aerdrv_core.c 21.7 KB
Newer Older
1 2 3 4 5 6 7
/*
 * drivers/pci/pcie/aer/aerdrv_core.c
 *
 * This file is subject to the terms and conditions of the GNU General Public
 * License.  See the file "COPYING" in the main directory of this archive
 * for more details.
 *
8
 * This file implements the core part of PCIe AER. When a PCIe
9 10
 * error is delivered, an error message will be collected and printed to
 * console, then, an error recovery procedure will be executed by following
11
 * the PCI error recovery rules.
12 13 14 15 16 17 18 19 20 21 22 23 24 25
 *
 * Copyright (C) 2006 Intel Corp.
 *	Tom Long Nguyen (tom.l.nguyen@intel.com)
 *	Zhang Yanmin (yanmin.zhang@intel.com)
 *
 */

#include <linux/module.h>
#include <linux/pci.h>
#include <linux/kernel.h>
#include <linux/errno.h>
#include <linux/pm.h>
#include <linux/suspend.h>
#include <linux/delay.h>
26
#include <linux/slab.h>
27
#include <linux/kfifo.h>
28 29
#include "aerdrv.h"

30 31 32
#define	PCI_EXP_AER_FLAGS	(PCI_EXP_DEVCTL_CERE | PCI_EXP_DEVCTL_NFERE | \
				 PCI_EXP_DEVCTL_FERE | PCI_EXP_DEVCTL_URRE)

33 34
int pci_enable_pcie_error_reporting(struct pci_dev *dev)
{
35
	if (pcie_aer_get_firmware_first(dev))
36 37
		return -EIO;

K
Keith Busch 已提交
38
	if (!dev->aer_cap)
39 40
		return -EIO;

41
	return pcie_capability_set_word(dev, PCI_EXP_DEVCTL, PCI_EXP_AER_FLAGS);
42
}
43
EXPORT_SYMBOL_GPL(pci_enable_pcie_error_reporting);
44 45 46

int pci_disable_pcie_error_reporting(struct pci_dev *dev)
{
47
	if (pcie_aer_get_firmware_first(dev))
48 49
		return -EIO;

50 51
	return pcie_capability_clear_word(dev, PCI_EXP_DEVCTL,
					  PCI_EXP_AER_FLAGS);
52
}
53
EXPORT_SYMBOL_GPL(pci_disable_pcie_error_reporting);
54 55 56 57

int pci_cleanup_aer_uncorrect_error_status(struct pci_dev *dev)
{
	int pos;
58
	u32 status;
59

K
Keith Busch 已提交
60
	pos = dev->aer_cap;
61 62 63 64
	if (!pos)
		return -EIO;

	pci_read_config_dword(dev, pos + PCI_ERR_UNCOR_STATUS, &status);
65 66
	if (status)
		pci_write_config_dword(dev, pos + PCI_ERR_UNCOR_STATUS, status);
67 68 69

	return 0;
}
70
EXPORT_SYMBOL_GPL(pci_cleanup_aer_uncorrect_error_status);
71

72 73 74 75 76 77 78 79 80
int pci_cleanup_aer_error_status_regs(struct pci_dev *dev)
{
	int pos;
	u32 status;
	int port_type;

	if (!pci_is_pcie(dev))
		return -ENODEV;

K
Keith Busch 已提交
81
	pos = dev->aer_cap;
82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99
	if (!pos)
		return -EIO;

	port_type = pci_pcie_type(dev);
	if (port_type == PCI_EXP_TYPE_ROOT_PORT) {
		pci_read_config_dword(dev, pos + PCI_ERR_ROOT_STATUS, &status);
		pci_write_config_dword(dev, pos + PCI_ERR_ROOT_STATUS, status);
	}

	pci_read_config_dword(dev, pos + PCI_ERR_COR_STATUS, &status);
	pci_write_config_dword(dev, pos + PCI_ERR_COR_STATUS, status);

	pci_read_config_dword(dev, pos + PCI_ERR_UNCOR_STATUS, &status);
	pci_write_config_dword(dev, pos + PCI_ERR_UNCOR_STATUS, status);

	return 0;
}

K
Keith Busch 已提交
100 101 102 103 104 105
int pci_aer_init(struct pci_dev *dev)
{
	dev->aer_cap = pci_find_ext_capability(dev, PCI_EXT_CAP_ID_ERR);
	return pci_cleanup_aer_error_status_regs(dev);
}

106 107 108 109 110
/**
 * add_error_device - list device to be handled
 * @e_info: pointer to error info
 * @dev: pointer to pci_dev to be added
 */
Z
Zhang, Yanmin 已提交
111 112 113 114 115
static int add_error_device(struct aer_err_info *e_info, struct pci_dev *dev)
{
	if (e_info->error_dev_num < AER_MAX_MULTI_ERR_DEVICES) {
		e_info->dev[e_info->error_dev_num] = dev;
		e_info->error_dev_num++;
116
		return 0;
117
	}
118
	return -ENOSPC;
Z
Zhang, Yanmin 已提交
119 120
}

121 122 123 124 125 126
/**
 * is_error_source - check whether the device is source of reported error
 * @dev: pointer to pci_dev to be checked
 * @e_info: pointer to reported error info
 */
static bool is_error_source(struct pci_dev *dev, struct aer_err_info *e_info)
127 128
{
	int pos;
129
	u32 status, mask;
130 131 132 133 134 135
	u16 reg16;

	/*
	 * When bus id is equal to 0, it might be a bad id
	 * reported by root port.
	 */
136
	if ((PCI_BUS_NUM(e_info->id) != 0) &&
137
	    !(dev->bus->bus_flags & PCI_BUS_FLAGS_NO_AERSID)) {
138 139
		/* Device ID match? */
		if (e_info->id == ((dev->bus->number << 8) | dev->devfn))
140
			return true;
Z
Zhang, Yanmin 已提交
141

142
		/* Continue id comparing if there is no multiple error */
H
Hidetoshi Seto 已提交
143
		if (!e_info->multi_error_valid)
144
			return false;
145 146 147
	}

	/*
Z
Zhang, Yanmin 已提交
148
	 * When either
149
	 *      1) bus id is equal to 0. Some ports might lose the bus
Z
Zhang, Yanmin 已提交
150
	 *              id of error source id;
151 152
	 *      2) bus flag PCI_BUS_FLAGS_NO_AERSID is set
	 *      3) There are multiple errors and prior ID comparing fails;
153
	 * We check AER status registers to find possible reporter.
154 155
	 */
	if (atomic_read(&dev->enable_cnt) == 0)
156 157
		return false;

158
	/* Check if AER is enabled */
159 160
	pcie_capability_read_word(dev, PCI_EXP_DEVCTL, &reg16);
	if (!(reg16 & PCI_EXP_AER_FLAGS))
161
		return false;
162

K
Keith Busch 已提交
163
	pos = dev->aer_cap;
164
	if (!pos)
165
		return false;
166

167
	/* Check if error is recorded */
168
	if (e_info->severity == AER_CORRECTABLE) {
169 170
		pci_read_config_dword(dev, pos + PCI_ERR_COR_STATUS, &status);
		pci_read_config_dword(dev, pos + PCI_ERR_COR_MASK, &mask);
171
	} else {
172 173
		pci_read_config_dword(dev, pos + PCI_ERR_UNCOR_STATUS, &status);
		pci_read_config_dword(dev, pos + PCI_ERR_UNCOR_MASK, &mask);
174
	}
175 176
	if (status & ~mask)
		return true;
177

178 179
	return false;
}
Z
Zhang, Yanmin 已提交
180

181 182 183 184 185
static int find_device_iter(struct pci_dev *dev, void *data)
{
	struct aer_err_info *e_info = (struct aer_err_info *)data;

	if (is_error_source(dev, e_info)) {
186 187 188 189 190 191
		/* List this device */
		if (add_error_device(e_info, dev)) {
			/* We cannot handle more... Stop iteration */
			/* TODO: Should print error message here? */
			return 1;
		}
192 193 194 195 196 197

		/* If there is only a single error, stop iteration */
		if (!e_info->multi_error_valid)
			return 1;
	}
	return 0;
198 199 200 201
}

/**
 * find_source_device - search through device hierarchy for source device
R
Randy Dunlap 已提交
202
 * @parent: pointer to Root Port pci_dev data structure
203
 * @e_info: including detailed error information such like id
204
 *
205 206 207
 * Return true if found.
 *
 * Invoked by DPC when error is detected at the Root Port.
208 209 210
 * Caller of this function must set id, severity, and multi_error_valid of
 * struct aer_err_info pointed by @e_info properly.  This function must fill
 * e_info->error_dev_num and e_info->dev[], based on the given information.
R
Randy Dunlap 已提交
211
 */
212
static bool find_source_device(struct pci_dev *parent,
213
		struct aer_err_info *e_info)
214 215
{
	struct pci_dev *dev = parent;
216
	int result;
217

218 219 220
	/* Must reset in this function */
	e_info->error_dev_num = 0;

221
	/* Is Root Port an agent that sends error message? */
222 223
	result = find_device_iter(dev, e_info);
	if (result)
224
		return true;
225

226
	pci_walk_bus(parent->subordinate, find_device_iter, e_info);
227 228 229 230 231 232 233 234

	if (!e_info->error_dev_num) {
		dev_printk(KERN_DEBUG, &parent->dev,
				"can't find device of ID%04x\n",
				e_info->id);
		return false;
	}
	return true;
235 236
}

237
static int report_error_detected(struct pci_dev *dev, void *data)
238 239
{
	pci_ers_result_t vote;
240
	const struct pci_error_handlers *err_handler;
241 242 243
	struct aer_broadcast_data *result_data;
	result_data = (struct aer_broadcast_data *) data;

244
	device_lock(&dev->dev);
245 246 247 248 249 250
	dev->error_state = result_data->state;

	if (!dev->driver ||
		!dev->driver->err_handler ||
		!dev->driver->err_handler->error_detected) {
		if (result_data->state == pci_channel_io_frozen &&
251
			dev->hdr_type != PCI_HEADER_TYPE_BRIDGE) {
252 253 254 255 256 257 258
			/*
			 * In case of fatal recovery, if one of down-
			 * stream device has no driver. We might be
			 * unable to recover because a later insmod
			 * of a driver for this device is unaware of
			 * its hw state.
			 */
259 260 261
			dev_printk(KERN_DEBUG, &dev->dev, "device has %s\n",
				   dev->driver ?
				   "no AER-aware driver" : "no driver");
262
		}
263 264 265 266 267 268 269 270 271 272 273

		/*
		 * If there's any device in the subtree that does not
		 * have an error_detected callback, returning
		 * PCI_ERS_RESULT_NO_AER_DRIVER prevents calling of
		 * the subsequent mmio_enabled/slot_reset/resume
		 * callbacks of "any" device in the subtree. All the
		 * devices in the subtree are left in the error state
		 * without recovery.
		 */

274
		if (dev->hdr_type != PCI_HEADER_TYPE_BRIDGE)
275 276 277 278 279 280
			vote = PCI_ERS_RESULT_NO_AER_DRIVER;
		else
			vote = PCI_ERS_RESULT_NONE;
	} else {
		err_handler = dev->driver->err_handler;
		vote = err_handler->error_detected(dev, result_data->state);
281 282 283
	}

	result_data->result = merge_result(result_data->result, vote);
284
	device_unlock(&dev->dev);
285
	return 0;
286 287
}

288
static int report_mmio_enabled(struct pci_dev *dev, void *data)
289 290
{
	pci_ers_result_t vote;
291
	const struct pci_error_handlers *err_handler;
292 293 294
	struct aer_broadcast_data *result_data;
	result_data = (struct aer_broadcast_data *) data;

295
	device_lock(&dev->dev);
296 297 298
	if (!dev->driver ||
		!dev->driver->err_handler ||
		!dev->driver->err_handler->mmio_enabled)
299
		goto out;
300 301 302 303

	err_handler = dev->driver->err_handler;
	vote = err_handler->mmio_enabled(dev);
	result_data->result = merge_result(result_data->result, vote);
304 305
out:
	device_unlock(&dev->dev);
306
	return 0;
307 308
}

309
static int report_slot_reset(struct pci_dev *dev, void *data)
310 311
{
	pci_ers_result_t vote;
312
	const struct pci_error_handlers *err_handler;
313 314 315
	struct aer_broadcast_data *result_data;
	result_data = (struct aer_broadcast_data *) data;

316
	device_lock(&dev->dev);
317 318 319
	if (!dev->driver ||
		!dev->driver->err_handler ||
		!dev->driver->err_handler->slot_reset)
320
		goto out;
321 322 323 324

	err_handler = dev->driver->err_handler;
	vote = err_handler->slot_reset(dev);
	result_data->result = merge_result(result_data->result, vote);
325 326
out:
	device_unlock(&dev->dev);
327
	return 0;
328 329
}

330
static int report_resume(struct pci_dev *dev, void *data)
331
{
332
	const struct pci_error_handlers *err_handler;
333

334
	device_lock(&dev->dev);
335 336 337 338
	dev->error_state = pci_channel_io_normal;

	if (!dev->driver ||
		!dev->driver->err_handler ||
H
Hidetoshi Seto 已提交
339
		!dev->driver->err_handler->resume)
340
		goto out;
341 342 343

	err_handler = dev->driver->err_handler;
	err_handler->resume(dev);
344 345
out:
	device_unlock(&dev->dev);
346
	return 0;
347 348 349 350
}

/**
 * broadcast_error_message - handle message broadcast to downstream drivers
R
Randy Dunlap 已提交
351
 * @dev: pointer to from where in a hierarchy message is broadcasted down
352
 * @state: error state
R
Randy Dunlap 已提交
353 354
 * @error_mesg: message to print
 * @cb: callback to be broadcasted
355 356 357 358
 *
 * Invoked during error recovery process. Once being invoked, the content
 * of error severity will be broadcasted to all downstream drivers in a
 * hierarchy in question.
R
Randy Dunlap 已提交
359
 */
360 361 362
static pci_ers_result_t broadcast_error_message(struct pci_dev *dev,
	enum pci_channel_state state,
	char *error_mesg,
363
	int (*cb)(struct pci_dev *, void *))
364 365 366
{
	struct aer_broadcast_data result_data;

367
	dev_printk(KERN_DEBUG, &dev->dev, "broadcast %s message\n", error_mesg);
368 369 370 371 372 373
	result_data.state = state;
	if (cb == report_error_detected)
		result_data.result = PCI_ERS_RESULT_CAN_RECOVER;
	else
		result_data.result = PCI_ERS_RESULT_RECOVERED;

374
	if (dev->hdr_type == PCI_HEADER_TYPE_BRIDGE) {
375 376 377 378 379 380 381 382 383 384 385 386 387
		/*
		 * If the error is reported by a bridge, we think this error
		 * is related to the downstream link of the bridge, so we
		 * do error recovery on all subordinates of the bridge instead
		 * of the bridge and clear the error status of the bridge.
		 */
		if (cb == report_error_detected)
			dev->error_state = state;
		pci_walk_bus(dev->subordinate, cb, &result_data);
		if (cb == report_resume) {
			pci_cleanup_aer_uncorrect_error_status(dev);
			dev->error_state = pci_channel_io_normal;
		}
388
	} else {
389 390 391 392
		/*
		 * If the error is reported by an end point, we think this
		 * error is related to the upstream link of the end point.
		 */
393 394 395 396 397 398 399 400
		if (state == pci_channel_io_normal)
			/*
			 * the error is non fatal so the bus is ok, just invoke
			 * the callback for the function that logged the error.
			 */
			cb(dev, &result_data);
		else
			pci_walk_bus(dev->bus, cb, &result_data);
401 402 403 404 405
	}

	return result_data.result;
}

406
/**
407 408
 * default_reset_link - default reset function
 * @dev: pointer to pci_dev data structure
409
 *
410 411
 * Invoked when performing link reset on a Downstream Port or a
 * Root Port with no aer driver.
412
 */
413
static pci_ers_result_t default_reset_link(struct pci_dev *dev)
414
{
415
	pci_reset_bridge_secondary_bus(dev);
416
	dev_printk(KERN_DEBUG, &dev->dev, "downstream link has been reset\n");
417 418 419
	return PCI_ERS_RESULT_RECOVERED;
}

420 421
static int find_aer_service_iter(struct device *device, void *data)
{
422
	struct pcie_port_service_driver *service_driver, **drv;
423

424
	drv = (struct pcie_port_service_driver **) data;
425

426 427 428
	if (device->bus == &pcie_port_bus_type && device->driver) {
		service_driver = to_service_driver(device->driver);
		if (service_driver->service == PCIE_PORT_SERVICE_AER) {
429
			*drv = service_driver;
430
			return 1;
431 432 433 434 435 436
		}
	}

	return 0;
}

437
static struct pcie_port_service_driver *find_aer_service(struct pci_dev *dev)
438
{
439 440 441 442 443
	struct pcie_port_service_driver *drv = NULL;

	device_for_each_child(&dev->dev, &drv, find_aer_service_iter);

	return drv;
444 445
}

446
static pci_ers_result_t reset_link(struct pci_dev *dev)
447 448 449
{
	struct pci_dev *udev;
	pci_ers_result_t status;
450
	struct pcie_port_service_driver *driver;
451

452
	if (dev->hdr_type == PCI_HEADER_TYPE_BRIDGE) {
453
		/* Reset this port for all subordinates */
454
		udev = dev;
455 456
	} else {
		/* Reset the upstream component (likely downstream port) */
457
		udev = dev->bus->self;
458
	}
459

460 461
	/* Use the aer driver of the component firstly */
	driver = find_aer_service(udev);
462

463 464
	if (driver && driver->reset_link) {
		status = driver->reset_link(udev);
465
	} else if (udev->has_secondary_link) {
466
		status = default_reset_link(udev);
467 468 469 470 471
	} else {
		dev_printk(KERN_DEBUG, &dev->dev,
			"no link-reset support at upstream device %s\n",
			pci_name(udev));
		return PCI_ERS_RESULT_DISCONNECT;
472 473 474
	}

	if (status != PCI_ERS_RESULT_RECOVERED) {
475 476 477
		dev_printk(KERN_DEBUG, &dev->dev,
			"link reset at upstream device %s failed\n",
			pci_name(udev));
478 479 480 481 482 483 484 485 486 487 488 489 490 491
		return PCI_ERS_RESULT_DISCONNECT;
	}

	return status;
}

/**
 * do_recovery - handle nonfatal/fatal error recovery process
 * @dev: pointer to a pci_dev data structure of agent detecting an error
 * @severity: error severity type
 *
 * Invoked when an error is nonfatal/fatal. Once being invoked, broadcast
 * error detected message to all downstream drivers within a hierarchy in
 * question and return the returned code.
R
Randy Dunlap 已提交
492
 */
493
static void do_recovery(struct pci_dev *dev, int severity)
494 495 496 497 498 499 500 501 502 503 504 505 506 507 508
{
	pci_ers_result_t status, result = PCI_ERS_RESULT_RECOVERED;
	enum pci_channel_state state;

	if (severity == AER_FATAL)
		state = pci_channel_io_frozen;
	else
		state = pci_channel_io_normal;

	status = broadcast_error_message(dev,
			state,
			"error_detected",
			report_error_detected);

	if (severity == AER_FATAL) {
509
		result = reset_link(dev);
H
Hidetoshi Seto 已提交
510 511
		if (result != PCI_ERS_RESULT_RECOVERED)
			goto failed;
512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531
	}

	if (status == PCI_ERS_RESULT_CAN_RECOVER)
		status = broadcast_error_message(dev,
				state,
				"mmio_enabled",
				report_mmio_enabled);

	if (status == PCI_ERS_RESULT_NEED_RESET) {
		/*
		 * TODO: Should call platform-specific
		 * functions to reset slot before calling
		 * drivers' slot_reset callbacks?
		 */
		status = broadcast_error_message(dev,
				state,
				"slot_reset",
				report_slot_reset);
	}

H
Hidetoshi Seto 已提交
532 533 534 535
	if (status != PCI_ERS_RESULT_RECOVERED)
		goto failed;

	broadcast_error_message(dev,
536 537 538 539
				state,
				"resume",
				report_resume);

540
	dev_info(&dev->dev, "AER: Device recovery successful\n");
H
Hidetoshi Seto 已提交
541 542 543 544
	return;

failed:
	/* TODO: Should kernel panic here? */
545
	dev_info(&dev->dev, "AER: Device recovery failed\n");
546 547 548 549 550 551 552 553 554
}

/**
 * handle_error_source - handle logging error into an event log
 * @aerdev: pointer to pcie_device data structure of the root port
 * @dev: pointer to pci_dev data structure of error source device
 * @info: comprehensive error information
 *
 * Invoked when an error being detected by Root Port.
R
Randy Dunlap 已提交
555
 */
556
static void handle_error_source(struct pcie_device *aerdev,
557
	struct pci_dev *dev,
558
	struct aer_err_info *info)
559 560 561
{
	int pos;

562
	if (info->severity == AER_CORRECTABLE) {
563
		/*
564
		 * Correctable error does not need software intervention.
565 566
		 * No need to go through error recovery process.
		 */
K
Keith Busch 已提交
567
		pos = dev->aer_cap;
568 569
		if (pos)
			pci_write_config_dword(dev, pos + PCI_ERR_COR_STATUS,
570
					info->status);
H
Hidetoshi Seto 已提交
571
	} else
572
		do_recovery(dev, info->severity);
573 574
}

575 576 577 578 579 580
#ifdef CONFIG_ACPI_APEI_PCIEAER
static void aer_recover_work_func(struct work_struct *work);

#define AER_RECOVER_RING_ORDER		4
#define AER_RECOVER_RING_SIZE		(1 << AER_RECOVER_RING_ORDER)

R
Ryan Desfosses 已提交
581
struct aer_recover_entry {
582 583 584 585
	u8	bus;
	u8	devfn;
	u16	domain;
	int	severity;
586
	struct aer_capability_regs *regs;
587 588 589 590 591 592 593 594 595 596 597 598 599
};

static DEFINE_KFIFO(aer_recover_ring, struct aer_recover_entry,
		    AER_RECOVER_RING_SIZE);
/*
 * Mutual exclusion for writers of aer_recover_ring, reader side don't
 * need lock, because there is only one reader and lock is not needed
 * between reader and writer.
 */
static DEFINE_SPINLOCK(aer_recover_ring_lock);
static DECLARE_WORK(aer_recover_work, aer_recover_work_func);

void aer_recover_queue(int domain, unsigned int bus, unsigned int devfn,
600
		       int severity, struct aer_capability_regs *aer_regs)
601 602 603 604 605 606 607
{
	unsigned long flags;
	struct aer_recover_entry entry = {
		.bus		= bus,
		.devfn		= devfn,
		.domain		= domain,
		.severity	= severity,
608
		.regs		= aer_regs,
609 610 611
	};

	spin_lock_irqsave(&aer_recover_ring_lock, flags);
S
Stefani Seibold 已提交
612
	if (kfifo_put(&aer_recover_ring, entry))
613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634
		schedule_work(&aer_recover_work);
	else
		pr_err("AER recover: Buffer overflow when recovering AER for %04x:%02x:%02x:%x\n",
		       domain, bus, PCI_SLOT(devfn), PCI_FUNC(devfn));
	spin_unlock_irqrestore(&aer_recover_ring_lock, flags);
}
EXPORT_SYMBOL_GPL(aer_recover_queue);

static void aer_recover_work_func(struct work_struct *work)
{
	struct aer_recover_entry entry;
	struct pci_dev *pdev;

	while (kfifo_get(&aer_recover_ring, &entry)) {
		pdev = pci_get_domain_bus_and_slot(entry.domain, entry.bus,
						   entry.devfn);
		if (!pdev) {
			pr_err("AER recover: Can not find pci_dev for %04x:%02x:%02x:%x\n",
			       entry.domain, entry.bus,
			       PCI_SLOT(entry.devfn), PCI_FUNC(entry.devfn));
			continue;
		}
635
		cper_print_aer(pdev, entry.severity, entry.regs);
636
		do_recovery(pdev, entry.severity);
637
		pci_dev_put(pdev);
638 639 640 641
	}
}
#endif

642 643 644 645 646 647
/**
 * get_device_error_info - read error status from dev and store it to info
 * @dev: pointer to the device expected to have a error record
 * @info: pointer to structure to store the error record
 *
 * Return 1 on success, 0 on error.
648 649
 *
 * Note that @info is reused among all error devices. Clear fields properly.
650
 */
651 652
static int get_device_error_info(struct pci_dev *dev, struct aer_err_info *info)
{
653
	int pos, temp;
654

655
	/* Must reset in this function */
656
	info->status = 0;
H
Hidetoshi Seto 已提交
657
	info->tlp_header_valid = 0;
658

K
Keith Busch 已提交
659
	pos = dev->aer_cap;
660 661 662

	/* The device might not support AER */
	if (!pos)
663
		return 1;
664 665 666 667

	if (info->severity == AER_CORRECTABLE) {
		pci_read_config_dword(dev, pos + PCI_ERR_COR_STATUS,
			&info->status);
668 669 670
		pci_read_config_dword(dev, pos + PCI_ERR_COR_MASK,
			&info->mask);
		if (!(info->status & ~info->mask))
671
			return 0;
672
	} else if (dev->hdr_type == PCI_HEADER_TYPE_BRIDGE ||
673 674 675 676 677
		info->severity == AER_NONFATAL) {

		/* Link is still healthy for IO reads */
		pci_read_config_dword(dev, pos + PCI_ERR_UNCOR_STATUS,
			&info->status);
678 679 680
		pci_read_config_dword(dev, pos + PCI_ERR_UNCOR_MASK,
			&info->mask);
		if (!(info->status & ~info->mask))
681
			return 0;
682

683 684
		/* Get First Error Pointer */
		pci_read_config_dword(dev, pos + PCI_ERR_CAP, &temp);
H
Hidetoshi Seto 已提交
685
		info->first_error = PCI_ERR_CAP_FEP(temp);
686

687
		if (info->status & AER_LOG_TLP_MASKS) {
H
Hidetoshi Seto 已提交
688
			info->tlp_header_valid = 1;
689 690 691 692 693 694 695 696 697 698 699
			pci_read_config_dword(dev,
				pos + PCI_ERR_HEADER_LOG, &info->tlp.dw0);
			pci_read_config_dword(dev,
				pos + PCI_ERR_HEADER_LOG + 4, &info->tlp.dw1);
			pci_read_config_dword(dev,
				pos + PCI_ERR_HEADER_LOG + 8, &info->tlp.dw2);
			pci_read_config_dword(dev,
				pos + PCI_ERR_HEADER_LOG + 12, &info->tlp.dw3);
		}
	}

700
	return 1;
701 702
}

Z
Zhang, Yanmin 已提交
703 704 705 706 707
static inline void aer_process_err_devices(struct pcie_device *p_device,
			struct aer_err_info *e_info)
{
	int i;

708
	/* Report all before handle them, not to lost records by reset etc. */
Z
Zhang, Yanmin 已提交
709
	for (i = 0; i < e_info->error_dev_num && e_info->dev[i]; i++) {
710
		if (get_device_error_info(e_info->dev[i], e_info))
Z
Zhang, Yanmin 已提交
711
			aer_print_error(e_info->dev[i], e_info);
712 713 714 715
	}
	for (i = 0; i < e_info->error_dev_num && e_info->dev[i]; i++) {
		if (get_device_error_info(e_info->dev[i], e_info))
			handle_error_source(p_device, e_info->dev[i], e_info);
Z
Zhang, Yanmin 已提交
716 717 718
	}
}

719 720 721 722
/**
 * aer_isr_one_error - consume an error detected by root port
 * @p_device: pointer to error root port service device
 * @e_src: pointer to an error source
R
Randy Dunlap 已提交
723
 */
724 725 726
static void aer_isr_one_error(struct pcie_device *p_device,
		struct aer_err_source *e_src)
{
727 728
	struct aer_rpc *rpc = get_service_data(p_device);
	struct aer_err_info *e_info = &rpc->e_info;
729 730 731 732 733

	/*
	 * There is a possibility that both correctable error and
	 * uncorrectable error being logged. Report correctable error first.
	 */
734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757
	if (e_src->status & PCI_ERR_ROOT_COR_RCV) {
		e_info->id = ERR_COR_ID(e_src->id);
		e_info->severity = AER_CORRECTABLE;

		if (e_src->status & PCI_ERR_ROOT_MULTI_COR_RCV)
			e_info->multi_error_valid = 1;
		else
			e_info->multi_error_valid = 0;

		aer_print_port_info(p_device->port, e_info);

		if (find_source_device(p_device->port, e_info))
			aer_process_err_devices(p_device, e_info);
	}

	if (e_src->status & PCI_ERR_ROOT_UNCOR_RCV) {
		e_info->id = ERR_UNCOR_ID(e_src->id);

		if (e_src->status & PCI_ERR_ROOT_FATAL_RCV)
			e_info->severity = AER_FATAL;
		else
			e_info->severity = AER_NONFATAL;

		if (e_src->status & PCI_ERR_ROOT_MULTI_UNCOR_RCV)
H
Hidetoshi Seto 已提交
758
			e_info->multi_error_valid = 1;
759 760
		else
			e_info->multi_error_valid = 0;
761

762 763
		aer_print_port_info(p_device->port, e_info);

764 765
		if (find_source_device(p_device->port, e_info))
			aer_process_err_devices(p_device, e_info);
766 767 768
	}
}

769 770 771 772 773 774 775 776 777 778 779 780 781 782 783
/**
 * get_e_source - retrieve an error source
 * @rpc: pointer to the root port which holds an error
 * @e_src: pointer to store retrieved error source
 *
 * Return 1 if an error source is retrieved, otherwise 0.
 *
 * Invoked by DPC handler to consume an error.
 */
static int get_e_source(struct aer_rpc *rpc, struct aer_err_source *e_src)
{
	unsigned long flags;

	/* Lock access to Root error producer/consumer index */
	spin_lock_irqsave(&rpc->e_lock, flags);
784 785 786
	if (rpc->prod_idx == rpc->cons_idx) {
		spin_unlock_irqrestore(&rpc->e_lock, flags);
		return 0;
787
	}
788 789 790 791 792

	*e_src = rpc->e_sources[rpc->cons_idx];
	rpc->cons_idx++;
	if (rpc->cons_idx == AER_ERROR_SOURCES_MAX)
		rpc->cons_idx = 0;
793 794
	spin_unlock_irqrestore(&rpc->e_lock, flags);

795
	return 1;
796 797
}

798 799
/**
 * aer_isr - consume errors detected by root port
800
 * @work: definition of this work item
801 802
 *
 * Invoked, as DPC, when root port records new detected error
R
Randy Dunlap 已提交
803
 */
804
void aer_isr(struct work_struct *work)
805
{
806 807
	struct aer_rpc *rpc = container_of(work, struct aer_rpc, dpc_handler);
	struct pcie_device *p_device = rpc->rpd;
808
	struct aer_err_source uninitialized_var(e_src);
809 810

	mutex_lock(&rpc->rpc_mutex);
811 812
	while (get_e_source(rpc, &e_src))
		aer_isr_one_error(p_device, &e_src);
813 814
	mutex_unlock(&rpc->rpc_mutex);
}