netdev.c 182.8 KB
Newer Older
1 2 3
/*******************************************************************************

  Intel PRO/1000 Linux driver
B
Bruce Allan 已提交
4
  Copyright(c) 1999 - 2012 Intel Corporation.
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

  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:
  Linux NICS <linux.nics@intel.com>
  e1000-devel Mailing List <e1000-devel@lists.sourceforge.net>
  Intel Corporation, 5200 N.E. Elam Young Parkway, Hillsboro, OR 97124-6497

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

29 30
#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt

31 32 33 34 35 36 37 38
#include <linux/module.h>
#include <linux/types.h>
#include <linux/init.h>
#include <linux/pci.h>
#include <linux/vmalloc.h>
#include <linux/pagemap.h>
#include <linux/delay.h>
#include <linux/netdevice.h>
39
#include <linux/interrupt.h>
40 41
#include <linux/tcp.h>
#include <linux/ipv6.h>
42
#include <linux/slab.h>
43 44 45 46 47 48 49
#include <net/checksum.h>
#include <net/ip6_checksum.h>
#include <linux/mii.h>
#include <linux/ethtool.h>
#include <linux/if_vlan.h>
#include <linux/cpu.h>
#include <linux/smp.h>
50
#include <linux/pm_qos.h>
51
#include <linux/pm_runtime.h>
J
Jesse Brandeburg 已提交
52
#include <linux/aer.h>
53
#include <linux/prefetch.h>
54 55 56

#include "e1000.h"

B
Bruce Allan 已提交
57
#define DRV_EXTRAVERSION "-k"
58

59
#define DRV_VERSION "1.10.6" DRV_EXTRAVERSION
60 61 62
char e1000e_driver_name[] = "e1000e";
const char e1000e_driver_version[] = DRV_VERSION;

63 64 65 66 67
#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)");

68 69
static void e1000e_disable_aspm(struct pci_dev *pdev, u16 state);

70 71 72 73
static const struct e1000_info *e1000_info_tbl[] = {
	[board_82571]		= &e1000_82571_info,
	[board_82572]		= &e1000_82572_info,
	[board_82573]		= &e1000_82573_info,
74
	[board_82574]		= &e1000_82574_info,
75
	[board_82583]		= &e1000_82583_info,
76 77 78
	[board_80003es2lan]	= &e1000_es2_info,
	[board_ich8lan]		= &e1000_ich8_info,
	[board_ich9lan]		= &e1000_ich9_info,
79
	[board_ich10lan]	= &e1000_ich10_info,
80
	[board_pchlan]		= &e1000_pch_info,
81
	[board_pch2lan]		= &e1000_pch2_info,
82 83
};

84 85 86 87 88
struct e1000_reg_info {
	u32 ofs;
	char *name;
};

89 90 91 92 93 94 95 96 97 98 99
#define E1000_RDFH	0x02410	/* Rx Data FIFO Head - RW */
#define E1000_RDFT	0x02418	/* Rx Data FIFO Tail - RW */
#define E1000_RDFHS	0x02420	/* Rx Data FIFO Head Saved - RW */
#define E1000_RDFTS	0x02428	/* Rx Data FIFO Tail Saved - RW */
#define E1000_RDFPC	0x02430	/* Rx Data FIFO Packet Count - RW */

#define E1000_TDFH	0x03410	/* Tx Data FIFO Head - RW */
#define E1000_TDFT	0x03418	/* Tx Data FIFO Tail - RW */
#define E1000_TDFHS	0x03420	/* Tx Data FIFO Head Saved - RW */
#define E1000_TDFTS	0x03428	/* Tx Data FIFO Tail Saved - RW */
#define E1000_TDFPC	0x03430	/* Tx Data FIFO Packet Count - RW */
100 101 102 103 104 105 106 107 108 109 110

static const struct e1000_reg_info e1000_reg_info_tbl[] = {

	/* General Registers */
	{E1000_CTRL, "CTRL"},
	{E1000_STATUS, "STATUS"},
	{E1000_CTRL_EXT, "CTRL_EXT"},

	/* Interrupt Registers */
	{E1000_ICR, "ICR"},

111
	/* Rx Registers */
112
	{E1000_RCTL, "RCTL"},
113 114 115
	{E1000_RDLEN(0), "RDLEN"},
	{E1000_RDH(0), "RDH"},
	{E1000_RDT(0), "RDT"},
116 117 118
	{E1000_RDTR, "RDTR"},
	{E1000_RXDCTL(0), "RXDCTL"},
	{E1000_ERT, "ERT"},
119 120
	{E1000_RDBAL(0), "RDBAL"},
	{E1000_RDBAH(0), "RDBAH"},
121 122 123 124 125 126
	{E1000_RDFH, "RDFH"},
	{E1000_RDFT, "RDFT"},
	{E1000_RDFHS, "RDFHS"},
	{E1000_RDFTS, "RDFTS"},
	{E1000_RDFPC, "RDFPC"},

127
	/* Tx Registers */
128
	{E1000_TCTL, "TCTL"},
129 130 131 132 133
	{E1000_TDBAL(0), "TDBAL"},
	{E1000_TDBAH(0), "TDBAH"},
	{E1000_TDLEN(0), "TDLEN"},
	{E1000_TDH(0), "TDH"},
	{E1000_TDT(0), "TDT"},
134 135 136 137 138 139 140 141 142 143 144
	{E1000_TIDV, "TIDV"},
	{E1000_TXDCTL(0), "TXDCTL"},
	{E1000_TADV, "TADV"},
	{E1000_TARC(0), "TARC"},
	{E1000_TDFH, "TDFH"},
	{E1000_TDFT, "TDFT"},
	{E1000_TDFHS, "TDFHS"},
	{E1000_TDFTS, "TDFTS"},
	{E1000_TDFPC, "TDFPC"},

	/* List Terminator */
145
	{0, NULL}
146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170
};

/*
 * e1000_regdump - register printout routine
 */
static void e1000_regdump(struct e1000_hw *hw, struct e1000_reg_info *reginfo)
{
	int n = 0;
	char rname[16];
	u32 regs[8];

	switch (reginfo->ofs) {
	case E1000_RXDCTL(0):
		for (n = 0; n < 2; n++)
			regs[n] = __er32(hw, E1000_RXDCTL(n));
		break;
	case E1000_TXDCTL(0):
		for (n = 0; n < 2; n++)
			regs[n] = __er32(hw, E1000_TXDCTL(n));
		break;
	case E1000_TARC(0):
		for (n = 0; n < 2; n++)
			regs[n] = __er32(hw, E1000_TARC(n));
		break;
	default:
171 172
		pr_info("%-15s %08x\n",
			reginfo->name, __er32(hw, reginfo->ofs));
173 174 175 176
		return;
	}

	snprintf(rname, 16, "%s%s", reginfo->name, "[0-1]");
177
	pr_info("%-15s %08x %08x\n", rname, regs[0], regs[1]);
178 179 180
}

/*
181
 * e1000e_dump - Print registers, Tx-ring and Rx-ring
182 183 184 185 186 187 188 189
 */
static void e1000e_dump(struct e1000_adapter *adapter)
{
	struct net_device *netdev = adapter->netdev;
	struct e1000_hw *hw = &adapter->hw;
	struct e1000_reg_info *reginfo;
	struct e1000_ring *tx_ring = adapter->tx_ring;
	struct e1000_tx_desc *tx_desc;
190
	struct my_u0 {
191 192
		__le64 a;
		__le64 b;
193
	} *u0;
194 195 196
	struct e1000_buffer *buffer_info;
	struct e1000_ring *rx_ring = adapter->rx_ring;
	union e1000_rx_desc_packet_split *rx_desc_ps;
197
	union e1000_rx_desc_extended *rx_desc;
198
	struct my_u1 {
199 200 201 202
		__le64 a;
		__le64 b;
		__le64 c;
		__le64 d;
203
	} *u1;
204 205 206 207 208 209 210 211 212
	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");
213 214 215 216
		pr_info("Device Name     state            trans_start      last_rx\n");
		pr_info("%-15s %016lX %016lX %016lX\n",
			netdev->name, netdev->state, netdev->trans_start,
			netdev->last_rx);
217 218 219 220
	}

	/* Print Registers */
	dev_info(&adapter->pdev->dev, "Register Dump\n");
221
	pr_info(" Register Name   Value\n");
222 223 224 225 226
	for (reginfo = (struct e1000_reg_info *)e1000_reg_info_tbl;
	     reginfo->name; reginfo++) {
		e1000_regdump(hw, reginfo);
	}

227
	/* Print Tx Ring Summary */
228
	if (!netdev || !netif_running(netdev))
229
		return;
230

231
	dev_info(&adapter->pdev->dev, "Tx Ring Summary\n");
232
	pr_info("Queue [NTU] [NTC] [bi(ntc)->dma  ] leng ntw timestamp\n");
233
	buffer_info = &tx_ring->buffer_info[tx_ring->next_to_clean];
234 235 236 237 238 239
	pr_info(" %5d %5X %5X %016llX %04X %3X %016llX\n",
		0, tx_ring->next_to_use, tx_ring->next_to_clean,
		(unsigned long long)buffer_info->dma,
		buffer_info->length,
		buffer_info->next_to_watch,
		(unsigned long long)buffer_info->time_stamp);
240

241
	/* Print Tx Ring */
242 243 244
	if (!netif_msg_tx_done(adapter))
		goto rx_ring_summary;

245
	dev_info(&adapter->pdev->dev, "Tx Ring Dump\n");
246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273

	/* Transmit Descriptor Formats - DEXT[29] is 0 (Legacy) or 1 (Extended)
	 *
	 * Legacy Transmit Descriptor
	 *   +--------------------------------------------------------------+
	 * 0 |         Buffer Address [63:0] (Reserved on Write Back)       |
	 *   +--------------------------------------------------------------+
	 * 8 | Special  |    CSS     | Status |  CMD    |  CSO   |  Length  |
	 *   +--------------------------------------------------------------+
	 *   63       48 47        36 35    32 31     24 23    16 15        0
	 *
	 * Extended Context Descriptor (DTYP=0x0) for TSO or checksum offload
	 *   63      48 47    40 39       32 31             16 15    8 7      0
	 *   +----------------------------------------------------------------+
	 * 0 |  TUCSE  | TUCS0  |   TUCSS   |     IPCSE       | IPCS0 | IPCSS |
	 *   +----------------------------------------------------------------+
	 * 8 |   MSS   | HDRLEN | RSV | STA | TUCMD | DTYP |      PAYLEN      |
	 *   +----------------------------------------------------------------+
	 *   63      48 47    40 39 36 35 32 31   24 23  20 19                0
	 *
	 * Extended Data Descriptor (DTYP=0x1)
	 *   +----------------------------------------------------------------+
	 * 0 |                     Buffer Address [63:0]                      |
	 *   +----------------------------------------------------------------+
	 * 8 | VLAN tag |  POPTS  | Rsvd | Status | Command | DTYP |  DTALEN  |
	 *   +----------------------------------------------------------------+
	 *   63       48 47     40 39  36 35    32 31     24 23  20 19        0
	 */
274 275 276
	pr_info("Tl[desc]     [address 63:0  ] [SpeCssSCmCsLen] [bi->dma       ] leng  ntw timestamp        bi->skb <-- Legacy format\n");
	pr_info("Tc[desc]     [Ce CoCsIpceCoS] [MssHlRSCm0Plen] [bi->dma       ] leng  ntw timestamp        bi->skb <-- Ext Context format\n");
	pr_info("Td[desc]     [address 63:0  ] [VlaPoRSCm1Dlen] [bi->dma       ] leng  ntw timestamp        bi->skb <-- Ext Data format\n");
277
	for (i = 0; tx_ring->desc && (i < tx_ring->count); i++) {
278
		const char *next_desc;
279 280 281 282
		tx_desc = E1000_TX_DESC(*tx_ring, i);
		buffer_info = &tx_ring->buffer_info[i];
		u0 = (struct my_u0 *)tx_desc;
		if (i == tx_ring->next_to_use && i == tx_ring->next_to_clean)
283
			next_desc = " NTC/U";
284
		else if (i == tx_ring->next_to_use)
285
			next_desc = " NTU";
286
		else if (i == tx_ring->next_to_clean)
287
			next_desc = " NTC";
288
		else
289 290 291 292 293 294 295 296 297 298 299
			next_desc = "";
		pr_info("T%c[0x%03X]    %016llX %016llX %016llX %04X  %3X %016llX %p%s\n",
			(!(le64_to_cpu(u0->b) & (1 << 29)) ? 'l' :
			 ((le64_to_cpu(u0->b) & (1 << 20)) ? 'd' : 'c')),
			i,
			(unsigned long long)le64_to_cpu(u0->a),
			(unsigned long long)le64_to_cpu(u0->b),
			(unsigned long long)buffer_info->dma,
			buffer_info->length, buffer_info->next_to_watch,
			(unsigned long long)buffer_info->time_stamp,
			buffer_info->skb, next_desc);
300 301 302

		if (netif_msg_pktdata(adapter) && buffer_info->dma != 0)
			print_hex_dump(KERN_INFO, "", DUMP_PREFIX_ADDRESS,
303 304
				       16, 1, phys_to_virt(buffer_info->dma),
				       buffer_info->length, true);
305 306
	}

307
	/* Print Rx Ring Summary */
308
rx_ring_summary:
309
	dev_info(&adapter->pdev->dev, "Rx Ring Summary\n");
310 311 312
	pr_info("Queue [NTU] [NTC]\n");
	pr_info(" %5d %5X %5X\n",
		0, rx_ring->next_to_use, rx_ring->next_to_clean);
313

314
	/* Print Rx Ring */
315
	if (!netif_msg_rx_status(adapter))
316
		return;
317

318
	dev_info(&adapter->pdev->dev, "Rx Ring Dump\n");
319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334
	switch (adapter->rx_ps_pages) {
	case 1:
	case 2:
	case 3:
		/* [Extended] Packet Split Receive Descriptor Format
		 *
		 *    +-----------------------------------------------------+
		 *  0 |                Buffer Address 0 [63:0]              |
		 *    +-----------------------------------------------------+
		 *  8 |                Buffer Address 1 [63:0]              |
		 *    +-----------------------------------------------------+
		 * 16 |                Buffer Address 2 [63:0]              |
		 *    +-----------------------------------------------------+
		 * 24 |                Buffer Address 3 [63:0]              |
		 *    +-----------------------------------------------------+
		 */
335
		pr_info("R  [desc]      [buffer 0 63:0 ] [buffer 1 63:0 ] [buffer 2 63:0 ] [buffer 3 63:0 ] [bi->dma       ] [bi->skb] <-- Ext Pkt Split format\n");
336 337 338 339 340 341 342 343 344 345 346
		/* [Extended] Receive Descriptor (Write-Back) Format
		 *
		 *   63       48 47    32 31     13 12    8 7    4 3        0
		 *   +------------------------------------------------------+
		 * 0 | Packet   | IP     |  Rsvd   | MRQ   | Rsvd | MRQ RSS |
		 *   | Checksum | Ident  |         | Queue |      |  Type   |
		 *   +------------------------------------------------------+
		 * 8 | VLAN Tag | Length | Extended Error | Extended Status |
		 *   +------------------------------------------------------+
		 *   63       48 47    32 31            20 19               0
		 */
347
		pr_info("RWB[desc]      [ck ipid mrqhsh] [vl   l0 ee  es] [ l3  l2  l1 hs] [reserved      ] ---------------- [bi->skb] <-- Ext Rx Write-Back format\n");
348
		for (i = 0; i < rx_ring->count; i++) {
349
			const char *next_desc;
350 351 352 353
			buffer_info = &rx_ring->buffer_info[i];
			rx_desc_ps = E1000_RX_DESC_PS(*rx_ring, i);
			u1 = (struct my_u1 *)rx_desc_ps;
			staterr =
354
			    le32_to_cpu(rx_desc_ps->wb.middle.status_error);
355 356 357 358 359 360 361 362

			if (i == rx_ring->next_to_use)
				next_desc = " NTU";
			else if (i == rx_ring->next_to_clean)
				next_desc = " NTC";
			else
				next_desc = "";

363 364
			if (staterr & E1000_RXD_STAT_DD) {
				/* Descriptor Done */
365 366 367 368 369 370 371
				pr_info("%s[0x%03X]     %016llX %016llX %016llX %016llX ---------------- %p%s\n",
					"RWB", i,
					(unsigned long long)le64_to_cpu(u1->a),
					(unsigned long long)le64_to_cpu(u1->b),
					(unsigned long long)le64_to_cpu(u1->c),
					(unsigned long long)le64_to_cpu(u1->d),
					buffer_info->skb, next_desc);
372
			} else {
373 374 375 376 377 378 379 380
				pr_info("%s[0x%03X]     %016llX %016llX %016llX %016llX %016llX %p%s\n",
					"R  ", i,
					(unsigned long long)le64_to_cpu(u1->a),
					(unsigned long long)le64_to_cpu(u1->b),
					(unsigned long long)le64_to_cpu(u1->c),
					(unsigned long long)le64_to_cpu(u1->d),
					(unsigned long long)buffer_info->dma,
					buffer_info->skb, next_desc);
381 382 383 384 385 386 387 388 389 390 391

				if (netif_msg_pktdata(adapter))
					print_hex_dump(KERN_INFO, "",
						DUMP_PREFIX_ADDRESS, 16, 1,
						phys_to_virt(buffer_info->dma),
						adapter->rx_ps_bsize0, true);
			}
		}
		break;
	default:
	case 0:
392
		/* Extended Receive Descriptor (Read) Format
393
		 *
394 395 396 397 398
		 *   +-----------------------------------------------------+
		 * 0 |                Buffer Address [63:0]                |
		 *   +-----------------------------------------------------+
		 * 8 |                      Reserved                       |
		 *   +-----------------------------------------------------+
399
		 */
400
		pr_info("R  [desc]      [buf addr 63:0 ] [reserved 63:0 ] [bi->dma       ] [bi->skb] <-- Ext (Read) format\n");
401 402 403 404 405 406 407 408 409 410 411 412 413
		/* Extended Receive Descriptor (Write-Back) Format
		 *
		 *   63       48 47    32 31    24 23            4 3        0
		 *   +------------------------------------------------------+
		 *   |     RSS Hash      |        |               |         |
		 * 0 +-------------------+  Rsvd  |   Reserved    | MRQ RSS |
		 *   | Packet   | IP     |        |               |  Type   |
		 *   | Checksum | Ident  |        |               |         |
		 *   +------------------------------------------------------+
		 * 8 | VLAN Tag | Length | Extended Error | Extended Status |
		 *   +------------------------------------------------------+
		 *   63       48 47    32 31            20 19               0
		 */
414
		pr_info("RWB[desc]      [cs ipid    mrq] [vt   ln xe  xs] [bi->skb] <-- Ext (Write-Back) format\n");
415 416

		for (i = 0; i < rx_ring->count; i++) {
417 418
			const char *next_desc;

419
			buffer_info = &rx_ring->buffer_info[i];
420 421 422
			rx_desc = E1000_RX_DESC_EXT(*rx_ring, i);
			u1 = (struct my_u1 *)rx_desc;
			staterr = le32_to_cpu(rx_desc->wb.upper.status_error);
423 424 425 426 427 428 429 430

			if (i == rx_ring->next_to_use)
				next_desc = " NTU";
			else if (i == rx_ring->next_to_clean)
				next_desc = " NTC";
			else
				next_desc = "";

431 432
			if (staterr & E1000_RXD_STAT_DD) {
				/* Descriptor Done */
433 434 435 436 437
				pr_info("%s[0x%03X]     %016llX %016llX ---------------- %p%s\n",
					"RWB", i,
					(unsigned long long)le64_to_cpu(u1->a),
					(unsigned long long)le64_to_cpu(u1->b),
					buffer_info->skb, next_desc);
438
			} else {
439 440 441 442 443 444
				pr_info("%s[0x%03X]     %016llX %016llX %016llX %p%s\n",
					"R  ", i,
					(unsigned long long)le64_to_cpu(u1->a),
					(unsigned long long)le64_to_cpu(u1->b),
					(unsigned long long)buffer_info->dma,
					buffer_info->skb, next_desc);
445 446 447 448 449 450 451 452 453 454

				if (netif_msg_pktdata(adapter))
					print_hex_dump(KERN_INFO, "",
						       DUMP_PREFIX_ADDRESS, 16,
						       1,
						       phys_to_virt
						       (buffer_info->dma),
						       adapter->rx_buffer_len,
						       true);
			}
455 456 457 458
		}
	}
}

459 460 461 462 463 464 465 466 467 468 469 470
/**
 * e1000_desc_unused - calculate if we have unused descriptors
 **/
static int e1000_desc_unused(struct e1000_ring *ring)
{
	if (ring->next_to_clean > ring->next_to_use)
		return ring->next_to_clean - ring->next_to_use - 1;

	return ring->count + ring->next_to_clean - ring->next_to_use - 1;
}

/**
471
 * e1000_receive_skb - helper function to handle Rx indications
472 473 474 475 476 477
 * @adapter: board private structure
 * @status: descriptor status field as written by hardware
 * @vlan: descriptor vlan field as written by hardware (no le/be conversion)
 * @skb: pointer to sk_buff to be indicated to stack
 **/
static void e1000_receive_skb(struct e1000_adapter *adapter,
478
			      struct net_device *netdev, struct sk_buff *skb,
A
Al Viro 已提交
479
			      u8 status, __le16 vlan)
480
{
J
Jeff Kirsher 已提交
481
	u16 tag = le16_to_cpu(vlan);
482 483
	skb->protocol = eth_type_trans(skb, netdev);

J
Jeff Kirsher 已提交
484 485 486 487
	if (status & E1000_RXD_STAT_VP)
		__vlan_hwaccel_put_tag(skb, tag);

	napi_gro_receive(&adapter->napi, skb);
488 489 490
}

/**
491
 * e1000_rx_checksum - Receive Checksum Offload
492 493 494 495
 * @adapter: board private structure
 * @status_err: receive descriptor status and error fields
 * @csum: receive descriptor csum field
 * @sk_buff: socket buffer with received data
496 497
 **/
static void e1000_rx_checksum(struct e1000_adapter *adapter, u32 status_err,
498
			      __le16 csum, struct sk_buff *skb)
499 500 501
{
	u16 status = (u16)status_err;
	u8 errors = (u8)(status_err >> 24);
502 503

	skb_checksum_none_assert(skb);
504

505 506 507 508
	/* Rx checksum disabled */
	if (!(adapter->netdev->features & NETIF_F_RXCSUM))
		return;

509 510 511
	/* Ignore Checksum bit is set */
	if (status & E1000_RXD_STAT_IXSM)
		return;
512

513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528
	/* TCP/UDP checksum error bit is set */
	if (errors & E1000_RXD_ERR_TCPE) {
		/* let the stack verify checksum errors */
		adapter->hw_csum_err++;
		return;
	}

	/* TCP/UDP Checksum has not been calculated */
	if (!(status & (E1000_RXD_STAT_TCPCS | E1000_RXD_STAT_UDPCS)))
		return;

	/* It must be a TCP or UDP packet with a valid checksum */
	if (status & E1000_RXD_STAT_TCPCS) {
		/* TCP checksum is good */
		skb->ip_summed = CHECKSUM_UNNECESSARY;
	} else {
529 530 531
		/*
		 * IP fragment with UDP payload
		 * Hardware complements the payload checksum, so we undo it
532 533
		 * and then put the value in host order for further stack use.
		 */
534
		__sum16 sum = (__force __sum16)swab16((__force u16)csum);
A
Al Viro 已提交
535
		skb->csum = csum_unfold(~sum);
536 537 538 539 540
		skb->ip_summed = CHECKSUM_COMPLETE;
	}
	adapter->hw_csum_good++;
}

541
static void e1000e_update_rdt_wa(struct e1000_ring *rx_ring, unsigned int i)
542
{
543
	struct e1000_adapter *adapter = rx_ring->adapter;
544
	struct e1000_hw *hw = &adapter->hw;
545 546 547
	s32 ret_val = __ew32_prepare(hw);

	writel(i, rx_ring->tail);
548

549
	if (unlikely(!ret_val && (i != readl(rx_ring->tail)))) {
550 551 552 553 554 555 556
		u32 rctl = er32(RCTL);
		ew32(RCTL, rctl & ~E1000_RCTL_EN);
		e_err("ME firmware caused invalid RDT - resetting\n");
		schedule_work(&adapter->reset_task);
	}
}

557
static void e1000e_update_tdt_wa(struct e1000_ring *tx_ring, unsigned int i)
558
{
559
	struct e1000_adapter *adapter = tx_ring->adapter;
560
	struct e1000_hw *hw = &adapter->hw;
561
	s32 ret_val = __ew32_prepare(hw);
562

563 564 565
	writel(i, tx_ring->tail);

	if (unlikely(!ret_val && (i != readl(tx_ring->tail)))) {
566 567 568 569 570 571 572
		u32 tctl = er32(TCTL);
		ew32(TCTL, tctl & ~E1000_TCTL_EN);
		e_err("ME firmware caused invalid TDT - resetting\n");
		schedule_work(&adapter->reset_task);
	}
}

573
/**
574
 * e1000_alloc_rx_buffers - Replace used receive buffers
575
 * @rx_ring: Rx descriptor ring
576
 **/
577
static void e1000_alloc_rx_buffers(struct e1000_ring *rx_ring,
578
				   int cleaned_count, gfp_t gfp)
579
{
580
	struct e1000_adapter *adapter = rx_ring->adapter;
581 582
	struct net_device *netdev = adapter->netdev;
	struct pci_dev *pdev = adapter->pdev;
583
	union e1000_rx_desc_extended *rx_desc;
584 585 586
	struct e1000_buffer *buffer_info;
	struct sk_buff *skb;
	unsigned int i;
587
	unsigned int bufsz = adapter->rx_buffer_len;
588 589 590 591 592 593 594 595 596 597 598

	i = rx_ring->next_to_use;
	buffer_info = &rx_ring->buffer_info[i];

	while (cleaned_count--) {
		skb = buffer_info->skb;
		if (skb) {
			skb_trim(skb, 0);
			goto map_skb;
		}

599
		skb = __netdev_alloc_skb_ip_align(netdev, bufsz, gfp);
600 601 602 603 604 605 606 607
		if (!skb) {
			/* Better luck next round */
			adapter->alloc_rx_buff_failed++;
			break;
		}

		buffer_info->skb = skb;
map_skb:
608
		buffer_info->dma = dma_map_single(&pdev->dev, skb->data,
609
						  adapter->rx_buffer_len,
610 611
						  DMA_FROM_DEVICE);
		if (dma_mapping_error(&pdev->dev, buffer_info->dma)) {
612
			dev_err(&pdev->dev, "Rx DMA map failed\n");
613 614 615 616
			adapter->rx_dma_failed++;
			break;
		}

617 618
		rx_desc = E1000_RX_DESC_EXT(*rx_ring, i);
		rx_desc->read.buffer_addr = cpu_to_le64(buffer_info->dma);
619

620 621 622 623 624 625 626 627
		if (unlikely(!(i & (E1000_RX_BUFFER_WRITE - 1)))) {
			/*
			 * 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();
628
			if (adapter->flags2 & FLAG2_PCIM2PCI_ARBITER_WA)
629
				e1000e_update_rdt_wa(rx_ring, i);
630
			else
631
				writel(i, rx_ring->tail);
632
		}
633 634 635 636 637 638
		i++;
		if (i == rx_ring->count)
			i = 0;
		buffer_info = &rx_ring->buffer_info[i];
	}

639
	rx_ring->next_to_use = i;
640 641 642 643
}

/**
 * e1000_alloc_rx_buffers_ps - Replace used receive buffers; packet split
644
 * @rx_ring: Rx descriptor ring
645
 **/
646
static void e1000_alloc_rx_buffers_ps(struct e1000_ring *rx_ring,
647
				      int cleaned_count, gfp_t gfp)
648
{
649
	struct e1000_adapter *adapter = rx_ring->adapter;
650 651 652 653 654 655 656 657 658 659 660 661 662 663 664
	struct net_device *netdev = adapter->netdev;
	struct pci_dev *pdev = adapter->pdev;
	union e1000_rx_desc_packet_split *rx_desc;
	struct e1000_buffer *buffer_info;
	struct e1000_ps_page *ps_page;
	struct sk_buff *skb;
	unsigned int i, j;

	i = rx_ring->next_to_use;
	buffer_info = &rx_ring->buffer_info[i];

	while (cleaned_count--) {
		rx_desc = E1000_RX_DESC_PS(*rx_ring, i);

		for (j = 0; j < PS_PAGE_BUFFERS; j++) {
A
Auke Kok 已提交
665 666 667
			ps_page = &buffer_info->ps_pages[j];
			if (j >= adapter->rx_ps_pages) {
				/* all unused desc entries get hw null ptr */
668 669
				rx_desc->read.buffer_addr[j + 1] =
				    ~cpu_to_le64(0);
A
Auke Kok 已提交
670 671 672
				continue;
			}
			if (!ps_page->page) {
673
				ps_page->page = alloc_page(gfp);
674
				if (!ps_page->page) {
A
Auke Kok 已提交
675 676 677
					adapter->alloc_rx_buff_failed++;
					goto no_buffers;
				}
678 679 680 681 682 683
				ps_page->dma = dma_map_page(&pdev->dev,
							    ps_page->page,
							    0, PAGE_SIZE,
							    DMA_FROM_DEVICE);
				if (dma_mapping_error(&pdev->dev,
						      ps_page->dma)) {
A
Auke Kok 已提交
684
					dev_err(&adapter->pdev->dev,
685
						"Rx DMA page map failed\n");
A
Auke Kok 已提交
686 687
					adapter->rx_dma_failed++;
					goto no_buffers;
688 689
				}
			}
A
Auke Kok 已提交
690 691 692 693 694
			/*
			 * Refresh the desc even if buffer_addrs
			 * didn't change because each write-back
			 * erases this info.
			 */
695 696
			rx_desc->read.buffer_addr[j + 1] =
			    cpu_to_le64(ps_page->dma);
697 698
		}

699 700 701
		skb = __netdev_alloc_skb_ip_align(netdev,
						  adapter->rx_ps_bsize0,
						  gfp);
702 703 704 705 706 707 708

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

		buffer_info->skb = skb;
709
		buffer_info->dma = dma_map_single(&pdev->dev, skb->data,
710
						  adapter->rx_ps_bsize0,
711 712
						  DMA_FROM_DEVICE);
		if (dma_mapping_error(&pdev->dev, buffer_info->dma)) {
713
			dev_err(&pdev->dev, "Rx DMA map failed\n");
714 715 716 717 718 719 720 721 722
			adapter->rx_dma_failed++;
			/* cleanup skb */
			dev_kfree_skb_any(skb);
			buffer_info->skb = NULL;
			break;
		}

		rx_desc->read.buffer_addr[0] = cpu_to_le64(buffer_info->dma);

723 724 725 726 727 728 729 730
		if (unlikely(!(i & (E1000_RX_BUFFER_WRITE - 1)))) {
			/*
			 * 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();
731
			if (adapter->flags2 & FLAG2_PCIM2PCI_ARBITER_WA)
732
				e1000e_update_rdt_wa(rx_ring, i << 1);
733
			else
734
				writel(i << 1, rx_ring->tail);
735 736
		}

737 738 739 740 741 742 743
		i++;
		if (i == rx_ring->count)
			i = 0;
		buffer_info = &rx_ring->buffer_info[i];
	}

no_buffers:
744
	rx_ring->next_to_use = i;
745 746
}

747 748
/**
 * e1000_alloc_jumbo_rx_buffers - Replace used jumbo receive buffers
749
 * @rx_ring: Rx descriptor ring
750 751 752
 * @cleaned_count: number of buffers to allocate this pass
 **/

753
static void e1000_alloc_jumbo_rx_buffers(struct e1000_ring *rx_ring,
754
					 int cleaned_count, gfp_t gfp)
755
{
756
	struct e1000_adapter *adapter = rx_ring->adapter;
757 758
	struct net_device *netdev = adapter->netdev;
	struct pci_dev *pdev = adapter->pdev;
759
	union e1000_rx_desc_extended *rx_desc;
760 761 762
	struct e1000_buffer *buffer_info;
	struct sk_buff *skb;
	unsigned int i;
763
	unsigned int bufsz = 256 - 16 /* for skb_reserve */;
764 765 766 767 768 769 770 771 772 773 774

	i = rx_ring->next_to_use;
	buffer_info = &rx_ring->buffer_info[i];

	while (cleaned_count--) {
		skb = buffer_info->skb;
		if (skb) {
			skb_trim(skb, 0);
			goto check_page;
		}

775
		skb = __netdev_alloc_skb_ip_align(netdev, bufsz, gfp);
776 777 778 779 780 781 782 783 784 785
		if (unlikely(!skb)) {
			/* Better luck next round */
			adapter->alloc_rx_buff_failed++;
			break;
		}

		buffer_info->skb = skb;
check_page:
		/* allocate a new page if necessary */
		if (!buffer_info->page) {
786
			buffer_info->page = alloc_page(gfp);
787 788 789 790 791 792 793
			if (unlikely(!buffer_info->page)) {
				adapter->alloc_rx_buff_failed++;
				break;
			}
		}

		if (!buffer_info->dma)
794
			buffer_info->dma = dma_map_page(&pdev->dev,
795 796
			                                buffer_info->page, 0,
			                                PAGE_SIZE,
797
							DMA_FROM_DEVICE);
798

799 800
		rx_desc = E1000_RX_DESC_EXT(*rx_ring, i);
		rx_desc->read.buffer_addr = cpu_to_le64(buffer_info->dma);
801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816

		if (unlikely(++i == rx_ring->count))
			i = 0;
		buffer_info = &rx_ring->buffer_info[i];
	}

	if (likely(rx_ring->next_to_use != i)) {
		rx_ring->next_to_use = i;
		if (unlikely(i-- == 0))
			i = (rx_ring->count - 1);

		/* 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();
817
		if (adapter->flags2 & FLAG2_PCIM2PCI_ARBITER_WA)
818
			e1000e_update_rdt_wa(rx_ring, i);
819
		else
820
			writel(i, rx_ring->tail);
821 822 823
	}
}

824 825 826 827 828 829 830
static inline void e1000_rx_hash(struct net_device *netdev, __le32 rss,
				 struct sk_buff *skb)
{
	if (netdev->features & NETIF_F_RXHASH)
		skb->rxhash = le32_to_cpu(rss);
}

831
/**
832 833
 * e1000_clean_rx_irq - Send received data up the network stack
 * @rx_ring: Rx descriptor ring
834 835 836 837
 *
 * the return value indicates whether actual cleaning was done, there
 * is no guarantee that everything was cleaned
 **/
838 839
static bool e1000_clean_rx_irq(struct e1000_ring *rx_ring, int *work_done,
			       int work_to_do)
840
{
841
	struct e1000_adapter *adapter = rx_ring->adapter;
842 843
	struct net_device *netdev = adapter->netdev;
	struct pci_dev *pdev = adapter->pdev;
844
	struct e1000_hw *hw = &adapter->hw;
845
	union e1000_rx_desc_extended *rx_desc, *next_rxd;
846
	struct e1000_buffer *buffer_info, *next_buffer;
847
	u32 length, staterr;
848 849
	unsigned int i;
	int cleaned_count = 0;
850
	bool cleaned = false;
851 852 853
	unsigned int total_rx_bytes = 0, total_rx_packets = 0;

	i = rx_ring->next_to_clean;
854 855
	rx_desc = E1000_RX_DESC_EXT(*rx_ring, i);
	staterr = le32_to_cpu(rx_desc->wb.upper.status_error);
856 857
	buffer_info = &rx_ring->buffer_info[i];

858
	while (staterr & E1000_RXD_STAT_DD) {
859 860 861 862 863
		struct sk_buff *skb;

		if (*work_done >= work_to_do)
			break;
		(*work_done)++;
864
		rmb();	/* read descriptor and rx_buffer_info after status DD */
865 866 867 868 869 870 871 872 873

		skb = buffer_info->skb;
		buffer_info->skb = NULL;

		prefetch(skb->data - NET_IP_ALIGN);

		i++;
		if (i == rx_ring->count)
			i = 0;
874
		next_rxd = E1000_RX_DESC_EXT(*rx_ring, i);
875 876 877 878
		prefetch(next_rxd);

		next_buffer = &rx_ring->buffer_info[i];

879
		cleaned = true;
880
		cleaned_count++;
881
		dma_unmap_single(&pdev->dev,
882 883
				 buffer_info->dma,
				 adapter->rx_buffer_len,
884
				 DMA_FROM_DEVICE);
885 886
		buffer_info->dma = 0;

887
		length = le16_to_cpu(rx_desc->wb.upper.length);
888

889 890 891 892 893 894 895
		/*
		 * !EOP means multiple descriptors were used to store a single
		 * packet, if that's the case we need to toss it.  In fact, we
		 * need to toss every packet with the EOP bit clear and the
		 * next frame that _does_ have the EOP bit set, as it is by
		 * definition only a frame fragment
		 */
896
		if (unlikely(!(staterr & E1000_RXD_STAT_EOP)))
897 898 899
			adapter->flags2 |= FLAG2_IS_DISCARDING;

		if (adapter->flags2 & FLAG2_IS_DISCARDING) {
900
			/* All receives must fit into a single buffer */
901
			e_dbg("Receive packet consumed multiple buffers\n");
902 903
			/* recycle */
			buffer_info->skb = skb;
904
			if (staterr & E1000_RXD_STAT_EOP)
905
				adapter->flags2 &= ~FLAG2_IS_DISCARDING;
906 907 908
			goto next_desc;
		}

B
Ben Greear 已提交
909 910
		if (unlikely((staterr & E1000_RXDEXT_ERR_FRAME_ERR_MASK) &&
			     !(netdev->features & NETIF_F_RXALL))) {
911 912 913 914 915
			/* recycle */
			buffer_info->skb = skb;
			goto next_desc;
		}

J
Jeff Kirsher 已提交
916
		/* adjust length to remove Ethernet CRC */
B
Ben Greear 已提交
917 918 919 920 921 922 923 924 925 926
		if (!(adapter->flags2 & FLAG2_CRC_STRIPPING)) {
			/* If configured to store CRC, don't subtract FCS,
			 * but keep the FCS bytes out of the total_rx_bytes
			 * counter
			 */
			if (netdev->features & NETIF_F_RXFCS)
				total_rx_bytes -= 4;
			else
				length -= 4;
		}
J
Jeff Kirsher 已提交
927

928 929 930
		total_rx_bytes += length;
		total_rx_packets++;

931 932
		/*
		 * code added for copybreak, this should improve
933
		 * performance for small packets with large amounts
934 935
		 * of reassembly being done in the stack
		 */
936 937
		if (length < copybreak) {
			struct sk_buff *new_skb =
938
			    netdev_alloc_skb_ip_align(netdev, length);
939
			if (new_skb) {
940 941 942 943 944 945
				skb_copy_to_linear_data_offset(new_skb,
							       -NET_IP_ALIGN,
							       (skb->data -
								NET_IP_ALIGN),
							       (length +
								NET_IP_ALIGN));
946 947 948 949 950 951 952 953 954 955
				/* save the skb in buffer_info as good */
				buffer_info->skb = skb;
				skb = new_skb;
			}
			/* else just continue with the old one */
		}
		/* end copybreak code */
		skb_put(skb, length);

		/* Receive Checksum Offload */
956
		e1000_rx_checksum(adapter, staterr,
957
				  rx_desc->wb.lower.hi_dword.csum_ip.csum, skb);
958

959 960
		e1000_rx_hash(netdev, rx_desc->wb.lower.hi_dword.rss, skb);

961 962
		e1000_receive_skb(adapter, netdev, skb, staterr,
				  rx_desc->wb.upper.vlan);
963 964

next_desc:
965
		rx_desc->wb.upper.status_error &= cpu_to_le32(~0xFF);
966 967 968

		/* return some buffers to hardware, one at a time is too slow */
		if (cleaned_count >= E1000_RX_BUFFER_WRITE) {
969
			adapter->alloc_rx_buf(rx_ring, cleaned_count,
970
					      GFP_ATOMIC);
971 972 973 974 975 976
			cleaned_count = 0;
		}

		/* use prefetched values */
		rx_desc = next_rxd;
		buffer_info = next_buffer;
977 978

		staterr = le32_to_cpu(rx_desc->wb.upper.status_error);
979 980 981 982 983
	}
	rx_ring->next_to_clean = i;

	cleaned_count = e1000_desc_unused(rx_ring);
	if (cleaned_count)
984
		adapter->alloc_rx_buf(rx_ring, cleaned_count, GFP_ATOMIC);
985 986

	adapter->total_rx_bytes += total_rx_bytes;
987
	adapter->total_rx_packets += total_rx_packets;
988 989 990
	return cleaned;
}

991 992
static void e1000_put_txbuf(struct e1000_ring *tx_ring,
			    struct e1000_buffer *buffer_info)
993
{
994 995
	struct e1000_adapter *adapter = tx_ring->adapter;

996 997
	if (buffer_info->dma) {
		if (buffer_info->mapped_as_page)
998 999
			dma_unmap_page(&adapter->pdev->dev, buffer_info->dma,
				       buffer_info->length, DMA_TO_DEVICE);
1000
		else
1001 1002
			dma_unmap_single(&adapter->pdev->dev, buffer_info->dma,
					 buffer_info->length, DMA_TO_DEVICE);
1003 1004
		buffer_info->dma = 0;
	}
1005 1006 1007 1008
	if (buffer_info->skb) {
		dev_kfree_skb_any(buffer_info->skb);
		buffer_info->skb = NULL;
	}
1009
	buffer_info->time_stamp = 0;
1010 1011
}

1012
static void e1000_print_hw_hang(struct work_struct *work)
1013
{
1014 1015 1016
	struct e1000_adapter *adapter = container_of(work,
	                                             struct e1000_adapter,
	                                             print_hang_task);
1017
	struct net_device *netdev = adapter->netdev;
1018 1019 1020 1021
	struct e1000_ring *tx_ring = adapter->tx_ring;
	unsigned int i = tx_ring->next_to_clean;
	unsigned int eop = tx_ring->buffer_info[i].next_to_watch;
	struct e1000_tx_desc *eop_desc = E1000_TX_DESC(*tx_ring, eop);
1022 1023 1024 1025
	struct e1000_hw *hw = &adapter->hw;
	u16 phy_status, phy_1000t_status, phy_ext_status;
	u16 pci_status;

1026 1027 1028
	if (test_bit(__E1000_DOWN, &adapter->state))
		return;

1029 1030
	if (!adapter->tx_hang_recheck &&
	    (adapter->flags2 & FLAG2_DMA_BURST)) {
M
Matthew Vick 已提交
1031 1032
		/*
		 * May be block on write-back, flush and detect again
1033 1034 1035 1036 1037
		 * flush pending descriptor writebacks to memory
		 */
		ew32(TIDV, adapter->tx_int_delay | E1000_TIDV_FPD);
		/* execute the writes immediately */
		e1e_flush();
1038 1039 1040 1041 1042 1043 1044
		/*
		 * Due to rare timing issues, write to TIDV again to ensure
		 * the write is successful
		 */
		ew32(TIDV, adapter->tx_int_delay | E1000_TIDV_FPD);
		/* execute the writes immediately */
		e1e_flush();
1045 1046 1047 1048 1049 1050 1051
		adapter->tx_hang_recheck = true;
		return;
	}
	/* Real hang detected */
	adapter->tx_hang_recheck = false;
	netif_stop_queue(netdev);

1052 1053 1054
	e1e_rphy(hw, PHY_STATUS, &phy_status);
	e1e_rphy(hw, PHY_1000T_STATUS, &phy_1000t_status);
	e1e_rphy(hw, PHY_EXT_STATUS, &phy_ext_status);
1055

1056 1057 1058 1059
	pci_read_config_word(adapter->pdev, PCI_STATUS, &pci_status);

	/* detected Hardware unit hang */
	e_err("Detected Hardware Unit Hang:\n"
1060 1061 1062 1063 1064 1065 1066 1067
	      "  TDH                  <%x>\n"
	      "  TDT                  <%x>\n"
	      "  next_to_use          <%x>\n"
	      "  next_to_clean        <%x>\n"
	      "buffer_info[next_to_clean]:\n"
	      "  time_stamp           <%lx>\n"
	      "  next_to_watch        <%x>\n"
	      "  jiffies              <%lx>\n"
1068 1069 1070 1071 1072 1073
	      "  next_to_watch.status <%x>\n"
	      "MAC Status             <%x>\n"
	      "PHY Status             <%x>\n"
	      "PHY 1000BASE-T Status  <%x>\n"
	      "PHY Extended Status    <%x>\n"
	      "PCI Status             <%x>\n",
1074 1075
	      readl(tx_ring->head),
	      readl(tx_ring->tail),
1076 1077 1078 1079 1080
	      tx_ring->next_to_use,
	      tx_ring->next_to_clean,
	      tx_ring->buffer_info[eop].time_stamp,
	      eop,
	      jiffies,
1081 1082 1083 1084 1085 1086
	      eop_desc->upper.fields.status,
	      er32(STATUS),
	      phy_status,
	      phy_1000t_status,
	      phy_ext_status,
	      pci_status);
1087 1088 1089 1090
}

/**
 * e1000_clean_tx_irq - Reclaim resources after transmit completes
1091
 * @tx_ring: Tx descriptor ring
1092 1093 1094 1095
 *
 * the return value indicates whether actual cleaning was done, there
 * is no guarantee that everything was cleaned
 **/
1096
static bool e1000_clean_tx_irq(struct e1000_ring *tx_ring)
1097
{
1098
	struct e1000_adapter *adapter = tx_ring->adapter;
1099 1100 1101 1102 1103 1104 1105
	struct net_device *netdev = adapter->netdev;
	struct e1000_hw *hw = &adapter->hw;
	struct e1000_tx_desc *tx_desc, *eop_desc;
	struct e1000_buffer *buffer_info;
	unsigned int i, eop;
	unsigned int count = 0;
	unsigned int total_tx_bytes = 0, total_tx_packets = 0;
1106
	unsigned int bytes_compl = 0, pkts_compl = 0;
1107 1108 1109 1110 1111

	i = tx_ring->next_to_clean;
	eop = tx_ring->buffer_info[i].next_to_watch;
	eop_desc = E1000_TX_DESC(*tx_ring, eop);

1112 1113
	while ((eop_desc->upper.data & cpu_to_le32(E1000_TXD_STAT_DD)) &&
	       (count < tx_ring->count)) {
1114
		bool cleaned = false;
1115
		rmb(); /* read buffer_info after eop_desc */
1116
		for (; !cleaned; count++) {
1117 1118 1119 1120 1121
			tx_desc = E1000_TX_DESC(*tx_ring, i);
			buffer_info = &tx_ring->buffer_info[i];
			cleaned = (i == eop);

			if (cleaned) {
1122 1123
				total_tx_packets += buffer_info->segs;
				total_tx_bytes += buffer_info->bytecount;
1124 1125 1126 1127
				if (buffer_info->skb) {
					bytes_compl += buffer_info->skb->len;
					pkts_compl++;
				}
1128 1129
			}

1130
			e1000_put_txbuf(tx_ring, buffer_info);
1131 1132 1133 1134 1135 1136 1137
			tx_desc->upper.data = 0;

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

1138 1139
		if (i == tx_ring->next_to_use)
			break;
1140 1141 1142 1143 1144 1145
		eop = tx_ring->buffer_info[i].next_to_watch;
		eop_desc = E1000_TX_DESC(*tx_ring, eop);
	}

	tx_ring->next_to_clean = i;

1146 1147
	netdev_completed_queue(netdev, pkts_compl, bytes_compl);

1148
#define TX_WAKE_THRESHOLD 32
1149 1150
	if (count && netif_carrier_ok(netdev) &&
	    e1000_desc_unused(tx_ring) >= TX_WAKE_THRESHOLD) {
1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163
		/* Make sure that anybody stopping the queue after this
		 * sees the new next_to_clean.
		 */
		smp_mb();

		if (netif_queue_stopped(netdev) &&
		    !(test_bit(__E1000_DOWN, &adapter->state))) {
			netif_wake_queue(netdev);
			++adapter->restart_queue;
		}
	}

	if (adapter->detect_tx_hung) {
1164 1165 1166 1167
		/*
		 * Detect a transmit hang in hardware, this serializes the
		 * check with the clearing of time_stamp and movement of i
		 */
1168
		adapter->detect_tx_hung = false;
1169 1170
		if (tx_ring->buffer_info[i].time_stamp &&
		    time_after(jiffies, tx_ring->buffer_info[i].time_stamp
1171
			       + (adapter->tx_timeout_factor * HZ)) &&
1172
		    !(er32(STATUS) & E1000_STATUS_TXOFF))
1173
			schedule_work(&adapter->print_hang_task);
1174 1175
		else
			adapter->tx_hang_recheck = false;
1176 1177 1178
	}
	adapter->total_tx_bytes += total_tx_bytes;
	adapter->total_tx_packets += total_tx_packets;
1179
	return count < tx_ring->count;
1180 1181 1182 1183
}

/**
 * e1000_clean_rx_irq_ps - Send received data up the network stack; packet split
1184
 * @rx_ring: Rx descriptor ring
1185 1186 1187 1188
 *
 * the return value indicates whether actual cleaning was done, there
 * is no guarantee that everything was cleaned
 **/
1189 1190
static bool e1000_clean_rx_irq_ps(struct e1000_ring *rx_ring, int *work_done,
				  int work_to_do)
1191
{
1192
	struct e1000_adapter *adapter = rx_ring->adapter;
1193
	struct e1000_hw *hw = &adapter->hw;
1194 1195 1196 1197 1198 1199 1200 1201 1202
	union e1000_rx_desc_packet_split *rx_desc, *next_rxd;
	struct net_device *netdev = adapter->netdev;
	struct pci_dev *pdev = adapter->pdev;
	struct e1000_buffer *buffer_info, *next_buffer;
	struct e1000_ps_page *ps_page;
	struct sk_buff *skb;
	unsigned int i, j;
	u32 length, staterr;
	int cleaned_count = 0;
1203
	bool cleaned = false;
1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215
	unsigned int total_rx_bytes = 0, total_rx_packets = 0;

	i = rx_ring->next_to_clean;
	rx_desc = E1000_RX_DESC_PS(*rx_ring, i);
	staterr = le32_to_cpu(rx_desc->wb.middle.status_error);
	buffer_info = &rx_ring->buffer_info[i];

	while (staterr & E1000_RXD_STAT_DD) {
		if (*work_done >= work_to_do)
			break;
		(*work_done)++;
		skb = buffer_info->skb;
1216
		rmb();	/* read descriptor and rx_buffer_info after status DD */
1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228

		/* in the packet split case this is header only */
		prefetch(skb->data - NET_IP_ALIGN);

		i++;
		if (i == rx_ring->count)
			i = 0;
		next_rxd = E1000_RX_DESC_PS(*rx_ring, i);
		prefetch(next_rxd);

		next_buffer = &rx_ring->buffer_info[i];

1229
		cleaned = true;
1230
		cleaned_count++;
1231
		dma_unmap_single(&pdev->dev, buffer_info->dma,
1232
				 adapter->rx_ps_bsize0, DMA_FROM_DEVICE);
1233 1234
		buffer_info->dma = 0;

1235
		/* see !EOP comment in other Rx routine */
1236 1237 1238 1239
		if (!(staterr & E1000_RXD_STAT_EOP))
			adapter->flags2 |= FLAG2_IS_DISCARDING;

		if (adapter->flags2 & FLAG2_IS_DISCARDING) {
1240
			e_dbg("Packet Split buffers didn't pick up the full packet\n");
1241
			dev_kfree_skb_irq(skb);
1242 1243
			if (staterr & E1000_RXD_STAT_EOP)
				adapter->flags2 &= ~FLAG2_IS_DISCARDING;
1244 1245 1246
			goto next_desc;
		}

B
Ben Greear 已提交
1247 1248
		if (unlikely((staterr & E1000_RXDEXT_ERR_FRAME_ERR_MASK) &&
			     !(netdev->features & NETIF_F_RXALL))) {
1249 1250 1251 1252 1253 1254 1255
			dev_kfree_skb_irq(skb);
			goto next_desc;
		}

		length = le16_to_cpu(rx_desc->wb.middle.length0);

		if (!length) {
1256
			e_dbg("Last part of the packet spanning multiple descriptors\n");
1257 1258 1259 1260 1261 1262 1263 1264
			dev_kfree_skb_irq(skb);
			goto next_desc;
		}

		/* Good Receive */
		skb_put(skb, length);

		{
1265 1266 1267 1268 1269
			/*
			 * this looks ugly, but it seems compiler issues make
			 * it more efficient than reusing j
			 */
			int l1 = le16_to_cpu(rx_desc->wb.upper.length[0]);
1270

1271
			/*
1272 1273 1274 1275
			 * page alloc/put takes too long and effects small
			 * packet throughput, so unsplit small packets and
			 * save the alloc/put only valid in softirq (napi)
			 * context to call kmap_*
1276
			 */
1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291
			if (l1 && (l1 <= copybreak) &&
			    ((length + l1) <= adapter->rx_ps_bsize0)) {
				u8 *vaddr;

				ps_page = &buffer_info->ps_pages[0];

				/*
				 * there is no documentation about how to call
				 * kmap_atomic, so we can't hold the mapping
				 * very long
				 */
				dma_sync_single_for_cpu(&pdev->dev,
							ps_page->dma,
							PAGE_SIZE,
							DMA_FROM_DEVICE);
1292
				vaddr = kmap_atomic(ps_page->page);
1293
				memcpy(skb_tail_pointer(skb), vaddr, l1);
1294
				kunmap_atomic(vaddr);
1295 1296 1297 1298 1299 1300
				dma_sync_single_for_device(&pdev->dev,
							   ps_page->dma,
							   PAGE_SIZE,
							   DMA_FROM_DEVICE);

				/* remove the CRC */
B
Ben Greear 已提交
1301 1302 1303 1304
				if (!(adapter->flags2 & FLAG2_CRC_STRIPPING)) {
					if (!(netdev->features & NETIF_F_RXFCS))
						l1 -= 4;
				}
1305 1306 1307 1308

				skb_put(skb, l1);
				goto copydone;
			} /* if */
1309 1310 1311 1312 1313 1314 1315
		}

		for (j = 0; j < PS_PAGE_BUFFERS; j++) {
			length = le16_to_cpu(rx_desc->wb.upper.length[j]);
			if (!length)
				break;

A
Auke Kok 已提交
1316
			ps_page = &buffer_info->ps_pages[j];
1317 1318
			dma_unmap_page(&pdev->dev, ps_page->dma, PAGE_SIZE,
				       DMA_FROM_DEVICE);
1319 1320 1321 1322 1323
			ps_page->dma = 0;
			skb_fill_page_desc(skb, j, ps_page->page, 0, length);
			ps_page->page = NULL;
			skb->len += length;
			skb->data_len += length;
1324
			skb->truesize += PAGE_SIZE;
1325 1326
		}

J
Jeff Kirsher 已提交
1327 1328 1329
		/* strip the ethernet crc, problem is we're using pages now so
		 * this whole operation can get a little cpu intensive
		 */
B
Ben Greear 已提交
1330 1331 1332 1333
		if (!(adapter->flags2 & FLAG2_CRC_STRIPPING)) {
			if (!(netdev->features & NETIF_F_RXFCS))
				pskb_trim(skb, skb->len - 4);
		}
J
Jeff Kirsher 已提交
1334

1335 1336 1337 1338
copydone:
		total_rx_bytes += skb->len;
		total_rx_packets++;

1339 1340
		e1000_rx_checksum(adapter, staterr,
				  rx_desc->wb.lower.hi_dword.csum_ip.csum, skb);
1341

1342 1343
		e1000_rx_hash(netdev, rx_desc->wb.lower.hi_dword.rss, skb);

1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356
		if (rx_desc->wb.upper.header_status &
			   cpu_to_le16(E1000_RXDPS_HDRSTAT_HDRSP))
			adapter->rx_hdr_split++;

		e1000_receive_skb(adapter, netdev, skb,
				  staterr, rx_desc->wb.middle.vlan);

next_desc:
		rx_desc->wb.middle.status_error &= cpu_to_le32(~0xFF);
		buffer_info->skb = NULL;

		/* return some buffers to hardware, one at a time is too slow */
		if (cleaned_count >= E1000_RX_BUFFER_WRITE) {
1357
			adapter->alloc_rx_buf(rx_ring, cleaned_count,
1358
					      GFP_ATOMIC);
1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371
			cleaned_count = 0;
		}

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

		staterr = le32_to_cpu(rx_desc->wb.middle.status_error);
	}
	rx_ring->next_to_clean = i;

	cleaned_count = e1000_desc_unused(rx_ring);
	if (cleaned_count)
1372
		adapter->alloc_rx_buf(rx_ring, cleaned_count, GFP_ATOMIC);
1373 1374

	adapter->total_rx_bytes += total_rx_bytes;
1375
	adapter->total_rx_packets += total_rx_packets;
1376 1377 1378
	return cleaned;
}

1379 1380 1381 1382 1383 1384 1385 1386 1387
/**
 * e1000_consume_page - helper function
 **/
static void e1000_consume_page(struct e1000_buffer *bi, struct sk_buff *skb,
                               u16 length)
{
	bi->page = NULL;
	skb->len += length;
	skb->data_len += length;
1388
	skb->truesize += PAGE_SIZE;
1389 1390 1391 1392 1393 1394 1395 1396 1397
}

/**
 * e1000_clean_jumbo_rx_irq - Send received data up the network stack; legacy
 * @adapter: board private structure
 *
 * the return value indicates whether actual cleaning was done, there
 * is no guarantee that everything was cleaned
 **/
1398 1399
static bool e1000_clean_jumbo_rx_irq(struct e1000_ring *rx_ring, int *work_done,
				     int work_to_do)
1400
{
1401
	struct e1000_adapter *adapter = rx_ring->adapter;
1402 1403
	struct net_device *netdev = adapter->netdev;
	struct pci_dev *pdev = adapter->pdev;
1404
	union e1000_rx_desc_extended *rx_desc, *next_rxd;
1405
	struct e1000_buffer *buffer_info, *next_buffer;
1406
	u32 length, staterr;
1407 1408 1409 1410 1411 1412
	unsigned int i;
	int cleaned_count = 0;
	bool cleaned = false;
	unsigned int total_rx_bytes=0, total_rx_packets=0;

	i = rx_ring->next_to_clean;
1413 1414
	rx_desc = E1000_RX_DESC_EXT(*rx_ring, i);
	staterr = le32_to_cpu(rx_desc->wb.upper.status_error);
1415 1416
	buffer_info = &rx_ring->buffer_info[i];

1417
	while (staterr & E1000_RXD_STAT_DD) {
1418 1419 1420 1421 1422
		struct sk_buff *skb;

		if (*work_done >= work_to_do)
			break;
		(*work_done)++;
1423
		rmb();	/* read descriptor and rx_buffer_info after status DD */
1424 1425 1426 1427 1428 1429 1430

		skb = buffer_info->skb;
		buffer_info->skb = NULL;

		++i;
		if (i == rx_ring->count)
			i = 0;
1431
		next_rxd = E1000_RX_DESC_EXT(*rx_ring, i);
1432 1433 1434 1435 1436 1437
		prefetch(next_rxd);

		next_buffer = &rx_ring->buffer_info[i];

		cleaned = true;
		cleaned_count++;
1438 1439
		dma_unmap_page(&pdev->dev, buffer_info->dma, PAGE_SIZE,
			       DMA_FROM_DEVICE);
1440 1441
		buffer_info->dma = 0;

1442
		length = le16_to_cpu(rx_desc->wb.upper.length);
1443 1444

		/* errors is only valid for DD + EOP descriptors */
1445
		if (unlikely((staterr & E1000_RXD_STAT_EOP) &&
B
Ben Greear 已提交
1446 1447
			     ((staterr & E1000_RXDEXT_ERR_FRAME_ERR_MASK) &&
			      !(netdev->features & NETIF_F_RXALL)))) {
1448 1449 1450 1451 1452 1453 1454
			/* recycle both page and skb */
			buffer_info->skb = skb;
			/* an error means any chain goes out the window too */
			if (rx_ring->rx_skb_top)
				dev_kfree_skb_irq(rx_ring->rx_skb_top);
			rx_ring->rx_skb_top = NULL;
			goto next_desc;
1455 1456
		}

1457
#define rxtop (rx_ring->rx_skb_top)
1458
		if (!(staterr & E1000_RXD_STAT_EOP)) {
1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492
			/* this descriptor is only the beginning (or middle) */
			if (!rxtop) {
				/* this is the beginning of a chain */
				rxtop = skb;
				skb_fill_page_desc(rxtop, 0, buffer_info->page,
				                   0, length);
			} else {
				/* this is the middle of a chain */
				skb_fill_page_desc(rxtop,
				    skb_shinfo(rxtop)->nr_frags,
				    buffer_info->page, 0, length);
				/* re-use the skb, only consumed the page */
				buffer_info->skb = skb;
			}
			e1000_consume_page(buffer_info, rxtop, length);
			goto next_desc;
		} else {
			if (rxtop) {
				/* end of the chain */
				skb_fill_page_desc(rxtop,
				    skb_shinfo(rxtop)->nr_frags,
				    buffer_info->page, 0, length);
				/* re-use the current skb, we only consumed the
				 * page */
				buffer_info->skb = skb;
				skb = rxtop;
				rxtop = NULL;
				e1000_consume_page(buffer_info, skb, length);
			} else {
				/* no chain, got EOP, this buf is the packet
				 * copybreak to save the put_page/alloc_page */
				if (length <= copybreak &&
				    skb_tailroom(skb) >= length) {
					u8 *vaddr;
1493
					vaddr = kmap_atomic(buffer_info->page);
1494 1495
					memcpy(skb_tail_pointer(skb), vaddr,
					       length);
1496
					kunmap_atomic(vaddr);
1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510
					/* re-use the page, so don't erase
					 * buffer_info->page */
					skb_put(skb, length);
				} else {
					skb_fill_page_desc(skb, 0,
					                   buffer_info->page, 0,
				                           length);
					e1000_consume_page(buffer_info, skb,
					                   length);
				}
			}
		}

		/* Receive Checksum Offload XXX recompute due to CRC strip? */
1511
		e1000_rx_checksum(adapter, staterr,
1512
				  rx_desc->wb.lower.hi_dword.csum_ip.csum, skb);
1513

1514 1515
		e1000_rx_hash(netdev, rx_desc->wb.lower.hi_dword.rss, skb);

1516 1517 1518 1519 1520 1521
		/* probably a little skewed due to removing CRC */
		total_rx_bytes += skb->len;
		total_rx_packets++;

		/* eth type trans needs skb->data to point to something */
		if (!pskb_may_pull(skb, ETH_HLEN)) {
1522
			e_err("pskb_may_pull failed.\n");
1523
			dev_kfree_skb_irq(skb);
1524 1525 1526
			goto next_desc;
		}

1527 1528
		e1000_receive_skb(adapter, netdev, skb, staterr,
				  rx_desc->wb.upper.vlan);
1529 1530

next_desc:
1531
		rx_desc->wb.upper.status_error &= cpu_to_le32(~0xFF);
1532 1533 1534

		/* return some buffers to hardware, one at a time is too slow */
		if (unlikely(cleaned_count >= E1000_RX_BUFFER_WRITE)) {
1535
			adapter->alloc_rx_buf(rx_ring, cleaned_count,
1536
					      GFP_ATOMIC);
1537 1538 1539 1540 1541 1542
			cleaned_count = 0;
		}

		/* use prefetched values */
		rx_desc = next_rxd;
		buffer_info = next_buffer;
1543 1544

		staterr = le32_to_cpu(rx_desc->wb.upper.status_error);
1545 1546 1547 1548 1549
	}
	rx_ring->next_to_clean = i;

	cleaned_count = e1000_desc_unused(rx_ring);
	if (cleaned_count)
1550
		adapter->alloc_rx_buf(rx_ring, cleaned_count, GFP_ATOMIC);
1551 1552 1553 1554 1555 1556

	adapter->total_rx_bytes += total_rx_bytes;
	adapter->total_rx_packets += total_rx_packets;
	return cleaned;
}

1557 1558
/**
 * e1000_clean_rx_ring - Free Rx Buffers per Queue
1559
 * @rx_ring: Rx descriptor ring
1560
 **/
1561
static void e1000_clean_rx_ring(struct e1000_ring *rx_ring)
1562
{
1563
	struct e1000_adapter *adapter = rx_ring->adapter;
1564 1565 1566 1567 1568 1569 1570 1571 1572 1573
	struct e1000_buffer *buffer_info;
	struct e1000_ps_page *ps_page;
	struct pci_dev *pdev = adapter->pdev;
	unsigned int i, j;

	/* Free all the Rx ring sk_buffs */
	for (i = 0; i < rx_ring->count; i++) {
		buffer_info = &rx_ring->buffer_info[i];
		if (buffer_info->dma) {
			if (adapter->clean_rx == e1000_clean_rx_irq)
1574
				dma_unmap_single(&pdev->dev, buffer_info->dma,
1575
						 adapter->rx_buffer_len,
1576
						 DMA_FROM_DEVICE);
1577
			else if (adapter->clean_rx == e1000_clean_jumbo_rx_irq)
1578
				dma_unmap_page(&pdev->dev, buffer_info->dma,
1579
				               PAGE_SIZE,
1580
					       DMA_FROM_DEVICE);
1581
			else if (adapter->clean_rx == e1000_clean_rx_irq_ps)
1582
				dma_unmap_single(&pdev->dev, buffer_info->dma,
1583
						 adapter->rx_ps_bsize0,
1584
						 DMA_FROM_DEVICE);
1585 1586 1587
			buffer_info->dma = 0;
		}

1588 1589 1590 1591 1592
		if (buffer_info->page) {
			put_page(buffer_info->page);
			buffer_info->page = NULL;
		}

1593 1594 1595 1596 1597 1598
		if (buffer_info->skb) {
			dev_kfree_skb(buffer_info->skb);
			buffer_info->skb = NULL;
		}

		for (j = 0; j < PS_PAGE_BUFFERS; j++) {
A
Auke Kok 已提交
1599
			ps_page = &buffer_info->ps_pages[j];
1600 1601
			if (!ps_page->page)
				break;
1602 1603
			dma_unmap_page(&pdev->dev, ps_page->dma, PAGE_SIZE,
				       DMA_FROM_DEVICE);
1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620
			ps_page->dma = 0;
			put_page(ps_page->page);
			ps_page->page = NULL;
		}
	}

	/* there also may be some cached data from a chained receive */
	if (rx_ring->rx_skb_top) {
		dev_kfree_skb(rx_ring->rx_skb_top);
		rx_ring->rx_skb_top = NULL;
	}

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

	rx_ring->next_to_clean = 0;
	rx_ring->next_to_use = 0;
1621
	adapter->flags2 &= ~FLAG2_IS_DISCARDING;
1622

1623
	writel(0, rx_ring->head);
1624 1625 1626 1627
	if (rx_ring->adapter->flags2 & FLAG2_PCIM2PCI_ARBITER_WA)
		e1000e_update_rdt_wa(rx_ring, 0);
	else
		writel(0, rx_ring->tail);
1628 1629
}

1630 1631 1632 1633 1634
static void e1000e_downshift_workaround(struct work_struct *work)
{
	struct e1000_adapter *adapter = container_of(work,
					struct e1000_adapter, downshift_task);

1635 1636 1637
	if (test_bit(__E1000_DOWN, &adapter->state))
		return;

1638 1639 1640
	e1000e_gig_downshift_workaround_ich8lan(&adapter->hw);
}

1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652
/**
 * e1000_intr_msi - Interrupt Handler
 * @irq: interrupt number
 * @data: pointer to a network interface device structure
 **/
static irqreturn_t e1000_intr_msi(int irq, void *data)
{
	struct net_device *netdev = data;
	struct e1000_adapter *adapter = netdev_priv(netdev);
	struct e1000_hw *hw = &adapter->hw;
	u32 icr = er32(ICR);

1653 1654 1655
	/*
	 * read ICR disables interrupts using IAM
	 */
1656

1657
	if (icr & E1000_ICR_LSC) {
1658
		hw->mac.get_link_status = true;
1659 1660 1661 1662
		/*
		 * ICH8 workaround-- Call gig speed drop workaround on cable
		 * disconnect (LSC) before accessing any PHY registers
		 */
1663 1664
		if ((adapter->flags & FLAG_LSC_GIG_SPEED_DROP) &&
		    (!(er32(STATUS) & E1000_STATUS_LU)))
1665
			schedule_work(&adapter->downshift_task);
1666

1667 1668
		/*
		 * 80003ES2LAN workaround-- For packet buffer work-around on
1669
		 * link down event; disable receives here in the ISR and reset
1670 1671
		 * adapter in watchdog
		 */
1672 1673 1674 1675 1676
		if (netif_carrier_ok(netdev) &&
		    adapter->flags & FLAG_RX_NEEDS_RESTART) {
			/* disable receives */
			u32 rctl = er32(RCTL);
			ew32(RCTL, rctl & ~E1000_RCTL_EN);
1677
			adapter->flags |= FLAG_RX_RESTART_NOW;
1678 1679 1680 1681 1682 1683
		}
		/* guard against interrupt when we're going down */
		if (!test_bit(__E1000_DOWN, &adapter->state))
			mod_timer(&adapter->watchdog_timer, jiffies + 1);
	}

1684
	if (napi_schedule_prep(&adapter->napi)) {
1685 1686 1687 1688
		adapter->total_tx_bytes = 0;
		adapter->total_tx_packets = 0;
		adapter->total_rx_bytes = 0;
		adapter->total_rx_packets = 0;
1689
		__napi_schedule(&adapter->napi);
1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705
	}

	return IRQ_HANDLED;
}

/**
 * e1000_intr - Interrupt Handler
 * @irq: interrupt number
 * @data: pointer to a network interface device structure
 **/
static irqreturn_t e1000_intr(int irq, void *data)
{
	struct net_device *netdev = data;
	struct e1000_adapter *adapter = netdev_priv(netdev);
	struct e1000_hw *hw = &adapter->hw;
	u32 rctl, icr = er32(ICR);
1706

1707
	if (!icr || test_bit(__E1000_DOWN, &adapter->state))
1708 1709
		return IRQ_NONE;  /* Not our interrupt */

1710 1711 1712 1713
	/*
	 * IMS will not auto-mask if INT_ASSERTED is not set, and if it is
	 * not set, then the adapter didn't send an interrupt
	 */
1714 1715 1716
	if (!(icr & E1000_ICR_INT_ASSERTED))
		return IRQ_NONE;

1717 1718 1719 1720 1721
	/*
	 * Interrupt Auto-Mask...upon reading ICR,
	 * interrupts are masked.  No need for the
	 * IMC write
	 */
1722

1723
	if (icr & E1000_ICR_LSC) {
1724
		hw->mac.get_link_status = true;
1725 1726 1727 1728
		/*
		 * ICH8 workaround-- Call gig speed drop workaround on cable
		 * disconnect (LSC) before accessing any PHY registers
		 */
1729 1730
		if ((adapter->flags & FLAG_LSC_GIG_SPEED_DROP) &&
		    (!(er32(STATUS) & E1000_STATUS_LU)))
1731
			schedule_work(&adapter->downshift_task);
1732

1733 1734
		/*
		 * 80003ES2LAN workaround--
1735 1736 1737 1738 1739 1740 1741 1742 1743
		 * For packet buffer work-around on link down event;
		 * disable receives here in the ISR and
		 * reset adapter in watchdog
		 */
		if (netif_carrier_ok(netdev) &&
		    (adapter->flags & FLAG_RX_NEEDS_RESTART)) {
			/* disable receives */
			rctl = er32(RCTL);
			ew32(RCTL, rctl & ~E1000_RCTL_EN);
1744
			adapter->flags |= FLAG_RX_RESTART_NOW;
1745 1746 1747 1748 1749 1750
		}
		/* guard against interrupt when we're going down */
		if (!test_bit(__E1000_DOWN, &adapter->state))
			mod_timer(&adapter->watchdog_timer, jiffies + 1);
	}

1751
	if (napi_schedule_prep(&adapter->napi)) {
1752 1753 1754 1755
		adapter->total_tx_bytes = 0;
		adapter->total_tx_packets = 0;
		adapter->total_rx_bytes = 0;
		adapter->total_rx_packets = 0;
1756
		__napi_schedule(&adapter->napi);
1757 1758 1759 1760 1761
	}

	return IRQ_HANDLED;
}

1762 1763 1764 1765 1766 1767 1768 1769
static irqreturn_t e1000_msix_other(int irq, void *data)
{
	struct net_device *netdev = data;
	struct e1000_adapter *adapter = netdev_priv(netdev);
	struct e1000_hw *hw = &adapter->hw;
	u32 icr = er32(ICR);

	if (!(icr & E1000_ICR_INT_ASSERTED)) {
1770 1771
		if (!test_bit(__E1000_DOWN, &adapter->state))
			ew32(IMS, E1000_IMS_OTHER);
1772 1773 1774 1775 1776 1777 1778 1779 1780
		return IRQ_NONE;
	}

	if (icr & adapter->eiac_mask)
		ew32(ICS, (icr & adapter->eiac_mask));

	if (icr & E1000_ICR_OTHER) {
		if (!(icr & E1000_ICR_LSC))
			goto no_link_interrupt;
1781
		hw->mac.get_link_status = true;
1782 1783 1784 1785 1786 1787
		/* guard against interrupt when we're going down */
		if (!test_bit(__E1000_DOWN, &adapter->state))
			mod_timer(&adapter->watchdog_timer, jiffies + 1);
	}

no_link_interrupt:
1788 1789
	if (!test_bit(__E1000_DOWN, &adapter->state))
		ew32(IMS, E1000_IMS_LSC | E1000_IMS_OTHER);
1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805

	return IRQ_HANDLED;
}


static irqreturn_t e1000_intr_msix_tx(int irq, void *data)
{
	struct net_device *netdev = data;
	struct e1000_adapter *adapter = netdev_priv(netdev);
	struct e1000_hw *hw = &adapter->hw;
	struct e1000_ring *tx_ring = adapter->tx_ring;


	adapter->total_tx_bytes = 0;
	adapter->total_tx_packets = 0;

1806
	if (!e1000_clean_tx_irq(tx_ring))
1807 1808 1809 1810 1811 1812 1813 1814 1815 1816
		/* Ring was not completely cleaned, so fire another interrupt */
		ew32(ICS, tx_ring->ims_val);

	return IRQ_HANDLED;
}

static irqreturn_t e1000_intr_msix_rx(int irq, void *data)
{
	struct net_device *netdev = data;
	struct e1000_adapter *adapter = netdev_priv(netdev);
1817
	struct e1000_ring *rx_ring = adapter->rx_ring;
1818 1819 1820 1821

	/* Write the ITR value calculated at the end of the
	 * previous interrupt.
	 */
1822 1823 1824 1825
	if (rx_ring->set_itr) {
		writel(1000000000 / (rx_ring->itr_val * 256),
		       rx_ring->itr_register);
		rx_ring->set_itr = 0;
1826 1827
	}

1828
	if (napi_schedule_prep(&adapter->napi)) {
1829 1830
		adapter->total_rx_bytes = 0;
		adapter->total_rx_packets = 0;
1831
		__napi_schedule(&adapter->napi);
1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864
	}
	return IRQ_HANDLED;
}

/**
 * e1000_configure_msix - Configure MSI-X hardware
 *
 * e1000_configure_msix sets up the hardware to properly
 * generate MSI-X interrupts.
 **/
static void e1000_configure_msix(struct e1000_adapter *adapter)
{
	struct e1000_hw *hw = &adapter->hw;
	struct e1000_ring *rx_ring = adapter->rx_ring;
	struct e1000_ring *tx_ring = adapter->tx_ring;
	int vector = 0;
	u32 ctrl_ext, ivar = 0;

	adapter->eiac_mask = 0;

	/* Workaround issue with spurious interrupts on 82574 in MSI-X mode */
	if (hw->mac.type == e1000_82574) {
		u32 rfctl = er32(RFCTL);
		rfctl |= E1000_RFCTL_ACK_DIS;
		ew32(RFCTL, rfctl);
	}

#define E1000_IVAR_INT_ALLOC_VALID	0x8
	/* Configure Rx vector */
	rx_ring->ims_val = E1000_IMS_RXQ0;
	adapter->eiac_mask |= rx_ring->ims_val;
	if (rx_ring->itr_val)
		writel(1000000000 / (rx_ring->itr_val * 256),
1865
		       rx_ring->itr_register);
1866
	else
1867
		writel(1, rx_ring->itr_register);
1868 1869 1870 1871 1872 1873 1874
	ivar = E1000_IVAR_INT_ALLOC_VALID | vector;

	/* Configure Tx vector */
	tx_ring->ims_val = E1000_IMS_TXQ0;
	vector++;
	if (tx_ring->itr_val)
		writel(1000000000 / (tx_ring->itr_val * 256),
1875
		       tx_ring->itr_register);
1876
	else
1877
		writel(1, tx_ring->itr_register);
1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927
	adapter->eiac_mask |= tx_ring->ims_val;
	ivar |= ((E1000_IVAR_INT_ALLOC_VALID | vector) << 8);

	/* set vector for Other Causes, e.g. link changes */
	vector++;
	ivar |= ((E1000_IVAR_INT_ALLOC_VALID | vector) << 16);
	if (rx_ring->itr_val)
		writel(1000000000 / (rx_ring->itr_val * 256),
		       hw->hw_addr + E1000_EITR_82574(vector));
	else
		writel(1, hw->hw_addr + E1000_EITR_82574(vector));

	/* Cause Tx interrupts on every write back */
	ivar |= (1 << 31);

	ew32(IVAR, ivar);

	/* enable MSI-X PBA support */
	ctrl_ext = er32(CTRL_EXT);
	ctrl_ext |= E1000_CTRL_EXT_PBA_CLR;

	/* Auto-Mask Other interrupts upon ICR read */
#define E1000_EIAC_MASK_82574   0x01F00000
	ew32(IAM, ~E1000_EIAC_MASK_82574 | E1000_IMS_OTHER);
	ctrl_ext |= E1000_CTRL_EXT_EIAME;
	ew32(CTRL_EXT, ctrl_ext);
	e1e_flush();
}

void e1000e_reset_interrupt_capability(struct e1000_adapter *adapter)
{
	if (adapter->msix_entries) {
		pci_disable_msix(adapter->pdev);
		kfree(adapter->msix_entries);
		adapter->msix_entries = NULL;
	} else if (adapter->flags & FLAG_MSI_ENABLED) {
		pci_disable_msi(adapter->pdev);
		adapter->flags &= ~FLAG_MSI_ENABLED;
	}
}

/**
 * e1000e_set_interrupt_capability - set MSI or MSI-X if supported
 *
 * Attempt to configure interrupts using the best available
 * capabilities of the hardware and kernel.
 **/
void e1000e_set_interrupt_capability(struct e1000_adapter *adapter)
{
	int err;
1928
	int i;
1929 1930 1931 1932

	switch (adapter->int_mode) {
	case E1000E_INT_MODE_MSIX:
		if (adapter->flags & FLAG_HAS_MSIX) {
1933 1934
			adapter->num_vectors = 3; /* RxQ0, TxQ0 and other */
			adapter->msix_entries = kcalloc(adapter->num_vectors,
1935 1936 1937
						      sizeof(struct msix_entry),
						      GFP_KERNEL);
			if (adapter->msix_entries) {
1938
				for (i = 0; i < adapter->num_vectors; i++)
1939 1940 1941 1942
					adapter->msix_entries[i].entry = i;

				err = pci_enable_msix(adapter->pdev,
						      adapter->msix_entries,
1943
						      adapter->num_vectors);
B
Bruce Allan 已提交
1944
				if (err == 0)
1945 1946 1947
					return;
			}
			/* MSI-X failed, so fall through and try MSI */
1948
			e_err("Failed to initialize MSI-X interrupts.  Falling back to MSI interrupts.\n");
1949 1950 1951 1952 1953 1954 1955 1956 1957
			e1000e_reset_interrupt_capability(adapter);
		}
		adapter->int_mode = E1000E_INT_MODE_MSI;
		/* Fall through */
	case E1000E_INT_MODE_MSI:
		if (!pci_enable_msi(adapter->pdev)) {
			adapter->flags |= FLAG_MSI_ENABLED;
		} else {
			adapter->int_mode = E1000E_INT_MODE_LEGACY;
1958
			e_err("Failed to initialize MSI interrupts.  Falling back to legacy interrupts.\n");
1959 1960 1961 1962 1963 1964
		}
		/* Fall through */
	case E1000E_INT_MODE_LEGACY:
		/* Don't do anything; this is the system default */
		break;
	}
1965 1966 1967

	/* store the number of vectors being used */
	adapter->num_vectors = 1;
1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981
}

/**
 * e1000_request_msix - Initialize MSI-X interrupts
 *
 * e1000_request_msix allocates MSI-X vectors and requests interrupts from the
 * kernel.
 **/
static int e1000_request_msix(struct e1000_adapter *adapter)
{
	struct net_device *netdev = adapter->netdev;
	int err = 0, vector = 0;

	if (strlen(netdev->name) < (IFNAMSIZ - 5))
1982 1983 1984
		snprintf(adapter->rx_ring->name,
			 sizeof(adapter->rx_ring->name) - 1,
			 "%s-rx-0", netdev->name);
1985 1986 1987
	else
		memcpy(adapter->rx_ring->name, netdev->name, IFNAMSIZ);
	err = request_irq(adapter->msix_entries[vector].vector,
1988
			  e1000_intr_msix_rx, 0, adapter->rx_ring->name,
1989 1990
			  netdev);
	if (err)
1991
		return err;
1992 1993
	adapter->rx_ring->itr_register = adapter->hw.hw_addr +
	    E1000_EITR_82574(vector);
1994 1995 1996 1997
	adapter->rx_ring->itr_val = adapter->itr;
	vector++;

	if (strlen(netdev->name) < (IFNAMSIZ - 5))
1998 1999 2000
		snprintf(adapter->tx_ring->name,
			 sizeof(adapter->tx_ring->name) - 1,
			 "%s-tx-0", netdev->name);
2001 2002 2003
	else
		memcpy(adapter->tx_ring->name, netdev->name, IFNAMSIZ);
	err = request_irq(adapter->msix_entries[vector].vector,
2004
			  e1000_intr_msix_tx, 0, adapter->tx_ring->name,
2005 2006
			  netdev);
	if (err)
2007
		return err;
2008 2009
	adapter->tx_ring->itr_register = adapter->hw.hw_addr +
	    E1000_EITR_82574(vector);
2010 2011 2012 2013
	adapter->tx_ring->itr_val = adapter->itr;
	vector++;

	err = request_irq(adapter->msix_entries[vector].vector,
2014
			  e1000_msix_other, 0, netdev->name, netdev);
2015
	if (err)
2016
		return err;
2017 2018

	e1000_configure_msix(adapter);
2019

2020 2021 2022
	return 0;
}

2023 2024 2025 2026 2027 2028
/**
 * e1000_request_irq - initialize interrupts
 *
 * Attempts to configure interrupts using the best available
 * capabilities of the hardware and kernel.
 **/
2029 2030 2031 2032 2033
static int e1000_request_irq(struct e1000_adapter *adapter)
{
	struct net_device *netdev = adapter->netdev;
	int err;

2034 2035 2036 2037 2038 2039 2040 2041
	if (adapter->msix_entries) {
		err = e1000_request_msix(adapter);
		if (!err)
			return err;
		/* fall back to MSI */
		e1000e_reset_interrupt_capability(adapter);
		adapter->int_mode = E1000E_INT_MODE_MSI;
		e1000e_set_interrupt_capability(adapter);
2042
	}
2043
	if (adapter->flags & FLAG_MSI_ENABLED) {
2044
		err = request_irq(adapter->pdev->irq, e1000_intr_msi, 0,
2045 2046 2047
				  netdev->name, netdev);
		if (!err)
			return err;
2048

2049 2050 2051
		/* fall back to legacy interrupt */
		e1000e_reset_interrupt_capability(adapter);
		adapter->int_mode = E1000E_INT_MODE_LEGACY;
2052 2053
	}

2054
	err = request_irq(adapter->pdev->irq, e1000_intr, IRQF_SHARED,
2055 2056 2057 2058
			  netdev->name, netdev);
	if (err)
		e_err("Unable to allocate interrupt, Error: %d\n", err);

2059 2060 2061 2062 2063 2064 2065
	return err;
}

static void e1000_free_irq(struct e1000_adapter *adapter)
{
	struct net_device *netdev = adapter->netdev;

2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077
	if (adapter->msix_entries) {
		int vector = 0;

		free_irq(adapter->msix_entries[vector].vector, netdev);
		vector++;

		free_irq(adapter->msix_entries[vector].vector, netdev);
		vector++;

		/* Other Causes interrupt vector */
		free_irq(adapter->msix_entries[vector].vector, netdev);
		return;
2078
	}
2079 2080

	free_irq(adapter->pdev->irq, netdev);
2081 2082 2083 2084 2085 2086 2087 2088 2089 2090
}

/**
 * e1000_irq_disable - Mask off interrupt generation on the NIC
 **/
static void e1000_irq_disable(struct e1000_adapter *adapter)
{
	struct e1000_hw *hw = &adapter->hw;

	ew32(IMC, ~0);
2091 2092
	if (adapter->msix_entries)
		ew32(EIAC_82574, 0);
2093
	e1e_flush();
2094 2095 2096 2097 2098 2099 2100 2101

	if (adapter->msix_entries) {
		int i;
		for (i = 0; i < adapter->num_vectors; i++)
			synchronize_irq(adapter->msix_entries[i].vector);
	} else {
		synchronize_irq(adapter->pdev->irq);
	}
2102 2103 2104 2105 2106 2107 2108 2109 2110
}

/**
 * e1000_irq_enable - Enable default interrupt generation settings
 **/
static void e1000_irq_enable(struct e1000_adapter *adapter)
{
	struct e1000_hw *hw = &adapter->hw;

2111 2112 2113 2114 2115 2116
	if (adapter->msix_entries) {
		ew32(EIAC_82574, adapter->eiac_mask & E1000_EIAC_MASK_82574);
		ew32(IMS, adapter->eiac_mask | E1000_IMS_OTHER | E1000_IMS_LSC);
	} else {
		ew32(IMS, IMS_ENABLE_MASK);
	}
J
Jesse Brandeburg 已提交
2117
	e1e_flush();
2118 2119 2120
}

/**
2121
 * e1000e_get_hw_control - get control of the h/w from f/w
2122 2123
 * @adapter: address of board private structure
 *
2124
 * e1000e_get_hw_control sets {CTRL_EXT|SWSM}:DRV_LOAD bit.
2125 2126 2127 2128
 * For ASF and Pass Through versions of f/w this means that
 * the driver is loaded. For AMT version (only with 82573)
 * of the f/w this means that the network i/f is open.
 **/
2129
void e1000e_get_hw_control(struct e1000_adapter *adapter)
2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140
{
	struct e1000_hw *hw = &adapter->hw;
	u32 ctrl_ext;
	u32 swsm;

	/* Let firmware know the driver has taken over */
	if (adapter->flags & FLAG_HAS_SWSM_ON_LOAD) {
		swsm = er32(SWSM);
		ew32(SWSM, swsm | E1000_SWSM_DRV_LOAD);
	} else if (adapter->flags & FLAG_HAS_CTRLEXT_ON_LOAD) {
		ctrl_ext = er32(CTRL_EXT);
2141
		ew32(CTRL_EXT, ctrl_ext | E1000_CTRL_EXT_DRV_LOAD);
2142 2143 2144 2145
	}
}

/**
2146
 * e1000e_release_hw_control - release control of the h/w to f/w
2147 2148
 * @adapter: address of board private structure
 *
2149
 * e1000e_release_hw_control resets {CTRL_EXT|SWSM}:DRV_LOAD bit.
2150 2151 2152 2153 2154
 * For ASF and Pass Through versions of f/w this means that the
 * driver is no longer loaded. For AMT version (only with 82573) i
 * of the f/w this means that the network i/f is closed.
 *
 **/
2155
void e1000e_release_hw_control(struct e1000_adapter *adapter)
2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166
{
	struct e1000_hw *hw = &adapter->hw;
	u32 ctrl_ext;
	u32 swsm;

	/* Let firmware taken over control of h/w */
	if (adapter->flags & FLAG_HAS_SWSM_ON_LOAD) {
		swsm = er32(SWSM);
		ew32(SWSM, swsm & ~E1000_SWSM_DRV_LOAD);
	} else if (adapter->flags & FLAG_HAS_CTRLEXT_ON_LOAD) {
		ctrl_ext = er32(CTRL_EXT);
2167
		ew32(CTRL_EXT, ctrl_ext & ~E1000_CTRL_EXT_DRV_LOAD);
2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188
	}
}

/**
 * @e1000_alloc_ring - allocate memory for a ring structure
 **/
static int e1000_alloc_ring_dma(struct e1000_adapter *adapter,
				struct e1000_ring *ring)
{
	struct pci_dev *pdev = adapter->pdev;

	ring->desc = dma_alloc_coherent(&pdev->dev, ring->size, &ring->dma,
					GFP_KERNEL);
	if (!ring->desc)
		return -ENOMEM;

	return 0;
}

/**
 * e1000e_setup_tx_resources - allocate Tx resources (Descriptors)
2189
 * @tx_ring: Tx descriptor ring
2190 2191 2192
 *
 * Return 0 on success, negative on failure
 **/
2193
int e1000e_setup_tx_resources(struct e1000_ring *tx_ring)
2194
{
2195
	struct e1000_adapter *adapter = tx_ring->adapter;
2196 2197 2198
	int err = -ENOMEM, size;

	size = sizeof(struct e1000_buffer) * tx_ring->count;
E
Eric Dumazet 已提交
2199
	tx_ring->buffer_info = vzalloc(size);
2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216
	if (!tx_ring->buffer_info)
		goto err;

	/* round up to nearest 4K */
	tx_ring->size = tx_ring->count * sizeof(struct e1000_tx_desc);
	tx_ring->size = ALIGN(tx_ring->size, 4096);

	err = e1000_alloc_ring_dma(adapter, tx_ring);
	if (err)
		goto err;

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

	return 0;
err:
	vfree(tx_ring->buffer_info);
2217
	e_err("Unable to allocate memory for the transmit descriptor ring\n");
2218 2219 2220 2221 2222
	return err;
}

/**
 * e1000e_setup_rx_resources - allocate Rx resources (Descriptors)
2223
 * @rx_ring: Rx descriptor ring
2224 2225 2226
 *
 * Returns 0 on success, negative on failure
 **/
2227
int e1000e_setup_rx_resources(struct e1000_ring *rx_ring)
2228
{
2229
	struct e1000_adapter *adapter = rx_ring->adapter;
A
Auke Kok 已提交
2230 2231
	struct e1000_buffer *buffer_info;
	int i, size, desc_len, err = -ENOMEM;
2232 2233

	size = sizeof(struct e1000_buffer) * rx_ring->count;
E
Eric Dumazet 已提交
2234
	rx_ring->buffer_info = vzalloc(size);
2235 2236 2237
	if (!rx_ring->buffer_info)
		goto err;

A
Auke Kok 已提交
2238 2239 2240 2241 2242 2243 2244 2245
	for (i = 0; i < rx_ring->count; i++) {
		buffer_info = &rx_ring->buffer_info[i];
		buffer_info->ps_pages = kcalloc(PS_PAGE_BUFFERS,
						sizeof(struct e1000_ps_page),
						GFP_KERNEL);
		if (!buffer_info->ps_pages)
			goto err_pages;
	}
2246 2247 2248 2249 2250 2251 2252 2253 2254

	desc_len = sizeof(union e1000_rx_desc_packet_split);

	/* Round up to nearest 4K */
	rx_ring->size = rx_ring->count * desc_len;
	rx_ring->size = ALIGN(rx_ring->size, 4096);

	err = e1000_alloc_ring_dma(adapter, rx_ring);
	if (err)
A
Auke Kok 已提交
2255
		goto err_pages;
2256 2257 2258 2259 2260 2261

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

	return 0;
A
Auke Kok 已提交
2262 2263 2264 2265 2266 2267

err_pages:
	for (i = 0; i < rx_ring->count; i++) {
		buffer_info = &rx_ring->buffer_info[i];
		kfree(buffer_info->ps_pages);
	}
2268 2269
err:
	vfree(rx_ring->buffer_info);
2270
	e_err("Unable to allocate memory for the receive descriptor ring\n");
2271 2272 2273 2274 2275
	return err;
}

/**
 * e1000_clean_tx_ring - Free Tx Buffers
2276
 * @tx_ring: Tx descriptor ring
2277
 **/
2278
static void e1000_clean_tx_ring(struct e1000_ring *tx_ring)
2279
{
2280
	struct e1000_adapter *adapter = tx_ring->adapter;
2281 2282 2283 2284 2285 2286
	struct e1000_buffer *buffer_info;
	unsigned long size;
	unsigned int i;

	for (i = 0; i < tx_ring->count; i++) {
		buffer_info = &tx_ring->buffer_info[i];
2287
		e1000_put_txbuf(tx_ring, buffer_info);
2288 2289
	}

2290
	netdev_reset_queue(adapter->netdev);
2291 2292 2293 2294 2295 2296 2297 2298
	size = sizeof(struct e1000_buffer) * tx_ring->count;
	memset(tx_ring->buffer_info, 0, size);

	memset(tx_ring->desc, 0, tx_ring->size);

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

2299
	writel(0, tx_ring->head);
2300 2301 2302 2303
	if (tx_ring->adapter->flags2 & FLAG2_PCIM2PCI_ARBITER_WA)
		e1000e_update_tdt_wa(tx_ring, 0);
	else
		writel(0, tx_ring->tail);
2304 2305 2306 2307
}

/**
 * e1000e_free_tx_resources - Free Tx Resources per Queue
2308
 * @tx_ring: Tx descriptor ring
2309 2310 2311
 *
 * Free all transmit software resources
 **/
2312
void e1000e_free_tx_resources(struct e1000_ring *tx_ring)
2313
{
2314
	struct e1000_adapter *adapter = tx_ring->adapter;
2315 2316
	struct pci_dev *pdev = adapter->pdev;

2317
	e1000_clean_tx_ring(tx_ring);
2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328

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

	dma_free_coherent(&pdev->dev, tx_ring->size, tx_ring->desc,
			  tx_ring->dma);
	tx_ring->desc = NULL;
}

/**
 * e1000e_free_rx_resources - Free Rx Resources
2329
 * @rx_ring: Rx descriptor ring
2330 2331 2332
 *
 * Free all receive software resources
 **/
2333
void e1000e_free_rx_resources(struct e1000_ring *rx_ring)
2334
{
2335
	struct e1000_adapter *adapter = rx_ring->adapter;
2336
	struct pci_dev *pdev = adapter->pdev;
A
Auke Kok 已提交
2337
	int i;
2338

2339
	e1000_clean_rx_ring(rx_ring);
2340

B
Bruce Allan 已提交
2341
	for (i = 0; i < rx_ring->count; i++)
A
Auke Kok 已提交
2342 2343
		kfree(rx_ring->buffer_info[i].ps_pages);

2344 2345 2346 2347 2348 2349 2350 2351 2352 2353
	vfree(rx_ring->buffer_info);
	rx_ring->buffer_info = NULL;

	dma_free_coherent(&pdev->dev, rx_ring->size, rx_ring->desc,
			  rx_ring->dma);
	rx_ring->desc = NULL;
}

/**
 * e1000_update_itr - update the dynamic ITR value based on statistics
2354 2355 2356 2357 2358
 * @adapter: pointer to adapter
 * @itr_setting: current adapter->itr
 * @packets: the number of packets during this measurement interval
 * @bytes: the number of bytes during this measurement interval
 *
2359 2360 2361 2362 2363 2364
 *      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
2365 2366
 *      while increasing bulk throughput.  This functionality is controlled
 *      by the InterruptThrottleRate module parameter.
2367 2368 2369 2370 2371 2372 2373 2374
 **/
static unsigned int e1000_update_itr(struct e1000_adapter *adapter,
				     u16 itr_setting, int packets,
				     int bytes)
{
	unsigned int retval = itr_setting;

	if (packets == 0)
2375
		return itr_setting;
2376 2377 2378 2379 2380 2381

	switch (itr_setting) {
	case lowest_latency:
		/* handle TSO and jumbo frames */
		if (bytes/packets > 8000)
			retval = bulk_latency;
B
Bruce Allan 已提交
2382
		else if ((packets < 5) && (bytes > 512))
2383 2384 2385 2386 2387
			retval = low_latency;
		break;
	case low_latency:  /* 50 usec aka 20000 ints/s */
		if (bytes > 10000) {
			/* this if handles the TSO accounting */
B
Bruce Allan 已提交
2388
			if (bytes/packets > 8000)
2389
				retval = bulk_latency;
B
Bruce Allan 已提交
2390
			else if ((packets < 10) || ((bytes/packets) > 1200))
2391
				retval = bulk_latency;
B
Bruce Allan 已提交
2392
			else if ((packets > 35))
2393 2394 2395 2396 2397 2398 2399 2400 2401
				retval = lowest_latency;
		} else if (bytes/packets > 2000) {
			retval = bulk_latency;
		} else if (packets <= 2 && bytes < 512) {
			retval = lowest_latency;
		}
		break;
	case bulk_latency: /* 250 usec aka 4000 ints/s */
		if (bytes > 25000) {
B
Bruce Allan 已提交
2402
			if (packets > 35)
2403 2404 2405 2406 2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 2425
				retval = low_latency;
		} else if (bytes < 6000) {
			retval = low_latency;
		}
		break;
	}

	return retval;
}

static void e1000_set_itr(struct e1000_adapter *adapter)
{
	struct e1000_hw *hw = &adapter->hw;
	u16 current_itr;
	u32 new_itr = adapter->itr;

	/* for non-gigabit speeds, just fix the interrupt rate at 4000 */
	if (adapter->link_speed != SPEED_1000) {
		current_itr = 0;
		new_itr = 4000;
		goto set_itr_now;
	}

2426 2427 2428 2429 2430
	if (adapter->flags2 & FLAG2_DISABLE_AIM) {
		new_itr = 0;
		goto set_itr_now;
	}

2431 2432 2433 2434 2435 2436 2437 2438 2439 2440 2441 2442 2443 2444 2445 2446 2447 2448 2449 2450 2451 2452 2453 2454 2455 2456 2457 2458 2459 2460 2461 2462 2463 2464 2465
	adapter->tx_itr = e1000_update_itr(adapter,
				    adapter->tx_itr,
				    adapter->total_tx_packets,
				    adapter->total_tx_bytes);
	/* conservative mode (itr 3) eliminates the lowest_latency setting */
	if (adapter->itr_setting == 3 && adapter->tx_itr == lowest_latency)
		adapter->tx_itr = low_latency;

	adapter->rx_itr = e1000_update_itr(adapter,
				    adapter->rx_itr,
				    adapter->total_rx_packets,
				    adapter->total_rx_bytes);
	/* conservative mode (itr 3) eliminates the lowest_latency setting */
	if (adapter->itr_setting == 3 && adapter->rx_itr == lowest_latency)
		adapter->rx_itr = low_latency;

	current_itr = max(adapter->rx_itr, adapter->tx_itr);

	switch (current_itr) {
	/* counts and packets in update_itr are dependent on these numbers */
	case lowest_latency:
		new_itr = 70000;
		break;
	case low_latency:
		new_itr = 20000; /* aka hwitr = ~200 */
		break;
	case bulk_latency:
		new_itr = 4000;
		break;
	default:
		break;
	}

set_itr_now:
	if (new_itr != adapter->itr) {
2466 2467
		/*
		 * this attempts to bias the interrupt rate towards Bulk
2468
		 * by adding intermediate steps when interrupt rate is
2469 2470
		 * increasing
		 */
2471 2472 2473 2474
		new_itr = new_itr > adapter->itr ?
			     min(adapter->itr + (new_itr >> 2), new_itr) :
			     new_itr;
		adapter->itr = new_itr;
2475 2476 2477 2478
		adapter->rx_ring->itr_val = new_itr;
		if (adapter->msix_entries)
			adapter->rx_ring->set_itr = 1;
		else
2479 2480 2481 2482
			if (new_itr)
				ew32(ITR, 1000000000 / (new_itr * 256));
			else
				ew32(ITR, 0);
2483 2484 2485
	}
}

2486 2487 2488 2489 2490 2491
/**
 * e1000_alloc_queues - Allocate memory for all rings
 * @adapter: board private structure to initialize
 **/
static int __devinit e1000_alloc_queues(struct e1000_adapter *adapter)
{
2492 2493 2494
	int size = sizeof(struct e1000_ring);

	adapter->tx_ring = kzalloc(size, GFP_KERNEL);
2495 2496
	if (!adapter->tx_ring)
		goto err;
2497 2498
	adapter->tx_ring->count = adapter->tx_ring_count;
	adapter->tx_ring->adapter = adapter;
2499

2500
	adapter->rx_ring = kzalloc(size, GFP_KERNEL);
2501 2502
	if (!adapter->rx_ring)
		goto err;
2503 2504
	adapter->rx_ring->count = adapter->rx_ring_count;
	adapter->rx_ring->adapter = adapter;
2505 2506 2507 2508 2509 2510 2511 2512 2513

	return 0;
err:
	e_err("Unable to allocate memory for queues\n");
	kfree(adapter->rx_ring);
	kfree(adapter->tx_ring);
	return -ENOMEM;
}

2514
/**
B
Bruce Allan 已提交
2515
 * e1000e_poll - NAPI Rx polling callback
2516
 * @napi: struct associated with this polling callback
B
Bruce Allan 已提交
2517
 * @weight: number of packets driver is allowed to process this poll
2518
 **/
B
Bruce Allan 已提交
2519
static int e1000e_poll(struct napi_struct *napi, int weight)
2520
{
B
Bruce Allan 已提交
2521 2522
	struct e1000_adapter *adapter = container_of(napi, struct e1000_adapter,
						     napi);
2523
	struct e1000_hw *hw = &adapter->hw;
2524
	struct net_device *poll_dev = adapter->netdev;
2525
	int tx_cleaned = 1, work_done = 0;
2526

2527
	adapter = netdev_priv(poll_dev);
2528

B
Bruce Allan 已提交
2529 2530 2531
	if (!adapter->msix_entries ||
	    (adapter->rx_ring->ims_val & adapter->tx_ring->ims_val))
		tx_cleaned = e1000_clean_tx_irq(adapter->tx_ring);
2532

B
Bruce Allan 已提交
2533
	adapter->clean_rx(adapter->rx_ring, &work_done, weight);
2534

2535
	if (!tx_cleaned)
B
Bruce Allan 已提交
2536
		work_done = weight;
2537

B
Bruce Allan 已提交
2538 2539
	/* If weight not fully consumed, exit the polling mode */
	if (work_done < weight) {
2540 2541
		if (adapter->itr_setting & 3)
			e1000_set_itr(adapter);
2542
		napi_complete(napi);
2543 2544 2545 2546 2547 2548
		if (!test_bit(__E1000_DOWN, &adapter->state)) {
			if (adapter->msix_entries)
				ew32(IMS, adapter->rx_ring->ims_val);
			else
				e1000_irq_enable(adapter);
		}
2549 2550 2551 2552 2553
	}

	return work_done;
}

2554
static int e1000_vlan_rx_add_vid(struct net_device *netdev, u16 vid)
2555 2556 2557 2558 2559 2560 2561 2562 2563
{
	struct e1000_adapter *adapter = netdev_priv(netdev);
	struct e1000_hw *hw = &adapter->hw;
	u32 vfta, index;

	/* don't update vlan cookie if already programmed */
	if ((adapter->hw.mng_cookie.status &
	     E1000_MNG_DHCP_COOKIE_STATUS_VLAN) &&
	    (vid == adapter->mng_vlan_id))
2564
		return 0;
2565

2566
	/* add VID to filter table */
2567 2568 2569 2570 2571 2572
	if (adapter->flags & FLAG_HAS_HW_VLAN_FILTER) {
		index = (vid >> 5) & 0x7F;
		vfta = E1000_READ_REG_ARRAY(hw, E1000_VFTA, index);
		vfta |= (1 << (vid & 0x1F));
		hw->mac.ops.write_vfta(hw, index, vfta);
	}
J
Jeff Kirsher 已提交
2573 2574

	set_bit(vid, adapter->active_vlans);
2575 2576

	return 0;
2577 2578
}

2579
static int e1000_vlan_rx_kill_vid(struct net_device *netdev, u16 vid)
2580 2581 2582 2583 2584 2585 2586 2587 2588
{
	struct e1000_adapter *adapter = netdev_priv(netdev);
	struct e1000_hw *hw = &adapter->hw;
	u32 vfta, index;

	if ((adapter->hw.mng_cookie.status &
	     E1000_MNG_DHCP_COOKIE_STATUS_VLAN) &&
	    (vid == adapter->mng_vlan_id)) {
		/* release control to f/w */
2589
		e1000e_release_hw_control(adapter);
2590
		return 0;
2591 2592 2593
	}

	/* remove VID from filter table */
2594 2595 2596 2597 2598 2599
	if (adapter->flags & FLAG_HAS_HW_VLAN_FILTER) {
		index = (vid >> 5) & 0x7F;
		vfta = E1000_READ_REG_ARRAY(hw, E1000_VFTA, index);
		vfta &= ~(1 << (vid & 0x1F));
		hw->mac.ops.write_vfta(hw, index, vfta);
	}
J
Jeff Kirsher 已提交
2600 2601

	clear_bit(vid, adapter->active_vlans);
2602 2603

	return 0;
2604 2605
}

J
Jeff Kirsher 已提交
2606 2607 2608 2609 2610
/**
 * e1000e_vlan_filter_disable - helper to disable hw VLAN filtering
 * @adapter: board private structure to initialize
 **/
static void e1000e_vlan_filter_disable(struct e1000_adapter *adapter)
2611 2612
{
	struct net_device *netdev = adapter->netdev;
J
Jeff Kirsher 已提交
2613 2614
	struct e1000_hw *hw = &adapter->hw;
	u32 rctl;
2615

J
Jeff Kirsher 已提交
2616 2617 2618 2619 2620 2621 2622 2623 2624
	if (adapter->flags & FLAG_HAS_HW_VLAN_FILTER) {
		/* disable VLAN receive filtering */
		rctl = er32(RCTL);
		rctl &= ~(E1000_RCTL_VFE | E1000_RCTL_CFIEN);
		ew32(RCTL, rctl);

		if (adapter->mng_vlan_id != (u16)E1000_MNG_VLAN_NONE) {
			e1000_vlan_rx_kill_vid(netdev, adapter->mng_vlan_id);
			adapter->mng_vlan_id = E1000_MNG_VLAN_NONE;
2625 2626 2627 2628
		}
	}
}

J
Jeff Kirsher 已提交
2629 2630 2631 2632 2633 2634 2635 2636 2637 2638 2639 2640 2641 2642 2643 2644 2645
/**
 * e1000e_vlan_filter_enable - helper to enable HW VLAN filtering
 * @adapter: board private structure to initialize
 **/
static void e1000e_vlan_filter_enable(struct e1000_adapter *adapter)
{
	struct e1000_hw *hw = &adapter->hw;
	u32 rctl;

	if (adapter->flags & FLAG_HAS_HW_VLAN_FILTER) {
		/* enable VLAN receive filtering */
		rctl = er32(RCTL);
		rctl |= E1000_RCTL_VFE;
		rctl &= ~E1000_RCTL_CFIEN;
		ew32(RCTL, rctl);
	}
}
2646

J
Jeff Kirsher 已提交
2647 2648 2649 2650 2651
/**
 * e1000e_vlan_strip_enable - helper to disable HW VLAN stripping
 * @adapter: board private structure to initialize
 **/
static void e1000e_vlan_strip_disable(struct e1000_adapter *adapter)
2652 2653
{
	struct e1000_hw *hw = &adapter->hw;
J
Jeff Kirsher 已提交
2654
	u32 ctrl;
2655

J
Jeff Kirsher 已提交
2656 2657 2658 2659 2660
	/* disable VLAN tag insert/strip */
	ctrl = er32(CTRL);
	ctrl &= ~E1000_CTRL_VME;
	ew32(CTRL, ctrl);
}
2661

J
Jeff Kirsher 已提交
2662 2663 2664 2665 2666 2667 2668 2669
/**
 * e1000e_vlan_strip_enable - helper to enable HW VLAN stripping
 * @adapter: board private structure to initialize
 **/
static void e1000e_vlan_strip_enable(struct e1000_adapter *adapter)
{
	struct e1000_hw *hw = &adapter->hw;
	u32 ctrl;
2670

J
Jeff Kirsher 已提交
2671 2672 2673 2674 2675
	/* enable VLAN tag insert/strip */
	ctrl = er32(CTRL);
	ctrl |= E1000_CTRL_VME;
	ew32(CTRL, ctrl);
}
2676

J
Jeff Kirsher 已提交
2677 2678 2679 2680 2681 2682 2683 2684 2685 2686
static void e1000_update_mng_vlan(struct e1000_adapter *adapter)
{
	struct net_device *netdev = adapter->netdev;
	u16 vid = adapter->hw.mng_cookie.vlan_id;
	u16 old_vid = adapter->mng_vlan_id;

	if (adapter->hw.mng_cookie.status &
	    E1000_MNG_DHCP_COOKIE_STATUS_VLAN) {
		e1000_vlan_rx_add_vid(netdev, vid);
		adapter->mng_vlan_id = vid;
2687 2688
	}

J
Jeff Kirsher 已提交
2689 2690
	if ((old_vid != (u16)E1000_MNG_VLAN_NONE) && (vid != old_vid))
		e1000_vlan_rx_kill_vid(netdev, old_vid);
2691 2692 2693 2694 2695 2696
}

static void e1000_restore_vlan(struct e1000_adapter *adapter)
{
	u16 vid;

J
Jeff Kirsher 已提交
2697
	e1000_vlan_rx_add_vid(adapter->netdev, 0);
2698

J
Jeff Kirsher 已提交
2699
	for_each_set_bit(vid, adapter->active_vlans, VLAN_N_VID)
2700 2701 2702
		e1000_vlan_rx_add_vid(adapter->netdev, vid);
}

2703
static void e1000_init_manageability_pt(struct e1000_adapter *adapter)
2704 2705
{
	struct e1000_hw *hw = &adapter->hw;
2706
	u32 manc, manc2h, mdef, i, j;
2707 2708 2709 2710 2711 2712

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

	manc = er32(MANC);

2713 2714
	/*
	 * enable receiving management packets to the host. this will probably
2715
	 * generate destination unreachable messages from the host OS, but
2716 2717
	 * the packets will be handled on SMBUS
	 */
2718 2719
	manc |= E1000_MANC_EN_MNG2HOST;
	manc2h = er32(MANC2H);
2720 2721 2722 2723 2724 2725 2726 2727 2728 2729 2730 2731 2732 2733 2734

	switch (hw->mac.type) {
	default:
		manc2h |= (E1000_MANC2H_PORT_623 | E1000_MANC2H_PORT_664);
		break;
	case e1000_82574:
	case e1000_82583:
		/*
		 * Check if IPMI pass-through decision filter already exists;
		 * if so, enable it.
		 */
		for (i = 0, j = 0; i < 8; i++) {
			mdef = er32(MDEF(i));

			/* Ignore filters with anything other than IPMI ports */
2735
			if (mdef & ~(E1000_MDEF_PORT_623 | E1000_MDEF_PORT_664))
2736 2737 2738 2739 2740 2741 2742 2743 2744 2745 2746 2747 2748 2749 2750 2751 2752 2753 2754 2755 2756 2757 2758 2759 2760 2761 2762
				continue;

			/* Enable this decision filter in MANC2H */
			if (mdef)
				manc2h |= (1 << i);

			j |= mdef;
		}

		if (j == (E1000_MDEF_PORT_623 | E1000_MDEF_PORT_664))
			break;

		/* Create new decision filter in an empty filter */
		for (i = 0, j = 0; i < 8; i++)
			if (er32(MDEF(i)) == 0) {
				ew32(MDEF(i), (E1000_MDEF_PORT_623 |
					       E1000_MDEF_PORT_664));
				manc2h |= (1 << 1);
				j++;
				break;
			}

		if (!j)
			e_warn("Unable to create IPMI pass-through filter\n");
		break;
	}

2763 2764 2765 2766 2767
	ew32(MANC2H, manc2h);
	ew32(MANC, manc);
}

/**
2768
 * e1000_configure_tx - Configure Transmit Unit after Reset
2769 2770 2771 2772 2773 2774 2775 2776 2777
 * @adapter: board private structure
 *
 * Configure the Tx unit of the MAC after a reset.
 **/
static void e1000_configure_tx(struct e1000_adapter *adapter)
{
	struct e1000_hw *hw = &adapter->hw;
	struct e1000_ring *tx_ring = adapter->tx_ring;
	u64 tdba;
2778
	u32 tdlen, tarc;
2779 2780 2781 2782

	/* Setup the HW Tx Head and Tail descriptor pointers */
	tdba = tx_ring->dma;
	tdlen = tx_ring->count * sizeof(struct e1000_tx_desc);
2783 2784 2785 2786 2787 2788 2789
	ew32(TDBAL(0), (tdba & DMA_BIT_MASK(32)));
	ew32(TDBAH(0), (tdba >> 32));
	ew32(TDLEN(0), tdlen);
	ew32(TDH(0), 0);
	ew32(TDT(0), 0);
	tx_ring->head = adapter->hw.hw_addr + E1000_TDH(0);
	tx_ring->tail = adapter->hw.hw_addr + E1000_TDT(0);
2790 2791 2792

	/* Set the Tx Interrupt Delay register */
	ew32(TIDV, adapter->tx_int_delay);
2793
	/* Tx irq moderation */
2794 2795
	ew32(TADV, adapter->tx_abs_int_delay);

2796 2797 2798 2799 2800 2801 2802 2803 2804 2805 2806 2807
	if (adapter->flags2 & FLAG2_DMA_BURST) {
		u32 txdctl = er32(TXDCTL(0));
		txdctl &= ~(E1000_TXDCTL_PTHRESH | E1000_TXDCTL_HTHRESH |
			    E1000_TXDCTL_WTHRESH);
		/*
		 * set up some performance related parameters to encourage the
		 * hardware to use the bus more efficiently in bursts, depends
		 * on the tx_int_delay to be enabled,
		 * wthresh = 5 ==> burst write a cacheline (64 bytes) at a time
		 * hthresh = 1 ==> prefetch when one or more available
		 * pthresh = 0x1f ==> prefetch if internal cache 31 or less
		 * BEWARE: this seems to work but should be considered first if
2808
		 * there are Tx hangs or other Tx related bugs
2809 2810 2811 2812
		 */
		txdctl |= E1000_TXDCTL_DMA_BURST_ENABLE;
		ew32(TXDCTL(0), txdctl);
	}
2813 2814
	/* erratum work around: set txdctl the same for both queues */
	ew32(TXDCTL(1), er32(TXDCTL(0)));
2815

2816
	if (adapter->flags & FLAG_TARC_SPEED_MODE_BIT) {
2817
		tarc = er32(TARC(0));
2818 2819 2820 2821
		/*
		 * set the speed mode bit, we'll clear it if we're not at
		 * gigabit link later
		 */
2822 2823
#define SPEED_MODE_BIT (1 << 21)
		tarc |= SPEED_MODE_BIT;
2824
		ew32(TARC(0), tarc);
2825 2826 2827 2828
	}

	/* errata: program both queues to unweighted RR */
	if (adapter->flags & FLAG_TARC_SET_BIT_ZERO) {
2829
		tarc = er32(TARC(0));
2830
		tarc |= 1;
2831 2832
		ew32(TARC(0), tarc);
		tarc = er32(TARC(1));
2833
		tarc |= 1;
2834
		ew32(TARC(1), tarc);
2835 2836 2837 2838 2839 2840 2841 2842 2843 2844 2845 2846
	}

	/* Setup Transmit Descriptor Settings for eop descriptor */
	adapter->txd_cmd = E1000_TXD_CMD_EOP | E1000_TXD_CMD_IFCS;

	/* only set IDE if we are delaying interrupts using the timers */
	if (adapter->tx_int_delay)
		adapter->txd_cmd |= E1000_TXD_CMD_IDE;

	/* enable Report Status bit */
	adapter->txd_cmd |= E1000_TXD_CMD_RS;

2847
	hw->mac.ops.config_collision_dist(hw);
2848 2849 2850 2851 2852 2853 2854 2855 2856 2857 2858 2859 2860 2861
}

/**
 * e1000_setup_rctl - configure the receive control registers
 * @adapter: Board private structure
 **/
#define PAGE_USE_COUNT(S) (((S) >> PAGE_SHIFT) + \
			   (((S) & (PAGE_SIZE - 1)) ? 1 : 0))
static void e1000_setup_rctl(struct e1000_adapter *adapter)
{
	struct e1000_hw *hw = &adapter->hw;
	u32 rctl, rfctl;
	u32 pages = 0;

2862 2863 2864 2865 2866 2867 2868 2869
	/* Workaround Si errata on 82579 - configure jumbo frame flow */
	if (hw->mac.type == e1000_pch2lan) {
		s32 ret_val;

		if (adapter->netdev->mtu > ETH_DATA_LEN)
			ret_val = e1000_lv_jumbo_workaround_ich8lan(hw, true);
		else
			ret_val = e1000_lv_jumbo_workaround_ich8lan(hw, false);
2870 2871 2872

		if (ret_val)
			e_dbg("failed to enable jumbo frame workaround mode\n");
2873 2874
	}

2875 2876 2877 2878 2879 2880 2881 2882 2883 2884 2885 2886 2887 2888 2889 2890
	/* Program MC offset vector base */
	rctl = er32(RCTL);
	rctl &= ~(3 << E1000_RCTL_MO_SHIFT);
	rctl |= E1000_RCTL_EN | E1000_RCTL_BAM |
		E1000_RCTL_LBM_NO | E1000_RCTL_RDMTS_HALF |
		(adapter->hw.mac.mc_filter_type << E1000_RCTL_MO_SHIFT);

	/* Do not Store bad packets */
	rctl &= ~E1000_RCTL_SBP;

	/* Enable Long Packet receive */
	if (adapter->netdev->mtu <= ETH_DATA_LEN)
		rctl &= ~E1000_RCTL_LPE;
	else
		rctl |= E1000_RCTL_LPE;

J
Jeff Kirsher 已提交
2891 2892 2893 2894 2895 2896
	/* Some systems expect that the CRC is included in SMBUS traffic. The
	 * hardware strips the CRC before sending to both SMBUS (BMC) and to
	 * host memory when this is enabled
	 */
	if (adapter->flags2 & FLAG2_CRC_STRIPPING)
		rctl |= E1000_RCTL_SECRC;
2897

2898 2899 2900 2901 2902 2903 2904 2905 2906 2907 2908 2909 2910 2911 2912 2913 2914
	/* Workaround Si errata on 82577 PHY - configure IPG for jumbos */
	if ((hw->phy.type == e1000_phy_82577) && (rctl & E1000_RCTL_LPE)) {
		u16 phy_data;

		e1e_rphy(hw, PHY_REG(770, 26), &phy_data);
		phy_data &= 0xfff8;
		phy_data |= (1 << 2);
		e1e_wphy(hw, PHY_REG(770, 26), phy_data);

		e1e_rphy(hw, 22, &phy_data);
		phy_data &= 0x0fff;
		phy_data |= (1 << 14);
		e1e_wphy(hw, 0x10, 0x2823);
		e1e_wphy(hw, 0x11, 0x0003);
		e1e_wphy(hw, 22, phy_data);
	}

2915 2916 2917 2918 2919 2920 2921 2922 2923 2924 2925 2926 2927 2928 2929 2930 2931 2932 2933 2934
	/* Setup buffer sizes */
	rctl &= ~E1000_RCTL_SZ_4096;
	rctl |= E1000_RCTL_BSEX;
	switch (adapter->rx_buffer_len) {
	case 2048:
	default:
		rctl |= E1000_RCTL_SZ_2048;
		rctl &= ~E1000_RCTL_BSEX;
		break;
	case 4096:
		rctl |= E1000_RCTL_SZ_4096;
		break;
	case 8192:
		rctl |= E1000_RCTL_SZ_8192;
		break;
	case 16384:
		rctl |= E1000_RCTL_SZ_16384;
		break;
	}

2935 2936 2937 2938
	/* Enable Extended Status in all Receive Descriptors */
	rfctl = er32(RFCTL);
	rfctl |= E1000_RFCTL_EXTEN;

2939 2940 2941 2942 2943 2944 2945 2946 2947 2948 2949 2950 2951 2952 2953 2954
	/*
	 * 82571 and greater support packet-split where the protocol
	 * header is placed in skb->data and the packet data is
	 * placed in pages hanging off of skb_shinfo(skb)->nr_frags.
	 * In the case of a non-split, skb->data is linearly filled,
	 * followed by the page buffers.  Therefore, skb->data is
	 * sized to hold the largest protocol header.
	 *
	 * allocations using alloc_page take too long for regular MTU
	 * so only enable packet split for jumbo frames
	 *
	 * Using pages when the page size is greater than 16k wastes
	 * a lot of memory, since we allocate 3 pages at all times
	 * per packet.
	 */
	pages = PAGE_USE_COUNT(adapter->netdev->mtu);
2955
	if ((pages <= 3) && (PAGE_SIZE <= 16384) && (rctl & E1000_RCTL_LPE))
2956
		adapter->rx_ps_pages = pages;
2957 2958
	else
		adapter->rx_ps_pages = 0;
2959 2960

	if (adapter->rx_ps_pages) {
2961 2962
		u32 psrctl = 0;

2963 2964 2965 2966
		/*
		 * disable packet split support for IPv6 extension headers,
		 * because some malformed IPv6 headers can hang the Rx
		 */
2967 2968 2969
		rfctl |= (E1000_RFCTL_IPV6_EX_DIS |
			  E1000_RFCTL_NEW_IPV6_EXT_DIS);

A
Auke Kok 已提交
2970 2971
		/* Enable Packet split descriptors */
		rctl |= E1000_RCTL_DTYP_PS;
2972 2973 2974 2975 2976 2977 2978 2979 2980 2981 2982 2983 2984 2985 2986 2987 2988 2989 2990 2991

		psrctl |= adapter->rx_ps_bsize0 >>
			E1000_PSRCTL_BSIZE0_SHIFT;

		switch (adapter->rx_ps_pages) {
		case 3:
			psrctl |= PAGE_SIZE <<
				E1000_PSRCTL_BSIZE3_SHIFT;
		case 2:
			psrctl |= PAGE_SIZE <<
				E1000_PSRCTL_BSIZE2_SHIFT;
		case 1:
			psrctl |= PAGE_SIZE >>
				E1000_PSRCTL_BSIZE1_SHIFT;
			break;
		}

		ew32(PSRCTL, psrctl);
	}

B
Ben Greear 已提交
2992 2993 2994 2995 2996 2997 2998 2999 3000 3001 3002 3003 3004 3005 3006 3007
	/* 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 */
		rctl |= (E1000_RCTL_SBP | /* Receive bad packets */
			 E1000_RCTL_BAM | /* RX All Bcast Pkts */
			 E1000_RCTL_PMCF); /* RX All MAC Ctrl Pkts */

		rctl &= ~(E1000_RCTL_VFE | /* Disable VLAN filter */
			  E1000_RCTL_DPF | /* Allow filtered pause */
			  E1000_RCTL_CFIEN); /* Dis VLAN CFIEN Filter */
		/* Do not mess with E1000_CTRL_VME, it affects transmit as well,
		 * and that breaks VLANs.
		 */
	}

3008
	ew32(RFCTL, rfctl);
3009
	ew32(RCTL, rctl);
3010 3011
	/* just started the receive unit, no need to restart */
	adapter->flags &= ~FLAG_RX_RESTART_NOW;
3012 3013 3014 3015 3016 3017 3018 3019 3020 3021 3022 3023 3024 3025 3026 3027 3028 3029
}

/**
 * e1000_configure_rx - Configure Receive Unit after Reset
 * @adapter: board private structure
 *
 * Configure the Rx unit of the MAC after a reset.
 **/
static void e1000_configure_rx(struct e1000_adapter *adapter)
{
	struct e1000_hw *hw = &adapter->hw;
	struct e1000_ring *rx_ring = adapter->rx_ring;
	u64 rdba;
	u32 rdlen, rctl, rxcsum, ctrl_ext;

	if (adapter->rx_ps_pages) {
		/* this is a 32 byte descriptor */
		rdlen = rx_ring->count *
3030
		    sizeof(union e1000_rx_desc_packet_split);
3031 3032
		adapter->clean_rx = e1000_clean_rx_irq_ps;
		adapter->alloc_rx_buf = e1000_alloc_rx_buffers_ps;
3033
	} else if (adapter->netdev->mtu > ETH_FRAME_LEN + ETH_FCS_LEN) {
3034
		rdlen = rx_ring->count * sizeof(union e1000_rx_desc_extended);
3035 3036
		adapter->clean_rx = e1000_clean_jumbo_rx_irq;
		adapter->alloc_rx_buf = e1000_alloc_jumbo_rx_buffers;
3037
	} else {
3038
		rdlen = rx_ring->count * sizeof(union e1000_rx_desc_extended);
3039 3040 3041 3042 3043 3044
		adapter->clean_rx = e1000_clean_rx_irq;
		adapter->alloc_rx_buf = e1000_alloc_rx_buffers;
	}

	/* disable receives while setting up the descriptors */
	rctl = er32(RCTL);
3045 3046
	if (!(adapter->flags2 & FLAG2_NO_DISABLE_RX))
		ew32(RCTL, rctl & ~E1000_RCTL_EN);
3047
	e1e_flush();
3048
	usleep_range(10000, 20000);
3049

3050 3051 3052 3053
	if (adapter->flags2 & FLAG2_DMA_BURST) {
		/*
		 * set the writeback threshold (only takes effect if the RDTR
		 * is set). set GRAN=1 and write back up to 0x4 worth, and
3054
		 * enable prefetching of 0x20 Rx descriptors
3055 3056 3057 3058 3059 3060 3061 3062 3063 3064 3065 3066 3067 3068 3069 3070 3071 3072
		 * granularity = 01
		 * wthresh = 04,
		 * hthresh = 04,
		 * pthresh = 0x20
		 */
		ew32(RXDCTL(0), E1000_RXDCTL_DMA_BURST_ENABLE);
		ew32(RXDCTL(1), E1000_RXDCTL_DMA_BURST_ENABLE);

		/*
		 * override the delay timers for enabling bursting, only if
		 * the value was not set by the user via module options
		 */
		if (adapter->rx_int_delay == DEFAULT_RDTR)
			adapter->rx_int_delay = BURST_RDTR;
		if (adapter->rx_abs_int_delay == DEFAULT_RADV)
			adapter->rx_abs_int_delay = BURST_RADV;
	}

3073 3074 3075 3076 3077
	/* set the Receive Delay Timer Register */
	ew32(RDTR, adapter->rx_int_delay);

	/* irq moderation */
	ew32(RADV, adapter->rx_abs_int_delay);
3078
	if ((adapter->itr_setting != 0) && (adapter->itr != 0))
3079
		ew32(ITR, 1000000000 / (adapter->itr * 256));
3080 3081 3082 3083 3084 3085 3086 3087

	ctrl_ext = er32(CTRL_EXT);
	/* Auto-Mask interrupts upon ICR access */
	ctrl_ext |= E1000_CTRL_EXT_IAME;
	ew32(IAM, 0xffffffff);
	ew32(CTRL_EXT, ctrl_ext);
	e1e_flush();

3088 3089 3090 3091
	/*
	 * Setup the HW Rx Head and Tail Descriptor Pointers and
	 * the Base and Length of the Rx Descriptor Ring
	 */
3092
	rdba = rx_ring->dma;
3093 3094 3095 3096 3097 3098 3099
	ew32(RDBAL(0), (rdba & DMA_BIT_MASK(32)));
	ew32(RDBAH(0), (rdba >> 32));
	ew32(RDLEN(0), rdlen);
	ew32(RDH(0), 0);
	ew32(RDT(0), 0);
	rx_ring->head = adapter->hw.hw_addr + E1000_RDH(0);
	rx_ring->tail = adapter->hw.hw_addr + E1000_RDT(0);
3100 3101 3102

	/* Enable Receive Checksum Offload for TCP and UDP */
	rxcsum = er32(RXCSUM);
3103
	if (adapter->netdev->features & NETIF_F_RXCSUM) {
3104 3105
		rxcsum |= E1000_RXCSUM_TUOFL;

3106 3107 3108 3109
		/*
		 * IPv4 payload checksum for UDP fragments must be
		 * used in conjunction with packet-split.
		 */
3110 3111 3112 3113 3114 3115 3116 3117
		if (adapter->rx_ps_pages)
			rxcsum |= E1000_RXCSUM_IPPCSE;
	} else {
		rxcsum &= ~E1000_RXCSUM_TUOFL;
		/* no need to clear IPPCSE as it defaults to 0 */
	}
	ew32(RXCSUM, rxcsum);

3118 3119 3120 3121 3122
	if (adapter->hw.mac.type == e1000_pch2lan) {
		/*
		 * With jumbo frames, excessive C-state transition
		 * latencies result in dropped transactions.
		 */
3123 3124 3125
		if (adapter->netdev->mtu > ETH_DATA_LEN) {
			u32 rxdctl = er32(RXDCTL(0));
			ew32(RXDCTL(0), rxdctl | 0x3);
3126
			pm_qos_update_request(&adapter->netdev->pm_qos_req, 55);
3127
		} else {
3128 3129
			pm_qos_update_request(&adapter->netdev->pm_qos_req,
					      PM_QOS_DEFAULT_VALUE);
3130
		}
3131
	}
3132 3133 3134 3135 3136 3137

	/* Enable Receives */
	ew32(RCTL, rctl);
}

/**
3138 3139
 * e1000e_write_mc_addr_list - write multicast addresses to MTA
 * @netdev: network interface device structure
3140
 *
3141 3142 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
 * 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 e1000e_write_mc_addr_list(struct net_device *netdev)
{
	struct e1000_adapter *adapter = netdev_priv(netdev);
	struct e1000_hw *hw = &adapter->hw;
	struct netdev_hw_addr *ha;
	u8 *mta_list;
	int i;

	if (netdev_mc_empty(netdev)) {
		/* nothing to program, so clear mc list */
		hw->mac.ops.update_mc_addr_list(hw, NULL, 0);
		return 0;
	}

	mta_list = kzalloc(netdev_mc_count(netdev) * ETH_ALEN, GFP_ATOMIC);
	if (!mta_list)
		return -ENOMEM;

	/* update_mc_addr_list expects a packed array of only addresses. */
	i = 0;
	netdev_for_each_mc_addr(ha, netdev)
		memcpy(mta_list + (i++ * ETH_ALEN), ha->addr, ETH_ALEN);

	hw->mac.ops.update_mc_addr_list(hw, mta_list, i);
	kfree(mta_list);

	return netdev_mc_count(netdev);
}

/**
 * e1000e_write_uc_addr_list - write unicast addresses to RAR table
 * @netdev: network interface device structure
3178
 *
3179 3180 3181 3182
 * 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
3183
 **/
3184
static int e1000e_write_uc_addr_list(struct net_device *netdev)
3185
{
3186 3187 3188 3189 3190 3191 3192 3193 3194 3195 3196 3197 3198 3199 3200 3201 3202 3203 3204 3205 3206 3207 3208 3209 3210 3211
	struct e1000_adapter *adapter = netdev_priv(netdev);
	struct e1000_hw *hw = &adapter->hw;
	unsigned int rar_entries = hw->mac.rar_entry_count;
	int count = 0;

	/* save a rar entry for our hardware address */
	rar_entries--;

	/* save a rar entry for the LAA workaround */
	if (adapter->flags & FLAG_RESET_OVERWRITES_LAA)
		rar_entries--;

	/* return ENOMEM indicating insufficient memory for addresses */
	if (netdev_uc_count(netdev) > rar_entries)
		return -ENOMEM;

	if (!netdev_uc_empty(netdev) && rar_entries) {
		struct netdev_hw_addr *ha;

		/*
		 * write the addresses in reverse order to avoid write
		 * combining
		 */
		netdev_for_each_uc_addr(ha, netdev) {
			if (!rar_entries)
				break;
3212
			hw->mac.ops.rar_set(hw, ha->addr, rar_entries--);
3213 3214 3215 3216 3217 3218 3219 3220 3221 3222 3223 3224
			count++;
		}
	}

	/* zero out the remaining RAR entries not used above */
	for (; rar_entries > 0; rar_entries--) {
		ew32(RAH(rar_entries), 0);
		ew32(RAL(rar_entries), 0);
	}
	e1e_flush();

	return count;
3225 3226 3227
}

/**
3228
 * e1000e_set_rx_mode - secondary unicast, Multicast and Promiscuous mode set
3229 3230
 * @netdev: network interface device structure
 *
3231 3232 3233
 * The ndo_set_rx_mode entry point is called whenever the unicast or multicast
 * address list or the network interface flags are updated.  This routine is
 * responsible for configuring the hardware for proper unicast, multicast,
3234 3235
 * promiscuous mode, and all-multi behavior.
 **/
3236
static void e1000e_set_rx_mode(struct net_device *netdev)
3237 3238 3239 3240 3241 3242 3243 3244
{
	struct e1000_adapter *adapter = netdev_priv(netdev);
	struct e1000_hw *hw = &adapter->hw;
	u32 rctl;

	/* Check for Promiscuous and All Multicast modes */
	rctl = er32(RCTL);

3245 3246 3247
	/* clear the affected bits */
	rctl &= ~(E1000_RCTL_UPE | E1000_RCTL_MPE);

3248 3249
	if (netdev->flags & IFF_PROMISC) {
		rctl |= (E1000_RCTL_UPE | E1000_RCTL_MPE);
J
Jeff Kirsher 已提交
3250 3251
		/* Do not hardware filter VLANs in promisc mode */
		e1000e_vlan_filter_disable(adapter);
3252
	} else {
3253
		int count;
3254

3255 3256 3257
		if (netdev->flags & IFF_ALLMULTI) {
			rctl |= E1000_RCTL_MPE;
		} else {
3258 3259 3260 3261 3262 3263 3264 3265
			/*
			 * 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
			 */
			count = e1000e_write_mc_addr_list(netdev);
			if (count < 0)
				rctl |= E1000_RCTL_MPE;
3266
		}
J
Jeff Kirsher 已提交
3267
		e1000e_vlan_filter_enable(adapter);
3268
		/*
3269 3270 3271
		 * Write addresses to available RAR registers, if there is not
		 * sufficient space to store all the addresses then enable
		 * unicast promiscuous mode
3272
		 */
3273 3274 3275
		count = e1000e_write_uc_addr_list(netdev);
		if (count < 0)
			rctl |= E1000_RCTL_UPE;
3276
	}
J
Jeff Kirsher 已提交
3277

3278 3279
	ew32(RCTL, rctl);

J
Jeff Kirsher 已提交
3280 3281 3282 3283
	if (netdev->features & NETIF_F_HW_VLAN_RX)
		e1000e_vlan_strip_enable(adapter);
	else
		e1000e_vlan_strip_disable(adapter);
3284 3285
}

3286 3287 3288 3289 3290 3291 3292 3293 3294 3295 3296 3297 3298 3299 3300 3301 3302 3303 3304 3305 3306 3307 3308 3309 3310 3311 3312 3313 3314 3315 3316 3317 3318 3319 3320 3321
static void e1000e_setup_rss_hash(struct e1000_adapter *adapter)
{
	struct e1000_hw *hw = &adapter->hw;
	u32 mrqc, rxcsum;
	int i;
	static const u32 rsskey[10] = {
		0xda565a6d, 0xc20e5b25, 0x3d256741, 0xb08fa343, 0xcb2bcad0,
		0xb4307bae, 0xa32dcb77, 0x0cf23080, 0x3bb7426a, 0xfa01acbe
	};

	/* Fill out hash function seed */
	for (i = 0; i < 10; i++)
		ew32(RSSRK(i), rsskey[i]);

	/* Direct all traffic to queue 0 */
	for (i = 0; i < 32; i++)
		ew32(RETA(i), 0);

	/*
	 * Disable raw packet checksumming so that RSS hash is placed in
	 * descriptor on writeback.
	 */
	rxcsum = er32(RXCSUM);
	rxcsum |= E1000_RXCSUM_PCSD;

	ew32(RXCSUM, rxcsum);

	mrqc = (E1000_MRQC_RSS_FIELD_IPV4 |
		E1000_MRQC_RSS_FIELD_IPV4_TCP |
		E1000_MRQC_RSS_FIELD_IPV6 |
		E1000_MRQC_RSS_FIELD_IPV6_TCP |
		E1000_MRQC_RSS_FIELD_IPV6_TCP_EX);

	ew32(MRQC, mrqc);
}

3322
/**
3323
 * e1000_configure - configure the hardware for Rx and Tx
3324 3325 3326 3327
 * @adapter: private board structure
 **/
static void e1000_configure(struct e1000_adapter *adapter)
{
3328 3329
	struct e1000_ring *rx_ring = adapter->rx_ring;

3330
	e1000e_set_rx_mode(adapter->netdev);
3331 3332

	e1000_restore_vlan(adapter);
3333
	e1000_init_manageability_pt(adapter);
3334 3335

	e1000_configure_tx(adapter);
3336 3337 3338

	if (adapter->netdev->features & NETIF_F_RXHASH)
		e1000e_setup_rss_hash(adapter);
3339 3340
	e1000_setup_rctl(adapter);
	e1000_configure_rx(adapter);
3341
	adapter->alloc_rx_buf(rx_ring, e1000_desc_unused(rx_ring), GFP_KERNEL);
3342 3343 3344 3345 3346 3347 3348 3349 3350 3351 3352 3353
}

/**
 * e1000e_power_up_phy - restore link in case the phy was powered down
 * @adapter: address of board private structure
 *
 * The phy may be powered down to save power and turn off link when the
 * driver is unloaded and wake on lan is not enabled (among others)
 * *** this routine MUST be followed by a call to e1000e_reset ***
 **/
void e1000e_power_up_phy(struct e1000_adapter *adapter)
{
3354 3355
	if (adapter->hw.phy.ops.power_up)
		adapter->hw.phy.ops.power_up(&adapter->hw);
3356 3357 3358 3359 3360 3361 3362

	adapter->hw.mac.ops.setup_link(&adapter->hw);
}

/**
 * e1000_power_down_phy - Power down the PHY
 *
3363 3364
 * Power down the PHY so no link is implied when interface is down.
 * The PHY cannot be powered down if management or WoL is active.
3365 3366 3367 3368
 */
static void e1000_power_down_phy(struct e1000_adapter *adapter)
{
	/* WoL is enabled */
3369
	if (adapter->wol)
3370 3371
		return;

3372 3373
	if (adapter->hw.phy.ops.power_down)
		adapter->hw.phy.ops.power_down(&adapter->hw);
3374 3375 3376 3377 3378 3379 3380 3381
}

/**
 * e1000e_reset - bring the hardware into a known good state
 *
 * This function boots the hardware and enables some settings that
 * require a configuration cycle of the hardware - those cannot be
 * set/changed during runtime. After reset the device needs to be
3382
 * properly configured for Rx, Tx etc.
3383 3384 3385 3386
 */
void e1000e_reset(struct e1000_adapter *adapter)
{
	struct e1000_mac_info *mac = &adapter->hw.mac;
3387
	struct e1000_fc_info *fc = &adapter->hw.fc;
3388 3389
	struct e1000_hw *hw = &adapter->hw;
	u32 tx_space, min_tx_space, min_rx_space;
3390
	u32 pba = adapter->pba;
3391 3392
	u16 hwm;

3393
	/* reset Packet Buffer Allocation to default */
3394
	ew32(PBA, pba);
3395

3396
	if (adapter->max_frame_size > ETH_FRAME_LEN + ETH_FCS_LEN) {
3397 3398
		/*
		 * To maintain wire speed transmits, the Tx FIFO should be
3399 3400 3401 3402
		 * large enough to accommodate two full transmit packets,
		 * rounded up to the next 1KB and expressed in KB.  Likewise,
		 * the Rx FIFO should be large enough to accommodate at least
		 * one full receive packet and is similarly rounded up and
3403 3404
		 * expressed in KB.
		 */
3405
		pba = er32(PBA);
3406
		/* upper 16 bits has Tx packet buffer allocation size in KB */
3407
		tx_space = pba >> 16;
3408
		/* lower 16 bits has Rx packet buffer allocation size in KB */
3409
		pba &= 0xffff;
3410
		/*
3411
		 * the Tx fifo also stores 16 bytes of information about the Tx
3412
		 * but don't include ethernet FCS because hardware appends it
3413 3414
		 */
		min_tx_space = (adapter->max_frame_size +
3415 3416 3417 3418 3419
				sizeof(struct e1000_tx_desc) -
				ETH_FCS_LEN) * 2;
		min_tx_space = ALIGN(min_tx_space, 1024);
		min_tx_space >>= 10;
		/* software strips receive CRC, so leave room for it */
3420
		min_rx_space = adapter->max_frame_size;
3421 3422 3423
		min_rx_space = ALIGN(min_rx_space, 1024);
		min_rx_space >>= 10;

3424 3425
		/*
		 * If current Tx allocation is less than the min Tx FIFO size,
3426
		 * and the min Tx FIFO size is less than the current Rx FIFO
3427 3428
		 * allocation, take space away from current Rx allocation
		 */
3429 3430 3431
		if ((tx_space < min_tx_space) &&
		    ((min_tx_space - tx_space) < pba)) {
			pba -= min_tx_space - tx_space;
3432

3433
			/*
3434
			 * if short on Rx space, Rx wins and must trump Tx
3435 3436
			 * adjustment or use Early Receive if available
			 */
3437
			if (pba < min_rx_space)
3438
				pba = min_rx_space;
3439
		}
3440 3441

		ew32(PBA, pba);
3442 3443
	}

3444 3445 3446
	/*
	 * flow control settings
	 *
3447
	 * The high water mark must be low enough to fit one full frame
3448 3449 3450
	 * (or the size used for early receive) above it in the Rx FIFO.
	 * Set it to the lower of:
	 * - 90% of the Rx FIFO size, and
3451
	 * - the full Rx FIFO size minus one full frame
3452
	 */
3453 3454 3455 3456
	if (adapter->flags & FLAG_DISABLE_FC_PAUSE_TIME)
		fc->pause_time = 0xFFFF;
	else
		fc->pause_time = E1000_FC_PAUSE_TIME;
3457
	fc->send_xon = true;
3458 3459 3460
	fc->current_mode = fc->requested_mode;

	switch (hw->mac.type) {
3461 3462 3463 3464 3465 3466 3467 3468 3469 3470
	case e1000_ich9lan:
	case e1000_ich10lan:
		if (adapter->netdev->mtu > ETH_DATA_LEN) {
			pba = 14;
			ew32(PBA, pba);
			fc->high_water = 0x2800;
			fc->low_water = fc->high_water - 8;
			break;
		}
		/* fall-through */
3471
	default:
3472 3473
		hwm = min(((pba << 10) * 9 / 10),
			  ((pba << 10) - adapter->max_frame_size));
3474 3475 3476 3477 3478

		fc->high_water = hwm & E1000_FCRTH_RTH; /* 8-byte granularity */
		fc->low_water = fc->high_water - 8;
		break;
	case e1000_pchlan:
3479 3480 3481 3482 3483 3484 3485 3486 3487 3488 3489
		/*
		 * Workaround PCH LOM adapter hangs with certain network
		 * loads.  If hangs persist, try disabling Tx flow control.
		 */
		if (adapter->netdev->mtu > ETH_DATA_LEN) {
			fc->high_water = 0x3500;
			fc->low_water  = 0x1500;
		} else {
			fc->high_water = 0x5000;
			fc->low_water  = 0x3000;
		}
3490
		fc->refresh_time = 0x1000;
3491 3492 3493 3494 3495 3496
		break;
	case e1000_pch2lan:
		fc->high_water = 0x05C20;
		fc->low_water = 0x05048;
		fc->pause_time = 0x0650;
		fc->refresh_time = 0x0400;
3497 3498 3499 3500
		if (adapter->netdev->mtu > ETH_DATA_LEN) {
			pba = 14;
			ew32(PBA, pba);
		}
3501
		break;
3502
	}
3503

3504 3505
	/*
	 * Disable Adaptive Interrupt Moderation if 2 full packets cannot
3506
	 * fit in receive buffer.
3507 3508
	 */
	if (adapter->itr_setting & 0x3) {
3509
		if ((adapter->max_frame_size * 2) > (pba << 10)) {
3510 3511 3512 3513 3514 3515 3516 3517 3518 3519 3520 3521 3522 3523 3524
			if (!(adapter->flags2 & FLAG2_DISABLE_AIM)) {
				dev_info(&adapter->pdev->dev,
					"Interrupt Throttle Rate turned off\n");
				adapter->flags2 |= FLAG2_DISABLE_AIM;
				ew32(ITR, 0);
			}
		} else if (adapter->flags2 & FLAG2_DISABLE_AIM) {
			dev_info(&adapter->pdev->dev,
				 "Interrupt Throttle Rate turned on\n");
			adapter->flags2 &= ~FLAG2_DISABLE_AIM;
			adapter->itr = 20000;
			ew32(ITR, 1000000000 / (adapter->itr * 256));
		}
	}

3525 3526
	/* Allow time for pending master requests to run */
	mac->ops.reset_hw(hw);
3527 3528 3529 3530 3531

	/*
	 * For parts with AMT enabled, let the firmware know
	 * that the network interface is in control
	 */
J
Jesse Brandeburg 已提交
3532
	if (adapter->flags & FLAG_HAS_AMT)
3533
		e1000e_get_hw_control(adapter);
3534

3535 3536 3537
	ew32(WUC, 0);

	if (mac->ops.init_hw(hw))
3538
		e_err("Hardware Error\n");
3539 3540 3541 3542 3543 3544 3545

	e1000_update_mng_vlan(adapter);

	/* Enable h/w to recognize an 802.1Q VLAN Ethernet packet */
	ew32(VET, ETH_P_8021Q);

	e1000e_reset_adaptive(hw);
3546 3547 3548 3549 3550 3551 3552

	if (!netif_running(adapter->netdev) &&
	    !test_bit(__E1000_TESTING, &adapter->state)) {
		e1000_power_down_phy(adapter);
		return;
	}

3553 3554
	e1000_get_phy_info(hw);

3555 3556
	if ((adapter->flags & FLAG_HAS_SMART_POWER_DOWN) &&
	    !(adapter->flags & FLAG_SMART_POWER_DOWN)) {
3557
		u16 phy_data = 0;
3558 3559
		/*
		 * speed up time to link by disabling smart power down, ignore
3560
		 * the return value of this function because there is nothing
3561 3562
		 * different we would do if it failed
		 */
3563 3564 3565 3566 3567 3568 3569 3570 3571 3572 3573 3574 3575 3576 3577
		e1e_rphy(hw, IGP02E1000_PHY_POWER_MGMT, &phy_data);
		phy_data &= ~IGP02E1000_PM_SPD;
		e1e_wphy(hw, IGP02E1000_PHY_POWER_MGMT, phy_data);
	}
}

int e1000e_up(struct e1000_adapter *adapter)
{
	struct e1000_hw *hw = &adapter->hw;

	/* hardware has been reset, we need to reload some things */
	e1000_configure(adapter);

	clear_bit(__E1000_DOWN, &adapter->state);

3578 3579
	if (adapter->msix_entries)
		e1000_configure_msix(adapter);
3580 3581
	e1000_irq_enable(adapter);

3582
	netif_start_queue(adapter->netdev);
3583

3584
	/* fire a link change interrupt to start the watchdog */
3585 3586 3587 3588 3589
	if (adapter->msix_entries)
		ew32(ICS, E1000_ICS_LSC | E1000_ICR_OTHER);
	else
		ew32(ICS, E1000_ICS_LSC);

3590 3591 3592
	return 0;
}

3593 3594 3595 3596 3597 3598 3599 3600 3601 3602 3603 3604 3605
static void e1000e_flush_descriptors(struct e1000_adapter *adapter)
{
	struct e1000_hw *hw = &adapter->hw;

	if (!(adapter->flags2 & FLAG2_DMA_BURST))
		return;

	/* flush pending descriptor writebacks to memory */
	ew32(TIDV, adapter->tx_int_delay | E1000_TIDV_FPD);
	ew32(RDTR, adapter->rx_int_delay | E1000_RDTR_FPD);

	/* execute the writes immediately */
	e1e_flush();
3606 3607 3608 3609 3610 3611 3612

	/*
	 * due to rare timing issues, write to TIDV/RDTR again to ensure the
	 * write is successful
	 */
	ew32(TIDV, adapter->tx_int_delay | E1000_TIDV_FPD);
	ew32(RDTR, adapter->rx_int_delay | E1000_RDTR_FPD);
3613 3614 3615 3616 3617

	/* execute the writes immediately */
	e1e_flush();
}

J
Jeff Kirsher 已提交
3618 3619
static void e1000e_update_stats(struct e1000_adapter *adapter);

3620 3621 3622 3623 3624 3625
void e1000e_down(struct e1000_adapter *adapter)
{
	struct net_device *netdev = adapter->netdev;
	struct e1000_hw *hw = &adapter->hw;
	u32 tctl, rctl;

3626 3627 3628 3629
	/*
	 * signal that we're down so the interrupt handler does not
	 * reschedule our watchdog timer
	 */
3630 3631 3632 3633
	set_bit(__E1000_DOWN, &adapter->state);

	/* disable receives in the hardware */
	rctl = er32(RCTL);
3634 3635
	if (!(adapter->flags2 & FLAG2_NO_DISABLE_RX))
		ew32(RCTL, rctl & ~E1000_RCTL_EN);
3636 3637
	/* flush and sleep below */

3638
	netif_stop_queue(netdev);
3639 3640 3641 3642 3643

	/* disable transmits in the hardware */
	tctl = er32(TCTL);
	tctl &= ~E1000_TCTL_EN;
	ew32(TCTL, tctl);
3644

3645 3646
	/* flush both disables and wait for them to finish */
	e1e_flush();
3647
	usleep_range(10000, 20000);
3648 3649 3650 3651 3652 3653 3654

	e1000_irq_disable(adapter);

	del_timer_sync(&adapter->watchdog_timer);
	del_timer_sync(&adapter->phy_info_timer);

	netif_carrier_off(netdev);
J
Jeff Kirsher 已提交
3655 3656 3657 3658 3659

	spin_lock(&adapter->stats64_lock);
	e1000e_update_stats(adapter);
	spin_unlock(&adapter->stats64_lock);

3660
	e1000e_flush_descriptors(adapter);
3661 3662
	e1000_clean_tx_ring(adapter->tx_ring);
	e1000_clean_rx_ring(adapter->rx_ring);
3663

3664 3665 3666
	adapter->link_speed = 0;
	adapter->link_duplex = 0;

3667 3668
	if (!pci_channel_offline(adapter->pdev))
		e1000e_reset(adapter);
3669

3670 3671 3672 3673 3674 3675 3676 3677 3678 3679
	/*
	 * TODO: for power management, we could drop the link and
	 * pci_disable_device here.
	 */
}

void e1000e_reinit_locked(struct e1000_adapter *adapter)
{
	might_sleep();
	while (test_and_set_bit(__E1000_RESETTING, &adapter->state))
3680
		usleep_range(1000, 2000);
3681 3682 3683 3684 3685 3686 3687 3688 3689 3690 3691 3692 3693 3694 3695 3696 3697 3698 3699
	e1000e_down(adapter);
	e1000e_up(adapter);
	clear_bit(__E1000_RESETTING, &adapter->state);
}

/**
 * e1000_sw_init - Initialize general software structures (struct e1000_adapter)
 * @adapter: board private structure to initialize
 *
 * e1000_sw_init initializes the Adapter private data structure.
 * Fields are initialized based on PCI device information and
 * OS network device settings (MTU size).
 **/
static int __devinit e1000_sw_init(struct e1000_adapter *adapter)
{
	struct net_device *netdev = adapter->netdev;

	adapter->rx_buffer_len = ETH_FRAME_LEN + VLAN_HLEN + ETH_FCS_LEN;
	adapter->rx_ps_bsize0 = 128;
3700 3701
	adapter->max_frame_size = netdev->mtu + ETH_HLEN + ETH_FCS_LEN;
	adapter->min_frame_size = ETH_ZLEN + ETH_FCS_LEN;
3702 3703
	adapter->tx_ring_count = E1000_DEFAULT_TXD;
	adapter->rx_ring_count = E1000_DEFAULT_RXD;
3704

J
Jeff Kirsher 已提交
3705 3706
	spin_lock_init(&adapter->stats64_lock);

3707
	e1000e_set_interrupt_capability(adapter);
3708

3709 3710
	if (e1000_alloc_queues(adapter))
		return -ENOMEM;
3711 3712 3713 3714 3715 3716 3717 3718

	/* Explicitly disable IRQ since the NIC can be in any state. */
	e1000_irq_disable(adapter);

	set_bit(__E1000_DOWN, &adapter->state);
	return 0;
}

3719 3720 3721 3722 3723 3724 3725 3726 3727 3728 3729 3730
/**
 * e1000_intr_msi_test - Interrupt Handler
 * @irq: interrupt number
 * @data: pointer to a network interface device structure
 **/
static irqreturn_t e1000_intr_msi_test(int irq, void *data)
{
	struct net_device *netdev = data;
	struct e1000_adapter *adapter = netdev_priv(netdev);
	struct e1000_hw *hw = &adapter->hw;
	u32 icr = er32(ICR);

3731
	e_dbg("icr is %08X\n", icr);
3732 3733 3734 3735 3736 3737 3738 3739 3740 3741 3742 3743 3744 3745 3746 3747 3748 3749 3750 3751 3752 3753 3754 3755 3756 3757
	if (icr & E1000_ICR_RXSEQ) {
		adapter->flags &= ~FLAG_MSI_TEST_FAILED;
		wmb();
	}

	return IRQ_HANDLED;
}

/**
 * e1000_test_msi_interrupt - Returns 0 for successful test
 * @adapter: board private struct
 *
 * code flow taken from tg3.c
 **/
static int e1000_test_msi_interrupt(struct e1000_adapter *adapter)
{
	struct net_device *netdev = adapter->netdev;
	struct e1000_hw *hw = &adapter->hw;
	int err;

	/* poll_enable hasn't been called yet, so don't need disable */
	/* clear any pending events */
	er32(ICR);

	/* free the real vector and request a test handler */
	e1000_free_irq(adapter);
3758
	e1000e_reset_interrupt_capability(adapter);
3759 3760 3761 3762 3763 3764 3765 3766 3767

	/* Assume that the test fails, if it succeeds then the test
	 * MSI irq handler will unset this flag */
	adapter->flags |= FLAG_MSI_TEST_FAILED;

	err = pci_enable_msi(adapter->pdev);
	if (err)
		goto msi_test_failed;

3768
	err = request_irq(adapter->pdev->irq, e1000_intr_msi_test, 0,
3769 3770 3771 3772 3773 3774 3775 3776 3777 3778 3779 3780 3781 3782 3783 3784 3785 3786 3787 3788
			  netdev->name, netdev);
	if (err) {
		pci_disable_msi(adapter->pdev);
		goto msi_test_failed;
	}

	wmb();

	e1000_irq_enable(adapter);

	/* fire an unusual interrupt on the test handler */
	ew32(ICS, E1000_ICS_RXSEQ);
	e1e_flush();
	msleep(50);

	e1000_irq_disable(adapter);

	rmb();

	if (adapter->flags & FLAG_MSI_TEST_FAILED) {
3789
		adapter->int_mode = E1000E_INT_MODE_LEGACY;
3790
		e_info("MSI interrupt test failed, using legacy interrupt.\n");
3791
	} else {
3792
		e_dbg("MSI interrupt test succeeded!\n");
3793
	}
3794 3795 3796 3797 3798

	free_irq(adapter->pdev->irq, netdev);
	pci_disable_msi(adapter->pdev);

msi_test_failed:
3799
	e1000e_set_interrupt_capability(adapter);
3800
	return e1000_request_irq(adapter);
3801 3802 3803 3804 3805 3806 3807 3808 3809 3810 3811 3812 3813 3814 3815 3816 3817 3818
}

/**
 * e1000_test_msi - Returns 0 if MSI test succeeds or INTx mode is restored
 * @adapter: board private struct
 *
 * code flow taken from tg3.c, called with e1000 interrupts disabled.
 **/
static int e1000_test_msi(struct e1000_adapter *adapter)
{
	int err;
	u16 pci_cmd;

	if (!(adapter->flags & FLAG_MSI_ENABLED))
		return 0;

	/* disable SERR in case the MSI write causes a master abort */
	pci_read_config_word(adapter->pdev, PCI_COMMAND, &pci_cmd);
3819 3820 3821
	if (pci_cmd & PCI_COMMAND_SERR)
		pci_write_config_word(adapter->pdev, PCI_COMMAND,
				      pci_cmd & ~PCI_COMMAND_SERR);
3822 3823 3824

	err = e1000_test_msi_interrupt(adapter);

3825 3826 3827 3828 3829 3830
	/* re-enable SERR */
	if (pci_cmd & PCI_COMMAND_SERR) {
		pci_read_config_word(adapter->pdev, PCI_COMMAND, &pci_cmd);
		pci_cmd |= PCI_COMMAND_SERR;
		pci_write_config_word(adapter->pdev, PCI_COMMAND, pci_cmd);
	}
3831 3832 3833 3834

	return err;
}

3835 3836 3837 3838 3839 3840 3841 3842 3843 3844 3845 3846 3847 3848 3849 3850
/**
 * e1000_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 e1000_open(struct net_device *netdev)
{
	struct e1000_adapter *adapter = netdev_priv(netdev);
	struct e1000_hw *hw = &adapter->hw;
3851
	struct pci_dev *pdev = adapter->pdev;
3852 3853 3854 3855 3856 3857
	int err;

	/* disallow open during test */
	if (test_bit(__E1000_TESTING, &adapter->state))
		return -EBUSY;

3858 3859
	pm_runtime_get_sync(&pdev->dev);

3860 3861
	netif_carrier_off(netdev);

3862
	/* allocate transmit descriptors */
3863
	err = e1000e_setup_tx_resources(adapter->tx_ring);
3864 3865 3866 3867
	if (err)
		goto err_setup_tx;

	/* allocate receive descriptors */
3868
	err = e1000e_setup_rx_resources(adapter->rx_ring);
3869 3870 3871
	if (err)
		goto err_setup_rx;

3872 3873 3874 3875 3876
	/*
	 * If AMT is enabled, let the firmware know that the network
	 * interface is now open and reset the part to a known state.
	 */
	if (adapter->flags & FLAG_HAS_AMT) {
3877
		e1000e_get_hw_control(adapter);
3878 3879 3880
		e1000e_reset(adapter);
	}

3881 3882 3883 3884 3885 3886 3887
	e1000e_power_up_phy(adapter);

	adapter->mng_vlan_id = E1000_MNG_VLAN_NONE;
	if ((adapter->hw.mng_cookie.status &
	     E1000_MNG_DHCP_COOKIE_STATUS_VLAN))
		e1000_update_mng_vlan(adapter);

3888 3889
	/* DMA latency requirement to workaround jumbo issue */
	if (adapter->hw.mac.type == e1000_pch2lan)
3890 3891 3892
		pm_qos_add_request(&adapter->netdev->pm_qos_req,
				   PM_QOS_CPU_DMA_LATENCY,
				   PM_QOS_DEFAULT_VALUE);
3893

3894 3895
	/*
	 * before we allocate an interrupt, we must be ready to handle it.
3896 3897
	 * Setting DEBUG_SHIRQ in the kernel makes it fire an interrupt
	 * as soon as we call pci_request_irq, so we have to setup our
3898 3899
	 * clean_rx handler before we do so.
	 */
3900 3901 3902 3903 3904 3905
	e1000_configure(adapter);

	err = e1000_request_irq(adapter);
	if (err)
		goto err_req_irq;

3906 3907 3908 3909 3910
	/*
	 * Work around PCIe errata with MSI interrupts causing some chipsets to
	 * ignore e1000e MSI messages, which means we need to test our MSI
	 * interrupt now
	 */
3911
	if (adapter->int_mode != E1000E_INT_MODE_LEGACY) {
3912 3913 3914 3915 3916 3917 3918
		err = e1000_test_msi(adapter);
		if (err) {
			e_err("Interrupt allocation failed\n");
			goto err_req_irq;
		}
	}

3919 3920 3921 3922 3923 3924 3925
	/* From here on the code is the same as e1000e_up() */
	clear_bit(__E1000_DOWN, &adapter->state);

	napi_enable(&adapter->napi);

	e1000_irq_enable(adapter);

3926
	adapter->tx_hang_recheck = false;
3927
	netif_start_queue(netdev);
3928

3929 3930 3931
	adapter->idle_check = true;
	pm_runtime_put(&pdev->dev);

3932
	/* fire a link status change interrupt to start the watchdog */
3933 3934 3935 3936
	if (adapter->msix_entries)
		ew32(ICS, E1000_ICS_LSC | E1000_ICR_OTHER);
	else
		ew32(ICS, E1000_ICS_LSC);
3937 3938 3939 3940

	return 0;

err_req_irq:
3941
	e1000e_release_hw_control(adapter);
3942
	e1000_power_down_phy(adapter);
3943
	e1000e_free_rx_resources(adapter->rx_ring);
3944
err_setup_rx:
3945
	e1000e_free_tx_resources(adapter->tx_ring);
3946 3947
err_setup_tx:
	e1000e_reset(adapter);
3948
	pm_runtime_put_sync(&pdev->dev);
3949 3950 3951 3952 3953 3954 3955 3956 3957 3958 3959 3960 3961 3962 3963 3964 3965 3966

	return err;
}

/**
 * e1000_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 e1000_close(struct net_device *netdev)
{
	struct e1000_adapter *adapter = netdev_priv(netdev);
3967
	struct pci_dev *pdev = adapter->pdev;
3968 3969 3970 3971
	int count = E1000_CHECK_RESET_COUNT;

	while (test_bit(__E1000_RESETTING, &adapter->state) && count--)
		usleep_range(10000, 20000);
3972 3973

	WARN_ON(test_bit(__E1000_RESETTING, &adapter->state));
3974 3975 3976

	pm_runtime_get_sync(&pdev->dev);

3977 3978
	napi_disable(&adapter->napi);

3979 3980 3981 3982
	if (!test_bit(__E1000_DOWN, &adapter->state)) {
		e1000e_down(adapter);
		e1000_free_irq(adapter);
	}
3983 3984
	e1000_power_down_phy(adapter);

3985 3986
	e1000e_free_tx_resources(adapter->tx_ring);
	e1000e_free_rx_resources(adapter->rx_ring);
3987

3988 3989 3990 3991
	/*
	 * kill manageability vlan ID if supported, but not if a vlan with
	 * the same ID is registered on the host OS (let 8021q kill it)
	 */
J
Jeff Kirsher 已提交
3992 3993
	if (adapter->hw.mng_cookie.status &
	    E1000_MNG_DHCP_COOKIE_STATUS_VLAN)
3994 3995
		e1000_vlan_rx_kill_vid(netdev, adapter->mng_vlan_id);

3996 3997 3998 3999
	/*
	 * If AMT is enabled, let the firmware know that the network
	 * interface is now closed
	 */
4000 4001 4002
	if ((adapter->flags & FLAG_HAS_AMT) &&
	    !test_bit(__E1000_TESTING, &adapter->state))
		e1000e_release_hw_control(adapter);
4003

4004
	if (adapter->hw.mac.type == e1000_pch2lan)
4005
		pm_qos_remove_request(&adapter->netdev->pm_qos_req);
4006

4007 4008
	pm_runtime_put_sync(&pdev->dev);

4009 4010 4011 4012 4013 4014 4015 4016 4017 4018 4019 4020
	return 0;
}
/**
 * e1000_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 e1000_set_mac(struct net_device *netdev, void *p)
{
	struct e1000_adapter *adapter = netdev_priv(netdev);
4021
	struct e1000_hw *hw = &adapter->hw;
4022 4023 4024 4025 4026 4027 4028 4029
	struct sockaddr *addr = p;

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

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

4030
	hw->mac.ops.rar_set(&adapter->hw, adapter->hw.mac.addr, 0);
4031 4032 4033 4034 4035

	if (adapter->flags & FLAG_RESET_OVERWRITES_LAA) {
		/* activate the work around */
		e1000e_set_laa_state_82571(&adapter->hw, 1);

4036 4037
		/*
		 * Hold a copy of the LAA in RAR[14] This is done so that
4038 4039 4040 4041
		 * between the time RAR[0] gets clobbered  and the time it
		 * gets fixed (in e1000_watchdog), the actual LAA is in one
		 * of the RARs and no incoming packets directed to this port
		 * are dropped. Eventually the LAA will be in RAR[0] and
4042 4043
		 * RAR[14]
		 */
4044 4045
		hw->mac.ops.rar_set(&adapter->hw, adapter->hw.mac.addr,
				    adapter->hw.mac.rar_entry_count - 1);
4046 4047 4048 4049 4050
	}

	return 0;
}

4051 4052 4053 4054 4055 4056 4057 4058 4059 4060 4061 4062
/**
 * e1000e_update_phy_task - work thread to update phy
 * @work: pointer to our work struct
 *
 * this worker thread exists because we must acquire a
 * semaphore to read the phy, which we could msleep while
 * waiting for it, and we can't msleep in a timer.
 **/
static void e1000e_update_phy_task(struct work_struct *work)
{
	struct e1000_adapter *adapter = container_of(work,
					struct e1000_adapter, update_phy_task);
4063 4064 4065 4066

	if (test_bit(__E1000_DOWN, &adapter->state))
		return;

4067 4068 4069
	e1000_get_phy_info(&adapter->hw);
}

4070 4071 4072 4073
/*
 * Need to wait a few seconds after link up to get diagnostic information from
 * the phy
 */
4074 4075 4076
static void e1000_update_phy_info(unsigned long data)
{
	struct e1000_adapter *adapter = (struct e1000_adapter *) data;
4077 4078 4079 4080

	if (test_bit(__E1000_DOWN, &adapter->state))
		return;

4081
	schedule_work(&adapter->update_phy_task);
4082 4083
}

4084 4085 4086
/**
 * e1000e_update_phy_stats - Update the PHY statistics counters
 * @adapter: board private structure
4087 4088
 *
 * Read/clear the upper 16-bit PHY registers and read/accumulate lower
4089 4090 4091 4092 4093 4094 4095 4096 4097 4098 4099 4100 4101 4102 4103
 **/
static void e1000e_update_phy_stats(struct e1000_adapter *adapter)
{
	struct e1000_hw *hw = &adapter->hw;
	s32 ret_val;
	u16 phy_data;

	ret_val = hw->phy.ops.acquire(hw);
	if (ret_val)
		return;

	/*
	 * A page set is expensive so check if already on desired page.
	 * If not, set to the page with the PHY status registers.
	 */
4104
	hw->phy.addr = 1;
4105 4106 4107 4108
	ret_val = e1000e_read_phy_reg_mdic(hw, IGP01E1000_PHY_PAGE_SELECT,
					   &phy_data);
	if (ret_val)
		goto release;
4109 4110 4111
	if (phy_data != (HV_STATS_PAGE << IGP_PAGE_SHIFT)) {
		ret_val = hw->phy.ops.set_page(hw,
					       HV_STATS_PAGE << IGP_PAGE_SHIFT);
4112 4113 4114 4115 4116
		if (ret_val)
			goto release;
	}

	/* Single Collision Count */
4117 4118
	hw->phy.ops.read_reg_page(hw, HV_SCC_UPPER, &phy_data);
	ret_val = hw->phy.ops.read_reg_page(hw, HV_SCC_LOWER, &phy_data);
4119 4120 4121 4122
	if (!ret_val)
		adapter->stats.scc += phy_data;

	/* Excessive Collision Count */
4123 4124
	hw->phy.ops.read_reg_page(hw, HV_ECOL_UPPER, &phy_data);
	ret_val = hw->phy.ops.read_reg_page(hw, HV_ECOL_LOWER, &phy_data);
4125 4126 4127 4128
	if (!ret_val)
		adapter->stats.ecol += phy_data;

	/* Multiple Collision Count */
4129 4130
	hw->phy.ops.read_reg_page(hw, HV_MCC_UPPER, &phy_data);
	ret_val = hw->phy.ops.read_reg_page(hw, HV_MCC_LOWER, &phy_data);
4131 4132 4133 4134
	if (!ret_val)
		adapter->stats.mcc += phy_data;

	/* Late Collision Count */
4135 4136
	hw->phy.ops.read_reg_page(hw, HV_LATECOL_UPPER, &phy_data);
	ret_val = hw->phy.ops.read_reg_page(hw, HV_LATECOL_LOWER, &phy_data);
4137 4138 4139 4140
	if (!ret_val)
		adapter->stats.latecol += phy_data;

	/* Collision Count - also used for adaptive IFS */
4141 4142
	hw->phy.ops.read_reg_page(hw, HV_COLC_UPPER, &phy_data);
	ret_val = hw->phy.ops.read_reg_page(hw, HV_COLC_LOWER, &phy_data);
4143 4144 4145 4146
	if (!ret_val)
		hw->mac.collision_delta = phy_data;

	/* Defer Count */
4147 4148
	hw->phy.ops.read_reg_page(hw, HV_DC_UPPER, &phy_data);
	ret_val = hw->phy.ops.read_reg_page(hw, HV_DC_LOWER, &phy_data);
4149 4150 4151 4152
	if (!ret_val)
		adapter->stats.dc += phy_data;

	/* Transmit with no CRS */
4153 4154
	hw->phy.ops.read_reg_page(hw, HV_TNCRS_UPPER, &phy_data);
	ret_val = hw->phy.ops.read_reg_page(hw, HV_TNCRS_LOWER, &phy_data);
4155 4156 4157 4158 4159 4160 4161
	if (!ret_val)
		adapter->stats.tncrs += phy_data;

release:
	hw->phy.ops.release(hw);
}

4162 4163 4164 4165
/**
 * e1000e_update_stats - Update the board statistics counters
 * @adapter: board private structure
 **/
J
Jeff Kirsher 已提交
4166
static void e1000e_update_stats(struct e1000_adapter *adapter)
4167
{
4168
	struct net_device *netdev = adapter->netdev;
4169 4170 4171 4172 4173 4174 4175 4176 4177 4178 4179 4180 4181 4182
	struct e1000_hw *hw = &adapter->hw;
	struct pci_dev *pdev = adapter->pdev;

	/*
	 * Prevent stats update while adapter is being reset, or if the pci
	 * connection is down.
	 */
	if (adapter->link_speed == 0)
		return;
	if (pci_channel_offline(pdev))
		return;

	adapter->stats.crcerrs += er32(CRCERRS);
	adapter->stats.gprc += er32(GPRC);
4183 4184
	adapter->stats.gorc += er32(GORCL);
	er32(GORCH); /* Clear gorc */
4185 4186 4187 4188 4189
	adapter->stats.bprc += er32(BPRC);
	adapter->stats.mprc += er32(MPRC);
	adapter->stats.roc += er32(ROC);

	adapter->stats.mpc += er32(MPC);
4190 4191 4192 4193 4194 4195 4196 4197 4198 4199 4200 4201 4202 4203 4204 4205 4206 4207 4208

	/* Half-duplex statistics */
	if (adapter->link_duplex == HALF_DUPLEX) {
		if (adapter->flags2 & FLAG2_HAS_PHY_STATS) {
			e1000e_update_phy_stats(adapter);
		} else {
			adapter->stats.scc += er32(SCC);
			adapter->stats.ecol += er32(ECOL);
			adapter->stats.mcc += er32(MCC);
			adapter->stats.latecol += er32(LATECOL);
			adapter->stats.dc += er32(DC);

			hw->mac.collision_delta = er32(COLC);

			if ((hw->mac.type != e1000_82574) &&
			    (hw->mac.type != e1000_82583))
				adapter->stats.tncrs += er32(TNCRS);
		}
		adapter->stats.colc += hw->mac.collision_delta;
4209
	}
4210

4211 4212 4213 4214 4215
	adapter->stats.xonrxc += er32(XONRXC);
	adapter->stats.xontxc += er32(XONTXC);
	adapter->stats.xoffrxc += er32(XOFFRXC);
	adapter->stats.xofftxc += er32(XOFFTXC);
	adapter->stats.gptc += er32(GPTC);
4216 4217
	adapter->stats.gotc += er32(GOTCL);
	er32(GOTCH); /* Clear gotc */
4218 4219 4220 4221 4222 4223 4224 4225 4226 4227 4228 4229 4230 4231 4232 4233 4234 4235
	adapter->stats.rnbc += er32(RNBC);
	adapter->stats.ruc += er32(RUC);

	adapter->stats.mptc += er32(MPTC);
	adapter->stats.bptc += er32(BPTC);

	/* used for adaptive IFS */

	hw->mac.tx_packet_delta = er32(TPT);
	adapter->stats.tpt += hw->mac.tx_packet_delta;

	adapter->stats.algnerrc += er32(ALGNERRC);
	adapter->stats.rxerrc += er32(RXERRC);
	adapter->stats.cexterr += er32(CEXTERR);
	adapter->stats.tsctc += er32(TSCTC);
	adapter->stats.tsctfc += er32(TSCTFC);

	/* Fill out the OS statistics structure */
4236 4237
	netdev->stats.multicast = adapter->stats.mprc;
	netdev->stats.collisions = adapter->stats.colc;
4238 4239 4240

	/* Rx Errors */

4241 4242 4243 4244
	/*
	 * RLEC on some newer hardware can be incorrect so build
	 * our own version based on RUC and ROC
	 */
4245
	netdev->stats.rx_errors = adapter->stats.rxerrc +
4246 4247 4248
		adapter->stats.crcerrs + adapter->stats.algnerrc +
		adapter->stats.ruc + adapter->stats.roc +
		adapter->stats.cexterr;
4249
	netdev->stats.rx_length_errors = adapter->stats.ruc +
4250
					      adapter->stats.roc;
4251 4252 4253
	netdev->stats.rx_crc_errors = adapter->stats.crcerrs;
	netdev->stats.rx_frame_errors = adapter->stats.algnerrc;
	netdev->stats.rx_missed_errors = adapter->stats.mpc;
4254 4255

	/* Tx Errors */
4256
	netdev->stats.tx_errors = adapter->stats.ecol +
4257
				       adapter->stats.latecol;
4258 4259 4260
	netdev->stats.tx_aborted_errors = adapter->stats.ecol;
	netdev->stats.tx_window_errors = adapter->stats.latecol;
	netdev->stats.tx_carrier_errors = adapter->stats.tncrs;
4261 4262 4263 4264 4265 4266 4267 4268 4269

	/* Tx Dropped needs to be maintained elsewhere */

	/* Management Stats */
	adapter->stats.mgptc += er32(MGTPTC);
	adapter->stats.mgprc += er32(MGTPRC);
	adapter->stats.mgpdc += er32(MGTPDC);
}

4270 4271 4272 4273 4274 4275 4276 4277 4278 4279 4280
/**
 * e1000_phy_read_status - Update the PHY register status snapshot
 * @adapter: board private structure
 **/
static void e1000_phy_read_status(struct e1000_adapter *adapter)
{
	struct e1000_hw *hw = &adapter->hw;
	struct e1000_phy_regs *phy = &adapter->phy_regs;

	if ((er32(STATUS) & E1000_STATUS_LU) &&
	    (adapter->hw.phy.media_type == e1000_media_type_copper)) {
4281 4282
		int ret_val;

4283 4284 4285 4286 4287 4288 4289 4290 4291
		ret_val  = e1e_rphy(hw, PHY_CONTROL, &phy->bmcr);
		ret_val |= e1e_rphy(hw, PHY_STATUS, &phy->bmsr);
		ret_val |= e1e_rphy(hw, PHY_AUTONEG_ADV, &phy->advertise);
		ret_val |= e1e_rphy(hw, PHY_LP_ABILITY, &phy->lpa);
		ret_val |= e1e_rphy(hw, PHY_AUTONEG_EXP, &phy->expansion);
		ret_val |= e1e_rphy(hw, PHY_1000T_CTRL, &phy->ctrl1000);
		ret_val |= e1e_rphy(hw, PHY_1000T_STATUS, &phy->stat1000);
		ret_val |= e1e_rphy(hw, PHY_EXT_STATUS, &phy->estatus);
		if (ret_val)
4292
			e_warn("Error reading PHY register\n");
4293 4294 4295 4296 4297 4298 4299 4300 4301 4302 4303 4304 4305 4306 4307 4308 4309 4310 4311
	} else {
		/*
		 * Do not read PHY registers if link is not up
		 * Set values to typical power-on defaults
		 */
		phy->bmcr = (BMCR_SPEED1000 | BMCR_ANENABLE | BMCR_FULLDPLX);
		phy->bmsr = (BMSR_100FULL | BMSR_100HALF | BMSR_10FULL |
			     BMSR_10HALF | BMSR_ESTATEN | BMSR_ANEGCAPABLE |
			     BMSR_ERCAP);
		phy->advertise = (ADVERTISE_PAUSE_ASYM | ADVERTISE_PAUSE_CAP |
				  ADVERTISE_ALL | ADVERTISE_CSMA);
		phy->lpa = 0;
		phy->expansion = EXPANSION_ENABLENPAGE;
		phy->ctrl1000 = ADVERTISE_1000FULL;
		phy->stat1000 = 0;
		phy->estatus = (ESTATUS_1000_TFULL | ESTATUS_1000_THALF);
	}
}

4312 4313 4314 4315 4316
static void e1000_print_link_info(struct e1000_adapter *adapter)
{
	struct e1000_hw *hw = &adapter->hw;
	u32 ctrl = er32(CTRL);

4317
	/* Link status message must follow this format for user tools */
4318 4319 4320 4321 4322 4323 4324
	printk(KERN_INFO "e1000e: %s NIC Link is Up %d Mbps %s Duplex, Flow Control: %s\n",
		adapter->netdev->name,
		adapter->link_speed,
		adapter->link_duplex == FULL_DUPLEX ? "Full" : "Half",
		(ctrl & E1000_CTRL_TFCE) && (ctrl & E1000_CTRL_RFCE) ? "Rx/Tx" :
		(ctrl & E1000_CTRL_RFCE) ? "Rx" :
		(ctrl & E1000_CTRL_TFCE) ? "Tx" : "None");
4325 4326
}

4327
static bool e1000e_has_link(struct e1000_adapter *adapter)
4328 4329
{
	struct e1000_hw *hw = &adapter->hw;
4330
	bool link_active = false;
4331 4332 4333 4334 4335 4336 4337 4338 4339 4340 4341 4342 4343 4344
	s32 ret_val = 0;

	/*
	 * get_link_status is set on LSC (link status) interrupt or
	 * Rx sequence error interrupt.  get_link_status will stay
	 * false until the check_for_link establishes link
	 * for copper adapters ONLY
	 */
	switch (hw->phy.media_type) {
	case e1000_media_type_copper:
		if (hw->mac.get_link_status) {
			ret_val = hw->mac.ops.check_for_link(hw);
			link_active = !hw->mac.get_link_status;
		} else {
4345
			link_active = true;
4346 4347 4348 4349 4350 4351 4352 4353 4354 4355 4356 4357 4358 4359 4360 4361 4362 4363
		}
		break;
	case e1000_media_type_fiber:
		ret_val = hw->mac.ops.check_for_link(hw);
		link_active = !!(er32(STATUS) & E1000_STATUS_LU);
		break;
	case e1000_media_type_internal_serdes:
		ret_val = hw->mac.ops.check_for_link(hw);
		link_active = adapter->hw.mac.serdes_has_link;
		break;
	default:
	case e1000_media_type_unknown:
		break;
	}

	if ((ret_val == E1000_ERR_PHY) && (hw->phy.type == e1000_phy_igp_3) &&
	    (er32(CTRL) & E1000_PHY_CTRL_GBE_DISABLE)) {
		/* See e1000_kmrn_lock_loss_workaround_ich8lan() */
4364
		e_info("Gigabit has been disabled, downgrading speed\n");
4365 4366 4367 4368 4369 4370 4371 4372 4373 4374 4375 4376 4377 4378 4379 4380 4381
	}

	return link_active;
}

static void e1000e_enable_receives(struct e1000_adapter *adapter)
{
	/* make sure the receive unit is started */
	if ((adapter->flags & FLAG_RX_NEEDS_RESTART) &&
	    (adapter->flags & FLAG_RX_RESTART_NOW)) {
		struct e1000_hw *hw = &adapter->hw;
		u32 rctl = er32(RCTL);
		ew32(RCTL, rctl | E1000_RCTL_EN);
		adapter->flags &= ~FLAG_RX_RESTART_NOW;
	}
}

4382 4383 4384 4385 4386 4387 4388 4389 4390 4391 4392 4393 4394 4395 4396 4397 4398 4399 4400
static void e1000e_check_82574_phy_workaround(struct e1000_adapter *adapter)
{
	struct e1000_hw *hw = &adapter->hw;

	/*
	 * With 82574 controllers, PHY needs to be checked periodically
	 * for hung state and reset, if two calls return true
	 */
	if (e1000_check_phy_82574(hw))
		adapter->phy_hang_count++;
	else
		adapter->phy_hang_count = 0;

	if (adapter->phy_hang_count > 1) {
		adapter->phy_hang_count = 0;
		schedule_work(&adapter->reset_task);
	}
}

4401 4402 4403 4404 4405 4406 4407 4408 4409 4410 4411 4412 4413 4414 4415 4416 4417 4418 4419 4420
/**
 * e1000_watchdog - Timer Call-back
 * @data: pointer to adapter cast into an unsigned long
 **/
static void e1000_watchdog(unsigned long data)
{
	struct e1000_adapter *adapter = (struct e1000_adapter *) data;

	/* Do the rest outside of interrupt context */
	schedule_work(&adapter->watchdog_task);

	/* TODO: make this use queue_delayed_work() */
}

static void e1000_watchdog_task(struct work_struct *work)
{
	struct e1000_adapter *adapter = container_of(work,
					struct e1000_adapter, watchdog_task);
	struct net_device *netdev = adapter->netdev;
	struct e1000_mac_info *mac = &adapter->hw.mac;
B
Bruce Allan 已提交
4421
	struct e1000_phy_info *phy = &adapter->hw.phy;
4422 4423 4424 4425
	struct e1000_ring *tx_ring = adapter->tx_ring;
	struct e1000_hw *hw = &adapter->hw;
	u32 link, tctl;

4426 4427 4428
	if (test_bit(__E1000_DOWN, &adapter->state))
		return;

4429
	link = e1000e_has_link(adapter);
4430
	if ((netif_carrier_ok(netdev)) && link) {
4431 4432 4433
		/* Cancel scheduled suspend requests. */
		pm_runtime_resume(netdev->dev.parent);

4434
		e1000e_enable_receives(adapter);
4435 4436 4437 4438 4439 4440 4441 4442 4443
		goto link_up;
	}

	if ((e1000e_enable_tx_pkt_filtering(hw)) &&
	    (adapter->mng_vlan_id != adapter->hw.mng_cookie.vlan_id))
		e1000_update_mng_vlan(adapter);

	if (link) {
		if (!netif_carrier_ok(netdev)) {
4444
			bool txb2b = true;
4445 4446 4447 4448

			/* Cancel scheduled suspend requests. */
			pm_runtime_resume(netdev->dev.parent);

4449
			/* update snapshot of PHY registers on LSC */
4450
			e1000_phy_read_status(adapter);
4451 4452 4453 4454
			mac->ops.get_link_up_info(&adapter->hw,
						   &adapter->link_speed,
						   &adapter->link_duplex);
			e1000_print_link_info(adapter);
4455 4456 4457 4458 4459 4460 4461 4462 4463 4464 4465 4466 4467 4468 4469
			/*
			 * On supported PHYs, check for duplex mismatch only
			 * if link has autonegotiated at 10/100 half
			 */
			if ((hw->phy.type == e1000_phy_igp_3 ||
			     hw->phy.type == e1000_phy_bm) &&
			    (hw->mac.autoneg == true) &&
			    (adapter->link_speed == SPEED_10 ||
			     adapter->link_speed == SPEED_100) &&
			    (adapter->link_duplex == HALF_DUPLEX)) {
				u16 autoneg_exp;

				e1e_rphy(hw, PHY_AUTONEG_EXP, &autoneg_exp);

				if (!(autoneg_exp & NWAY_ER_LP_NWAY_CAPS))
4470
					e_info("Autonegotiated half duplex but link partner cannot autoneg.  Try forcing full duplex if link gets many collisions.\n");
4471 4472
			}

4473
			/* adjust timeout factor according to speed/duplex */
4474 4475 4476
			adapter->tx_timeout_factor = 1;
			switch (adapter->link_speed) {
			case SPEED_10:
4477
				txb2b = false;
4478
				adapter->tx_timeout_factor = 16;
4479 4480
				break;
			case SPEED_100:
4481
				txb2b = false;
4482
				adapter->tx_timeout_factor = 10;
4483 4484 4485
				break;
			}

4486 4487 4488 4489
			/*
			 * workaround: re-program speed mode bit after
			 * link-up event
			 */
4490 4491 4492
			if ((adapter->flags & FLAG_TARC_SPEED_MODE_BIT) &&
			    !txb2b) {
				u32 tarc0;
4493
				tarc0 = er32(TARC(0));
4494
				tarc0 &= ~SPEED_MODE_BIT;
4495
				ew32(TARC(0), tarc0);
4496 4497
			}

4498 4499 4500 4501
			/*
			 * disable TSO for pcie and 10/100 speeds, to avoid
			 * some hardware issues
			 */
4502 4503 4504 4505
			if (!(adapter->flags & FLAG_TSO_FORCE)) {
				switch (adapter->link_speed) {
				case SPEED_10:
				case SPEED_100:
4506
					e_info("10/100 speed: disabling TSO\n");
4507 4508 4509 4510 4511 4512 4513 4514 4515 4516 4517 4518 4519
					netdev->features &= ~NETIF_F_TSO;
					netdev->features &= ~NETIF_F_TSO6;
					break;
				case SPEED_1000:
					netdev->features |= NETIF_F_TSO;
					netdev->features |= NETIF_F_TSO6;
					break;
				default:
					/* oops */
					break;
				}
			}

4520 4521 4522 4523
			/*
			 * enable transmits in the hardware, need to do this
			 * after setting TARC(0)
			 */
4524 4525 4526 4527
			tctl = er32(TCTL);
			tctl |= E1000_TCTL_EN;
			ew32(TCTL, tctl);

B
Bruce Allan 已提交
4528 4529 4530 4531 4532 4533 4534
                        /*
			 * Perform any post-link-up configuration before
			 * reporting link up.
			 */
			if (phy->ops.cfg_on_link_up)
				phy->ops.cfg_on_link_up(hw);

4535 4536 4537 4538 4539 4540 4541 4542 4543 4544
			netif_carrier_on(netdev);

			if (!test_bit(__E1000_DOWN, &adapter->state))
				mod_timer(&adapter->phy_info_timer,
					  round_jiffies(jiffies + 2 * HZ));
		}
	} else {
		if (netif_carrier_ok(netdev)) {
			adapter->link_speed = 0;
			adapter->link_duplex = 0;
4545 4546 4547
			/* Link status message must follow this format */
			printk(KERN_INFO "e1000e: %s NIC Link is Down\n",
			       adapter->netdev->name);
4548 4549 4550 4551 4552 4553 4554
			netif_carrier_off(netdev);
			if (!test_bit(__E1000_DOWN, &adapter->state))
				mod_timer(&adapter->phy_info_timer,
					  round_jiffies(jiffies + 2 * HZ));

			if (adapter->flags & FLAG_RX_NEEDS_RESTART)
				schedule_work(&adapter->reset_task);
4555 4556 4557
			else
				pm_schedule_suspend(netdev->dev.parent,
							LINK_TIMEOUT);
4558 4559 4560 4561
		}
	}

link_up:
J
Jeff Kirsher 已提交
4562
	spin_lock(&adapter->stats64_lock);
4563 4564 4565 4566 4567 4568 4569
	e1000e_update_stats(adapter);

	mac->tx_packet_delta = adapter->stats.tpt - adapter->tpt_old;
	adapter->tpt_old = adapter->stats.tpt;
	mac->collision_delta = adapter->stats.colc - adapter->colc_old;
	adapter->colc_old = adapter->stats.colc;

4570 4571 4572 4573
	adapter->gorc = adapter->stats.gorc - adapter->gorc_old;
	adapter->gorc_old = adapter->stats.gorc;
	adapter->gotc = adapter->stats.gotc - adapter->gotc_old;
	adapter->gotc_old = adapter->stats.gotc;
4574
	spin_unlock(&adapter->stats64_lock);
4575 4576 4577

	e1000e_update_adaptive(&adapter->hw);

4578 4579 4580 4581 4582 4583 4584 4585 4586 4587 4588
	if (!netif_carrier_ok(netdev) &&
	    (e1000_desc_unused(tx_ring) + 1 < tx_ring->count)) {
		/*
		 * 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).
		 */
		schedule_work(&adapter->reset_task);
		/* return immediately since reset is imminent */
		return;
4589 4590
	}

4591 4592 4593 4594 4595 4596 4597 4598 4599 4600 4601 4602 4603 4604 4605 4606
	/* Simple mode for Interrupt Throttle Rate (ITR) */
	if (adapter->itr_setting == 4) {
		/*
		 * Symmetric Tx/Rx gets a reduced ITR=2000;
		 * Total asymmetrical Tx or Rx gets ITR=8000;
		 * everyone else is between 2000-8000.
		 */
		u32 goc = (adapter->gotc + adapter->gorc) / 10000;
		u32 dif = (adapter->gotc > adapter->gorc ?
			    adapter->gotc - adapter->gorc :
			    adapter->gorc - adapter->gotc) / 10000;
		u32 itr = goc > 0 ? (dif * 6000 / goc + 2000) : 8000;

		ew32(ITR, 1000000000 / (itr * 256));
	}

4607
	/* Cause software interrupt to ensure Rx ring is cleaned */
4608 4609 4610 4611
	if (adapter->msix_entries)
		ew32(ICS, adapter->rx_ring->ims_val);
	else
		ew32(ICS, E1000_ICS_RXDMT0);
4612

4613 4614 4615
	/* flush pending descriptors to memory before detecting Tx hang */
	e1000e_flush_descriptors(adapter);

4616
	/* Force detection of hung controller every watchdog period */
4617
	adapter->detect_tx_hung = true;
4618

4619 4620 4621 4622
	/*
	 * With 82571 controllers, LAA may be overwritten due to controller
	 * reset from the other port. Set the appropriate LAA in RAR[0]
	 */
4623
	if (e1000e_get_laa_state_82571(hw))
4624
		hw->mac.ops.rar_set(hw, adapter->hw.mac.addr, 0);
4625

4626 4627 4628
	if (adapter->flags2 & FLAG2_CHECK_PHY_HANG)
		e1000e_check_82574_phy_workaround(adapter);

4629 4630 4631 4632 4633 4634 4635 4636 4637 4638
	/* Reset the timer */
	if (!test_bit(__E1000_DOWN, &adapter->state))
		mod_timer(&adapter->watchdog_timer,
			  round_jiffies(jiffies + 2 * HZ));
}

#define E1000_TX_FLAGS_CSUM		0x00000001
#define E1000_TX_FLAGS_VLAN		0x00000002
#define E1000_TX_FLAGS_TSO		0x00000004
#define E1000_TX_FLAGS_IPV4		0x00000008
4639
#define E1000_TX_FLAGS_NO_FCS		0x00000010
4640 4641 4642
#define E1000_TX_FLAGS_VLAN_MASK	0xffff0000
#define E1000_TX_FLAGS_VLAN_SHIFT	16

4643
static int e1000_tso(struct e1000_ring *tx_ring, struct sk_buff *skb)
4644 4645 4646 4647 4648 4649 4650 4651
{
	struct e1000_context_desc *context_desc;
	struct e1000_buffer *buffer_info;
	unsigned int i;
	u32 cmd_length = 0;
	u16 ipcse = 0, tucse, mss;
	u8 ipcss, ipcso, tucss, tucso, hdr_len;

4652 4653
	if (!skb_is_gso(skb))
		return 0;
4654

4655
	if (skb_header_cloned(skb)) {
4656 4657
		int err = pskb_expand_head(skb, 0, 0, GFP_ATOMIC);

4658 4659
		if (err)
			return err;
4660 4661
	}

4662 4663 4664 4665 4666 4667 4668 4669 4670 4671
	hdr_len = skb_transport_offset(skb) + tcp_hdrlen(skb);
	mss = skb_shinfo(skb)->gso_size;
	if (skb->protocol == htons(ETH_P_IP)) {
		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);
		cmd_length = E1000_TXD_CMD_IP;
		ipcse = skb_transport_offset(skb) - 1;
4672
	} else if (skb_is_gso_v6(skb)) {
4673 4674 4675 4676 4677 4678 4679 4680 4681 4682 4683 4684 4685 4686 4687 4688 4689 4690 4691 4692 4693 4694 4695 4696 4697 4698 4699 4700 4701 4702 4703 4704 4705 4706 4707 4708 4709 4710
		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);
		ipcse = 0;
	}
	ipcss = skb_network_offset(skb);
	ipcso = (void *)&(ip_hdr(skb)->check) - (void *)skb->data;
	tucss = skb_transport_offset(skb);
	tucso = (void *)&(tcp_hdr(skb)->check) - (void *)skb->data;
	tucse = 0;

	cmd_length |= (E1000_TXD_CMD_DEXT | E1000_TXD_CMD_TSE |
	               E1000_TXD_CMD_TCP | (skb->len - (hdr_len)));

	i = tx_ring->next_to_use;
	context_desc = E1000_CONTEXT_DESC(*tx_ring, i);
	buffer_info = &tx_ring->buffer_info[i];

	context_desc->lower_setup.ip_fields.ipcss  = ipcss;
	context_desc->lower_setup.ip_fields.ipcso  = ipcso;
	context_desc->lower_setup.ip_fields.ipcse  = cpu_to_le16(ipcse);
	context_desc->upper_setup.tcp_fields.tucss = tucss;
	context_desc->upper_setup.tcp_fields.tucso = tucso;
	context_desc->upper_setup.tcp_fields.tucse = cpu_to_le16(tucse);
	context_desc->tcp_seg_setup.fields.mss     = cpu_to_le16(mss);
	context_desc->tcp_seg_setup.fields.hdr_len = hdr_len;
	context_desc->cmd_and_length = cpu_to_le32(cmd_length);

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

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

	return 1;
4711 4712
}

4713
static bool e1000_tx_csum(struct e1000_ring *tx_ring, struct sk_buff *skb)
4714
{
4715
	struct e1000_adapter *adapter = tx_ring->adapter;
4716 4717 4718 4719
	struct e1000_context_desc *context_desc;
	struct e1000_buffer *buffer_info;
	unsigned int i;
	u8 css;
4720
	u32 cmd_len = E1000_TXD_CMD_DEXT;
4721
	__be16 protocol;
4722

4723 4724
	if (skb->ip_summed != CHECKSUM_PARTIAL)
		return 0;
4725

4726 4727 4728 4729 4730
	if (skb->protocol == cpu_to_be16(ETH_P_8021Q))
		protocol = vlan_eth_hdr(skb)->h_vlan_encapsulated_proto;
	else
		protocol = skb->protocol;

A
Arthur Jones 已提交
4731
	switch (protocol) {
4732
	case cpu_to_be16(ETH_P_IP):
4733 4734 4735
		if (ip_hdr(skb)->protocol == IPPROTO_TCP)
			cmd_len |= E1000_TXD_CMD_TCP;
		break;
4736
	case cpu_to_be16(ETH_P_IPV6):
4737 4738 4739 4740 4741 4742
		/* XXX not handling all IPV6 headers */
		if (ipv6_hdr(skb)->nexthdr == IPPROTO_TCP)
			cmd_len |= E1000_TXD_CMD_TCP;
		break;
	default:
		if (unlikely(net_ratelimit()))
4743 4744
			e_warn("checksum_partial proto=%x!\n",
			       be16_to_cpu(protocol));
4745
		break;
4746 4747
	}

4748
	css = skb_checksum_start_offset(skb);
4749 4750 4751 4752 4753 4754 4755 4756 4757 4758 4759 4760 4761 4762 4763 4764 4765 4766 4767 4768 4769 4770

	i = tx_ring->next_to_use;
	buffer_info = &tx_ring->buffer_info[i];
	context_desc = E1000_CONTEXT_DESC(*tx_ring, i);

	context_desc->lower_setup.ip_config = 0;
	context_desc->upper_setup.tcp_fields.tucss = css;
	context_desc->upper_setup.tcp_fields.tucso =
				css + skb->csum_offset;
	context_desc->upper_setup.tcp_fields.tucse = 0;
	context_desc->tcp_seg_setup.data = 0;
	context_desc->cmd_and_length = cpu_to_le32(cmd_len);

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

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

	return 1;
4771 4772 4773 4774 4775
}

#define E1000_MAX_PER_TXD	8192
#define E1000_MAX_TXD_PWR	12

4776 4777 4778
static int e1000_tx_map(struct e1000_ring *tx_ring, struct sk_buff *skb,
			unsigned int first, unsigned int max_per_txd,
			unsigned int nr_frags, unsigned int mss)
4779
{
4780
	struct e1000_adapter *adapter = tx_ring->adapter;
4781
	struct pci_dev *pdev = adapter->pdev;
4782
	struct e1000_buffer *buffer_info;
J
Jesse Brandeburg 已提交
4783
	unsigned int len = skb_headlen(skb);
4784
	unsigned int offset = 0, size, count = 0, i;
4785
	unsigned int f, bytecount, segs;
4786 4787 4788 4789

	i = tx_ring->next_to_use;

	while (len) {
4790
		buffer_info = &tx_ring->buffer_info[i];
4791 4792 4793 4794 4795
		size = min(len, max_per_txd);

		buffer_info->length = size;
		buffer_info->time_stamp = jiffies;
		buffer_info->next_to_watch = i;
4796 4797
		buffer_info->dma = dma_map_single(&pdev->dev,
						  skb->data + offset,
4798
						  size, DMA_TO_DEVICE);
4799
		buffer_info->mapped_as_page = false;
4800
		if (dma_mapping_error(&pdev->dev, buffer_info->dma))
4801
			goto dma_error;
4802 4803 4804

		len -= size;
		offset += size;
4805
		count++;
4806 4807 4808 4809 4810 4811

		if (len) {
			i++;
			if (i == tx_ring->count)
				i = 0;
		}
4812 4813 4814
	}

	for (f = 0; f < nr_frags; f++) {
E
Eric Dumazet 已提交
4815
		const struct skb_frag_struct *frag;
4816 4817

		frag = &skb_shinfo(skb)->frags[f];
E
Eric Dumazet 已提交
4818
		len = skb_frag_size(frag);
4819
		offset = 0;
4820 4821

		while (len) {
4822 4823 4824 4825
			i++;
			if (i == tx_ring->count)
				i = 0;

4826 4827 4828 4829 4830 4831
			buffer_info = &tx_ring->buffer_info[i];
			size = min(len, max_per_txd);

			buffer_info->length = size;
			buffer_info->time_stamp = jiffies;
			buffer_info->next_to_watch = i;
4832 4833
			buffer_info->dma = skb_frag_dma_map(&pdev->dev, frag,
						offset, size, DMA_TO_DEVICE);
4834
			buffer_info->mapped_as_page = true;
4835
			if (dma_mapping_error(&pdev->dev, buffer_info->dma))
4836
				goto dma_error;
4837 4838 4839 4840 4841 4842 4843

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

4844
	segs = skb_shinfo(skb)->gso_segs ? : 1;
4845 4846 4847
	/* multiply data chunks by size of headers */
	bytecount = ((segs - 1) * skb_headlen(skb)) + skb->len;

4848
	tx_ring->buffer_info[i].skb = skb;
4849 4850
	tx_ring->buffer_info[i].segs = segs;
	tx_ring->buffer_info[i].bytecount = bytecount;
4851 4852 4853
	tx_ring->buffer_info[first].next_to_watch = i;

	return count;
4854 4855

dma_error:
4856
	dev_err(&pdev->dev, "Tx DMA map failed\n");
4857
	buffer_info->dma = 0;
4858
	if (count)
4859
		count--;
4860 4861

	while (count--) {
4862
		if (i == 0)
4863
			i += tx_ring->count;
4864
		i--;
4865
		buffer_info = &tx_ring->buffer_info[i];
4866
		e1000_put_txbuf(tx_ring, buffer_info);
4867 4868 4869
	}

	return 0;
4870 4871
}

4872
static void e1000_tx_queue(struct e1000_ring *tx_ring, int tx_flags, int count)
4873
{
4874
	struct e1000_adapter *adapter = tx_ring->adapter;
4875 4876 4877 4878 4879 4880 4881 4882 4883 4884 4885 4886 4887 4888 4889 4890 4891 4892 4893 4894 4895 4896 4897 4898
	struct e1000_tx_desc *tx_desc = NULL;
	struct e1000_buffer *buffer_info;
	u32 txd_upper = 0, txd_lower = E1000_TXD_CMD_IFCS;
	unsigned int i;

	if (tx_flags & E1000_TX_FLAGS_TSO) {
		txd_lower |= E1000_TXD_CMD_DEXT | E1000_TXD_DTYP_D |
			     E1000_TXD_CMD_TSE;
		txd_upper |= E1000_TXD_POPTS_TXSM << 8;

		if (tx_flags & E1000_TX_FLAGS_IPV4)
			txd_upper |= E1000_TXD_POPTS_IXSM << 8;
	}

	if (tx_flags & E1000_TX_FLAGS_CSUM) {
		txd_lower |= E1000_TXD_CMD_DEXT | E1000_TXD_DTYP_D;
		txd_upper |= E1000_TXD_POPTS_TXSM << 8;
	}

	if (tx_flags & E1000_TX_FLAGS_VLAN) {
		txd_lower |= E1000_TXD_CMD_VLE;
		txd_upper |= (tx_flags & E1000_TX_FLAGS_VLAN_MASK);
	}

4899 4900 4901
	if (unlikely(tx_flags & E1000_TX_FLAGS_NO_FCS))
		txd_lower &= ~(E1000_TXD_CMD_IFCS);

4902 4903
	i = tx_ring->next_to_use;

4904
	do {
4905 4906 4907 4908 4909 4910 4911 4912 4913 4914
		buffer_info = &tx_ring->buffer_info[i];
		tx_desc = E1000_TX_DESC(*tx_ring, i);
		tx_desc->buffer_addr = cpu_to_le64(buffer_info->dma);
		tx_desc->lower.data =
			cpu_to_le32(txd_lower | buffer_info->length);
		tx_desc->upper.data = cpu_to_le32(txd_upper);

		i++;
		if (i == tx_ring->count)
			i = 0;
4915
	} while (--count > 0);
4916 4917 4918

	tx_desc->lower.data |= cpu_to_le32(adapter->txd_cmd);

4919 4920 4921 4922
	/* txd_cmd re-enables FCS, so we'll re-disable it here as desired. */
	if (unlikely(tx_flags & E1000_TX_FLAGS_NO_FCS))
		tx_desc->lower.data &= ~(cpu_to_le32(E1000_TXD_CMD_IFCS));

4923 4924
	/*
	 * Force memory writes to complete before letting h/w
4925 4926
	 * know there are new descriptors to fetch.  (Only
	 * applicable for weak-ordered memory model archs,
4927 4928
	 * such as IA-64).
	 */
4929 4930 4931
	wmb();

	tx_ring->next_to_use = i;
4932 4933

	if (adapter->flags2 & FLAG2_PCIM2PCI_ARBITER_WA)
4934
		e1000e_update_tdt_wa(tx_ring, i);
4935
	else
4936
		writel(i, tx_ring->tail);
4937

4938 4939 4940 4941
	/*
	 * we need this if more than one processor can write to our tail
	 * at a time, it synchronizes IO on IA64/Altix systems
	 */
4942 4943 4944 4945 4946 4947 4948 4949 4950 4951 4952
	mmiowb();
}

#define MINIMUM_DHCP_PACKET_SIZE 282
static int e1000_transfer_dhcp_info(struct e1000_adapter *adapter,
				    struct sk_buff *skb)
{
	struct e1000_hw *hw =  &adapter->hw;
	u16 length, offset;

	if (vlan_tx_tag_present(skb)) {
4953 4954
		if (!((vlan_tx_tag_get(skb) == adapter->hw.mng_cookie.vlan_id) &&
		    (adapter->hw.mng_cookie.status &
4955 4956 4957 4958 4959 4960 4961 4962 4963 4964 4965 4966 4967 4968 4969 4970 4971 4972 4973 4974 4975 4976 4977 4978 4979 4980 4981 4982 4983
			E1000_MNG_DHCP_COOKIE_STATUS_VLAN)))
			return 0;
	}

	if (skb->len <= MINIMUM_DHCP_PACKET_SIZE)
		return 0;

	if (((struct ethhdr *) skb->data)->h_proto != htons(ETH_P_IP))
		return 0;

	{
		const struct iphdr *ip = (struct iphdr *)((u8 *)skb->data+14);
		struct udphdr *udp;

		if (ip->protocol != IPPROTO_UDP)
			return 0;

		udp = (struct udphdr *)((u8 *)ip + (ip->ihl << 2));
		if (ntohs(udp->dest) != 67)
			return 0;

		offset = (u8 *)udp + 8 - skb->data;
		length = skb->len - offset;
		return e1000e_mng_write_dhcp_info(hw, (u8 *)udp + 8, length);
	}

	return 0;
}

4984
static int __e1000_maybe_stop_tx(struct e1000_ring *tx_ring, int size)
4985
{
4986
	struct e1000_adapter *adapter = tx_ring->adapter;
4987

4988
	netif_stop_queue(adapter->netdev);
4989 4990
	/*
	 * Herbert's original patch had:
4991
	 *  smp_mb__after_netif_stop_queue();
4992 4993
	 * but since that doesn't exist yet, just open code it.
	 */
4994 4995
	smp_mb();

4996 4997 4998 4999
	/*
	 * We need to check again in a case another CPU has just
	 * made room available.
	 */
5000
	if (e1000_desc_unused(tx_ring) < size)
5001 5002 5003
		return -EBUSY;

	/* A reprieve! */
5004
	netif_start_queue(adapter->netdev);
5005 5006 5007 5008
	++adapter->restart_queue;
	return 0;
}

5009
static int e1000_maybe_stop_tx(struct e1000_ring *tx_ring, int size)
5010
{
5011
	if (e1000_desc_unused(tx_ring) >= size)
5012
		return 0;
5013
	return __e1000_maybe_stop_tx(tx_ring, size);
5014 5015
}

5016
#define TXD_USE_COUNT(S, X) (((S) >> (X)) + 1)
5017 5018
static netdev_tx_t e1000_xmit_frame(struct sk_buff *skb,
				    struct net_device *netdev)
5019 5020 5021 5022 5023 5024 5025
{
	struct e1000_adapter *adapter = netdev_priv(netdev);
	struct e1000_ring *tx_ring = adapter->tx_ring;
	unsigned int first;
	unsigned int max_per_txd = E1000_MAX_PER_TXD;
	unsigned int max_txd_pwr = E1000_MAX_TXD_PWR;
	unsigned int tx_flags = 0;
E
Eric Dumazet 已提交
5026
	unsigned int len = skb_headlen(skb);
5027 5028
	unsigned int nr_frags;
	unsigned int mss;
5029 5030 5031 5032 5033 5034 5035 5036 5037 5038 5039 5040 5041 5042 5043
	int count = 0;
	int tso;
	unsigned int f;

	if (test_bit(__E1000_DOWN, &adapter->state)) {
		dev_kfree_skb_any(skb);
		return NETDEV_TX_OK;
	}

	if (skb->len <= 0) {
		dev_kfree_skb_any(skb);
		return NETDEV_TX_OK;
	}

	mss = skb_shinfo(skb)->gso_size;
5044 5045
	/*
	 * The controller does a simple calculation to
5046 5047 5048 5049
	 * make sure there is enough room in the FIFO before
	 * initiating the DMA for each buffer.  The calc is:
	 * 4 = ceil(buffer len/mss).  To make sure we don't
	 * overrun the FIFO, adjust the max buffer len if mss
5050 5051
	 * drops.
	 */
5052 5053 5054 5055 5056
	if (mss) {
		u8 hdr_len;
		max_per_txd = min(mss << 2, max_per_txd);
		max_txd_pwr = fls(max_per_txd) - 1;

5057 5058 5059 5060 5061
		/*
		 * TSO Workaround for 82571/2/3 Controllers -- if skb->data
		 * points to just header, pull a few bytes of payload from
		 * frags into skb->data
		 */
5062
		hdr_len = skb_transport_offset(skb) + tcp_hdrlen(skb);
5063 5064 5065 5066
		/*
		 * we do this workaround for ES2LAN, but it is un-necessary,
		 * avoiding it could save a lot of cycles
		 */
5067
		if (skb->data_len && (hdr_len == len)) {
5068 5069
			unsigned int pull_size;

5070
			pull_size = min_t(unsigned int, 4, skb->data_len);
5071
			if (!__pskb_pull_tail(skb, pull_size)) {
5072
				e_err("__pskb_pull_tail failed.\n");
5073 5074 5075
				dev_kfree_skb_any(skb);
				return NETDEV_TX_OK;
			}
E
Eric Dumazet 已提交
5076
			len = skb_headlen(skb);
5077 5078 5079 5080 5081 5082 5083 5084 5085 5086 5087 5088
		}
	}

	/* reserve a descriptor for the offload context */
	if ((mss) || (skb->ip_summed == CHECKSUM_PARTIAL))
		count++;
	count++;

	count += TXD_USE_COUNT(len, max_txd_pwr);

	nr_frags = skb_shinfo(skb)->nr_frags;
	for (f = 0; f < nr_frags; f++)
E
Eric Dumazet 已提交
5089
		count += TXD_USE_COUNT(skb_frag_size(&skb_shinfo(skb)->frags[f]),
5090 5091 5092 5093 5094
				       max_txd_pwr);

	if (adapter->hw.mac.tx_pkt_filtering)
		e1000_transfer_dhcp_info(adapter, skb);

5095 5096 5097 5098
	/*
	 * need: count + 2 desc gap to keep tail from touching
	 * head, otherwise try next time
	 */
5099
	if (e1000_maybe_stop_tx(tx_ring, count + 2))
5100 5101
		return NETDEV_TX_BUSY;

5102
	if (vlan_tx_tag_present(skb)) {
5103 5104 5105 5106 5107 5108
		tx_flags |= E1000_TX_FLAGS_VLAN;
		tx_flags |= (vlan_tx_tag_get(skb) << E1000_TX_FLAGS_VLAN_SHIFT);
	}

	first = tx_ring->next_to_use;

5109
	tso = e1000_tso(tx_ring, skb);
5110 5111 5112 5113 5114 5115 5116
	if (tso < 0) {
		dev_kfree_skb_any(skb);
		return NETDEV_TX_OK;
	}

	if (tso)
		tx_flags |= E1000_TX_FLAGS_TSO;
5117
	else if (e1000_tx_csum(tx_ring, skb))
5118 5119
		tx_flags |= E1000_TX_FLAGS_CSUM;

5120 5121
	/*
	 * Old method was to assume IPv4 packet by default if TSO was enabled.
5122
	 * 82571 hardware supports TSO capabilities for IPv6 as well...
5123 5124
	 * no longer assume, we must.
	 */
5125 5126 5127
	if (skb->protocol == htons(ETH_P_IP))
		tx_flags |= E1000_TX_FLAGS_IPV4;

5128 5129 5130
	if (unlikely(skb->no_fcs))
		tx_flags |= E1000_TX_FLAGS_NO_FCS;

L
Lucas De Marchi 已提交
5131
	/* if count is 0 then mapping error has occurred */
5132
	count = e1000_tx_map(tx_ring, skb, first, max_per_txd, nr_frags, mss);
5133
	if (count) {
5134 5135
		skb_tx_timestamp(skb);

5136
		netdev_sent_queue(netdev, skb->len);
5137
		e1000_tx_queue(tx_ring, tx_flags, count);
5138
		/* Make sure there is space in the ring for the next send. */
5139
		e1000_maybe_stop_tx(tx_ring, MAX_SKB_FRAGS + 2);
5140 5141

	} else {
5142
		dev_kfree_skb_any(skb);
5143 5144
		tx_ring->buffer_info[first].time_stamp = 0;
		tx_ring->next_to_use = first;
5145 5146 5147 5148 5149 5150 5151 5152 5153 5154 5155 5156 5157 5158 5159 5160 5161 5162 5163 5164 5165 5166 5167
	}

	return NETDEV_TX_OK;
}

/**
 * e1000_tx_timeout - Respond to a Tx Hang
 * @netdev: network interface device structure
 **/
static void e1000_tx_timeout(struct net_device *netdev)
{
	struct e1000_adapter *adapter = netdev_priv(netdev);

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

static void e1000_reset_task(struct work_struct *work)
{
	struct e1000_adapter *adapter;
	adapter = container_of(work, struct e1000_adapter, reset_task);

5168 5169 5170 5171
	/* don't run the task if already down */
	if (test_bit(__E1000_DOWN, &adapter->state))
		return;

5172 5173 5174 5175 5176
	if (!((adapter->flags & FLAG_RX_NEEDS_RESTART) &&
	      (adapter->flags & FLAG_RX_RESTART_NOW))) {
		e1000e_dump(adapter);
		e_err("Reset adapter\n");
	}
5177 5178 5179 5180
	e1000e_reinit_locked(adapter);
}

/**
J
Jeff Kirsher 已提交
5181
 * e1000_get_stats64 - Get System Network Statistics
5182
 * @netdev: network interface device structure
J
Jeff Kirsher 已提交
5183
 * @stats: rtnl_link_stats64 pointer
5184 5185 5186
 *
 * Returns the address of the device statistics structure.
 **/
J
Jeff Kirsher 已提交
5187 5188
struct rtnl_link_stats64 *e1000e_get_stats64(struct net_device *netdev,
                                             struct rtnl_link_stats64 *stats)
5189
{
J
Jeff Kirsher 已提交
5190 5191 5192 5193 5194 5195 5196 5197 5198 5199 5200 5201 5202 5203 5204 5205 5206 5207 5208 5209 5210 5211 5212 5213 5214 5215 5216 5217 5218 5219 5220 5221 5222 5223 5224 5225 5226 5227 5228 5229
	struct e1000_adapter *adapter = netdev_priv(netdev);

	memset(stats, 0, sizeof(struct rtnl_link_stats64));
	spin_lock(&adapter->stats64_lock);
	e1000e_update_stats(adapter);
	/* Fill out the OS statistics structure */
	stats->rx_bytes = adapter->stats.gorc;
	stats->rx_packets = adapter->stats.gprc;
	stats->tx_bytes = adapter->stats.gotc;
	stats->tx_packets = adapter->stats.gptc;
	stats->multicast = adapter->stats.mprc;
	stats->collisions = adapter->stats.colc;

	/* Rx Errors */

	/*
	 * RLEC on some newer hardware can be incorrect so build
	 * our own version based on RUC and ROC
	 */
	stats->rx_errors = adapter->stats.rxerrc +
		adapter->stats.crcerrs + adapter->stats.algnerrc +
		adapter->stats.ruc + adapter->stats.roc +
		adapter->stats.cexterr;
	stats->rx_length_errors = adapter->stats.ruc +
					      adapter->stats.roc;
	stats->rx_crc_errors = adapter->stats.crcerrs;
	stats->rx_frame_errors = adapter->stats.algnerrc;
	stats->rx_missed_errors = adapter->stats.mpc;

	/* Tx Errors */
	stats->tx_errors = adapter->stats.ecol +
				       adapter->stats.latecol;
	stats->tx_aborted_errors = adapter->stats.ecol;
	stats->tx_window_errors = adapter->stats.latecol;
	stats->tx_carrier_errors = adapter->stats.tncrs;

	/* Tx Dropped needs to be maintained elsewhere */

	spin_unlock(&adapter->stats64_lock);
	return stats;
5230 5231 5232 5233 5234 5235 5236 5237 5238 5239 5240 5241 5242 5243
}

/**
 * e1000_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 e1000_change_mtu(struct net_device *netdev, int new_mtu)
{
	struct e1000_adapter *adapter = netdev_priv(netdev);
	int max_frame = new_mtu + ETH_HLEN + ETH_FCS_LEN;

5244
	/* Jumbo frame support */
5245 5246 5247 5248 5249 5250 5251 5252 5253 5254 5255 5256 5257 5258 5259 5260
	if (max_frame > ETH_FRAME_LEN + ETH_FCS_LEN) {
		if (!(adapter->flags & FLAG_HAS_JUMBO_FRAMES)) {
			e_err("Jumbo Frames not supported.\n");
			return -EINVAL;
		}

		/*
		 * IP payload checksum (enabled with jumbos/packet-split when
		 * Rx checksum is enabled) and generation of RSS hash is
		 * mutually exclusive in the hardware.
		 */
		if ((netdev->features & NETIF_F_RXCSUM) &&
		    (netdev->features & NETIF_F_RXHASH)) {
			e_err("Jumbo frames cannot be enabled when both receive checksum offload and receive hashing are enabled.  Disable one of the receive offload features before enabling jumbos.\n");
			return -EINVAL;
		}
5261 5262
	}

5263 5264 5265 5266
	/* Supported frame sizes */
	if ((new_mtu < ETH_ZLEN + ETH_FCS_LEN + VLAN_HLEN) ||
	    (max_frame > adapter->max_hw_frame_size)) {
		e_err("Unsupported MTU setting\n");
5267 5268 5269
		return -EINVAL;
	}

5270 5271 5272 5273
	/* Jumbo frame workaround on 82579 requires CRC be stripped */
	if ((adapter->hw.mac.type == e1000_pch2lan) &&
	    !(adapter->flags2 & FLAG2_CRC_STRIPPING) &&
	    (new_mtu > ETH_DATA_LEN)) {
5274
		e_err("Jumbo Frames not supported on 82579 when CRC stripping is disabled.\n");
5275 5276 5277
		return -EINVAL;
	}

5278 5279 5280 5281 5282 5283 5284 5285
	/* 82573 Errata 17 */
	if (((adapter->hw.mac.type == e1000_82573) ||
	     (adapter->hw.mac.type == e1000_82574)) &&
	    (max_frame > ETH_FRAME_LEN + ETH_FCS_LEN)) {
		adapter->flags2 |= FLAG2_DISABLE_ASPM_L1;
		e1000e_disable_aspm(adapter->pdev, PCIE_LINK_STATE_L1);
	}

5286
	while (test_and_set_bit(__E1000_RESETTING, &adapter->state))
5287
		usleep_range(1000, 2000);
5288
	/* e1000e_down -> e1000e_reset dependent on max_frame_size & mtu */
5289
	adapter->max_frame_size = max_frame;
5290 5291
	e_info("changing MTU from %d to %d\n", netdev->mtu, new_mtu);
	netdev->mtu = new_mtu;
5292 5293 5294
	if (netif_running(netdev))
		e1000e_down(adapter);

5295 5296
	/*
	 * NOTE: netdev_alloc_skb reserves 16 bytes, and typically NET_IP_ALIGN
5297 5298
	 * means we reserve 2 more, this pushes us to allocate from the next
	 * larger slab size.
5299
	 * i.e. RXBUFFER_2048 --> size-4096 slab
5300 5301
	 * However with the new *_jumbo_rx* routines, jumbo receives will use
	 * fragmented skbs
5302
	 */
5303

5304
	if (max_frame <= 2048)
5305 5306 5307 5308 5309 5310 5311 5312
		adapter->rx_buffer_len = 2048;
	else
		adapter->rx_buffer_len = 4096;

	/* adjust allocation if LPE protects us, and we aren't using SBP */
	if ((max_frame == ETH_FRAME_LEN + ETH_FCS_LEN) ||
	     (max_frame == ETH_FRAME_LEN + VLAN_HLEN + ETH_FCS_LEN))
		adapter->rx_buffer_len = ETH_FRAME_LEN + VLAN_HLEN
5313
					 + ETH_FCS_LEN;
5314 5315 5316 5317 5318 5319 5320 5321 5322 5323 5324 5325 5326 5327 5328 5329 5330

	if (netif_running(netdev))
		e1000e_up(adapter);
	else
		e1000e_reset(adapter);

	clear_bit(__E1000_RESETTING, &adapter->state);

	return 0;
}

static int e1000_mii_ioctl(struct net_device *netdev, struct ifreq *ifr,
			   int cmd)
{
	struct e1000_adapter *adapter = netdev_priv(netdev);
	struct mii_ioctl_data *data = if_mii(ifr);

5331
	if (adapter->hw.phy.media_type != e1000_media_type_copper)
5332 5333 5334 5335 5336 5337 5338
		return -EOPNOTSUPP;

	switch (cmd) {
	case SIOCGMIIPHY:
		data->phy_id = adapter->hw.phy.addr;
		break;
	case SIOCGMIIREG:
5339 5340
		e1000_phy_read_status(adapter);

5341 5342 5343 5344 5345 5346 5347 5348 5349 5350 5351 5352 5353 5354 5355 5356 5357 5358 5359 5360 5361 5362 5363 5364 5365 5366 5367 5368 5369 5370 5371 5372
		switch (data->reg_num & 0x1F) {
		case MII_BMCR:
			data->val_out = adapter->phy_regs.bmcr;
			break;
		case MII_BMSR:
			data->val_out = adapter->phy_regs.bmsr;
			break;
		case MII_PHYSID1:
			data->val_out = (adapter->hw.phy.id >> 16);
			break;
		case MII_PHYSID2:
			data->val_out = (adapter->hw.phy.id & 0xFFFF);
			break;
		case MII_ADVERTISE:
			data->val_out = adapter->phy_regs.advertise;
			break;
		case MII_LPA:
			data->val_out = adapter->phy_regs.lpa;
			break;
		case MII_EXPANSION:
			data->val_out = adapter->phy_regs.expansion;
			break;
		case MII_CTRL1000:
			data->val_out = adapter->phy_regs.ctrl1000;
			break;
		case MII_STAT1000:
			data->val_out = adapter->phy_regs.stat1000;
			break;
		case MII_ESTATUS:
			data->val_out = adapter->phy_regs.estatus;
			break;
		default:
5373 5374 5375 5376 5377 5378 5379 5380 5381 5382 5383 5384 5385 5386 5387 5388 5389 5390 5391 5392 5393 5394
			return -EIO;
		}
		break;
	case SIOCSMIIREG:
	default:
		return -EOPNOTSUPP;
	}
	return 0;
}

static int e1000_ioctl(struct net_device *netdev, struct ifreq *ifr, int cmd)
{
	switch (cmd) {
	case SIOCGMIIPHY:
	case SIOCGMIIREG:
	case SIOCSMIIREG:
		return e1000_mii_ioctl(netdev, ifr, cmd);
	default:
		return -EOPNOTSUPP;
	}
}

5395 5396 5397 5398
static int e1000_init_phy_wakeup(struct e1000_adapter *adapter, u32 wufc)
{
	struct e1000_hw *hw = &adapter->hw;
	u32 i, mac_reg;
5399
	u16 phy_reg, wuc_enable;
5400 5401 5402
	int retval = 0;

	/* copy MAC RARs to PHY RARs */
5403
	e1000_copy_rx_addrs_to_phy_ich8lan(hw);
5404

5405 5406 5407 5408 5409 5410 5411 5412 5413
	retval = hw->phy.ops.acquire(hw);
	if (retval) {
		e_err("Could not acquire PHY\n");
		return retval;
	}

	/* Enable access to wakeup registers on and set page to BM_WUC_PAGE */
	retval = e1000_enable_phy_wakeup_reg_access_bm(hw, &wuc_enable);
	if (retval)
5414
		goto release;
5415 5416

	/* copy MAC MTA to PHY MTA - only needed for pchlan */
5417 5418
	for (i = 0; i < adapter->hw.mac.mta_reg_count; i++) {
		mac_reg = E1000_READ_REG_ARRAY(hw, E1000_MTA, i);
5419 5420 5421 5422
		hw->phy.ops.write_reg_page(hw, BM_MTA(i),
					   (u16)(mac_reg & 0xFFFF));
		hw->phy.ops.write_reg_page(hw, BM_MTA(i) + 1,
					   (u16)((mac_reg >> 16) & 0xFFFF));
5423 5424 5425
	}

	/* configure PHY Rx Control register */
5426
	hw->phy.ops.read_reg_page(&adapter->hw, BM_RCTL, &phy_reg);
5427 5428 5429 5430 5431 5432 5433 5434 5435 5436 5437 5438 5439 5440 5441 5442
	mac_reg = er32(RCTL);
	if (mac_reg & E1000_RCTL_UPE)
		phy_reg |= BM_RCTL_UPE;
	if (mac_reg & E1000_RCTL_MPE)
		phy_reg |= BM_RCTL_MPE;
	phy_reg &= ~(BM_RCTL_MO_MASK);
	if (mac_reg & E1000_RCTL_MO_3)
		phy_reg |= (((mac_reg & E1000_RCTL_MO_3) >> E1000_RCTL_MO_SHIFT)
				<< BM_RCTL_MO_SHIFT);
	if (mac_reg & E1000_RCTL_BAM)
		phy_reg |= BM_RCTL_BAM;
	if (mac_reg & E1000_RCTL_PMCF)
		phy_reg |= BM_RCTL_PMCF;
	mac_reg = er32(CTRL);
	if (mac_reg & E1000_CTRL_RFCE)
		phy_reg |= BM_RCTL_RFCE;
5443
	hw->phy.ops.write_reg_page(&adapter->hw, BM_RCTL, phy_reg);
5444 5445 5446 5447 5448 5449

	/* enable PHY wakeup in MAC register */
	ew32(WUFC, wufc);
	ew32(WUC, E1000_WUC_PHY_WAKE | E1000_WUC_PME_EN);

	/* configure and enable PHY wakeup in PHY registers */
5450 5451
	hw->phy.ops.write_reg_page(&adapter->hw, BM_WUFC, wufc);
	hw->phy.ops.write_reg_page(&adapter->hw, BM_WUC, E1000_WUC_PME_EN);
5452 5453

	/* activate PHY wakeup */
5454 5455
	wuc_enable |= BM_WUC_ENABLE_BIT | BM_WUC_HOST_WU_BIT;
	retval = e1000_disable_phy_wakeup_reg_access_bm(hw, &wuc_enable);
5456 5457
	if (retval)
		e_err("Could not set PHY Host Wakeup bit\n");
5458
release:
5459
	hw->phy.ops.release(hw);
5460 5461 5462 5463

	return retval;
}

5464 5465
static int __e1000_shutdown(struct pci_dev *pdev, bool *enable_wake,
			    bool runtime)
5466 5467 5468 5469 5470
{
	struct net_device *netdev = pci_get_drvdata(pdev);
	struct e1000_adapter *adapter = netdev_priv(netdev);
	struct e1000_hw *hw = &adapter->hw;
	u32 ctrl, ctrl_ext, rctl, status;
5471 5472
	/* Runtime suspend should only enable wakeup for link changes */
	u32 wufc = runtime ? E1000_WUFC_LNKC : adapter->wol;
5473 5474 5475 5476 5477
	int retval = 0;

	netif_device_detach(netdev);

	if (netif_running(netdev)) {
5478 5479 5480 5481 5482
		int count = E1000_CHECK_RESET_COUNT;

		while (test_bit(__E1000_RESETTING, &adapter->state) && count--)
			usleep_range(10000, 20000);

5483 5484 5485 5486
		WARN_ON(test_bit(__E1000_RESETTING, &adapter->state));
		e1000e_down(adapter);
		e1000_free_irq(adapter);
	}
5487
	e1000e_reset_interrupt_capability(adapter);
5488 5489 5490 5491 5492 5493 5494 5495 5496 5497 5498

	retval = pci_save_state(pdev);
	if (retval)
		return retval;

	status = er32(STATUS);
	if (status & E1000_STATUS_LU)
		wufc &= ~E1000_WUFC_LNKC;

	if (wufc) {
		e1000_setup_rctl(adapter);
5499
		e1000e_set_rx_mode(netdev);
5500 5501 5502 5503 5504 5505 5506 5507 5508 5509 5510 5511 5512

		/* turn on all-multi mode if wake on multicast is enabled */
		if (wufc & E1000_WUFC_MC) {
			rctl = er32(RCTL);
			rctl |= E1000_RCTL_MPE;
			ew32(RCTL, rctl);
		}

		ctrl = er32(CTRL);
		/* advertise wake from D3Cold */
		#define E1000_CTRL_ADVD3WUC 0x00100000
		/* phy power management enable */
		#define E1000_CTRL_EN_PHY_PWR_MGMT 0x00200000
5513 5514 5515
		ctrl |= E1000_CTRL_ADVD3WUC;
		if (!(adapter->flags2 & FLAG2_HAS_PHY_WAKEUP))
			ctrl |= E1000_CTRL_EN_PHY_PWR_MGMT;
5516 5517
		ew32(CTRL, ctrl);

5518 5519 5520
		if (adapter->hw.phy.media_type == e1000_media_type_fiber ||
		    adapter->hw.phy.media_type ==
		    e1000_media_type_internal_serdes) {
5521 5522
			/* keep the laser running in D3 */
			ctrl_ext = er32(CTRL_EXT);
5523
			ctrl_ext |= E1000_CTRL_EXT_SDP3_DATA;
5524 5525 5526
			ew32(CTRL_EXT, ctrl_ext);
		}

5527
		if (adapter->flags & FLAG_IS_ICH)
5528
			e1000_suspend_workarounds_ich8lan(&adapter->hw);
5529

5530 5531 5532
		/* Allow time for pending master requests to run */
		e1000e_disable_pcie_master(&adapter->hw);

5533
		if (adapter->flags2 & FLAG2_HAS_PHY_WAKEUP) {
5534 5535 5536 5537 5538 5539 5540 5541 5542
			/* enable wakeup by the PHY */
			retval = e1000_init_phy_wakeup(adapter, wufc);
			if (retval)
				return retval;
		} else {
			/* enable wakeup by the MAC */
			ew32(WUFC, wufc);
			ew32(WUC, E1000_WUC_PME_EN);
		}
5543 5544 5545 5546 5547
	} else {
		ew32(WUC, 0);
		ew32(WUFC, 0);
	}

5548 5549
	*enable_wake = !!wufc;

5550
	/* make sure adapter isn't asleep if manageability is enabled */
5551 5552
	if ((adapter->flags & FLAG_MNG_PT_ENABLED) ||
	    (hw->mac.ops.check_mng_mode(hw)))
5553
		*enable_wake = true;
5554 5555 5556 5557

	if (adapter->hw.phy.type == e1000_phy_igp_3)
		e1000e_igp3_phy_powerdown_workaround_ich8lan(&adapter->hw);

5558 5559 5560 5561
	/*
	 * Release control of h/w to f/w.  If f/w is AMT enabled, this
	 * would have already happened in close and is redundant.
	 */
5562
	e1000e_release_hw_control(adapter);
5563 5564 5565

	pci_disable_device(pdev);

5566 5567 5568 5569 5570 5571 5572 5573 5574 5575 5576 5577 5578 5579 5580 5581 5582 5583 5584 5585
	return 0;
}

static void e1000_power_off(struct pci_dev *pdev, bool sleep, bool wake)
{
	if (sleep && wake) {
		pci_prepare_to_sleep(pdev);
		return;
	}

	pci_wake_from_d3(pdev, wake);
	pci_set_power_state(pdev, PCI_D3hot);
}

static void e1000_complete_shutdown(struct pci_dev *pdev, bool sleep,
                                    bool wake)
{
	struct net_device *netdev = pci_get_drvdata(pdev);
	struct e1000_adapter *adapter = netdev_priv(netdev);

5586 5587 5588 5589 5590 5591 5592 5593
	/*
	 * The pci-e switch on some quad port adapters will report a
	 * correctable error when the MAC transitions from D0 to D3.  To
	 * prevent this we need to mask off the correctable errors on the
	 * downstream port of the pci-e switch.
	 */
	if (adapter->flags & FLAG_IS_QUAD_PORT) {
		struct pci_dev *us_dev = pdev->bus->self;
5594
		int pos = pci_pcie_cap(us_dev);
5595 5596 5597 5598 5599 5600
		u16 devctl;

		pci_read_config_word(us_dev, pos + PCI_EXP_DEVCTL, &devctl);
		pci_write_config_word(us_dev, pos + PCI_EXP_DEVCTL,
		                      (devctl & ~PCI_EXP_DEVCTL_CERE));

5601
		e1000_power_off(pdev, sleep, wake);
5602 5603 5604

		pci_write_config_word(us_dev, pos + PCI_EXP_DEVCTL, devctl);
	} else {
5605
		e1000_power_off(pdev, sleep, wake);
5606
	}
5607 5608
}

5609 5610 5611
#ifdef CONFIG_PCIEASPM
static void __e1000e_disable_aspm(struct pci_dev *pdev, u16 state)
{
5612
	pci_disable_link_state_locked(pdev, state);
5613 5614 5615
}
#else
static void __e1000e_disable_aspm(struct pci_dev *pdev, u16 state)
5616 5617
{
	int pos;
5618
	u16 reg16;
5619 5620

	/*
5621 5622
	 * Both device and parent should have the same ASPM setting.
	 * Disable ASPM in downstream component first and then upstream.
5623
	 */
5624 5625 5626 5627 5628
	pos = pci_pcie_cap(pdev);
	pci_read_config_word(pdev, pos + PCI_EXP_LNKCTL, &reg16);
	reg16 &= ~state;
	pci_write_config_word(pdev, pos + PCI_EXP_LNKCTL, reg16);

5629 5630 5631
	if (!pdev->bus->self)
		return;

5632 5633 5634 5635 5636 5637
	pos = pci_pcie_cap(pdev->bus->self);
	pci_read_config_word(pdev->bus->self, pos + PCI_EXP_LNKCTL, &reg16);
	reg16 &= ~state;
	pci_write_config_word(pdev->bus->self, pos + PCI_EXP_LNKCTL, reg16);
}
#endif
5638
static void e1000e_disable_aspm(struct pci_dev *pdev, u16 state)
5639 5640 5641 5642 5643 5644
{
	dev_info(&pdev->dev, "Disabling ASPM %s %s\n",
		 (state & PCIE_LINK_STATE_L0S) ? "L0s" : "",
		 (state & PCIE_LINK_STATE_L1) ? "L1" : "");

	__e1000e_disable_aspm(pdev, state);
5645 5646
}

R
Rafael J. Wysocki 已提交
5647
#ifdef CONFIG_PM
5648
static bool e1000e_pm_ready(struct e1000_adapter *adapter)
5649
{
5650
	return !!adapter->tx_ring->buffer_info;
5651 5652
}

5653
static int __e1000_resume(struct pci_dev *pdev)
5654 5655 5656 5657
{
	struct net_device *netdev = pci_get_drvdata(pdev);
	struct e1000_adapter *adapter = netdev_priv(netdev);
	struct e1000_hw *hw = &adapter->hw;
5658
	u16 aspm_disable_flag = 0;
5659 5660
	u32 err;

5661 5662 5663 5664 5665 5666 5667
	if (adapter->flags2 & FLAG2_DISABLE_ASPM_L0S)
		aspm_disable_flag = PCIE_LINK_STATE_L0S;
	if (adapter->flags2 & FLAG2_DISABLE_ASPM_L1)
		aspm_disable_flag |= PCIE_LINK_STATE_L1;
	if (aspm_disable_flag)
		e1000e_disable_aspm(pdev, aspm_disable_flag);

5668 5669
	pci_set_power_state(pdev, PCI_D0);
	pci_restore_state(pdev);
5670
	pci_save_state(pdev);
T
Taku Izumi 已提交
5671

5672
	e1000e_set_interrupt_capability(adapter);
5673 5674 5675 5676 5677 5678
	if (netif_running(netdev)) {
		err = e1000_request_irq(adapter);
		if (err)
			return err;
	}

5679 5680 5681
	if (hw->mac.type == e1000_pch2lan)
		e1000_resume_workarounds_pchlan(&adapter->hw);

5682
	e1000e_power_up_phy(adapter);
5683 5684 5685 5686 5687 5688 5689 5690 5691 5692 5693 5694

	/* report the system wakeup cause from S3/S4 */
	if (adapter->flags2 & FLAG2_HAS_PHY_WAKEUP) {
		u16 phy_data;

		e1e_rphy(&adapter->hw, BM_WUS, &phy_data);
		if (phy_data) {
			e_info("PHY Wakeup cause - %s\n",
				phy_data & E1000_WUS_EX ? "Unicast Packet" :
				phy_data & E1000_WUS_MC ? "Multicast Packet" :
				phy_data & E1000_WUS_BC ? "Broadcast Packet" :
				phy_data & E1000_WUS_MAG ? "Magic Packet" :
5695 5696
				phy_data & E1000_WUS_LNKC ?
				"Link Status Change" : "other");
5697 5698 5699 5700 5701 5702 5703 5704 5705 5706 5707 5708 5709 5710 5711 5712
		}
		e1e_wphy(&adapter->hw, BM_WUS, ~0);
	} else {
		u32 wus = er32(WUS);
		if (wus) {
			e_info("MAC Wakeup cause - %s\n",
				wus & E1000_WUS_EX ? "Unicast Packet" :
				wus & E1000_WUS_MC ? "Multicast Packet" :
				wus & E1000_WUS_BC ? "Broadcast Packet" :
				wus & E1000_WUS_MAG ? "Magic Packet" :
				wus & E1000_WUS_LNKC ? "Link Status Change" :
				"other");
		}
		ew32(WUS, ~0);
	}

5713 5714
	e1000e_reset(adapter);

5715
	e1000_init_manageability_pt(adapter);
5716 5717 5718 5719 5720 5721

	if (netif_running(netdev))
		e1000e_up(adapter);

	netif_device_attach(netdev);

5722 5723
	/*
	 * If the controller has AMT, do not set DRV_LOAD until the interface
5724
	 * is up.  For all other cases, let the f/w know that the h/w is now
5725 5726
	 * under the control of the driver.
	 */
J
Jesse Brandeburg 已提交
5727
	if (!(adapter->flags & FLAG_HAS_AMT))
5728
		e1000e_get_hw_control(adapter);
5729 5730 5731

	return 0;
}
5732

5733 5734 5735 5736 5737 5738 5739 5740 5741 5742 5743 5744 5745 5746
#ifdef CONFIG_PM_SLEEP
static int e1000_suspend(struct device *dev)
{
	struct pci_dev *pdev = to_pci_dev(dev);
	int retval;
	bool wake;

	retval = __e1000_shutdown(pdev, &wake, false);
	if (!retval)
		e1000_complete_shutdown(pdev, true, wake);

	return retval;
}

5747 5748 5749 5750 5751 5752 5753 5754 5755 5756 5757
static int e1000_resume(struct device *dev)
{
	struct pci_dev *pdev = to_pci_dev(dev);
	struct net_device *netdev = pci_get_drvdata(pdev);
	struct e1000_adapter *adapter = netdev_priv(netdev);

	if (e1000e_pm_ready(adapter))
		adapter->idle_check = true;

	return __e1000_resume(pdev);
}
5758 5759 5760 5761 5762 5763 5764 5765 5766 5767 5768 5769 5770 5771 5772 5773 5774 5775 5776 5777 5778 5779 5780 5781 5782 5783 5784 5785 5786 5787 5788 5789 5790 5791 5792
#endif /* CONFIG_PM_SLEEP */

#ifdef CONFIG_PM_RUNTIME
static int e1000_runtime_suspend(struct device *dev)
{
	struct pci_dev *pdev = to_pci_dev(dev);
	struct net_device *netdev = pci_get_drvdata(pdev);
	struct e1000_adapter *adapter = netdev_priv(netdev);

	if (e1000e_pm_ready(adapter)) {
		bool wake;

		__e1000_shutdown(pdev, &wake, true);
	}

	return 0;
}

static int e1000_idle(struct device *dev)
{
	struct pci_dev *pdev = to_pci_dev(dev);
	struct net_device *netdev = pci_get_drvdata(pdev);
	struct e1000_adapter *adapter = netdev_priv(netdev);

	if (!e1000e_pm_ready(adapter))
		return 0;

	if (adapter->idle_check) {
		adapter->idle_check = false;
		if (!e1000e_has_link(adapter))
			pm_schedule_suspend(dev, MSEC_PER_SEC);
	}

	return -EBUSY;
}
5793 5794 5795 5796 5797 5798 5799 5800 5801 5802 5803 5804 5805

static int e1000_runtime_resume(struct device *dev)
{
	struct pci_dev *pdev = to_pci_dev(dev);
	struct net_device *netdev = pci_get_drvdata(pdev);
	struct e1000_adapter *adapter = netdev_priv(netdev);

	if (!e1000e_pm_ready(adapter))
		return 0;

	adapter->idle_check = !dev->power.runtime_auto;
	return __e1000_resume(pdev);
}
5806
#endif /* CONFIG_PM_RUNTIME */
R
Rafael J. Wysocki 已提交
5807
#endif /* CONFIG_PM */
5808 5809 5810

static void e1000_shutdown(struct pci_dev *pdev)
{
5811 5812
	bool wake = false;

5813
	__e1000_shutdown(pdev, &wake, false);
5814 5815 5816

	if (system_state == SYSTEM_POWER_OFF)
		e1000_complete_shutdown(pdev, false, wake);
5817 5818 5819
}

#ifdef CONFIG_NET_POLL_CONTROLLER
5820 5821 5822 5823 5824 5825 5826

static irqreturn_t e1000_intr_msix(int irq, void *data)
{
	struct net_device *netdev = data;
	struct e1000_adapter *adapter = netdev_priv(netdev);

	if (adapter->msix_entries) {
5827 5828
		int vector, msix_irq;

5829 5830 5831 5832 5833 5834 5835 5836 5837 5838 5839 5840 5841 5842 5843 5844 5845 5846 5847 5848 5849 5850
		vector = 0;
		msix_irq = adapter->msix_entries[vector].vector;
		disable_irq(msix_irq);
		e1000_intr_msix_rx(msix_irq, netdev);
		enable_irq(msix_irq);

		vector++;
		msix_irq = adapter->msix_entries[vector].vector;
		disable_irq(msix_irq);
		e1000_intr_msix_tx(msix_irq, netdev);
		enable_irq(msix_irq);

		vector++;
		msix_irq = adapter->msix_entries[vector].vector;
		disable_irq(msix_irq);
		e1000_msix_other(msix_irq, netdev);
		enable_irq(msix_irq);
	}

	return IRQ_HANDLED;
}

5851 5852 5853 5854 5855 5856 5857 5858 5859
/*
 * 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 e1000_netpoll(struct net_device *netdev)
{
	struct e1000_adapter *adapter = netdev_priv(netdev);

5860 5861 5862 5863 5864 5865 5866 5867 5868 5869 5870 5871 5872 5873 5874
	switch (adapter->int_mode) {
	case E1000E_INT_MODE_MSIX:
		e1000_intr_msix(adapter->pdev->irq, netdev);
		break;
	case E1000E_INT_MODE_MSI:
		disable_irq(adapter->pdev->irq);
		e1000_intr_msi(adapter->pdev->irq, netdev);
		enable_irq(adapter->pdev->irq);
		break;
	default: /* E1000E_INT_MODE_LEGACY */
		disable_irq(adapter->pdev->irq);
		e1000_intr(adapter->pdev->irq, netdev);
		enable_irq(adapter->pdev->irq);
		break;
	}
5875 5876 5877 5878 5879 5880 5881 5882 5883 5884 5885 5886 5887 5888 5889 5890 5891 5892 5893
}
#endif

/**
 * e1000_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 e1000_io_error_detected(struct pci_dev *pdev,
						pci_channel_state_t state)
{
	struct net_device *netdev = pci_get_drvdata(pdev);
	struct e1000_adapter *adapter = netdev_priv(netdev);

	netif_device_detach(netdev);

5894 5895 5896
	if (state == pci_channel_io_perm_failure)
		return PCI_ERS_RESULT_DISCONNECT;

5897 5898 5899 5900 5901 5902 5903 5904 5905 5906 5907 5908 5909 5910 5911 5912 5913 5914 5915 5916
	if (netif_running(netdev))
		e1000e_down(adapter);
	pci_disable_device(pdev);

	/* Request a slot slot reset. */
	return PCI_ERS_RESULT_NEED_RESET;
}

/**
 * e1000_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. Implementation
 * resembles the first-half of the e1000_resume routine.
 */
static pci_ers_result_t e1000_io_slot_reset(struct pci_dev *pdev)
{
	struct net_device *netdev = pci_get_drvdata(pdev);
	struct e1000_adapter *adapter = netdev_priv(netdev);
	struct e1000_hw *hw = &adapter->hw;
5917
	u16 aspm_disable_flag = 0;
T
Taku Izumi 已提交
5918
	int err;
J
Jesse Brandeburg 已提交
5919
	pci_ers_result_t result;
5920

5921 5922
	if (adapter->flags2 & FLAG2_DISABLE_ASPM_L0S)
		aspm_disable_flag = PCIE_LINK_STATE_L0S;
5923
	if (adapter->flags2 & FLAG2_DISABLE_ASPM_L1)
5924 5925 5926 5927
		aspm_disable_flag |= PCIE_LINK_STATE_L1;
	if (aspm_disable_flag)
		e1000e_disable_aspm(pdev, aspm_disable_flag);

5928
	err = pci_enable_device_mem(pdev);
T
Taku Izumi 已提交
5929
	if (err) {
5930 5931
		dev_err(&pdev->dev,
			"Cannot re-enable PCI device after reset.\n");
J
Jesse Brandeburg 已提交
5932 5933 5934
		result = PCI_ERS_RESULT_DISCONNECT;
	} else {
		pci_set_master(pdev);
5935
		pdev->state_saved = true;
J
Jesse Brandeburg 已提交
5936
		pci_restore_state(pdev);
5937

J
Jesse Brandeburg 已提交
5938 5939
		pci_enable_wake(pdev, PCI_D3hot, 0);
		pci_enable_wake(pdev, PCI_D3cold, 0);
5940

J
Jesse Brandeburg 已提交
5941 5942 5943 5944
		e1000e_reset(adapter);
		ew32(WUS, ~0);
		result = PCI_ERS_RESULT_RECOVERED;
	}
5945

J
Jesse Brandeburg 已提交
5946 5947 5948
	pci_cleanup_aer_uncorrect_error_status(pdev);

	return result;
5949 5950 5951 5952 5953 5954 5955 5956 5957 5958 5959 5960 5961 5962 5963
}

/**
 * e1000_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. Implementation resembles the
 * second-half of the e1000_resume routine.
 */
static void e1000_io_resume(struct pci_dev *pdev)
{
	struct net_device *netdev = pci_get_drvdata(pdev);
	struct e1000_adapter *adapter = netdev_priv(netdev);

5964
	e1000_init_manageability_pt(adapter);
5965 5966 5967 5968 5969 5970 5971 5972 5973 5974 5975

	if (netif_running(netdev)) {
		if (e1000e_up(adapter)) {
			dev_err(&pdev->dev,
				"can't bring device back up after reset\n");
			return;
		}
	}

	netif_device_attach(netdev);

5976 5977
	/*
	 * If the controller has AMT, do not set DRV_LOAD until the interface
5978
	 * is up.  For all other cases, let the f/w know that the h/w is now
5979 5980
	 * under the control of the driver.
	 */
J
Jesse Brandeburg 已提交
5981
	if (!(adapter->flags & FLAG_HAS_AMT))
5982
		e1000e_get_hw_control(adapter);
5983 5984 5985 5986 5987 5988 5989

}

static void e1000_print_device_info(struct e1000_adapter *adapter)
{
	struct e1000_hw *hw = &adapter->hw;
	struct net_device *netdev = adapter->netdev;
5990 5991
	u32 ret_val;
	u8 pba_str[E1000_PBANUM_LENGTH];
5992 5993

	/* print bus type/speed/width info */
5994
	e_info("(PCI Express:2.5GT/s:%s) %pM\n",
5995 5996 5997 5998
	       /* bus width */
	       ((hw->bus.width == e1000_bus_width_pcie_x4) ? "Width x4" :
	        "Width x1"),
	       /* MAC address */
J
Johannes Berg 已提交
5999
	       netdev->dev_addr);
6000 6001
	e_info("Intel(R) PRO/%s Network Connection\n",
	       (hw->phy.type == e1000_phy_ife) ? "10/100" : "1000");
6002 6003 6004
	ret_val = e1000_read_pba_string_generic(hw, pba_str,
						E1000_PBANUM_LENGTH);
	if (ret_val)
6005
		strlcpy((char *)pba_str, "Unknown", sizeof(pba_str));
6006 6007
	e_info("MAC: %d, PHY: %d, PBA No: %s\n",
	       hw->mac.type, hw->phy.type, pba_str);
6008 6009
}

6010 6011 6012 6013 6014 6015 6016 6017 6018 6019
static void e1000_eeprom_checks(struct e1000_adapter *adapter)
{
	struct e1000_hw *hw = &adapter->hw;
	int ret_val;
	u16 buf = 0;

	if (hw->mac.type != e1000_82573)
		return;

	ret_val = e1000_read_nvm(hw, NVM_INIT_CONTROL2_REG, 1, &buf);
6020 6021
	le16_to_cpus(&buf);
	if (!ret_val && (!(buf & (1 << 0)))) {
6022
		/* Deep Smart Power Down (DSPD) */
6023 6024
		dev_warn(&adapter->pdev->dev,
			 "Warning: detected DSPD enabled in EEPROM\n");
6025 6026 6027
	}
}

6028
static int e1000_set_features(struct net_device *netdev,
6029
			      netdev_features_t features)
6030 6031
{
	struct e1000_adapter *adapter = netdev_priv(netdev);
6032
	netdev_features_t changed = features ^ netdev->features;
6033 6034 6035 6036 6037

	if (changed & (NETIF_F_TSO | NETIF_F_TSO6))
		adapter->flags |= FLAG_TSO_FORCE;

	if (!(changed & (NETIF_F_HW_VLAN_RX | NETIF_F_HW_VLAN_TX |
B
Ben Greear 已提交
6038 6039
			 NETIF_F_RXCSUM | NETIF_F_RXHASH | NETIF_F_RXFCS |
			 NETIF_F_RXALL)))
6040 6041
		return 0;

6042 6043 6044 6045 6046 6047 6048 6049 6050 6051 6052
	/*
	 * IP payload checksum (enabled with jumbos/packet-split when Rx
	 * checksum is enabled) and generation of RSS hash is mutually
	 * exclusive in the hardware.
	 */
	if (adapter->rx_ps_pages &&
	    (features & NETIF_F_RXCSUM) && (features & NETIF_F_RXHASH)) {
		e_err("Enabling both receive checksum offload and receive hashing is not possible with jumbo frames.  Disable jumbos or enable only one of the receive offload features.\n");
		return -EINVAL;
	}

B
Ben Greear 已提交
6053 6054 6055 6056 6057 6058 6059 6060 6061 6062 6063 6064 6065 6066
	if (changed & NETIF_F_RXFCS) {
		if (features & NETIF_F_RXFCS) {
			adapter->flags2 &= ~FLAG2_CRC_STRIPPING;
		} else {
			/* We need to take it back to defaults, which might mean
			 * stripping is still disabled at the adapter level.
			 */
			if (adapter->flags2 & FLAG2_DFLT_CRC_STRIPPING)
				adapter->flags2 |= FLAG2_CRC_STRIPPING;
			else
				adapter->flags2 &= ~FLAG2_CRC_STRIPPING;
		}
	}

6067 6068
	netdev->features = features;

6069 6070 6071 6072 6073 6074 6075 6076
	if (netif_running(netdev))
		e1000e_reinit_locked(adapter);
	else
		e1000e_reset(adapter);

	return 0;
}

6077 6078 6079
static const struct net_device_ops e1000e_netdev_ops = {
	.ndo_open		= e1000_open,
	.ndo_stop		= e1000_close,
6080
	.ndo_start_xmit		= e1000_xmit_frame,
J
Jeff Kirsher 已提交
6081
	.ndo_get_stats64	= e1000e_get_stats64,
6082
	.ndo_set_rx_mode	= e1000e_set_rx_mode,
6083 6084 6085 6086 6087 6088 6089 6090 6091 6092 6093
	.ndo_set_mac_address	= e1000_set_mac,
	.ndo_change_mtu		= e1000_change_mtu,
	.ndo_do_ioctl		= e1000_ioctl,
	.ndo_tx_timeout		= e1000_tx_timeout,
	.ndo_validate_addr	= eth_validate_addr,

	.ndo_vlan_rx_add_vid	= e1000_vlan_rx_add_vid,
	.ndo_vlan_rx_kill_vid	= e1000_vlan_rx_kill_vid,
#ifdef CONFIG_NET_POLL_CONTROLLER
	.ndo_poll_controller	= e1000_netpoll,
#endif
6094
	.ndo_set_features = e1000_set_features,
6095 6096
};

6097 6098 6099 6100 6101 6102 6103 6104 6105 6106 6107 6108 6109 6110 6111 6112 6113 6114
/**
 * e1000_probe - Device Initialization Routine
 * @pdev: PCI device information struct
 * @ent: entry in e1000_pci_tbl
 *
 * Returns 0 on success, negative on failure
 *
 * e1000_probe initializes an adapter identified by a pci_dev structure.
 * The OS initialization, configuring of the adapter private structure,
 * and a hardware reset occur.
 **/
static int __devinit e1000_probe(struct pci_dev *pdev,
				 const struct pci_device_id *ent)
{
	struct net_device *netdev;
	struct e1000_adapter *adapter;
	struct e1000_hw *hw;
	const struct e1000_info *ei = e1000_info_tbl[ent->driver_data];
6115 6116
	resource_size_t mmio_start, mmio_len;
	resource_size_t flash_start, flash_len;
6117
	static int cards_found;
6118
	u16 aspm_disable_flag = 0;
6119 6120 6121 6122
	int i, err, pci_using_dac;
	u16 eeprom_data = 0;
	u16 eeprom_apme_mask = E1000_EEPROM_APME;

6123 6124
	if (ei->flags2 & FLAG2_DISABLE_ASPM_L0S)
		aspm_disable_flag = PCIE_LINK_STATE_L0S;
6125
	if (ei->flags2 & FLAG2_DISABLE_ASPM_L1)
6126 6127 6128
		aspm_disable_flag |= PCIE_LINK_STATE_L1;
	if (aspm_disable_flag)
		e1000e_disable_aspm(pdev, aspm_disable_flag);
T
Taku Izumi 已提交
6129

6130
	err = pci_enable_device_mem(pdev);
6131 6132 6133 6134
	if (err)
		return err;

	pci_using_dac = 0;
6135
	err = dma_set_mask(&pdev->dev, DMA_BIT_MASK(64));
6136
	if (!err) {
6137
		err = dma_set_coherent_mask(&pdev->dev, DMA_BIT_MASK(64));
6138 6139 6140
		if (!err)
			pci_using_dac = 1;
	} else {
6141
		err = dma_set_mask(&pdev->dev, DMA_BIT_MASK(32));
6142
		if (err) {
6143 6144
			err = dma_set_coherent_mask(&pdev->dev,
						    DMA_BIT_MASK(32));
6145
			if (err) {
6146
				dev_err(&pdev->dev, "No usable DMA configuration, aborting\n");
6147 6148 6149 6150 6151
				goto err_dma;
			}
		}
	}

6152
	err = pci_request_selected_regions_exclusive(pdev,
6153 6154
	                                  pci_select_bars(pdev, IORESOURCE_MEM),
	                                  e1000e_driver_name);
6155 6156 6157
	if (err)
		goto err_pci_reg;

6158
	/* AER (Advanced Error Reporting) hooks */
6159
	pci_enable_pcie_error_reporting(pdev);
6160

6161
	pci_set_master(pdev);
6162 6163 6164 6165
	/* PCI config space info */
	err = pci_save_state(pdev);
	if (err)
		goto err_alloc_etherdev;
6166 6167 6168 6169 6170 6171 6172 6173

	err = -ENOMEM;
	netdev = alloc_etherdev(sizeof(struct e1000_adapter));
	if (!netdev)
		goto err_alloc_etherdev;

	SET_NETDEV_DEV(netdev, &pdev->dev);

6174 6175
	netdev->irq = pdev->irq;

6176 6177 6178 6179 6180 6181 6182 6183
	pci_set_drvdata(pdev, netdev);
	adapter = netdev_priv(netdev);
	hw = &adapter->hw;
	adapter->netdev = netdev;
	adapter->pdev = pdev;
	adapter->ei = ei;
	adapter->pba = ei->pba;
	adapter->flags = ei->flags;
J
Jeff Kirsher 已提交
6184
	adapter->flags2 = ei->flags2;
6185 6186
	adapter->hw.adapter = adapter;
	adapter->hw.mac.type = ei->mac;
6187
	adapter->max_hw_frame_size = ei->max_hw_frame_size;
6188
	adapter->msg_enable = netif_msg_init(debug, DEFAULT_MSG_ENABLE);
6189 6190 6191 6192 6193 6194 6195 6196 6197 6198 6199 6200 6201 6202 6203 6204 6205 6206 6207

	mmio_start = pci_resource_start(pdev, 0);
	mmio_len = pci_resource_len(pdev, 0);

	err = -EIO;
	adapter->hw.hw_addr = ioremap(mmio_start, mmio_len);
	if (!adapter->hw.hw_addr)
		goto err_ioremap;

	if ((adapter->flags & FLAG_HAS_FLASH) &&
	    (pci_resource_flags(pdev, 1) & IORESOURCE_MEM)) {
		flash_start = pci_resource_start(pdev, 1);
		flash_len = pci_resource_len(pdev, 1);
		adapter->hw.flash_address = ioremap(flash_start, flash_len);
		if (!adapter->hw.flash_address)
			goto err_flashmap;
	}

	/* construct the net_device struct */
6208
	netdev->netdev_ops		= &e1000e_netdev_ops;
6209 6210
	e1000e_set_ethtool_ops(netdev);
	netdev->watchdog_timeo		= 5 * HZ;
B
Bruce Allan 已提交
6211
	netif_napi_add(netdev, &adapter->napi, e1000e_poll, 64);
6212
	strlcpy(netdev->name, pci_name(pdev), sizeof(netdev->name));
6213 6214 6215 6216 6217 6218

	netdev->mem_start = mmio_start;
	netdev->mem_end = mmio_start + mmio_len;

	adapter->bd_number = cards_found++;

6219 6220
	e1000e_check_options(adapter);

6221 6222 6223 6224 6225 6226 6227 6228 6229
	/* setup adapter struct */
	err = e1000_sw_init(adapter);
	if (err)
		goto err_sw_init;

	memcpy(&hw->mac.ops, ei->mac_ops, sizeof(hw->mac.ops));
	memcpy(&hw->nvm.ops, ei->nvm_ops, sizeof(hw->nvm.ops));
	memcpy(&hw->phy.ops, ei->phy_ops, sizeof(hw->phy.ops));

J
Jeff Kirsher 已提交
6230
	err = ei->get_variants(adapter);
6231 6232 6233
	if (err)
		goto err_hw_init;

6234 6235 6236 6237
	if ((adapter->flags & FLAG_IS_ICH) &&
	    (adapter->flags & FLAG_READ_ONLY_NVM))
		e1000e_write_protect_nvm_ich8lan(&adapter->hw);

6238 6239
	hw->mac.ops.get_bus_info(&adapter->hw);

6240
	adapter->hw.phy.autoneg_wait_to_complete = 0;
6241 6242

	/* Copper options */
6243
	if (adapter->hw.phy.media_type == e1000_media_type_copper) {
6244 6245 6246 6247 6248
		adapter->hw.phy.mdix = AUTO_ALL_MODES;
		adapter->hw.phy.disable_polarity_correction = 0;
		adapter->hw.phy.ms_type = e1000_ms_hw_default;
	}

6249
	if (hw->phy.ops.check_reset_block(hw))
6250
		e_info("PHY reset is blocked due to SOL/IDER session.\n");
6251

6252 6253 6254 6255 6256 6257
	/* Set initial default active device features */
	netdev->features = (NETIF_F_SG |
			    NETIF_F_HW_VLAN_RX |
			    NETIF_F_HW_VLAN_TX |
			    NETIF_F_TSO |
			    NETIF_F_TSO6 |
6258
			    NETIF_F_RXHASH |
6259 6260 6261 6262 6263
			    NETIF_F_RXCSUM |
			    NETIF_F_HW_CSUM);

	/* Set user-changeable features (subset of all device features) */
	netdev->hw_features = netdev->features;
B
Ben Greear 已提交
6264
	netdev->hw_features |= NETIF_F_RXFCS;
6265
	netdev->priv_flags |= IFF_SUPP_NOFCS;
B
Ben Greear 已提交
6266
	netdev->hw_features |= NETIF_F_RXALL;
6267 6268 6269 6270

	if (adapter->flags & FLAG_HAS_HW_VLAN_FILTER)
		netdev->features |= NETIF_F_HW_VLAN_FILTER;

6271 6272 6273 6274
	netdev->vlan_features |= (NETIF_F_SG |
				  NETIF_F_TSO |
				  NETIF_F_TSO6 |
				  NETIF_F_HW_CSUM);
6275

6276 6277
	netdev->priv_flags |= IFF_UNICAST_FLT;

6278
	if (pci_using_dac) {
6279
		netdev->features |= NETIF_F_HIGHDMA;
6280 6281
		netdev->vlan_features |= NETIF_F_HIGHDMA;
	}
6282 6283 6284 6285

	if (e1000e_enable_mng_pass_thru(&adapter->hw))
		adapter->flags |= FLAG_MNG_PT_ENABLED;

6286 6287 6288 6289
	/*
	 * before reading the NVM, reset the controller to
	 * put the device in a known good starting state
	 */
6290 6291 6292 6293 6294 6295 6296 6297 6298 6299
	adapter->hw.mac.ops.reset_hw(&adapter->hw);

	/*
	 * systems with ASPM and others may see the checksum fail on the first
	 * attempt. Let's give it a few tries
	 */
	for (i = 0;; i++) {
		if (e1000_validate_nvm_checksum(&adapter->hw) >= 0)
			break;
		if (i == 2) {
6300
			e_err("The NVM Checksum Is Not Valid\n");
6301 6302 6303 6304 6305
			err = -EIO;
			goto err_eeprom;
		}
	}

6306 6307
	e1000_eeprom_checks(adapter);

6308
	/* copy the MAC address */
6309
	if (e1000e_read_mac_addr(&adapter->hw))
6310
		e_err("NVM Read Error while reading MAC address\n");
6311 6312 6313 6314 6315

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

	if (!is_valid_ether_addr(netdev->perm_addr)) {
J
Johannes Berg 已提交
6316
		e_err("Invalid MAC Address: %pM\n", netdev->perm_addr);
6317 6318 6319 6320 6321
		err = -EIO;
		goto err_eeprom;
	}

	init_timer(&adapter->watchdog_timer);
6322
	adapter->watchdog_timer.function = e1000_watchdog;
6323 6324 6325
	adapter->watchdog_timer.data = (unsigned long) adapter;

	init_timer(&adapter->phy_info_timer);
6326
	adapter->phy_info_timer.function = e1000_update_phy_info;
6327 6328 6329 6330
	adapter->phy_info_timer.data = (unsigned long) adapter;

	INIT_WORK(&adapter->reset_task, e1000_reset_task);
	INIT_WORK(&adapter->watchdog_task, e1000_watchdog_task);
6331 6332
	INIT_WORK(&adapter->downshift_task, e1000e_downshift_workaround);
	INIT_WORK(&adapter->update_phy_task, e1000e_update_phy_task);
6333
	INIT_WORK(&adapter->print_hang_task, e1000_print_hw_hang);
6334 6335 6336

	/* Initialize link parameters. User can change them with ethtool */
	adapter->hw.mac.autoneg = 1;
6337
	adapter->fc_autoneg = true;
6338 6339
	adapter->hw.fc.requested_mode = e1000_fc_default;
	adapter->hw.fc.current_mode = e1000_fc_default;
6340 6341 6342 6343 6344 6345 6346 6347 6348 6349 6350 6351 6352 6353
	adapter->hw.phy.autoneg_advertised = 0x2f;

	/* ring size defaults */
	adapter->rx_ring->count = 256;
	adapter->tx_ring->count = 256;

	/*
	 * Initial Wake on LAN setting - If APM wake is enabled in
	 * the EEPROM, enable the ACPI Magic Packet filter
	 */
	if (adapter->flags & FLAG_APME_IN_WUC) {
		/* APME bit in EEPROM is mapped to WUC.APME */
		eeprom_data = er32(WUC);
		eeprom_apme_mask = E1000_WUC_APME;
6354 6355
		if ((hw->mac.type > e1000_ich10lan) &&
		    (eeprom_data & E1000_WUC_PHY_WAKE))
6356
			adapter->flags2 |= FLAG2_HAS_PHY_WAKEUP;
6357 6358 6359
	} else if (adapter->flags & FLAG_APME_IN_CTRL3) {
		if (adapter->flags & FLAG_APME_CHECK_PORT_B &&
		    (adapter->hw.bus.func == 1))
6360 6361
			e1000_read_nvm(&adapter->hw, NVM_INIT_CONTROL3_PORT_B,
				       1, &eeprom_data);
6362
		else
6363 6364
			e1000_read_nvm(&adapter->hw, NVM_INIT_CONTROL3_PORT_A,
				       1, &eeprom_data);
6365 6366 6367 6368 6369 6370 6371 6372 6373 6374 6375 6376 6377 6378 6379 6380
	}

	/* fetch WoL from EEPROM */
	if (eeprom_data & eeprom_apme_mask)
		adapter->eeprom_wol |= E1000_WUFC_MAG;

	/*
	 * now that we have the eeprom settings, apply the special cases
	 * where the eeprom may be wrong or the board simply won't support
	 * wake on lan on a particular port
	 */
	if (!(adapter->flags & FLAG_HAS_WOL))
		adapter->eeprom_wol = 0;

	/* initialize the wol settings based on the eeprom settings */
	adapter->wol = adapter->eeprom_wol;
6381
	device_set_wakeup_enable(&adapter->pdev->dev, adapter->wol);
6382

6383 6384 6385
	/* save off EEPROM version number */
	e1000_read_nvm(&adapter->hw, 5, 1, &adapter->eeprom_vers);

6386 6387 6388
	/* reset the hardware with the new settings */
	e1000e_reset(adapter);

6389 6390
	/*
	 * If the controller has AMT, do not set DRV_LOAD until the interface
6391
	 * is up.  For all other cases, let the f/w know that the h/w is now
6392 6393
	 * under the control of the driver.
	 */
J
Jesse Brandeburg 已提交
6394
	if (!(adapter->flags & FLAG_HAS_AMT))
6395
		e1000e_get_hw_control(adapter);
6396

6397
	strlcpy(netdev->name, "eth%d", sizeof(netdev->name));
6398 6399 6400 6401
	err = register_netdev(netdev);
	if (err)
		goto err_register;

6402 6403 6404
	/* carrier off reporting is important to ethtool even BEFORE open */
	netif_carrier_off(netdev);

6405 6406
	e1000_print_device_info(adapter);

6407 6408
	if (pci_dev_run_wake(pdev))
		pm_runtime_put_noidle(&pdev->dev);
6409

6410 6411 6412
	return 0;

err_register:
J
Jesse Brandeburg 已提交
6413
	if (!(adapter->flags & FLAG_HAS_AMT))
6414
		e1000e_release_hw_control(adapter);
6415
err_eeprom:
6416
	if (!hw->phy.ops.check_reset_block(hw))
6417
		e1000_phy_hw_reset(&adapter->hw);
J
Jesse Brandeburg 已提交
6418
err_hw_init:
6419 6420 6421
	kfree(adapter->tx_ring);
	kfree(adapter->rx_ring);
err_sw_init:
J
Jesse Brandeburg 已提交
6422 6423
	if (adapter->hw.flash_address)
		iounmap(adapter->hw.flash_address);
6424
	e1000e_reset_interrupt_capability(adapter);
J
Jesse Brandeburg 已提交
6425
err_flashmap:
6426 6427 6428 6429
	iounmap(adapter->hw.hw_addr);
err_ioremap:
	free_netdev(netdev);
err_alloc_etherdev:
6430 6431
	pci_release_selected_regions(pdev,
	                             pci_select_bars(pdev, IORESOURCE_MEM));
6432 6433 6434 6435 6436 6437 6438 6439 6440 6441 6442 6443 6444 6445 6446 6447 6448 6449 6450
err_pci_reg:
err_dma:
	pci_disable_device(pdev);
	return err;
}

/**
 * e1000_remove - Device Removal Routine
 * @pdev: PCI device information struct
 *
 * e1000_remove is called by the PCI subsystem to alert the driver
 * that it should release a PCI device.  The could be caused by a
 * Hot-Plug event, or because the driver is going to be removed from
 * memory.
 **/
static void __devexit e1000_remove(struct pci_dev *pdev)
{
	struct net_device *netdev = pci_get_drvdata(pdev);
	struct e1000_adapter *adapter = netdev_priv(netdev);
6451 6452
	bool down = test_bit(__E1000_DOWN, &adapter->state);

6453
	/*
6454 6455
	 * The timers may be rescheduled, so explicitly disable them
	 * from being rescheduled.
6456
	 */
6457 6458
	if (!down)
		set_bit(__E1000_DOWN, &adapter->state);
6459 6460 6461
	del_timer_sync(&adapter->watchdog_timer);
	del_timer_sync(&adapter->phy_info_timer);

6462 6463 6464 6465 6466
	cancel_work_sync(&adapter->reset_task);
	cancel_work_sync(&adapter->watchdog_task);
	cancel_work_sync(&adapter->downshift_task);
	cancel_work_sync(&adapter->update_phy_task);
	cancel_work_sync(&adapter->print_hang_task);
6467

6468 6469 6470
	if (!(netdev->flags & IFF_UP))
		e1000_power_down_phy(adapter);

6471 6472 6473
	/* Don't lie to e1000_close() down the road. */
	if (!down)
		clear_bit(__E1000_DOWN, &adapter->state);
6474 6475
	unregister_netdev(netdev);

6476 6477
	if (pci_dev_run_wake(pdev))
		pm_runtime_get_noresume(&pdev->dev);
6478

6479 6480 6481 6482
	/*
	 * Release control of h/w to f/w.  If f/w is AMT enabled, this
	 * would have already happened in close and is redundant.
	 */
6483
	e1000e_release_hw_control(adapter);
6484

6485
	e1000e_reset_interrupt_capability(adapter);
6486 6487 6488 6489 6490 6491
	kfree(adapter->tx_ring);
	kfree(adapter->rx_ring);

	iounmap(adapter->hw.hw_addr);
	if (adapter->hw.flash_address)
		iounmap(adapter->hw.flash_address);
6492 6493
	pci_release_selected_regions(pdev,
	                             pci_select_bars(pdev, IORESOURCE_MEM));
6494 6495 6496

	free_netdev(netdev);

J
Jesse Brandeburg 已提交
6497
	/* AER disable */
6498
	pci_disable_pcie_error_reporting(pdev);
J
Jesse Brandeburg 已提交
6499

6500 6501 6502 6503 6504 6505 6506 6507 6508 6509
	pci_disable_device(pdev);
}

/* PCI Error Recovery (ERS) */
static struct pci_error_handlers e1000_err_handler = {
	.error_detected = e1000_io_error_detected,
	.slot_reset = e1000_io_slot_reset,
	.resume = e1000_io_resume,
};

6510
static DEFINE_PCI_DEVICE_TABLE(e1000_pci_tbl) = {
6511 6512 6513 6514 6515 6516
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_82571EB_COPPER), board_82571 },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_82571EB_FIBER), board_82571 },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_82571EB_QUAD_COPPER), board_82571 },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_82571EB_QUAD_COPPER_LP), board_82571 },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_82571EB_QUAD_FIBER), board_82571 },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_82571EB_SERDES), board_82571 },
6517 6518 6519
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_82571EB_SERDES_DUAL), board_82571 },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_82571EB_SERDES_QUAD), board_82571 },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_82571PT_QUAD_COPPER), board_82571 },
6520

6521 6522 6523 6524
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_82572EI), board_82572 },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_82572EI_COPPER), board_82572 },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_82572EI_FIBER), board_82572 },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_82572EI_SERDES), board_82572 },
6525

6526 6527 6528
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_82573E), board_82573 },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_82573E_IAMT), board_82573 },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_82573L), board_82573 },
6529

6530
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_82574L), board_82574 },
6531
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_82574LA), board_82574 },
6532
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_82583V), board_82583 },
6533

6534 6535 6536 6537 6538 6539 6540 6541
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_80003ES2LAN_COPPER_DPT),
	  board_80003es2lan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_80003ES2LAN_COPPER_SPT),
	  board_80003es2lan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_80003ES2LAN_SERDES_DPT),
	  board_80003es2lan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_80003ES2LAN_SERDES_SPT),
	  board_80003es2lan },
6542

6543 6544 6545 6546 6547 6548 6549
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH8_IFE), board_ich8lan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH8_IFE_G), board_ich8lan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH8_IFE_GT), board_ich8lan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH8_IGP_AMT), board_ich8lan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH8_IGP_C), board_ich8lan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH8_IGP_M), board_ich8lan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH8_IGP_M_AMT), board_ich8lan },
B
Bruce Allan 已提交
6550
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH8_82567V_3), board_ich8lan },
6551

6552 6553 6554 6555 6556
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH9_IFE), board_ich9lan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH9_IFE_G), board_ich9lan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH9_IFE_GT), board_ich9lan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH9_IGP_AMT), board_ich9lan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH9_IGP_C), board_ich9lan },
6557
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH9_BM), board_ich9lan },
6558 6559 6560 6561 6562 6563 6564
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH9_IGP_M), board_ich9lan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH9_IGP_M_AMT), board_ich9lan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH9_IGP_M_V), board_ich9lan },

	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH10_R_BM_LM), board_ich9lan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH10_R_BM_LF), board_ich9lan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH10_R_BM_V), board_ich9lan },
6565

6566 6567
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH10_D_BM_LM), board_ich10lan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH10_D_BM_LF), board_ich10lan },
6568
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_ICH10_D_BM_V), board_ich10lan },
6569

6570 6571 6572 6573 6574
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_PCH_M_HV_LM), board_pchlan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_PCH_M_HV_LC), board_pchlan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_PCH_D_HV_DM), board_pchlan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_PCH_D_HV_DC), board_pchlan },

6575 6576 6577
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_PCH2_LV_LM), board_pch2lan },
	{ PCI_VDEVICE(INTEL, E1000_DEV_ID_PCH2_LV_V), board_pch2lan },

6578
	{ 0, 0, 0, 0, 0, 0, 0 }	/* terminate list */
6579 6580 6581
};
MODULE_DEVICE_TABLE(pci, e1000_pci_tbl);

R
Rafael J. Wysocki 已提交
6582
#ifdef CONFIG_PM
6583
static const struct dev_pm_ops e1000_pm_ops = {
6584 6585 6586
	SET_SYSTEM_SLEEP_PM_OPS(e1000_suspend, e1000_resume)
	SET_RUNTIME_PM_OPS(e1000_runtime_suspend,
				e1000_runtime_resume, e1000_idle)
6587
};
6588
#endif
6589

6590 6591 6592 6593 6594 6595
/* PCI Device API Driver */
static struct pci_driver e1000_driver = {
	.name     = e1000e_driver_name,
	.id_table = e1000_pci_tbl,
	.probe    = e1000_probe,
	.remove   = __devexit_p(e1000_remove),
R
Rafael J. Wysocki 已提交
6596
#ifdef CONFIG_PM
6597 6598 6599
	.driver   = {
		.pm = &e1000_pm_ops,
	},
6600 6601 6602 6603 6604 6605 6606 6607 6608 6609 6610 6611 6612 6613
#endif
	.shutdown = e1000_shutdown,
	.err_handler = &e1000_err_handler
};

/**
 * e1000_init_module - Driver Registration Routine
 *
 * e1000_init_module is the first routine called when the driver is
 * loaded. All it does is register with the PCI subsystem.
 **/
static int __init e1000_init_module(void)
{
	int ret;
6614 6615
	pr_info("Intel(R) PRO/1000 Network Driver - %s\n",
		e1000e_driver_version);
B
Bruce Allan 已提交
6616
	pr_info("Copyright(c) 1999 - 2012 Intel Corporation.\n");
6617
	ret = pci_register_driver(&e1000_driver);
6618

6619 6620 6621 6622 6623 6624 6625 6626 6627 6628 6629 6630 6631 6632 6633 6634 6635 6636 6637 6638 6639 6640
	return ret;
}
module_init(e1000_init_module);

/**
 * e1000_exit_module - Driver Exit Cleanup Routine
 *
 * e1000_exit_module is called just before the driver is removed
 * from memory.
 **/
static void __exit e1000_exit_module(void)
{
	pci_unregister_driver(&e1000_driver);
}
module_exit(e1000_exit_module);


MODULE_AUTHOR("Intel Corporation, <linux.nics@intel.com>");
MODULE_DESCRIPTION("Intel(R) PRO/1000 Network Driver");
MODULE_LICENSE("GPL");
MODULE_VERSION(DRV_VERSION);

6641
/* netdev.c */