ixgbe_main.c 247.0 KB
Newer Older
1 2 3
/*******************************************************************************

  Intel 10 Gigabit PCI Express Linux driver
4
  Copyright(c) 1999 - 2014 Intel Corporation.
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

  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:
23
  Linux NICS <linux.nics@intel.com>
24 25 26 27 28 29 30 31 32 33 34 35
  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>
36
#include <linux/interrupt.h>
37 38
#include <linux/ip.h>
#include <linux/tcp.h>
39
#include <linux/sctp.h>
40
#include <linux/pkt_sched.h>
41
#include <linux/ipv6.h>
42
#include <linux/slab.h>
43 44
#include <net/checksum.h>
#include <net/ip6_checksum.h>
45
#include <linux/etherdevice.h>
46
#include <linux/ethtool.h>
47
#include <linux/if.h>
48
#include <linux/if_vlan.h>
49
#include <linux/if_macvlan.h>
50
#include <linux/if_bridge.h>
51
#include <linux/prefetch.h>
52
#include <scsi/fc/fc_fcoe.h>
53
#include <net/vxlan.h>
54

55 56 57 58 59 60 61 62 63
#ifdef CONFIG_OF
#include <linux/of_net.h>
#endif

#ifdef CONFIG_SPARC
#include <asm/idprom.h>
#include <asm/prom.h>
#endif

64 65
#include "ixgbe.h"
#include "ixgbe_common.h"
66
#include "ixgbe_dcb_82599.h"
67
#include "ixgbe_sriov.h"
68 69

char ixgbe_driver_name[] = "ixgbe";
S
Stephen Hemminger 已提交
70
static const char ixgbe_driver_string[] =
71
			      "Intel(R) 10 Gigabit PCI Express Network Driver";
72
#ifdef IXGBE_FCOE
73 74
char ixgbe_default_device_descr[] =
			      "Intel(R) 10 Gigabit Network Connection";
75 76 77 78
#else
static char ixgbe_default_device_descr[] =
			      "Intel(R) 10 Gigabit Network Connection";
#endif
D
Don Skidmore 已提交
79
#define DRV_VERSION "4.0.1-k"
S
Stephen Hemminger 已提交
80
const char ixgbe_driver_version[] = DRV_VERSION;
81
static const char ixgbe_copyright[] =
82
				"Copyright (c) 1999-2014 Intel Corporation.";
83 84

static const struct ixgbe_info *ixgbe_info_tbl[] = {
85 86 87 88 89
	[board_82598]		= &ixgbe_82598_info,
	[board_82599]		= &ixgbe_82599_info,
	[board_X540]		= &ixgbe_X540_info,
	[board_X550]		= &ixgbe_X550_info,
	[board_X550EM_x]	= &ixgbe_X550EM_x_info,
90 91 92 93 94 95 96 97 98 99
};

/* 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) }
 */
100
static const struct pci_device_id ixgbe_pci_tbl[] = {
101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_82598), board_82598 },
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_82598AF_DUAL_PORT), board_82598 },
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_82598AF_SINGLE_PORT), board_82598 },
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_82598AT), board_82598 },
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_82598AT2), board_82598 },
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_82598EB_CX4), board_82598 },
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_82598_CX4_DUAL_PORT), board_82598 },
	{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 },
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_82598EB_XF_LR), board_82598 },
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_82598EB_SFP_LOM), board_82598 },
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_82598_BX), board_82598 },
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_82599_KX4), board_82599 },
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_82599_XAUI_LOM), board_82599 },
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_82599_KR), board_82599 },
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_82599_SFP), board_82599 },
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_82599_SFP_EM), board_82599 },
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_82599_KX4_MEZZ), board_82599 },
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_82599_CX4), board_82599 },
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_82599_BACKPLANE_FCOE), board_82599 },
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_82599_SFP_FCOE), board_82599 },
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_82599_T3_LOM), board_82599 },
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_82599_COMBO_BACKPLANE), board_82599 },
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_X540T), board_X540 },
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_82599_SFP_SF2), board_82599 },
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_82599_LS), board_82599 },
127
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_82599_QSFP_SF_QP), board_82599 },
128
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_82599EN_SFP), board_82599 },
129
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_82599_SFP_SF_QP), board_82599 },
130
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_X540T1), board_X540 },
131 132 133
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_X550T), board_X550},
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_X550EM_X_KX4), board_X550EM_x},
	{PCI_VDEVICE(INTEL, IXGBE_DEV_ID_X550EM_X_KR), board_X550EM_x},
134 135 136 137 138
	/* required last entry */
	{0, }
};
MODULE_DEVICE_TABLE(pci, ixgbe_pci_tbl);

139
#ifdef CONFIG_IXGBE_DCA
140
static int ixgbe_notify_dca(struct notifier_block *, unsigned long event,
141
			    void *p);
142 143 144 145 146 147 148
static struct notifier_block dca_notifier = {
	.notifier_call = ixgbe_notify_dca,
	.next          = NULL,
	.priority      = 0
};
#endif

149 150 151
#ifdef CONFIG_PCI_IOV
static unsigned int max_vfs;
module_param(max_vfs, uint, 0);
152
MODULE_PARM_DESC(max_vfs,
153
		 "Maximum number of virtual functions to allocate per physical function - default is zero and maximum value is 63. (Deprecated)");
154 155
#endif /* CONFIG_PCI_IOV */

156 157 158 159 160
static unsigned int allow_unsupported_sfp;
module_param(allow_unsupported_sfp, uint, 0);
MODULE_PARM_DESC(allow_unsupported_sfp,
		 "Allow unsupported and untested SFP+ modules on 82599-based adapters");

161 162 163 164 165
#define DEFAULT_MSG_ENABLE (NETIF_MSG_DRV|NETIF_MSG_PROBE|NETIF_MSG_LINK)
static int debug = -1;
module_param(debug, int, 0);
MODULE_PARM_DESC(debug, "Debug level (0=none,...,16=all)");

166 167 168 169 170
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);

171 172
static bool ixgbe_check_cfg_remove(struct ixgbe_hw *hw, struct pci_dev *pdev);

173 174 175 176 177 178 179 180 181 182 183 184 185 186
static int ixgbe_read_pci_cfg_word_parent(struct ixgbe_adapter *adapter,
					  u32 reg, u16 *value)
{
	struct pci_dev *parent_dev;
	struct pci_bus *parent_bus;

	parent_bus = adapter->pdev->bus->parent;
	if (!parent_bus)
		return -1;

	parent_dev = parent_bus->self;
	if (!parent_dev)
		return -1;

187
	if (!pci_is_pcie(parent_dev))
188 189
		return -1;

190
	pcie_capability_read_word(parent_dev, reg, value);
191 192 193
	if (*value == IXGBE_FAILED_READ_CFG_WORD &&
	    ixgbe_check_cfg_remove(&adapter->hw, parent_dev))
		return -1;
194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219
	return 0;
}

static s32 ixgbe_get_parent_bus_info(struct ixgbe_adapter *adapter)
{
	struct ixgbe_hw *hw = &adapter->hw;
	u16 link_status = 0;
	int err;

	hw->bus.type = ixgbe_bus_type_pci_express;

	/* Get the negotiated link width and speed from PCI config space of the
	 * parent, as this device is behind a switch
	 */
	err = ixgbe_read_pci_cfg_word_parent(adapter, 18, &link_status);

	/* assume caller will handle error case */
	if (err)
		return err;

	hw->bus.width = ixgbe_convert_bus_width(link_status);
	hw->bus.speed = ixgbe_convert_bus_speed(link_status);

	return 0;
}

220 221 222 223 224 225 226 227 228 229 230 231 232
/**
 * ixgbe_check_from_parent - Determine whether PCIe info should come from parent
 * @hw: hw specific details
 *
 * This function is used by probe to determine whether a device's PCI-Express
 * bandwidth details should be gathered from the parent bus instead of from the
 * device. Used to ensure that various locations all have the correct device ID
 * checks.
 */
static inline bool ixgbe_pcie_from_parent(struct ixgbe_hw *hw)
{
	switch (hw->device_id) {
	case IXGBE_DEV_ID_82599_SFP_SF_QP:
233
	case IXGBE_DEV_ID_82599_QSFP_SF_QP:
234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270
		return true;
	default:
		return false;
	}
}

static void ixgbe_check_minimum_link(struct ixgbe_adapter *adapter,
				     int expected_gts)
{
	int max_gts = 0;
	enum pci_bus_speed speed = PCI_SPEED_UNKNOWN;
	enum pcie_link_width width = PCIE_LNK_WIDTH_UNKNOWN;
	struct pci_dev *pdev;

	/* determine whether to use the the parent device
	 */
	if (ixgbe_pcie_from_parent(&adapter->hw))
		pdev = adapter->pdev->bus->parent->self;
	else
		pdev = adapter->pdev;

	if (pcie_get_minimum_link(pdev, &speed, &width) ||
	    speed == PCI_SPEED_UNKNOWN || width == PCIE_LNK_WIDTH_UNKNOWN) {
		e_dev_warn("Unable to determine PCI Express bandwidth.\n");
		return;
	}

	switch (speed) {
	case PCIE_SPEED_2_5GT:
		/* 8b/10b encoding reduces max throughput by 20% */
		max_gts = 2 * width;
		break;
	case PCIE_SPEED_5_0GT:
		/* 8b/10b encoding reduces max throughput by 20% */
		max_gts = 4 * width;
		break;
	case PCIE_SPEED_8_0GT:
271
		/* 128b/130b encoding reduces throughput by less than 2% */
272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288
		max_gts = 8 * width;
		break;
	default:
		e_dev_warn("Unable to determine PCI Express bandwidth.\n");
		return;
	}

	e_dev_info("PCI Express bandwidth of %dGT/s available\n",
		   max_gts);
	e_dev_info("(Speed:%s, Width: x%d, Encoding Loss:%s)\n",
		   (speed == PCIE_SPEED_8_0GT ? "8.0GT/s" :
		    speed == PCIE_SPEED_5_0GT ? "5.0GT/s" :
		    speed == PCIE_SPEED_2_5GT ? "2.5GT/s" :
		    "Unknown"),
		   width,
		   (speed == PCIE_SPEED_2_5GT ? "20%" :
		    speed == PCIE_SPEED_5_0GT ? "20%" :
289
		    speed == PCIE_SPEED_8_0GT ? "<2%" :
290 291 292 293 294 295 296 297 298 299
		    "Unknown"));

	if (max_gts < expected_gts) {
		e_dev_warn("This is not sufficient for optimal performance of this card.\n");
		e_dev_warn("For optimal performance, at least %dGT/s of bandwidth is required.\n",
			expected_gts);
		e_dev_warn("A slot with more lanes and/or higher speed is suggested.\n");
	}
}

300 301 302
static void ixgbe_service_event_schedule(struct ixgbe_adapter *adapter)
{
	if (!test_bit(__IXGBE_DOWN, &adapter->state) &&
303
	    !test_bit(__IXGBE_REMOVING, &adapter->state) &&
304 305 306 307
	    !test_and_set_bit(__IXGBE_SERVICE_SCHED, &adapter->state))
		schedule_work(&adapter->service_task);
}

308 309 310 311 312 313 314 315
static void ixgbe_remove_adapter(struct ixgbe_hw *hw)
{
	struct ixgbe_adapter *adapter = hw->back;

	if (!hw->hw_addr)
		return;
	hw->hw_addr = NULL;
	e_dev_err("Adapter removed\n");
316 317
	if (test_bit(__IXGBE_SERVICE_INITED, &adapter->state))
		ixgbe_service_event_schedule(adapter);
318 319
}

320
static void ixgbe_check_remove(struct ixgbe_hw *hw, u32 reg)
321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338
{
	u32 value;

	/* The following check not only optimizes a bit by not
	 * performing a read on the status register when the
	 * register just read was a status register read that
	 * returned IXGBE_FAILED_READ_REG. It also blocks any
	 * potential recursion.
	 */
	if (reg == IXGBE_STATUS) {
		ixgbe_remove_adapter(hw);
		return;
	}
	value = ixgbe_read_reg(hw, IXGBE_STATUS);
	if (value == IXGBE_FAILED_READ_REG)
		ixgbe_remove_adapter(hw);
}

339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364
/**
 * ixgbe_read_reg - Read from device register
 * @hw: hw specific details
 * @reg: offset of register to read
 *
 * Returns : value read or IXGBE_FAILED_READ_REG if removed
 *
 * This function is used to read device registers. It checks for device
 * removal by confirming any read that returns all ones by checking the
 * status register value for all ones. This function avoids reading from
 * the hardware if a removal was previously detected in which case it
 * returns IXGBE_FAILED_READ_REG (all ones).
 */
u32 ixgbe_read_reg(struct ixgbe_hw *hw, u32 reg)
{
	u8 __iomem *reg_addr = ACCESS_ONCE(hw->hw_addr);
	u32 value;

	if (ixgbe_removed(reg_addr))
		return IXGBE_FAILED_READ_REG;
	value = readl(reg_addr + reg);
	if (unlikely(value == IXGBE_FAILED_READ_REG))
		ixgbe_check_remove(hw, reg);
	return value;
}

365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406
static bool ixgbe_check_cfg_remove(struct ixgbe_hw *hw, struct pci_dev *pdev)
{
	u16 value;

	pci_read_config_word(pdev, PCI_VENDOR_ID, &value);
	if (value == IXGBE_FAILED_READ_CFG_WORD) {
		ixgbe_remove_adapter(hw);
		return true;
	}
	return false;
}

u16 ixgbe_read_pci_cfg_word(struct ixgbe_hw *hw, u32 reg)
{
	struct ixgbe_adapter *adapter = hw->back;
	u16 value;

	if (ixgbe_removed(hw->hw_addr))
		return IXGBE_FAILED_READ_CFG_WORD;
	pci_read_config_word(adapter->pdev, reg, &value);
	if (value == IXGBE_FAILED_READ_CFG_WORD &&
	    ixgbe_check_cfg_remove(hw, adapter->pdev))
		return IXGBE_FAILED_READ_CFG_WORD;
	return value;
}

#ifdef CONFIG_PCI_IOV
static u32 ixgbe_read_pci_cfg_dword(struct ixgbe_hw *hw, u32 reg)
{
	struct ixgbe_adapter *adapter = hw->back;
	u32 value;

	if (ixgbe_removed(hw->hw_addr))
		return IXGBE_FAILED_READ_CFG_DWORD;
	pci_read_config_dword(adapter->pdev, reg, &value);
	if (value == IXGBE_FAILED_READ_CFG_DWORD &&
	    ixgbe_check_cfg_remove(hw, adapter->pdev))
		return IXGBE_FAILED_READ_CFG_DWORD;
	return value;
}
#endif /* CONFIG_PCI_IOV */

407 408 409 410 411 412 413 414 415
void ixgbe_write_pci_cfg_word(struct ixgbe_hw *hw, u32 reg, u16 value)
{
	struct ixgbe_adapter *adapter = hw->back;

	if (ixgbe_removed(hw->hw_addr))
		return;
	pci_write_config_word(adapter->pdev, reg, value);
}

416 417 418 419
static void ixgbe_service_event_complete(struct ixgbe_adapter *adapter)
{
	BUG_ON(!test_bit(__IXGBE_SERVICE_SCHED, &adapter->state));

S
Stephen Hemminger 已提交
420
	/* flush memory to make sure state is correct before next watchdog */
421
	smp_mb__before_atomic();
422 423 424
	clear_bit(__IXGBE_SERVICE_SCHED, &adapter->state);
}

425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458
struct ixgbe_reg_info {
	u32 ofs;
	char *name;
};

static const struct ixgbe_reg_info ixgbe_reg_info_tbl[] = {

	/* General Registers */
	{IXGBE_CTRL, "CTRL"},
	{IXGBE_STATUS, "STATUS"},
	{IXGBE_CTRL_EXT, "CTRL_EXT"},

	/* Interrupt Registers */
	{IXGBE_EICR, "EICR"},

	/* RX Registers */
	{IXGBE_SRRCTL(0), "SRRCTL"},
	{IXGBE_DCA_RXCTRL(0), "DRXCTL"},
	{IXGBE_RDLEN(0), "RDLEN"},
	{IXGBE_RDH(0), "RDH"},
	{IXGBE_RDT(0), "RDT"},
	{IXGBE_RXDCTL(0), "RXDCTL"},
	{IXGBE_RDBAL(0), "RDBAL"},
	{IXGBE_RDBAH(0), "RDBAH"},

	/* TX Registers */
	{IXGBE_TDBAL(0), "TDBAL"},
	{IXGBE_TDBAH(0), "TDBAH"},
	{IXGBE_TDLEN(0), "TDLEN"},
	{IXGBE_TDH(0), "TDH"},
	{IXGBE_TDT(0), "TDT"},
	{IXGBE_TXDCTL(0), "TXDCTL"},

	/* List Terminator */
459
	{ .name = NULL }
460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529
};


/*
 * ixgbe_regdump - register printout routine
 */
static void ixgbe_regdump(struct ixgbe_hw *hw, struct ixgbe_reg_info *reginfo)
{
	int i = 0, j = 0;
	char rname[16];
	u32 regs[64];

	switch (reginfo->ofs) {
	case IXGBE_SRRCTL(0):
		for (i = 0; i < 64; i++)
			regs[i] = IXGBE_READ_REG(hw, IXGBE_SRRCTL(i));
		break;
	case IXGBE_DCA_RXCTRL(0):
		for (i = 0; i < 64; i++)
			regs[i] = IXGBE_READ_REG(hw, IXGBE_DCA_RXCTRL(i));
		break;
	case IXGBE_RDLEN(0):
		for (i = 0; i < 64; i++)
			regs[i] = IXGBE_READ_REG(hw, IXGBE_RDLEN(i));
		break;
	case IXGBE_RDH(0):
		for (i = 0; i < 64; i++)
			regs[i] = IXGBE_READ_REG(hw, IXGBE_RDH(i));
		break;
	case IXGBE_RDT(0):
		for (i = 0; i < 64; i++)
			regs[i] = IXGBE_READ_REG(hw, IXGBE_RDT(i));
		break;
	case IXGBE_RXDCTL(0):
		for (i = 0; i < 64; i++)
			regs[i] = IXGBE_READ_REG(hw, IXGBE_RXDCTL(i));
		break;
	case IXGBE_RDBAL(0):
		for (i = 0; i < 64; i++)
			regs[i] = IXGBE_READ_REG(hw, IXGBE_RDBAL(i));
		break;
	case IXGBE_RDBAH(0):
		for (i = 0; i < 64; i++)
			regs[i] = IXGBE_READ_REG(hw, IXGBE_RDBAH(i));
		break;
	case IXGBE_TDBAL(0):
		for (i = 0; i < 64; i++)
			regs[i] = IXGBE_READ_REG(hw, IXGBE_TDBAL(i));
		break;
	case IXGBE_TDBAH(0):
		for (i = 0; i < 64; i++)
			regs[i] = IXGBE_READ_REG(hw, IXGBE_TDBAH(i));
		break;
	case IXGBE_TDLEN(0):
		for (i = 0; i < 64; i++)
			regs[i] = IXGBE_READ_REG(hw, IXGBE_TDLEN(i));
		break;
	case IXGBE_TDH(0):
		for (i = 0; i < 64; i++)
			regs[i] = IXGBE_READ_REG(hw, IXGBE_TDH(i));
		break;
	case IXGBE_TDT(0):
		for (i = 0; i < 64; i++)
			regs[i] = IXGBE_READ_REG(hw, IXGBE_TDT(i));
		break;
	case IXGBE_TXDCTL(0):
		for (i = 0; i < 64; i++)
			regs[i] = IXGBE_READ_REG(hw, IXGBE_TXDCTL(i));
		break;
	default:
530
		pr_info("%-15s %08x\n", reginfo->name,
531 532 533 534 535 536
			IXGBE_READ_REG(hw, reginfo->ofs));
		return;
	}

	for (i = 0; i < 8; i++) {
		snprintf(rname, 16, "%s[%d-%d]", reginfo->name, i*8, i*8+7);
537
		pr_err("%-15s", rname);
538
		for (j = 0; j < 8; j++)
539 540
			pr_cont(" %08x", regs[i*8+j]);
		pr_cont("\n");
541 542 543 544 545 546 547 548 549 550 551 552 553 554
	}

}

/*
 * ixgbe_dump - Print registers, tx-rings and rx-rings
 */
static void ixgbe_dump(struct ixgbe_adapter *adapter)
{
	struct net_device *netdev = adapter->netdev;
	struct ixgbe_hw *hw = &adapter->hw;
	struct ixgbe_reg_info *reginfo;
	int n = 0;
	struct ixgbe_ring *tx_ring;
555
	struct ixgbe_tx_buffer *tx_buffer;
556 557 558 559 560 561 562 563 564 565 566 567 568 569
	union ixgbe_adv_tx_desc *tx_desc;
	struct my_u0 { u64 a; u64 b; } *u0;
	struct ixgbe_ring *rx_ring;
	union ixgbe_adv_rx_desc *rx_desc;
	struct ixgbe_rx_buffer *rx_buffer_info;
	u32 staterr;
	int i = 0;

	if (!netif_msg_hw(adapter))
		return;

	/* Print netdevice Info */
	if (netdev) {
		dev_info(&adapter->pdev->dev, "Net device Info\n");
570
		pr_info("Device Name     state            "
571
			"trans_start      last_rx\n");
572 573 574 575 576
		pr_info("%-15s %016lX %016lX %016lX\n",
			netdev->name,
			netdev->state,
			netdev->trans_start,
			netdev->last_rx);
577 578 579 580
	}

	/* Print Registers */
	dev_info(&adapter->pdev->dev, "Register Dump\n");
581
	pr_info(" Register Name   Value\n");
582 583 584 585 586 587 588
	for (reginfo = (struct ixgbe_reg_info *)ixgbe_reg_info_tbl;
	     reginfo->name; reginfo++) {
		ixgbe_regdump(hw, reginfo);
	}

	/* Print TX Ring Summary */
	if (!netdev || !netif_running(netdev))
589
		return;
590 591

	dev_info(&adapter->pdev->dev, "TX Rings Summary\n");
J
Josh Hay 已提交
592 593 594
	pr_info(" %s     %s              %s        %s\n",
		"Queue [NTU] [NTC] [bi(ntc)->dma  ]",
		"leng", "ntw", "timestamp");
595 596
	for (n = 0; n < adapter->num_tx_queues; n++) {
		tx_ring = adapter->tx_ring[n];
597
		tx_buffer = &tx_ring->tx_buffer_info[tx_ring->next_to_clean];
J
Josh Hay 已提交
598
		pr_info(" %5d %5X %5X %016llX %08X %p %016llX\n",
599
			   n, tx_ring->next_to_use, tx_ring->next_to_clean,
600 601 602 603
			   (u64)dma_unmap_addr(tx_buffer, dma),
			   dma_unmap_len(tx_buffer, len),
			   tx_buffer->next_to_watch,
			   (u64)tx_buffer->time_stamp);
604 605 606 607 608 609 610 611 612 613
	}

	/* Print TX Rings */
	if (!netif_msg_tx_done(adapter))
		goto rx_ring_summary;

	dev_info(&adapter->pdev->dev, "TX Rings Dump\n");

	/* Transmit Descriptor Formats
	 *
614
	 * 82598 Advanced Transmit Descriptor
615 616 617
	 *   +--------------------------------------------------------------+
	 * 0 |         Buffer Address [63:0]                                |
	 *   +--------------------------------------------------------------+
618
	 * 8 |  PAYLEN  | POPTS  | IDX | STA | DCMD  |DTYP |  RSV |  DTALEN |
619 620
	 *   +--------------------------------------------------------------+
	 *   63       46 45    40 39 36 35 32 31   24 23 20 19              0
621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644
	 *
	 * 82598 Advanced Transmit Descriptor (Write-Back Format)
	 *   +--------------------------------------------------------------+
	 * 0 |                          RSV [63:0]                          |
	 *   +--------------------------------------------------------------+
	 * 8 |            RSV           |  STA  |          NXTSEQ           |
	 *   +--------------------------------------------------------------+
	 *   63                       36 35   32 31                         0
	 *
	 * 82599+ Advanced Transmit Descriptor
	 *   +--------------------------------------------------------------+
	 * 0 |         Buffer Address [63:0]                                |
	 *   +--------------------------------------------------------------+
	 * 8 |PAYLEN  |POPTS|CC|IDX  |STA  |DCMD  |DTYP |MAC  |RSV  |DTALEN |
	 *   +--------------------------------------------------------------+
	 *   63     46 45 40 39 38 36 35 32 31  24 23 20 19 18 17 16 15     0
	 *
	 * 82599+ Advanced Transmit Descriptor (Write-Back Format)
	 *   +--------------------------------------------------------------+
	 * 0 |                          RSV [63:0]                          |
	 *   +--------------------------------------------------------------+
	 * 8 |            RSV           |  STA  |           RSV             |
	 *   +--------------------------------------------------------------+
	 *   63                       36 35   32 31                         0
645 646 647 648
	 */

	for (n = 0; n < adapter->num_tx_queues; n++) {
		tx_ring = adapter->tx_ring[n];
649 650 651
		pr_info("------------------------------------\n");
		pr_info("TX QUEUE INDEX = %d\n", tx_ring->queue_index);
		pr_info("------------------------------------\n");
J
Josh Hay 已提交
652 653 654 655
		pr_info("%s%s    %s              %s        %s          %s\n",
			"T [desc]     [address 63:0  ] ",
			"[PlPOIdStDDt Ln] [bi->dma       ] ",
			"leng", "ntw", "timestamp", "bi->skb");
656 657

		for (i = 0; tx_ring->desc && (i < tx_ring->count); i++) {
658
			tx_desc = IXGBE_TX_DESC(tx_ring, i);
659
			tx_buffer = &tx_ring->tx_buffer_info[i];
660
			u0 = (struct my_u0 *)tx_desc;
J
Josh Hay 已提交
661 662 663 664 665 666
			if (dma_unmap_len(tx_buffer, len) > 0) {
				pr_info("T [0x%03X]    %016llX %016llX %016llX %08X %p %016llX %p",
					i,
					le64_to_cpu(u0->a),
					le64_to_cpu(u0->b),
					(u64)dma_unmap_addr(tx_buffer, dma),
667
					dma_unmap_len(tx_buffer, len),
J
Josh Hay 已提交
668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688
					tx_buffer->next_to_watch,
					(u64)tx_buffer->time_stamp,
					tx_buffer->skb);
				if (i == tx_ring->next_to_use &&
					i == tx_ring->next_to_clean)
					pr_cont(" NTC/U\n");
				else if (i == tx_ring->next_to_use)
					pr_cont(" NTU\n");
				else if (i == tx_ring->next_to_clean)
					pr_cont(" NTC\n");
				else
					pr_cont("\n");

				if (netif_msg_pktdata(adapter) &&
				    tx_buffer->skb)
					print_hex_dump(KERN_INFO, "",
						DUMP_PREFIX_ADDRESS, 16, 1,
						tx_buffer->skb->data,
						dma_unmap_len(tx_buffer, len),
						true);
			}
689 690 691 692 693 694
		}
	}

	/* Print RX Rings Summary */
rx_ring_summary:
	dev_info(&adapter->pdev->dev, "RX Rings Summary\n");
695
	pr_info("Queue [NTU] [NTC]\n");
696 697
	for (n = 0; n < adapter->num_rx_queues; n++) {
		rx_ring = adapter->rx_ring[n];
698 699
		pr_info("%5d %5X %5X\n",
			n, rx_ring->next_to_use, rx_ring->next_to_clean);
700 701 702 703
	}

	/* Print RX Rings */
	if (!netif_msg_rx_status(adapter))
704
		return;
705 706 707

	dev_info(&adapter->pdev->dev, "RX Rings Dump\n");

708 709 710
	/* Receive Descriptor Formats
	 *
	 * 82598 Advanced Receive Descriptor (Read) Format
711 712 713 714 715 716 717 718
	 *    63                                           1        0
	 *    +-----------------------------------------------------+
	 *  0 |       Packet Buffer Address [63:1]           |A0/NSE|
	 *    +----------------------------------------------+------+
	 *  8 |       Header Buffer Address [63:1]           |  DD  |
	 *    +-----------------------------------------------------+
	 *
	 *
719
	 * 82598 Advanced Receive Descriptor (Write-Back) Format
720 721 722
	 *
	 *   63       48 47    32 31  30      21 20 16 15   4 3     0
	 *   +------------------------------------------------------+
723 724 725
	 * 0 |       RSS Hash /  |SPH| HDR_LEN  | RSV |Packet|  RSS |
	 *   | Packet   | IP     |   |          |     | Type | Type |
	 *   | Checksum | Ident  |   |          |     |      |      |
726 727 728 729
	 *   +------------------------------------------------------+
	 * 8 | VLAN Tag | Length | Extended Error | Extended Status |
	 *   +------------------------------------------------------+
	 *   63       48 47    32 31            20 19               0
730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750
	 *
	 * 82599+ Advanced Receive Descriptor (Read) Format
	 *    63                                           1        0
	 *    +-----------------------------------------------------+
	 *  0 |       Packet Buffer Address [63:1]           |A0/NSE|
	 *    +----------------------------------------------+------+
	 *  8 |       Header Buffer Address [63:1]           |  DD  |
	 *    +-----------------------------------------------------+
	 *
	 *
	 * 82599+ Advanced Receive Descriptor (Write-Back) Format
	 *
	 *   63       48 47    32 31  30      21 20 17 16   4 3     0
	 *   +------------------------------------------------------+
	 * 0 |RSS / Frag Checksum|SPH| HDR_LEN  |RSC- |Packet|  RSS |
	 *   |/ RTT / PCoE_PARAM |   |          | CNT | Type | Type |
	 *   |/ Flow Dir Flt ID  |   |          |     |      |      |
	 *   +------------------------------------------------------+
	 * 8 | VLAN Tag | Length |Extended Error| Xtnd Status/NEXTP |
	 *   +------------------------------------------------------+
	 *   63       48 47    32 31          20 19                 0
751
	 */
752

753 754
	for (n = 0; n < adapter->num_rx_queues; n++) {
		rx_ring = adapter->rx_ring[n];
755 756 757
		pr_info("------------------------------------\n");
		pr_info("RX QUEUE INDEX = %d\n", rx_ring->queue_index);
		pr_info("------------------------------------\n");
J
Josh Hay 已提交
758 759 760
		pr_info("%s%s%s",
			"R  [desc]      [ PktBuf     A0] ",
			"[  HeadBuf   DD] [bi->dma       ] [bi->skb       ] ",
761
			"<-- Adv Rx Read format\n");
J
Josh Hay 已提交
762 763 764
		pr_info("%s%s%s",
			"RWB[desc]      [PcsmIpSHl PtRs] ",
			"[vl er S cks ln] ---------------- [bi->skb       ] ",
765 766 767 768
			"<-- Adv Rx Write-Back format\n");

		for (i = 0; i < rx_ring->count; i++) {
			rx_buffer_info = &rx_ring->rx_buffer_info[i];
769
			rx_desc = IXGBE_RX_DESC(rx_ring, i);
770 771 772 773
			u0 = (struct my_u0 *)rx_desc;
			staterr = le32_to_cpu(rx_desc->wb.upper.status_error);
			if (staterr & IXGBE_RXD_STAT_DD) {
				/* Descriptor Done */
774
				pr_info("RWB[0x%03X]     %016llX "
775 776 777 778 779
					"%016llX ---------------- %p", i,
					le64_to_cpu(u0->a),
					le64_to_cpu(u0->b),
					rx_buffer_info->skb);
			} else {
780
				pr_info("R  [0x%03X]     %016llX "
781 782 783 784 785 786
					"%016llX %016llX %p", i,
					le64_to_cpu(u0->a),
					le64_to_cpu(u0->b),
					(u64)rx_buffer_info->dma,
					rx_buffer_info->skb);

787 788
				if (netif_msg_pktdata(adapter) &&
				    rx_buffer_info->dma) {
789 790
					print_hex_dump(KERN_INFO, "",
					   DUMP_PREFIX_ADDRESS, 16, 1,
791 792
					   page_address(rx_buffer_info->page) +
						    rx_buffer_info->page_offset,
793
					   ixgbe_rx_bufsz(rx_ring), true);
794 795 796 797
				}
			}

			if (i == rx_ring->next_to_use)
798
				pr_cont(" NTU\n");
799
			else if (i == rx_ring->next_to_clean)
800
				pr_cont(" NTC\n");
801
			else
802
				pr_cont("\n");
803 804 805 806 807

		}
	}
}

808 809 810 811 812 813 814
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,
815
			ctrl_ext & ~IXGBE_CTRL_EXT_DRV_LOAD);
816 817 818 819 820 821 822 823 824
}

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,
825
			ctrl_ext | IXGBE_CTRL_EXT_DRV_LOAD);
826
}
827

828
/**
829 830 831 832 833 834 835 836
 * 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,
837
			   u8 queue, u8 msix_vector)
838 839
{
	u32 ivar, index;
840 841 842 843 844 845 846 847 848 849 850 851 852
	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:
D
Don Skidmore 已提交
853
	case ixgbe_mac_X540:
854 855
	case ixgbe_mac_X550:
	case ixgbe_mac_X550EM_x:
856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877
		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;
	}
878 879
}

880
static inline void ixgbe_irq_rearm_queues(struct ixgbe_adapter *adapter,
881
					  u64 qmask)
882 883 884
{
	u32 mask;

885 886
	switch (adapter->hw.mac.type) {
	case ixgbe_mac_82598EB:
887 888
		mask = (IXGBE_EIMS_RTX_QUEUE & qmask);
		IXGBE_WRITE_REG(&adapter->hw, IXGBE_EICS, mask);
889 890
		break;
	case ixgbe_mac_82599EB:
D
Don Skidmore 已提交
891
	case ixgbe_mac_X540:
892 893
	case ixgbe_mac_X550:
	case ixgbe_mac_X550EM_x:
894 895 896 897
		mask = (qmask & 0xFFFFFFFF);
		IXGBE_WRITE_REG(&adapter->hw, IXGBE_EICS_EX(0), mask);
		mask = (qmask >> 32);
		IXGBE_WRITE_REG(&adapter->hw, IXGBE_EICS_EX(1), mask);
898 899 900
		break;
	default:
		break;
901 902 903
	}
}

904 905
void ixgbe_unmap_and_free_tx_resource(struct ixgbe_ring *ring,
				      struct ixgbe_tx_buffer *tx_buffer)
906
{
907 908 909
	if (tx_buffer->skb) {
		dev_kfree_skb_any(tx_buffer->skb);
		if (dma_unmap_len(tx_buffer, len))
910
			dma_unmap_single(ring->dev,
911 912 913 914 915 916 917 918
					 dma_unmap_addr(tx_buffer, dma),
					 dma_unmap_len(tx_buffer, len),
					 DMA_TO_DEVICE);
	} else if (dma_unmap_len(tx_buffer, len)) {
		dma_unmap_page(ring->dev,
			       dma_unmap_addr(tx_buffer, dma),
			       dma_unmap_len(tx_buffer, len),
			       DMA_TO_DEVICE);
919
	}
920 921 922 923
	tx_buffer->next_to_watch = NULL;
	tx_buffer->skb = NULL;
	dma_unmap_len_set(tx_buffer, len, 0);
	/* tx_buffer must be completely set up in the transmit path */
924 925
}

926
static void ixgbe_update_xoff_rx_lfc(struct ixgbe_adapter *adapter)
927 928 929 930
{
	struct ixgbe_hw *hw = &adapter->hw;
	struct ixgbe_hw_stats *hwstats = &adapter->stats;
	int i;
931
	u32 data;
932

933 934 935
	if ((hw->fc.current_mode != ixgbe_fc_full) &&
	    (hw->fc.current_mode != ixgbe_fc_rx_pause))
		return;
936

937 938 939 940 941 942 943 944
	switch (hw->mac.type) {
	case ixgbe_mac_82598EB:
		data = IXGBE_READ_REG(hw, IXGBE_LXOFFRXC);
		break;
	default:
		data = IXGBE_READ_REG(hw, IXGBE_LXOFFRXCNT);
	}
	hwstats->lxoffrxc += data;
945

946 947
	/* refill credits (no tx hang) if we received xoff */
	if (!data)
948
		return;
949 950 951 952 953 954 955 956 957 958 959

	for (i = 0; i < adapter->num_tx_queues; i++)
		clear_bit(__IXGBE_HANG_CHECK_ARMED,
			  &adapter->tx_ring[i]->state);
}

static void ixgbe_update_xoff_received(struct ixgbe_adapter *adapter)
{
	struct ixgbe_hw *hw = &adapter->hw;
	struct ixgbe_hw_stats *hwstats = &adapter->stats;
	u32 xoff[8] = {0};
960
	u8 tc;
961 962 963 964 965 966 967 968
	int i;
	bool pfc_en = adapter->dcb_cfg.pfc_mode_enable;

	if (adapter->ixgbe_ieee_pfc)
		pfc_en |= !!(adapter->ixgbe_ieee_pfc->pfc_en);

	if (!(adapter->flags & IXGBE_FLAG_DCB_ENABLED) || !pfc_en) {
		ixgbe_update_xoff_rx_lfc(adapter);
969
		return;
970
	}
971 972 973

	/* update stats for each tc, only valid with PFC enabled */
	for (i = 0; i < MAX_TX_PACKET_BUFFERS; i++) {
974 975
		u32 pxoffrxc;

976 977
		switch (hw->mac.type) {
		case ixgbe_mac_82598EB:
978
			pxoffrxc = IXGBE_READ_REG(hw, IXGBE_PXOFFRXC(i));
979
			break;
980
		default:
981
			pxoffrxc = IXGBE_READ_REG(hw, IXGBE_PXOFFRXCNT(i));
982
		}
983 984 985 986
		hwstats->pxoffrxc[i] += pxoffrxc;
		/* Get the TC for given UP */
		tc = netdev_get_prio_tc_map(adapter->netdev, i);
		xoff[tc] += pxoffrxc;
987 988 989 990 991 992
	}

	/* disarm tx queues that have received xoff frames */
	for (i = 0; i < adapter->num_tx_queues; i++) {
		struct ixgbe_ring *tx_ring = adapter->tx_ring[i];

993
		tc = tx_ring->dcb_tc;
994 995
		if (xoff[tc])
			clear_bit(__IXGBE_HANG_CHECK_ARMED, &tx_ring->state);
996 997 998
	}
}

999
static u64 ixgbe_get_tx_completed(struct ixgbe_ring *ring)
1000
{
1001
	return ring->stats.packets;
1002 1003 1004 1005
}

static u64 ixgbe_get_tx_pending(struct ixgbe_ring *ring)
{
1006 1007 1008 1009 1010 1011 1012 1013
	struct ixgbe_adapter *adapter;
	struct ixgbe_hw *hw;
	u32 head, tail;

	if (ring->l2_accel_priv)
		adapter = ring->l2_accel_priv->real_adapter;
	else
		adapter = netdev_priv(ring->netdev);
1014

1015 1016 1017
	hw = &adapter->hw;
	head = IXGBE_READ_REG(hw, IXGBE_TDH(ring->reg_idx));
	tail = IXGBE_READ_REG(hw, IXGBE_TDT(ring->reg_idx));
1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031

	if (head != tail)
		return (head < tail) ?
			tail - head : (tail + ring->count - head);

	return 0;
}

static inline bool ixgbe_check_tx_hang(struct ixgbe_ring *tx_ring)
{
	u32 tx_done = ixgbe_get_tx_completed(tx_ring);
	u32 tx_done_old = tx_ring->tx_stats.tx_done_old;
	u32 tx_pending = ixgbe_get_tx_pending(tx_ring);

A
Alexander Duyck 已提交
1032
	clear_check_for_tx_hang(tx_ring);
1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045

	/*
	 * Check for a hung queue, but be thorough. This verifies
	 * that a transmit has been completed since the previous
	 * check AND there is at least one packet pending. The
	 * ARMED bit is set to indicate a potential hang. The
	 * bit is cleared if a pause frame is received to remove
	 * false hang detection due to PFC or 802.3x frames. By
	 * requiring this to fail twice we avoid races with
	 * pfc clearing the ARMED bit and conditions where we
	 * run the check_tx_hang logic with a transmit completion
	 * pending but without time to complete it yet.
	 */
1046
	if (tx_done_old == tx_done && tx_pending)
1047
		/* make sure it is true for two checks in a row */
1048 1049 1050 1051 1052 1053
		return test_and_set_bit(__IXGBE_HANG_CHECK_ARMED,
					&tx_ring->state);
	/* update completed stats and continue */
	tx_ring->tx_stats.tx_done_old = tx_done;
	/* reset the countdown */
	clear_bit(__IXGBE_HANG_CHECK_ARMED, &tx_ring->state);
1054

1055
	return false;
1056 1057
}

1058 1059 1060 1061 1062 1063 1064 1065 1066 1067
/**
 * ixgbe_tx_timeout_reset - initiate reset due to Tx timeout
 * @adapter: driver private struct
 **/
static void ixgbe_tx_timeout_reset(struct ixgbe_adapter *adapter)
{

	/* Do the reset outside of interrupt context */
	if (!test_bit(__IXGBE_DOWN, &adapter->state)) {
		adapter->flags2 |= IXGBE_FLAG2_RESET_REQUESTED;
1068
		e_warn(drv, "initiating reset due to tx timeout\n");
1069 1070 1071
		ixgbe_service_event_schedule(adapter);
	}
}
1072

1073 1074
/**
 * ixgbe_clean_tx_irq - Reclaim resources after transmit completes
1075
 * @q_vector: structure containing interrupt and ring information
1076
 * @tx_ring: tx ring to clean
1077
 **/
1078
static bool ixgbe_clean_tx_irq(struct ixgbe_q_vector *q_vector,
1079
			       struct ixgbe_ring *tx_ring)
1080
{
1081
	struct ixgbe_adapter *adapter = q_vector->adapter;
1082 1083
	struct ixgbe_tx_buffer *tx_buffer;
	union ixgbe_adv_tx_desc *tx_desc;
1084
	unsigned int total_bytes = 0, total_packets = 0;
1085
	unsigned int budget = q_vector->tx.work_limit;
1086 1087 1088 1089
	unsigned int i = tx_ring->next_to_clean;

	if (test_bit(__IXGBE_DOWN, &adapter->state))
		return true;
1090

1091
	tx_buffer = &tx_ring->tx_buffer_info[i];
1092
	tx_desc = IXGBE_TX_DESC(tx_ring, i);
1093
	i -= tx_ring->count;
1094

1095
	do {
1096 1097 1098 1099 1100 1101
		union ixgbe_adv_tx_desc *eop_desc = tx_buffer->next_to_watch;

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

1102
		/* prevent any other reads prior to eop_desc */
1103
		read_barrier_depends();
1104

1105 1106 1107
		/* if DD is not set pending work has not been completed */
		if (!(eop_desc->wb.status & cpu_to_le32(IXGBE_TXD_STAT_DD)))
			break;
1108

1109 1110
		/* clear next_to_watch to prevent false hangs */
		tx_buffer->next_to_watch = NULL;
1111

1112 1113 1114 1115
		/* update the statistics for this packet */
		total_bytes += tx_buffer->bytecount;
		total_packets += tx_buffer->gso_segs;

1116
		/* free the skb */
1117
		dev_consume_skb_any(tx_buffer->skb);
1118

1119 1120 1121 1122 1123 1124
		/* unmap skb header data */
		dma_unmap_single(tx_ring->dev,
				 dma_unmap_addr(tx_buffer, dma),
				 dma_unmap_len(tx_buffer, len),
				 DMA_TO_DEVICE);

1125 1126
		/* clear tx_buffer data */
		tx_buffer->skb = NULL;
1127
		dma_unmap_len_set(tx_buffer, len, 0);
1128

1129 1130
		/* unmap remaining buffers */
		while (tx_desc != eop_desc) {
1131 1132
			tx_buffer++;
			tx_desc++;
1133
			i++;
1134 1135
			if (unlikely(!i)) {
				i -= tx_ring->count;
1136
				tx_buffer = tx_ring->tx_buffer_info;
1137
				tx_desc = IXGBE_TX_DESC(tx_ring, 0);
1138
			}
1139

1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161
			/* unmap any remaining paged data */
			if (dma_unmap_len(tx_buffer, len)) {
				dma_unmap_page(tx_ring->dev,
					       dma_unmap_addr(tx_buffer, dma),
					       dma_unmap_len(tx_buffer, len),
					       DMA_TO_DEVICE);
				dma_unmap_len_set(tx_buffer, len, 0);
			}
		}

		/* move us one more past the eop_desc for start of next pkt */
		tx_buffer++;
		tx_desc++;
		i++;
		if (unlikely(!i)) {
			i -= tx_ring->count;
			tx_buffer = tx_ring->tx_buffer_info;
			tx_desc = IXGBE_TX_DESC(tx_ring, 0);
		}

		/* issue prefetch for next Tx descriptor */
		prefetch(tx_desc);
1162

1163 1164 1165 1166 1167
		/* update budget accounting */
		budget--;
	} while (likely(budget));

	i += tx_ring->count;
1168
	tx_ring->next_to_clean = i;
1169
	u64_stats_update_begin(&tx_ring->syncp);
1170
	tx_ring->stats.bytes += total_bytes;
1171
	tx_ring->stats.packets += total_packets;
1172
	u64_stats_update_end(&tx_ring->syncp);
1173 1174
	q_vector->tx.total_bytes += total_bytes;
	q_vector->tx.total_packets += total_packets;
1175

1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189
	if (check_for_tx_hang(tx_ring) && ixgbe_check_tx_hang(tx_ring)) {
		/* schedule immediate reset if we believe we hung */
		struct ixgbe_hw *hw = &adapter->hw;
		e_err(drv, "Detected Tx Unit Hang\n"
			"  Tx Queue             <%d>\n"
			"  TDH, TDT             <%x>, <%x>\n"
			"  next_to_use          <%x>\n"
			"  next_to_clean        <%x>\n"
			"tx_buffer_info[next_to_clean]\n"
			"  time_stamp           <%lx>\n"
			"  jiffies              <%lx>\n",
			tx_ring->queue_index,
			IXGBE_READ_REG(hw, IXGBE_TDH(tx_ring->reg_idx)),
			IXGBE_READ_REG(hw, IXGBE_TDT(tx_ring->reg_idx)),
1190 1191
			tx_ring->next_to_use, i,
			tx_ring->tx_buffer_info[i].time_stamp, jiffies);
1192 1193 1194 1195 1196 1197 1198

		netif_stop_subqueue(tx_ring->netdev, tx_ring->queue_index);

		e_info(probe,
		       "tx hang %d detected on queue %d, resetting adapter\n",
			adapter->tx_timeout_count + 1, tx_ring->queue_index);

1199
		/* schedule immediate reset if we believe we hung */
1200
		ixgbe_tx_timeout_reset(adapter);
1201 1202

		/* the adapter is about to reset, no point in enabling stuff */
1203
		return true;
1204
	}
1205

1206 1207 1208
	netdev_tx_completed_queue(txring_txq(tx_ring),
				  total_packets, total_bytes);

1209
#define TX_WAKE_THRESHOLD (DESC_NEEDED * 2)
1210
	if (unlikely(total_packets && netif_carrier_ok(tx_ring->netdev) &&
1211
		     (ixgbe_desc_unused(tx_ring) >= TX_WAKE_THRESHOLD))) {
1212 1213 1214 1215
		/* Make sure that anybody stopping the queue after this
		 * sees the new next_to_clean.
		 */
		smp_mb();
1216 1217 1218 1219 1220
		if (__netif_subqueue_stopped(tx_ring->netdev,
					     tx_ring->queue_index)
		    && !test_bit(__IXGBE_DOWN, &adapter->state)) {
			netif_wake_subqueue(tx_ring->netdev,
					    tx_ring->queue_index);
1221
			++tx_ring->tx_stats.restart_queue;
1222
		}
1223
	}
1224

1225
	return !!budget;
1226 1227
}

1228
#ifdef CONFIG_IXGBE_DCA
1229 1230
static void ixgbe_update_tx_dca(struct ixgbe_adapter *adapter,
				struct ixgbe_ring *tx_ring,
1231
				int cpu)
1232
{
1233
	struct ixgbe_hw *hw = &adapter->hw;
1234 1235
	u32 txctrl = dca3_get_tag(tx_ring->dev, cpu);
	u16 reg_offset;
1236 1237 1238

	switch (hw->mac.type) {
	case ixgbe_mac_82598EB:
1239
		reg_offset = IXGBE_DCA_TXCTRL(tx_ring->reg_idx);
1240 1241
		break;
	case ixgbe_mac_82599EB:
D
Don Skidmore 已提交
1242
	case ixgbe_mac_X540:
1243 1244
		reg_offset = IXGBE_DCA_TXCTRL_82599(tx_ring->reg_idx);
		txctrl <<= IXGBE_DCA_TXCTRL_CPUID_SHIFT_82599;
1245 1246
		break;
	default:
1247 1248
		/* for unknown hardware do not write register */
		return;
1249
	}
1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260

	/*
	 * We can enable relaxed ordering for reads, but not writes when
	 * DCA is enabled.  This is due to a known issue in some chipsets
	 * which will cause the DCA tag to be cleared.
	 */
	txctrl |= IXGBE_DCA_TXCTRL_DESC_RRO_EN |
		  IXGBE_DCA_TXCTRL_DATA_RRO_EN |
		  IXGBE_DCA_TXCTRL_DESC_DCA_EN;

	IXGBE_WRITE_REG(hw, reg_offset, txctrl);
1261 1262
}

1263 1264
static void ixgbe_update_rx_dca(struct ixgbe_adapter *adapter,
				struct ixgbe_ring *rx_ring,
1265
				int cpu)
1266
{
1267
	struct ixgbe_hw *hw = &adapter->hw;
1268 1269 1270
	u32 rxctrl = dca3_get_tag(rx_ring->dev, cpu);
	u8 reg_idx = rx_ring->reg_idx;

1271 1272 1273

	switch (hw->mac.type) {
	case ixgbe_mac_82599EB:
D
Don Skidmore 已提交
1274
	case ixgbe_mac_X540:
1275
		rxctrl <<= IXGBE_DCA_RXCTRL_CPUID_SHIFT_82599;
1276 1277 1278 1279
		break;
	default:
		break;
	}
1280 1281 1282 1283 1284 1285 1286 1287 1288 1289

	/*
	 * We can enable relaxed ordering for reads, but not writes when
	 * DCA is enabled.  This is due to a known issue in some chipsets
	 * which will cause the DCA tag to be cleared.
	 */
	rxctrl |= IXGBE_DCA_RXCTRL_DESC_RRO_EN |
		  IXGBE_DCA_RXCTRL_DESC_DCA_EN;

	IXGBE_WRITE_REG(hw, IXGBE_DCA_RXCTRL(reg_idx), rxctrl);
1290 1291 1292 1293 1294
}

static void ixgbe_update_dca(struct ixgbe_q_vector *q_vector)
{
	struct ixgbe_adapter *adapter = q_vector->adapter;
1295
	struct ixgbe_ring *ring;
1296 1297
	int cpu = get_cpu();

1298 1299 1300
	if (q_vector->cpu == cpu)
		goto out_no_update;

1301
	ixgbe_for_each_ring(ring, q_vector->tx)
1302
		ixgbe_update_tx_dca(adapter, ring, cpu);
1303

1304
	ixgbe_for_each_ring(ring, q_vector->rx)
1305
		ixgbe_update_rx_dca(adapter, ring, cpu);
1306 1307 1308

	q_vector->cpu = cpu;
out_no_update:
1309 1310 1311 1312 1313 1314 1315 1316 1317 1318
	put_cpu();
}

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

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

1319 1320 1321
	/* always use CB2 mode, difference is masked in the CB driver */
	IXGBE_WRITE_REG(&adapter->hw, IXGBE_DCA_CTRL, 2);

1322
	for (i = 0; i < adapter->num_q_vectors; i++) {
1323 1324
		adapter->q_vector[i]->cpu = -1;
		ixgbe_update_dca(adapter->q_vector[i]);
1325 1326 1327 1328 1329
	}
}

static int __ixgbe_notify_dca(struct device *dev, void *data)
{
1330
	struct ixgbe_adapter *adapter = dev_get_drvdata(dev);
1331 1332
	unsigned long event = *(unsigned long *)data;

1333
	if (!(adapter->flags & IXGBE_FLAG_DCA_CAPABLE))
1334 1335
		return 0;

1336 1337
	switch (event) {
	case DCA_PROVIDER_ADD:
1338 1339 1340
		/* if we're already enabled, don't do it again */
		if (adapter->flags & IXGBE_FLAG_DCA_ENABLED)
			break;
1341
		if (dca_add_requester(dev) == 0) {
1342
			adapter->flags |= IXGBE_FLAG_DCA_ENABLED;
1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355
			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;
	}

1356
	return 0;
1357
}
E
Emil Tantilov 已提交
1358

1359
#endif /* CONFIG_IXGBE_DCA */
1360 1361
static inline void ixgbe_rx_hash(struct ixgbe_ring *ring,
				 union ixgbe_adv_rx_desc *rx_desc,
E
Emil Tantilov 已提交
1362 1363
				 struct sk_buff *skb)
{
1364
	if (ring->netdev->features & NETIF_F_RXHASH)
T
Tom Herbert 已提交
1365 1366 1367
		skb_set_hash(skb,
			     le32_to_cpu(rx_desc->wb.lower.hi_dword.rss),
			     PKT_HASH_TYPE_L3);
E
Emil Tantilov 已提交
1368 1369
}

1370
#ifdef IXGBE_FCOE
1371 1372
/**
 * ixgbe_rx_is_fcoe - check the rx desc for incoming pkt type
1373
 * @ring: structure containing ring specific data
1374 1375 1376 1377
 * @rx_desc: advanced rx descriptor
 *
 * Returns : true if it is FCoE pkt
 */
1378
static inline bool ixgbe_rx_is_fcoe(struct ixgbe_ring *ring,
1379 1380 1381 1382
				    union ixgbe_adv_rx_desc *rx_desc)
{
	__le16 pkt_info = rx_desc->wb.lower.lo_dword.hs_rss.pkt_info;

1383
	return test_bit(__IXGBE_RX_FCOE, &ring->state) &&
1384 1385 1386 1387 1388
	       ((pkt_info & cpu_to_le16(IXGBE_RXDADV_PKTTYPE_ETQF_MASK)) ==
		(cpu_to_le16(IXGBE_ETQF_FILTER_FCOE <<
			     IXGBE_RXDADV_PKTTYPE_ETQF_SHIFT)));
}

1389
#endif /* IXGBE_FCOE */
1390 1391
/**
 * ixgbe_rx_checksum - indicate in skb if hw indicated a good cksum
1392 1393
 * @ring: structure containing ring specific data
 * @rx_desc: current Rx descriptor being processed
1394 1395
 * @skb: skb currently being received and modified
 **/
1396
static inline void ixgbe_rx_checksum(struct ixgbe_ring *ring,
1397
				     union ixgbe_adv_rx_desc *rx_desc,
1398
				     struct sk_buff *skb)
1399
{
1400 1401 1402 1403
	__le16 pkt_info = rx_desc->wb.lower.lo_dword.hs_rss.pkt_info;
	__le16 hdr_info = rx_desc->wb.lower.lo_dword.hs_rss.hdr_info;
	bool encap_pkt = false;

1404
	skb_checksum_none_assert(skb);
1405

1406
	/* Rx csum disabled */
1407
	if (!(ring->netdev->features & NETIF_F_RXCSUM))
1408
		return;
1409

1410 1411 1412 1413 1414 1415 1416
	if ((pkt_info & cpu_to_le16(IXGBE_RXDADV_PKTTYPE_VXLAN)) &&
	    (hdr_info & cpu_to_le16(IXGBE_RXDADV_PKTTYPE_TUNNEL >> 16))) {
		encap_pkt = true;
		skb->encapsulation = 1;
		skb->ip_summed = CHECKSUM_NONE;
	}

1417
	/* if IP and error */
1418 1419
	if (ixgbe_test_staterr(rx_desc, IXGBE_RXD_STAT_IPCS) &&
	    ixgbe_test_staterr(rx_desc, IXGBE_RXDADV_ERR_IPE)) {
1420
		ring->rx_stats.csum_err++;
1421 1422
		return;
	}
1423

1424
	if (!ixgbe_test_staterr(rx_desc, IXGBE_RXD_STAT_L4CS))
1425 1426
		return;

1427
	if (ixgbe_test_staterr(rx_desc, IXGBE_RXDADV_ERR_TCPE)) {
1428 1429 1430 1431
		/*
		 * 82599 errata, UDP frames with a 0 checksum can be marked as
		 * checksum errors.
		 */
1432 1433
		if ((pkt_info & cpu_to_le16(IXGBE_RXDADV_PKTTYPE_UDP)) &&
		    test_bit(__IXGBE_RX_CSUM_UDP_ZERO_ERR, &ring->state))
1434 1435
			return;

1436
		ring->rx_stats.csum_err++;
1437 1438 1439
		return;
	}

1440
	/* It must be a TCP or UDP packet with a valid checksum */
1441
	skb->ip_summed = CHECKSUM_UNNECESSARY;
1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452
	if (encap_pkt) {
		if (!ixgbe_test_staterr(rx_desc, IXGBE_RXD_STAT_OUTERIPCS))
			return;

		if (ixgbe_test_staterr(rx_desc, IXGBE_RXDADV_ERR_OUTERIPER)) {
			ring->rx_stats.csum_err++;
			return;
		}
		/* If we checked the outer header let the stack know */
		skb->csum_level = 1;
	}
1453 1454
}

1455 1456 1457 1458
static bool ixgbe_alloc_mapped_page(struct ixgbe_ring *rx_ring,
				    struct ixgbe_rx_buffer *bi)
{
	struct page *page = bi->page;
A
Alexander Duyck 已提交
1459
	dma_addr_t dma;
1460

1461
	/* since we are recycling buffers we should seldom need to alloc */
A
Alexander Duyck 已提交
1462
	if (likely(page))
1463 1464
		return true;

1465
	/* alloc new page for storage */
A
Alexander Duyck 已提交
1466 1467 1468 1469
	page = dev_alloc_pages(ixgbe_rx_pg_order(rx_ring));
	if (unlikely(!page)) {
		rx_ring->rx_stats.alloc_rx_page_failed++;
		return false;
1470 1471
	}

1472 1473 1474 1475 1476 1477 1478 1479 1480
	/* map page for use */
	dma = dma_map_page(rx_ring->dev, page, 0,
			   ixgbe_rx_pg_size(rx_ring), DMA_FROM_DEVICE);

	/*
	 * if mapping failed free memory back to system since
	 * there isn't much point in holding memory we can't use
	 */
	if (dma_mapping_error(rx_ring->dev, dma)) {
1481
		__free_pages(page, ixgbe_rx_pg_order(rx_ring));
1482 1483 1484 1485 1486

		rx_ring->rx_stats.alloc_rx_page_failed++;
		return false;
	}

1487
	bi->dma = dma;
A
Alexander Duyck 已提交
1488
	bi->page = page;
1489
	bi->page_offset = 0;
1490

1491 1492 1493
	return true;
}

1494
/**
1495
 * ixgbe_alloc_rx_buffers - Replace used receive buffers
1496 1497
 * @rx_ring: ring to place buffers on
 * @cleaned_count: number of buffers to replace
1498
 **/
1499
void ixgbe_alloc_rx_buffers(struct ixgbe_ring *rx_ring, u16 cleaned_count)
1500 1501
{
	union ixgbe_adv_rx_desc *rx_desc;
1502
	struct ixgbe_rx_buffer *bi;
1503
	u16 i = rx_ring->next_to_use;
1504

1505 1506
	/* nothing to do */
	if (!cleaned_count)
1507 1508
		return;

1509
	rx_desc = IXGBE_RX_DESC(rx_ring, i);
1510 1511
	bi = &rx_ring->rx_buffer_info[i];
	i -= rx_ring->count;
1512

1513 1514
	do {
		if (!ixgbe_alloc_mapped_page(rx_ring, bi))
1515
			break;
1516

1517 1518 1519 1520 1521
		/*
		 * Refresh the desc even if buffer_addrs didn't change
		 * because each write-back erases this info.
		 */
		rx_desc->read.pkt_addr = cpu_to_le64(bi->dma + bi->page_offset);
1522

1523 1524
		rx_desc++;
		bi++;
1525
		i++;
1526
		if (unlikely(!i)) {
1527
			rx_desc = IXGBE_RX_DESC(rx_ring, 0);
1528 1529 1530 1531
			bi = rx_ring->rx_buffer_info;
			i -= rx_ring->count;
		}

A
Alexander Duyck 已提交
1532 1533
		/* clear the status bits for the next_to_use descriptor */
		rx_desc->wb.upper.status_error = 0;
1534 1535 1536

		cleaned_count--;
	} while (cleaned_count);
1537

1538 1539
	i += rx_ring->count;

1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553
	if (rx_ring->next_to_use != i) {
		rx_ring->next_to_use = i;

		/* update next to alloc since we have filled the ring */
		rx_ring->next_to_alloc = i;

		/* 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();
		writel(i, rx_ring->tail);
	}
1554 1555
}

1556 1557 1558
static void ixgbe_set_rsc_gso_size(struct ixgbe_ring *ring,
				   struct sk_buff *skb)
{
1559
	u16 hdr_len = skb_headlen(skb);
1560 1561 1562 1563

	/* set gso_size to avoid messing up TCP MSS */
	skb_shinfo(skb)->gso_size = DIV_ROUND_UP((skb->len - hdr_len),
						 IXGBE_CB(skb)->append_cnt);
1564
	skb_shinfo(skb)->gso_type = SKB_GSO_TCPV4;
1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582
}

static void ixgbe_update_rsc_stats(struct ixgbe_ring *rx_ring,
				   struct sk_buff *skb)
{
	/* if append_cnt is 0 then frame is not RSC */
	if (!IXGBE_CB(skb)->append_cnt)
		return;

	rx_ring->rx_stats.rsc_count += IXGBE_CB(skb)->append_cnt;
	rx_ring->rx_stats.rsc_flush++;

	ixgbe_set_rsc_gso_size(rx_ring, skb);

	/* gso_size is computed using append_cnt so always clear it last */
	IXGBE_CB(skb)->append_cnt = 0;
}

1583 1584 1585 1586 1587
/**
 * ixgbe_process_skb_fields - Populate skb header fields from Rx descriptor
 * @rx_ring: rx descriptor ring packet is being transacted on
 * @rx_desc: pointer to the EOP Rx descriptor
 * @skb: pointer to current skb being populated
A
Alexander Duyck 已提交
1588
 *
1589 1590 1591
 * This function checks the ring, descriptor, and packet information in
 * order to populate the hash, checksum, VLAN, timestamp, protocol, and
 * other fields within the skb.
A
Alexander Duyck 已提交
1592
 **/
1593 1594 1595
static void ixgbe_process_skb_fields(struct ixgbe_ring *rx_ring,
				     union ixgbe_adv_rx_desc *rx_desc,
				     struct sk_buff *skb)
A
Alexander Duyck 已提交
1596
{
1597 1598
	struct net_device *dev = rx_ring->netdev;

1599 1600 1601
	ixgbe_update_rsc_stats(rx_ring, skb);

	ixgbe_rx_hash(rx_ring, rx_desc, skb);
A
Alexander Duyck 已提交
1602

1603 1604
	ixgbe_rx_checksum(rx_ring, rx_desc, skb);

1605 1606
	if (unlikely(ixgbe_test_staterr(rx_desc, IXGBE_RXDADV_STAT_TS)))
		ixgbe_ptp_rx_hwtstamp(rx_ring->q_vector->adapter, skb);
1607

1608
	if ((dev->features & NETIF_F_HW_VLAN_CTAG_RX) &&
1609
	    ixgbe_test_staterr(rx_desc, IXGBE_RXD_STAT_VP)) {
1610
		u16 vid = le16_to_cpu(rx_desc->wb.upper.vlan);
1611
		__vlan_hwaccel_put_tag(skb, htons(ETH_P_8021Q), vid);
A
Alexander Duyck 已提交
1612 1613
	}

1614
	skb_record_rx_queue(skb, rx_ring->queue_index);
1615

1616
	skb->protocol = eth_type_trans(skb, dev);
A
Alexander Duyck 已提交
1617 1618
}

1619 1620
static void ixgbe_rx_skb(struct ixgbe_q_vector *q_vector,
			 struct sk_buff *skb)
1621
{
1622
	if (ixgbe_qv_busy_polling(q_vector))
1623
		netif_receive_skb(skb);
1624
	else
1625
		napi_gro_receive(&q_vector->napi, skb);
1626
}
1627

1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650
/**
 * ixgbe_is_non_eop - process handling of non-EOP buffers
 * @rx_ring: Rx ring being processed
 * @rx_desc: Rx descriptor for current buffer
 * @skb: Current socket buffer containing buffer in progress
 *
 * This function updates next to clean.  If the buffer is an EOP buffer
 * this function exits returning false, otherwise it will place the
 * sk_buff in the next buffer to be chained and return true indicating
 * that this is in fact a non-EOP buffer.
 **/
static bool ixgbe_is_non_eop(struct ixgbe_ring *rx_ring,
			     union ixgbe_adv_rx_desc *rx_desc,
			     struct sk_buff *skb)
{
	u32 ntc = rx_ring->next_to_clean + 1;

	/* fetch, update, and store next to clean */
	ntc = (ntc < rx_ring->count) ? ntc : 0;
	rx_ring->next_to_clean = ntc;

	prefetch(IXGBE_RX_DESC(rx_ring, ntc));

1651 1652 1653 1654 1655 1656 1657 1658 1659 1660
	/* update RSC append count if present */
	if (ring_is_rsc_enabled(rx_ring)) {
		__le32 rsc_enabled = rx_desc->wb.lower.lo_dword.data &
				     cpu_to_le32(IXGBE_RXDADV_RSCCNT_MASK);

		if (unlikely(rsc_enabled)) {
			u32 rsc_cnt = le32_to_cpu(rsc_enabled);

			rsc_cnt >>= IXGBE_RXDADV_RSCCNT_SHIFT;
			IXGBE_CB(skb)->append_cnt += rsc_cnt - 1;
1661

1662 1663 1664 1665 1666
			/* update ntc based on RSC value */
			ntc = le32_to_cpu(rx_desc->wb.upper.status_error);
			ntc &= IXGBE_RXDADV_NEXTP_MASK;
			ntc >>= IXGBE_RXDADV_NEXTP_SHIFT;
		}
1667 1668
	}

1669 1670 1671 1672
	/* if we are the last buffer then there is nothing else to do */
	if (likely(ixgbe_test_staterr(rx_desc, IXGBE_RXD_STAT_EOP)))
		return false;

1673 1674 1675 1676 1677 1678 1679
	/* place skb in next buffer to be received */
	rx_ring->rx_buffer_info[ntc].skb = skb;
	rx_ring->rx_stats.non_eop_descs++;

	return true;
}

1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709
/**
 * ixgbe_pull_tail - ixgbe specific version of skb_pull_tail
 * @rx_ring: rx descriptor ring packet is being transacted on
 * @skb: pointer to current skb being adjusted
 *
 * This function is an ixgbe specific version of __pskb_pull_tail.  The
 * main difference between this version and the original function is that
 * this function can make several assumptions about the state of things
 * that allow for significant optimizations versus the standard function.
 * As a result we can do things like drop a frag and maintain an accurate
 * truesize for the skb.
 */
static void ixgbe_pull_tail(struct ixgbe_ring *rx_ring,
			    struct sk_buff *skb)
{
	struct skb_frag_struct *frag = &skb_shinfo(skb)->frags[0];
	unsigned char *va;
	unsigned int pull_len;

	/*
	 * it is valid to use page_address instead of kmap since we are
	 * working with pages allocated out of the lomem pool per
	 * alloc_page(GFP_ATOMIC)
	 */
	va = skb_frag_address(frag);

	/*
	 * we need the header to contain the greater of either ETH_HLEN or
	 * 60 bytes if the skb->len is less than 60 for skb_pad.
	 */
1710
	pull_len = eth_get_headlen(va, IXGBE_RX_HDR_SIZE);
1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721

	/* align pull length to size of long to optimize memcpy performance */
	skb_copy_to_linear_data(skb, va, ALIGN(pull_len, sizeof(long)));

	/* update all of the pointers */
	skb_frag_size_sub(frag, pull_len);
	frag->page_offset += pull_len;
	skb->data_len -= pull_len;
	skb->tail += pull_len;
}

1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751
/**
 * ixgbe_dma_sync_frag - perform DMA sync for first frag of SKB
 * @rx_ring: rx descriptor ring packet is being transacted on
 * @skb: pointer to current skb being updated
 *
 * This function provides a basic DMA sync up for the first fragment of an
 * skb.  The reason for doing this is that the first fragment cannot be
 * unmapped until we have reached the end of packet descriptor for a buffer
 * chain.
 */
static void ixgbe_dma_sync_frag(struct ixgbe_ring *rx_ring,
				struct sk_buff *skb)
{
	/* if the page was released unmap it, else just sync our portion */
	if (unlikely(IXGBE_CB(skb)->page_released)) {
		dma_unmap_page(rx_ring->dev, IXGBE_CB(skb)->dma,
			       ixgbe_rx_pg_size(rx_ring), DMA_FROM_DEVICE);
		IXGBE_CB(skb)->page_released = false;
	} else {
		struct skb_frag_struct *frag = &skb_shinfo(skb)->frags[0];

		dma_sync_single_range_for_cpu(rx_ring->dev,
					      IXGBE_CB(skb)->dma,
					      frag->page_offset,
					      ixgbe_rx_bufsz(rx_ring),
					      DMA_FROM_DEVICE);
	}
	IXGBE_CB(skb)->dma = 0;
}

1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783
/**
 * ixgbe_cleanup_headers - Correct corrupted or empty headers
 * @rx_ring: rx descriptor ring packet is being transacted on
 * @rx_desc: pointer to the EOP Rx descriptor
 * @skb: pointer to current skb being fixed
 *
 * Check for corrupted packet headers caused by senders on the local L2
 * embedded NIC switch not setting up their Tx Descriptors right.  These
 * should be very rare.
 *
 * Also address the case where we are pulling data in on pages only
 * and as such no data is present in the skb header.
 *
 * In addition if skb is not at least 60 bytes we need to pad it so that
 * it is large enough to qualify as a valid Ethernet frame.
 *
 * Returns true if an error was encountered and skb was freed.
 **/
static bool ixgbe_cleanup_headers(struct ixgbe_ring *rx_ring,
				  union ixgbe_adv_rx_desc *rx_desc,
				  struct sk_buff *skb)
{
	struct net_device *netdev = rx_ring->netdev;

	/* verify that the packet does not have any known errors */
	if (unlikely(ixgbe_test_staterr(rx_desc,
					IXGBE_RXDADV_ERR_FRAME_ERR_MASK) &&
	    !(netdev->features & NETIF_F_RXALL))) {
		dev_kfree_skb_any(skb);
		return true;
	}

1784
	/* place header in linear portion of buffer */
1785 1786
	if (skb_is_nonlinear(skb))
		ixgbe_pull_tail(rx_ring, skb);
1787

1788 1789 1790 1791 1792 1793
#ifdef IXGBE_FCOE
	/* do not attempt to pad FCoE Frames as this will disrupt DDP */
	if (ixgbe_rx_is_fcoe(rx_ring, rx_desc))
		return false;

#endif
1794 1795 1796
	/* if eth_skb_pad returns an error the skb was freed */
	if (eth_skb_pad(skb))
		return true;
1797 1798 1799 1800 1801 1802 1803 1804 1805

	return false;
}

/**
 * ixgbe_reuse_rx_page - page flip buffer and store it back on the ring
 * @rx_ring: rx descriptor ring to store buffers on
 * @old_buff: donor buffer to have page reused
 *
1806
 * Synchronizes page for reuse by the adapter
1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820
 **/
static void ixgbe_reuse_rx_page(struct ixgbe_ring *rx_ring,
				struct ixgbe_rx_buffer *old_buff)
{
	struct ixgbe_rx_buffer *new_buff;
	u16 nta = rx_ring->next_to_alloc;

	new_buff = &rx_ring->rx_buffer_info[nta];

	/* update, and store next to alloc */
	nta++;
	rx_ring->next_to_alloc = (nta < rx_ring->count) ? nta : 0;

	/* transfer page from old buffer to new buffer */
A
Alexander Duyck 已提交
1821
	*new_buff = *old_buff;
1822 1823 1824

	/* sync the buffer for use by the device */
	dma_sync_single_range_for_device(rx_ring->dev, new_buff->dma,
1825 1826
					 new_buff->page_offset,
					 ixgbe_rx_bufsz(rx_ring),
1827 1828 1829
					 DMA_FROM_DEVICE);
}

A
Alexander Duyck 已提交
1830 1831 1832 1833 1834
static inline bool ixgbe_page_is_reserved(struct page *page)
{
	return (page_to_nid(page) != numa_mem_id()) || page->pfmemalloc;
}

1835 1836 1837 1838 1839 1840 1841
/**
 * ixgbe_add_rx_frag - Add contents of Rx buffer to sk_buff
 * @rx_ring: rx descriptor ring to transact packets on
 * @rx_buffer: buffer containing page to add
 * @rx_desc: descriptor containing length of buffer written by hardware
 * @skb: sk_buff to place the data into
 *
1842 1843 1844 1845 1846 1847 1848
 * This function will add the data contained in rx_buffer->page to the skb.
 * This is done either through a direct copy if the data in the buffer is
 * less than the skb header size, otherwise it will just attach the page as
 * a frag to the skb.
 *
 * The function will then update the page offset if necessary and return
 * true if the buffer can be reused by the adapter.
1849
 **/
1850
static bool ixgbe_add_rx_frag(struct ixgbe_ring *rx_ring,
1851
			      struct ixgbe_rx_buffer *rx_buffer,
1852 1853
			      union ixgbe_adv_rx_desc *rx_desc,
			      struct sk_buff *skb)
1854
{
1855 1856
	struct page *page = rx_buffer->page;
	unsigned int size = le16_to_cpu(rx_desc->wb.upper.length);
1857
#if (PAGE_SIZE < 8192)
1858
	unsigned int truesize = ixgbe_rx_bufsz(rx_ring);
1859 1860 1861 1862 1863
#else
	unsigned int truesize = ALIGN(size, L1_CACHE_BYTES);
	unsigned int last_offset = ixgbe_rx_pg_size(rx_ring) -
				   ixgbe_rx_bufsz(rx_ring);
#endif
1864

1865 1866 1867 1868 1869
	if ((size <= IXGBE_RX_HDR_SIZE) && !skb_is_nonlinear(skb)) {
		unsigned char *va = page_address(page) + rx_buffer->page_offset;

		memcpy(__skb_put(skb, size), va, ALIGN(size, sizeof(long)));

A
Alexander Duyck 已提交
1870 1871
		/* page is not reserved, we can reuse buffer as-is */
		if (likely(!ixgbe_page_is_reserved(page)))
1872 1873 1874
			return true;

		/* this page cannot be reused so discard it */
A
Alexander Duyck 已提交
1875
		__free_pages(page, ixgbe_rx_pg_order(rx_ring));
1876 1877 1878
		return false;
	}

1879 1880 1881
	skb_add_rx_frag(skb, skb_shinfo(skb)->nr_frags, page,
			rx_buffer->page_offset, size, truesize);

1882
	/* avoid re-using remote pages */
A
Alexander Duyck 已提交
1883
	if (unlikely(ixgbe_page_is_reserved(page)))
1884 1885 1886 1887 1888
		return false;

#if (PAGE_SIZE < 8192)
	/* if we are only owner of page we can reuse it */
	if (unlikely(page_count(page) != 1))
1889 1890 1891 1892
		return false;

	/* flip page offset to other buffer */
	rx_buffer->page_offset ^= truesize;
1893 1894 1895 1896 1897 1898 1899
#else
	/* move offset up to the next cache line */
	rx_buffer->page_offset += truesize;

	if (rx_buffer->page_offset > last_offset)
		return false;
#endif
1900

A
Alexander Duyck 已提交
1901 1902 1903 1904 1905
	/* Even if we own the page, we are not allowed to use atomic_set()
	 * This would break get_page_unless_zero() users.
	 */
	atomic_inc(&page->_count);

1906
	return true;
1907 1908
}

1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932
static struct sk_buff *ixgbe_fetch_rx_buffer(struct ixgbe_ring *rx_ring,
					     union ixgbe_adv_rx_desc *rx_desc)
{
	struct ixgbe_rx_buffer *rx_buffer;
	struct sk_buff *skb;
	struct page *page;

	rx_buffer = &rx_ring->rx_buffer_info[rx_ring->next_to_clean];
	page = rx_buffer->page;
	prefetchw(page);

	skb = rx_buffer->skb;

	if (likely(!skb)) {
		void *page_addr = page_address(page) +
				  rx_buffer->page_offset;

		/* prefetch first cache line of first page */
		prefetch(page_addr);
#if L1_CACHE_BYTES < 128
		prefetch(page_addr + L1_CACHE_BYTES);
#endif

		/* allocate a skb to store the frags */
1933 1934
		skb = napi_alloc_skb(&rx_ring->q_vector->napi,
				     IXGBE_RX_HDR_SIZE);
1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967
		if (unlikely(!skb)) {
			rx_ring->rx_stats.alloc_rx_buff_failed++;
			return NULL;
		}

		/*
		 * we will be copying header into skb->data in
		 * pskb_may_pull so it is in our interest to prefetch
		 * it now to avoid a possible cache miss
		 */
		prefetchw(skb->data);

		/*
		 * Delay unmapping of the first packet. It carries the
		 * header information, HW may still access the header
		 * after the writeback.  Only unmap it when EOP is
		 * reached
		 */
		if (likely(ixgbe_test_staterr(rx_desc, IXGBE_RXD_STAT_EOP)))
			goto dma_sync;

		IXGBE_CB(skb)->dma = rx_buffer->dma;
	} else {
		if (ixgbe_test_staterr(rx_desc, IXGBE_RXD_STAT_EOP))
			ixgbe_dma_sync_frag(rx_ring, skb);

dma_sync:
		/* we are reusing so sync this buffer for CPU use */
		dma_sync_single_range_for_cpu(rx_ring->dev,
					      rx_buffer->dma,
					      rx_buffer->page_offset,
					      ixgbe_rx_bufsz(rx_ring),
					      DMA_FROM_DEVICE);
A
Alexander Duyck 已提交
1968 1969

		rx_buffer->skb = NULL;
1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989
	}

	/* pull page into skb */
	if (ixgbe_add_rx_frag(rx_ring, rx_buffer, rx_desc, skb)) {
		/* hand second half of page back to the ring */
		ixgbe_reuse_rx_page(rx_ring, rx_buffer);
	} else if (IXGBE_CB(skb)->dma == rx_buffer->dma) {
		/* the page has been released from the ring */
		IXGBE_CB(skb)->page_released = true;
	} else {
		/* we are not reusing the buffer so unmap it */
		dma_unmap_page(rx_ring->dev, rx_buffer->dma,
			       ixgbe_rx_pg_size(rx_ring),
			       DMA_FROM_DEVICE);
	}

	/* clear contents of buffer_info */
	rx_buffer->page = NULL;

	return skb;
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002
}

/**
 * ixgbe_clean_rx_irq - Clean completed descriptors from Rx ring - bounce buf
 * @q_vector: structure containing interrupt and ring information
 * @rx_ring: rx descriptor ring to transact packets on
 * @budget: Total limit on number of packets to process
 *
 * This function provides a "bounce buffer" approach to Rx interrupt
 * processing.  The advantage to this is that on systems that have
 * expensive overhead for IOMMU access this provides a means of avoiding
 * it by maintaining the mapping of the page to the syste.
 *
2003
 * Returns amount of work completed
2004
 **/
2005
static int ixgbe_clean_rx_irq(struct ixgbe_q_vector *q_vector,
2006
			       struct ixgbe_ring *rx_ring,
2007
			       const int budget)
2008
{
2009
	unsigned int total_rx_bytes = 0, total_rx_packets = 0;
B
Ben Greear 已提交
2010
#ifdef IXGBE_FCOE
2011
	struct ixgbe_adapter *adapter = q_vector->adapter;
2012 2013
	int ddp_bytes;
	unsigned int mss = 0;
2014
#endif /* IXGBE_FCOE */
2015
	u16 cleaned_count = ixgbe_desc_unused(rx_ring);
2016

2017
	while (likely(total_rx_packets < budget)) {
2018 2019 2020 2021 2022 2023 2024 2025 2026
		union ixgbe_adv_rx_desc *rx_desc;
		struct sk_buff *skb;

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

2027
		rx_desc = IXGBE_RX_DESC(rx_ring, rx_ring->next_to_clean);
2028

2029
		if (!rx_desc->wb.upper.status_error)
2030
			break;
2031

2032
		/* This memory barrier is needed to keep us from reading
2033
		 * any other fields out of the rx_desc until we know the
2034
		 * descriptor has been written back
2035
		 */
2036
		dma_rmb();
2037

2038 2039
		/* retrieve a buffer from the ring */
		skb = ixgbe_fetch_rx_buffer(rx_ring, rx_desc);
2040

2041 2042 2043
		/* exit if we failed to retrieve a buffer */
		if (!skb)
			break;
2044 2045

		cleaned_count++;
A
Alexander Duyck 已提交
2046

2047 2048 2049
		/* place incomplete frames back on ring for completion */
		if (ixgbe_is_non_eop(rx_ring, rx_desc, skb))
			continue;
2050

2051 2052 2053
		/* verify the packet layout is correct */
		if (ixgbe_cleanup_headers(rx_ring, rx_desc, skb))
			continue;
2054

2055 2056 2057
		/* probably a little skewed due to removing CRC */
		total_rx_bytes += skb->len;

2058 2059 2060
		/* populate checksum, timestamp, VLAN, and protocol */
		ixgbe_process_skb_fields(rx_ring, rx_desc, skb);

2061 2062
#ifdef IXGBE_FCOE
		/* if ddp, not passing to ULD unless for FCP_RSP or error */
2063
		if (ixgbe_rx_is_fcoe(rx_ring, rx_desc)) {
2064
			ddp_bytes = ixgbe_fcoe_ddp(adapter, rx_desc, skb);
2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078
			/* include DDPed FCoE data */
			if (ddp_bytes > 0) {
				if (!mss) {
					mss = rx_ring->netdev->mtu -
						sizeof(struct fcoe_hdr) -
						sizeof(struct fc_frame_header) -
						sizeof(struct fcoe_crc_eof);
					if (mss > 512)
						mss &= ~511;
				}
				total_rx_bytes += ddp_bytes;
				total_rx_packets += DIV_ROUND_UP(ddp_bytes,
								 mss);
			}
2079 2080
			if (!ddp_bytes) {
				dev_kfree_skb_any(skb);
2081
				continue;
2082
			}
2083
		}
2084

2085
#endif /* IXGBE_FCOE */
2086
		skb_mark_napi_id(skb, &q_vector->napi);
2087
		ixgbe_rx_skb(q_vector, skb);
2088

2089
		/* update budget accounting */
2090
		total_rx_packets++;
2091
	}
2092

2093 2094 2095 2096
	u64_stats_update_begin(&rx_ring->syncp);
	rx_ring->stats.packets += total_rx_packets;
	rx_ring->stats.bytes += total_rx_bytes;
	u64_stats_update_end(&rx_ring->syncp);
2097 2098
	q_vector->rx.total_packets += total_rx_packets;
	q_vector->rx.total_bytes += total_rx_bytes;
2099

2100
	return total_rx_packets;
2101 2102
}

2103
#ifdef CONFIG_NET_RX_BUSY_POLL
2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120
/* must be called with local_bh_disable()d */
static int ixgbe_low_latency_recv(struct napi_struct *napi)
{
	struct ixgbe_q_vector *q_vector =
			container_of(napi, struct ixgbe_q_vector, napi);
	struct ixgbe_adapter *adapter = q_vector->adapter;
	struct ixgbe_ring  *ring;
	int found = 0;

	if (test_bit(__IXGBE_DOWN, &adapter->state))
		return LL_FLUSH_FAILED;

	if (!ixgbe_qv_lock_poll(q_vector))
		return LL_FLUSH_BUSY;

	ixgbe_for_each_ring(ring, q_vector->rx) {
		found = ixgbe_clean_rx_irq(q_vector, ring, 4);
2121
#ifdef BP_EXTENDED_STATS
2122 2123 2124 2125 2126
		if (found)
			ring->stats.cleaned += found;
		else
			ring->stats.misses++;
#endif
2127 2128 2129 2130 2131 2132 2133 2134
		if (found)
			break;
	}

	ixgbe_qv_unlock_poll(q_vector);

	return found;
}
2135
#endif	/* CONFIG_NET_RX_BUSY_POLL */
2136

2137 2138 2139 2140 2141 2142 2143 2144 2145
/**
 * 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)
{
2146
	struct ixgbe_q_vector *q_vector;
2147
	int v_idx;
2148
	u32 mask;
2149

2150 2151 2152 2153 2154 2155
	/* Populate MSIX to EITR Select */
	if (adapter->num_vfs > 32) {
		u32 eitrsel = (1 << (adapter->num_vfs - 32)) - 1;
		IXGBE_WRITE_REG(&adapter->hw, IXGBE_EITRSEL, eitrsel);
	}

2156 2157
	/*
	 * Populate the IVAR table and set the ITR values to the
2158 2159
	 * corresponding register.
	 */
2160
	for (v_idx = 0; v_idx < adapter->num_q_vectors; v_idx++) {
2161
		struct ixgbe_ring *ring;
2162
		q_vector = adapter->q_vector[v_idx];
2163

2164
		ixgbe_for_each_ring(ring, q_vector->rx)
2165 2166
			ixgbe_set_ivar(adapter, 0, ring->reg_idx, v_idx);

2167
		ixgbe_for_each_ring(ring, q_vector->tx)
2168 2169
			ixgbe_set_ivar(adapter, 1, ring->reg_idx, v_idx);

2170
		ixgbe_write_eitr(q_vector);
2171 2172
	}

2173 2174
	switch (adapter->hw.mac.type) {
	case ixgbe_mac_82598EB:
2175
		ixgbe_set_ivar(adapter, -1, IXGBE_IVAR_OTHER_CAUSES_INDEX,
2176
			       v_idx);
2177 2178
		break;
	case ixgbe_mac_82599EB:
D
Don Skidmore 已提交
2179
	case ixgbe_mac_X540:
2180 2181
	case ixgbe_mac_X550:
	case ixgbe_mac_X550EM_x:
2182
		ixgbe_set_ivar(adapter, -1, 1, v_idx);
2183 2184 2185 2186
		break;
	default:
		break;
	}
2187 2188
	IXGBE_WRITE_REG(&adapter->hw, IXGBE_EITR(v_idx), 1950);

2189
	/* set up to autoclear timer, and the vectors */
2190
	mask = IXGBE_EIMS_ENABLE_MASK;
2191 2192 2193 2194
	mask &= ~(IXGBE_EIMS_OTHER |
		  IXGBE_EIMS_MAILBOX |
		  IXGBE_EIMS_LSC);

2195
	IXGBE_WRITE_REG(&adapter->hw, IXGBE_EIAC, mask);
2196 2197
}

2198 2199 2200 2201 2202 2203 2204 2205 2206
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
2207 2208
 * @q_vector: structure containing interrupt and ring information
 * @ring_container: structure containing ring performance data
2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219
 *
 *      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)
 **/
2220 2221
static void ixgbe_update_itr(struct ixgbe_q_vector *q_vector,
			     struct ixgbe_ring_container *ring_container)
2222
{
2223 2224 2225
	int bytes = ring_container->total_bytes;
	int packets = ring_container->total_packets;
	u32 timepassed_us;
2226
	u64 bytes_perint;
2227
	u8 itr_setting = ring_container->itr;
2228 2229

	if (packets == 0)
2230
		return;
2231 2232

	/* simple throttlerate management
2233 2234 2235
	 *   0-10MB/s   lowest (100000 ints/s)
	 *  10-20MB/s   low    (20000 ints/s)
	 *  20-1249MB/s bulk   (8000 ints/s)
2236 2237
	 */
	/* what was last interrupt timeslice? */
2238
	timepassed_us = q_vector->itr >> 2;
2239 2240 2241
	if (timepassed_us == 0)
		return;

2242 2243 2244 2245
	bytes_perint = bytes / timepassed_us; /* bytes/usec */

	switch (itr_setting) {
	case lowest_latency:
2246
		if (bytes_perint > 10)
2247
			itr_setting = low_latency;
2248 2249
		break;
	case low_latency:
2250
		if (bytes_perint > 20)
2251
			itr_setting = bulk_latency;
2252
		else if (bytes_perint <= 10)
2253
			itr_setting = lowest_latency;
2254 2255
		break;
	case bulk_latency:
2256
		if (bytes_perint <= 20)
2257
			itr_setting = low_latency;
2258 2259 2260
		break;
	}

2261 2262 2263 2264 2265 2266
	/* clear work counters since we have the values we need */
	ring_container->total_bytes = 0;
	ring_container->total_packets = 0;

	/* write updated itr to ring container */
	ring_container->itr = itr_setting;
2267 2268
}

2269 2270
/**
 * ixgbe_write_eitr - write EITR register in hardware specific way
2271
 * @q_vector: structure containing interrupt and ring information
2272 2273 2274 2275 2276
 *
 * 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.
 */
2277
void ixgbe_write_eitr(struct ixgbe_q_vector *q_vector)
2278
{
2279
	struct ixgbe_adapter *adapter = q_vector->adapter;
2280
	struct ixgbe_hw *hw = &adapter->hw;
2281
	int v_idx = q_vector->v_idx;
2282
	u32 itr_reg = q_vector->itr & IXGBE_MAX_EITR;
2283

2284 2285
	switch (adapter->hw.mac.type) {
	case ixgbe_mac_82598EB:
2286 2287
		/* must write high and low 16 bits to reset counter */
		itr_reg |= (itr_reg << 16);
2288 2289
		break;
	case ixgbe_mac_82599EB:
D
Don Skidmore 已提交
2290
	case ixgbe_mac_X540:
2291 2292
	case ixgbe_mac_X550:
	case ixgbe_mac_X550EM_x:
2293 2294 2295 2296 2297
		/*
		 * set the WDIS bit to not clear the timer bits and cause an
		 * immediate assertion of the interrupt
		 */
		itr_reg |= IXGBE_EITR_CNT_WDIS;
2298 2299 2300
		break;
	default:
		break;
2301 2302 2303 2304
	}
	IXGBE_WRITE_REG(hw, IXGBE_EITR(v_idx), itr_reg);
}

2305
static void ixgbe_set_itr(struct ixgbe_q_vector *q_vector)
2306
{
2307
	u32 new_itr = q_vector->itr;
2308
	u8 current_itr;
2309

2310 2311
	ixgbe_update_itr(q_vector, &q_vector->tx);
	ixgbe_update_itr(q_vector, &q_vector->rx);
2312

2313
	current_itr = max(q_vector->rx.itr, q_vector->tx.itr);
2314 2315 2316 2317

	switch (current_itr) {
	/* counts and packets in update_itr are dependent on these numbers */
	case lowest_latency:
2318
		new_itr = IXGBE_100K_ITR;
2319 2320
		break;
	case low_latency:
2321
		new_itr = IXGBE_20K_ITR;
2322 2323
		break;
	case bulk_latency:
2324
		new_itr = IXGBE_8K_ITR;
2325
		break;
2326 2327
	default:
		break;
2328 2329
	}

2330
	if (new_itr != q_vector->itr) {
2331
		/* do an exponential smoothing */
2332 2333
		new_itr = (10 * new_itr * q_vector->itr) /
			  ((9 * new_itr) + q_vector->itr);
2334

2335
		/* save the algorithm value here */
2336
		q_vector->itr = new_itr;
2337 2338

		ixgbe_write_eitr(q_vector);
2339 2340 2341
	}
}

2342
/**
2343
 * ixgbe_check_overtemp_subtask - check for over temperature
2344
 * @adapter: pointer to adapter
2345
 **/
2346
static void ixgbe_check_overtemp_subtask(struct ixgbe_adapter *adapter)
2347 2348 2349 2350
{
	struct ixgbe_hw *hw = &adapter->hw;
	u32 eicr = adapter->interrupt_event;

2351
	if (test_bit(__IXGBE_DOWN, &adapter->state))
2352 2353
		return;

2354 2355 2356 2357 2358 2359
	if (!(adapter->flags2 & IXGBE_FLAG2_TEMP_SENSOR_CAPABLE) &&
	    !(adapter->flags2 & IXGBE_FLAG2_TEMP_SENSOR_EVENT))
		return;

	adapter->flags2 &= ~IXGBE_FLAG2_TEMP_SENSOR_EVENT;

2360
	switch (hw->device_id) {
2361 2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373
	case IXGBE_DEV_ID_82599_T3_LOM:
		/*
		 * Since the warning interrupt is for both ports
		 * we don't have to check if:
		 *  - This interrupt wasn't for our port.
		 *  - We may have missed the interrupt so always have to
		 *    check if we  got a LSC
		 */
		if (!(eicr & IXGBE_EICR_GPI_SDP0) &&
		    !(eicr & IXGBE_EICR_LSC))
			return;

		if (!(eicr & IXGBE_EICR_LSC) && hw->mac.ops.check_link) {
J
Josh Hay 已提交
2374
			u32 speed;
2375
			bool link_up = false;
2376

J
Josh Hay 已提交
2377
			hw->mac.ops.check_link(hw, &speed, &link_up, false);
2378

2379 2380 2381 2382 2383 2384 2385 2386 2387
			if (link_up)
				return;
		}

		/* Check if this is not due to overtemp */
		if (hw->phy.ops.check_overtemp(hw) != IXGBE_ERR_OVERTEMP)
			return;

		break;
2388 2389
	default:
		if (!(eicr & IXGBE_EICR_GPI_SDP0))
2390
			return;
2391
		break;
2392
	}
2393 2394 2395 2396
	e_crit(drv,
	       "Network adapter has been stopped because it has over heated. "
	       "Restart the computer. If the problem persists, "
	       "power off the system and replace the adapter\n");
2397 2398

	adapter->interrupt_event = 0;
2399 2400
}

2401 2402 2403 2404 2405 2406
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)) {
2407
		e_crit(probe, "Fan has stopped, replace the adapter\n");
2408 2409 2410 2411
		/* write to clear the interrupt */
		IXGBE_WRITE_REG(hw, IXGBE_EICR, IXGBE_EICR_GPI_SDP1);
	}
}
2412

2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435 2436 2437 2438 2439 2440 2441 2442 2443 2444 2445
static void ixgbe_check_overtemp_event(struct ixgbe_adapter *adapter, u32 eicr)
{
	if (!(adapter->flags2 & IXGBE_FLAG2_TEMP_SENSOR_CAPABLE))
		return;

	switch (adapter->hw.mac.type) {
	case ixgbe_mac_82599EB:
		/*
		 * Need to check link state so complete overtemp check
		 * on service task
		 */
		if (((eicr & IXGBE_EICR_GPI_SDP0) || (eicr & IXGBE_EICR_LSC)) &&
		    (!test_bit(__IXGBE_DOWN, &adapter->state))) {
			adapter->interrupt_event = eicr;
			adapter->flags2 |= IXGBE_FLAG2_TEMP_SENSOR_EVENT;
			ixgbe_service_event_schedule(adapter);
			return;
		}
		return;
	case ixgbe_mac_X540:
		if (!(eicr & IXGBE_EICR_TS))
			return;
		break;
	default:
		return;
	}

	e_crit(drv,
	       "Network adapter has been stopped because it has over heated. "
	       "Restart the computer. If the problem persists, "
	       "power off the system and replace the adapter\n");
}

2446 2447 2448 2449
static void ixgbe_check_sfp_event(struct ixgbe_adapter *adapter, u32 eicr)
{
	struct ixgbe_hw *hw = &adapter->hw;

2450 2451 2452
	if (eicr & IXGBE_EICR_GPI_SDP2) {
		/* Clear the interrupt */
		IXGBE_WRITE_REG(hw, IXGBE_EICR, IXGBE_EICR_GPI_SDP2);
2453 2454 2455 2456
		if (!test_bit(__IXGBE_DOWN, &adapter->state)) {
			adapter->flags2 |= IXGBE_FLAG2_SFP_NEEDS_RESET;
			ixgbe_service_event_schedule(adapter);
		}
2457 2458
	}

2459 2460 2461
	if (eicr & IXGBE_EICR_GPI_SDP1) {
		/* Clear the interrupt */
		IXGBE_WRITE_REG(hw, IXGBE_EICR, IXGBE_EICR_GPI_SDP1);
2462 2463 2464 2465
		if (!test_bit(__IXGBE_DOWN, &adapter->state)) {
			adapter->flags |= IXGBE_FLAG_NEED_LINK_CONFIG;
			ixgbe_service_event_schedule(adapter);
		}
2466 2467 2468
	}
}

2469 2470 2471 2472 2473 2474 2475 2476 2477
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);
2478
		IXGBE_WRITE_FLUSH(hw);
2479
		ixgbe_service_event_schedule(adapter);
2480 2481 2482
	}
}

2483 2484 2485 2486
static inline void ixgbe_irq_enable_queues(struct ixgbe_adapter *adapter,
					   u64 qmask)
{
	u32 mask;
2487
	struct ixgbe_hw *hw = &adapter->hw;
2488

2489 2490
	switch (hw->mac.type) {
	case ixgbe_mac_82598EB:
2491
		mask = (IXGBE_EIMS_RTX_QUEUE & qmask);
2492 2493 2494
		IXGBE_WRITE_REG(hw, IXGBE_EIMS, mask);
		break;
	case ixgbe_mac_82599EB:
D
Don Skidmore 已提交
2495
	case ixgbe_mac_X540:
2496 2497
	case ixgbe_mac_X550:
	case ixgbe_mac_X550EM_x:
2498
		mask = (qmask & 0xFFFFFFFF);
2499 2500
		if (mask)
			IXGBE_WRITE_REG(hw, IXGBE_EIMS_EX(0), mask);
2501
		mask = (qmask >> 32);
2502 2503 2504 2505 2506
		if (mask)
			IXGBE_WRITE_REG(hw, IXGBE_EIMS_EX(1), mask);
		break;
	default:
		break;
2507 2508 2509 2510 2511
	}
	/* skip the flush */
}

static inline void ixgbe_irq_disable_queues(struct ixgbe_adapter *adapter,
2512
					    u64 qmask)
2513 2514
{
	u32 mask;
2515
	struct ixgbe_hw *hw = &adapter->hw;
2516

2517 2518
	switch (hw->mac.type) {
	case ixgbe_mac_82598EB:
2519
		mask = (IXGBE_EIMS_RTX_QUEUE & qmask);
2520 2521 2522
		IXGBE_WRITE_REG(hw, IXGBE_EIMC, mask);
		break;
	case ixgbe_mac_82599EB:
D
Don Skidmore 已提交
2523
	case ixgbe_mac_X540:
2524 2525
	case ixgbe_mac_X550:
	case ixgbe_mac_X550EM_x:
2526
		mask = (qmask & 0xFFFFFFFF);
2527 2528
		if (mask)
			IXGBE_WRITE_REG(hw, IXGBE_EIMC_EX(0), mask);
2529
		mask = (qmask >> 32);
2530 2531 2532 2533 2534
		if (mask)
			IXGBE_WRITE_REG(hw, IXGBE_EIMC_EX(1), mask);
		break;
	default:
		break;
2535 2536 2537 2538
	}
	/* skip the flush */
}

2539
/**
2540 2541
 * ixgbe_irq_enable - Enable default interrupt generation settings
 * @adapter: board private structure
2542
 **/
2543 2544
static inline void ixgbe_irq_enable(struct ixgbe_adapter *adapter, bool queues,
				    bool flush)
2545
{
2546
	u32 mask = (IXGBE_EIMS_ENABLE_MASK & ~IXGBE_EIMS_RTX_QUEUE);
2547

2548 2549 2550
	/* don't reenable LSC while waiting for link */
	if (adapter->flags & IXGBE_FLAG_NEED_LINK_UPDATE)
		mask &= ~IXGBE_EIMS_LSC;
2551

2552
	if (adapter->flags2 & IXGBE_FLAG2_TEMP_SENSOR_CAPABLE)
2553 2554 2555 2556 2557
		switch (adapter->hw.mac.type) {
		case ixgbe_mac_82599EB:
			mask |= IXGBE_EIMS_GPI_SDP0;
			break;
		case ixgbe_mac_X540:
2558 2559
		case ixgbe_mac_X550:
		case ixgbe_mac_X550EM_x:
2560 2561 2562 2563 2564
			mask |= IXGBE_EIMS_TS;
			break;
		default:
			break;
		}
2565 2566 2567 2568 2569 2570
	if (adapter->flags & IXGBE_FLAG_FAN_FAIL_CAPABLE)
		mask |= IXGBE_EIMS_GPI_SDP1;
	switch (adapter->hw.mac.type) {
	case ixgbe_mac_82599EB:
		mask |= IXGBE_EIMS_GPI_SDP1;
		mask |= IXGBE_EIMS_GPI_SDP2;
2571
		/* fall through */
2572
	case ixgbe_mac_X540:
2573 2574
	case ixgbe_mac_X550:
	case ixgbe_mac_X550EM_x:
2575
		mask |= IXGBE_EIMS_ECC;
2576 2577 2578 2579
		mask |= IXGBE_EIMS_MAILBOX;
		break;
	default:
		break;
2580
	}
J
Jacob Keller 已提交
2581

2582 2583 2584
	if ((adapter->flags & IXGBE_FLAG_FDIR_HASH_CAPABLE) &&
	    !(adapter->flags2 & IXGBE_FLAG2_FDIR_REQUIRES_REINIT))
		mask |= IXGBE_EIMS_FLOW_DIR;
2585

2586 2587 2588 2589 2590
	IXGBE_WRITE_REG(&adapter->hw, IXGBE_EIMS, mask);
	if (queues)
		ixgbe_irq_enable_queues(adapter, ~0);
	if (flush)
		IXGBE_WRITE_FLUSH(&adapter->hw);
2591 2592
}

2593
static irqreturn_t ixgbe_msix_other(int irq, void *data)
2594
{
2595
	struct ixgbe_adapter *adapter = data;
2596
	struct ixgbe_hw *hw = &adapter->hw;
2597
	u32 eicr;
2598

2599 2600 2601 2602 2603 2604 2605
	/*
	 * 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);
2606 2607

	/* The lower 16bits of the EICR register are for the queue interrupts
2608
	 * which should be masked here in order to not accidentally clear them if
2609 2610 2611 2612 2613 2614 2615
	 * the bits are high when ixgbe_msix_other is called. There is a race
	 * condition otherwise which results in possible performance loss
	 * especially if the ixgbe_msix_other interrupt is triggering
	 * consistently (as it would when PPS is turned on for the X540 device)
	 */
	eicr &= 0xFFFF0000;

2616
	IXGBE_WRITE_REG(hw, IXGBE_EICR, eicr);
2617

2618 2619
	if (eicr & IXGBE_EICR_LSC)
		ixgbe_check_lsc(adapter);
2620

2621 2622
	if (eicr & IXGBE_EICR_MAILBOX)
		ixgbe_msg_task(adapter);
2623

2624 2625
	switch (hw->mac.type) {
	case ixgbe_mac_82599EB:
D
Don Skidmore 已提交
2626
	case ixgbe_mac_X540:
2627 2628
	case ixgbe_mac_X550:
	case ixgbe_mac_X550EM_x:
2629 2630 2631 2632 2633 2634
		if (eicr & IXGBE_EICR_ECC) {
			e_info(link, "Received ECC Err, initiating reset\n");
			adapter->flags2 |= IXGBE_FLAG2_RESET_REQUESTED;
			ixgbe_service_event_schedule(adapter);
			IXGBE_WRITE_REG(hw, IXGBE_EICR, IXGBE_EICR_ECC);
		}
2635 2636
		/* Handle Flow Director Full threshold interrupt */
		if (eicr & IXGBE_EICR_FLOW_DIR) {
2637
			int reinit_count = 0;
2638 2639
			int i;
			for (i = 0; i < adapter->num_tx_queues; i++) {
2640
				struct ixgbe_ring *ring = adapter->tx_ring[i];
A
Alexander Duyck 已提交
2641
				if (test_and_clear_bit(__IXGBE_TX_FDIR_INIT_DONE,
2642 2643 2644 2645 2646 2647 2648 2649
						       &ring->state))
					reinit_count++;
			}
			if (reinit_count) {
				/* no more flow director interrupts until after init */
				IXGBE_WRITE_REG(hw, IXGBE_EIMC, IXGBE_EIMC_FLOW_DIR);
				adapter->flags2 |= IXGBE_FLAG2_FDIR_REQUIRES_REINIT;
				ixgbe_service_event_schedule(adapter);
2650 2651
			}
		}
2652
		ixgbe_check_sfp_event(adapter, eicr);
2653
		ixgbe_check_overtemp_event(adapter, eicr);
2654 2655 2656
		break;
	default:
		break;
2657
	}
2658

2659
	ixgbe_check_fan_failure(adapter, eicr);
J
Jacob Keller 已提交
2660 2661 2662

	if (unlikely(eicr & IXGBE_EICR_TIMESYNC))
		ixgbe_ptp_check_pps_event(adapter, eicr);
2663

2664
	/* re-enable the original interrupt state, no lsc, no queues */
2665
	if (!test_bit(__IXGBE_DOWN, &adapter->state))
2666
		ixgbe_irq_enable(adapter, false, false);
2667

2668
	return IRQ_HANDLED;
2669
}
2670

2671
static irqreturn_t ixgbe_msix_clean_rings(int irq, void *data)
2672
{
2673
	struct ixgbe_q_vector *q_vector = data;
2674

2675
	/* EIAM disabled interrupts (on this vector) for us */
2676

2677 2678
	if (q_vector->rx.ring || q_vector->tx.ring)
		napi_schedule(&q_vector->napi);
2679

2680
	return IRQ_HANDLED;
2681 2682
}

2683 2684 2685 2686 2687 2688 2689
/**
 * 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
 **/
2690
int ixgbe_poll(struct napi_struct *napi, int budget)
2691 2692 2693 2694 2695 2696 2697 2698 2699 2700 2701 2702 2703 2704 2705 2706
{
	struct ixgbe_q_vector *q_vector =
				container_of(napi, struct ixgbe_q_vector, napi);
	struct ixgbe_adapter *adapter = q_vector->adapter;
	struct ixgbe_ring *ring;
	int per_ring_budget;
	bool clean_complete = true;

#ifdef CONFIG_IXGBE_DCA
	if (adapter->flags & IXGBE_FLAG_DCA_ENABLED)
		ixgbe_update_dca(q_vector);
#endif

	ixgbe_for_each_ring(ring, q_vector->tx)
		clean_complete &= !!ixgbe_clean_tx_irq(q_vector, ring);

2707 2708 2709
	if (!ixgbe_qv_lock_napi(q_vector))
		return budget;

2710 2711 2712 2713 2714 2715 2716 2717
	/* attempt to distribute budget to each queue fairly, but don't allow
	 * the budget to go below 1 because we'll exit polling */
	if (q_vector->rx.count > 1)
		per_ring_budget = max(budget/q_vector->rx.count, 1);
	else
		per_ring_budget = budget;

	ixgbe_for_each_ring(ring, q_vector->rx)
2718 2719
		clean_complete &= (ixgbe_clean_rx_irq(q_vector, ring,
				   per_ring_budget) < per_ring_budget);
2720

2721
	ixgbe_qv_unlock_napi(q_vector);
2722 2723 2724 2725 2726 2727 2728 2729 2730 2731 2732 2733 2734 2735
	/* If all work not completed, return budget and keep polling */
	if (!clean_complete)
		return budget;

	/* all work done, exit the polling mode */
	napi_complete(napi);
	if (adapter->rx_itr_setting & 1)
		ixgbe_set_itr(q_vector);
	if (!test_bit(__IXGBE_DOWN, &adapter->state))
		ixgbe_irq_enable_queues(adapter, ((u64)1 << q_vector->v_idx));

	return 0;
}

2736 2737 2738 2739 2740 2741 2742 2743 2744 2745
/**
 * 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;
2746
	int vector, err;
2747
	int ri = 0, ti = 0;
2748

2749
	for (vector = 0; vector < adapter->num_q_vectors; vector++) {
2750
		struct ixgbe_q_vector *q_vector = adapter->q_vector[vector];
2751
		struct msix_entry *entry = &adapter->msix_entries[vector];
R
Robert Olsson 已提交
2752

2753
		if (q_vector->tx.ring && q_vector->rx.ring) {
2754
			snprintf(q_vector->name, sizeof(q_vector->name) - 1,
2755 2756 2757
				 "%s-%s-%d", netdev->name, "TxRx", ri++);
			ti++;
		} else if (q_vector->rx.ring) {
2758
			snprintf(q_vector->name, sizeof(q_vector->name) - 1,
2759 2760
				 "%s-%s-%d", netdev->name, "rx", ri++);
		} else if (q_vector->tx.ring) {
2761
			snprintf(q_vector->name, sizeof(q_vector->name) - 1,
2762
				 "%s-%s-%d", netdev->name, "tx", ti++);
2763 2764 2765
		} else {
			/* skip this unused q_vector */
			continue;
2766
		}
2767 2768
		err = request_irq(entry->vector, &ixgbe_msix_clean_rings, 0,
				  q_vector->name, q_vector);
2769
		if (err) {
2770
			e_err(probe, "request_irq failed for MSIX interrupt "
2771
			      "Error: %d\n", err);
2772
			goto free_queue_irqs;
2773
		}
2774 2775 2776 2777
		/* If Flow Director is enabled, set interrupt affinity */
		if (adapter->flags & IXGBE_FLAG_FDIR_HASH_CAPABLE) {
			/* assign the mask for this irq */
			irq_set_affinity_hint(entry->vector,
2778
					      &q_vector->affinity_mask);
2779
		}
2780 2781
	}

2782
	err = request_irq(adapter->msix_entries[vector].vector,
2783
			  ixgbe_msix_other, 0, netdev->name, adapter);
2784
	if (err) {
2785
		e_err(probe, "request_irq for msix_other failed: %d\n", err);
2786
		goto free_queue_irqs;
2787 2788 2789 2790
	}

	return 0;

2791
free_queue_irqs:
2792 2793 2794 2795 2796 2797 2798
	while (vector) {
		vector--;
		irq_set_affinity_hint(adapter->msix_entries[vector].vector,
				      NULL);
		free_irq(adapter->msix_entries[vector].vector,
			 adapter->q_vector[vector]);
	}
2799 2800
	adapter->flags &= ~IXGBE_FLAG_MSIX_ENABLED;
	pci_disable_msix(adapter->pdev);
2801 2802 2803 2804 2805 2806
	kfree(adapter->msix_entries);
	adapter->msix_entries = NULL;
	return err;
}

/**
2807
 * ixgbe_intr - legacy mode Interrupt Handler
2808 2809 2810 2811 2812
 * @irq: interrupt number
 * @data: pointer to a network interface device structure
 **/
static irqreturn_t ixgbe_intr(int irq, void *data)
{
2813
	struct ixgbe_adapter *adapter = data;
2814
	struct ixgbe_hw *hw = &adapter->hw;
2815
	struct ixgbe_q_vector *q_vector = adapter->q_vector[0];
2816 2817
	u32 eicr;

2818
	/*
2819
	 * Workaround for silicon errata #26 on 82598.  Mask the interrupt
2820 2821 2822 2823
	 * before the read of EICR.
	 */
	IXGBE_WRITE_REG(hw, IXGBE_EIMC, IXGBE_IRQ_CLEAR_MASK);

2824
	/* for NAPI, using EIAM to auto-mask tx/rx interrupt bits on read
S
Stephen Hemminger 已提交
2825
	 * therefore no explicit interrupt disable is necessary */
2826
	eicr = IXGBE_READ_REG(hw, IXGBE_EICR);
2827
	if (!eicr) {
2828 2829
		/*
		 * shared interrupt alert!
2830
		 * make sure interrupts are enabled because the read will
2831 2832 2833 2834 2835 2836
		 * have disabled interrupts due to EIAM
		 * finish the workaround of silicon errata on 82598.  Unmask
		 * the interrupt that we masked before the EICR read.
		 */
		if (!test_bit(__IXGBE_DOWN, &adapter->state))
			ixgbe_irq_enable(adapter, true, true);
2837
		return IRQ_NONE;	/* Not our interrupt */
2838
	}
2839

2840 2841
	if (eicr & IXGBE_EICR_LSC)
		ixgbe_check_lsc(adapter);
2842

2843 2844
	switch (hw->mac.type) {
	case ixgbe_mac_82599EB:
2845
		ixgbe_check_sfp_event(adapter, eicr);
2846 2847
		/* Fall through */
	case ixgbe_mac_X540:
2848 2849
	case ixgbe_mac_X550:
	case ixgbe_mac_X550EM_x:
2850 2851 2852 2853 2854 2855
		if (eicr & IXGBE_EICR_ECC) {
			e_info(link, "Received ECC Err, initiating reset\n");
			adapter->flags2 |= IXGBE_FLAG2_RESET_REQUESTED;
			ixgbe_service_event_schedule(adapter);
			IXGBE_WRITE_REG(hw, IXGBE_EICR, IXGBE_EICR_ECC);
		}
2856
		ixgbe_check_overtemp_event(adapter, eicr);
2857 2858 2859 2860
		break;
	default:
		break;
	}
2861

2862
	ixgbe_check_fan_failure(adapter, eicr);
J
Jacob Keller 已提交
2863 2864
	if (unlikely(eicr & IXGBE_EICR_TIMESYNC))
		ixgbe_ptp_check_pps_event(adapter, eicr);
2865

2866 2867
	/* would disable interrupts here but EIAM disabled it */
	napi_schedule(&q_vector->napi);
2868

2869 2870 2871 2872 2873 2874 2875
	/*
	 * re-enable link(maybe) and non-queue interrupts, no flush.
	 * ixgbe_poll will re-enable the queue interrupts
	 */
	if (!test_bit(__IXGBE_DOWN, &adapter->state))
		ixgbe_irq_enable(adapter, false, false);

2876 2877 2878 2879 2880 2881 2882 2883 2884 2885
	return IRQ_HANDLED;
}

/**
 * ixgbe_request_irq - initialize interrupts
 * @adapter: board private structure
 *
 * Attempts to configure interrupts using the best available
 * capabilities of the hardware and kernel.
 **/
2886
static int ixgbe_request_irq(struct ixgbe_adapter *adapter)
2887 2888
{
	struct net_device *netdev = adapter->netdev;
2889
	int err;
2890

2891
	if (adapter->flags & IXGBE_FLAG_MSIX_ENABLED)
2892
		err = ixgbe_request_msix_irqs(adapter);
2893
	else if (adapter->flags & IXGBE_FLAG_MSI_ENABLED)
2894
		err = request_irq(adapter->pdev->irq, ixgbe_intr, 0,
2895
				  netdev->name, adapter);
2896
	else
2897
		err = request_irq(adapter->pdev->irq, ixgbe_intr, IRQF_SHARED,
2898
				  netdev->name, adapter);
2899

2900
	if (err)
2901
		e_err(probe, "request_irq failed, Error %d\n", err);
2902 2903 2904 2905 2906 2907

	return err;
}

static void ixgbe_free_irq(struct ixgbe_adapter *adapter)
{
2908
	int vector;
2909

2910 2911 2912 2913
	if (!(adapter->flags & IXGBE_FLAG_MSIX_ENABLED)) {
		free_irq(adapter->pdev->irq, adapter);
		return;
	}
2914

2915 2916 2917
	for (vector = 0; vector < adapter->num_q_vectors; vector++) {
		struct ixgbe_q_vector *q_vector = adapter->q_vector[vector];
		struct msix_entry *entry = &adapter->msix_entries[vector];
2918

2919 2920 2921
		/* free only the irqs that were actually requested */
		if (!q_vector->rx.ring && !q_vector->tx.ring)
			continue;
2922

2923 2924 2925 2926
		/* clear the affinity_mask in the IRQ descriptor */
		irq_set_affinity_hint(entry->vector, NULL);

		free_irq(entry->vector, q_vector);
2927
	}
2928 2929

	free_irq(adapter->msix_entries[vector++].vector, adapter);
2930 2931
}

2932 2933 2934 2935 2936 2937
/**
 * ixgbe_irq_disable - Mask off interrupt generation on the NIC
 * @adapter: board private structure
 **/
static inline void ixgbe_irq_disable(struct ixgbe_adapter *adapter)
{
2938 2939
	switch (adapter->hw.mac.type) {
	case ixgbe_mac_82598EB:
2940
		IXGBE_WRITE_REG(&adapter->hw, IXGBE_EIMC, ~0);
2941 2942
		break;
	case ixgbe_mac_82599EB:
D
Don Skidmore 已提交
2943
	case ixgbe_mac_X540:
2944 2945
	case ixgbe_mac_X550:
	case ixgbe_mac_X550EM_x:
2946 2947
		IXGBE_WRITE_REG(&adapter->hw, IXGBE_EIMC, 0xFFFF0000);
		IXGBE_WRITE_REG(&adapter->hw, IXGBE_EIMC_EX(0), ~0);
2948
		IXGBE_WRITE_REG(&adapter->hw, IXGBE_EIMC_EX(1), ~0);
2949 2950 2951
		break;
	default:
		break;
2952 2953 2954
	}
	IXGBE_WRITE_FLUSH(&adapter->hw);
	if (adapter->flags & IXGBE_FLAG_MSIX_ENABLED) {
2955 2956 2957 2958 2959 2960
		int vector;

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

		synchronize_irq(adapter->msix_entries[vector++].vector);
2961 2962 2963 2964 2965
	} else {
		synchronize_irq(adapter->pdev->irq);
	}
}

2966 2967 2968 2969 2970 2971
/**
 * ixgbe_configure_msi_and_legacy - Initialize PIN (INTA...) and MSI interrupts
 *
 **/
static void ixgbe_configure_msi_and_legacy(struct ixgbe_adapter *adapter)
{
2972
	struct ixgbe_q_vector *q_vector = adapter->q_vector[0];
2973

2974
	ixgbe_write_eitr(q_vector);
2975

2976 2977
	ixgbe_set_ivar(adapter, 0, 0, 0);
	ixgbe_set_ivar(adapter, 1, 0, 0);
2978

2979
	e_info(hw, "Legacy interrupt IVAR setup done\n");
2980 2981
}

2982 2983 2984 2985 2986 2987 2988
/**
 * ixgbe_configure_tx_ring - Configure 8259x Tx ring after Reset
 * @adapter: board private structure
 * @ring: structure containing ring specific data
 *
 * Configure the Tx descriptor ring after a reset.
 **/
2989 2990
void ixgbe_configure_tx_ring(struct ixgbe_adapter *adapter,
			     struct ixgbe_ring *ring)
2991 2992 2993
{
	struct ixgbe_hw *hw = &adapter->hw;
	u64 tdba = ring->dma;
2994
	int wait_loop = 10;
2995
	u32 txdctl = IXGBE_TXDCTL_ENABLE;
2996
	u8 reg_idx = ring->reg_idx;
2997

2998
	/* disable queue to avoid issues while updating state */
2999
	IXGBE_WRITE_REG(hw, IXGBE_TXDCTL(reg_idx), 0);
3000 3001
	IXGBE_WRITE_FLUSH(hw);

3002
	IXGBE_WRITE_REG(hw, IXGBE_TDBAL(reg_idx),
3003
			(tdba & DMA_BIT_MASK(32)));
3004 3005 3006 3007 3008
	IXGBE_WRITE_REG(hw, IXGBE_TDBAH(reg_idx), (tdba >> 32));
	IXGBE_WRITE_REG(hw, IXGBE_TDLEN(reg_idx),
			ring->count * sizeof(union ixgbe_adv_tx_desc));
	IXGBE_WRITE_REG(hw, IXGBE_TDH(reg_idx), 0);
	IXGBE_WRITE_REG(hw, IXGBE_TDT(reg_idx), 0);
3009
	ring->tail = adapter->io_addr + IXGBE_TDT(reg_idx);
3010

3011 3012
	/*
	 * set WTHRESH to encourage burst writeback, it should not be set
E
Emil Tantilov 已提交
3013 3014 3015
	 * higher than 1 when:
	 * - ITR is 0 as it could cause false TX hangs
	 * - ITR is set to > 100k int/sec and BQL is enabled
3016 3017 3018 3019 3020
	 *
	 * In order to avoid issues WTHRESH + PTHRESH should always be equal
	 * to or less than the number of on chip descriptors, which is
	 * currently 40.
	 */
E
Emil Tantilov 已提交
3021
	if (!ring->q_vector || (ring->q_vector->itr < IXGBE_100K_ITR))
3022 3023 3024 3025
		txdctl |= (1 << 16);	/* WTHRESH = 1 */
	else
		txdctl |= (8 << 16);	/* WTHRESH = 8 */

3026 3027 3028 3029
	/*
	 * Setting PTHRESH to 32 both improves performance
	 * and avoids a TX hang with DFP enabled
	 */
3030 3031
	txdctl |= (1 << 8) |	/* HTHRESH = 1 */
		   32;		/* PTHRESH = 32 */
3032 3033

	/* reinitialize flowdirector state */
3034
	if (adapter->flags & IXGBE_FLAG_FDIR_HASH_CAPABLE) {
3035 3036 3037 3038 3039 3040
		ring->atr_sample_rate = adapter->atr_sample_rate;
		ring->atr_count = 0;
		set_bit(__IXGBE_TX_FDIR_INIT_DONE, &ring->state);
	} else {
		ring->atr_sample_rate = 0;
	}
3041

3042 3043 3044 3045 3046
	/* initialize XPS */
	if (!test_and_set_bit(__IXGBE_TX_XPS_INIT_DONE, &ring->state)) {
		struct ixgbe_q_vector *q_vector = ring->q_vector;

		if (q_vector)
3047
			netif_set_xps_queue(ring->netdev,
3048 3049 3050 3051
					    &q_vector->affinity_mask,
					    ring->queue_index);
	}

3052 3053
	clear_bit(__IXGBE_HANG_CHECK_ARMED, &ring->state);

3054 3055 3056 3057 3058 3059 3060 3061 3062 3063
	/* enable queue */
	IXGBE_WRITE_REG(hw, IXGBE_TXDCTL(reg_idx), txdctl);

	/* TXDCTL.EN will return 0 on 82598 if link is down, so skip it */
	if (hw->mac.type == ixgbe_mac_82598EB &&
	    !(IXGBE_READ_REG(hw, IXGBE_LINKS) & IXGBE_LINKS_UP))
		return;

	/* poll to verify queue is enabled */
	do {
3064
		usleep_range(1000, 2000);
3065 3066 3067 3068
		txdctl = IXGBE_READ_REG(hw, IXGBE_TXDCTL(reg_idx));
	} while (--wait_loop && !(txdctl & IXGBE_TXDCTL_ENABLE));
	if (!wait_loop)
		e_err(drv, "Could not enable Tx Queue %d\n", reg_idx);
3069 3070
}

3071 3072 3073
static void ixgbe_setup_mtqc(struct ixgbe_adapter *adapter)
{
	struct ixgbe_hw *hw = &adapter->hw;
3074
	u32 rttdcs, mtqc;
3075
	u8 tcs = netdev_get_num_tc(adapter->netdev);
3076 3077 3078 3079 3080 3081 3082 3083 3084 3085

	if (hw->mac.type == ixgbe_mac_82598EB)
		return;

	/* disable the arbiter while setting MTQC */
	rttdcs = IXGBE_READ_REG(hw, IXGBE_RTTDCS);
	rttdcs |= IXGBE_RTTDCS_ARBDIS;
	IXGBE_WRITE_REG(hw, IXGBE_RTTDCS, rttdcs);

	/* set transmit pool layout */
3086 3087 3088 3089 3090 3091 3092 3093 3094 3095 3096 3097 3098 3099 3100
	if (adapter->flags & IXGBE_FLAG_SRIOV_ENABLED) {
		mtqc = IXGBE_MTQC_VT_ENA;
		if (tcs > 4)
			mtqc |= IXGBE_MTQC_RT_ENA | IXGBE_MTQC_8TC_8TQ;
		else if (tcs > 1)
			mtqc |= IXGBE_MTQC_RT_ENA | IXGBE_MTQC_4TC_4TQ;
		else if (adapter->ring_feature[RING_F_RSS].indices == 4)
			mtqc |= IXGBE_MTQC_32VF;
		else
			mtqc |= IXGBE_MTQC_64VF;
	} else {
		if (tcs > 4)
			mtqc = IXGBE_MTQC_RT_ENA | IXGBE_MTQC_8TC_8TQ;
		else if (tcs > 1)
			mtqc = IXGBE_MTQC_RT_ENA | IXGBE_MTQC_4TC_4TQ;
3101
		else
3102 3103
			mtqc = IXGBE_MTQC_64Q_1PB;
	}
3104

3105
	IXGBE_WRITE_REG(hw, IXGBE_MTQC, mtqc);
3106

3107 3108 3109 3110 3111
	/* Enable Security TX Buffer IFG for multiple pb */
	if (tcs) {
		u32 sectx = IXGBE_READ_REG(hw, IXGBE_SECTXMINIFG);
		sectx |= IXGBE_SECTX_DCB;
		IXGBE_WRITE_REG(hw, IXGBE_SECTXMINIFG, sectx);
3112 3113 3114 3115 3116 3117 3118
	}

	/* re-enable the arbiter */
	rttdcs &= ~IXGBE_RTTDCS_ARBDIS;
	IXGBE_WRITE_REG(hw, IXGBE_RTTDCS, rttdcs);
}

3119
/**
3120
 * ixgbe_configure_tx - Configure 8259x Transmit Unit after Reset
3121 3122 3123 3124 3125 3126
 * @adapter: board private structure
 *
 * Configure the Tx unit of the MAC after a reset.
 **/
static void ixgbe_configure_tx(struct ixgbe_adapter *adapter)
{
3127 3128
	struct ixgbe_hw *hw = &adapter->hw;
	u32 dmatxctl;
3129
	u32 i;
3130

3131 3132 3133 3134 3135 3136 3137 3138 3139
	ixgbe_setup_mtqc(adapter);

	if (hw->mac.type != ixgbe_mac_82598EB) {
		/* DMATXCTL.EN must be before Tx queues are enabled */
		dmatxctl = IXGBE_READ_REG(hw, IXGBE_DMATXCTL);
		dmatxctl |= IXGBE_DMATXCTL_TE;
		IXGBE_WRITE_REG(hw, IXGBE_DMATXCTL, dmatxctl);
	}

3140
	/* Setup the HW Tx Head and Tail descriptor pointers */
3141 3142
	for (i = 0; i < adapter->num_tx_queues; i++)
		ixgbe_configure_tx_ring(adapter, adapter->tx_ring[i]);
3143 3144
}

3145 3146 3147 3148 3149 3150 3151 3152 3153 3154 3155 3156 3157 3158 3159 3160 3161 3162 3163 3164 3165 3166 3167 3168 3169 3170 3171 3172 3173 3174 3175 3176 3177 3178 3179 3180 3181 3182 3183 3184 3185 3186 3187 3188 3189 3190 3191 3192 3193 3194 3195 3196 3197 3198 3199
static void ixgbe_enable_rx_drop(struct ixgbe_adapter *adapter,
				 struct ixgbe_ring *ring)
{
	struct ixgbe_hw *hw = &adapter->hw;
	u8 reg_idx = ring->reg_idx;
	u32 srrctl = IXGBE_READ_REG(hw, IXGBE_SRRCTL(reg_idx));

	srrctl |= IXGBE_SRRCTL_DROP_EN;

	IXGBE_WRITE_REG(hw, IXGBE_SRRCTL(reg_idx), srrctl);
}

static void ixgbe_disable_rx_drop(struct ixgbe_adapter *adapter,
				  struct ixgbe_ring *ring)
{
	struct ixgbe_hw *hw = &adapter->hw;
	u8 reg_idx = ring->reg_idx;
	u32 srrctl = IXGBE_READ_REG(hw, IXGBE_SRRCTL(reg_idx));

	srrctl &= ~IXGBE_SRRCTL_DROP_EN;

	IXGBE_WRITE_REG(hw, IXGBE_SRRCTL(reg_idx), srrctl);
}

#ifdef CONFIG_IXGBE_DCB
void ixgbe_set_rx_drop_en(struct ixgbe_adapter *adapter)
#else
static void ixgbe_set_rx_drop_en(struct ixgbe_adapter *adapter)
#endif
{
	int i;
	bool pfc_en = adapter->dcb_cfg.pfc_mode_enable;

	if (adapter->ixgbe_ieee_pfc)
		pfc_en |= !!(adapter->ixgbe_ieee_pfc->pfc_en);

	/*
	 * We should set the drop enable bit if:
	 *  SR-IOV is enabled
	 *   or
	 *  Number of Rx queues > 1 and flow control is disabled
	 *
	 *  This allows us to avoid head of line blocking for security
	 *  and performance reasons.
	 */
	if (adapter->num_vfs || (adapter->num_rx_queues > 1 &&
	    !(adapter->hw.fc.current_mode & ixgbe_fc_tx_pause) && !pfc_en)) {
		for (i = 0; i < adapter->num_rx_queues; i++)
			ixgbe_enable_rx_drop(adapter, adapter->rx_ring[i]);
	} else {
		for (i = 0; i < adapter->num_rx_queues; i++)
			ixgbe_disable_rx_drop(adapter, adapter->rx_ring[i]);
	}
}

3200
#define IXGBE_SRRCTL_BSIZEHDRSIZE_SHIFT 2
3201

3202
static void ixgbe_configure_srrctl(struct ixgbe_adapter *adapter,
3203
				   struct ixgbe_ring *rx_ring)
3204
{
3205
	struct ixgbe_hw *hw = &adapter->hw;
3206
	u32 srrctl;
3207
	u8 reg_idx = rx_ring->reg_idx;
3208

3209 3210
	if (hw->mac.type == ixgbe_mac_82598EB) {
		u16 mask = adapter->ring_feature[RING_F_RSS].mask;
3211

3212 3213 3214 3215 3216 3217
		/*
		 * if VMDq is not active we must program one srrctl register
		 * per RSS queue since we have enabled RDRXCTL.MVMEN
		 */
		reg_idx &= mask;
	}
3218

3219 3220
	/* configure header buffer length, needed for RSC */
	srrctl = IXGBE_RX_HDR_SIZE << IXGBE_SRRCTL_BSIZEHDRSIZE_SHIFT;
3221

3222
	/* configure the packet buffer length */
3223
	srrctl |= ixgbe_rx_bufsz(rx_ring) >> IXGBE_SRRCTL_BSIZEPKT_SHIFT;
3224 3225

	/* configure descriptor type */
3226
	srrctl |= IXGBE_SRRCTL_DESCTYPE_ADV_ONEBUF;
3227

3228
	IXGBE_WRITE_REG(hw, IXGBE_SRRCTL(reg_idx), srrctl);
3229
}
3230

3231
static void ixgbe_setup_reta(struct ixgbe_adapter *adapter, const u32 *seed)
3232
{
3233
	struct ixgbe_hw *hw = &adapter->hw;
3234
	u32 reta = 0;
3235
	int i, j;
3236
	int reta_entries = 128;
3237
	u16 rss_i = adapter->ring_feature[RING_F_RSS].indices;
3238
	int indices_multi;
3239 3240 3241 3242 3243 3244 3245 3246

	/*
	 * Program table for at least 2 queues w/ SR-IOV so that VFs can
	 * make full use of any rings they may have.  We will use the
	 * PSRTYPE register to control how many rings we use within the PF.
	 */
	if ((adapter->flags & IXGBE_FLAG_SRIOV_ENABLED) && (rss_i < 2))
		rss_i = 2;
3247

3248 3249 3250 3251
	/* Fill out hash function seeds */
	for (i = 0; i < 10; i++)
		IXGBE_WRITE_REG(hw, IXGBE_RSSRK(i), seed[i]);

3252 3253 3254 3255 3256 3257 3258 3259 3260 3261 3262 3263 3264 3265 3266 3267 3268 3269 3270
	/* Fill out the redirection table as follows:
	 * 82598: 128 (8 bit wide) entries containing pair of 4 bit RSS indices
	 * 82599/X540: 128 (8 bit wide) entries containing 4 bit RSS index
	 * X550: 512 (8 bit wide) entries containing 6 bit RSS index
	 */
	if (adapter->hw.mac.type == ixgbe_mac_82598EB)
		indices_multi = 0x11;
	else
		indices_multi = 0x1;

	switch (adapter->hw.mac.type) {
	case ixgbe_mac_X550:
	case ixgbe_mac_X550EM_x:
		if (!(adapter->flags & IXGBE_FLAG_SRIOV_ENABLED))
			reta_entries = 512;
	default:
		break;
	}

3271
	/* Fill out redirection table */
3272 3273 3274 3275 3276 3277 3278 3279 3280 3281 3282 3283 3284 3285 3286 3287 3288 3289 3290 3291 3292 3293 3294 3295 3296 3297 3298 3299
	for (i = 0, j = 0; i < reta_entries; i++, j++) {
		if (j == rss_i)
			j = 0;
		reta = (reta << 8) | (j * indices_multi);
		if ((i & 3) == 3) {
			if (i < 128)
				IXGBE_WRITE_REG(hw, IXGBE_RETA(i >> 2), reta);
			else
				IXGBE_WRITE_REG(hw, IXGBE_ERETA((i >> 2) - 32),
						reta);
		}
	}
}

static void ixgbe_setup_vfreta(struct ixgbe_adapter *adapter, const u32 *seed)
{
	struct ixgbe_hw *hw = &adapter->hw;
	u32 vfreta = 0;
	u16 rss_i = adapter->ring_feature[RING_F_RSS].indices;
	unsigned int pf_pool = adapter->num_vfs;
	int i, j;

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

	/* Fill out the redirection table */
	for (i = 0, j = 0; i < 64; i++, j++) {
3300
		if (j == rss_i)
3301
			j = 0;
3302
		vfreta = (vfreta << 8) | j;
3303
		if ((i & 3) == 3)
3304 3305
			IXGBE_WRITE_REG(hw, IXGBE_PFVFRETA(i >> 2, pf_pool),
					vfreta);
3306
	}
3307 3308 3309 3310 3311
}

static void ixgbe_setup_mrqc(struct ixgbe_adapter *adapter)
{
	struct ixgbe_hw *hw = &adapter->hw;
3312
	u32 mrqc = 0, rss_field = 0, vfmrqc = 0;
3313
	u32 rss_key[10];
3314
	u32 rxcsum;
3315

3316 3317 3318 3319 3320
	/* Disable indicating checksum in descriptor, enables RSS hash */
	rxcsum = IXGBE_READ_REG(hw, IXGBE_RXCSUM);
	rxcsum |= IXGBE_RXCSUM_PCSD;
	IXGBE_WRITE_REG(hw, IXGBE_RXCSUM, rxcsum);

3321
	if (adapter->hw.mac.type == ixgbe_mac_82598EB) {
3322
		if (adapter->ring_feature[RING_F_RSS].mask)
3323
			mrqc = IXGBE_MRQC_RSSEN;
3324
	} else {
3325 3326 3327 3328 3329 3330 3331 3332 3333
		u8 tcs = netdev_get_num_tc(adapter->netdev);

		if (adapter->flags & IXGBE_FLAG_SRIOV_ENABLED) {
			if (tcs > 4)
				mrqc = IXGBE_MRQC_VMDQRT8TCEN;	/* 8 TCs */
			else if (tcs > 1)
				mrqc = IXGBE_MRQC_VMDQRT4TCEN;	/* 4 TCs */
			else if (adapter->ring_feature[RING_F_RSS].indices == 4)
				mrqc = IXGBE_MRQC_VMDQRSS32EN;
3334
			else
3335 3336 3337
				mrqc = IXGBE_MRQC_VMDQRSS64EN;
		} else {
			if (tcs > 4)
3338
				mrqc = IXGBE_MRQC_RTRSS8TCEN;
3339 3340 3341 3342
			else if (tcs > 1)
				mrqc = IXGBE_MRQC_RTRSS4TCEN;
			else
				mrqc = IXGBE_MRQC_RSSEN;
3343
		}
3344 3345
	}

3346
	/* Perform hash on these packet types */
3347 3348 3349 3350
	rss_field |= IXGBE_MRQC_RSS_FIELD_IPV4 |
		     IXGBE_MRQC_RSS_FIELD_IPV4_TCP |
		     IXGBE_MRQC_RSS_FIELD_IPV6 |
		     IXGBE_MRQC_RSS_FIELD_IPV6_TCP;
3351

3352
	if (adapter->flags2 & IXGBE_FLAG2_RSS_FIELD_IPV4_UDP)
3353
		rss_field |= IXGBE_MRQC_RSS_FIELD_IPV4_UDP;
3354
	if (adapter->flags2 & IXGBE_FLAG2_RSS_FIELD_IPV6_UDP)
3355
		rss_field |= IXGBE_MRQC_RSS_FIELD_IPV6_UDP;
3356

3357
	netdev_rss_key_fill(rss_key, sizeof(rss_key));
3358 3359 3360 3361 3362 3363 3364 3365 3366 3367 3368 3369 3370 3371 3372 3373 3374 3375
	if ((hw->mac.type >= ixgbe_mac_X550) &&
	    (adapter->flags & IXGBE_FLAG_SRIOV_ENABLED)) {
		unsigned int pf_pool = adapter->num_vfs;

		/* Enable VF RSS mode */
		mrqc |= IXGBE_MRQC_MULTIPLE_RSS;
		IXGBE_WRITE_REG(hw, IXGBE_MRQC, mrqc);

		/* Setup RSS through the VF registers */
		ixgbe_setup_vfreta(adapter, rss_key);
		vfmrqc = IXGBE_MRQC_RSSEN;
		vfmrqc |= rss_field;
		IXGBE_WRITE_REG(hw, IXGBE_PFVFMRQC(pf_pool), vfmrqc);
	} else {
		ixgbe_setup_reta(adapter, rss_key);
		mrqc |= rss_field;
		IXGBE_WRITE_REG(hw, IXGBE_MRQC, mrqc);
	}
3376 3377
}

3378 3379 3380 3381 3382
/**
 * ixgbe_configure_rscctl - enable RSC for the indicated ring
 * @adapter:    address of board private structure
 * @index:      index of ring to set
 **/
3383
static void ixgbe_configure_rscctl(struct ixgbe_adapter *adapter,
3384
				   struct ixgbe_ring *ring)
3385 3386 3387
{
	struct ixgbe_hw *hw = &adapter->hw;
	u32 rscctrl;
3388
	u8 reg_idx = ring->reg_idx;
3389

A
Alexander Duyck 已提交
3390
	if (!ring_is_rsc_enabled(ring))
3391
		return;
3392

3393
	rscctrl = IXGBE_READ_REG(hw, IXGBE_RSCCTL(reg_idx));
3394 3395 3396 3397
	rscctrl |= IXGBE_RSCCTL_RSCEN;
	/*
	 * we must limit the number of descriptors so that the
	 * total size of max desc * buf_len is not greater
3398
	 * than 65536
3399
	 */
3400
	rscctrl |= IXGBE_RSCCTL_MAXDESC_16;
3401
	IXGBE_WRITE_REG(hw, IXGBE_RSCCTL(reg_idx), rscctrl);
3402 3403
}

3404 3405 3406 3407 3408 3409 3410
#define IXGBE_MAX_RX_DESC_POLL 10
static void ixgbe_rx_desc_queue_enable(struct ixgbe_adapter *adapter,
				       struct ixgbe_ring *ring)
{
	struct ixgbe_hw *hw = &adapter->hw;
	int wait_loop = IXGBE_MAX_RX_DESC_POLL;
	u32 rxdctl;
3411
	u8 reg_idx = ring->reg_idx;
3412

3413 3414
	if (ixgbe_removed(hw->hw_addr))
		return;
3415 3416 3417 3418 3419 3420
	/* RXDCTL.EN will return 0 on 82598 if link is down, so skip it */
	if (hw->mac.type == ixgbe_mac_82598EB &&
	    !(IXGBE_READ_REG(hw, IXGBE_LINKS) & IXGBE_LINKS_UP))
		return;

	do {
3421
		usleep_range(1000, 2000);
3422 3423 3424 3425 3426 3427 3428 3429 3430
		rxdctl = IXGBE_READ_REG(hw, IXGBE_RXDCTL(reg_idx));
	} while (--wait_loop && !(rxdctl & IXGBE_RXDCTL_ENABLE));

	if (!wait_loop) {
		e_err(drv, "RXDCTL.ENABLE on Rx queue %d not set within "
		      "the polling period\n", reg_idx);
	}
}

3431 3432 3433 3434 3435 3436 3437 3438
void ixgbe_disable_rx_queue(struct ixgbe_adapter *adapter,
			    struct ixgbe_ring *ring)
{
	struct ixgbe_hw *hw = &adapter->hw;
	int wait_loop = IXGBE_MAX_RX_DESC_POLL;
	u32 rxdctl;
	u8 reg_idx = ring->reg_idx;

3439 3440
	if (ixgbe_removed(hw->hw_addr))
		return;
3441 3442 3443 3444 3445 3446 3447 3448 3449 3450 3451 3452 3453 3454 3455 3456 3457 3458 3459 3460 3461 3462
	rxdctl = IXGBE_READ_REG(hw, IXGBE_RXDCTL(reg_idx));
	rxdctl &= ~IXGBE_RXDCTL_ENABLE;

	/* write value back with RXDCTL.ENABLE bit cleared */
	IXGBE_WRITE_REG(hw, IXGBE_RXDCTL(reg_idx), rxdctl);

	if (hw->mac.type == ixgbe_mac_82598EB &&
	    !(IXGBE_READ_REG(hw, IXGBE_LINKS) & IXGBE_LINKS_UP))
		return;

	/* the hardware may take up to 100us to really disable the rx queue */
	do {
		udelay(10);
		rxdctl = IXGBE_READ_REG(hw, IXGBE_RXDCTL(reg_idx));
	} while (--wait_loop && (rxdctl & IXGBE_RXDCTL_ENABLE));

	if (!wait_loop) {
		e_err(drv, "RXDCTL.ENABLE on Rx queue %d not cleared within "
		      "the polling period\n", reg_idx);
	}
}

3463 3464
void ixgbe_configure_rx_ring(struct ixgbe_adapter *adapter,
			     struct ixgbe_ring *ring)
3465 3466 3467
{
	struct ixgbe_hw *hw = &adapter->hw;
	u64 rdba = ring->dma;
3468
	u32 rxdctl;
3469
	u8 reg_idx = ring->reg_idx;
3470

3471 3472
	/* disable queue to avoid issues while updating state */
	rxdctl = IXGBE_READ_REG(hw, IXGBE_RXDCTL(reg_idx));
3473
	ixgbe_disable_rx_queue(adapter, ring);
3474

3475 3476 3477 3478 3479 3480
	IXGBE_WRITE_REG(hw, IXGBE_RDBAL(reg_idx), (rdba & DMA_BIT_MASK(32)));
	IXGBE_WRITE_REG(hw, IXGBE_RDBAH(reg_idx), (rdba >> 32));
	IXGBE_WRITE_REG(hw, IXGBE_RDLEN(reg_idx),
			ring->count * sizeof(union ixgbe_adv_rx_desc));
	IXGBE_WRITE_REG(hw, IXGBE_RDH(reg_idx), 0);
	IXGBE_WRITE_REG(hw, IXGBE_RDT(reg_idx), 0);
3481
	ring->tail = adapter->io_addr + IXGBE_RDT(reg_idx);
3482 3483 3484 3485 3486 3487 3488 3489 3490 3491 3492 3493 3494 3495 3496 3497 3498 3499 3500 3501 3502

	ixgbe_configure_srrctl(adapter, ring);
	ixgbe_configure_rscctl(adapter, ring);

	if (hw->mac.type == ixgbe_mac_82598EB) {
		/*
		 * enable cache line friendly hardware writes:
		 * PTHRESH=32 descriptors (half the internal cache),
		 * this also removes ugly rx_no_buffer_count increment
		 * HTHRESH=4 descriptors (to minimize latency on fetch)
		 * WTHRESH=8 burst writeback up to two cache lines
		 */
		rxdctl &= ~0x3FFFFF;
		rxdctl |=  0x080420;
	}

	/* enable receive descriptor ring */
	rxdctl |= IXGBE_RXDCTL_ENABLE;
	IXGBE_WRITE_REG(hw, IXGBE_RXDCTL(reg_idx), rxdctl);

	ixgbe_rx_desc_queue_enable(adapter, ring);
3503
	ixgbe_alloc_rx_buffers(ring, ixgbe_desc_unused(ring));
3504 3505
}

3506 3507 3508
static void ixgbe_setup_psrtype(struct ixgbe_adapter *adapter)
{
	struct ixgbe_hw *hw = &adapter->hw;
3509
	int rss_i = adapter->ring_feature[RING_F_RSS].indices;
3510
	u16 pool;
3511 3512 3513

	/* PSRTYPE must be initialized in non 82598 adapters */
	u32 psrtype = IXGBE_PSRTYPE_TCPHDR |
3514 3515
		      IXGBE_PSRTYPE_UDPHDR |
		      IXGBE_PSRTYPE_IPV4HDR |
3516
		      IXGBE_PSRTYPE_L2HDR |
3517
		      IXGBE_PSRTYPE_IPV6HDR;
3518 3519 3520 3521

	if (hw->mac.type == ixgbe_mac_82598EB)
		return;

3522 3523 3524 3525
	if (rss_i > 3)
		psrtype |= 2 << 29;
	else if (rss_i > 1)
		psrtype |= 1 << 29;
3526

3527 3528
	for_each_set_bit(pool, &adapter->fwd_bitmask, 32)
		IXGBE_WRITE_REG(hw, IXGBE_PSRTYPE(VMDQ_P(pool)), psrtype);
3529 3530
}

3531 3532 3533 3534
static void ixgbe_configure_virtualization(struct ixgbe_adapter *adapter)
{
	struct ixgbe_hw *hw = &adapter->hw;
	u32 reg_offset, vf_shift;
3535
	u32 gcr_ext, vmdctl;
3536
	int i;
3537 3538 3539 3540 3541

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

	vmdctl = IXGBE_READ_REG(hw, IXGBE_VT_CTL);
3542 3543
	vmdctl |= IXGBE_VMD_CTL_VMDQ_EN;
	vmdctl &= ~IXGBE_VT_CTL_POOL_MASK;
3544
	vmdctl |= VMDQ_P(0) << IXGBE_VT_CTL_POOL_SHIFT;
3545 3546
	vmdctl |= IXGBE_VT_CTL_REPLEN;
	IXGBE_WRITE_REG(hw, IXGBE_VT_CTL, vmdctl);
3547

3548 3549
	vf_shift = VMDQ_P(0) % 32;
	reg_offset = (VMDQ_P(0) >= 32) ? 1 : 0;
3550 3551

	/* Enable only the PF's pool for Tx/Rx */
3552 3553 3554 3555
	IXGBE_WRITE_REG(hw, IXGBE_VFRE(reg_offset), (~0) << vf_shift);
	IXGBE_WRITE_REG(hw, IXGBE_VFRE(reg_offset ^ 1), reg_offset - 1);
	IXGBE_WRITE_REG(hw, IXGBE_VFTE(reg_offset), (~0) << vf_shift);
	IXGBE_WRITE_REG(hw, IXGBE_VFTE(reg_offset ^ 1), reg_offset - 1);
3556
	if (adapter->bridge_mode == BRIDGE_MODE_VEB)
3557
		IXGBE_WRITE_REG(hw, IXGBE_PFDTXGSWC, IXGBE_PFDTXGSWC_VT_LBEN);
3558 3559

	/* Map PF MAC address in RAR Entry 0 to first pool following VFs */
3560
	hw->mac.ops.set_vmdq(hw, 0, VMDQ_P(0));
3561 3562 3563 3564 3565

	/*
	 * Set up VF register offsets for selected VT Mode,
	 * i.e. 32 or 64 VFs for SR-IOV
	 */
3566 3567 3568 3569 3570 3571 3572 3573 3574 3575 3576 3577
	switch (adapter->ring_feature[RING_F_VMDQ].mask) {
	case IXGBE_82599_VMDQ_8Q_MASK:
		gcr_ext = IXGBE_GCR_EXT_VT_MODE_16;
		break;
	case IXGBE_82599_VMDQ_4Q_MASK:
		gcr_ext = IXGBE_GCR_EXT_VT_MODE_32;
		break;
	default:
		gcr_ext = IXGBE_GCR_EXT_VT_MODE_64;
		break;
	}

3578 3579
	IXGBE_WRITE_REG(hw, IXGBE_GCR_EXT, gcr_ext);

3580

3581
	/* Enable MAC Anti-Spoofing */
3582
	hw->mac.ops.set_mac_anti_spoofing(hw, (adapter->num_vfs != 0),
3583
					  adapter->num_vfs);
3584 3585 3586 3587 3588 3589 3590 3591 3592 3593

	/* Ensure LLDP is set for Ethertype Antispoofing if we will be
	 * calling set_ethertype_anti_spoofing for each VF in loop below
	 */
	if (hw->mac.ops.set_ethertype_anti_spoofing)
		IXGBE_WRITE_REG(hw, IXGBE_ETQF(IXGBE_ETQF_FILTER_LLDP),
				(IXGBE_ETQF_FILTER_EN    | /* enable filter */
				 IXGBE_ETQF_TX_ANTISPOOF | /* tx antispoof */
				 IXGBE_ETH_P_LLDP));	   /* LLDP eth type */

3594 3595 3596 3597
	/* For VFs that have spoof checking turned off */
	for (i = 0; i < adapter->num_vfs; i++) {
		if (!adapter->vfinfo[i].spoofchk_enabled)
			ixgbe_ndo_set_vf_spoofchk(adapter->netdev, i, false);
3598 3599 3600 3601

		/* enable ethertype anti spoofing if hw supports it */
		if (hw->mac.ops.set_ethertype_anti_spoofing)
			hw->mac.ops.set_ethertype_anti_spoofing(hw, true, i);
3602
	}
3603 3604
}

3605
static void ixgbe_set_rx_buffer_len(struct ixgbe_adapter *adapter)
3606 3607 3608 3609
{
	struct ixgbe_hw *hw = &adapter->hw;
	struct net_device *netdev = adapter->netdev;
	int max_frame = netdev->mtu + ETH_HLEN + ETH_FCS_LEN;
3610 3611 3612
	struct ixgbe_ring *rx_ring;
	int i;
	u32 mhadd, hlreg0;
3613

3614
#ifdef IXGBE_FCOE
3615 3616 3617 3618
	/* adjust max frame to be able to do baby jumbo for FCoE */
	if ((adapter->flags & IXGBE_FLAG_FCOE_ENABLED) &&
	    (max_frame < IXGBE_FCOE_JUMBO_FRAME_SIZE))
		max_frame = IXGBE_FCOE_JUMBO_FRAME_SIZE;
3619

3620
#endif /* IXGBE_FCOE */
3621 3622 3623 3624 3625

	/* adjust max frame to be at least the size of a standard frame */
	if (max_frame < (ETH_FRAME_LEN + ETH_FCS_LEN))
		max_frame = (ETH_FRAME_LEN + ETH_FCS_LEN);

3626 3627 3628 3629 3630 3631 3632 3633 3634 3635 3636 3637
	mhadd = IXGBE_READ_REG(hw, IXGBE_MHADD);
	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);
	}

	hlreg0 = IXGBE_READ_REG(hw, IXGBE_HLREG0);
	/* set jumbo enable since MHADD.MFS is keeping size locked at max_frame */
	hlreg0 |= IXGBE_HLREG0_JUMBOEN;
	IXGBE_WRITE_REG(hw, IXGBE_HLREG0, hlreg0);
3638

3639 3640 3641 3642
	/*
	 * Setup the HW Rx Head and Tail Descriptor Pointers and
	 * the Base and Length of the Rx Descriptor Ring
	 */
3643
	for (i = 0; i < adapter->num_rx_queues; i++) {
3644
		rx_ring = adapter->rx_ring[i];
A
Alexander Duyck 已提交
3645 3646
		if (adapter->flags2 & IXGBE_FLAG2_RSC_ENABLED)
			set_ring_rsc_enabled(rx_ring);
3647
		else
A
Alexander Duyck 已提交
3648
			clear_ring_rsc_enabled(rx_ring);
3649 3650 3651
	}
}

3652 3653 3654 3655 3656 3657
static void ixgbe_setup_rdrxctl(struct ixgbe_adapter *adapter)
{
	struct ixgbe_hw *hw = &adapter->hw;
	u32 rdrxctl = IXGBE_READ_REG(hw, IXGBE_RDRXCTL);

	switch (hw->mac.type) {
3658 3659
	case ixgbe_mac_X550:
	case ixgbe_mac_X550EM_x:
3660 3661 3662 3663 3664 3665 3666 3667 3668 3669 3670 3671 3672 3673
	case 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]
		 */
		rdrxctl |= IXGBE_RDRXCTL_MVMEN;
		break;
	case ixgbe_mac_82599EB:
D
Don Skidmore 已提交
3674
	case ixgbe_mac_X540:
3675 3676 3677 3678 3679 3680 3681 3682 3683 3684 3685 3686 3687 3688 3689 3690
		/* Disable RSC for ACK packets */
		IXGBE_WRITE_REG(hw, IXGBE_RSCDBU,
		   (IXGBE_RSCDBU_RSCACKDIS | IXGBE_READ_REG(hw, IXGBE_RSCDBU)));
		rdrxctl &= ~IXGBE_RDRXCTL_RSCFRSTSIZE;
		/* hardware requires some bits to be set by default */
		rdrxctl |= (IXGBE_RDRXCTL_RSCACKC | IXGBE_RDRXCTL_FCOE_WRFIX);
		rdrxctl |= IXGBE_RDRXCTL_CRCSTRIP;
		break;
	default:
		/* We should do nothing since we don't know this hardware */
		return;
	}

	IXGBE_WRITE_REG(hw, IXGBE_RDRXCTL, rdrxctl);
}

3691 3692 3693 3694 3695 3696 3697 3698 3699 3700
/**
 * ixgbe_configure_rx - Configure 8259x Receive Unit after Reset
 * @adapter: board private structure
 *
 * Configure the Rx unit of the MAC after a reset.
 **/
static void ixgbe_configure_rx(struct ixgbe_adapter *adapter)
{
	struct ixgbe_hw *hw = &adapter->hw;
	int i;
3701
	u32 rxctrl, rfctl;
3702 3703

	/* disable receives while setting up the descriptors */
3704
	hw->mac.ops.disable_rx(hw);
3705 3706

	ixgbe_setup_psrtype(adapter);
3707
	ixgbe_setup_rdrxctl(adapter);
3708

3709 3710 3711 3712 3713 3714 3715
	/* RSC Setup */
	rfctl = IXGBE_READ_REG(hw, IXGBE_RFCTL);
	rfctl &= ~IXGBE_RFCTL_RSC_DIS;
	if (!(adapter->flags2 & IXGBE_FLAG2_RSC_ENABLED))
		rfctl |= IXGBE_RFCTL_RSC_DIS;
	IXGBE_WRITE_REG(hw, IXGBE_RFCTL, rfctl);

3716
	/* Program registers for the distribution of queues */
3717 3718
	ixgbe_setup_mrqc(adapter);

3719 3720 3721 3722 3723 3724 3725
	/* set_rx_buffer_len must be called before ring initialization */
	ixgbe_set_rx_buffer_len(adapter);

	/*
	 * Setup the HW Rx Head and Tail Descriptor Pointers and
	 * the Base and Length of the Rx Descriptor Ring
	 */
3726 3727
	for (i = 0; i < adapter->num_rx_queues; i++)
		ixgbe_configure_rx_ring(adapter, adapter->rx_ring[i]);
3728

3729
	rxctrl = IXGBE_READ_REG(hw, IXGBE_RXCTRL);
3730 3731 3732 3733 3734 3735 3736
	/* disable drop enable for 82598 parts */
	if (hw->mac.type == ixgbe_mac_82598EB)
		rxctrl |= IXGBE_RXCTRL_DMBYPS;

	/* enable all receives */
	rxctrl |= IXGBE_RXCTRL_RXEN;
	hw->mac.ops.enable_rx_dma(hw, rxctrl);
3737 3738
}

3739 3740
static int ixgbe_vlan_rx_add_vid(struct net_device *netdev,
				 __be16 proto, u16 vid)
3741 3742 3743 3744 3745
{
	struct ixgbe_adapter *adapter = netdev_priv(netdev);
	struct ixgbe_hw *hw = &adapter->hw;

	/* add VID to filter table */
3746
	hw->mac.ops.set_vfta(&adapter->hw, vid, VMDQ_P(0), true);
3747
	set_bit(vid, adapter->active_vlans);
3748 3749

	return 0;
3750 3751
}

3752 3753
static int ixgbe_vlan_rx_kill_vid(struct net_device *netdev,
				  __be16 proto, u16 vid)
3754 3755 3756 3757 3758
{
	struct ixgbe_adapter *adapter = netdev_priv(netdev);
	struct ixgbe_hw *hw = &adapter->hw;

	/* remove VID from filter table */
3759
	hw->mac.ops.set_vfta(&adapter->hw, vid, VMDQ_P(0), false);
3760
	clear_bit(vid, adapter->active_vlans);
3761 3762

	return 0;
3763 3764
}

3765 3766 3767 3768 3769 3770 3771 3772
/**
 * ixgbe_vlan_strip_disable - helper to disable hw vlan stripping
 * @adapter: driver data
 */
static void ixgbe_vlan_strip_disable(struct ixgbe_adapter *adapter)
{
	struct ixgbe_hw *hw = &adapter->hw;
	u32 vlnctrl;
3773 3774 3775 3776
	int i, j;

	switch (hw->mac.type) {
	case ixgbe_mac_82598EB:
3777 3778
		vlnctrl = IXGBE_READ_REG(hw, IXGBE_VLNCTRL);
		vlnctrl &= ~IXGBE_VLNCTRL_VME;
3779 3780 3781
		IXGBE_WRITE_REG(hw, IXGBE_VLNCTRL, vlnctrl);
		break;
	case ixgbe_mac_82599EB:
D
Don Skidmore 已提交
3782
	case ixgbe_mac_X540:
3783 3784
	case ixgbe_mac_X550:
	case ixgbe_mac_X550EM_x:
3785
		for (i = 0; i < adapter->num_rx_queues; i++) {
3786 3787 3788 3789 3790
			struct ixgbe_ring *ring = adapter->rx_ring[i];

			if (ring->l2_accel_priv)
				continue;
			j = ring->reg_idx;
3791 3792 3793 3794 3795 3796 3797 3798 3799 3800 3801
			vlnctrl = IXGBE_READ_REG(hw, IXGBE_RXDCTL(j));
			vlnctrl &= ~IXGBE_RXDCTL_VME;
			IXGBE_WRITE_REG(hw, IXGBE_RXDCTL(j), vlnctrl);
		}
		break;
	default:
		break;
	}
}

/**
3802
 * ixgbe_vlan_strip_enable - helper to enable hw vlan stripping
3803 3804
 * @adapter: driver data
 */
3805
static void ixgbe_vlan_strip_enable(struct ixgbe_adapter *adapter)
3806 3807
{
	struct ixgbe_hw *hw = &adapter->hw;
3808
	u32 vlnctrl;
3809 3810 3811 3812
	int i, j;

	switch (hw->mac.type) {
	case ixgbe_mac_82598EB:
3813 3814
		vlnctrl = IXGBE_READ_REG(hw, IXGBE_VLNCTRL);
		vlnctrl |= IXGBE_VLNCTRL_VME;
3815 3816 3817
		IXGBE_WRITE_REG(hw, IXGBE_VLNCTRL, vlnctrl);
		break;
	case ixgbe_mac_82599EB:
D
Don Skidmore 已提交
3818
	case ixgbe_mac_X540:
3819 3820
	case ixgbe_mac_X550:
	case ixgbe_mac_X550EM_x:
3821
		for (i = 0; i < adapter->num_rx_queues; i++) {
3822 3823 3824 3825 3826
			struct ixgbe_ring *ring = adapter->rx_ring[i];

			if (ring->l2_accel_priv)
				continue;
			j = ring->reg_idx;
3827 3828 3829 3830 3831 3832 3833 3834 3835 3836
			vlnctrl = IXGBE_READ_REG(hw, IXGBE_RXDCTL(j));
			vlnctrl |= IXGBE_RXDCTL_VME;
			IXGBE_WRITE_REG(hw, IXGBE_RXDCTL(j), vlnctrl);
		}
		break;
	default:
		break;
	}
}

3837 3838
static void ixgbe_restore_vlan(struct ixgbe_adapter *adapter)
{
3839
	u16 vid;
3840

3841
	ixgbe_vlan_rx_add_vid(adapter->netdev, htons(ETH_P_8021Q), 0);
3842 3843

	for_each_set_bit(vid, adapter->active_vlans, VLAN_N_VID)
3844
		ixgbe_vlan_rx_add_vid(adapter->netdev, htons(ETH_P_8021Q), vid);
3845 3846
}

3847 3848 3849 3850 3851 3852 3853 3854 3855 3856 3857 3858 3859 3860 3861 3862 3863 3864 3865 3866 3867 3868 3869
/**
 * ixgbe_write_mc_addr_list - write multicast addresses to MTA
 * @netdev: network interface device structure
 *
 * Writes multicast address list to the MTA hash table.
 * Returns: -ENOMEM on failure
 *                0 on no addresses written
 *                X on writing X addresses to MTA
 **/
static int ixgbe_write_mc_addr_list(struct net_device *netdev)
{
	struct ixgbe_adapter *adapter = netdev_priv(netdev);
	struct ixgbe_hw *hw = &adapter->hw;

	if (!netif_running(netdev))
		return 0;

	if (hw->mac.ops.update_mc_addr_list)
		hw->mac.ops.update_mc_addr_list(hw, netdev);
	else
		return -ENOMEM;

#ifdef CONFIG_PCI_IOV
3870
	ixgbe_restore_vf_multicasts(adapter);
3871 3872 3873 3874 3875
#endif

	return netdev_mc_count(netdev);
}

3876 3877 3878 3879 3880 3881 3882 3883 3884 3885 3886 3887 3888 3889 3890 3891 3892 3893 3894 3895 3896 3897 3898 3899 3900 3901 3902 3903 3904 3905 3906 3907 3908 3909 3910 3911 3912 3913 3914 3915 3916 3917 3918 3919 3920 3921 3922
#ifdef CONFIG_PCI_IOV
void ixgbe_full_sync_mac_table(struct ixgbe_adapter *adapter)
{
	struct ixgbe_hw *hw = &adapter->hw;
	int i;
	for (i = 0; i < hw->mac.num_rar_entries; i++) {
		if (adapter->mac_table[i].state & IXGBE_MAC_STATE_IN_USE)
			hw->mac.ops.set_rar(hw, i, adapter->mac_table[i].addr,
					    adapter->mac_table[i].queue,
					    IXGBE_RAH_AV);
		else
			hw->mac.ops.clear_rar(hw, i);

		adapter->mac_table[i].state &= ~(IXGBE_MAC_STATE_MODIFIED);
	}
}
#endif

static void ixgbe_sync_mac_table(struct ixgbe_adapter *adapter)
{
	struct ixgbe_hw *hw = &adapter->hw;
	int i;
	for (i = 0; i < hw->mac.num_rar_entries; i++) {
		if (adapter->mac_table[i].state & IXGBE_MAC_STATE_MODIFIED) {
			if (adapter->mac_table[i].state &
			    IXGBE_MAC_STATE_IN_USE)
				hw->mac.ops.set_rar(hw, i,
						adapter->mac_table[i].addr,
						adapter->mac_table[i].queue,
						IXGBE_RAH_AV);
			else
				hw->mac.ops.clear_rar(hw, i);

			adapter->mac_table[i].state &=
						~(IXGBE_MAC_STATE_MODIFIED);
		}
	}
}

static void ixgbe_flush_sw_mac_table(struct ixgbe_adapter *adapter)
{
	int i;
	struct ixgbe_hw *hw = &adapter->hw;

	for (i = 0; i < hw->mac.num_rar_entries; i++) {
		adapter->mac_table[i].state |= IXGBE_MAC_STATE_MODIFIED;
		adapter->mac_table[i].state &= ~IXGBE_MAC_STATE_IN_USE;
3923
		eth_zero_addr(adapter->mac_table[i].addr);
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 3949 3950 3951 3952 3953 3954 3955 3956 3957 3958 3959 3960 3961 3962 3963 3964 3965 3966 3967 3968 3969 3970 3971 3972 3973 3974 3975 3976 3977 3978 3979 3980 3981 3982 3983 3984 3985 3986 3987 3988 3989 3990
		adapter->mac_table[i].queue = 0;
	}
	ixgbe_sync_mac_table(adapter);
}

static int ixgbe_available_rars(struct ixgbe_adapter *adapter)
{
	struct ixgbe_hw *hw = &adapter->hw;
	int i, count = 0;

	for (i = 0; i < hw->mac.num_rar_entries; i++) {
		if (adapter->mac_table[i].state == 0)
			count++;
	}
	return count;
}

/* this function destroys the first RAR entry */
static void ixgbe_mac_set_default_filter(struct ixgbe_adapter *adapter,
					 u8 *addr)
{
	struct ixgbe_hw *hw = &adapter->hw;

	memcpy(&adapter->mac_table[0].addr, addr, ETH_ALEN);
	adapter->mac_table[0].queue = VMDQ_P(0);
	adapter->mac_table[0].state = (IXGBE_MAC_STATE_DEFAULT |
				       IXGBE_MAC_STATE_IN_USE);
	hw->mac.ops.set_rar(hw, 0, adapter->mac_table[0].addr,
			    adapter->mac_table[0].queue,
			    IXGBE_RAH_AV);
}

int ixgbe_add_mac_filter(struct ixgbe_adapter *adapter, u8 *addr, u16 queue)
{
	struct ixgbe_hw *hw = &adapter->hw;
	int i;

	if (is_zero_ether_addr(addr))
		return -EINVAL;

	for (i = 0; i < hw->mac.num_rar_entries; i++) {
		if (adapter->mac_table[i].state & IXGBE_MAC_STATE_IN_USE)
			continue;
		adapter->mac_table[i].state |= (IXGBE_MAC_STATE_MODIFIED |
						IXGBE_MAC_STATE_IN_USE);
		ether_addr_copy(adapter->mac_table[i].addr, addr);
		adapter->mac_table[i].queue = queue;
		ixgbe_sync_mac_table(adapter);
		return i;
	}
	return -ENOMEM;
}

int ixgbe_del_mac_filter(struct ixgbe_adapter *adapter, u8 *addr, u16 queue)
{
	/* search table for addr, if found, set to 0 and sync */
	int i;
	struct ixgbe_hw *hw = &adapter->hw;

	if (is_zero_ether_addr(addr))
		return -EINVAL;

	for (i = 0; i < hw->mac.num_rar_entries; i++) {
		if (ether_addr_equal(addr, adapter->mac_table[i].addr) &&
		    adapter->mac_table[i].queue == queue) {
			adapter->mac_table[i].state |= IXGBE_MAC_STATE_MODIFIED;
			adapter->mac_table[i].state &= ~IXGBE_MAC_STATE_IN_USE;
3991
			eth_zero_addr(adapter->mac_table[i].addr);
3992 3993 3994 3995 3996 3997 3998
			adapter->mac_table[i].queue = 0;
			ixgbe_sync_mac_table(adapter);
			return 0;
		}
	}
	return -ENOMEM;
}
3999 4000 4001 4002 4003 4004 4005 4006 4007
/**
 * ixgbe_write_uc_addr_list - write unicast addresses to RAR table
 * @netdev: network interface device structure
 *
 * Writes unicast address list to the RAR table.
 * Returns: -ENOMEM on failure/insufficient address space
 *                0 on no addresses written
 *                X on writing X addresses to the RAR table
 **/
4008
static int ixgbe_write_uc_addr_list(struct net_device *netdev, int vfn)
4009 4010 4011 4012 4013
{
	struct ixgbe_adapter *adapter = netdev_priv(netdev);
	int count = 0;

	/* return ENOMEM indicating insufficient memory for addresses */
4014
	if (netdev_uc_count(netdev) > ixgbe_available_rars(adapter))
4015 4016
		return -ENOMEM;

4017
	if (!netdev_uc_empty(netdev)) {
4018 4019
		struct netdev_hw_addr *ha;
		netdev_for_each_uc_addr(ha, netdev) {
4020 4021
			ixgbe_del_mac_filter(adapter, ha->addr, vfn);
			ixgbe_add_mac_filter(adapter, ha->addr, vfn);
4022 4023 4024 4025 4026 4027
			count++;
		}
	}
	return count;
}

4028
/**
4029
 * ixgbe_set_rx_mode - Unicast, Multicast and Promiscuous mode set
4030 4031
 * @netdev: network interface device structure
 *
4032 4033 4034 4035
 * 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.
4036
 **/
4037
void ixgbe_set_rx_mode(struct net_device *netdev)
4038 4039 4040
{
	struct ixgbe_adapter *adapter = netdev_priv(netdev);
	struct ixgbe_hw *hw = &adapter->hw;
4041
	u32 fctrl, vmolr = IXGBE_VMOLR_BAM | IXGBE_VMOLR_AUPE;
4042
	u32 vlnctrl;
4043
	int count;
4044 4045 4046

	/* Check for Promiscuous and All Multicast modes */
	fctrl = IXGBE_READ_REG(hw, IXGBE_FCTRL);
4047
	vlnctrl = IXGBE_READ_REG(hw, IXGBE_VLNCTRL);
4048

4049
	/* set all bits that we expect to always be set */
B
Ben Greear 已提交
4050
	fctrl &= ~IXGBE_FCTRL_SBP; /* disable store-bad-packets */
4051 4052 4053 4054
	fctrl |= IXGBE_FCTRL_BAM;
	fctrl |= IXGBE_FCTRL_DPF; /* discard pause frames when FC enabled */
	fctrl |= IXGBE_FCTRL_PMCF;

4055 4056
	/* clear the bits we are changing the status of */
	fctrl &= ~(IXGBE_FCTRL_UPE | IXGBE_FCTRL_MPE);
4057
	vlnctrl &= ~(IXGBE_VLNCTRL_VFE | IXGBE_VLNCTRL_CFIEN);
4058
	if (netdev->flags & IFF_PROMISC) {
4059
		hw->addr_ctrl.user_set_promisc = true;
4060
		fctrl |= (IXGBE_FCTRL_UPE | IXGBE_FCTRL_MPE);
4061
		vmolr |= IXGBE_VMOLR_MPE;
4062 4063 4064 4065
		/* Only disable hardware filter vlans in promiscuous mode
		 * if SR-IOV and VMDQ are disabled - otherwise ensure
		 * that hardware VLAN filters remain enabled.
		 */
4066 4067
		if (adapter->flags & (IXGBE_FLAG_VMDQ_ENABLED |
				      IXGBE_FLAG_SRIOV_ENABLED))
4068
			vlnctrl |= (IXGBE_VLNCTRL_VFE | IXGBE_VLNCTRL_CFIEN);
4069
	} else {
4070 4071
		if (netdev->flags & IFF_ALLMULTI) {
			fctrl |= IXGBE_FCTRL_MPE;
4072
			vmolr |= IXGBE_VMOLR_MPE;
4073
		}
4074
		vlnctrl |= IXGBE_VLNCTRL_VFE;
4075
		hw->addr_ctrl.user_set_promisc = false;
4076 4077 4078 4079 4080 4081 4082
	}

	/*
	 * Write addresses to available RAR registers, if there is not
	 * sufficient space to store all the addresses then enable
	 * unicast promiscuous mode
	 */
4083
	count = ixgbe_write_uc_addr_list(netdev, VMDQ_P(0));
4084 4085 4086
	if (count < 0) {
		fctrl |= IXGBE_FCTRL_UPE;
		vmolr |= IXGBE_VMOLR_ROPE;
4087 4088
	}

4089 4090 4091 4092
	/* Write addresses to the MTA, if the attempt fails
	 * then we should just turn on promiscuous mode so
	 * that we can at least receive multicast traffic
	 */
4093 4094 4095 4096 4097 4098 4099
	count = ixgbe_write_mc_addr_list(netdev);
	if (count < 0) {
		fctrl |= IXGBE_FCTRL_MPE;
		vmolr |= IXGBE_VMOLR_MPE;
	} else if (count) {
		vmolr |= IXGBE_VMOLR_ROMPE;
	}
4100 4101 4102

	if (hw->mac.type != ixgbe_mac_82598EB) {
		vmolr |= IXGBE_READ_REG(hw, IXGBE_VMOLR(VMDQ_P(0))) &
4103 4104
			 ~(IXGBE_VMOLR_MPE | IXGBE_VMOLR_ROMPE |
			   IXGBE_VMOLR_ROPE);
4105
		IXGBE_WRITE_REG(hw, IXGBE_VMOLR(VMDQ_P(0)), vmolr);
4106 4107
	}

B
Ben Greear 已提交
4108 4109 4110 4111 4112 4113 4114 4115 4116 4117 4118 4119
	/* This is useful for sniffing bad packets. */
	if (adapter->netdev->features & NETIF_F_RXALL) {
		/* UPE and MPE will be handled by normal PROMISC logic
		 * in e1000e_set_rx_mode */
		fctrl |= (IXGBE_FCTRL_SBP | /* Receive bad packets */
			  IXGBE_FCTRL_BAM | /* RX All Bcast Pkts */
			  IXGBE_FCTRL_PMCF); /* RX All MAC Ctrl Pkts */

		fctrl &= ~(IXGBE_FCTRL_DPF);
		/* NOTE:  VLAN filtering is disabled by setting PROMISC */
	}

4120
	IXGBE_WRITE_REG(hw, IXGBE_VLNCTRL, vlnctrl);
4121
	IXGBE_WRITE_REG(hw, IXGBE_FCTRL, fctrl);
4122

4123
	if (netdev->features & NETIF_F_HW_VLAN_CTAG_RX)
4124 4125 4126
		ixgbe_vlan_strip_enable(adapter);
	else
		ixgbe_vlan_strip_disable(adapter);
4127 4128
}

4129 4130 4131 4132
static void ixgbe_napi_enable_all(struct ixgbe_adapter *adapter)
{
	int q_idx;

4133 4134
	for (q_idx = 0; q_idx < adapter->num_q_vectors; q_idx++) {
		ixgbe_qv_init_lock(adapter->q_vector[q_idx]);
4135
		napi_enable(&adapter->q_vector[q_idx]->napi);
4136
	}
4137 4138 4139 4140 4141 4142
}

static void ixgbe_napi_disable_all(struct ixgbe_adapter *adapter)
{
	int q_idx;

4143
	for (q_idx = 0; q_idx < adapter->num_q_vectors; q_idx++) {
4144
		napi_disable(&adapter->q_vector[q_idx]->napi);
4145
		while (!ixgbe_qv_disable(adapter->q_vector[q_idx])) {
4146
			pr_info("QV %d locked\n", q_idx);
4147
			usleep_range(1000, 20000);
4148 4149
		}
	}
4150 4151
}

J
Jeff Kirsher 已提交
4152
#ifdef CONFIG_IXGBE_DCB
4153
/**
4154 4155 4156 4157 4158 4159 4160 4161 4162 4163
 * 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;
4164
	int max_frame = adapter->netdev->mtu + ETH_HLEN + ETH_FCS_LEN;
4165

4166 4167 4168 4169 4170 4171 4172 4173 4174
	if (!(adapter->flags & IXGBE_FLAG_DCB_ENABLED)) {
		if (hw->mac.type == ixgbe_mac_82598EB)
			netif_set_gso_max_size(adapter->netdev, 65536);
		return;
	}

	if (hw->mac.type == ixgbe_mac_82598EB)
		netif_set_gso_max_size(adapter->netdev, 32768);

4175
#ifdef IXGBE_FCOE
4176 4177
	if (adapter->netdev->features & NETIF_F_FCOE_MTU)
		max_frame = max(max_frame, IXGBE_FCOE_JUMBO_FRAME_SIZE);
4178
#endif
4179 4180 4181

	/* reconfigure the hardware */
	if (adapter->dcbx_cap & DCB_CAP_DCBX_VER_CEE) {
4182 4183 4184 4185 4186
		ixgbe_dcb_calculate_tc_credits(hw, &adapter->dcb_cfg, max_frame,
						DCB_TX_CONFIG);
		ixgbe_dcb_calculate_tc_credits(hw, &adapter->dcb_cfg, max_frame,
						DCB_RX_CONFIG);
		ixgbe_dcb_hw_config(hw, &adapter->dcb_cfg);
4187 4188 4189 4190 4191 4192 4193
	} else if (adapter->ixgbe_ieee_ets && adapter->ixgbe_ieee_pfc) {
		ixgbe_dcb_hw_ets(&adapter->hw,
				 adapter->ixgbe_ieee_ets,
				 max_frame);
		ixgbe_dcb_hw_pfc_config(&adapter->hw,
					adapter->ixgbe_ieee_pfc->pfc_en,
					adapter->ixgbe_ieee_ets->prio_tc);
4194
	}
4195 4196 4197

	/* Enable RSS Hash per TC */
	if (hw->mac.type != ixgbe_mac_82598EB) {
4198 4199
		u32 msb = 0;
		u16 rss_i = adapter->ring_feature[RING_F_RSS].indices - 1;
4200

4201 4202 4203 4204
		while (rss_i) {
			msb++;
			rss_i >>= 1;
		}
4205

4206 4207
		/* write msb to all 8 TCs in one write */
		IXGBE_WRITE_REG(hw, IXGBE_RQTC, msb * 0x11111111);
4208
	}
4209
}
4210 4211 4212 4213 4214
#endif

/* Additional bittime to account for IXGBE framing */
#define IXGBE_ETH_FRAMING 20

4215
/**
4216 4217 4218
 * ixgbe_hpbthresh - calculate high water mark for flow control
 *
 * @adapter: board private structure to calculate for
4219
 * @pb: packet buffer to calculate
4220 4221 4222 4223 4224 4225 4226 4227 4228 4229 4230 4231 4232
 */
static int ixgbe_hpbthresh(struct ixgbe_adapter *adapter, int pb)
{
	struct ixgbe_hw *hw = &adapter->hw;
	struct net_device *dev = adapter->netdev;
	int link, tc, kb, marker;
	u32 dv_id, rx_pba;

	/* Calculate max LAN frame size */
	tc = link = dev->mtu + ETH_HLEN + ETH_FCS_LEN + IXGBE_ETH_FRAMING;

#ifdef IXGBE_FCOE
	/* FCoE traffic class uses FCOE jumbo frames */
4233 4234 4235 4236
	if ((dev->features & NETIF_F_FCOE_MTU) &&
	    (tc < IXGBE_FCOE_JUMBO_FRAME_SIZE) &&
	    (pb == ixgbe_fcoe_get_tc(adapter)))
		tc = IXGBE_FCOE_JUMBO_FRAME_SIZE;
4237
#endif
4238

4239 4240 4241
	/* Calculate delay value for device */
	switch (hw->mac.type) {
	case ixgbe_mac_X540:
4242 4243
	case ixgbe_mac_X550:
	case ixgbe_mac_X550EM_x:
4244 4245 4246 4247 4248 4249 4250 4251 4252 4253 4254 4255 4256 4257 4258 4259 4260 4261 4262 4263 4264 4265 4266 4267 4268 4269 4270 4271 4272 4273 4274
		dv_id = IXGBE_DV_X540(link, tc);
		break;
	default:
		dv_id = IXGBE_DV(link, tc);
		break;
	}

	/* Loopback switch introduces additional latency */
	if (adapter->flags & IXGBE_FLAG_SRIOV_ENABLED)
		dv_id += IXGBE_B2BT(tc);

	/* Delay value is calculated in bit times convert to KB */
	kb = IXGBE_BT2KB(dv_id);
	rx_pba = IXGBE_READ_REG(hw, IXGBE_RXPBSIZE(pb)) >> 10;

	marker = rx_pba - kb;

	/* It is possible that the packet buffer is not large enough
	 * to provide required headroom. In this case throw an error
	 * to user and a do the best we can.
	 */
	if (marker < 0) {
		e_warn(drv, "Packet Buffer(%i) can not provide enough"
			    "headroom to support flow control."
			    "Decrease MTU or number of traffic classes\n", pb);
		marker = tc + 1;
	}

	return marker;
}

4275
/**
4276 4277 4278
 * ixgbe_lpbthresh - calculate low water mark for for flow control
 *
 * @adapter: board private structure to calculate for
4279
 * @pb: packet buffer to calculate
4280
 */
4281
static int ixgbe_lpbthresh(struct ixgbe_adapter *adapter, int pb)
4282 4283 4284 4285 4286 4287 4288 4289 4290
{
	struct ixgbe_hw *hw = &adapter->hw;
	struct net_device *dev = adapter->netdev;
	int tc;
	u32 dv_id;

	/* Calculate max LAN frame size */
	tc = dev->mtu + ETH_HLEN + ETH_FCS_LEN;

4291 4292 4293 4294 4295 4296 4297 4298
#ifdef IXGBE_FCOE
	/* FCoE traffic class uses FCOE jumbo frames */
	if ((dev->features & NETIF_F_FCOE_MTU) &&
	    (tc < IXGBE_FCOE_JUMBO_FRAME_SIZE) &&
	    (pb == netdev_get_prio_tc_map(dev, adapter->fcoe.up)))
		tc = IXGBE_FCOE_JUMBO_FRAME_SIZE;
#endif

4299 4300 4301
	/* Calculate delay value for device */
	switch (hw->mac.type) {
	case ixgbe_mac_X540:
4302 4303
	case ixgbe_mac_X550:
	case ixgbe_mac_X550EM_x:
4304 4305 4306 4307 4308 4309 4310 4311 4312 4313 4314 4315 4316 4317 4318 4319 4320 4321 4322 4323 4324 4325 4326 4327 4328
		dv_id = IXGBE_LOW_DV_X540(tc);
		break;
	default:
		dv_id = IXGBE_LOW_DV(tc);
		break;
	}

	/* Delay value is calculated in bit times convert to KB */
	return IXGBE_BT2KB(dv_id);
}

/*
 * ixgbe_pbthresh_setup - calculate and setup high low water marks
 */
static void ixgbe_pbthresh_setup(struct ixgbe_adapter *adapter)
{
	struct ixgbe_hw *hw = &adapter->hw;
	int num_tc = netdev_get_num_tc(adapter->netdev);
	int i;

	if (!num_tc)
		num_tc = 1;

	for (i = 0; i < num_tc; i++) {
		hw->fc.high_water[i] = ixgbe_hpbthresh(adapter, i);
4329
		hw->fc.low_water[i] = ixgbe_lpbthresh(adapter, i);
4330 4331

		/* Low water marks must not be larger than high water marks */
4332 4333
		if (hw->fc.low_water[i] > hw->fc.high_water[i])
			hw->fc.low_water[i] = 0;
4334
	}
4335 4336 4337

	for (; i < MAX_TRAFFIC_CLASS; i++)
		hw->fc.high_water[i] = 0;
4338 4339
}

4340 4341 4342
static void ixgbe_configure_pb(struct ixgbe_adapter *adapter)
{
	struct ixgbe_hw *hw = &adapter->hw;
4343 4344
	int hdrm;
	u8 tc = netdev_get_num_tc(adapter->netdev);
4345 4346 4347

	if (adapter->flags & IXGBE_FLAG_FDIR_HASH_CAPABLE ||
	    adapter->flags & IXGBE_FLAG_FDIR_PERFECT_CAPABLE)
4348 4349 4350
		hdrm = 32 << adapter->fdir_pballoc;
	else
		hdrm = 0;
4351

4352
	hw->mac.ops.set_rxpba(hw, tc, hdrm, PBA_STRATEGY_EQUAL);
4353
	ixgbe_pbthresh_setup(adapter);
4354 4355
}

4356 4357 4358
static void ixgbe_fdir_filter_restore(struct ixgbe_adapter *adapter)
{
	struct ixgbe_hw *hw = &adapter->hw;
4359
	struct hlist_node *node2;
4360 4361 4362 4363 4364 4365 4366
	struct ixgbe_fdir_filter *filter;

	spin_lock(&adapter->fdir_perfect_lock);

	if (!hlist_empty(&adapter->fdir_filter_list))
		ixgbe_fdir_set_input_mask_82599(hw, &adapter->fdir_mask);

4367
	hlist_for_each_entry_safe(filter, node2,
4368 4369
				  &adapter->fdir_filter_list, fdir_node) {
		ixgbe_fdir_write_perfect_filter_82599(hw,
4370 4371 4372 4373 4374
				&filter->filter,
				filter->sw_idx,
				(filter->action == IXGBE_FDIR_DROP_QUEUE) ?
				IXGBE_FDIR_DROP_QUEUE :
				adapter->rx_ring[filter->action]->reg_idx);
4375 4376 4377 4378 4379
	}

	spin_unlock(&adapter->fdir_perfect_lock);
}

4380 4381 4382 4383 4384 4385 4386 4387 4388 4389 4390 4391 4392 4393 4394 4395 4396 4397 4398
static void ixgbe_macvlan_set_rx_mode(struct net_device *dev, unsigned int pool,
				      struct ixgbe_adapter *adapter)
{
	struct ixgbe_hw *hw = &adapter->hw;
	u32 vmolr;

	/* No unicast promiscuous support for VMDQ devices. */
	vmolr = IXGBE_READ_REG(hw, IXGBE_VMOLR(pool));
	vmolr |= (IXGBE_VMOLR_ROMPE | IXGBE_VMOLR_BAM | IXGBE_VMOLR_AUPE);

	/* clear the affected bit */
	vmolr &= ~IXGBE_VMOLR_MPE;

	if (dev->flags & IFF_ALLMULTI) {
		vmolr |= IXGBE_VMOLR_MPE;
	} else {
		vmolr |= IXGBE_VMOLR_ROMPE;
		hw->mac.ops.update_mc_addr_list(hw, dev);
	}
4399
	ixgbe_write_uc_addr_list(adapter->netdev, pool);
4400 4401 4402 4403 4404 4405
	IXGBE_WRITE_REG(hw, IXGBE_VMOLR(pool), vmolr);
}

static void ixgbe_fwd_psrtype(struct ixgbe_fwd_adapter *vadapter)
{
	struct ixgbe_adapter *adapter = vadapter->real_adapter;
4406
	int rss_i = adapter->num_rx_queues_per_pool;
4407 4408 4409 4410 4411 4412 4413 4414 4415 4416 4417 4418 4419 4420 4421 4422 4423 4424 4425 4426 4427 4428 4429 4430 4431 4432 4433 4434 4435 4436 4437 4438 4439 4440 4441
	struct ixgbe_hw *hw = &adapter->hw;
	u16 pool = vadapter->pool;
	u32 psrtype = IXGBE_PSRTYPE_TCPHDR |
		      IXGBE_PSRTYPE_UDPHDR |
		      IXGBE_PSRTYPE_IPV4HDR |
		      IXGBE_PSRTYPE_L2HDR |
		      IXGBE_PSRTYPE_IPV6HDR;

	if (hw->mac.type == ixgbe_mac_82598EB)
		return;

	if (rss_i > 3)
		psrtype |= 2 << 29;
	else if (rss_i > 1)
		psrtype |= 1 << 29;

	IXGBE_WRITE_REG(hw, IXGBE_PSRTYPE(VMDQ_P(pool)), psrtype);
}

/**
 * ixgbe_clean_rx_ring - Free Rx Buffers per Queue
 * @rx_ring: ring to free buffers from
 **/
static void ixgbe_clean_rx_ring(struct ixgbe_ring *rx_ring)
{
	struct device *dev = rx_ring->dev;
	unsigned long size;
	u16 i;

	/* ring already cleared, nothing to do */
	if (!rx_ring->rx_buffer_info)
		return;

	/* Free all the Rx ring sk_buffs */
	for (i = 0; i < rx_ring->count; i++) {
A
Alexander Duyck 已提交
4442
		struct ixgbe_rx_buffer *rx_buffer = &rx_ring->rx_buffer_info[i];
4443 4444 4445

		if (rx_buffer->skb) {
			struct sk_buff *skb = rx_buffer->skb;
A
Alexander Duyck 已提交
4446
			if (IXGBE_CB(skb)->page_released)
4447 4448 4449 4450 4451
				dma_unmap_page(dev,
					       IXGBE_CB(skb)->dma,
					       ixgbe_rx_bufsz(rx_ring),
					       DMA_FROM_DEVICE);
			dev_kfree_skb(skb);
4452
			rx_buffer->skb = NULL;
4453
		}
A
Alexander Duyck 已提交
4454 4455 4456 4457 4458 4459 4460 4461

		if (!rx_buffer->page)
			continue;

		dma_unmap_page(dev, rx_buffer->dma,
			       ixgbe_rx_pg_size(rx_ring), DMA_FROM_DEVICE);
		__free_pages(rx_buffer->page, ixgbe_rx_pg_order(rx_ring));

4462 4463 4464 4465 4466 4467 4468 4469 4470 4471 4472 4473 4474 4475 4476 4477 4478 4479 4480 4481 4482 4483 4484 4485 4486 4487 4488 4489
		rx_buffer->page = NULL;
	}

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

static void ixgbe_disable_fwd_ring(struct ixgbe_fwd_adapter *vadapter,
				   struct ixgbe_ring *rx_ring)
{
	struct ixgbe_adapter *adapter = vadapter->real_adapter;
	int index = rx_ring->queue_index + vadapter->rx_base_queue;

	/* shutdown specific queue receive and wait for dma to settle */
	ixgbe_disable_rx_queue(adapter, rx_ring);
	usleep_range(10000, 20000);
	ixgbe_irq_disable_queues(adapter, ((u64)1 << index));
	ixgbe_clean_rx_ring(rx_ring);
	rx_ring->l2_accel_priv = NULL;
}

4490 4491
static int ixgbe_fwd_ring_down(struct net_device *vdev,
			       struct ixgbe_fwd_adapter *accel)
4492 4493 4494 4495 4496 4497 4498 4499 4500 4501 4502 4503 4504 4505 4506 4507 4508 4509 4510 4511 4512 4513 4514 4515 4516 4517 4518 4519 4520 4521 4522 4523 4524 4525 4526 4527 4528 4529 4530 4531 4532 4533 4534 4535 4536 4537 4538 4539 4540 4541 4542 4543 4544 4545 4546 4547 4548 4549 4550 4551 4552 4553 4554 4555 4556 4557 4558 4559 4560 4561 4562 4563 4564 4565 4566 4567 4568 4569 4570 4571 4572 4573 4574 4575 4576 4577 4578 4579 4580 4581 4582 4583 4584 4585 4586 4587 4588
{
	struct ixgbe_adapter *adapter = accel->real_adapter;
	unsigned int rxbase = accel->rx_base_queue;
	unsigned int txbase = accel->tx_base_queue;
	int i;

	netif_tx_stop_all_queues(vdev);

	for (i = 0; i < adapter->num_rx_queues_per_pool; i++) {
		ixgbe_disable_fwd_ring(accel, adapter->rx_ring[rxbase + i]);
		adapter->rx_ring[rxbase + i]->netdev = adapter->netdev;
	}

	for (i = 0; i < adapter->num_rx_queues_per_pool; i++) {
		adapter->tx_ring[txbase + i]->l2_accel_priv = NULL;
		adapter->tx_ring[txbase + i]->netdev = adapter->netdev;
	}


	return 0;
}

static int ixgbe_fwd_ring_up(struct net_device *vdev,
			     struct ixgbe_fwd_adapter *accel)
{
	struct ixgbe_adapter *adapter = accel->real_adapter;
	unsigned int rxbase, txbase, queues;
	int i, baseq, err = 0;

	if (!test_bit(accel->pool, &adapter->fwd_bitmask))
		return 0;

	baseq = accel->pool * adapter->num_rx_queues_per_pool;
	netdev_dbg(vdev, "pool %i:%i queues %i:%i VSI bitmask %lx\n",
		   accel->pool, adapter->num_rx_pools,
		   baseq, baseq + adapter->num_rx_queues_per_pool,
		   adapter->fwd_bitmask);

	accel->netdev = vdev;
	accel->rx_base_queue = rxbase = baseq;
	accel->tx_base_queue = txbase = baseq;

	for (i = 0; i < adapter->num_rx_queues_per_pool; i++)
		ixgbe_disable_fwd_ring(accel, adapter->rx_ring[rxbase + i]);

	for (i = 0; i < adapter->num_rx_queues_per_pool; i++) {
		adapter->rx_ring[rxbase + i]->netdev = vdev;
		adapter->rx_ring[rxbase + i]->l2_accel_priv = accel;
		ixgbe_configure_rx_ring(adapter, adapter->rx_ring[rxbase + i]);
	}

	for (i = 0; i < adapter->num_rx_queues_per_pool; i++) {
		adapter->tx_ring[txbase + i]->netdev = vdev;
		adapter->tx_ring[txbase + i]->l2_accel_priv = accel;
	}

	queues = min_t(unsigned int,
		       adapter->num_rx_queues_per_pool, vdev->num_tx_queues);
	err = netif_set_real_num_tx_queues(vdev, queues);
	if (err)
		goto fwd_queue_err;

	err = netif_set_real_num_rx_queues(vdev, queues);
	if (err)
		goto fwd_queue_err;

	if (is_valid_ether_addr(vdev->dev_addr))
		ixgbe_add_mac_filter(adapter, vdev->dev_addr, accel->pool);

	ixgbe_fwd_psrtype(accel);
	ixgbe_macvlan_set_rx_mode(vdev, accel->pool, adapter);
	return err;
fwd_queue_err:
	ixgbe_fwd_ring_down(vdev, accel);
	return err;
}

static void ixgbe_configure_dfwd(struct ixgbe_adapter *adapter)
{
	struct net_device *upper;
	struct list_head *iter;
	int err;

	netdev_for_each_all_upper_dev_rcu(adapter->netdev, upper, iter) {
		if (netif_is_macvlan(upper)) {
			struct macvlan_dev *dfwd = netdev_priv(upper);
			struct ixgbe_fwd_adapter *vadapter = dfwd->fwd_priv;

			if (dfwd->fwd_priv) {
				err = ixgbe_fwd_ring_up(upper, vadapter);
				if (err)
					continue;
			}
		}
	}
}

4589 4590
static void ixgbe_configure(struct ixgbe_adapter *adapter)
{
4591 4592
	struct ixgbe_hw *hw = &adapter->hw;

4593
	ixgbe_configure_pb(adapter);
J
Jeff Kirsher 已提交
4594
#ifdef CONFIG_IXGBE_DCB
4595
	ixgbe_configure_dcb(adapter);
4596
#endif
4597 4598 4599 4600 4601
	/*
	 * We must restore virtualization before VLANs or else
	 * the VLVF registers will not be populated
	 */
	ixgbe_configure_virtualization(adapter);
4602

4603
	ixgbe_set_rx_mode(adapter->netdev);
4604 4605
	ixgbe_restore_vlan(adapter);

4606 4607 4608 4609 4610 4611 4612 4613 4614
	switch (hw->mac.type) {
	case ixgbe_mac_82599EB:
	case ixgbe_mac_X540:
		hw->mac.ops.disable_rx_buff(hw);
		break;
	default:
		break;
	}

4615
	if (adapter->flags & IXGBE_FLAG_FDIR_HASH_CAPABLE) {
4616 4617
		ixgbe_init_fdir_signature_82599(&adapter->hw,
						adapter->fdir_pballoc);
4618 4619 4620 4621
	} else if (adapter->flags & IXGBE_FLAG_FDIR_PERFECT_CAPABLE) {
		ixgbe_init_fdir_perfect_82599(&adapter->hw,
					      adapter->fdir_pballoc);
		ixgbe_fdir_filter_restore(adapter);
4622
	}
4623

4624 4625 4626 4627 4628 4629 4630 4631 4632
	switch (hw->mac.type) {
	case ixgbe_mac_82599EB:
	case ixgbe_mac_X540:
		hw->mac.ops.enable_rx_buff(hw);
		break;
	default:
		break;
	}

4633 4634 4635 4636 4637
#ifdef IXGBE_FCOE
	/* configure FCoE L2 filters, redirection table, and Rx control */
	ixgbe_configure_fcoe(adapter);

#endif /* IXGBE_FCOE */
4638 4639
	ixgbe_configure_tx(adapter);
	ixgbe_configure_rx(adapter);
4640
	ixgbe_configure_dfwd(adapter);
4641 4642
}

4643 4644 4645 4646 4647 4648 4649
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:
4650 4651 4652 4653
	case ixgbe_phy_sfp_passive_tyco:
	case ixgbe_phy_sfp_passive_unknown:
	case ixgbe_phy_sfp_active_unknown:
	case ixgbe_phy_sfp_ftl_active:
4654 4655 4656 4657
	case ixgbe_phy_qsfp_passive_unknown:
	case ixgbe_phy_qsfp_active_unknown:
	case ixgbe_phy_qsfp_intel:
	case ixgbe_phy_qsfp_unknown:
4658 4659
	/* ixgbe_phy_none is set when no SFP module is present */
	case ixgbe_phy_none:
4660
		return true;
4661 4662 4663
	case ixgbe_phy_nl:
		if (hw->mac.type == ixgbe_mac_82598EB)
			return true;
4664 4665 4666 4667 4668
	default:
		return false;
	}
}

4669
/**
4670 4671 4672 4673 4674
 * ixgbe_sfp_link_config - set up SFP+ link
 * @adapter: pointer to private adapter struct
 **/
static void ixgbe_sfp_link_config(struct ixgbe_adapter *adapter)
{
4675
	/*
S
Stephen Hemminger 已提交
4676
	 * We are assuming the worst case scenario here, and that
4677 4678 4679 4680 4681 4682
	 * is that an SFP was inserted/removed after the reset
	 * but before SFP detection was enabled.  As such the best
	 * solution is to just start searching as soon as we start
	 */
	if (adapter->hw.mac.type == ixgbe_mac_82598EB)
		adapter->flags2 |= IXGBE_FLAG2_SEARCH_FOR_SFP;
4683

4684
	adapter->flags2 |= IXGBE_FLAG2_SFP_NEEDS_RESET;
4685 4686 4687 4688
}

/**
 * ixgbe_non_sfp_link_config - set up non-SFP+ link
4689 4690 4691 4692
 * @hw: pointer to private hardware struct
 *
 * Returns 0 on success, negative on failure
 **/
4693
static int ixgbe_non_sfp_link_config(struct ixgbe_hw *hw)
4694
{
J
Josh Hay 已提交
4695 4696
	u32 speed;
	bool autoneg, link_up = false;
4697 4698 4699
	u32 ret = IXGBE_ERR_LINK_SETUP;

	if (hw->mac.ops.check_link)
J
Josh Hay 已提交
4700
		ret = hw->mac.ops.check_link(hw, &speed, &link_up, false);
4701 4702

	if (ret)
4703
		return ret;
4704

J
Josh Hay 已提交
4705 4706 4707 4708
	speed = hw->phy.autoneg_advertised;
	if ((!speed) && (hw->mac.ops.get_link_capabilities))
		ret = hw->mac.ops.get_link_capabilities(hw, &speed,
							&autoneg);
4709
	if (ret)
4710
		return ret;
4711

4712
	if (hw->mac.ops.setup_link)
J
Josh Hay 已提交
4713
		ret = hw->mac.ops.setup_link(hw, speed, link_up);
4714

4715 4716 4717
	return ret;
}

4718
static void ixgbe_setup_gpie(struct ixgbe_adapter *adapter)
4719 4720
{
	struct ixgbe_hw *hw = &adapter->hw;
4721
	u32 gpie = 0;
4722

4723
	if (adapter->flags & IXGBE_FLAG_MSIX_ENABLED) {
4724 4725 4726
		gpie = IXGBE_GPIE_MSIX_MODE | IXGBE_GPIE_PBA_SUPPORT |
		       IXGBE_GPIE_OCD;
		gpie |= IXGBE_GPIE_EIAME;
4727 4728 4729 4730 4731 4732 4733 4734 4735
		/*
		 * use EIAM to auto-mask when MSI-X interrupt is asserted
		 * this saves a register write for every interrupt
		 */
		switch (hw->mac.type) {
		case ixgbe_mac_82598EB:
			IXGBE_WRITE_REG(hw, IXGBE_EIAM, IXGBE_EICS_RTX_QUEUE);
			break;
		case ixgbe_mac_82599EB:
D
Don Skidmore 已提交
4736
		case ixgbe_mac_X540:
4737 4738
		case ixgbe_mac_X550:
		case ixgbe_mac_X550EM_x:
D
Don Skidmore 已提交
4739
		default:
4740 4741 4742 4743 4744
			IXGBE_WRITE_REG(hw, IXGBE_EIAM_EX(0), 0xFFFFFFFF);
			IXGBE_WRITE_REG(hw, IXGBE_EIAM_EX(1), 0xFFFFFFFF);
			break;
		}
	} else {
4745 4746 4747 4748
		/* 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);
	}
4749

4750 4751 4752 4753 4754
	/* XXX: to interrupt immediately for EICS writes, enable this */
	/* gpie |= IXGBE_GPIE_EIMEN; */

	if (adapter->flags & IXGBE_FLAG_SRIOV_ENABLED) {
		gpie &= ~IXGBE_GPIE_VTMODE_MASK;
4755 4756 4757 4758 4759 4760 4761 4762 4763 4764 4765 4766

		switch (adapter->ring_feature[RING_F_VMDQ].mask) {
		case IXGBE_82599_VMDQ_8Q_MASK:
			gpie |= IXGBE_GPIE_VTMODE_16;
			break;
		case IXGBE_82599_VMDQ_4Q_MASK:
			gpie |= IXGBE_GPIE_VTMODE_32;
			break;
		default:
			gpie |= IXGBE_GPIE_VTMODE_64;
			break;
		}
4767 4768
	}

4769
	/* Enable Thermal over heat sensor interrupt */
4770 4771 4772 4773 4774 4775 4776 4777 4778 4779 4780 4781
	if (adapter->flags2 & IXGBE_FLAG2_TEMP_SENSOR_CAPABLE) {
		switch (adapter->hw.mac.type) {
		case ixgbe_mac_82599EB:
			gpie |= IXGBE_SDP0_GPIEN;
			break;
		case ixgbe_mac_X540:
			gpie |= IXGBE_EIMS_TS;
			break;
		default:
			break;
		}
	}
4782

4783 4784
	/* Enable fan failure interrupt */
	if (adapter->flags & IXGBE_FLAG_FAN_FAIL_CAPABLE)
4785 4786
		gpie |= IXGBE_SDP1_GPIEN;

4787
	if (hw->mac.type == ixgbe_mac_82599EB) {
4788 4789
		gpie |= IXGBE_SDP1_GPIEN;
		gpie |= IXGBE_SDP2_GPIEN;
4790
	}
4791 4792 4793 4794

	IXGBE_WRITE_REG(hw, IXGBE_GPIE, gpie);
}

4795
static void ixgbe_up_complete(struct ixgbe_adapter *adapter)
4796 4797 4798 4799 4800 4801 4802
{
	struct ixgbe_hw *hw = &adapter->hw;
	int err;
	u32 ctrl_ext;

	ixgbe_get_hw_control(adapter);
	ixgbe_setup_gpie(adapter);
4803

4804 4805 4806 4807 4808
	if (adapter->flags & IXGBE_FLAG_MSIX_ENABLED)
		ixgbe_configure_msix(adapter);
	else
		ixgbe_configure_msi_and_legacy(adapter);

4809 4810
	/* enable the optics for 82599 SFP+ fiber */
	if (hw->mac.ops.enable_tx_laser)
4811 4812
		hw->mac.ops.enable_tx_laser(hw);

4813
	smp_mb__before_atomic();
4814
	clear_bit(__IXGBE_DOWN, &adapter->state);
4815 4816
	ixgbe_napi_enable_all(adapter);

4817 4818 4819 4820 4821 4822 4823 4824
	if (ixgbe_is_sfp(hw)) {
		ixgbe_sfp_link_config(adapter);
	} else {
		err = ixgbe_non_sfp_link_config(hw);
		if (err)
			e_err(probe, "link_config FAILED %d\n", err);
	}

4825 4826
	/* clear any pending interrupts, may auto mask */
	IXGBE_READ_REG(hw, IXGBE_EICR);
4827
	ixgbe_irq_enable(adapter, true, true);
4828

4829 4830 4831 4832 4833 4834 4835
	/*
	 * If this adapter has a fan, check to see if we had a failure
	 * before we enabled the interrupt.
	 */
	if (adapter->flags & IXGBE_FLAG_FAN_FAIL_CAPABLE) {
		u32 esdp = IXGBE_READ_REG(hw, IXGBE_ESDP);
		if (esdp & IXGBE_ESDP_SDP1)
4836
			e_crit(drv, "Fan has stopped, replace the adapter\n");
4837 4838
	}

4839 4840
	/* bring the link up in the watchdog, this could race with our first
	 * link up interrupt but shouldn't be a problem */
4841 4842
	adapter->flags |= IXGBE_FLAG_NEED_LINK_UPDATE;
	adapter->link_check_timeout = jiffies;
4843
	mod_timer(&adapter->service_timer, jiffies);
4844 4845 4846 4847 4848

	/* Set PF Reset Done bit so PF/VF Mail Ops can work */
	ctrl_ext = IXGBE_READ_REG(hw, IXGBE_CTRL_EXT);
	ctrl_ext |= IXGBE_CTRL_EXT_PFRSTD;
	IXGBE_WRITE_REG(hw, IXGBE_CTRL_EXT, ctrl_ext);
4849 4850
}

4851 4852 4853
void ixgbe_reinit_locked(struct ixgbe_adapter *adapter)
{
	WARN_ON(in_interrupt());
4854 4855 4856
	/* put off any impending NetWatchDogTimeout */
	adapter->netdev->trans_start = jiffies;

4857
	while (test_and_set_bit(__IXGBE_RESETTING, &adapter->state))
4858
		usleep_range(1000, 2000);
4859
	ixgbe_down(adapter);
4860 4861 4862 4863 4864 4865 4866 4867
	/*
	 * If SR-IOV enabled then wait a bit before bringing the adapter
	 * back up to give the VFs time to respond to the reset.  The
	 * two second wait is based upon the watchdog timer cycle in
	 * the VF driver.
	 */
	if (adapter->flags & IXGBE_FLAG_SRIOV_ENABLED)
		msleep(2000);
4868 4869 4870 4871
	ixgbe_up(adapter);
	clear_bit(__IXGBE_RESETTING, &adapter->state);
}

4872
void ixgbe_up(struct ixgbe_adapter *adapter)
4873 4874 4875 4876
{
	/* hardware has been reset, we need to reload some things */
	ixgbe_configure(adapter);

4877
	ixgbe_up_complete(adapter);
4878 4879 4880 4881
}

void ixgbe_reset(struct ixgbe_adapter *adapter)
{
4882
	struct ixgbe_hw *hw = &adapter->hw;
4883
	struct net_device *netdev = adapter->netdev;
4884
	int err;
4885
	u8 old_addr[ETH_ALEN];
4886

4887 4888
	if (ixgbe_removed(hw->hw_addr))
		return;
4889 4890 4891 4892 4893 4894 4895 4896 4897
	/* lock SFP init bit to prevent race conditions with the watchdog */
	while (test_and_set_bit(__IXGBE_IN_SFP_INIT, &adapter->state))
		usleep_range(1000, 2000);

	/* clear all SFP and link config related flags while holding SFP_INIT */
	adapter->flags2 &= ~(IXGBE_FLAG2_SEARCH_FOR_SFP |
			     IXGBE_FLAG2_SFP_NEEDS_RESET);
	adapter->flags &= ~IXGBE_FLAG_NEED_LINK_CONFIG;

4898
	err = hw->mac.ops.init_hw(hw);
4899 4900 4901
	switch (err) {
	case 0:
	case IXGBE_ERR_SFP_NOT_PRESENT:
4902
	case IXGBE_ERR_SFP_NOT_SUPPORTED:
4903 4904
		break;
	case IXGBE_ERR_MASTER_REQUESTS_PENDING:
4905
		e_dev_err("master disable timed out\n");
4906
		break;
4907 4908
	case IXGBE_ERR_EEPROM_VERSION:
		/* We are running on a pre-production device, log a warning */
4909
		e_dev_warn("This device is a pre-production adapter/LOM. "
S
Stephen Hemminger 已提交
4910
			   "Please be aware there may be issues associated with "
4911 4912 4913 4914
			   "your hardware.  If you are experiencing problems "
			   "please contact your Intel or hardware "
			   "representative who provided you with this "
			   "hardware.\n");
4915
		break;
4916
	default:
4917
		e_dev_err("Hardware Error: %d\n", err);
4918
	}
4919

4920
	clear_bit(__IXGBE_IN_SFP_INIT, &adapter->state);
4921 4922 4923 4924
	/* do not flush user set addresses */
	memcpy(old_addr, &adapter->mac_table[0].addr, netdev->addr_len);
	ixgbe_flush_sw_mac_table(adapter);
	ixgbe_mac_set_default_filter(adapter, old_addr);
4925 4926 4927 4928

	/* update SAN MAC vmdq pool selection */
	if (hw->mac.san_mac_rar_index)
		hw->mac.ops.set_vmdq_san_mac(hw, VMDQ_P(0));
4929

4930
	if (test_bit(__IXGBE_PTP_RUNNING, &adapter->state))
4931
		ixgbe_ptp_reset(adapter);
4932 4933 4934 4935 4936 4937
}

/**
 * ixgbe_clean_tx_ring - Free Tx Buffers
 * @tx_ring: ring to be cleaned
 **/
4938
static void ixgbe_clean_tx_ring(struct ixgbe_ring *tx_ring)
4939 4940 4941
{
	struct ixgbe_tx_buffer *tx_buffer_info;
	unsigned long size;
4942
	u16 i;
4943

4944 4945 4946
	/* ring already cleared, nothing to do */
	if (!tx_ring->tx_buffer_info)
		return;
4947

4948
	/* Free all the Tx ring sk_buffs */
4949 4950
	for (i = 0; i < tx_ring->count; i++) {
		tx_buffer_info = &tx_ring->tx_buffer_info[i];
4951
		ixgbe_unmap_and_free_tx_resource(tx_ring, tx_buffer_info);
4952 4953
	}

4954 4955
	netdev_tx_reset_queue(txring_txq(tx_ring));

4956 4957 4958 4959 4960 4961 4962 4963 4964 4965 4966
	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;
}

/**
4967
 * ixgbe_clean_all_rx_rings - Free Rx Buffers for all queues
4968 4969
 * @adapter: board private structure
 **/
4970
static void ixgbe_clean_all_rx_rings(struct ixgbe_adapter *adapter)
4971 4972 4973
{
	int i;

4974
	for (i = 0; i < adapter->num_rx_queues; i++)
4975
		ixgbe_clean_rx_ring(adapter->rx_ring[i]);
4976 4977 4978
}

/**
4979
 * ixgbe_clean_all_tx_rings - Free Tx Buffers for all queues
4980 4981
 * @adapter: board private structure
 **/
4982
static void ixgbe_clean_all_tx_rings(struct ixgbe_adapter *adapter)
4983 4984 4985
{
	int i;

4986
	for (i = 0; i < adapter->num_tx_queues; i++)
4987
		ixgbe_clean_tx_ring(adapter->tx_ring[i]);
4988 4989
}

4990 4991
static void ixgbe_fdir_filter_exit(struct ixgbe_adapter *adapter)
{
4992
	struct hlist_node *node2;
4993 4994 4995 4996
	struct ixgbe_fdir_filter *filter;

	spin_lock(&adapter->fdir_perfect_lock);

4997
	hlist_for_each_entry_safe(filter, node2,
4998 4999 5000 5001 5002 5003 5004 5005 5006
				  &adapter->fdir_filter_list, fdir_node) {
		hlist_del(&filter->fdir_node);
		kfree(filter);
	}
	adapter->fdir_filter_count = 0;

	spin_unlock(&adapter->fdir_perfect_lock);
}

5007 5008 5009
void ixgbe_down(struct ixgbe_adapter *adapter)
{
	struct net_device *netdev = adapter->netdev;
5010
	struct ixgbe_hw *hw = &adapter->hw;
5011 5012
	struct net_device *upper;
	struct list_head *iter;
5013
	int i;
5014 5015

	/* signal that we are down to the interrupt handler */
5016 5017
	if (test_and_set_bit(__IXGBE_DOWN, &adapter->state))
		return; /* do nothing if already down */
5018 5019

	/* disable receives */
5020
	hw->mac.ops.disable_rx(hw);
5021

5022 5023 5024 5025 5026
	/* disable all enabled rx queues */
	for (i = 0; i < adapter->num_rx_queues; i++)
		/* this call also flushes the previous write */
		ixgbe_disable_rx_queue(adapter, adapter->rx_ring[i]);

5027
	usleep_range(10000, 20000);
5028

5029 5030
	netif_tx_stop_all_queues(netdev);

5031
	/* call carrier off first to avoid false dev_watchdog timeouts */
5032 5033 5034
	netif_carrier_off(netdev);
	netif_tx_disable(netdev);

5035 5036 5037 5038 5039 5040 5041 5042 5043 5044 5045 5046 5047
	/* disable any upper devices */
	netdev_for_each_all_upper_dev_rcu(adapter->netdev, upper, iter) {
		if (netif_is_macvlan(upper)) {
			struct macvlan_dev *vlan = netdev_priv(upper);

			if (vlan->fwd_priv) {
				netif_tx_stop_all_queues(upper);
				netif_carrier_off(upper);
				netif_tx_disable(upper);
			}
		}
	}

5048 5049 5050 5051
	ixgbe_irq_disable(adapter);

	ixgbe_napi_disable_all(adapter);

5052 5053
	adapter->flags2 &= ~(IXGBE_FLAG2_FDIR_REQUIRES_REINIT |
			     IXGBE_FLAG2_RESET_REQUESTED);
5054 5055 5056 5057
	adapter->flags &= ~IXGBE_FLAG_NEED_LINK_UPDATE;

	del_timer_sync(&adapter->service_timer);

5058
	if (adapter->num_vfs) {
5059 5060
		/* Clear EITR Select mapping */
		IXGBE_WRITE_REG(&adapter->hw, IXGBE_EITRSEL, 0);
5061 5062 5063

		/* Mark all the VFs as inactive */
		for (i = 0 ; i < adapter->num_vfs; i++)
5064
			adapter->vfinfo[i].clear_to_send = false;
5065 5066 5067 5068 5069 5070

		/* ping all the active vfs to let them know we are going down */
		ixgbe_ping_all_vfs(adapter);

		/* Disable all VFTE/VFRE TX/RX */
		ixgbe_disable_tx_rx(adapter);
5071 5072
	}

5073 5074
	/* disable transmits in the hardware now that interrupts are off */
	for (i = 0; i < adapter->num_tx_queues; i++) {
5075
		u8 reg_idx = adapter->tx_ring[i]->reg_idx;
5076
		IXGBE_WRITE_REG(hw, IXGBE_TXDCTL(reg_idx), IXGBE_TXDCTL_SWFLSH);
5077
	}
5078

5079
	/* Disable the Tx DMA engine on 82599 and later MAC */
5080 5081
	switch (hw->mac.type) {
	case ixgbe_mac_82599EB:
D
Don Skidmore 已提交
5082
	case ixgbe_mac_X540:
5083 5084
	case ixgbe_mac_X550:
	case ixgbe_mac_X550EM_x:
5085
		IXGBE_WRITE_REG(hw, IXGBE_DMATXCTL,
5086 5087
				(IXGBE_READ_REG(hw, IXGBE_DMATXCTL) &
				 ~IXGBE_DMATXCTL_TE));
5088 5089 5090 5091
		break;
	default:
		break;
	}
5092

5093 5094
	if (!pci_channel_offline(adapter->pdev))
		ixgbe_reset(adapter);
5095

5096 5097
	/* power down the optics for 82599 SFP+ fiber */
	if (hw->mac.ops.disable_tx_laser)
5098 5099
		hw->mac.ops.disable_tx_laser(hw);

5100 5101 5102
	ixgbe_clean_all_tx_rings(adapter);
	ixgbe_clean_all_rx_rings(adapter);

5103
#ifdef CONFIG_IXGBE_DCA
5104
	/* since we reset the hardware DCA settings were cleared */
5105
	ixgbe_setup_dca(adapter);
5106
#endif
5107 5108 5109 5110 5111 5112 5113 5114 5115 5116 5117
}

/**
 * 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 */
5118
	ixgbe_tx_timeout_reset(adapter);
5119 5120 5121 5122 5123 5124 5125 5126 5127 5128
}

/**
 * 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).
 **/
5129
static int ixgbe_sw_init(struct ixgbe_adapter *adapter)
5130 5131 5132
{
	struct ixgbe_hw *hw = &adapter->hw;
	struct pci_dev *pdev = adapter->pdev;
5133
	unsigned int rss, fdir;
5134
	u32 fwsm;
J
Jeff Kirsher 已提交
5135
#ifdef CONFIG_IXGBE_DCB
5136 5137 5138
	int j;
	struct tc_configuration *tc;
#endif
5139

5140 5141 5142 5143 5144 5145 5146 5147
	/* 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;

5148
	/* Set common capability flags and settings */
5149
	rss = min_t(int, ixgbe_max_rss_indices(adapter), num_online_cpus());
5150
	adapter->ring_feature[RING_F_RSS].limit = rss;
5151 5152 5153 5154
	adapter->flags2 |= IXGBE_FLAG2_RSC_CAPABLE;
	adapter->flags2 |= IXGBE_FLAG2_RSC_ENABLED;
	adapter->max_q_vectors = MAX_Q_VECTORS_82599;
	adapter->atr_sample_rate = 20;
5155 5156
	fdir = min_t(int, IXGBE_MAX_FDIR_INDICES, num_online_cpus());
	adapter->ring_feature[RING_F_FDIR].limit = fdir;
5157 5158 5159 5160 5161 5162 5163 5164 5165 5166 5167 5168 5169
	adapter->fdir_pballoc = IXGBE_FDIR_PBALLOC_64K;
#ifdef CONFIG_IXGBE_DCA
	adapter->flags |= IXGBE_FLAG_DCA_CAPABLE;
#endif
#ifdef IXGBE_FCOE
	adapter->flags |= IXGBE_FLAG_FCOE_CAPABLE;
	adapter->flags &= ~IXGBE_FLAG_FCOE_ENABLED;
#ifdef CONFIG_IXGBE_DCB
	/* Default traffic class to use for FCoE */
	adapter->fcoe.up = IXGBE_FCOE_DEFTC;
#endif /* CONFIG_IXGBE_DCB */
#endif /* IXGBE_FCOE */

5170 5171 5172 5173
	adapter->mac_table = kzalloc(sizeof(struct ixgbe_mac_addr) *
				     hw->mac.num_rar_entries,
				     GFP_ATOMIC);

5174
	/* Set MAC specific capability flags and exceptions */
5175 5176
	switch (hw->mac.type) {
	case ixgbe_mac_82598EB:
5177 5178 5179
		adapter->flags2 &= ~IXGBE_FLAG2_RSC_CAPABLE;
		adapter->flags2 &= ~IXGBE_FLAG2_RSC_ENABLED;

5180 5181
		if (hw->device_id == IXGBE_DEV_ID_82598AT)
			adapter->flags |= IXGBE_FLAG_FAN_FAIL_CAPABLE;
5182

5183
		adapter->max_q_vectors = MAX_Q_VECTORS_82598;
5184 5185 5186 5187 5188 5189 5190 5191 5192 5193 5194 5195 5196 5197
		adapter->ring_feature[RING_F_FDIR].limit = 0;
		adapter->atr_sample_rate = 0;
		adapter->fdir_pballoc = 0;
#ifdef IXGBE_FCOE
		adapter->flags &= ~IXGBE_FLAG_FCOE_CAPABLE;
		adapter->flags &= ~IXGBE_FLAG_FCOE_ENABLED;
#ifdef CONFIG_IXGBE_DCB
		adapter->fcoe.up = 0;
#endif /* IXGBE_DCB */
#endif /* IXGBE_FCOE */
		break;
	case ixgbe_mac_82599EB:
		if (hw->device_id == IXGBE_DEV_ID_82599_T3_LOM)
			adapter->flags2 |= IXGBE_FLAG2_TEMP_SENSOR_CAPABLE;
5198
		break;
D
Don Skidmore 已提交
5199
	case ixgbe_mac_X540:
5200 5201 5202
		fwsm = IXGBE_READ_REG(hw, IXGBE_FWSM);
		if (fwsm & IXGBE_FWSM_TS_ENABLED)
			adapter->flags2 |= IXGBE_FLAG2_TEMP_SENSOR_CAPABLE;
5203
		break;
5204 5205 5206 5207 5208 5209
	case ixgbe_mac_X550EM_x:
	case ixgbe_mac_X550:
#ifdef CONFIG_IXGBE_DCA
		adapter->flags &= ~IXGBE_FLAG_DCA_CAPABLE;
#endif
		break;
5210 5211
	default:
		break;
A
Alexander Duyck 已提交
5212
	}
5213

5214 5215 5216 5217 5218
#ifdef IXGBE_FCOE
	/* FCoE support exists, always init the FCoE lock */
	spin_lock_init(&adapter->fcoe.lock);

#endif
5219 5220 5221
	/* n-tuple support exists, always init our spinlock */
	spin_lock_init(&adapter->fdir_perfect_lock);

J
Jeff Kirsher 已提交
5222
#ifdef CONFIG_IXGBE_DCB
5223 5224
	switch (hw->mac.type) {
	case ixgbe_mac_X540:
5225 5226
	case ixgbe_mac_X550:
	case ixgbe_mac_X550EM_x:
5227 5228 5229 5230 5231 5232 5233 5234 5235
		adapter->dcb_cfg.num_tcs.pg_tcs = X540_TRAFFIC_CLASS;
		adapter->dcb_cfg.num_tcs.pfc_tcs = X540_TRAFFIC_CLASS;
		break;
	default:
		adapter->dcb_cfg.num_tcs.pg_tcs = MAX_TRAFFIC_CLASS;
		adapter->dcb_cfg.num_tcs.pfc_tcs = MAX_TRAFFIC_CLASS;
		break;
	}

5236 5237 5238 5239 5240 5241 5242 5243 5244
	/* 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;
	}
5245 5246 5247 5248 5249 5250

	/* Initialize default user to priority mapping, UPx->TC0 */
	tc = &adapter->dcb_cfg.tc_config[0];
	tc->path[DCB_TX_CONFIG].up_to_tc_bitmap = 0xFF;
	tc->path[DCB_RX_CONFIG].up_to_tc_bitmap = 0xFF;

5251 5252
	adapter->dcb_cfg.bw_percentage[DCB_TX_CONFIG][0] = 100;
	adapter->dcb_cfg.bw_percentage[DCB_RX_CONFIG][0] = 100;
5253
	adapter->dcb_cfg.pfc_mode_enable = false;
5254
	adapter->dcb_set_bitmap = 0x00;
5255
	adapter->dcbx_cap = DCB_CAP_DCBX_HOST | DCB_CAP_DCBX_VER_CEE;
5256 5257
	memcpy(&adapter->temp_dcb_cfg, &adapter->dcb_cfg,
	       sizeof(adapter->temp_dcb_cfg));
5258 5259

#endif
5260 5261

	/* default flow control settings */
5262
	hw->fc.requested_mode = ixgbe_fc_full;
D
Don Skidmore 已提交
5263
	hw->fc.current_mode = ixgbe_fc_full;	/* init for ethtool output */
5264
	ixgbe_pbthresh_setup(adapter);
5265 5266
	hw->fc.pause_time = IXGBE_DEFAULT_FCPAUSE;
	hw->fc.send_xon = true;
5267
	hw->fc.disable_fc_autoneg = ixgbe_device_supports_autoneg_fc(hw);
5268

5269
#ifdef CONFIG_PCI_IOV
5270 5271 5272
	if (max_vfs > 0)
		e_dev_warn("Enabling SR-IOV VFs using the max_vfs module parameter is deprecated - please use the pci sysfs interface instead.\n");

5273
	/* assign number of SR-IOV VFs */
5274
	if (hw->mac.type != ixgbe_mac_82598EB) {
5275
		if (max_vfs > IXGBE_MAX_VFS_DRV_LIMIT) {
5276 5277 5278 5279 5280 5281 5282
			adapter->num_vfs = 0;
			e_dev_warn("max_vfs parameter out of range. Not assigning any SR-IOV VFs\n");
		} else {
			adapter->num_vfs = max_vfs;
		}
	}
#endif /* CONFIG_PCI_IOV */
5283

5284
	/* enable itr by default in dynamic mode */
5285 5286
	adapter->rx_itr_setting = 1;
	adapter->tx_itr_setting = 1;
5287 5288 5289 5290 5291

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

5292
	/* set default work limits */
5293
	adapter->tx_work_limit = IXGBE_DEFAULT_TX_WORK;
5294

5295
	/* initialize eeprom parameters */
5296
	if (ixgbe_init_eeprom_params_generic(hw)) {
5297
		e_dev_err("EEPROM initialization failed\n");
5298 5299 5300
		return -EIO;
	}

5301 5302
	/* PF holds first pool slot */
	set_bit(0, &adapter->fwd_bitmask);
5303 5304 5305 5306 5307 5308 5309
	set_bit(__IXGBE_DOWN, &adapter->state);

	return 0;
}

/**
 * ixgbe_setup_tx_resources - allocate Tx resources (Descriptors)
5310
 * @tx_ring:    tx descriptor ring (for a specific queue) to setup
5311 5312 5313
 *
 * Return 0 on success, negative on failure
 **/
5314
int ixgbe_setup_tx_resources(struct ixgbe_ring *tx_ring)
5315
{
5316
	struct device *dev = tx_ring->dev;
5317
	int orig_node = dev_to_node(dev);
5318
	int ring_node = -1;
5319 5320
	int size;

5321
	size = sizeof(struct ixgbe_tx_buffer) * tx_ring->count;
5322 5323

	if (tx_ring->q_vector)
5324
		ring_node = tx_ring->q_vector->numa_node;
5325

5326
	tx_ring->tx_buffer_info = vzalloc_node(size, ring_node);
5327
	if (!tx_ring->tx_buffer_info)
E
Eric Dumazet 已提交
5328
		tx_ring->tx_buffer_info = vzalloc(size);
5329 5330
	if (!tx_ring->tx_buffer_info)
		goto err;
5331

5332 5333
	u64_stats_init(&tx_ring->syncp);

5334
	/* round up to nearest 4K */
5335
	tx_ring->size = tx_ring->count * sizeof(union ixgbe_adv_tx_desc);
5336
	tx_ring->size = ALIGN(tx_ring->size, 4096);
5337

5338
	set_dev_node(dev, ring_node);
5339 5340 5341 5342 5343 5344 5345 5346
	tx_ring->desc = dma_alloc_coherent(dev,
					   tx_ring->size,
					   &tx_ring->dma,
					   GFP_KERNEL);
	set_dev_node(dev, orig_node);
	if (!tx_ring->desc)
		tx_ring->desc = dma_alloc_coherent(dev, tx_ring->size,
						   &tx_ring->dma, GFP_KERNEL);
5347 5348
	if (!tx_ring->desc)
		goto err;
5349

5350 5351
	tx_ring->next_to_use = 0;
	tx_ring->next_to_clean = 0;
5352
	return 0;
5353 5354 5355 5356

err:
	vfree(tx_ring->tx_buffer_info);
	tx_ring->tx_buffer_info = NULL;
5357
	dev_err(dev, "Unable to allocate memory for the Tx descriptor ring\n");
5358
	return -ENOMEM;
5359 5360
}

5361 5362 5363 5364 5365 5366 5367 5368 5369 5370 5371 5372 5373 5374 5375
/**
 * 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++) {
5376
		err = ixgbe_setup_tx_resources(adapter->tx_ring[i]);
5377 5378
		if (!err)
			continue;
5379

5380
		e_err(probe, "Allocation for Tx Queue %u failed\n", i);
5381
		goto err_setup_tx;
5382 5383
	}

5384 5385 5386 5387 5388
	return 0;
err_setup_tx:
	/* rewind the index freeing the rings as we go */
	while (i--)
		ixgbe_free_tx_resources(adapter->tx_ring[i]);
5389 5390 5391
	return err;
}

5392 5393
/**
 * ixgbe_setup_rx_resources - allocate Rx resources (Descriptors)
5394
 * @rx_ring:    rx descriptor ring (for a specific queue) to setup
5395 5396 5397
 *
 * Returns 0 on success, negative on failure
 **/
5398
int ixgbe_setup_rx_resources(struct ixgbe_ring *rx_ring)
5399
{
5400
	struct device *dev = rx_ring->dev;
5401
	int orig_node = dev_to_node(dev);
5402
	int ring_node = -1;
5403
	int size;
5404

5405
	size = sizeof(struct ixgbe_rx_buffer) * rx_ring->count;
5406 5407

	if (rx_ring->q_vector)
5408
		ring_node = rx_ring->q_vector->numa_node;
5409

5410
	rx_ring->rx_buffer_info = vzalloc_node(size, ring_node);
5411
	if (!rx_ring->rx_buffer_info)
E
Eric Dumazet 已提交
5412
		rx_ring->rx_buffer_info = vzalloc(size);
5413 5414
	if (!rx_ring->rx_buffer_info)
		goto err;
5415

5416 5417
	u64_stats_init(&rx_ring->syncp);

5418
	/* Round up to nearest 4K */
5419 5420
	rx_ring->size = rx_ring->count * sizeof(union ixgbe_adv_rx_desc);
	rx_ring->size = ALIGN(rx_ring->size, 4096);
5421

5422
	set_dev_node(dev, ring_node);
5423 5424 5425 5426 5427 5428 5429 5430
	rx_ring->desc = dma_alloc_coherent(dev,
					   rx_ring->size,
					   &rx_ring->dma,
					   GFP_KERNEL);
	set_dev_node(dev, orig_node);
	if (!rx_ring->desc)
		rx_ring->desc = dma_alloc_coherent(dev, rx_ring->size,
						   &rx_ring->dma, GFP_KERNEL);
5431 5432
	if (!rx_ring->desc)
		goto err;
5433

5434 5435
	rx_ring->next_to_clean = 0;
	rx_ring->next_to_use = 0;
5436 5437

	return 0;
5438 5439 5440 5441
err:
	vfree(rx_ring->rx_buffer_info);
	rx_ring->rx_buffer_info = NULL;
	dev_err(dev, "Unable to allocate memory for the Rx descriptor ring\n");
5442
	return -ENOMEM;
5443 5444
}

5445 5446 5447 5448 5449 5450 5451 5452 5453 5454 5455 5456 5457 5458 5459
/**
 * 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++) {
5460
		err = ixgbe_setup_rx_resources(adapter->rx_ring[i]);
5461 5462
		if (!err)
			continue;
5463

5464
		e_err(probe, "Allocation for Rx Queue %u failed\n", i);
5465
		goto err_setup_rx;
5466 5467
	}

5468 5469 5470 5471 5472
#ifdef IXGBE_FCOE
	err = ixgbe_setup_fcoe_ddp_resources(adapter);
	if (!err)
#endif
		return 0;
5473 5474 5475 5476
err_setup_rx:
	/* rewind the index freeing the rings as we go */
	while (i--)
		ixgbe_free_rx_resources(adapter->rx_ring[i]);
5477 5478 5479
	return err;
}

5480 5481 5482 5483 5484 5485
/**
 * ixgbe_free_tx_resources - Free Tx Resources per Queue
 * @tx_ring: Tx descriptor ring for a specific queue
 *
 * Free all transmit software resources
 **/
5486
void ixgbe_free_tx_resources(struct ixgbe_ring *tx_ring)
5487
{
5488
	ixgbe_clean_tx_ring(tx_ring);
5489 5490 5491 5492

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

5493 5494 5495 5496 5497 5498
	/* if not set, then don't free */
	if (!tx_ring->desc)
		return;

	dma_free_coherent(tx_ring->dev, tx_ring->size,
			  tx_ring->desc, tx_ring->dma);
5499 5500 5501 5502 5503 5504 5505 5506 5507 5508 5509 5510 5511 5512 5513

	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++)
5514
		if (adapter->tx_ring[i]->desc)
5515
			ixgbe_free_tx_resources(adapter->tx_ring[i]);
5516 5517 5518
}

/**
5519
 * ixgbe_free_rx_resources - Free Rx Resources
5520 5521 5522 5523
 * @rx_ring: ring to clean the resources from
 *
 * Free all receive software resources
 **/
5524
void ixgbe_free_rx_resources(struct ixgbe_ring *rx_ring)
5525
{
5526
	ixgbe_clean_rx_ring(rx_ring);
5527 5528 5529 5530

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

5531 5532 5533 5534 5535 5536
	/* if not set, then don't free */
	if (!rx_ring->desc)
		return;

	dma_free_coherent(rx_ring->dev, rx_ring->size,
			  rx_ring->desc, rx_ring->dma);
5537 5538 5539 5540 5541 5542 5543 5544 5545 5546 5547 5548 5549 5550

	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;

5551 5552 5553 5554
#ifdef IXGBE_FCOE
	ixgbe_free_fcoe_ddp_resources(adapter);

#endif
5555
	for (i = 0; i < adapter->num_rx_queues; i++)
5556
		if (adapter->rx_ring[i]->desc)
5557
			ixgbe_free_rx_resources(adapter->rx_ring[i]);
5558 5559 5560 5561 5562 5563 5564 5565 5566 5567 5568 5569 5570 5571
}

/**
 * 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;

5572
	/* MTU < 68 is an error and causes problems on some kernels */
5573 5574 5575 5576
	if ((new_mtu < 68) || (max_frame > IXGBE_MAX_JUMBO_FRAME_SIZE))
		return -EINVAL;

	/*
5577 5578 5579
	 * For 82599EB we cannot allow legacy VFs to enable their receive
	 * paths when MTU greater than 1500 is configured.  So display a
	 * warning that legacy VFs will be disabled.
5580 5581 5582
	 */
	if ((adapter->flags & IXGBE_FLAG_SRIOV_ENABLED) &&
	    (adapter->hw.mac.type == ixgbe_mac_82599EB) &&
A
Alexander Duyck 已提交
5583
	    (max_frame > (ETH_FRAME_LEN + ETH_FCS_LEN)))
5584
		e_warn(probe, "Setting MTU > 1500 will disable legacy VFs\n");
5585

5586
	e_info(probe, "changing MTU from %d to %d\n", netdev->mtu, new_mtu);
5587

5588
	/* must set new MTU before calling down or up */
5589 5590
	netdev->mtu = new_mtu;

5591 5592
	if (netif_running(netdev))
		ixgbe_reinit_locked(adapter);
5593 5594 5595 5596 5597 5598 5599 5600 5601 5602 5603 5604 5605 5606 5607 5608 5609 5610 5611

	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);
5612
	int err, queues;
5613 5614 5615 5616

	/* disallow open during test */
	if (test_bit(__IXGBE_TESTING, &adapter->state))
		return -EBUSY;
5617

5618 5619
	netif_carrier_off(netdev);

5620 5621 5622 5623 5624 5625 5626 5627 5628 5629 5630 5631
	/* 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);

5632
	err = ixgbe_request_irq(adapter);
5633 5634 5635
	if (err)
		goto err_req_irq;

5636
	/* Notify the stack of the actual queue counts. */
5637 5638 5639 5640 5641 5642
	if (adapter->num_rx_pools > 1)
		queues = adapter->num_rx_queues_per_pool;
	else
		queues = adapter->num_tx_queues;

	err = netif_set_real_num_tx_queues(netdev, queues);
5643 5644 5645
	if (err)
		goto err_set_queues;

5646 5647 5648 5649 5650 5651
	if (adapter->num_rx_pools > 1 &&
	    adapter->num_rx_queues > IXGBE_MAX_L2A_QUEUES)
		queues = IXGBE_MAX_L2A_QUEUES;
	else
		queues = adapter->num_rx_queues;
	err = netif_set_real_num_rx_queues(netdev, queues);
5652 5653 5654
	if (err)
		goto err_set_queues;

5655 5656
	ixgbe_ptp_init(adapter);

5657
	ixgbe_up_complete(adapter);
5658

5659 5660 5661 5662
#if IS_ENABLED(CONFIG_IXGBE_VXLAN)
	vxlan_get_rx_port(netdev);

#endif
5663 5664
	return 0;

5665 5666
err_set_queues:
	ixgbe_free_irq(adapter);
5667
err_req_irq:
5668
	ixgbe_free_all_rx_resources(adapter);
5669
err_setup_rx:
5670
	ixgbe_free_all_tx_resources(adapter);
5671
err_setup_tx:
5672 5673 5674 5675 5676
	ixgbe_reset(adapter);

	return err;
}

5677 5678 5679 5680 5681 5682 5683 5684 5685 5686 5687
static void ixgbe_close_suspend(struct ixgbe_adapter *adapter)
{
	ixgbe_ptp_suspend(adapter);

	ixgbe_down(adapter);
	ixgbe_free_irq(adapter);

	ixgbe_free_all_tx_resources(adapter);
	ixgbe_free_all_rx_resources(adapter);
}

5688 5689 5690 5691 5692 5693 5694 5695 5696 5697 5698 5699 5700 5701 5702
/**
 * 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);

5703 5704
	ixgbe_ptp_stop(adapter);

5705
	ixgbe_close_suspend(adapter);
5706

5707 5708
	ixgbe_fdir_filter_exit(adapter);

5709
	ixgbe_release_hw_control(adapter);
5710 5711 5712 5713

	return 0;
}

5714 5715 5716
#ifdef CONFIG_PM
static int ixgbe_resume(struct pci_dev *pdev)
{
5717 5718
	struct ixgbe_adapter *adapter = pci_get_drvdata(pdev);
	struct net_device *netdev = adapter->netdev;
5719 5720
	u32 err;

5721
	adapter->hw.hw_addr = adapter->io_addr;
5722 5723
	pci_set_power_state(pdev, PCI_D0);
	pci_restore_state(pdev);
5724 5725 5726 5727 5728
	/*
	 * pci_restore_state clears dev->state_saved so call
	 * pci_save_state to restore it.
	 */
	pci_save_state(pdev);
5729 5730

	err = pci_enable_device_mem(pdev);
5731
	if (err) {
5732
		e_dev_err("Cannot enable PCI device from suspend\n");
5733 5734
		return err;
	}
5735
	smp_mb__before_atomic();
5736
	clear_bit(__IXGBE_DISABLED, &adapter->state);
5737 5738
	pci_set_master(pdev);

5739
	pci_wake_from_d3(pdev, false);
5740 5741 5742

	ixgbe_reset(adapter);

5743 5744
	IXGBE_WRITE_REG(&adapter->hw, IXGBE_WUS, ~0);

5745 5746 5747
	rtnl_lock();
	err = ixgbe_init_interrupt_scheme(adapter);
	if (!err && netif_running(netdev))
5748
		err = ixgbe_open(netdev);
5749 5750 5751 5752 5753

	rtnl_unlock();

	if (err)
		return err;
5754 5755 5756 5757 5758 5759

	netif_device_attach(netdev);

	return 0;
}
#endif /* CONFIG_PM */
5760 5761

static int __ixgbe_shutdown(struct pci_dev *pdev, bool *enable_wake)
5762
{
5763 5764
	struct ixgbe_adapter *adapter = pci_get_drvdata(pdev);
	struct net_device *netdev = adapter->netdev;
5765 5766 5767
	struct ixgbe_hw *hw = &adapter->hw;
	u32 ctrl, fctrl;
	u32 wufc = adapter->wol;
5768 5769 5770 5771 5772 5773
#ifdef CONFIG_PM
	int retval = 0;
#endif

	netif_device_detach(netdev);

5774
	rtnl_lock();
5775 5776
	if (netif_running(netdev))
		ixgbe_close_suspend(adapter);
5777
	rtnl_unlock();
5778

5779 5780
	ixgbe_clear_interrupt_scheme(adapter);

5781 5782 5783 5784
#ifdef CONFIG_PM
	retval = pci_save_state(pdev);
	if (retval)
		return retval;
5785

5786
#endif
5787 5788 5789
	if (hw->mac.ops.stop_link_on_d3)
		hw->mac.ops.stop_link_on_d3(hw);

5790 5791
	if (wufc) {
		ixgbe_set_rx_mode(netdev);
5792

5793 5794
		/* enable the optics for 82599 SFP+ fiber as we can WoL */
		if (hw->mac.ops.enable_tx_laser)
D
Don Skidmore 已提交
5795 5796
			hw->mac.ops.enable_tx_laser(hw);

5797 5798 5799 5800 5801 5802 5803 5804 5805 5806 5807 5808 5809 5810 5811 5812 5813
		/* 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);
	}

5814 5815
	switch (hw->mac.type) {
	case ixgbe_mac_82598EB:
5816
		pci_wake_from_d3(pdev, false);
5817 5818
		break;
	case ixgbe_mac_82599EB:
D
Don Skidmore 已提交
5819
	case ixgbe_mac_X540:
5820 5821
	case ixgbe_mac_X550:
	case ixgbe_mac_X550EM_x:
5822 5823 5824 5825 5826
		pci_wake_from_d3(pdev, !!wufc);
		break;
	default:
		break;
	}
5827

5828 5829
	*enable_wake = !!wufc;

5830 5831
	ixgbe_release_hw_control(adapter);

5832 5833
	if (!test_and_set_bit(__IXGBE_DISABLED, &adapter->state))
		pci_disable_device(pdev);
5834

5835 5836 5837 5838 5839 5840 5841 5842 5843 5844 5845 5846 5847 5848 5849 5850 5851 5852 5853
	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);
	}
5854 5855 5856

	return 0;
}
5857
#endif /* CONFIG_PM */
5858 5859 5860

static void ixgbe_shutdown(struct pci_dev *pdev)
{
5861 5862 5863 5864 5865 5866 5867 5868
	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);
	}
5869 5870
}

5871 5872 5873 5874 5875 5876
/**
 * ixgbe_update_stats - Update the board statistics counters.
 * @adapter: board private structure
 **/
void ixgbe_update_stats(struct ixgbe_adapter *adapter)
{
5877
	struct net_device *netdev = adapter->netdev;
5878
	struct ixgbe_hw *hw = &adapter->hw;
5879
	struct ixgbe_hw_stats *hwstats = &adapter->stats;
5880 5881
	u64 total_mpc = 0;
	u32 i, missed_rx = 0, mpc, bprc, lxon, lxoff, xon_off_tot;
5882 5883
	u64 non_eop_descs = 0, restart_queue = 0, tx_busy = 0;
	u64 alloc_rx_page_failed = 0, alloc_rx_buff_failed = 0;
5884
	u64 bytes = 0, packets = 0, hw_csum_rx_error = 0;
5885

5886 5887 5888 5889
	if (test_bit(__IXGBE_DOWN, &adapter->state) ||
	    test_bit(__IXGBE_RESETTING, &adapter->state))
		return;

5890
	if (adapter->flags2 & IXGBE_FLAG2_RSC_ENABLED) {
A
Alexander Duyck 已提交
5891
		u64 rsc_count = 0;
5892 5893
		u64 rsc_flush = 0;
		for (i = 0; i < adapter->num_rx_queues; i++) {
5894 5895
			rsc_count += adapter->rx_ring[i]->rx_stats.rsc_count;
			rsc_flush += adapter->rx_ring[i]->rx_stats.rsc_flush;
5896 5897 5898
		}
		adapter->rsc_total_count = rsc_count;
		adapter->rsc_total_flush = rsc_flush;
5899 5900
	}

5901 5902 5903 5904 5905
	for (i = 0; i < adapter->num_rx_queues; i++) {
		struct ixgbe_ring *rx_ring = adapter->rx_ring[i];
		non_eop_descs += rx_ring->rx_stats.non_eop_descs;
		alloc_rx_page_failed += rx_ring->rx_stats.alloc_rx_page_failed;
		alloc_rx_buff_failed += rx_ring->rx_stats.alloc_rx_buff_failed;
5906
		hw_csum_rx_error += rx_ring->rx_stats.csum_err;
5907 5908 5909 5910 5911 5912
		bytes += rx_ring->stats.bytes;
		packets += rx_ring->stats.packets;
	}
	adapter->non_eop_descs = non_eop_descs;
	adapter->alloc_rx_page_failed = alloc_rx_page_failed;
	adapter->alloc_rx_buff_failed = alloc_rx_buff_failed;
5913
	adapter->hw_csum_rx_error = hw_csum_rx_error;
5914 5915 5916 5917 5918
	netdev->stats.rx_bytes = bytes;
	netdev->stats.rx_packets = packets;

	bytes = 0;
	packets = 0;
J
Jesse Brandeburg 已提交
5919
	/* gather some stats to the adapter struct that are per queue */
5920 5921 5922 5923 5924 5925 5926
	for (i = 0; i < adapter->num_tx_queues; i++) {
		struct ixgbe_ring *tx_ring = adapter->tx_ring[i];
		restart_queue += tx_ring->tx_stats.restart_queue;
		tx_busy += tx_ring->tx_stats.tx_busy;
		bytes += tx_ring->stats.bytes;
		packets += tx_ring->stats.packets;
	}
5927
	adapter->restart_queue = restart_queue;
5928 5929 5930
	adapter->tx_busy = tx_busy;
	netdev->stats.tx_bytes = bytes;
	netdev->stats.tx_packets = packets;
J
Jesse Brandeburg 已提交
5931

5932
	hwstats->crcerrs += IXGBE_READ_REG(hw, IXGBE_CRCERRS);
5933 5934

	/* 8 register reads */
5935 5936 5937 5938
	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;
5939 5940
		hwstats->mpc[i] += mpc;
		total_mpc += hwstats->mpc[i];
5941 5942
		hwstats->pxontxc[i] += IXGBE_READ_REG(hw, IXGBE_PXONTXC(i));
		hwstats->pxofftxc[i] += IXGBE_READ_REG(hw, IXGBE_PXOFFTXC(i));
5943 5944
		switch (hw->mac.type) {
		case ixgbe_mac_82598EB:
5945 5946 5947
			hwstats->rnbc[i] += IXGBE_READ_REG(hw, IXGBE_RNBC(i));
			hwstats->qbtc[i] += IXGBE_READ_REG(hw, IXGBE_QBTC(i));
			hwstats->qbrc[i] += IXGBE_READ_REG(hw, IXGBE_QBRC(i));
5948 5949
			hwstats->pxonrxc[i] +=
				IXGBE_READ_REG(hw, IXGBE_PXONRXC(i));
5950 5951
			break;
		case ixgbe_mac_82599EB:
D
Don Skidmore 已提交
5952
		case ixgbe_mac_X540:
5953 5954
		case ixgbe_mac_X550:
		case ixgbe_mac_X550EM_x:
5955 5956 5957 5958 5959
			hwstats->pxonrxc[i] +=
				IXGBE_READ_REG(hw, IXGBE_PXONRXCNT(i));
			break;
		default:
			break;
5960
		}
5961
	}
5962 5963 5964 5965 5966 5967

	/*16 register reads */
	for (i = 0; i < 16; i++) {
		hwstats->qptc[i] += IXGBE_READ_REG(hw, IXGBE_QPTC(i));
		hwstats->qprc[i] += IXGBE_READ_REG(hw, IXGBE_QPRC(i));
		if ((hw->mac.type == ixgbe_mac_82599EB) ||
5968 5969 5970
		    (hw->mac.type == ixgbe_mac_X540) ||
		    (hw->mac.type == ixgbe_mac_X550) ||
		    (hw->mac.type == ixgbe_mac_X550EM_x)) {
5971 5972 5973 5974 5975 5976 5977
			hwstats->qbtc[i] += IXGBE_READ_REG(hw, IXGBE_QBTC_L(i));
			IXGBE_READ_REG(hw, IXGBE_QBTC_H(i)); /* to clear */
			hwstats->qbrc[i] += IXGBE_READ_REG(hw, IXGBE_QBRC_L(i));
			IXGBE_READ_REG(hw, IXGBE_QBRC_H(i)); /* to clear */
		}
	}

5978
	hwstats->gprc += IXGBE_READ_REG(hw, IXGBE_GPRC);
5979
	/* work around hardware counting issue */
5980
	hwstats->gprc -= missed_rx;
5981

5982 5983
	ixgbe_update_xoff_received(adapter);

5984
	/* 82598 hardware only has a 32 bit counter in the high register */
5985 5986 5987 5988 5989 5990 5991
	switch (hw->mac.type) {
	case ixgbe_mac_82598EB:
		hwstats->lxonrxc += IXGBE_READ_REG(hw, IXGBE_LXONRXC);
		hwstats->gorc += IXGBE_READ_REG(hw, IXGBE_GORCH);
		hwstats->gotc += IXGBE_READ_REG(hw, IXGBE_GOTCH);
		hwstats->tor += IXGBE_READ_REG(hw, IXGBE_TORH);
		break;
D
Don Skidmore 已提交
5992
	case ixgbe_mac_X540:
5993 5994 5995
	case ixgbe_mac_X550:
	case ixgbe_mac_X550EM_x:
		/* OS2BMC stats are X540 and later */
5996 5997 5998 5999 6000
		hwstats->o2bgptc += IXGBE_READ_REG(hw, IXGBE_O2BGPTC);
		hwstats->o2bspc += IXGBE_READ_REG(hw, IXGBE_O2BSPC);
		hwstats->b2ospc += IXGBE_READ_REG(hw, IXGBE_B2OSPC);
		hwstats->b2ogprc += IXGBE_READ_REG(hw, IXGBE_B2OGPRC);
	case ixgbe_mac_82599EB:
6001 6002 6003
		for (i = 0; i < 16; i++)
			adapter->hw_rx_no_dma_resources +=
					     IXGBE_READ_REG(hw, IXGBE_QPRDC(i));
6004
		hwstats->gorc += IXGBE_READ_REG(hw, IXGBE_GORCL);
6005
		IXGBE_READ_REG(hw, IXGBE_GORCH); /* to clear */
6006
		hwstats->gotc += IXGBE_READ_REG(hw, IXGBE_GOTCL);
6007
		IXGBE_READ_REG(hw, IXGBE_GOTCH); /* to clear */
6008
		hwstats->tor += IXGBE_READ_REG(hw, IXGBE_TORL);
6009
		IXGBE_READ_REG(hw, IXGBE_TORH); /* to clear */
6010 6011 6012
		hwstats->lxonrxc += IXGBE_READ_REG(hw, IXGBE_LXONRXCNT);
		hwstats->fdirmatch += IXGBE_READ_REG(hw, IXGBE_FDIRMATCH);
		hwstats->fdirmiss += IXGBE_READ_REG(hw, IXGBE_FDIRMISS);
6013
#ifdef IXGBE_FCOE
6014 6015 6016 6017 6018 6019
		hwstats->fccrc += IXGBE_READ_REG(hw, IXGBE_FCCRC);
		hwstats->fcoerpdc += IXGBE_READ_REG(hw, IXGBE_FCOERPDC);
		hwstats->fcoeprc += IXGBE_READ_REG(hw, IXGBE_FCOEPRC);
		hwstats->fcoeptc += IXGBE_READ_REG(hw, IXGBE_FCOEPTC);
		hwstats->fcoedwrc += IXGBE_READ_REG(hw, IXGBE_FCOEDWRC);
		hwstats->fcoedwtc += IXGBE_READ_REG(hw, IXGBE_FCOEDWTC);
6020
		/* Add up per cpu counters for total ddp aloc fail */
6021 6022 6023 6024 6025
		if (adapter->fcoe.ddp_pool) {
			struct ixgbe_fcoe *fcoe = &adapter->fcoe;
			struct ixgbe_fcoe_ddp_pool *ddp_pool;
			unsigned int cpu;
			u64 noddp = 0, noddp_ext_buff = 0;
6026
			for_each_possible_cpu(cpu) {
6027 6028 6029
				ddp_pool = per_cpu_ptr(fcoe->ddp_pool, cpu);
				noddp += ddp_pool->noddp;
				noddp_ext_buff += ddp_pool->noddp_ext_buff;
6030
			}
6031 6032
			hwstats->fcoe_noddp = noddp;
			hwstats->fcoe_noddp_ext_buff = noddp_ext_buff;
6033
		}
6034
#endif /* IXGBE_FCOE */
6035 6036 6037
		break;
	default:
		break;
6038
	}
6039
	bprc = IXGBE_READ_REG(hw, IXGBE_BPRC);
6040 6041
	hwstats->bprc += bprc;
	hwstats->mprc += IXGBE_READ_REG(hw, IXGBE_MPRC);
6042
	if (hw->mac.type == ixgbe_mac_82598EB)
6043 6044 6045 6046 6047 6048 6049 6050 6051
		hwstats->mprc -= bprc;
	hwstats->roc += IXGBE_READ_REG(hw, IXGBE_ROC);
	hwstats->prc64 += IXGBE_READ_REG(hw, IXGBE_PRC64);
	hwstats->prc127 += IXGBE_READ_REG(hw, IXGBE_PRC127);
	hwstats->prc255 += IXGBE_READ_REG(hw, IXGBE_PRC255);
	hwstats->prc511 += IXGBE_READ_REG(hw, IXGBE_PRC511);
	hwstats->prc1023 += IXGBE_READ_REG(hw, IXGBE_PRC1023);
	hwstats->prc1522 += IXGBE_READ_REG(hw, IXGBE_PRC1522);
	hwstats->rlec += IXGBE_READ_REG(hw, IXGBE_RLEC);
6052
	lxon = IXGBE_READ_REG(hw, IXGBE_LXONTXC);
6053
	hwstats->lxontxc += lxon;
6054
	lxoff = IXGBE_READ_REG(hw, IXGBE_LXOFFTXC);
6055 6056 6057
	hwstats->lxofftxc += lxoff;
	hwstats->gptc += IXGBE_READ_REG(hw, IXGBE_GPTC);
	hwstats->mptc += IXGBE_READ_REG(hw, IXGBE_MPTC);
6058 6059 6060 6061
	/*
	 * 82598 errata - tx of flow control packets is included in tx counters
	 */
	xon_off_tot = lxon + lxoff;
6062 6063 6064 6065 6066 6067 6068 6069 6070 6071 6072 6073 6074 6075 6076
	hwstats->gptc -= xon_off_tot;
	hwstats->mptc -= xon_off_tot;
	hwstats->gotc -= (xon_off_tot * (ETH_ZLEN + ETH_FCS_LEN));
	hwstats->ruc += IXGBE_READ_REG(hw, IXGBE_RUC);
	hwstats->rfc += IXGBE_READ_REG(hw, IXGBE_RFC);
	hwstats->rjc += IXGBE_READ_REG(hw, IXGBE_RJC);
	hwstats->tpr += IXGBE_READ_REG(hw, IXGBE_TPR);
	hwstats->ptc64 += IXGBE_READ_REG(hw, IXGBE_PTC64);
	hwstats->ptc64 -= xon_off_tot;
	hwstats->ptc127 += IXGBE_READ_REG(hw, IXGBE_PTC127);
	hwstats->ptc255 += IXGBE_READ_REG(hw, IXGBE_PTC255);
	hwstats->ptc511 += IXGBE_READ_REG(hw, IXGBE_PTC511);
	hwstats->ptc1023 += IXGBE_READ_REG(hw, IXGBE_PTC1023);
	hwstats->ptc1522 += IXGBE_READ_REG(hw, IXGBE_PTC1522);
	hwstats->bptc += IXGBE_READ_REG(hw, IXGBE_BPTC);
6077 6078

	/* Fill out the OS statistics structure */
6079
	netdev->stats.multicast = hwstats->mprc;
6080 6081

	/* Rx Errors */
6082
	netdev->stats.rx_errors = hwstats->crcerrs + hwstats->rlec;
6083
	netdev->stats.rx_dropped = 0;
6084 6085
	netdev->stats.rx_length_errors = hwstats->rlec;
	netdev->stats.rx_crc_errors = hwstats->crcerrs;
6086
	netdev->stats.rx_missed_errors = total_mpc;
6087 6088 6089
}

/**
6090
 * ixgbe_fdir_reinit_subtask - worker thread to reinit FDIR filter table
6091
 * @adapter: pointer to the device adapter structure
6092
 **/
6093
static void ixgbe_fdir_reinit_subtask(struct ixgbe_adapter *adapter)
6094
{
6095
	struct ixgbe_hw *hw = &adapter->hw;
6096
	int i;
6097

6098 6099 6100 6101
	if (!(adapter->flags2 & IXGBE_FLAG2_FDIR_REQUIRES_REINIT))
		return;

	adapter->flags2 &= ~IXGBE_FLAG2_FDIR_REQUIRES_REINIT;
6102

6103
	/* if interface is down do nothing */
6104
	if (test_bit(__IXGBE_DOWN, &adapter->state))
6105 6106 6107 6108 6109 6110 6111 6112
		return;

	/* do nothing if we are not using signature filters */
	if (!(adapter->flags & IXGBE_FLAG_FDIR_HASH_CAPABLE))
		return;

	adapter->fdir_overflow++;

6113 6114 6115
	if (ixgbe_reinit_fdir_tables_82599(hw) == 0) {
		for (i = 0; i < adapter->num_tx_queues; i++)
			set_bit(__IXGBE_TX_FDIR_INIT_DONE,
6116
				&(adapter->tx_ring[i]->state));
6117 6118
		/* re-enable flow director interrupts */
		IXGBE_WRITE_REG(hw, IXGBE_EIMS, IXGBE_EIMS_FLOW_DIR);
6119 6120 6121 6122 6123 6124 6125 6126
	} else {
		e_err(probe, "failed to finish FDIR re-initialization, "
		      "ignored adding FDIR ATR filters\n");
	}
}

/**
 * ixgbe_check_hang_subtask - check for hung queues and dropped interrupts
6127
 * @adapter: pointer to the device adapter structure
6128 6129
 *
 * This function serves two purposes.  First it strobes the interrupt lines
S
Stephen Hemminger 已提交
6130
 * in order to make certain interrupts are occurring.  Secondly it sets the
6131
 * bits needed to check for TX hangs.  As a result we should immediately
S
Stephen Hemminger 已提交
6132
 * determine if a hang has occurred.
6133 6134
 */
static void ixgbe_check_hang_subtask(struct ixgbe_adapter *adapter)
6135
{
6136
	struct ixgbe_hw *hw = &adapter->hw;
6137 6138
	u64 eics = 0;
	int i;
6139

6140
	/* If we're down, removing or resetting, just bail */
6141
	if (test_bit(__IXGBE_DOWN, &adapter->state) ||
6142
	    test_bit(__IXGBE_REMOVING, &adapter->state) ||
6143 6144
	    test_bit(__IXGBE_RESETTING, &adapter->state))
		return;
6145

6146 6147 6148 6149 6150
	/* Force detection of hung controller */
	if (netif_carrier_ok(adapter->netdev)) {
		for (i = 0; i < adapter->num_tx_queues; i++)
			set_check_for_tx_hang(adapter->tx_ring[i]);
	}
6151

6152 6153 6154 6155 6156 6157 6158 6159
	if (!(adapter->flags & IXGBE_FLAG_MSIX_ENABLED)) {
		/*
		 * 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));
6160 6161
	} else {
		/* get one bit for every active tx/rx interrupt vector */
6162
		for (i = 0; i < adapter->num_q_vectors; i++) {
6163
			struct ixgbe_q_vector *qv = adapter->q_vector[i];
6164
			if (qv->rx.ring || qv->tx.ring)
6165 6166
				eics |= ((u64)1 << i);
		}
6167
	}
6168

6169
	/* Cause software interrupt to ensure rings are cleaned */
6170
	ixgbe_irq_rearm_queues(adapter, eics);
6171 6172
}

6173
/**
6174
 * ixgbe_watchdog_update_link - update the link status
6175 6176
 * @adapter: pointer to the device adapter structure
 * @link_speed: pointer to a u32 to store the link_speed
6177
 **/
6178
static void ixgbe_watchdog_update_link(struct ixgbe_adapter *adapter)
6179 6180
{
	struct ixgbe_hw *hw = &adapter->hw;
6181 6182
	u32 link_speed = adapter->link_speed;
	bool link_up = adapter->link_up;
6183
	bool pfc_en = adapter->dcb_cfg.pfc_mode_enable;
6184

6185 6186 6187 6188 6189
	if (!(adapter->flags & IXGBE_FLAG_NEED_LINK_UPDATE))
		return;

	if (hw->mac.ops.check_link) {
		hw->mac.ops.check_link(hw, &link_speed, &link_up, false);
6190
	} else {
6191 6192 6193
		/* always assume link is up, if no check link function */
		link_speed = IXGBE_LINK_SPEED_10GB_FULL;
		link_up = true;
6194
	}
6195 6196 6197 6198

	if (adapter->ixgbe_ieee_pfc)
		pfc_en |= !!(adapter->ixgbe_ieee_pfc->pfc_en);

6199
	if (link_up && !((adapter->flags & IXGBE_FLAG_DCB_ENABLED) && pfc_en)) {
6200
		hw->mac.ops.fc_enable(hw);
6201 6202
		ixgbe_set_rx_drop_en(adapter);
	}
6203 6204 6205 6206 6207 6208 6209 6210 6211 6212 6213

	if (link_up ||
	    time_after(jiffies, (adapter->link_check_timeout +
				 IXGBE_TRY_LINK_TIMEOUT))) {
		adapter->flags &= ~IXGBE_FLAG_NEED_LINK_UPDATE;
		IXGBE_WRITE_REG(hw, IXGBE_EIMS, IXGBE_EIMC_LSC);
		IXGBE_WRITE_FLUSH(hw);
	}

	adapter->link_up = link_up;
	adapter->link_speed = link_speed;
6214 6215
}

6216 6217 6218 6219 6220 6221 6222 6223 6224 6225 6226 6227 6228 6229 6230 6231 6232
static void ixgbe_update_default_up(struct ixgbe_adapter *adapter)
{
#ifdef CONFIG_IXGBE_DCB
	struct net_device *netdev = adapter->netdev;
	struct dcb_app app = {
			      .selector = IEEE_8021QAZ_APP_SEL_ETHERTYPE,
			      .protocol = 0,
			     };
	u8 up = 0;

	if (adapter->dcbx_cap & DCB_CAP_DCBX_VER_IEEE)
		up = dcb_ieee_getapp_mask(netdev, &app);

	adapter->default_up = (up > 1) ? (ffs(up) - 1) : 0;
#endif
}

6233
/**
6234 6235
 * ixgbe_watchdog_link_is_up - update netif_carrier status and
 *                             print link up message
6236
 * @adapter: pointer to the device adapter structure
6237
 **/
6238
static void ixgbe_watchdog_link_is_up(struct ixgbe_adapter *adapter)
6239
{
6240
	struct net_device *netdev = adapter->netdev;
6241
	struct ixgbe_hw *hw = &adapter->hw;
6242 6243
	struct net_device *upper;
	struct list_head *iter;
6244 6245
	u32 link_speed = adapter->link_speed;
	bool flow_rx, flow_tx;
6246

6247 6248
	/* only continue if link was previously down */
	if (netif_carrier_ok(netdev))
6249
		return;
6250

6251
	adapter->flags2 &= ~IXGBE_FLAG2_SEARCH_FOR_SFP;
6252

6253 6254 6255 6256 6257 6258 6259 6260 6261
	switch (hw->mac.type) {
	case ixgbe_mac_82598EB: {
		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);
	}
		break;
	case ixgbe_mac_X540:
6262 6263
	case ixgbe_mac_X550:
	case ixgbe_mac_X550EM_x:
6264 6265 6266 6267 6268 6269 6270 6271 6272 6273 6274
	case 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);
	}
		break;
	default:
		flow_tx = false;
		flow_rx = false;
		break;
6275
	}
6276

6277 6278
	adapter->last_rx_ptp_check = jiffies;

6279
	if (test_bit(__IXGBE_PTP_RUNNING, &adapter->state))
6280
		ixgbe_ptp_start_cyclecounter(adapter);
6281

6282 6283 6284 6285 6286 6287 6288 6289 6290 6291 6292
	e_info(drv, "NIC Link is Up %s, Flow Control: %s\n",
	       (link_speed == IXGBE_LINK_SPEED_10GB_FULL ?
	       "10 Gbps" :
	       (link_speed == IXGBE_LINK_SPEED_1GB_FULL ?
	       "1 Gbps" :
	       (link_speed == IXGBE_LINK_SPEED_100_FULL ?
	       "100 Mbps" :
	       "unknown speed"))),
	       ((flow_rx && flow_tx) ? "RX/TX" :
	       (flow_rx ? "RX" :
	       (flow_tx ? "TX" : "None"))));
6293

6294 6295
	netif_carrier_on(netdev);
	ixgbe_check_vf_rate_limit(adapter);
6296

6297 6298 6299 6300 6301 6302 6303 6304 6305 6306 6307 6308 6309 6310 6311
	/* enable transmits */
	netif_tx_wake_all_queues(adapter->netdev);

	/* enable any upper devices */
	rtnl_lock();
	netdev_for_each_all_upper_dev_rcu(adapter->netdev, upper, iter) {
		if (netif_is_macvlan(upper)) {
			struct macvlan_dev *vlan = netdev_priv(upper);

			if (vlan->fwd_priv)
				netif_tx_wake_all_queues(upper);
		}
	}
	rtnl_unlock();

6312 6313 6314
	/* update the default user priority for VFs */
	ixgbe_update_default_up(adapter);

6315 6316
	/* ping all the active vfs to let them know link has changed */
	ixgbe_ping_all_vfs(adapter);
6317 6318
}

6319
/**
6320 6321
 * ixgbe_watchdog_link_is_down - update netif_carrier status and
 *                               print link down message
6322
 * @adapter: pointer to the adapter structure
6323
 **/
A
Alexander Duyck 已提交
6324
static void ixgbe_watchdog_link_is_down(struct ixgbe_adapter *adapter)
6325
{
6326
	struct net_device *netdev = adapter->netdev;
6327
	struct ixgbe_hw *hw = &adapter->hw;
6328

6329 6330
	adapter->link_up = false;
	adapter->link_speed = 0;
6331

6332 6333 6334
	/* only continue if link was up previously */
	if (!netif_carrier_ok(netdev))
		return;
6335

6336 6337 6338
	/* poll for SFP+ cable when link is down */
	if (ixgbe_is_sfp(hw) && hw->mac.type == ixgbe_mac_82598EB)
		adapter->flags2 |= IXGBE_FLAG2_SEARCH_FOR_SFP;
6339

6340
	if (test_bit(__IXGBE_PTP_RUNNING, &adapter->state))
6341
		ixgbe_ptp_start_cyclecounter(adapter);
6342

6343 6344
	e_info(drv, "NIC Link is Down\n");
	netif_carrier_off(netdev);
6345 6346 6347

	/* ping all the active vfs to let them know link has changed */
	ixgbe_ping_all_vfs(adapter);
6348
}
6349

6350 6351 6352 6353 6354 6355 6356 6357 6358 6359 6360 6361 6362 6363 6364 6365 6366 6367 6368 6369 6370 6371 6372 6373 6374
static bool ixgbe_ring_tx_pending(struct ixgbe_adapter *adapter)
{
	int i;

	for (i = 0; i < adapter->num_tx_queues; i++) {
		struct ixgbe_ring *tx_ring = adapter->tx_ring[i];

		if (tx_ring->next_to_use != tx_ring->next_to_clean)
			return true;
	}

	return false;
}

static bool ixgbe_vf_tx_pending(struct ixgbe_adapter *adapter)
{
	struct ixgbe_hw *hw = &adapter->hw;
	struct ixgbe_ring_feature *vmdq = &adapter->ring_feature[RING_F_VMDQ];
	u32 q_per_pool = __ALIGN_MASK(1, ~vmdq->mask);

	int i, j;

	if (!adapter->num_vfs)
		return false;

6375 6376 6377 6378
	/* resetting the PF is only needed for MAC before X550 */
	if (hw->mac.type >= ixgbe_mac_X550)
		return false;

6379 6380 6381 6382 6383 6384 6385 6386 6387 6388 6389 6390 6391 6392 6393
	for (i = 0; i < adapter->num_vfs; i++) {
		for (j = 0; j < q_per_pool; j++) {
			u32 h, t;

			h = IXGBE_READ_REG(hw, IXGBE_PVFTDHN(q_per_pool, i, j));
			t = IXGBE_READ_REG(hw, IXGBE_PVFTDTN(q_per_pool, i, j));

			if (h != t)
				return true;
		}
	}

	return false;
}

6394 6395
/**
 * ixgbe_watchdog_flush_tx - flush queues on link down
6396
 * @adapter: pointer to the device adapter structure
6397 6398 6399 6400
 **/
static void ixgbe_watchdog_flush_tx(struct ixgbe_adapter *adapter)
{
	if (!netif_carrier_ok(adapter->netdev)) {
6401 6402
		if (ixgbe_ring_tx_pending(adapter) ||
		    ixgbe_vf_tx_pending(adapter)) {
6403 6404 6405 6406 6407
			/* We've lost link, so the controller stops DMA,
			 * but we've got queued Tx work that's never going
			 * to get done, so reset controller to flush Tx.
			 * (Do the reset outside of interrupt context).
			 */
6408
			e_warn(drv, "initiating reset to clear Tx work after link loss\n");
6409
			adapter->flags2 |= IXGBE_FLAG2_RESET_REQUESTED;
6410
		}
6411 6412 6413
	}
}

6414 6415 6416 6417 6418 6419 6420 6421 6422 6423 6424 6425 6426 6427 6428 6429 6430 6431 6432 6433 6434 6435 6436 6437 6438 6439 6440 6441 6442 6443 6444 6445 6446 6447 6448 6449 6450 6451 6452 6453 6454 6455 6456 6457 6458 6459 6460 6461 6462 6463 6464 6465 6466 6467 6468 6469 6470 6471 6472 6473
#ifdef CONFIG_PCI_IOV
static inline void ixgbe_issue_vf_flr(struct ixgbe_adapter *adapter,
				      struct pci_dev *vfdev)
{
	if (!pci_wait_for_pending_transaction(vfdev))
		e_dev_warn("Issuing VFLR with pending transactions\n");

	e_dev_err("Issuing VFLR for VF %s\n", pci_name(vfdev));
	pcie_capability_set_word(vfdev, PCI_EXP_DEVCTL, PCI_EXP_DEVCTL_BCR_FLR);

	msleep(100);
}

static void ixgbe_check_for_bad_vf(struct ixgbe_adapter *adapter)
{
	struct ixgbe_hw *hw = &adapter->hw;
	struct pci_dev *pdev = adapter->pdev;
	struct pci_dev *vfdev;
	u32 gpc;
	int pos;
	unsigned short vf_id;

	if (!(netif_carrier_ok(adapter->netdev)))
		return;

	gpc = IXGBE_READ_REG(hw, IXGBE_TXDGPC);
	if (gpc) /* If incrementing then no need for the check below */
		return;
	/* Check to see if a bad DMA write target from an errant or
	 * malicious VF has caused a PCIe error.  If so then we can
	 * issue a VFLR to the offending VF(s) and then resume without
	 * requesting a full slot reset.
	 */

	if (!pdev)
		return;

	pos = pci_find_ext_capability(pdev, PCI_EXT_CAP_ID_SRIOV);
	if (!pos)
		return;

	/* get the device ID for the VF */
	pci_read_config_word(pdev, pos + PCI_SRIOV_VF_DID, &vf_id);

	/* check status reg for all VFs owned by this PF */
	vfdev = pci_get_device(pdev->vendor, vf_id, NULL);
	while (vfdev) {
		if (vfdev->is_virtfn && (vfdev->physfn == pdev)) {
			u16 status_reg;

			pci_read_config_word(vfdev, PCI_STATUS, &status_reg);
			if (status_reg & PCI_STATUS_REC_MASTER_ABORT)
				/* issue VFLR */
				ixgbe_issue_vf_flr(adapter, vfdev);
		}

		vfdev = pci_get_device(pdev->vendor, vf_id, vfdev);
	}
}

6474 6475 6476 6477
static void ixgbe_spoof_check(struct ixgbe_adapter *adapter)
{
	u32 ssvpc;

6478 6479 6480
	/* Do not perform spoof check for 82598 or if not in IOV mode */
	if (adapter->hw.mac.type == ixgbe_mac_82598EB ||
	    adapter->num_vfs == 0)
6481 6482 6483 6484 6485 6486 6487 6488 6489 6490 6491
		return;

	ssvpc = IXGBE_READ_REG(&adapter->hw, IXGBE_SSVPC);

	/*
	 * ssvpc register is cleared on read, if zero then no
	 * spoofed packets in the last interval.
	 */
	if (!ssvpc)
		return;

6492
	e_warn(drv, "%u Spoofed packets detected\n", ssvpc);
6493
}
6494 6495 6496 6497 6498 6499 6500 6501 6502 6503 6504
#else
static void ixgbe_spoof_check(struct ixgbe_adapter __always_unused *adapter)
{
}

static void
ixgbe_check_for_bad_vf(struct ixgbe_adapter __always_unused *adapter)
{
}
#endif /* CONFIG_PCI_IOV */

6505

6506 6507
/**
 * ixgbe_watchdog_subtask - check and bring link up
6508
 * @adapter: pointer to the device adapter structure
6509 6510 6511
 **/
static void ixgbe_watchdog_subtask(struct ixgbe_adapter *adapter)
{
6512
	/* if interface is down, removing or resetting, do nothing */
6513
	if (test_bit(__IXGBE_DOWN, &adapter->state) ||
6514
	    test_bit(__IXGBE_REMOVING, &adapter->state) ||
6515
	    test_bit(__IXGBE_RESETTING, &adapter->state))
6516 6517 6518 6519 6520 6521 6522 6523
		return;

	ixgbe_watchdog_update_link(adapter);

	if (adapter->link_up)
		ixgbe_watchdog_link_is_up(adapter);
	else
		ixgbe_watchdog_link_is_down(adapter);
6524

6525
	ixgbe_check_for_bad_vf(adapter);
6526
	ixgbe_spoof_check(adapter);
6527
	ixgbe_update_stats(adapter);
6528 6529

	ixgbe_watchdog_flush_tx(adapter);
6530
}
6531

6532
/**
6533
 * ixgbe_sfp_detection_subtask - poll for SFP+ cable
6534
 * @adapter: the ixgbe adapter structure
6535
 **/
6536
static void ixgbe_sfp_detection_subtask(struct ixgbe_adapter *adapter)
6537 6538
{
	struct ixgbe_hw *hw = &adapter->hw;
6539
	s32 err;
6540

6541 6542 6543 6544
	/* not searching for SFP so there is nothing to do here */
	if (!(adapter->flags2 & IXGBE_FLAG2_SEARCH_FOR_SFP) &&
	    !(adapter->flags2 & IXGBE_FLAG2_SFP_NEEDS_RESET))
		return;
6545

6546 6547 6548
	/* someone else is in init, wait until next service event */
	if (test_and_set_bit(__IXGBE_IN_SFP_INIT, &adapter->state))
		return;
6549

6550 6551 6552
	err = hw->phy.ops.identify_sfp(hw);
	if (err == IXGBE_ERR_SFP_NOT_SUPPORTED)
		goto sfp_out;
6553

6554 6555 6556 6557
	if (err == IXGBE_ERR_SFP_NOT_PRESENT) {
		/* If no cable is present, then we need to reset
		 * the next time we find a good cable. */
		adapter->flags2 |= IXGBE_FLAG2_SFP_NEEDS_RESET;
6558
	}
6559

6560 6561 6562
	/* exit on error */
	if (err)
		goto sfp_out;
6563

6564 6565 6566
	/* exit if reset not needed */
	if (!(adapter->flags2 & IXGBE_FLAG2_SFP_NEEDS_RESET))
		goto sfp_out;
6567

6568
	adapter->flags2 &= ~IXGBE_FLAG2_SFP_NEEDS_RESET;
6569

6570 6571 6572 6573 6574 6575 6576 6577 6578 6579 6580 6581 6582 6583 6584 6585 6586 6587 6588 6589 6590 6591 6592 6593 6594 6595
	/*
	 * A module may be identified correctly, but the EEPROM may not have
	 * support for that module.  setup_sfp() will fail in that case, so
	 * we should not allow that module to load.
	 */
	if (hw->mac.type == ixgbe_mac_82598EB)
		err = hw->phy.ops.reset(hw);
	else
		err = hw->mac.ops.setup_sfp(hw);

	if (err == IXGBE_ERR_SFP_NOT_SUPPORTED)
		goto sfp_out;

	adapter->flags |= IXGBE_FLAG_NEED_LINK_CONFIG;
	e_info(probe, "detected SFP+: %d\n", hw->phy.sfp_type);

sfp_out:
	clear_bit(__IXGBE_IN_SFP_INIT, &adapter->state);

	if ((err == IXGBE_ERR_SFP_NOT_SUPPORTED) &&
	    (adapter->netdev->reg_state == NETREG_REGISTERED)) {
		e_dev_err("failed to initialize because an unsupported "
			  "SFP+ module type was detected.\n");
		e_dev_err("Reload the driver after installing a "
			  "supported module.\n");
		unregister_netdev(adapter->netdev);
6596
	}
6597
}
6598

6599 6600
/**
 * ixgbe_sfp_link_config_subtask - set up link SFP after module install
6601
 * @adapter: the ixgbe adapter structure
6602 6603 6604 6605
 **/
static void ixgbe_sfp_link_config_subtask(struct ixgbe_adapter *adapter)
{
	struct ixgbe_hw *hw = &adapter->hw;
J
Josh Hay 已提交
6606 6607
	u32 speed;
	bool autoneg = false;
6608 6609 6610 6611 6612 6613 6614 6615 6616 6617

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

	/* someone else is in init, wait until next service event */
	if (test_and_set_bit(__IXGBE_IN_SFP_INIT, &adapter->state))
		return;

	adapter->flags &= ~IXGBE_FLAG_NEED_LINK_CONFIG;

J
Josh Hay 已提交
6618
	speed = hw->phy.autoneg_advertised;
6619
	if ((!speed) && (hw->mac.ops.get_link_capabilities)) {
J
Josh Hay 已提交
6620
		hw->mac.ops.get_link_capabilities(hw, &speed, &autoneg);
6621 6622 6623 6624 6625 6626 6627 6628

		/* setup the highest link when no autoneg */
		if (!autoneg) {
			if (speed & IXGBE_LINK_SPEED_10GB_FULL)
				speed = IXGBE_LINK_SPEED_10GB_FULL;
		}
	}

6629
	if (hw->mac.ops.setup_link)
J
Josh Hay 已提交
6630
		hw->mac.ops.setup_link(hw, speed, true);
6631 6632 6633 6634 6635 6636 6637 6638 6639 6640 6641 6642 6643 6644 6645

	adapter->flags |= IXGBE_FLAG_NEED_LINK_UPDATE;
	adapter->link_check_timeout = jiffies;
	clear_bit(__IXGBE_IN_SFP_INIT, &adapter->state);
}

/**
 * ixgbe_service_timer - Timer Call-back
 * @data: pointer to adapter cast into an unsigned long
 **/
static void ixgbe_service_timer(unsigned long data)
{
	struct ixgbe_adapter *adapter = (struct ixgbe_adapter *)data;
	unsigned long next_event_offset;

6646 6647 6648 6649 6650
	/* poll faster when waiting for link */
	if (adapter->flags & IXGBE_FLAG_NEED_LINK_UPDATE)
		next_event_offset = HZ / 10;
	else
		next_event_offset = HZ * 2;
6651

6652 6653 6654
	/* Reset the timer */
	mod_timer(&adapter->service_timer, next_event_offset + jiffies);

6655
	ixgbe_service_event_schedule(adapter);
6656 6657
}

6658 6659 6660 6661 6662 6663 6664
static void ixgbe_reset_subtask(struct ixgbe_adapter *adapter)
{
	if (!(adapter->flags2 & IXGBE_FLAG2_RESET_REQUESTED))
		return;

	adapter->flags2 &= ~IXGBE_FLAG2_RESET_REQUESTED;

6665
	/* If we're already down, removing or resetting, just bail */
6666
	if (test_bit(__IXGBE_DOWN, &adapter->state) ||
6667
	    test_bit(__IXGBE_REMOVING, &adapter->state) ||
6668 6669 6670 6671 6672 6673 6674
	    test_bit(__IXGBE_RESETTING, &adapter->state))
		return;

	ixgbe_dump(adapter);
	netdev_err(adapter->netdev, "Reset adapter\n");
	adapter->tx_timeout_count++;

6675
	rtnl_lock();
6676
	ixgbe_reinit_locked(adapter);
6677
	rtnl_unlock();
6678 6679
}

6680 6681 6682 6683 6684 6685 6686 6687 6688
/**
 * ixgbe_service_task - manages and runs subtasks
 * @work: pointer to work_struct containing our data
 **/
static void ixgbe_service_task(struct work_struct *work)
{
	struct ixgbe_adapter *adapter = container_of(work,
						     struct ixgbe_adapter,
						     service_task);
6689 6690 6691 6692 6693 6694 6695 6696 6697
	if (ixgbe_removed(adapter->hw.hw_addr)) {
		if (!test_bit(__IXGBE_DOWN, &adapter->state)) {
			rtnl_lock();
			ixgbe_down(adapter);
			rtnl_unlock();
		}
		ixgbe_service_event_complete(adapter);
		return;
	}
6698
	ixgbe_reset_subtask(adapter);
6699 6700
	ixgbe_sfp_detection_subtask(adapter);
	ixgbe_sfp_link_config_subtask(adapter);
6701
	ixgbe_check_overtemp_subtask(adapter);
6702
	ixgbe_watchdog_subtask(adapter);
6703
	ixgbe_fdir_reinit_subtask(adapter);
6704
	ixgbe_check_hang_subtask(adapter);
6705

6706
	if (test_bit(__IXGBE_PTP_RUNNING, &adapter->state)) {
6707 6708 6709
		ixgbe_ptp_overflow_check(adapter);
		ixgbe_ptp_rx_hang(adapter);
	}
6710 6711

	ixgbe_service_event_complete(adapter);
6712 6713
}

6714 6715
static int ixgbe_tso(struct ixgbe_ring *tx_ring,
		     struct ixgbe_tx_buffer *first,
6716
		     u8 *hdr_len)
6717
{
6718
	struct sk_buff *skb = first->skb;
6719 6720
	u32 vlan_macip_lens, type_tucmd;
	u32 mss_l4len_idx, l4len;
6721
	int err;
6722

6723 6724 6725
	if (skb->ip_summed != CHECKSUM_PARTIAL)
		return 0;

6726 6727
	if (!skb_is_gso(skb))
		return 0;
6728

6729 6730 6731
	err = skb_cow_head(skb, 0);
	if (err < 0)
		return err;
6732

6733 6734 6735
	/* ADV DTYP TUCMD MKRLOC/ISCSIHEDLEN */
	type_tucmd = IXGBE_ADVTXD_TUCMD_L4T_TCP;

6736
	if (first->protocol == htons(ETH_P_IP)) {
6737 6738 6739 6740 6741 6742 6743 6744
		struct iphdr *iph = ip_hdr(skb);
		iph->tot_len = 0;
		iph->check = 0;
		tcp_hdr(skb)->check = ~csum_tcpudp_magic(iph->saddr,
							 iph->daddr, 0,
							 IPPROTO_TCP,
							 0);
		type_tucmd |= IXGBE_ADVTXD_TUCMD_IPV4;
6745 6746 6747
		first->tx_flags |= IXGBE_TX_FLAGS_TSO |
				   IXGBE_TX_FLAGS_CSUM |
				   IXGBE_TX_FLAGS_IPV4;
6748 6749 6750 6751 6752 6753
	} else if (skb_is_gso_v6(skb)) {
		ipv6_hdr(skb)->payload_len = 0;
		tcp_hdr(skb)->check =
		    ~csum_ipv6_magic(&ipv6_hdr(skb)->saddr,
				     &ipv6_hdr(skb)->daddr,
				     0, IPPROTO_TCP, 0);
6754 6755
		first->tx_flags |= IXGBE_TX_FLAGS_TSO |
				   IXGBE_TX_FLAGS_CSUM;
6756 6757
	}

6758
	/* compute header lengths */
6759 6760 6761
	l4len = tcp_hdrlen(skb);
	*hdr_len = skb_transport_offset(skb) + l4len;

6762 6763 6764 6765
	/* update gso size and bytecount with header size */
	first->gso_segs = skb_shinfo(skb)->gso_segs;
	first->bytecount += (first->gso_segs - 1) * *hdr_len;

6766
	/* mss_l4len_id: use 0 as index for TSO */
6767 6768 6769 6770 6771 6772
	mss_l4len_idx = l4len << IXGBE_ADVTXD_L4LEN_SHIFT;
	mss_l4len_idx |= skb_shinfo(skb)->gso_size << IXGBE_ADVTXD_MSS_SHIFT;

	/* vlan_macip_lens: HEADLEN, MACLEN, VLAN tag */
	vlan_macip_lens = skb_network_header_len(skb);
	vlan_macip_lens |= skb_network_offset(skb) << IXGBE_ADVTXD_MACLEN_SHIFT;
6773
	vlan_macip_lens |= first->tx_flags & IXGBE_TX_FLAGS_VLAN_MASK;
6774 6775

	ixgbe_tx_ctxtdesc(tx_ring, vlan_macip_lens, 0, type_tucmd,
6776
			  mss_l4len_idx);
6777 6778 6779 6780

	return 1;
}

6781 6782
static void ixgbe_tx_csum(struct ixgbe_ring *tx_ring,
			  struct ixgbe_tx_buffer *first)
6783
{
6784
	struct sk_buff *skb = first->skb;
6785 6786 6787
	u32 vlan_macip_lens = 0;
	u32 mss_l4len_idx = 0;
	u32 type_tucmd = 0;
6788

6789
	if (skb->ip_summed != CHECKSUM_PARTIAL) {
6790 6791 6792
		if (!(first->tx_flags & IXGBE_TX_FLAGS_HW_VLAN) &&
		    !(first->tx_flags & IXGBE_TX_FLAGS_CC))
			return;
6793 6794
	} else {
		u8 l4_hdr = 0;
6795
		switch (first->protocol) {
6796
		case htons(ETH_P_IP):
6797 6798 6799
			vlan_macip_lens |= skb_network_header_len(skb);
			type_tucmd |= IXGBE_ADVTXD_TUCMD_IPV4;
			l4_hdr = ip_hdr(skb)->protocol;
6800
			break;
6801
		case htons(ETH_P_IPV6):
6802 6803 6804 6805 6806 6807 6808
			vlan_macip_lens |= skb_network_header_len(skb);
			l4_hdr = ipv6_hdr(skb)->nexthdr;
			break;
		default:
			if (unlikely(net_ratelimit())) {
				dev_warn(tx_ring->dev,
				 "partial checksum but proto=%x!\n",
6809
				 first->protocol);
6810
			}
6811 6812
			break;
		}
6813 6814

		switch (l4_hdr) {
6815
		case IPPROTO_TCP:
6816 6817 6818
			type_tucmd |= IXGBE_ADVTXD_TUCMD_L4T_TCP;
			mss_l4len_idx = tcp_hdrlen(skb) <<
					IXGBE_ADVTXD_L4LEN_SHIFT;
6819 6820
			break;
		case IPPROTO_SCTP:
6821 6822 6823 6824 6825 6826 6827 6828 6829 6830 6831 6832
			type_tucmd |= IXGBE_ADVTXD_TUCMD_L4T_SCTP;
			mss_l4len_idx = sizeof(struct sctphdr) <<
					IXGBE_ADVTXD_L4LEN_SHIFT;
			break;
		case IPPROTO_UDP:
			mss_l4len_idx = sizeof(struct udphdr) <<
					IXGBE_ADVTXD_L4LEN_SHIFT;
			break;
		default:
			if (unlikely(net_ratelimit())) {
				dev_warn(tx_ring->dev,
				 "partial checksum but l4 proto=%x!\n",
6833
				 l4_hdr);
6834
			}
6835 6836
			break;
		}
6837 6838 6839

		/* update TX checksum flag */
		first->tx_flags |= IXGBE_TX_FLAGS_CSUM;
6840 6841
	}

6842
	/* vlan_macip_lens: MACLEN, VLAN tag */
6843
	vlan_macip_lens |= skb_network_offset(skb) << IXGBE_ADVTXD_MACLEN_SHIFT;
6844
	vlan_macip_lens |= first->tx_flags & IXGBE_TX_FLAGS_VLAN_MASK;
6845

6846 6847
	ixgbe_tx_ctxtdesc(tx_ring, vlan_macip_lens, 0,
			  type_tucmd, mss_l4len_idx);
6848 6849
}

6850 6851 6852 6853 6854 6855
#define IXGBE_SET_FLAG(_input, _flag, _result) \
	((_flag <= _result) ? \
	 ((u32)(_input & _flag) * (_result / _flag)) : \
	 ((u32)(_input & _flag) / (_flag / _result)))

static u32 ixgbe_tx_cmd_type(struct sk_buff *skb, u32 tx_flags)
6856
{
6857
	/* set type for advanced descriptor with frame checksum insertion */
6858 6859 6860
	u32 cmd_type = IXGBE_ADVTXD_DTYP_DATA |
		       IXGBE_ADVTXD_DCMD_DEXT |
		       IXGBE_ADVTXD_DCMD_IFCS;
6861

6862
	/* set HW vlan bit if vlan is present */
6863 6864
	cmd_type |= IXGBE_SET_FLAG(tx_flags, IXGBE_TX_FLAGS_HW_VLAN,
				   IXGBE_ADVTXD_DCMD_VLE);
6865

6866
	/* set segmentation enable bits for TSO/FSO */
6867 6868 6869 6870 6871 6872
	cmd_type |= IXGBE_SET_FLAG(tx_flags, IXGBE_TX_FLAGS_TSO,
				   IXGBE_ADVTXD_DCMD_TSE);

	/* set timestamp bit if present */
	cmd_type |= IXGBE_SET_FLAG(tx_flags, IXGBE_TX_FLAGS_TSTAMP,
				   IXGBE_ADVTXD_MAC_TSTAMP);
6873

6874
	/* insert frame checksum */
6875
	cmd_type ^= IXGBE_SET_FLAG(skb->no_fcs, 1, IXGBE_ADVTXD_DCMD_IFCS);
6876

6877 6878
	return cmd_type;
}
6879

6880 6881
static void ixgbe_tx_olinfo_status(union ixgbe_adv_tx_desc *tx_desc,
				   u32 tx_flags, unsigned int paylen)
6882
{
6883
	u32 olinfo_status = paylen << IXGBE_ADVTXD_PAYLEN_SHIFT;
6884

6885
	/* enable L4 checksum for TSO and TX checksum offload */
6886 6887 6888
	olinfo_status |= IXGBE_SET_FLAG(tx_flags,
					IXGBE_TX_FLAGS_CSUM,
					IXGBE_ADVTXD_POPTS_TXSM);
6889

6890
	/* enble IPv4 checksum for TSO */
6891 6892 6893
	olinfo_status |= IXGBE_SET_FLAG(tx_flags,
					IXGBE_TX_FLAGS_IPV4,
					IXGBE_ADVTXD_POPTS_IXSM);
6894

6895 6896 6897 6898
	/*
	 * Check Context must be set if Tx switch is enabled, which it
	 * always is for case where virtual functions are running
	 */
6899 6900 6901
	olinfo_status |= IXGBE_SET_FLAG(tx_flags,
					IXGBE_TX_FLAGS_CC,
					IXGBE_ADVTXD_CC);
6902

6903
	tx_desc->read.olinfo_status = cpu_to_le32(olinfo_status);
6904
}
6905

6906 6907 6908 6909 6910 6911 6912 6913 6914 6915 6916 6917 6918 6919 6920 6921 6922 6923 6924 6925 6926 6927 6928 6929 6930 6931 6932 6933 6934 6935
static int __ixgbe_maybe_stop_tx(struct ixgbe_ring *tx_ring, u16 size)
{
	netif_stop_subqueue(tx_ring->netdev, tx_ring->queue_index);

	/* 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 */
	netif_start_subqueue(tx_ring->netdev, tx_ring->queue_index);
	++tx_ring->tx_stats.restart_queue;
	return 0;
}

static inline int ixgbe_maybe_stop_tx(struct ixgbe_ring *tx_ring, u16 size)
{
	if (likely(ixgbe_desc_unused(tx_ring) >= size))
		return 0;

	return __ixgbe_maybe_stop_tx(tx_ring, size);
}

6936 6937 6938 6939 6940 6941 6942
#define IXGBE_TXD_CMD (IXGBE_TXD_CMD_EOP | \
		       IXGBE_TXD_CMD_RS)

static void ixgbe_tx_map(struct ixgbe_ring *tx_ring,
			 struct ixgbe_tx_buffer *first,
			 const u8 hdr_len)
{
6943
	struct sk_buff *skb = first->skb;
6944
	struct ixgbe_tx_buffer *tx_buffer;
6945
	union ixgbe_adv_tx_desc *tx_desc;
6946 6947 6948
	struct skb_frag_struct *frag;
	dma_addr_t dma;
	unsigned int data_len, size;
6949
	u32 tx_flags = first->tx_flags;
6950
	u32 cmd_type = ixgbe_tx_cmd_type(skb, tx_flags);
6951 6952
	u16 i = tx_ring->next_to_use;

6953 6954
	tx_desc = IXGBE_TX_DESC(tx_ring, i);

6955 6956 6957 6958
	ixgbe_tx_olinfo_status(tx_desc, tx_flags, skb->len - hdr_len);

	size = skb_headlen(skb);
	data_len = skb->data_len;
6959

6960 6961
#ifdef IXGBE_FCOE
	if (tx_flags & IXGBE_TX_FLAGS_FCOE) {
6962
		if (data_len < sizeof(struct fcoe_crc_eof)) {
6963 6964
			size -= sizeof(struct fcoe_crc_eof) - data_len;
			data_len = 0;
6965 6966
		} else {
			data_len -= sizeof(struct fcoe_crc_eof);
6967 6968
		}
	}
6969

6970
#endif
6971
	dma = dma_map_single(tx_ring->dev, skb->data, size, DMA_TO_DEVICE);
6972

6973
	tx_buffer = first;
6974

6975 6976 6977 6978 6979 6980 6981 6982 6983
	for (frag = &skb_shinfo(skb)->frags[0];; frag++) {
		if (dma_mapping_error(tx_ring->dev, dma))
			goto dma_error;

		/* record length, and DMA address */
		dma_unmap_len_set(tx_buffer, len, size);
		dma_unmap_addr_set(tx_buffer, dma, dma);

		tx_desc->read.buffer_addr = cpu_to_le64(dma);
6984

6985
		while (unlikely(size > IXGBE_MAX_DATA_PER_TXD)) {
6986
			tx_desc->read.cmd_type_len =
6987
				cpu_to_le32(cmd_type ^ IXGBE_MAX_DATA_PER_TXD);
6988

6989
			i++;
6990
			tx_desc++;
6991
			if (i == tx_ring->count) {
6992
				tx_desc = IXGBE_TX_DESC(tx_ring, 0);
6993 6994
				i = 0;
			}
6995
			tx_desc->read.olinfo_status = 0;
6996 6997 6998 6999 7000

			dma += IXGBE_MAX_DATA_PER_TXD;
			size -= IXGBE_MAX_DATA_PER_TXD;

			tx_desc->read.buffer_addr = cpu_to_le64(dma);
7001
		}
7002

7003 7004
		if (likely(!data_len))
			break;
7005

7006
		tx_desc->read.cmd_type_len = cpu_to_le32(cmd_type ^ size);
7007

7008 7009 7010 7011 7012 7013
		i++;
		tx_desc++;
		if (i == tx_ring->count) {
			tx_desc = IXGBE_TX_DESC(tx_ring, 0);
			i = 0;
		}
7014
		tx_desc->read.olinfo_status = 0;
7015

7016
#ifdef IXGBE_FCOE
E
Eric Dumazet 已提交
7017
		size = min_t(unsigned int, data_len, skb_frag_size(frag));
7018
#else
E
Eric Dumazet 已提交
7019
		size = skb_frag_size(frag);
7020 7021
#endif
		data_len -= size;
7022

7023 7024
		dma = skb_frag_dma_map(tx_ring->dev, frag, 0, size,
				       DMA_TO_DEVICE);
7025

7026 7027
		tx_buffer = &tx_ring->tx_buffer_info[i];
	}
7028

7029
	/* write last descriptor with RS and EOP bits */
7030 7031
	cmd_type |= size | IXGBE_TXD_CMD;
	tx_desc->read.cmd_type_len = cpu_to_le32(cmd_type);
7032

7033
	netdev_tx_sent_queue(txring_txq(tx_ring), first->bytecount);
7034

7035 7036
	/* set the timestamp */
	first->time_stamp = jiffies;
7037 7038

	/*
7039 7040 7041 7042 7043 7044
	 * 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).
	 *
	 * We also need this memory barrier to make certain all of the
	 * status bits have been updated before next_to_watch is written.
7045 7046 7047
	 */
	wmb();

7048 7049 7050
	/* set next_to_watch value indicating a packet is present */
	first->next_to_watch = tx_desc;

7051 7052 7053 7054 7055 7056
	i++;
	if (i == tx_ring->count)
		i = 0;

	tx_ring->next_to_use = i;

7057 7058 7059
	ixgbe_maybe_stop_tx(tx_ring, DESC_NEEDED);

	if (netif_xmit_stopped(txring_txq(tx_ring)) || !skb->xmit_more) {
7060 7061 7062 7063 7064 7065
		writel(i, tx_ring->tail);

		/* we need this if more than one processor can write to our tail
		 * at a time, it synchronizes IO on IA64/Altix systems
		 */
		mmiowb();
7066
	}
7067

7068 7069
	return;
dma_error:
7070
	dev_err(tx_ring->dev, "TX DMA map failed\n");
7071 7072 7073

	/* clear dma mappings for failed tx_buffer_info map */
	for (;;) {
7074 7075 7076
		tx_buffer = &tx_ring->tx_buffer_info[i];
		ixgbe_unmap_and_free_tx_resource(tx_ring, tx_buffer);
		if (tx_buffer == first)
7077 7078 7079 7080 7081 7082 7083
			break;
		if (i == 0)
			i = tx_ring->count;
		i--;
	}

	tx_ring->next_to_use = i;
7084 7085
}

7086
static void ixgbe_atr(struct ixgbe_ring *ring,
7087
		      struct ixgbe_tx_buffer *first)
7088 7089 7090 7091 7092 7093 7094 7095 7096
{
	struct ixgbe_q_vector *q_vector = ring->q_vector;
	union ixgbe_atr_hash_dword input = { .dword = 0 };
	union ixgbe_atr_hash_dword common = { .dword = 0 };
	union {
		unsigned char *network;
		struct iphdr *ipv4;
		struct ipv6hdr *ipv6;
	} hdr;
7097
	struct tcphdr *th;
7098
	__be16 vlan_id;
7099

7100 7101 7102 7103 7104 7105
	/* if ring doesn't have a interrupt vector, cannot perform ATR */
	if (!q_vector)
		return;

	/* do nothing if sampling is disabled */
	if (!ring->atr_sample_rate)
7106
		return;
7107

7108
	ring->atr_count++;
7109

7110
	/* snag network header to get L4 type and address */
7111
	hdr.network = skb_network_header(first->skb);
7112 7113

	/* Currently only IPv4/IPv6 with TCP is supported */
7114
	if ((first->protocol != htons(ETH_P_IPV6) ||
7115
	     hdr.ipv6->nexthdr != IPPROTO_TCP) &&
7116
	    (first->protocol != htons(ETH_P_IP) ||
7117 7118
	     hdr.ipv4->protocol != IPPROTO_TCP))
		return;
7119

7120
	th = tcp_hdr(first->skb);
7121

7122 7123
	/* skip this packet since it is invalid or the socket is closing */
	if (!th || th->fin)
7124 7125 7126 7127 7128 7129 7130 7131 7132
		return;

	/* sample on all syn packets or once every atr sample count */
	if (!th->syn && (ring->atr_count < ring->atr_sample_rate))
		return;

	/* reset sample count */
	ring->atr_count = 0;

7133
	vlan_id = htons(first->tx_flags >> IXGBE_TX_FLAGS_VLAN_SHIFT);
7134 7135 7136 7137 7138 7139 7140 7141 7142 7143 7144 7145 7146 7147

	/*
	 * src and dst are inverted, think how the receiver sees them
	 *
	 * The input is broken into two sections, a non-compressed section
	 * containing vm_pool, vlan_id, and flow_type.  The rest of the data
	 * is XORed together and stored in the compressed dword.
	 */
	input.formatted.vlan_id = vlan_id;

	/*
	 * since src port and flex bytes occupy the same word XOR them together
	 * and write the value to source port portion of compressed dword
	 */
7148
	if (first->tx_flags & (IXGBE_TX_FLAGS_SW_VLAN | IXGBE_TX_FLAGS_HW_VLAN))
7149
		common.port.src ^= th->dest ^ htons(ETH_P_8021Q);
7150
	else
7151
		common.port.src ^= th->dest ^ first->protocol;
7152 7153
	common.port.dst ^= th->source;

7154
	if (first->protocol == htons(ETH_P_IP)) {
7155 7156 7157 7158 7159 7160 7161 7162 7163 7164 7165 7166 7167
		input.formatted.flow_type = IXGBE_ATR_FLOW_TYPE_TCPV4;
		common.ip ^= hdr.ipv4->saddr ^ hdr.ipv4->daddr;
	} else {
		input.formatted.flow_type = IXGBE_ATR_FLOW_TYPE_TCPV6;
		common.ip ^= hdr.ipv6->saddr.s6_addr32[0] ^
			     hdr.ipv6->saddr.s6_addr32[1] ^
			     hdr.ipv6->saddr.s6_addr32[2] ^
			     hdr.ipv6->saddr.s6_addr32[3] ^
			     hdr.ipv6->daddr.s6_addr32[0] ^
			     hdr.ipv6->daddr.s6_addr32[1] ^
			     hdr.ipv6->daddr.s6_addr32[2] ^
			     hdr.ipv6->daddr.s6_addr32[3];
	}
7168 7169

	/* This assumes the Rx queue and Tx queue are bound to the same CPU */
7170 7171
	ixgbe_fdir_add_signature_filter_82599(&q_vector->adapter->hw,
					      input, common, ring->queue_index);
7172 7173
}

7174
static u16 ixgbe_select_queue(struct net_device *dev, struct sk_buff *skb,
7175
			      void *accel_priv, select_queue_fallback_t fallback)
7176
{
7177 7178
	struct ixgbe_fwd_adapter *fwd_adapter = accel_priv;
#ifdef IXGBE_FCOE
7179 7180 7181
	struct ixgbe_adapter *adapter;
	struct ixgbe_ring_feature *f;
	int txq;
7182 7183 7184 7185 7186 7187
#endif

	if (fwd_adapter)
		return skb->queue_mapping + fwd_adapter->tx_base_queue;

#ifdef IXGBE_FCOE
7188

7189 7190 7191 7192 7193
	/*
	 * only execute the code below if protocol is FCoE
	 * or FIP and we have FCoE enabled on the adapter
	 */
	switch (vlan_get_protocol(skb)) {
7194 7195
	case htons(ETH_P_FCOE):
	case htons(ETH_P_FIP):
7196
		adapter = netdev_priv(dev);
7197

7198 7199 7200
		if (adapter->flags & IXGBE_FLAG_FCOE_ENABLED)
			break;
	default:
7201
		return fallback(dev, skb);
7202
	}
7203

7204
	f = &adapter->ring_feature[RING_F_FCOE];
7205

7206 7207
	txq = skb_rx_queue_recorded(skb) ? skb_get_rx_queue(skb) :
					   smp_processor_id();
7208

7209 7210
	while (txq >= f->indices)
		txq -= f->indices;
7211

7212
	return txq + f->offset;
7213
#else
7214
	return fallback(dev, skb);
7215
#endif
7216 7217
}

7218
netdev_tx_t ixgbe_xmit_frame_ring(struct sk_buff *skb,
7219 7220
			  struct ixgbe_adapter *adapter,
			  struct ixgbe_ring *tx_ring)
7221
{
7222
	struct ixgbe_tx_buffer *first;
7223
	int tso;
7224
	u32 tx_flags = 0;
7225 7226
	unsigned short f;
	u16 count = TXD_USE_COUNT(skb_headlen(skb));
7227
	__be16 protocol = skb->protocol;
7228
	u8 hdr_len = 0;
7229

7230 7231
	/*
	 * need: 1 descriptor per page * PAGE_SIZE/IXGBE_MAX_DATA_PER_TXD,
7232
	 *       + 1 desc for skb_headlen/IXGBE_MAX_DATA_PER_TXD,
7233 7234 7235 7236 7237 7238
	 *       + 2 desc gap to keep tail from touching head,
	 *       + 1 desc for context descriptor,
	 * otherwise try next time
	 */
	for (f = 0; f < skb_shinfo(skb)->nr_frags; f++)
		count += TXD_USE_COUNT(skb_shinfo(skb)->frags[f].size);
7239

7240 7241 7242 7243 7244
	if (ixgbe_maybe_stop_tx(tx_ring, count + 3)) {
		tx_ring->tx_stats.tx_busy++;
		return NETDEV_TX_BUSY;
	}

7245 7246 7247
	/* record the location of the first descriptor for this packet */
	first = &tx_ring->tx_buffer_info[tx_ring->next_to_use];
	first->skb = skb;
7248 7249
	first->bytecount = skb->len;
	first->gso_segs = 1;
7250

7251
	/* if we have a HW VLAN tag being added default to the HW one */
7252 7253
	if (skb_vlan_tag_present(skb)) {
		tx_flags |= skb_vlan_tag_get(skb) << IXGBE_TX_FLAGS_VLAN_SHIFT;
7254 7255
		tx_flags |= IXGBE_TX_FLAGS_HW_VLAN;
	/* else if it is a SW VLAN check the next protocol and store the tag */
7256
	} else if (protocol == htons(ETH_P_8021Q)) {
7257 7258 7259 7260 7261
		struct vlan_hdr *vhdr, _vhdr;
		vhdr = skb_header_pointer(skb, ETH_HLEN, sizeof(_vhdr), &_vhdr);
		if (!vhdr)
			goto out_drop;

7262 7263
		tx_flags |= ntohs(vhdr->h_vlan_TCI) <<
				  IXGBE_TX_FLAGS_VLAN_SHIFT;
7264 7265
		tx_flags |= IXGBE_TX_FLAGS_SW_VLAN;
	}
7266
	protocol = vlan_get_protocol(skb);
7267

7268 7269 7270 7271
	if (unlikely(skb_shinfo(skb)->tx_flags & SKBTX_HW_TSTAMP) &&
	    adapter->ptp_clock &&
	    !test_and_set_bit_lock(__IXGBE_PTP_TX_IN_PROGRESS,
				   &adapter->state)) {
7272 7273
		skb_shinfo(skb)->tx_flags |= SKBTX_IN_PROGRESS;
		tx_flags |= IXGBE_TX_FLAGS_TSTAMP;
7274 7275 7276 7277 7278

		/* schedule check for Tx timestamp */
		adapter->ptp_tx_skb = skb_get(skb);
		adapter->ptp_tx_start = jiffies;
		schedule_work(&adapter->ptp_tx_work);
7279 7280
	}

7281 7282
	skb_tx_timestamp(skb);

7283 7284 7285 7286 7287 7288
#ifdef CONFIG_PCI_IOV
	/*
	 * Use the l2switch_enable flag - would be false if the DMA
	 * Tx switch had been disabled.
	 */
	if (adapter->flags & IXGBE_FLAG_SRIOV_ENABLED)
7289
		tx_flags |= IXGBE_TX_FLAGS_CC;
7290 7291

#endif
7292
	/* DCB maps skb priorities 0-7 onto 3 bit PCP of VLAN tag. */
7293
	if ((adapter->flags & IXGBE_FLAG_DCB_ENABLED) &&
7294 7295
	    ((tx_flags & (IXGBE_TX_FLAGS_HW_VLAN | IXGBE_TX_FLAGS_SW_VLAN)) ||
	     (skb->priority != TC_PRIO_CONTROL))) {
7296
		tx_flags &= ~IXGBE_TX_FLAGS_VLAN_PRIO_MASK;
7297 7298
		tx_flags |= (skb->priority & 0x7) <<
					IXGBE_TX_FLAGS_VLAN_PRIO_SHIFT;
7299 7300
		if (tx_flags & IXGBE_TX_FLAGS_SW_VLAN) {
			struct vlan_ethhdr *vhdr;
7301 7302

			if (skb_cow_head(skb, 0))
7303 7304 7305 7306 7307 7308
				goto out_drop;
			vhdr = (struct vlan_ethhdr *)skb->data;
			vhdr->h_vlan_TCI = htons(tx_flags >>
						 IXGBE_TX_FLAGS_VLAN_SHIFT);
		} else {
			tx_flags |= IXGBE_TX_FLAGS_HW_VLAN;
7309
		}
7310
	}
7311

7312 7313 7314 7315
	/* record initial flags and protocol */
	first->tx_flags = tx_flags;
	first->protocol = protocol;

7316
#ifdef IXGBE_FCOE
7317
	/* setup tx offload for FCoE */
7318
	if ((protocol == htons(ETH_P_FCOE)) &&
7319
	    (tx_ring->netdev->features & (NETIF_F_FSO | NETIF_F_FCOE_CRC))) {
7320
		tso = ixgbe_fso(tx_ring, first, &hdr_len);
7321 7322
		if (tso < 0)
			goto out_drop;
7323

7324
		goto xmit_fcoe;
7325
	}
7326

7327
#endif /* IXGBE_FCOE */
7328
	tso = ixgbe_tso(tx_ring, first, &hdr_len);
7329
	if (tso < 0)
7330
		goto out_drop;
7331 7332
	else if (!tso)
		ixgbe_tx_csum(tx_ring, first);
7333 7334 7335

	/* add the ATR filter if ATR is on */
	if (test_bit(__IXGBE_TX_FDIR_INIT_DONE, &tx_ring->state))
7336
		ixgbe_atr(tx_ring, first);
7337 7338 7339 7340

#ifdef IXGBE_FCOE
xmit_fcoe:
#endif /* IXGBE_FCOE */
7341
	ixgbe_tx_map(tx_ring, first, hdr_len);
7342

7343
	return NETDEV_TX_OK;
7344 7345

out_drop:
7346 7347 7348
	dev_kfree_skb_any(first->skb);
	first->skb = NULL;

7349
	return NETDEV_TX_OK;
7350 7351
}

7352 7353 7354
static netdev_tx_t __ixgbe_xmit_frame(struct sk_buff *skb,
				      struct net_device *netdev,
				      struct ixgbe_ring *ring)
7355 7356 7357 7358
{
	struct ixgbe_adapter *adapter = netdev_priv(netdev);
	struct ixgbe_ring *tx_ring;

7359 7360 7361 7362
	/*
	 * The minimum packet size for olinfo paylen is 17 so pad the skb
	 * in order to meet this minimum size requirement.
	 */
7363 7364
	if (skb_put_padto(skb, 17))
		return NETDEV_TX_OK;
7365

7366 7367
	tx_ring = ring ? ring : adapter->tx_ring[skb->queue_mapping];

7368
	return ixgbe_xmit_frame_ring(skb, adapter, tx_ring);
7369 7370
}

7371 7372 7373 7374 7375 7376
static netdev_tx_t ixgbe_xmit_frame(struct sk_buff *skb,
				    struct net_device *netdev)
{
	return __ixgbe_xmit_frame(skb, netdev, NULL);
}

7377 7378 7379 7380 7381 7382 7383 7384 7385 7386
/**
 * 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);
7387
	struct ixgbe_hw *hw = &adapter->hw;
7388
	struct sockaddr *addr = p;
7389
	int ret;
7390 7391 7392 7393

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

7394
	ixgbe_del_mac_filter(adapter, hw->mac.addr, VMDQ_P(0));
7395
	memcpy(netdev->dev_addr, addr->sa_data, netdev->addr_len);
7396
	memcpy(hw->mac.addr, addr->sa_data, netdev->addr_len);
7397

7398 7399
	ret = ixgbe_add_mac_filter(adapter, hw->mac.addr, VMDQ_P(0));
	return ret > 0 ? 0 : ret;
7400 7401
}

7402 7403 7404 7405 7406 7407 7408 7409 7410 7411 7412 7413 7414 7415 7416 7417 7418 7419 7420 7421 7422 7423 7424 7425 7426 7427 7428 7429 7430 7431 7432
static int
ixgbe_mdio_read(struct net_device *netdev, int prtad, int devad, u16 addr)
{
	struct ixgbe_adapter *adapter = netdev_priv(netdev);
	struct ixgbe_hw *hw = &adapter->hw;
	u16 value;
	int rc;

	if (prtad != hw->phy.mdio.prtad)
		return -EINVAL;
	rc = hw->phy.ops.read_reg(hw, addr, devad, &value);
	if (!rc)
		rc = value;
	return rc;
}

static int ixgbe_mdio_write(struct net_device *netdev, int prtad, int devad,
			    u16 addr, u16 value)
{
	struct ixgbe_adapter *adapter = netdev_priv(netdev);
	struct ixgbe_hw *hw = &adapter->hw;

	if (prtad != hw->phy.mdio.prtad)
		return -EINVAL;
	return hw->phy.ops.write_reg(hw, addr, devad, value);
}

static int ixgbe_ioctl(struct net_device *netdev, struct ifreq *req, int cmd)
{
	struct ixgbe_adapter *adapter = netdev_priv(netdev);

7433 7434
	switch (cmd) {
	case SIOCSHWTSTAMP:
7435 7436 7437
		return ixgbe_ptp_set_ts_config(adapter, req);
	case SIOCGHWTSTAMP:
		return ixgbe_ptp_get_ts_config(adapter, req);
7438 7439 7440
	default:
		return mdio_mii_ioctl(&adapter->hw.phy.mdio, if_mii(req), cmd);
	}
7441 7442
}

7443 7444
/**
 * ixgbe_add_sanmac_netdev - Add the SAN MAC address to the corresponding
7445
 * netdev->dev_addrs
7446 7447 7448 7449 7450 7451 7452 7453
 * @netdev: network interface device structure
 *
 * Returns non-zero on failure
 **/
static int ixgbe_add_sanmac_netdev(struct net_device *dev)
{
	int err = 0;
	struct ixgbe_adapter *adapter = netdev_priv(dev);
7454
	struct ixgbe_hw *hw = &adapter->hw;
7455

7456
	if (is_valid_ether_addr(hw->mac.san_addr)) {
7457
		rtnl_lock();
7458
		err = dev_addr_add(dev, hw->mac.san_addr, NETDEV_HW_ADDR_T_SAN);
7459
		rtnl_unlock();
7460 7461 7462

		/* update SAN MAC vmdq pool selection */
		hw->mac.ops.set_vmdq_san_mac(hw, VMDQ_P(0));
7463 7464 7465 7466 7467 7468
	}
	return err;
}

/**
 * ixgbe_del_sanmac_netdev - Removes the SAN MAC address to the corresponding
7469
 * netdev->dev_addrs
7470 7471 7472 7473 7474 7475 7476 7477 7478 7479 7480 7481 7482 7483 7484 7485 7486 7487
 * @netdev: network interface device structure
 *
 * Returns non-zero on failure
 **/
static int ixgbe_del_sanmac_netdev(struct net_device *dev)
{
	int err = 0;
	struct ixgbe_adapter *adapter = netdev_priv(dev);
	struct ixgbe_mac_info *mac = &adapter->hw.mac;

	if (is_valid_ether_addr(mac->san_addr)) {
		rtnl_lock();
		err = dev_addr_del(dev, mac->san_addr, NETDEV_HW_ADDR_T_SAN);
		rtnl_unlock();
	}
	return err;
}

7488 7489 7490 7491 7492 7493 7494 7495 7496
#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);
7497
	int i;
7498

7499 7500 7501 7502
	/* if interface is down do nothing */
	if (test_bit(__IXGBE_DOWN, &adapter->state))
		return;

7503 7504 7505
	/* loop through and schedule all active queues */
	for (i = 0; i < adapter->num_q_vectors; i++)
		ixgbe_msix_clean_rings(0, adapter->q_vector[i]);
7506 7507
}

A
Alexander Duyck 已提交
7508
#endif
E
Eric Dumazet 已提交
7509 7510 7511 7512 7513 7514
static struct rtnl_link_stats64 *ixgbe_get_stats64(struct net_device *netdev,
						   struct rtnl_link_stats64 *stats)
{
	struct ixgbe_adapter *adapter = netdev_priv(netdev);
	int i;

E
Eric Dumazet 已提交
7515
	rcu_read_lock();
E
Eric Dumazet 已提交
7516
	for (i = 0; i < adapter->num_rx_queues; i++) {
E
Eric Dumazet 已提交
7517
		struct ixgbe_ring *ring = ACCESS_ONCE(adapter->rx_ring[i]);
E
Eric Dumazet 已提交
7518 7519 7520
		u64 bytes, packets;
		unsigned int start;

E
Eric Dumazet 已提交
7521 7522
		if (ring) {
			do {
7523
				start = u64_stats_fetch_begin_irq(&ring->syncp);
E
Eric Dumazet 已提交
7524 7525
				packets = ring->stats.packets;
				bytes   = ring->stats.bytes;
7526
			} while (u64_stats_fetch_retry_irq(&ring->syncp, start));
E
Eric Dumazet 已提交
7527 7528 7529
			stats->rx_packets += packets;
			stats->rx_bytes   += bytes;
		}
E
Eric Dumazet 已提交
7530
	}
E
Eric Dumazet 已提交
7531 7532 7533 7534 7535 7536 7537 7538

	for (i = 0; i < adapter->num_tx_queues; i++) {
		struct ixgbe_ring *ring = ACCESS_ONCE(adapter->tx_ring[i]);
		u64 bytes, packets;
		unsigned int start;

		if (ring) {
			do {
7539
				start = u64_stats_fetch_begin_irq(&ring->syncp);
E
Eric Dumazet 已提交
7540 7541
				packets = ring->stats.packets;
				bytes   = ring->stats.bytes;
7542
			} while (u64_stats_fetch_retry_irq(&ring->syncp, start));
E
Eric Dumazet 已提交
7543 7544 7545 7546
			stats->tx_packets += packets;
			stats->tx_bytes   += bytes;
		}
	}
E
Eric Dumazet 已提交
7547
	rcu_read_unlock();
E
Eric Dumazet 已提交
7548 7549 7550 7551 7552 7553 7554 7555 7556
	/* following stats updated by ixgbe_watchdog_task() */
	stats->multicast	= netdev->stats.multicast;
	stats->rx_errors	= netdev->stats.rx_errors;
	stats->rx_length_errors	= netdev->stats.rx_length_errors;
	stats->rx_crc_errors	= netdev->stats.rx_crc_errors;
	stats->rx_missed_errors	= netdev->stats.rx_missed_errors;
	return stats;
}

7557
#ifdef CONFIG_IXGBE_DCB
7558 7559 7560
/**
 * ixgbe_validate_rtr - verify 802.1Qp to Rx packet buffer mapping is valid.
 * @adapter: pointer to ixgbe_adapter
7561 7562 7563 7564 7565 7566 7567 7568 7569 7570 7571 7572 7573 7574 7575 7576 7577 7578 7579 7580 7581 7582 7583 7584 7585 7586 7587 7588 7589 7590 7591 7592 7593 7594
 * @tc: number of traffic classes currently enabled
 *
 * Configure a valid 802.1Qp to Rx packet buffer mapping ie confirm
 * 802.1Q priority maps to a packet buffer that exists.
 */
static void ixgbe_validate_rtr(struct ixgbe_adapter *adapter, u8 tc)
{
	struct ixgbe_hw *hw = &adapter->hw;
	u32 reg, rsave;
	int i;

	/* 82598 have a static priority to TC mapping that can not
	 * be changed so no validation is needed.
	 */
	if (hw->mac.type == ixgbe_mac_82598EB)
		return;

	reg = IXGBE_READ_REG(hw, IXGBE_RTRUP2TC);
	rsave = reg;

	for (i = 0; i < MAX_TRAFFIC_CLASS; i++) {
		u8 up2tc = reg >> (i * IXGBE_RTRUP2TC_UP_SHIFT);

		/* If up2tc is out of bounds default to zero */
		if (up2tc > tc)
			reg &= ~(0x7 << IXGBE_RTRUP2TC_UP_SHIFT);
	}

	if (reg != rsave)
		IXGBE_WRITE_REG(hw, IXGBE_RTRUP2TC, reg);

	return;
}

7595 7596 7597 7598 7599 7600 7601 7602 7603 7604 7605 7606 7607 7608 7609 7610 7611 7612 7613 7614 7615 7616 7617 7618 7619
/**
 * ixgbe_set_prio_tc_map - Configure netdev prio tc map
 * @adapter: Pointer to adapter struct
 *
 * Populate the netdev user priority to tc map
 */
static void ixgbe_set_prio_tc_map(struct ixgbe_adapter *adapter)
{
	struct net_device *dev = adapter->netdev;
	struct ixgbe_dcb_config *dcb_cfg = &adapter->dcb_cfg;
	struct ieee_ets *ets = adapter->ixgbe_ieee_ets;
	u8 prio;

	for (prio = 0; prio < MAX_USER_PRIORITY; prio++) {
		u8 tc = 0;

		if (adapter->dcbx_cap & DCB_CAP_DCBX_VER_CEE)
			tc = ixgbe_dcb_get_tc_from_up(dcb_cfg, 0, prio);
		else if (ets)
			tc = ets->prio_tc[prio];

		netdev_set_prio_tc_map(dev, prio, tc);
	}
}

7620
#endif /* CONFIG_IXGBE_DCB */
7621 7622
/**
 * ixgbe_setup_tc - configure net_device for multiple traffic classes
7623 7624 7625 7626 7627 7628 7629 7630
 *
 * @netdev: net device to configure
 * @tc: number of traffic classes to enable
 */
int ixgbe_setup_tc(struct net_device *dev, u8 tc)
{
	struct ixgbe_adapter *adapter = netdev_priv(dev);
	struct ixgbe_hw *hw = &adapter->hw;
7631
	bool pools;
7632 7633

	/* Hardware supports up to 8 traffic classes */
7634
	if (tc > adapter->dcb_cfg.num_tcs.pg_tcs ||
A
Alexander Duyck 已提交
7635 7636
	    (hw->mac.type == ixgbe_mac_82598EB &&
	     tc < MAX_TRAFFIC_CLASS))
7637 7638
		return -EINVAL;

7639 7640 7641 7642
	pools = (find_first_zero_bit(&adapter->fwd_bitmask, 32) > 1);
	if (tc && pools && adapter->num_rx_pools > IXGBE_MAX_DCBMACVLANS)
		return -EBUSY;

7643
	/* Hardware has to reinitialize queues and interrupts to
S
Stephen Hemminger 已提交
7644
	 * match packet buffer alignment. Unfortunately, the
7645 7646 7647 7648 7649 7650
	 * hardware is not flexible enough to do this dynamically.
	 */
	if (netif_running(dev))
		ixgbe_close(dev);
	ixgbe_clear_interrupt_scheme(adapter);

7651
#ifdef CONFIG_IXGBE_DCB
7652
	if (tc) {
7653
		netdev_set_num_tc(dev, tc);
7654 7655
		ixgbe_set_prio_tc_map(adapter);

7656 7657
		adapter->flags |= IXGBE_FLAG_DCB_ENABLED;

7658 7659
		if (adapter->hw.mac.type == ixgbe_mac_82598EB) {
			adapter->last_lfc_mode = adapter->hw.fc.requested_mode;
7660
			adapter->hw.fc.requested_mode = ixgbe_fc_none;
7661
		}
7662
	} else {
7663
		netdev_reset_tc(dev);
7664

7665 7666
		if (adapter->hw.mac.type == ixgbe_mac_82598EB)
			adapter->hw.fc.requested_mode = adapter->last_lfc_mode;
7667 7668 7669 7670 7671 7672 7673

		adapter->flags &= ~IXGBE_FLAG_DCB_ENABLED;

		adapter->temp_dcb_cfg.pfc_mode_enable = false;
		adapter->dcb_cfg.pfc_mode_enable = false;
	}

7674
	ixgbe_validate_rtr(adapter, tc);
7675 7676 7677 7678

#endif /* CONFIG_IXGBE_DCB */
	ixgbe_init_interrupt_scheme(adapter);

7679
	if (netif_running(dev))
7680
		return ixgbe_open(dev);
7681 7682 7683

	return 0;
}
E
Eric Dumazet 已提交
7684

7685 7686 7687 7688 7689 7690 7691 7692 7693 7694 7695
#ifdef CONFIG_PCI_IOV
void ixgbe_sriov_reinit(struct ixgbe_adapter *adapter)
{
	struct net_device *netdev = adapter->netdev;

	rtnl_lock();
	ixgbe_setup_tc(netdev, netdev_get_num_tc(netdev));
	rtnl_unlock();
}

#endif
7696 7697 7698 7699 7700 7701 7702 7703 7704 7705
void ixgbe_do_reset(struct net_device *netdev)
{
	struct ixgbe_adapter *adapter = netdev_priv(netdev);

	if (netif_running(netdev))
		ixgbe_reinit_locked(adapter);
	else
		ixgbe_reset(adapter);
}

7706
static netdev_features_t ixgbe_fix_features(struct net_device *netdev,
7707
					    netdev_features_t features)
7708 7709 7710 7711
{
	struct ixgbe_adapter *adapter = netdev_priv(netdev);

	/* If Rx checksum is disabled, then RSC/LRO should also be disabled */
7712 7713
	if (!(features & NETIF_F_RXCSUM))
		features &= ~NETIF_F_LRO;
7714

7715 7716 7717
	/* Turn off LRO if not RSC capable */
	if (!(adapter->flags2 & IXGBE_FLAG2_RSC_CAPABLE))
		features &= ~NETIF_F_LRO;
7718

7719
	return features;
7720 7721
}

7722
static int ixgbe_set_features(struct net_device *netdev,
7723
			      netdev_features_t features)
7724 7725
{
	struct ixgbe_adapter *adapter = netdev_priv(netdev);
7726
	netdev_features_t changed = netdev->features ^ features;
7727 7728 7729
	bool need_reset = false;

	/* Make sure RSC matches LRO, reset if change */
7730 7731
	if (!(features & NETIF_F_LRO)) {
		if (adapter->flags2 & IXGBE_FLAG2_RSC_ENABLED)
7732
			need_reset = true;
7733 7734 7735 7736 7737 7738 7739 7740 7741 7742
		adapter->flags2 &= ~IXGBE_FLAG2_RSC_ENABLED;
	} else if ((adapter->flags2 & IXGBE_FLAG2_RSC_CAPABLE) &&
		   !(adapter->flags2 & IXGBE_FLAG2_RSC_ENABLED)) {
		if (adapter->rx_itr_setting == 1 ||
		    adapter->rx_itr_setting > IXGBE_MIN_RSC_ITR) {
			adapter->flags2 |= IXGBE_FLAG2_RSC_ENABLED;
			need_reset = true;
		} else if ((changed ^ features) & NETIF_F_LRO) {
			e_info(probe, "rx-usecs set too low, "
			       "disabling RSC\n");
7743 7744 7745 7746 7747 7748 7749
		}
	}

	/*
	 * Check if Flow Director n-tuple support was enabled or disabled.  If
	 * the state changed, we need to reset.
	 */
7750 7751
	switch (features & NETIF_F_NTUPLE) {
	case NETIF_F_NTUPLE:
7752
		/* turn off ATR, enable perfect filters and reset */
7753 7754 7755
		if (!(adapter->flags & IXGBE_FLAG_FDIR_PERFECT_CAPABLE))
			need_reset = true;

7756 7757
		adapter->flags &= ~IXGBE_FLAG_FDIR_HASH_CAPABLE;
		adapter->flags |= IXGBE_FLAG_FDIR_PERFECT_CAPABLE;
7758 7759 7760 7761 7762 7763 7764 7765 7766 7767 7768 7769 7770 7771 7772 7773 7774 7775 7776 7777 7778 7779 7780 7781 7782 7783
		break;
	default:
		/* turn off perfect filters, enable ATR and reset */
		if (adapter->flags & IXGBE_FLAG_FDIR_PERFECT_CAPABLE)
			need_reset = true;

		adapter->flags &= ~IXGBE_FLAG_FDIR_PERFECT_CAPABLE;

		/* We cannot enable ATR if SR-IOV is enabled */
		if (adapter->flags & IXGBE_FLAG_SRIOV_ENABLED)
			break;

		/* We cannot enable ATR if we have 2 or more traffic classes */
		if (netdev_get_num_tc(netdev) > 1)
			break;

		/* We cannot enable ATR if RSS is disabled */
		if (adapter->ring_feature[RING_F_RSS].limit <= 1)
			break;

		/* A sample rate of 0 indicates ATR disabled */
		if (!adapter->atr_sample_rate)
			break;

		adapter->flags |= IXGBE_FLAG_FDIR_HASH_CAPABLE;
		break;
7784 7785
	}

7786
	if (features & NETIF_F_HW_VLAN_CTAG_RX)
7787 7788 7789 7790
		ixgbe_vlan_strip_enable(adapter);
	else
		ixgbe_vlan_strip_disable(adapter);

B
Ben Greear 已提交
7791 7792 7793
	if (changed & NETIF_F_RXALL)
		need_reset = true;

7794
	netdev->features = features;
7795 7796 7797 7798 7799 7800
	if (need_reset)
		ixgbe_do_reset(netdev);

	return 0;
}

7801 7802 7803 7804 7805 7806 7807 7808 7809 7810 7811 7812 7813 7814 7815 7816 7817 7818 7819 7820 7821 7822 7823 7824 7825 7826 7827 7828 7829 7830 7831 7832 7833 7834 7835 7836 7837 7838 7839 7840 7841 7842 7843 7844 7845 7846 7847 7848 7849 7850 7851 7852 7853 7854 7855 7856 7857 7858
/**
 * ixgbe_add_vxlan_port - Get notifications about VXLAN ports that come up
 * @dev: The port's netdev
 * @sa_family: Socket Family that VXLAN is notifiying us about
 * @port: New UDP port number that VXLAN started listening to
 **/
static void ixgbe_add_vxlan_port(struct net_device *dev, sa_family_t sa_family,
				 __be16 port)
{
	struct ixgbe_adapter *adapter = netdev_priv(dev);
	struct ixgbe_hw *hw = &adapter->hw;
	u16 new_port = ntohs(port);

	if (sa_family == AF_INET6)
		return;

	if (adapter->vxlan_port == new_port) {
		netdev_info(dev, "Port %d already offloaded\n", new_port);
		return;
	}

	if (adapter->vxlan_port) {
		netdev_info(dev,
			    "Hit Max num of UDP ports, not adding port %d\n",
			    new_port);
		return;
	}

	adapter->vxlan_port = new_port;
	IXGBE_WRITE_REG(hw, IXGBE_VXLANCTRL, new_port);
}

/**
 * ixgbe_del_vxlan_port - Get notifications about VXLAN ports that go away
 * @dev: The port's netdev
 * @sa_family: Socket Family that VXLAN is notifying us about
 * @port: UDP port number that VXLAN stopped listening to
 **/
static void ixgbe_del_vxlan_port(struct net_device *dev, sa_family_t sa_family,
				 __be16 port)
{
	struct ixgbe_adapter *adapter = netdev_priv(dev);
	struct ixgbe_hw *hw = &adapter->hw;
	u16 new_port = ntohs(port);

	if (sa_family == AF_INET6)
		return;

	if (adapter->vxlan_port != new_port) {
		netdev_info(dev, "Port %d was not found, not deleting\n",
			    new_port);
		return;
	}

	adapter->vxlan_port = 0;
	IXGBE_WRITE_REG(hw, IXGBE_VXLANCTRL, 0);
}

7859
static int ixgbe_ndo_fdb_add(struct ndmsg *ndm, struct nlattr *tb[],
J
John Fastabend 已提交
7860
			     struct net_device *dev,
7861
			     const unsigned char *addr, u16 vid,
J
John Fastabend 已提交
7862 7863
			     u16 flags)
{
7864
	/* guarantee we can provide a unique filter for the unicast address */
7865
	if (is_unicast_ether_addr(addr) || is_link_local_ether_addr(addr)) {
7866 7867
		if (IXGBE_MAX_PF_MACVLANS <= netdev_uc_count(dev))
			return -ENOMEM;
J
John Fastabend 已提交
7868 7869
	}

7870
	return ndo_dflt_fdb_add(ndm, tb, dev, addr, vid, flags);
J
John Fastabend 已提交
7871 7872
}

7873 7874 7875 7876 7877 7878 7879 7880 7881 7882
/**
 * ixgbe_configure_bridge_mode - set various bridge modes
 * @adapter - the private structure
 * @mode - requested bridge mode
 *
 * Configure some settings require for various bridge modes.
 **/
static int ixgbe_configure_bridge_mode(struct ixgbe_adapter *adapter,
				       __u16 mode)
{
7883 7884 7885 7886
	struct ixgbe_hw *hw = &adapter->hw;
	unsigned int p, num_pools;
	u32 vmdctl;

7887 7888
	switch (mode) {
	case BRIDGE_MODE_VEPA:
7889
		/* disable Tx loopback, rely on switch hairpin mode */
7890
		IXGBE_WRITE_REG(&adapter->hw, IXGBE_PFDTXGSWC, 0);
7891 7892 7893 7894 7895 7896 7897 7898 7899 7900 7901 7902 7903 7904 7905 7906 7907 7908 7909

		/* must enable Rx switching replication to allow multicast
		 * packet reception on all VFs, and to enable source address
		 * pruning.
		 */
		vmdctl = IXGBE_READ_REG(hw, IXGBE_VMD_CTL);
		vmdctl |= IXGBE_VT_CTL_REPLEN;
		IXGBE_WRITE_REG(hw, IXGBE_VMD_CTL, vmdctl);

		/* enable Rx source address pruning. Note, this requires
		 * replication to be enabled or else it does nothing.
		 */
		num_pools = adapter->num_vfs + adapter->num_rx_pools;
		for (p = 0; p < num_pools; p++) {
			if (hw->mac.ops.set_source_address_pruning)
				hw->mac.ops.set_source_address_pruning(hw,
								       true,
								       p);
		}
7910 7911
		break;
	case BRIDGE_MODE_VEB:
7912
		/* enable Tx loopback for internal VF/PF communication */
7913 7914
		IXGBE_WRITE_REG(&adapter->hw, IXGBE_PFDTXGSWC,
				IXGBE_PFDTXGSWC_VT_LBEN);
7915 7916 7917 7918 7919 7920 7921 7922 7923 7924 7925 7926 7927 7928 7929 7930 7931 7932 7933

		/* disable Rx switching replication unless we have SR-IOV
		 * virtual functions
		 */
		vmdctl = IXGBE_READ_REG(hw, IXGBE_VMD_CTL);
		if (!adapter->num_vfs)
			vmdctl &= ~IXGBE_VT_CTL_REPLEN;
		IXGBE_WRITE_REG(hw, IXGBE_VMD_CTL, vmdctl);

		/* disable Rx source address pruning, since we don't expect to
		 * be receiving external loopback of our transmitted frames.
		 */
		num_pools = adapter->num_vfs + adapter->num_rx_pools;
		for (p = 0; p < num_pools; p++) {
			if (hw->mac.ops.set_source_address_pruning)
				hw->mac.ops.set_source_address_pruning(hw,
								       false,
								       p);
		}
7934 7935 7936 7937 7938 7939 7940 7941 7942 7943 7944 7945 7946
		break;
	default:
		return -EINVAL;
	}

	adapter->bridge_mode = mode;

	e_info(drv, "enabling bridge mode: %s\n",
	       mode == BRIDGE_MODE_VEPA ? "VEPA" : "VEB");

	return 0;
}

7947
static int ixgbe_ndo_bridge_setlink(struct net_device *dev,
7948
				    struct nlmsghdr *nlh, u16 flags)
7949 7950 7951 7952 7953 7954 7955 7956 7957
{
	struct ixgbe_adapter *adapter = netdev_priv(dev);
	struct nlattr *attr, *br_spec;
	int rem;

	if (!(adapter->flags & IXGBE_FLAG_SRIOV_ENABLED))
		return -EOPNOTSUPP;

	br_spec = nlmsg_find_attr(nlh, sizeof(struct ifinfomsg), IFLA_AF_SPEC);
7958 7959
	if (!br_spec)
		return -EINVAL;
7960 7961

	nla_for_each_nested(attr, br_spec, rem) {
7962
		u32 status;
7963 7964 7965 7966 7967
		__u16 mode;

		if (nla_type(attr) != IFLA_BRIDGE_MODE)
			continue;

7968 7969 7970
		if (nla_len(attr) < sizeof(mode))
			return -EINVAL;

7971
		mode = nla_get_u16(attr);
7972 7973 7974
		status = ixgbe_configure_bridge_mode(adapter, mode);
		if (status)
			return status;
7975 7976

		break;
7977 7978 7979 7980 7981 7982
	}

	return 0;
}

static int ixgbe_ndo_bridge_getlink(struct sk_buff *skb, u32 pid, u32 seq,
7983 7984
				    struct net_device *dev,
				    u32 filter_mask)
7985 7986 7987 7988 7989 7990
{
	struct ixgbe_adapter *adapter = netdev_priv(dev);

	if (!(adapter->flags & IXGBE_FLAG_SRIOV_ENABLED))
		return 0;

7991 7992
	return ndo_dflt_bridge_getlink(skb, pid, seq, dev,
				       adapter->bridge_mode, 0, 0);
7993 7994
}

7995 7996 7997 7998
static void *ixgbe_fwd_add(struct net_device *pdev, struct net_device *vdev)
{
	struct ixgbe_fwd_adapter *fwd_adapter = NULL;
	struct ixgbe_adapter *adapter = netdev_priv(pdev);
7999
	int used_pools = adapter->num_vfs + adapter->num_rx_pools;
8000
	unsigned int limit;
8001 8002
	int pool, err;

8003 8004 8005 8006 8007 8008 8009
	/* Hardware has a limited number of available pools. Each VF, and the
	 * PF require a pool. Check to ensure we don't attempt to use more
	 * then the available number of pools.
	 */
	if (used_pools >= IXGBE_MAX_VF_FUNCTIONS)
		return ERR_PTR(-EINVAL);

8010 8011 8012 8013 8014 8015 8016
#ifdef CONFIG_RPS
	if (vdev->num_rx_queues != vdev->num_tx_queues) {
		netdev_info(pdev, "%s: Only supports a single queue count for TX and RX\n",
			    vdev->name);
		return ERR_PTR(-EINVAL);
	}
#endif
8017
	/* Check for hardware restriction on number of rx/tx queues */
8018
	if (vdev->num_tx_queues > IXGBE_MAX_L2A_QUEUES ||
8019 8020 8021 8022 8023 8024 8025 8026 8027 8028 8029 8030 8031 8032 8033 8034 8035 8036 8037
	    vdev->num_tx_queues == IXGBE_BAD_L2A_QUEUE) {
		netdev_info(pdev,
			    "%s: Supports RX/TX Queue counts 1,2, and 4\n",
			    pdev->name);
		return ERR_PTR(-EINVAL);
	}

	if (((adapter->flags & IXGBE_FLAG_DCB_ENABLED) &&
	      adapter->num_rx_pools > IXGBE_MAX_DCBMACVLANS - 1) ||
	    (adapter->num_rx_pools > IXGBE_MAX_MACVLANS))
		return ERR_PTR(-EBUSY);

	fwd_adapter = kcalloc(1, sizeof(struct ixgbe_fwd_adapter), GFP_KERNEL);
	if (!fwd_adapter)
		return ERR_PTR(-ENOMEM);

	pool = find_first_zero_bit(&adapter->fwd_bitmask, 32);
	adapter->num_rx_pools++;
	set_bit(pool, &adapter->fwd_bitmask);
8038
	limit = find_last_bit(&adapter->fwd_bitmask, 32);
8039 8040 8041

	/* Enable VMDq flag so device will be set in VM mode */
	adapter->flags |= IXGBE_FLAG_VMDQ_ENABLED | IXGBE_FLAG_SRIOV_ENABLED;
8042
	adapter->ring_feature[RING_F_VMDQ].limit = limit + 1;
8043
	adapter->ring_feature[RING_F_RSS].limit = vdev->num_tx_queues;
8044 8045 8046 8047 8048 8049 8050 8051 8052 8053 8054 8055 8056 8057 8058 8059 8060 8061 8062 8063 8064 8065 8066 8067 8068 8069

	/* Force reinit of ring allocation with VMDQ enabled */
	err = ixgbe_setup_tc(pdev, netdev_get_num_tc(pdev));
	if (err)
		goto fwd_add_err;
	fwd_adapter->pool = pool;
	fwd_adapter->real_adapter = adapter;
	err = ixgbe_fwd_ring_up(vdev, fwd_adapter);
	if (err)
		goto fwd_add_err;
	netif_tx_start_all_queues(vdev);
	return fwd_adapter;
fwd_add_err:
	/* unwind counter and free adapter struct */
	netdev_info(pdev,
		    "%s: dfwd hardware acceleration failed\n", vdev->name);
	clear_bit(pool, &adapter->fwd_bitmask);
	adapter->num_rx_pools--;
	kfree(fwd_adapter);
	return ERR_PTR(err);
}

static void ixgbe_fwd_del(struct net_device *pdev, void *priv)
{
	struct ixgbe_fwd_adapter *fwd_adapter = priv;
	struct ixgbe_adapter *adapter = fwd_adapter->real_adapter;
8070
	unsigned int limit;
8071 8072 8073 8074

	clear_bit(fwd_adapter->pool, &adapter->fwd_bitmask);
	adapter->num_rx_pools--;

8075 8076
	limit = find_last_bit(&adapter->fwd_bitmask, 32);
	adapter->ring_feature[RING_F_VMDQ].limit = limit + 1;
8077 8078 8079 8080 8081 8082 8083 8084 8085 8086
	ixgbe_fwd_ring_down(fwd_adapter->netdev, fwd_adapter);
	ixgbe_setup_tc(pdev, netdev_get_num_tc(pdev));
	netdev_dbg(pdev, "pool %i:%i queues %i:%i VSI bitmask %lx\n",
		   fwd_adapter->pool, adapter->num_rx_pools,
		   fwd_adapter->rx_base_queue,
		   fwd_adapter->rx_base_queue + adapter->num_rx_queues_per_pool,
		   adapter->fwd_bitmask);
	kfree(fwd_adapter);
}

8087
static const struct net_device_ops ixgbe_netdev_ops = {
8088
	.ndo_open		= ixgbe_open,
8089
	.ndo_stop		= ixgbe_close,
8090
	.ndo_start_xmit		= ixgbe_xmit_frame,
8091
	.ndo_select_queue	= ixgbe_select_queue,
A
Alexander Duyck 已提交
8092
	.ndo_set_rx_mode	= ixgbe_set_rx_mode,
8093 8094 8095 8096 8097 8098
	.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_add_vid	= ixgbe_vlan_rx_add_vid,
	.ndo_vlan_rx_kill_vid	= ixgbe_vlan_rx_kill_vid,
8099
	.ndo_do_ioctl		= ixgbe_ioctl,
8100 8101
	.ndo_set_vf_mac		= ixgbe_ndo_set_vf_mac,
	.ndo_set_vf_vlan	= ixgbe_ndo_set_vf_vlan,
8102
	.ndo_set_vf_rate	= ixgbe_ndo_set_vf_bw,
A
Alexander Duyck 已提交
8103
	.ndo_set_vf_spoofchk	= ixgbe_ndo_set_vf_spoofchk,
8104
	.ndo_get_vf_config	= ixgbe_ndo_get_vf_config,
E
Eric Dumazet 已提交
8105
	.ndo_get_stats64	= ixgbe_get_stats64,
8106
#ifdef CONFIG_IXGBE_DCB
J
John Fastabend 已提交
8107
	.ndo_setup_tc		= ixgbe_setup_tc,
8108
#endif
8109 8110 8111
#ifdef CONFIG_NET_POLL_CONTROLLER
	.ndo_poll_controller	= ixgbe_netpoll,
#endif
8112
#ifdef CONFIG_NET_RX_BUSY_POLL
8113
	.ndo_busy_poll		= ixgbe_low_latency_recv,
8114
#endif
8115 8116
#ifdef IXGBE_FCOE
	.ndo_fcoe_ddp_setup = ixgbe_fcoe_ddp_get,
8117
	.ndo_fcoe_ddp_target = ixgbe_fcoe_ddp_target,
8118
	.ndo_fcoe_ddp_done = ixgbe_fcoe_ddp_put,
8119 8120
	.ndo_fcoe_enable = ixgbe_fcoe_enable,
	.ndo_fcoe_disable = ixgbe_fcoe_disable,
8121
	.ndo_fcoe_get_wwn = ixgbe_fcoe_get_wwn,
8122
	.ndo_fcoe_get_hbainfo = ixgbe_fcoe_get_hbainfo,
8123
#endif /* IXGBE_FCOE */
8124 8125
	.ndo_set_features = ixgbe_set_features,
	.ndo_fix_features = ixgbe_fix_features,
J
John Fastabend 已提交
8126
	.ndo_fdb_add		= ixgbe_ndo_fdb_add,
8127 8128
	.ndo_bridge_setlink	= ixgbe_ndo_bridge_setlink,
	.ndo_bridge_getlink	= ixgbe_ndo_bridge_getlink,
8129 8130
	.ndo_dfwd_add_station	= ixgbe_fwd_add,
	.ndo_dfwd_del_station	= ixgbe_fwd_del,
8131 8132
	.ndo_add_vxlan_port	= ixgbe_add_vxlan_port,
	.ndo_del_vxlan_port	= ixgbe_del_vxlan_port,
8133 8134
};

8135 8136 8137 8138 8139 8140 8141 8142 8143 8144 8145
/**
 * ixgbe_enumerate_functions - Get the number of ports this device has
 * @adapter: adapter structure
 *
 * This function enumerates the phsyical functions co-located on a single slot,
 * in order to determine how many ports a device has. This is most useful in
 * determining the required GT/s of PCIe bandwidth necessary for optimal
 * performance.
 **/
static inline int ixgbe_enumerate_functions(struct ixgbe_adapter *adapter)
{
8146
	struct pci_dev *entry, *pdev = adapter->pdev;
8147 8148
	int physfns = 0;

8149 8150 8151
	/* Some cards can not use the generic count PCIe functions method,
	 * because they are behind a parent switch, so we hardcode these with
	 * the correct number of functions.
8152
	 */
8153
	if (ixgbe_pcie_from_parent(&adapter->hw))
8154
		physfns = 4;
8155 8156 8157

	list_for_each_entry(entry, &adapter->pdev->bus->devices, bus_list) {
		/* don't count virtual functions */
8158 8159 8160 8161 8162 8163 8164 8165 8166 8167 8168 8169 8170 8171
		if (entry->is_virtfn)
			continue;

		/* When the devices on the bus don't all match our device ID,
		 * we can't reliably determine the correct number of
		 * functions. This can occur if a function has been direct
		 * attached to a virtual machine using VT-d, for example. In
		 * this case, simply return -1 to indicate this.
		 */
		if ((entry->vendor != pdev->vendor) ||
		    (entry->device != pdev->device))
			return -1;

		physfns++;
8172 8173 8174 8175 8176
	}

	return physfns;
}

8177 8178 8179 8180 8181 8182 8183 8184 8185 8186 8187 8188 8189 8190 8191 8192 8193 8194 8195 8196 8197
/**
 * ixgbe_wol_supported - Check whether device supports WoL
 * @hw: hw specific details
 * @device_id: the device ID
 * @subdev_id: the subsystem device ID
 *
 * This function is used by probe and ethtool to determine
 * which devices have WoL support
 *
 **/
int ixgbe_wol_supported(struct ixgbe_adapter *adapter, u16 device_id,
			u16 subdevice_id)
{
	struct ixgbe_hw *hw = &adapter->hw;
	u16 wol_cap = adapter->eeprom_cap & IXGBE_DEVICE_CAPS_WOL_MASK;
	int is_wol_supported = 0;

	switch (device_id) {
	case IXGBE_DEV_ID_82599_SFP:
		/* Only these subdevices could supports WOL */
		switch (subdevice_id) {
8198
		case IXGBE_SUBDEV_ID_82599_SFP_WOL0:
8199 8200 8201 8202
		case IXGBE_SUBDEV_ID_82599_560FLR:
			/* only support first port */
			if (hw->bus.func != 0)
				break;
8203
		case IXGBE_SUBDEV_ID_82599_SP_560FLR:
8204
		case IXGBE_SUBDEV_ID_82599_SFP:
8205
		case IXGBE_SUBDEV_ID_82599_RNDC:
8206
		case IXGBE_SUBDEV_ID_82599_ECNA_DP:
8207
		case IXGBE_SUBDEV_ID_82599_LOM_SFP:
8208 8209 8210 8211
			is_wol_supported = 1;
			break;
		}
		break;
8212 8213 8214 8215 8216 8217 8218 8219
	case IXGBE_DEV_ID_82599EN_SFP:
		/* Only this subdevice supports WOL */
		switch (subdevice_id) {
		case IXGBE_SUBDEV_ID_82599EN_SFP_OCP1:
			is_wol_supported = 1;
			break;
		}
		break;
8220 8221 8222 8223 8224 8225 8226 8227 8228
	case IXGBE_DEV_ID_82599_COMBO_BACKPLANE:
		/* All except this subdevice support WOL */
		if (subdevice_id != IXGBE_SUBDEV_ID_82599_KX4_KR_MEZZ)
			is_wol_supported = 1;
		break;
	case IXGBE_DEV_ID_82599_KX4:
		is_wol_supported = 1;
		break;
	case IXGBE_DEV_ID_X540T:
8229
	case IXGBE_DEV_ID_X540T1:
8230 8231 8232 8233 8234 8235 8236 8237 8238 8239 8240 8241
		/* check eeprom to see if enabled wol */
		if ((wol_cap == IXGBE_DEVICE_CAPS_WOL_PORT0_1) ||
		    ((wol_cap == IXGBE_DEVICE_CAPS_WOL_PORT0) &&
		     (hw->bus.func == 0))) {
			is_wol_supported = 1;
		}
		break;
	}

	return is_wol_supported;
}

8242 8243 8244 8245 8246 8247 8248 8249 8250 8251 8252 8253 8254 8255 8256 8257 8258 8259 8260 8261 8262 8263 8264
/**
 * ixgbe_get_platform_mac_addr - Look up MAC address in Open Firmware / IDPROM
 * @adapter: Pointer to adapter struct
 */
static void ixgbe_get_platform_mac_addr(struct ixgbe_adapter *adapter)
{
#ifdef CONFIG_OF
	struct device_node *dp = pci_device_to_OF_node(adapter->pdev);
	struct ixgbe_hw *hw = &adapter->hw;
	const unsigned char *addr;

	addr = of_get_mac_address(dp);
	if (addr) {
		ether_addr_copy(hw->mac.perm_addr, addr);
		return;
	}
#endif /* CONFIG_OF */

#ifdef CONFIG_SPARC
	ether_addr_copy(hw->mac.perm_addr, idprom->id_ethaddr);
#endif /* CONFIG_SPARC */
}

8265 8266 8267 8268 8269 8270 8271 8272 8273 8274 8275
/**
 * 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.
 **/
8276
static int ixgbe_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
8277 8278 8279 8280 8281
{
	struct net_device *netdev;
	struct ixgbe_adapter *adapter = NULL;
	struct ixgbe_hw *hw;
	const struct ixgbe_info *ii = ixgbe_info_tbl[ent->driver_data];
8282
	int i, err, pci_using_dac, expected_gts;
8283
	unsigned int indices = MAX_TX_QUEUES;
8284
	u8 part_str[IXGBE_PBANUM_LENGTH];
8285
	bool disable_dev = false;
8286 8287 8288
#ifdef IXGBE_FCOE
	u16 device_caps;
#endif
8289
	u32 eec;
8290

8291 8292 8293 8294 8295 8296 8297 8298 8299
	/* Catch broken hardware that put the wrong VF device ID in
	 * the PCIe SR-IOV capability.
	 */
	if (pdev->is_virtfn) {
		WARN(1, KERN_ERR "%s (%hx:%hx) should not be a VF!\n",
		     pci_name(pdev), pdev->vendor, pdev->device);
		return -EINVAL;
	}

8300
	err = pci_enable_device_mem(pdev);
8301 8302 8303
	if (err)
		return err;

8304
	if (!dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(64))) {
8305 8306
		pci_using_dac = 1;
	} else {
8307
		err = dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(32));
8308
		if (err) {
8309 8310 8311
			dev_err(&pdev->dev,
				"No usable DMA configuration, aborting\n");
			goto err_dma;
8312 8313 8314 8315
		}
		pci_using_dac = 0;
	}

8316
	err = pci_request_selected_regions(pdev, pci_select_bars(pdev,
8317
					   IORESOURCE_MEM), ixgbe_driver_name);
8318
	if (err) {
8319 8320
		dev_err(&pdev->dev,
			"pci_request_selected_regions failed 0x%x\n", err);
8321 8322 8323
		goto err_pci_reg;
	}

8324
	pci_enable_pcie_error_reporting(pdev);
8325

8326
	pci_set_master(pdev);
8327
	pci_save_state(pdev);
8328

8329
	if (ii->mac == ixgbe_mac_82598EB) {
8330
#ifdef CONFIG_IXGBE_DCB
8331 8332 8333 8334
		/* 8 TC w/ 4 queues per TC */
		indices = 4 * MAX_TRAFFIC_CLASS;
#else
		indices = IXGBE_MAX_RSS_INDICES;
8335
#endif
8336
	}
8337

8338
	netdev = alloc_etherdev_mq(sizeof(struct ixgbe_adapter), indices);
8339 8340 8341 8342 8343 8344 8345 8346 8347 8348 8349 8350 8351
	if (!netdev) {
		err = -ENOMEM;
		goto err_alloc_etherdev;
	}

	SET_NETDEV_DEV(netdev, &pdev->dev);

	adapter = netdev_priv(netdev);

	adapter->netdev = netdev;
	adapter->pdev = pdev;
	hw = &adapter->hw;
	hw->back = adapter;
8352
	adapter->msg_enable = netif_msg_init(debug, DEFAULT_MSG_ENABLE);
8353

8354
	hw->hw_addr = ioremap(pci_resource_start(pdev, 0),
8355
			      pci_resource_len(pdev, 0));
8356
	adapter->io_addr = hw->hw_addr;
8357 8358 8359 8360 8361
	if (!hw->hw_addr) {
		err = -EIO;
		goto err_ioremap;
	}

8362
	netdev->netdev_ops = &ixgbe_netdev_ops;
8363 8364
	ixgbe_set_ethtool_ops(netdev);
	netdev->watchdog_timeo = 5 * HZ;
8365
	strlcpy(netdev->name, pci_name(pdev), sizeof(netdev->name));
8366 8367 8368

	/* Setup hw api */
	memcpy(&hw->mac.ops, ii->mac_ops, sizeof(hw->mac.ops));
8369
	hw->mac.type  = ii->mac;
8370

8371 8372 8373
	/* EEPROM */
	memcpy(&hw->eeprom.ops, ii->eeprom_ops, sizeof(hw->eeprom.ops));
	eec = IXGBE_READ_REG(hw, IXGBE_EEC);
8374 8375 8376 8377
	if (ixgbe_removed(hw->hw_addr)) {
		err = -EIO;
		goto err_ioremap;
	}
8378 8379 8380 8381 8382 8383
	/* 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
8384
	hw->phy.sfp_type = ixgbe_sfp_type_unknown;
8385 8386 8387 8388 8389 8390 8391
	/* ixgbe_identify_phy_generic will set prtad and mmds properly */
	hw->phy.mdio.prtad = MDIO_PRTAD_NONE;
	hw->phy.mdio.mmds = 0;
	hw->phy.mdio.mode_support = MDIO_SUPPORTS_C45 | MDIO_EMULATE_C22;
	hw->phy.mdio.dev = netdev;
	hw->phy.mdio.mdio_read = ixgbe_mdio_read;
	hw->phy.mdio.mdio_write = ixgbe_mdio_write;
D
Donald Skidmore 已提交
8392

8393
	ii->get_invariants(hw);
8394 8395 8396 8397 8398 8399

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

8400
	/* Make it possible the adapter to be woken up via WOL */
D
Don Skidmore 已提交
8401 8402 8403
	switch (adapter->hw.mac.type) {
	case ixgbe_mac_82599EB:
	case ixgbe_mac_X540:
8404 8405
	case ixgbe_mac_X550:
	case ixgbe_mac_X550EM_x:
8406
		IXGBE_WRITE_REG(&adapter->hw, IXGBE_WUS, ~0);
D
Don Skidmore 已提交
8407 8408 8409 8410
		break;
	default:
		break;
	}
8411

8412 8413 8414 8415 8416 8417 8418
	/*
	 * If there is a fan on this device and it has failed log the
	 * failure.
	 */
	if (adapter->flags & IXGBE_FLAG_FAN_FAIL_CAPABLE) {
		u32 esdp = IXGBE_READ_REG(hw, IXGBE_ESDP);
		if (esdp & IXGBE_ESDP_SDP1)
8419
			e_crit(probe, "Fan has stopped, replace the adapter\n");
8420 8421
	}

8422 8423 8424
	if (allow_unsupported_sfp)
		hw->allow_unsupported_sfp = allow_unsupported_sfp;

8425
	/* reset_hw fills in the perm_addr as well */
8426
	hw->phy.reset_if_overtemp = true;
8427
	err = hw->mac.ops.reset_hw(hw);
8428
	hw->phy.reset_if_overtemp = false;
8429 8430 8431 8432
	if (err == IXGBE_ERR_SFP_NOT_PRESENT &&
	    hw->mac.type == ixgbe_mac_82598EB) {
		err = 0;
	} else if (err == IXGBE_ERR_SFP_NOT_SUPPORTED) {
D
Don Skidmore 已提交
8433 8434
		e_dev_err("failed to load because an unsupported SFP+ or QSFP module type was detected.\n");
		e_dev_err("Reload the driver after installing a supported module.\n");
8435 8436
		goto err_sw_init;
	} else if (err) {
8437
		e_dev_err("HW Init failed: %d\n", err);
8438 8439 8440
		goto err_sw_init;
	}

8441
#ifdef CONFIG_PCI_IOV
8442 8443 8444 8445 8446 8447
	/* SR-IOV not supported on the 82598 */
	if (adapter->hw.mac.type == ixgbe_mac_82598EB)
		goto skip_sriov;
	/* Mailbox */
	ixgbe_init_mbx_params_pf(hw);
	memcpy(&hw->mbx.ops, ii->mbx_ops, sizeof(hw->mbx.ops));
8448
	pci_sriov_set_totalvfs(pdev, IXGBE_MAX_VFS_DRV_LIMIT);
8449
	ixgbe_enable_sriov(adapter);
8450
skip_sriov:
8451

8452
#endif
8453
	netdev->features = NETIF_F_SG |
8454
			   NETIF_F_IP_CSUM |
8455
			   NETIF_F_IPV6_CSUM |
8456 8457
			   NETIF_F_HW_VLAN_CTAG_TX |
			   NETIF_F_HW_VLAN_CTAG_RX |
8458 8459 8460
			   NETIF_F_TSO |
			   NETIF_F_TSO6 |
			   NETIF_F_RXHASH |
8461
			   NETIF_F_RXCSUM;
8462

8463
	netdev->hw_features = netdev->features | NETIF_F_HW_L2FW_DOFFLOAD;
8464

8465 8466 8467
	switch (adapter->hw.mac.type) {
	case ixgbe_mac_82599EB:
	case ixgbe_mac_X540:
8468 8469
	case ixgbe_mac_X550:
	case ixgbe_mac_X550EM_x:
8470
		netdev->features |= NETIF_F_SCTP_CSUM;
8471 8472
		netdev->hw_features |= NETIF_F_SCTP_CSUM |
				       NETIF_F_NTUPLE;
8473 8474 8475 8476
		break;
	default:
		break;
	}
8477

B
Ben Greear 已提交
8478
	netdev->hw_features |= NETIF_F_RXALL;
8479
	netdev->features |= NETIF_F_HW_VLAN_CTAG_FILTER;
B
Ben Greear 已提交
8480

8481 8482
	netdev->vlan_features |= NETIF_F_TSO;
	netdev->vlan_features |= NETIF_F_TSO6;
8483
	netdev->vlan_features |= NETIF_F_IP_CSUM;
8484
	netdev->vlan_features |= NETIF_F_IPV6_CSUM;
8485 8486
	netdev->vlan_features |= NETIF_F_SG;

8487
	netdev->priv_flags |= IFF_UNICAST_FLT;
8488
	netdev->priv_flags |= IFF_SUPP_NOFCS;
8489

8490 8491 8492 8493 8494 8495 8496 8497 8498
	switch (adapter->hw.mac.type) {
	case ixgbe_mac_X550:
	case ixgbe_mac_X550EM_x:
		netdev->hw_enc_features |= NETIF_F_RXCSUM;
		break;
	default:
		break;
	}

J
Jeff Kirsher 已提交
8499
#ifdef CONFIG_IXGBE_DCB
8500 8501 8502
	netdev->dcbnl_ops = &dcbnl_ops;
#endif

8503
#ifdef IXGBE_FCOE
8504
	if (adapter->flags & IXGBE_FLAG_FCOE_CAPABLE) {
8505 8506
		unsigned int fcoe_l;

8507 8508
		if (hw->mac.ops.get_device_caps) {
			hw->mac.ops.get_device_caps(hw, &device_caps);
8509 8510
			if (device_caps & IXGBE_DEVICE_CAPS_FCOE_OFFLOADS)
				adapter->flags &= ~IXGBE_FLAG_FCOE_CAPABLE;
8511
		}
8512

8513 8514 8515

		fcoe_l = min_t(int, IXGBE_FCRETA_SIZE, num_online_cpus());
		adapter->ring_feature[RING_F_FCOE].limit = fcoe_l;
8516

8517 8518 8519
		netdev->features |= NETIF_F_FSO |
				    NETIF_F_FCOE_CRC;

8520 8521 8522
		netdev->vlan_features |= NETIF_F_FSO |
					 NETIF_F_FCOE_CRC |
					 NETIF_F_FCOE_MTU;
8523
	}
8524
#endif /* IXGBE_FCOE */
8525
	if (pci_using_dac) {
8526
		netdev->features |= NETIF_F_HIGHDMA;
8527 8528
		netdev->vlan_features |= NETIF_F_HIGHDMA;
	}
8529

8530 8531
	if (adapter->flags2 & IXGBE_FLAG2_RSC_CAPABLE)
		netdev->hw_features |= NETIF_F_LRO;
8532
	if (adapter->flags2 & IXGBE_FLAG2_RSC_ENABLED)
A
Alexander Duyck 已提交
8533 8534
		netdev->features |= NETIF_F_LRO;

8535
	/* make sure the EEPROM is good */
8536
	if (hw->eeprom.ops.validate_checksum(hw, NULL) < 0) {
8537
		e_dev_err("The EEPROM Checksum Is Not Valid\n");
8538
		err = -EIO;
8539
		goto err_sw_init;
8540 8541
	}

8542 8543
	ixgbe_get_platform_mac_addr(adapter);

8544 8545
	memcpy(netdev->dev_addr, hw->mac.perm_addr, netdev->addr_len);

8546
	if (!is_valid_ether_addr(netdev->dev_addr)) {
8547
		e_dev_err("invalid MAC address\n");
8548
		err = -EIO;
8549
		goto err_sw_init;
8550 8551
	}

8552 8553
	ixgbe_mac_set_default_filter(adapter, hw->mac.perm_addr);

8554
	setup_timer(&adapter->service_timer, &ixgbe_service_timer,
A
Alexander Duyck 已提交
8555
		    (unsigned long) adapter);
8556

8557 8558 8559 8560
	if (ixgbe_removed(hw->hw_addr)) {
		err = -EIO;
		goto err_sw_init;
	}
8561
	INIT_WORK(&adapter->service_task, ixgbe_service_task);
8562
	set_bit(__IXGBE_SERVICE_INITED, &adapter->state);
8563
	clear_bit(__IXGBE_SERVICE_SCHED, &adapter->state);
8564

8565 8566 8567
	err = ixgbe_init_interrupt_scheme(adapter);
	if (err)
		goto err_sw_init;
8568

8569
	/* WOL not supported for all devices */
E
Emil Tantilov 已提交
8570
	adapter->wol = 0;
8571
	hw->eeprom.ops.read(hw, 0x2c, &adapter->eeprom_cap);
8572
	hw->wol_enabled = ixgbe_wol_supported(adapter, pdev->device,
D
Don Skidmore 已提交
8573
						pdev->subsystem_device);
8574
	if (hw->wol_enabled)
8575
		adapter->wol = IXGBE_WUFC_MAG;
E
Emil Tantilov 已提交
8576

8577 8578
	device_set_wakeup_enable(&adapter->pdev->dev, adapter->wol);

8579 8580 8581 8582
	/* save off EEPROM version number */
	hw->eeprom.ops.read(hw, 0x2e, &adapter->eeprom_verh);
	hw->eeprom.ops.read(hw, 0x2d, &adapter->eeprom_verl);

8583 8584
	/* pick up the PCI bus settings for reporting later */
	hw->mac.ops.get_bus_info(hw);
8585
	if (ixgbe_pcie_from_parent(hw))
8586
		ixgbe_get_parent_bus_info(adapter);
8587

8588 8589 8590 8591 8592 8593 8594 8595 8596 8597 8598 8599
	/* calculate the expected PCIe bandwidth required for optimal
	 * performance. Note that some older parts will never have enough
	 * bandwidth due to being older generation PCIe parts. We clamp these
	 * parts to ensure no warning is displayed if it can't be fixed.
	 */
	switch (hw->mac.type) {
	case ixgbe_mac_82598EB:
		expected_gts = min(ixgbe_enumerate_functions(adapter) * 10, 16);
		break;
	default:
		expected_gts = ixgbe_enumerate_functions(adapter) * 10;
		break;
8600
	}
8601 8602 8603 8604

	/* don't check link if we failed to enumerate functions */
	if (expected_gts > 0)
		ixgbe_check_minimum_link(adapter, expected_gts);
8605

8606
	err = ixgbe_read_pba_string_generic(hw, part_str, sizeof(part_str));
8607
	if (err)
8608
		strlcpy(part_str, "Unknown", sizeof(part_str));
8609 8610 8611
	if (ixgbe_is_sfp(hw) && hw->phy.sfp_type != ixgbe_sfp_type_not_present)
		e_dev_info("MAC: %d, PHY: %d, SFP+: %d, PBA No: %s\n",
			   hw->mac.type, hw->phy.type, hw->phy.sfp_type,
8612
			   part_str);
8613 8614 8615 8616 8617 8618
	else
		e_dev_info("MAC: %d, PHY: %d, PBA No: %s\n",
			   hw->mac.type, hw->phy.type, part_str);

	e_dev_info("%pM\n", netdev->dev_addr);

8619
	/* reset the hardware with the new settings */
8620 8621 8622
	err = hw->mac.ops.start_hw(hw);
	if (err == IXGBE_ERR_EEPROM_VERSION) {
		/* We are running on a pre-production device, log a warning */
8623 8624 8625 8626 8627 8628
		e_dev_warn("This device is a pre-production adapter/LOM. "
			   "Please be aware there may be issues associated "
			   "with your hardware.  If you are experiencing "
			   "problems please contact your Intel or hardware "
			   "representative who provided you with this "
			   "hardware.\n");
8629
	}
8630 8631 8632 8633 8634
	strcpy(netdev->name, "eth%d");
	err = register_netdev(netdev);
	if (err)
		goto err_register;

8635 8636
	pci_set_drvdata(pdev, adapter);

8637 8638
	/* power down the optics for 82599 SFP+ fiber */
	if (hw->mac.ops.disable_tx_laser)
8639 8640
		hw->mac.ops.disable_tx_laser(hw);

8641 8642 8643
	/* carrier off reporting is important to ethtool even BEFORE open */
	netif_carrier_off(netdev);

8644
#ifdef CONFIG_IXGBE_DCA
8645
	if (dca_add_requester(&pdev->dev) == 0) {
8646 8647 8648 8649
		adapter->flags |= IXGBE_FLAG_DCA_ENABLED;
		ixgbe_setup_dca(adapter);
	}
#endif
8650
	if (adapter->flags & IXGBE_FLAG_SRIOV_ENABLED) {
8651
		e_info(probe, "IOV is enabled with %d VFs\n", adapter->num_vfs);
8652 8653 8654 8655
		for (i = 0; i < adapter->num_vfs; i++)
			ixgbe_vf_configuration(pdev, (i | 0x10000000));
	}

8656 8657 8658
	/* firmware requires driver version to be 0xFFFFFFFF
	 * since os does not support feature
	 */
E
Emil Tantilov 已提交
8659
	if (hw->mac.ops.set_fw_drv_ver)
8660 8661
		hw->mac.ops.set_fw_drv_ver(hw, 0xFF, 0xFF, 0xFF,
					   0xFF);
E
Emil Tantilov 已提交
8662

8663 8664
	/* add san mac addr to netdev */
	ixgbe_add_sanmac_netdev(netdev);
8665

8666
	e_dev_info("%s\n", ixgbe_default_device_descr);
8667

8668
#ifdef CONFIG_IXGBE_HWMON
8669 8670
	if (ixgbe_sysfs_init(adapter))
		e_err(probe, "failed to allocate sysfs resources\n");
8671
#endif /* CONFIG_IXGBE_HWMON */
8672

C
Catherine Sullivan 已提交
8673 8674
	ixgbe_dbg_adapter_init(adapter);

8675 8676
	/* setup link for SFP devices with MNG FW, else wait for IXGBE_UP */
	if (ixgbe_mng_enabled(hw) && ixgbe_is_sfp(hw) && hw->mac.ops.setup_link)
8677 8678 8679 8680
		hw->mac.ops.setup_link(hw,
			IXGBE_LINK_SPEED_10GB_FULL | IXGBE_LINK_SPEED_1GB_FULL,
			true);

8681 8682 8683
	return 0;

err_register:
8684
	ixgbe_release_hw_control(adapter);
8685
	ixgbe_clear_interrupt_scheme(adapter);
8686
err_sw_init:
8687
	ixgbe_disable_sriov(adapter);
8688
	adapter->flags2 &= ~IXGBE_FLAG2_SEARCH_FOR_SFP;
8689
	iounmap(adapter->io_addr);
8690
	kfree(adapter->mac_table);
8691
err_ioremap:
8692
	disable_dev = !test_and_set_bit(__IXGBE_DISABLED, &adapter->state);
8693 8694
	free_netdev(netdev);
err_alloc_etherdev:
8695 8696
	pci_release_selected_regions(pdev,
				     pci_select_bars(pdev, IORESOURCE_MEM));
8697 8698
err_pci_reg:
err_dma:
8699
	if (!adapter || disable_dev)
8700
		pci_disable_device(pdev);
8701 8702 8703 8704 8705 8706 8707 8708 8709 8710 8711 8712
	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.
 **/
8713
static void ixgbe_remove(struct pci_dev *pdev)
8714
{
8715
	struct ixgbe_adapter *adapter = pci_get_drvdata(pdev);
8716
	struct net_device *netdev;
8717
	bool disable_dev;
8718

8719 8720 8721 8722 8723
	/* if !adapter then we already cleaned up in probe */
	if (!adapter)
		return;

	netdev  = adapter->netdev;
C
Catherine Sullivan 已提交
8724 8725
	ixgbe_dbg_adapter_exit(adapter);

8726
	set_bit(__IXGBE_REMOVING, &adapter->state);
8727
	cancel_work_sync(&adapter->service_task);
8728

8729

8730
#ifdef CONFIG_IXGBE_DCA
8731 8732 8733 8734 8735 8736 8737
	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
8738
#ifdef CONFIG_IXGBE_HWMON
8739
	ixgbe_sysfs_exit(adapter);
8740
#endif /* CONFIG_IXGBE_HWMON */
8741

8742 8743 8744
	/* remove the added san mac */
	ixgbe_del_sanmac_netdev(netdev);

D
Donald Skidmore 已提交
8745 8746
	if (netdev->reg_state == NETREG_REGISTERED)
		unregister_netdev(netdev);
8747

8748 8749 8750 8751 8752 8753 8754 8755
#ifdef CONFIG_PCI_IOV
	/*
	 * Only disable SR-IOV on unload if the user specified the now
	 * deprecated max_vfs module parameter.
	 */
	if (max_vfs)
		ixgbe_disable_sriov(adapter);
#endif
8756
	ixgbe_clear_interrupt_scheme(adapter);
8757

8758
	ixgbe_release_hw_control(adapter);
8759

8760 8761 8762 8763 8764
#ifdef CONFIG_DCB
	kfree(adapter->ixgbe_ieee_pfc);
	kfree(adapter->ixgbe_ieee_ets);

#endif
8765
	iounmap(adapter->io_addr);
8766
	pci_release_selected_regions(pdev, pci_select_bars(pdev,
8767
				     IORESOURCE_MEM));
8768

8769
	e_dev_info("complete\n");
8770

8771
	kfree(adapter->mac_table);
8772
	disable_dev = !test_and_set_bit(__IXGBE_DISABLED, &adapter->state);
8773 8774
	free_netdev(netdev);

8775
	pci_disable_pcie_error_reporting(pdev);
8776

8777
	if (disable_dev)
8778
		pci_disable_device(pdev);
8779 8780 8781 8782 8783 8784 8785 8786 8787 8788 8789
}

/**
 * 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,
8790
						pci_channel_state_t state)
8791
{
8792 8793
	struct ixgbe_adapter *adapter = pci_get_drvdata(pdev);
	struct net_device *netdev = adapter->netdev;
8794

8795
#ifdef CONFIG_PCI_IOV
8796
	struct ixgbe_hw *hw = &adapter->hw;
8797 8798 8799 8800 8801 8802 8803 8804 8805 8806
	struct pci_dev *bdev, *vfdev;
	u32 dw0, dw1, dw2, dw3;
	int vf, pos;
	u16 req_id, pf_func;

	if (adapter->hw.mac.type == ixgbe_mac_82598EB ||
	    adapter->num_vfs == 0)
		goto skip_bad_vf_detection;

	bdev = pdev->bus->self;
8807
	while (bdev && (pci_pcie_type(bdev) != PCI_EXP_TYPE_ROOT_PORT))
8808 8809 8810 8811 8812 8813 8814 8815 8816
		bdev = bdev->bus->self;

	if (!bdev)
		goto skip_bad_vf_detection;

	pos = pci_find_ext_capability(bdev, PCI_EXT_CAP_ID_ERR);
	if (!pos)
		goto skip_bad_vf_detection;

8817 8818 8819 8820 8821 8822
	dw0 = ixgbe_read_pci_cfg_dword(hw, pos + PCI_ERR_HEADER_LOG);
	dw1 = ixgbe_read_pci_cfg_dword(hw, pos + PCI_ERR_HEADER_LOG + 4);
	dw2 = ixgbe_read_pci_cfg_dword(hw, pos + PCI_ERR_HEADER_LOG + 8);
	dw3 = ixgbe_read_pci_cfg_dword(hw, pos + PCI_ERR_HEADER_LOG + 12);
	if (ixgbe_removed(hw->hw_addr))
		goto skip_bad_vf_detection;
8823 8824 8825 8826 8827 8828 8829 8830 8831 8832 8833 8834 8835 8836 8837 8838 8839 8840 8841 8842 8843 8844

	req_id = dw1 >> 16;
	/* On the 82599 if bit 7 of the requestor ID is set then it's a VF */
	if (!(req_id & 0x0080))
		goto skip_bad_vf_detection;

	pf_func = req_id & 0x01;
	if ((pf_func & 1) == (pdev->devfn & 1)) {
		unsigned int device_id;

		vf = (req_id & 0x7F) >> 1;
		e_dev_err("VF %d has caused a PCIe error\n", vf);
		e_dev_err("TLP: dw0: %8.8x\tdw1: %8.8x\tdw2: "
				"%8.8x\tdw3: %8.8x\n",
		dw0, dw1, dw2, dw3);
		switch (adapter->hw.mac.type) {
		case ixgbe_mac_82599EB:
			device_id = IXGBE_82599_VF_DEVICE_ID;
			break;
		case ixgbe_mac_X540:
			device_id = IXGBE_X540_VF_DEVICE_ID;
			break;
8845 8846 8847 8848 8849 8850
		case ixgbe_mac_X550:
			device_id = IXGBE_DEV_ID_X550_VF;
			break;
		case ixgbe_mac_X550EM_x:
			device_id = IXGBE_DEV_ID_X550EM_X_VF;
			break;
8851 8852 8853 8854 8855 8856
		default:
			device_id = 0;
			break;
		}

		/* Find the pci device of the offending VF */
J
Jon Mason 已提交
8857
		vfdev = pci_get_device(PCI_VENDOR_ID_INTEL, device_id, NULL);
8858 8859 8860
		while (vfdev) {
			if (vfdev->devfn == (req_id & 0xFF))
				break;
J
Jon Mason 已提交
8861
			vfdev = pci_get_device(PCI_VENDOR_ID_INTEL,
8862 8863 8864 8865 8866 8867 8868 8869
					       device_id, vfdev);
		}
		/*
		 * There's a slim chance the VF could have been hot plugged,
		 * so if it is no longer present we don't need to issue the
		 * VFLR.  Just clean up the AER in that case.
		 */
		if (vfdev) {
8870
			ixgbe_issue_vf_flr(adapter, vfdev);
G
Greg Rose 已提交
8871 8872
			/* Free device reference count */
			pci_dev_put(vfdev);
8873 8874 8875 8876 8877 8878 8879 8880 8881 8882 8883 8884 8885 8886 8887 8888 8889
		}

		pci_cleanup_aer_uncorrect_error_status(pdev);
	}

	/*
	 * Even though the error may have occurred on the other port
	 * we still need to increment the vf error reference count for
	 * both ports because the I/O resume function will be called
	 * for both of them.
	 */
	adapter->vferr_refcount++;

	return PCI_ERS_RESULT_RECOVERED;

skip_bad_vf_detection:
#endif /* CONFIG_PCI_IOV */
8890 8891 8892
	if (!test_bit(__IXGBE_SERVICE_INITED, &adapter->state))
		return PCI_ERS_RESULT_DISCONNECT;

8893
	rtnl_lock();
8894 8895
	netif_device_detach(netdev);

8896 8897
	if (state == pci_channel_io_perm_failure) {
		rtnl_unlock();
8898
		return PCI_ERS_RESULT_DISCONNECT;
8899
	}
8900

8901 8902
	if (netif_running(netdev))
		ixgbe_down(adapter);
8903 8904 8905 8906

	if (!test_and_set_bit(__IXGBE_DISABLED, &adapter->state))
		pci_disable_device(pdev);
	rtnl_unlock();
8907

8908
	/* Request a slot reset. */
8909 8910 8911 8912 8913 8914 8915 8916 8917 8918 8919
	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)
{
8920
	struct ixgbe_adapter *adapter = pci_get_drvdata(pdev);
8921 8922
	pci_ers_result_t result;
	int err;
8923

8924
	if (pci_enable_device_mem(pdev)) {
8925
		e_err(probe, "Cannot re-enable PCI device after reset.\n");
8926 8927
		result = PCI_ERS_RESULT_DISCONNECT;
	} else {
8928
		smp_mb__before_atomic();
8929
		clear_bit(__IXGBE_DISABLED, &adapter->state);
8930
		adapter->hw.hw_addr = adapter->io_addr;
8931 8932
		pci_set_master(pdev);
		pci_restore_state(pdev);
8933
		pci_save_state(pdev);
8934

8935
		pci_wake_from_d3(pdev, false);
8936

8937
		ixgbe_reset(adapter);
8938
		IXGBE_WRITE_REG(&adapter->hw, IXGBE_WUS, ~0);
8939 8940 8941 8942 8943
		result = PCI_ERS_RESULT_RECOVERED;
	}

	err = pci_cleanup_aer_uncorrect_error_status(pdev);
	if (err) {
8944 8945
		e_dev_err("pci_cleanup_aer_uncorrect_error_status "
			  "failed 0x%0x\n", err);
8946 8947
		/* non-fatal, continue */
	}
8948

8949
	return result;
8950 8951 8952 8953 8954 8955 8956 8957 8958 8959 8960
}

/**
 * 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)
{
8961 8962
	struct ixgbe_adapter *adapter = pci_get_drvdata(pdev);
	struct net_device *netdev = adapter->netdev;
8963

8964 8965 8966 8967 8968 8969 8970 8971
#ifdef CONFIG_PCI_IOV
	if (adapter->vferr_refcount) {
		e_info(drv, "Resuming after VF err\n");
		adapter->vferr_refcount--;
		return;
	}

#endif
8972 8973
	if (netif_running(netdev))
		ixgbe_up(adapter);
8974 8975 8976 8977

	netif_device_attach(netdev);
}

8978
static const struct pci_error_handlers ixgbe_err_handler = {
8979 8980 8981 8982 8983 8984 8985 8986 8987
	.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,
8988
	.remove   = ixgbe_remove,
8989 8990 8991 8992 8993
#ifdef CONFIG_PM
	.suspend  = ixgbe_suspend,
	.resume   = ixgbe_resume,
#endif
	.shutdown = ixgbe_shutdown,
8994
	.sriov_configure = ixgbe_pci_sriov_configure,
8995 8996 8997 8998 8999 9000 9001 9002 9003 9004 9005 9006
	.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;
9007
	pr_info("%s - version %s\n", ixgbe_driver_string, ixgbe_driver_version);
9008
	pr_info("%s\n", ixgbe_copyright);
9009

C
Catherine Sullivan 已提交
9010 9011
	ixgbe_dbg_init();

9012 9013 9014 9015 9016 9017
	ret = pci_register_driver(&ixgbe_driver);
	if (ret) {
		ixgbe_dbg_exit();
		return ret;
	}

9018
#ifdef CONFIG_IXGBE_DCA
9019 9020
	dca_register_notify(&dca_notifier);
#endif
9021

9022
	return 0;
9023
}
9024

9025 9026 9027 9028 9029 9030 9031 9032 9033 9034
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)
{
9035
#ifdef CONFIG_IXGBE_DCA
9036 9037
	dca_unregister_notify(&dca_notifier);
#endif
9038
	pci_unregister_driver(&ixgbe_driver);
C
Catherine Sullivan 已提交
9039 9040

	ixgbe_dbg_exit();
9041
}
9042

9043
#ifdef CONFIG_IXGBE_DCA
9044
static int ixgbe_notify_dca(struct notifier_block *nb, unsigned long event,
9045
			    void *p)
9046 9047 9048 9049
{
	int ret_val;

	ret_val = driver_for_each_device(&ixgbe_driver.driver, NULL, &event,
9050
					 __ixgbe_notify_dca);
9051 9052 9053

	return ret_val ? NOTIFY_BAD : NOTIFY_DONE;
}
9054

9055
#endif /* CONFIG_IXGBE_DCA */
9056

9057 9058 9059
module_exit(ixgbe_exit_module);

/* ixgbe_main.c */