM |
	                   NETIF_F_HW_VLAN_TX |
	                   NETIF_F_HW_VLAN_RX |
	                   NETIF_F_HW_VLAN_FILTER;
4676

4677
	netdev->features |= NETIF_F_IPV6_CSUM;
4678 4679
	netdev->features |= NETIF_F_TSO;
	netdev->features |= NETIF_F_TSO6;
H
Herbert Xu 已提交
4680
	netdev->features |= NETIF_F_GRO;
4681 4682 4683

	netdev->vlan_features |= NETIF_F_TSO;
	netdev->vlan_features |= NETIF_F_TSO6;
4684
	netdev->vlan_features |= NETIF_F_IP_CSUM;
4685 4686
	netdev->vlan_features |= NETIF_F_SG;

4687 4688 4689
	if (adapter->flags & IXGBE_FLAG_DCB_ENABLED)
		adapter->flags &= ~IXGBE_FLAG_RSS_ENABLED;

J
Jeff Kirsher 已提交
4690
#ifdef CONFIG_IXGBE_DCB
4691 4692 4693
	netdev->dcbnl_ops = &dcbnl_ops;
#endif

4694 4695 4696 4697
	if (pci_using_dac)
		netdev->features |= NETIF_F_HIGHDMA;

	/* make sure the EEPROM is good */
4698
	if (hw->eeprom.ops.validate_checksum(hw, NULL) < 0) {
4699 4700 4701 4702 4703 4704 4705 4706
		dev_err(&pdev->dev, "The EEPROM Checksum Is Not Valid\n");
		err = -EIO;
		goto err_eeprom;
	}

	memcpy(netdev->dev_addr, hw->mac.perm_addr, netdev->addr_len);
	memcpy(netdev->perm_addr, hw->mac.perm_addr, netdev->addr_len);

4707 4708
	if (ixgbe_validate_mac_addr(netdev->perm_addr)) {
		dev_err(&pdev->dev, "invalid MAC address\n");
4709 4710 4711 4712 4713 4714 4715 4716 4717
		err = -EIO;
		goto err_eeprom;
	}

	init_timer(&adapter->watchdog_timer);
	adapter->watchdog_timer.function = &ixgbe_watchdog;
	adapter->watchdog_timer.data = (unsigned long)adapter;

	INIT_WORK(&adapter->reset_task, ixgbe_reset_task);
4718
	INIT_WORK(&adapter->watchdog_task, ixgbe_watchdog_task);
4719

4720 4721 4722
	err = ixgbe_init_interrupt_scheme(adapter);
	if (err)
		goto err_sw_init;
4723

4724 4725 4726 4727 4728 4729 4730 4731 4732 4733 4734 4735 4736 4737 4738
	switch (pdev->device) {
	case IXGBE_DEV_ID_82599_KX4:
#define IXGBE_PCIE_PMCSR 0x44
		adapter->wol = IXGBE_WUFC_MAG;
		pci_read_config_word(pdev, IXGBE_PCIE_PMCSR, &pm_value);
		pci_write_config_word(pdev, IXGBE_PCIE_PMCSR,
		                      (pm_value | (1 << 8)));
		break;
	default:
		adapter->wol = 0;
		break;
	}
	device_init_wakeup(&adapter->pdev->dev, true);
	device_set_wakeup_enable(&adapter->pdev->dev, adapter->wol);

4739 4740 4741
	/* pick up the PCI bus settings for reporting later */
	hw->mac.ops.get_bus_info(hw);

4742
	/* print bus type/speed/width info */
J
Johannes Berg 已提交
4743
	dev_info(&pdev->dev, "(PCI Express:%s:%s) %pM\n",
4744 4745 4746 4747 4748
	        ((hw->bus.speed == ixgbe_bus_speed_5000) ? "5.0Gb/s":
	         (hw->bus.speed == ixgbe_bus_speed_2500) ? "2.5Gb/s":"Unknown"),
	        ((hw->bus.width == ixgbe_bus_width_pcie_x8) ? "Width x8" :
	         (hw->bus.width == ixgbe_bus_width_pcie_x4) ? "Width x4" :
	         (hw->bus.width == ixgbe_bus_width_pcie_x1) ? "Width x1" :
4749
	         "Unknown"),
J
Johannes Berg 已提交
4750
	        netdev->dev_addr);
4751
	ixgbe_read_pba_num_generic(hw, &part_num);
4752 4753 4754 4755 4756 4757 4758 4759
	if (ixgbe_is_sfp(hw) && hw->phy.sfp_type != ixgbe_sfp_type_not_present)
		dev_info(&pdev->dev, "MAC: %d, PHY: %d, SFP+: %d, PBA No: %06x-%03x\n",
		         hw->mac.type, hw->phy.type, hw->phy.sfp_type,
		         (part_num >> 8), (part_num & 0xff));
	else
		dev_info(&pdev->dev, "MAC: %d, PHY: %d, PBA No: %06x-%03x\n",
		         hw->mac.type, hw->phy.type,
		         (part_num >> 8), (part_num & 0xff));
4760

4761
	if (hw->bus.width <= ixgbe_bus_width_pcie_x4) {
4762
		dev_warn(&pdev->dev, "PCI-Express bandwidth available for "
4763 4764
		         "this card is not sufficient for optimal "
		         "performance.\n");
4765
		dev_warn(&pdev->dev, "For optimal performance a x8 "
4766
		         "PCI-Express slot is required.\n");
4767 4768
	}

4769 4770 4771
	/* save off EEPROM version number */
	hw->eeprom.ops.read(hw, 0x29, &adapter->eeprom_version);

4772
	/* reset the hardware with the new settings */
4773 4774
	hw->mac.ops.start_hw(hw);

4775 4776 4777 4778 4779 4780 4781
	netif_carrier_off(netdev);

	strcpy(netdev->name, "eth%d");
	err = register_netdev(netdev);
	if (err)
		goto err_register;

4782
#ifdef CONFIG_IXGBE_DCA
4783
	if (dca_add_requester(&pdev->dev) == 0) {
4784 4785 4786 4787 4788 4789 4790
		adapter->flags |= IXGBE_FLAG_DCA_ENABLED;
		/* always use CB2 mode, difference is masked
		 * in the CB driver */
		IXGBE_WRITE_REG(hw, IXGBE_DCA_CTRL, 2);
		ixgbe_setup_dca(adapter);
	}
#endif
4791 4792 4793 4794 4795 4796

	dev_info(&pdev->dev, "Intel(R) 10 Gigabit Network Connection\n");
	cards_found++;
	return 0;

err_register:
4797
	ixgbe_release_hw_control(adapter);
4798 4799
err_hw_init:
err_sw_init:
4800
	ixgbe_reset_interrupt_capability(adapter);
4801
err_eeprom:
D
Donald Skidmore 已提交
4802 4803 4804
	clear_bit(__IXGBE_SFP_MODULE_NOT_FOUND, &adapter->state);
	del_timer_sync(&adapter->sfp_timer);
	cancel_work_sync(&adapter->sfp_task);
4805 4806
	cancel_work_sync(&adapter->multispeed_fiber_task);
	cancel_work_sync(&adapter->sfp_config_module_task);
4807 4808 4809 4810 4811 4812 4813 4814 4815 4816 4817 4818 4819 4820 4821 4822 4823 4824 4825 4826 4827 4828 4829 4830
	iounmap(hw->hw_addr);
err_ioremap:
	free_netdev(netdev);
err_alloc_etherdev:
	pci_release_regions(pdev);
err_pci_reg:
err_dma:
	pci_disable_device(pdev);
	return err;
}

/**
 * ixgbe_remove - Device Removal Routine
 * @pdev: PCI device information struct
 *
 * ixgbe_remove is called by the PCI subsystem to alert the driver
 * that it should release a PCI device.  The could be caused by a
 * Hot-Plug event, or because the driver is going to be removed from
 * memory.
 **/
static void __devexit ixgbe_remove(struct pci_dev *pdev)
{
	struct net_device *netdev = pci_get_drvdata(pdev);
	struct ixgbe_adapter *adapter = netdev_priv(netdev);
4831
	int err;
4832 4833

	set_bit(__IXGBE_DOWN, &adapter->state);
D
Donald Skidmore 已提交
4834 4835 4836 4837
	/* clear the module not found bit to make sure the worker won't
	 * reschedule
	 */
	clear_bit(__IXGBE_SFP_MODULE_NOT_FOUND, &adapter->state);
4838 4839
	del_timer_sync(&adapter->watchdog_timer);

D
Donald Skidmore 已提交
4840 4841 4842
	del_timer_sync(&adapter->sfp_timer);
	cancel_work_sync(&adapter->watchdog_task);
	cancel_work_sync(&adapter->sfp_task);
4843 4844
	cancel_work_sync(&adapter->multispeed_fiber_task);
	cancel_work_sync(&adapter->sfp_config_module_task);
4845 4846
	flush_scheduled_work();

4847
#ifdef CONFIG_IXGBE_DCA
4848 4849 4850 4851 4852 4853 4854
	if (adapter->flags & IXGBE_FLAG_DCA_ENABLED) {
		adapter->flags &= ~IXGBE_FLAG_DCA_ENABLED;
		dca_remove_requester(&pdev->dev);
		IXGBE_WRITE_REG(&adapter->hw, IXGBE_DCA_CTRL, 1);
	}

#endif
D
Donald Skidmore 已提交
4855 4856
	if (netdev->reg_state == NETREG_REGISTERED)
		unregister_netdev(netdev);
4857

4858
	ixgbe_reset_interrupt_capability(adapter);
4859

4860
	ixgbe_release_hw_control(adapter);
4861 4862 4863 4864

	iounmap(adapter->hw.hw_addr);
	pci_release_regions(pdev);

4865 4866 4867 4868
	DPRINTK(PROBE, INFO, "complete\n");
	kfree(adapter->tx_ring);
	kfree(adapter->rx_ring);

4869 4870
	free_netdev(netdev);

4871 4872 4873 4874 4875
	err = pci_disable_pcie_error_reporting(pdev);
	if (err)
		dev_err(&pdev->dev,
		        "pci_disable_pcie_error_reporting failed 0x%x\n", err);

4876 4877 4878 4879 4880 4881 4882 4883 4884 4885 4886 4887
	pci_disable_device(pdev);
}

/**
 * ixgbe_io_error_detected - called when PCI error is detected
 * @pdev: Pointer to PCI device
 * @state: The current pci connection state
 *
 * This function is called after a PCI bus error affecting
 * this device has been detected.
 */
static pci_ers_result_t ixgbe_io_error_detected(struct pci_dev *pdev,
4888
                                                pci_channel_state_t state)
4889 4890
{
	struct net_device *netdev = pci_get_drvdata(pdev);
4891
	struct ixgbe_adapter *adapter = netdev_priv(netdev);
4892 4893 4894 4895 4896 4897 4898

	netif_device_detach(netdev);

	if (netif_running(netdev))
		ixgbe_down(adapter);
	pci_disable_device(pdev);

4899
	/* Request a slot reset. */
4900 4901 4902 4903 4904 4905 4906 4907 4908 4909 4910 4911
	return PCI_ERS_RESULT_NEED_RESET;
}

/**
 * ixgbe_io_slot_reset - called after the pci bus has been reset.
 * @pdev: Pointer to PCI device
 *
 * Restart the card from scratch, as if from a cold-boot.
 */
static pci_ers_result_t ixgbe_io_slot_reset(struct pci_dev *pdev)
{
	struct net_device *netdev = pci_get_drvdata(pdev);
4912
	struct ixgbe_adapter *adapter = netdev_priv(netdev);
4913 4914
	pci_ers_result_t result;
	int err;
4915 4916 4917

	if (pci_enable_device(pdev)) {
		DPRINTK(PROBE, ERR,
4918
		        "Cannot re-enable PCI device after reset.\n");
4919 4920 4921 4922
		result = PCI_ERS_RESULT_DISCONNECT;
	} else {
		pci_set_master(pdev);
		pci_restore_state(pdev);
4923

4924 4925
		pci_enable_wake(pdev, PCI_D3hot, 0);
		pci_enable_wake(pdev, PCI_D3cold, 0);
4926

4927
		ixgbe_reset(adapter);
4928
		IXGBE_WRITE_REG(&adapter->hw, IXGBE_WUS, ~0);
4929 4930 4931 4932 4933 4934 4935 4936 4937
		result = PCI_ERS_RESULT_RECOVERED;
	}

	err = pci_cleanup_aer_uncorrect_error_status(pdev);
	if (err) {
		dev_err(&pdev->dev,
		  "pci_cleanup_aer_uncorrect_error_status failed 0x%0x\n", err);
		/* non-fatal, continue */
	}
4938

4939
	return result;
4940 4941 4942 4943 4944 4945 4946 4947 4948 4949 4950 4951
}

/**
 * ixgbe_io_resume - called when traffic can start flowing again.
 * @pdev: Pointer to PCI device
 *
 * This callback is called when the error recovery driver tells us that
 * its OK to resume normal operation.
 */
static void ixgbe_io_resume(struct pci_dev *pdev)
{
	struct net_device *netdev = pci_get_drvdata(pdev);
4952
	struct ixgbe_adapter *adapter = netdev_priv(netdev);
4953 4954 4955 4956 4957 4958 4959 4960 4961 4962 4963 4964 4965 4966 4967 4968 4969 4970 4971 4972 4973 4974 4975 4976 4977 4978 4979 4980 4981 4982 4983 4984 4985 4986 4987 4988 4989 4990 4991 4992 4993 4994 4995 4996

	if (netif_running(netdev)) {
		if (ixgbe_up(adapter)) {
			DPRINTK(PROBE, INFO, "ixgbe_up failed after reset\n");
			return;
		}
	}

	netif_device_attach(netdev);
}

static struct pci_error_handlers ixgbe_err_handler = {
	.error_detected = ixgbe_io_error_detected,
	.slot_reset = ixgbe_io_slot_reset,
	.resume = ixgbe_io_resume,
};

static struct pci_driver ixgbe_driver = {
	.name     = ixgbe_driver_name,
	.id_table = ixgbe_pci_tbl,
	.probe    = ixgbe_probe,
	.remove   = __devexit_p(ixgbe_remove),
#ifdef CONFIG_PM
	.suspend  = ixgbe_suspend,
	.resume   = ixgbe_resume,
#endif
	.shutdown = ixgbe_shutdown,
	.err_handler = &ixgbe_err_handler
};

/**
 * ixgbe_init_module - Driver Registration Routine
 *
 * ixgbe_init_module is the first routine called when the driver is
 * loaded. All it does is register with the PCI subsystem.
 **/
static int __init ixgbe_init_module(void)
{
	int ret;
	printk(KERN_INFO "%s: %s - version %s\n", ixgbe_driver_name,
	       ixgbe_driver_string, ixgbe_driver_version);

	printk(KERN_INFO "%s: %s\n", ixgbe_driver_name, ixgbe_copyright);

4997
#ifdef CONFIG_IXGBE_DCA
4998 4999
	dca_register_notify(&dca_notifier);
#endif
5000

5001 5002 5003
	ret = pci_register_driver(&ixgbe_driver);
	return ret;
}
5004

5005 5006 5007 5008 5009 5010 5011 5012 5013 5014
module_init(ixgbe_init_module);

/**
 * ixgbe_exit_module - Driver Exit Cleanup Routine
 *
 * ixgbe_exit_module is called just before the driver is removed
 * from memory.
 **/
static void __exit ixgbe_exit_module(void)
{
5015
#ifdef CONFIG_IXGBE_DCA
5016 5017
	dca_unregister_notify(&dca_notifier);
#endif
5018 5019
	pci_unregister_driver(&ixgbe_driver);
}
5020

5021
#ifdef CONFIG_IXGBE_DCA
5022
static int ixgbe_notify_dca(struct notifier_block *nb, unsigned long event,
5023
                            void *p)
5024 5025 5026 5027
{
	int ret_val;

	ret_val = driver_for_each_device(&ixgbe_driver.driver, NULL, &event,
5028
	                                 __ixgbe_notify_dca);
5029 5030 5031

	return ret_val ? NOTIFY_BAD : NOTIFY_DONE;
}
5032

5033
#endif /* CONFIG_IXGBE_DCA */
5034 5035 5036 5037 5038 5039 5040 5041 5042 5043
#ifdef DEBUG
/**
 * ixgbe_get_hw_dev_name - return device name string
 * used by hardware layer to print debugging information
 **/
char *ixgbe_get_hw_dev_name(struct ixgbe_hw *hw)
{
	struct ixgbe_adapter *adapter = hw->back;
	return adapter->netdev->name;
}
5044

5045
#endif
5046 5047 5048
module_exit(ixgbe_exit_module);

/* ixgbe_main.c */