watchdog.c 21.4 KB
Newer Older
1 2 3 4 5
/*
 * Detect hard and soft lockups on a system
 *
 * started by Don Zickus, Copyright (C) 2010 Red Hat, Inc.
 *
6 7 8
 * Note: Most of this code is borrowed heavily from the original softlockup
 * detector, so thanks to Ingo for the initial implementation.
 * Some chunks also taken from the old x86-specific nmi watchdog code, thanks
9 10 11
 * to those contributors as well.
 */

12
#define pr_fmt(fmt) "watchdog: " fmt
13

14 15 16 17 18 19
#include <linux/mm.h>
#include <linux/cpu.h>
#include <linux/nmi.h>
#include <linux/init.h>
#include <linux/module.h>
#include <linux/sysctl.h>
20
#include <linux/smpboot.h>
21
#include <linux/sched/rt.h>
22
#include <uapi/linux/sched/types.h>
23
#include <linux/tick.h>
24
#include <linux/workqueue.h>
25
#include <linux/sched/clock.h>
26
#include <linux/sched/debug.h>
27 28

#include <asm/irq_regs.h>
29
#include <linux/kvm_para.h>
30
#include <linux/kthread.h>
31

32
static DEFINE_MUTEX(watchdog_mutex);
33

34 35 36 37 38
int __read_mostly nmi_watchdog_enabled;

#if defined(CONFIG_HARDLOCKUP_DETECTOR) || defined(CONFIG_HAVE_NMI_WATCHDOG)
unsigned long __read_mostly watchdog_enabled = SOFT_WATCHDOG_ENABLED |
						NMI_WATCHDOG_ENABLED;
39
#else
40
unsigned long __read_mostly watchdog_enabled = SOFT_WATCHDOG_ENABLED;
41
#endif
42 43 44 45 46 47 48 49 50 51 52 53 54 55 56

#ifdef CONFIG_HARDLOCKUP_DETECTOR
/*
 * Should we panic when a soft-lockup or hard-lockup occurs:
 */
unsigned int __read_mostly hardlockup_panic =
			CONFIG_BOOTPARAM_HARDLOCKUP_PANIC_VALUE;
/*
 * We may not want to enable hard lockup detection by default in all cases,
 * for example when running the kernel as a guest on a hypervisor. In these
 * cases this function can be called to disable hard lockup detection. This
 * function should only be executed once by the boot processor before the
 * kernel command line parameters are parsed, because otherwise it is not
 * possible to override this in hardlockup_panic_setup().
 */
57
void __init hardlockup_detector_disable(void)
58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75
{
	watchdog_enabled &= ~NMI_WATCHDOG_ENABLED;
}

static int __init hardlockup_panic_setup(char *str)
{
	if (!strncmp(str, "panic", 5))
		hardlockup_panic = 1;
	else if (!strncmp(str, "nopanic", 7))
		hardlockup_panic = 0;
	else if (!strncmp(str, "0", 1))
		watchdog_enabled &= ~NMI_WATCHDOG_ENABLED;
	else if (!strncmp(str, "1", 1))
		watchdog_enabled |= NMI_WATCHDOG_ENABLED;
	return 1;
}
__setup("nmi_watchdog=", hardlockup_panic_setup);

76 77
# ifdef CONFIG_SMP
int __read_mostly sysctl_hardlockup_all_cpu_backtrace;
78

79 80 81 82 83 84 85 86
static int __init hardlockup_all_cpu_backtrace_setup(char *str)
{
	sysctl_hardlockup_all_cpu_backtrace = !!simple_strtol(str, NULL, 0);
	return 1;
}
__setup("hardlockup_all_cpu_backtrace=", hardlockup_all_cpu_backtrace_setup);
# endif /* CONFIG_SMP */
#endif /* CONFIG_HARDLOCKUP_DETECTOR */
87

88
int __read_mostly watchdog_user_enabled;
89
int __read_mostly watchdog_thresh = 10;
90

91
struct cpumask watchdog_cpumask __read_mostly;
92 93
unsigned long *watchdog_cpumask_bits = cpumask_bits(&watchdog_cpumask);

94 95 96
/*
 * These functions can be overridden if an architecture implements its
 * own hardlockup detector.
97 98 99 100
 *
 * watchdog_nmi_enable/disable can be implemented to start and stop when
 * softlockup watchdog threads start and stop. The arch must select the
 * SOFTLOCKUP_DETECTOR Kconfig.
101 102 103 104 105
 */
int __weak watchdog_nmi_enable(unsigned int cpu)
{
	return 0;
}
106

107 108
void __weak watchdog_nmi_disable(unsigned int cpu)
{
109
	hardlockup_detector_perf_disable();
110 111
}

112 113 114 115 116 117 118 119 120 121
/*
 * watchdog_nmi_reconfigure can be implemented to be notified after any
 * watchdog configuration change. The arch hardlockup watchdog should
 * respond to the following variables:
 * - nmi_watchdog_enabled
 * - watchdog_thresh
 * - watchdog_cpumask
 * - sysctl_hardlockup_all_cpu_backtrace
 * - hardlockup_panic
 */
122
void __weak watchdog_nmi_reconfigure(void) { }
123

124 125
#ifdef CONFIG_SOFTLOCKUP_DETECTOR

126 127 128 129 130
/* Global variables, exported for sysctl */
unsigned int __read_mostly softlockup_panic =
			CONFIG_BOOTPARAM_SOFTLOCKUP_PANIC_VALUE;
int __read_mostly soft_watchdog_enabled;

131 132 133
struct cpumask watchdog_allowed_mask __read_mostly;
static bool softlockup_threads_initialized __read_mostly;

134
static u64 __read_mostly sample_period;
135 136 137 138 139 140

static DEFINE_PER_CPU(unsigned long, watchdog_touch_ts);
static DEFINE_PER_CPU(struct task_struct *, softlockup_watchdog);
static DEFINE_PER_CPU(struct hrtimer, watchdog_hrtimer);
static DEFINE_PER_CPU(bool, softlockup_touch_sync);
static DEFINE_PER_CPU(bool, soft_watchdog_warn);
141 142
static DEFINE_PER_CPU(unsigned long, hrtimer_interrupts);
static DEFINE_PER_CPU(unsigned long, soft_lockup_hrtimer_cnt);
143
static DEFINE_PER_CPU(struct task_struct *, softlockup_task_ptr_saved);
144
static DEFINE_PER_CPU(unsigned long, hrtimer_interrupts_saved);
145
static unsigned long soft_lockup_nmi_warn;
146 147 148 149 150 151 152 153 154 155

static int __init softlockup_panic_setup(char *str)
{
	softlockup_panic = simple_strtoul(str, NULL, 0);
	return 1;
}
__setup("softlockup_panic=", softlockup_panic_setup);

static int __init nowatchdog_setup(char *str)
{
156
	watchdog_enabled = 0;
157 158 159 160 161 162
	return 1;
}
__setup("nowatchdog", nowatchdog_setup);

static int __init nosoftlockup_setup(char *str)
{
163
	watchdog_enabled &= ~SOFT_WATCHDOG_ENABLED;
164 165 166
	return 1;
}
__setup("nosoftlockup", nosoftlockup_setup);
167

168
#ifdef CONFIG_SMP
169 170
int __read_mostly sysctl_softlockup_all_cpu_backtrace;

171 172
static int __init softlockup_all_cpu_backtrace_setup(char *str)
{
173
	sysctl_softlockup_all_cpu_backtrace = !!simple_strtol(str, NULL, 0);
174 175 176
	return 1;
}
__setup("softlockup_all_cpu_backtrace=", softlockup_all_cpu_backtrace_setup);
177
#endif
178

179 180
static void __lockup_detector_cleanup(void);

181 182 183 184 185 186 187
/*
 * Hard-lockup warnings should be triggered after just a few seconds. Soft-
 * lockups can have false positives under extreme conditions. So we generally
 * want a higher threshold for soft lockups than for hard lockups. So we couple
 * the thresholds with a factor: we make the soft threshold twice the amount of
 * time the hard threshold is.
 */
188
static int get_softlockup_thresh(void)
189 190 191
{
	return watchdog_thresh * 2;
}
192 193 194 195 196 197

/*
 * Returns seconds, approximately.  We don't need nanosecond
 * resolution, and we don't need to waste time with a big divide when
 * 2^30ns == 1.074s.
 */
198
static unsigned long get_timestamp(void)
199
{
200
	return running_clock() >> 30LL;  /* 2^30 ~= 10^9 */
201 202
}

203
static void set_sample_period(void)
204 205
{
	/*
206
	 * convert watchdog_thresh from seconds to ns
207 208 209 210
	 * the divide by 5 is to give hrtimer several chances (two
	 * or three with the current relation between the soft
	 * and hard thresholds) to increment before the
	 * hardlockup detector generates a warning
211
	 */
212
	sample_period = get_softlockup_thresh() * ((u64)NSEC_PER_SEC / 5);
213
	watchdog_update_hrtimer_threshold(sample_period);
214 215 216 217 218
}

/* Commands for resetting the watchdog */
static void __touch_watchdog(void)
{
219
	__this_cpu_write(watchdog_touch_ts, get_timestamp());
220 221
}

222 223 224 225 226 227 228 229 230
/**
 * touch_softlockup_watchdog_sched - touch watchdog on scheduler stalls
 *
 * Call when the scheduler may have stalled for legitimate reasons
 * preventing the watchdog task from executing - e.g. the scheduler
 * entering idle state.  This should only be used for scheduler events.
 * Use touch_softlockup_watchdog() for everything else.
 */
void touch_softlockup_watchdog_sched(void)
231
{
232 233 234 235 236
	/*
	 * Preemption can be enabled.  It doesn't matter which CPU's timestamp
	 * gets zeroed here, so use the raw_ operation.
	 */
	raw_cpu_write(watchdog_touch_ts, 0);
237
}
238 239 240 241

void touch_softlockup_watchdog(void)
{
	touch_softlockup_watchdog_sched();
242
	wq_watchdog_touch(raw_smp_processor_id());
243
}
244
EXPORT_SYMBOL(touch_softlockup_watchdog);
245

246
void touch_all_softlockup_watchdogs(void)
247 248 249 250
{
	int cpu;

	/*
251 252 253 254 255 256 257
	 * watchdog_mutex cannpt be taken here, as this might be called
	 * from (soft)interrupt context, so the access to
	 * watchdog_allowed_cpumask might race with a concurrent update.
	 *
	 * The watchdog time stamp can race against a concurrent real
	 * update as well, the only side effect might be a cycle delay for
	 * the softlockup check.
258
	 */
259
	for_each_cpu(cpu, &watchdog_allowed_mask)
260
		per_cpu(watchdog_touch_ts, cpu) = 0;
261
	wq_watchdog_touch(-1);
262 263 264 265
}

void touch_softlockup_watchdog_sync(void)
{
266 267
	__this_cpu_write(softlockup_touch_sync, true);
	__this_cpu_write(watchdog_touch_ts, 0);
268 269
}

270
static int is_softlockup(unsigned long touch_ts)
271
{
272
	unsigned long now = get_timestamp();
273

274
	if ((watchdog_enabled & SOFT_WATCHDOG_ENABLED) && watchdog_thresh){
275 276 277 278
		/* Warn about unreasonable delays. */
		if (time_after(now, touch_ts + get_softlockup_thresh()))
			return now - touch_ts;
	}
279 280 281
	return 0;
}

282 283
/* watchdog detector functions */
bool is_hardlockup(void)
284
{
285
	unsigned long hrint = __this_cpu_read(hrtimer_interrupts);
286

287 288 289 290 291
	if (__this_cpu_read(hrtimer_interrupts_saved) == hrint)
		return true;

	__this_cpu_write(hrtimer_interrupts_saved, hrint);
	return false;
292
}
293 294

static void watchdog_interrupt_count(void)
295
{
296
	__this_cpu_inc(hrtimer_interrupts);
297
}
298 299 300 301

/* watchdog kicker functions */
static enum hrtimer_restart watchdog_timer_fn(struct hrtimer *hrtimer)
{
302
	unsigned long touch_ts = __this_cpu_read(watchdog_touch_ts);
303 304
	struct pt_regs *regs = get_irq_regs();
	int duration;
305
	int softlockup_all_cpu_backtrace = sysctl_softlockup_all_cpu_backtrace;
306

307
	if (!watchdog_enabled)
308 309
		return HRTIMER_NORESTART;

310 311 312 313
	/* kick the hardlockup detector */
	watchdog_interrupt_count();

	/* kick the softlockup detector */
314
	wake_up_process(__this_cpu_read(softlockup_watchdog));
315 316

	/* .. and repeat */
317
	hrtimer_forward_now(hrtimer, ns_to_ktime(sample_period));
318 319

	if (touch_ts == 0) {
320
		if (unlikely(__this_cpu_read(softlockup_touch_sync))) {
321 322 323 324
			/*
			 * If the time stamp was touched atomically
			 * make sure the scheduler tick is up to date.
			 */
325
			__this_cpu_write(softlockup_touch_sync, false);
326 327
			sched_clock_tick();
		}
328 329 330

		/* Clear the guest paused flag on watchdog reset */
		kvm_check_and_clear_guest_paused();
331 332 333 334 335 336 337 338 339 340
		__touch_watchdog();
		return HRTIMER_RESTART;
	}

	/* check for a softlockup
	 * This is done by making sure a high priority task is
	 * being scheduled.  The task touches the watchdog to
	 * indicate it is getting cpu time.  If it hasn't then
	 * this is a good indication some task is hogging the cpu
	 */
341
	duration = is_softlockup(touch_ts);
342
	if (unlikely(duration)) {
343 344 345 346 347 348 349 350
		/*
		 * If a virtual machine is stopped by the host it can look to
		 * the watchdog like a soft lockup, check to see if the host
		 * stopped the vm before we issue the warning
		 */
		if (kvm_check_and_clear_guest_paused())
			return HRTIMER_RESTART;

351
		/* only warn once */
352 353 354 355 356 357 358 359 360 361 362 363 364 365
		if (__this_cpu_read(soft_watchdog_warn) == true) {
			/*
			 * When multiple processes are causing softlockups the
			 * softlockup detector only warns on the first one
			 * because the code relies on a full quiet cycle to
			 * re-arm.  The second process prevents the quiet cycle
			 * and never gets reported.  Use task pointers to detect
			 * this.
			 */
			if (__this_cpu_read(softlockup_task_ptr_saved) !=
			    current) {
				__this_cpu_write(soft_watchdog_warn, false);
				__touch_watchdog();
			}
366
			return HRTIMER_RESTART;
367
		}
368

369 370 371 372 373 374 375 376 377 378 379
		if (softlockup_all_cpu_backtrace) {
			/* Prevent multiple soft-lockup reports if one cpu is already
			 * engaged in dumping cpu back traces
			 */
			if (test_and_set_bit(0, &soft_lockup_nmi_warn)) {
				/* Someone else will report us. Let's give up */
				__this_cpu_write(soft_watchdog_warn, true);
				return HRTIMER_RESTART;
			}
		}

380
		pr_emerg("BUG: soft lockup - CPU#%d stuck for %us! [%s:%d]\n",
381
			smp_processor_id(), duration,
382
			current->comm, task_pid_nr(current));
383
		__this_cpu_write(softlockup_task_ptr_saved, current);
384 385 386 387 388 389 390
		print_modules();
		print_irqtrace_events(current);
		if (regs)
			show_regs(regs);
		else
			dump_stack();

391 392 393 394 395 396 397 398 399 400 401
		if (softlockup_all_cpu_backtrace) {
			/* Avoid generating two back traces for current
			 * given that one is already made above
			 */
			trigger_allbutself_cpu_backtrace();

			clear_bit(0, &soft_lockup_nmi_warn);
			/* Barrier to sync with other cpus */
			smp_mb__after_atomic();
		}

Josh Hunt's avatar
Josh Hunt committed
402
		add_taint(TAINT_SOFTLOCKUP, LOCKDEP_STILL_OK);
403 404
		if (softlockup_panic)
			panic("softlockup: hung tasks");
405
		__this_cpu_write(soft_watchdog_warn, true);
406
	} else
407
		__this_cpu_write(soft_watchdog_warn, false);
408 409 410 411

	return HRTIMER_RESTART;
}

412 413 414
static void watchdog_set_prio(unsigned int policy, unsigned int prio)
{
	struct sched_param param = { .sched_priority = prio };
415

416 417 418 419
	sched_setscheduler(current, policy, &param);
}

static void watchdog_enable(unsigned int cpu)
420
{
421
	struct hrtimer *hrtimer = this_cpu_ptr(&watchdog_hrtimer);
422

423 424 425 426
	/*
	 * Start the timer first to prevent the NMI watchdog triggering
	 * before the timer has a chance to fire.
	 */
427 428
	hrtimer_init(hrtimer, CLOCK_MONOTONIC, HRTIMER_MODE_REL);
	hrtimer->function = watchdog_timer_fn;
429 430
	hrtimer_start(hrtimer, ns_to_ktime(sample_period),
		      HRTIMER_MODE_REL_PINNED);
431

432 433
	/* Initialize timestamp */
	__touch_watchdog();
434 435
	/* Enable the perf event */
	watchdog_nmi_enable(cpu);
436

437 438
	watchdog_set_prio(SCHED_FIFO, MAX_RT_PRIO - 1);
}
439

440 441
static void watchdog_disable(unsigned int cpu)
{
442
	struct hrtimer *hrtimer = this_cpu_ptr(&watchdog_hrtimer);
443

444
	watchdog_set_prio(SCHED_NORMAL, 0);
445 446 447 448 449
	/*
	 * Disable the perf event first. That prevents that a large delay
	 * between disabling the timer and disabling the perf event causes
	 * the perf NMI to detect a false positive.
	 */
450
	watchdog_nmi_disable(cpu);
451
	hrtimer_cancel(hrtimer);
452 453
}

454 455 456 457 458
static void watchdog_cleanup(unsigned int cpu, bool online)
{
	watchdog_disable(cpu);
}

459 460 461 462 463 464 465 466 467
static int watchdog_should_run(unsigned int cpu)
{
	return __this_cpu_read(hrtimer_interrupts) !=
		__this_cpu_read(soft_lockup_hrtimer_cnt);
}

/*
 * The watchdog thread function - touches the timestamp.
 *
468
 * It only runs once every sample_period seconds (4 seconds by
469 470 471 472 473 474 475 476 477 478
 * default) to reset the softlockup timestamp. If this gets delayed
 * for more than 2*watchdog_thresh seconds then the debug-printout
 * triggers in watchdog_timer_fn().
 */
static void watchdog(unsigned int cpu)
{
	__this_cpu_write(soft_lockup_hrtimer_cnt,
			 __this_cpu_read(hrtimer_interrupts));
	__touch_watchdog();
}
479

480 481 482 483 484 485 486 487 488 489 490
static struct smp_hotplug_thread watchdog_threads = {
	.store			= &softlockup_watchdog,
	.thread_should_run	= watchdog_should_run,
	.thread_fn		= watchdog,
	.thread_comm		= "watchdog/%u",
	.setup			= watchdog_enable,
	.cleanup		= watchdog_cleanup,
	.park			= watchdog_disable,
	.unpark			= watchdog_enable,
};

491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559
static void softlockup_update_smpboot_threads(void)
{
	lockdep_assert_held(&watchdog_mutex);

	if (!softlockup_threads_initialized)
		return;

	smpboot_update_cpumask_percpu_thread(&watchdog_threads,
					     &watchdog_allowed_mask);
	__lockup_detector_cleanup();
}

/* Temporarily park all watchdog threads */
static void softlockup_park_all_threads(void)
{
	cpumask_clear(&watchdog_allowed_mask);
	softlockup_update_smpboot_threads();
}

/*
 * Park threads which are not longer enabled and unpark threads which have
 * been newly enabled.
 */
static void softlockup_update_threads(void)
{
	cpumask_copy(&watchdog_allowed_mask, &watchdog_cpumask);
	softlockup_update_smpboot_threads();
}

static void softlockup_reconfigure_threads(bool enabled)
{
	softlockup_park_all_threads();
	set_sample_period();
	if (enabled)
		softlockup_update_threads();
}

/*
 * Create the watchdog thread infrastructure.
 *
 * The threads are not unparked as watchdog_allowed_mask is empty.  When
 * the threads are sucessfully initialized, take the proper locks and
 * unpark the threads in the watchdog_cpumask if the watchdog is enabled.
 */
static __init void softlockup_init_threads(void)
{
	int ret;

	/*
	 * If sysctl is off and watchdog got disabled on the command line,
	 * nothing to do here.
	 */
	if (!IS_ENABLED(CONFIG_SYSCTL) &&
	    !(watchdog_enabled && watchdog_thresh))
		return;

	ret = smpboot_register_percpu_thread_cpumask(&watchdog_threads,
						     &watchdog_allowed_mask);
	if (ret) {
		pr_err("Failed to initialize soft lockup detector threads\n");
		return;
	}

	mutex_lock(&watchdog_mutex);
	softlockup_threads_initialized = true;
	softlockup_reconfigure_threads(watchdog_enabled && watchdog_thresh);
	mutex_unlock(&watchdog_mutex);
}

560 561 562 563 564
#else /* CONFIG_SOFTLOCKUP_DETECTOR */
static inline int watchdog_park_threads(void) { return 0; }
static inline void watchdog_unpark_threads(void) { }
static inline int watchdog_enable_all_cpus(void) { return 0; }
static inline void watchdog_disable_all_cpus(void) { }
565 566 567
static inline void softlockup_init_threads(void) { }
static inline void softlockup_update_threads(void) { }
static inline void softlockup_reconfigure_threads(bool enabled) { }
568
#endif /* !CONFIG_SOFTLOCKUP_DETECTOR */
569

570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587
static void __lockup_detector_cleanup(void)
{
	lockdep_assert_held(&watchdog_mutex);
	hardlockup_detector_perf_cleanup();
}

/**
 * lockup_detector_cleanup - Cleanup after cpu hotplug or sysctl changes
 *
 * Caller must not hold the cpu hotplug rwsem.
 */
void lockup_detector_cleanup(void)
{
	mutex_lock(&watchdog_mutex);
	__lockup_detector_cleanup();
	mutex_unlock(&watchdog_mutex);
}

588 589 590 591 592 593 594 595 596 597 598
/**
 * lockup_detector_soft_poweroff - Interface to stop lockup detector(s)
 *
 * Special interface for parisc. It prevents lockup detector warnings from
 * the default pm_poweroff() function which busy loops forever.
 */
void lockup_detector_soft_poweroff(void)
{
	watchdog_enabled = 0;
}

599 600
#ifdef CONFIG_SYSCTL

601
/*
602 603
 * Update the run state of the lockup detectors.
 */
604
static void proc_watchdog_update(void)
605
{
606
	softlockup_reconfigure_threads(watchdog_enabled && watchdog_thresh);
607
	watchdog_nmi_reconfigure();
608 609
}

610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627
/*
 * common function for watchdog, nmi_watchdog and soft_watchdog parameter
 *
 * caller             | table->data points to | 'which' contains the flag(s)
 * -------------------|-----------------------|-----------------------------
 * proc_watchdog      | watchdog_user_enabled | NMI_WATCHDOG_ENABLED or'ed
 *                    |                       | with SOFT_WATCHDOG_ENABLED
 * -------------------|-----------------------|-----------------------------
 * proc_nmi_watchdog  | nmi_watchdog_enabled  | NMI_WATCHDOG_ENABLED
 * -------------------|-----------------------|-----------------------------
 * proc_soft_watchdog | soft_watchdog_enabled | SOFT_WATCHDOG_ENABLED
 */
static int proc_watchdog_common(int which, struct ctl_table *table, int write,
				void __user *buffer, size_t *lenp, loff_t *ppos)
{
	int err, old, new;
	int *watchdog_param = (int *)table->data;

628
	cpu_hotplug_disable();
629
	mutex_lock(&watchdog_mutex);
630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662

	/*
	 * If the parameter is being read return the state of the corresponding
	 * bit(s) in 'watchdog_enabled', else update 'watchdog_enabled' and the
	 * run state of the lockup detectors.
	 */
	if (!write) {
		*watchdog_param = (watchdog_enabled & which) != 0;
		err = proc_dointvec_minmax(table, write, buffer, lenp, ppos);
	} else {
		err = proc_dointvec_minmax(table, write, buffer, lenp, ppos);
		if (err)
			goto out;

		/*
		 * There is a race window between fetching the current value
		 * from 'watchdog_enabled' and storing the new value. During
		 * this race window, watchdog_nmi_enable() can sneak in and
		 * clear the NMI_WATCHDOG_ENABLED bit in 'watchdog_enabled'.
		 * The 'cmpxchg' detects this race and the loop retries.
		 */
		do {
			old = watchdog_enabled;
			/*
			 * If the parameter value is not zero set the
			 * corresponding bit(s), else clear it(them).
			 */
			if (*watchdog_param)
				new = old | which;
			else
				new = old & ~which;
		} while (cmpxchg(&watchdog_enabled, old, new) != old);

663 664
		if (old != new)
			proc_watchdog_update();
665 666
	}
out:
667
	mutex_unlock(&watchdog_mutex);
668
	cpu_hotplug_enable();
669 670 671
	return err;
}

672 673 674 675 676 677 678 679 680 681 682 683
/*
 * /proc/sys/kernel/watchdog
 */
int proc_watchdog(struct ctl_table *table, int write,
		  void __user *buffer, size_t *lenp, loff_t *ppos)
{
	return proc_watchdog_common(NMI_WATCHDOG_ENABLED|SOFT_WATCHDOG_ENABLED,
				    table, write, buffer, lenp, ppos);
}

/*
 * /proc/sys/kernel/nmi_watchdog
684
 */
685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700
int proc_nmi_watchdog(struct ctl_table *table, int write,
		      void __user *buffer, size_t *lenp, loff_t *ppos)
{
	return proc_watchdog_common(NMI_WATCHDOG_ENABLED,
				    table, write, buffer, lenp, ppos);
}

/*
 * /proc/sys/kernel/soft_watchdog
 */
int proc_soft_watchdog(struct ctl_table *table, int write,
			void __user *buffer, size_t *lenp, loff_t *ppos)
{
	return proc_watchdog_common(SOFT_WATCHDOG_ENABLED,
				    table, write, buffer, lenp, ppos);
}
701

702 703 704 705 706
/*
 * /proc/sys/kernel/watchdog_thresh
 */
int proc_watchdog_thresh(struct ctl_table *table, int write,
			 void __user *buffer, size_t *lenp, loff_t *ppos)
707
{
708
	int err, old;
709

710
	cpu_hotplug_disable();
711
	mutex_lock(&watchdog_mutex);
712

713
	old = READ_ONCE(watchdog_thresh);
714
	err = proc_dointvec_minmax(table, write, buffer, lenp, ppos);
715

716 717
	if (!err && write && old != READ_ONCE(watchdog_thresh))
		proc_watchdog_update();
718

719
	mutex_unlock(&watchdog_mutex);
720
	cpu_hotplug_enable();
721
	return err;
722
}
723

724 725 726 727
static void proc_watchdog_cpumask_update(void)
{
	/* Remove impossible cpus to keep sysctl output clean. */
	cpumask_and(&watchdog_cpumask, &watchdog_cpumask, cpu_possible_mask);
728 729

	softlockup_update_threads();
730 731 732
	watchdog_nmi_reconfigure();
}

733 734 735 736 737 738 739 740 741 742 743
/*
 * The cpumask is the mask of possible cpus that the watchdog can run
 * on, not the mask of cpus it is actually running on.  This allows the
 * user to specify a mask that will include cpus that have not yet
 * been brought online, if desired.
 */
int proc_watchdog_cpumask(struct ctl_table *table, int write,
			  void __user *buffer, size_t *lenp, loff_t *ppos)
{
	int err;

744
	cpu_hotplug_disable();
745
	mutex_lock(&watchdog_mutex);
746

747
	err = proc_do_large_bitmap(table, write, buffer, lenp, ppos);
748 749
	if (!err && write)
		proc_watchdog_cpumask_update();
750

751
	mutex_unlock(&watchdog_mutex);
752
	cpu_hotplug_enable();
753 754
	return err;
}
755 756
#endif /* CONFIG_SYSCTL */

757
void __init lockup_detector_init(void)
758
{
759 760
#ifdef CONFIG_NO_HZ_FULL
	if (tick_nohz_full_enabled()) {
761 762
		pr_info("Disabling watchdog on nohz_full cores by default\n");
		cpumask_copy(&watchdog_cpumask, housekeeping_mask);
763 764 765 766 767 768
	} else
		cpumask_copy(&watchdog_cpumask, cpu_possible_mask);
#else
	cpumask_copy(&watchdog_cpumask, cpu_possible_mask);
#endif

769
	softlockup_init_threads();
770
}