Improve watchdog config for arch watchdogs

index aa3cd0878270..5e2e57536d98 100644
@@ -12,9 +12,6 @@ extern void touch_softlockup_watchdog_sched(void);
 extern void touch_softlockup_watchdog(void);
 extern void touch_softlockup_watchdog_sync(void);
 extern void touch_all_softlockup_watchdogs(void);
-extern int proc_dowatchdog_thresh(struct ctl_table *table, int write,
-				  void __user *buffer,
-				  size_t *lenp, loff_t *ppos);
 extern unsigned int  softlockup_panic;
 extern unsigned int  hardlockup_panic;
 void lockup_detector_init(void);

index b9caac4fcfd8..107d23705f46 100644
@@ -9,4 +9,6 @@
 
 #include <linux/nmi.h>
 
+extern void arch_touch_nmi_watchdog(void);
+
 #endif

index 633c37083e87..1e714329fe8a 100644
@@ -190,7 +190,7 @@ static int __init init_nmi_wdt(void)
 }
 device_initcall(init_nmi_wdt);
 
-void touch_nmi_watchdog(void)
+void arch_touch_nmi_watchdog(void)
 {
 	atomic_set(&nmi_touched[smp_processor_id()], 1);
 }

index f3671cbbc117..b05627597b1b 100644
@@ -11,4 +11,6 @@
 #ifndef _ASM_NMI_H
 #define _ASM_NMI_H
 
+extern void arch_touch_nmi_watchdog(void);
+
 #endif /* _ASM_NMI_H */

index a2d8e6938d67..0d5641beadf5 100644
@@ -31,7 +31,7 @@ static unsigned int watchdog;
 static unsigned int watchdog_hz = 1;
 unsigned int watchdog_alert_counter[NR_CPUS];
 
-EXPORT_SYMBOL(touch_nmi_watchdog);
+EXPORT_SYMBOL(arch_touch_nmi_watchdog);
 
 /*
  * the best way to detect whether a CPU has a 'hard lockup' problem

index 26ad2b2607c6..284eac3ffaf2 100644
@@ -7,6 +7,7 @@ void nmi_adjust_hz(unsigned int new_hz);
 
 extern atomic_t nmi_active;
 
+void arch_touch_nmi_watchdog(void);
 void start_nmi_watchdog(void *unused);
 void stop_nmi_watchdog(void *unused);
 

index 95e73c63c99d..048ad783ea3f 100644
@@ -51,7 +51,7 @@ static DEFINE_PER_CPU(unsigned int, last_irq_sum);
 static DEFINE_PER_CPU(long, alert_counter);
 static DEFINE_PER_CPU(int, nmi_touch);
 
-void touch_nmi_watchdog(void)
+void arch_touch_nmi_watchdog(void)
 {
 	if (atomic_read(&nmi_active)) {
 		int cpu;

@@ -61,10 +61,8 @@ void touch_nmi_watchdog(void)
 				per_cpu(nmi_touch, cpu) = 1;
 		}
 	}
-
-	touch_softlockup_watchdog();
 }
-EXPORT_SYMBOL(touch_nmi_watchdog);
+EXPORT_SYMBOL(arch_touch_nmi_watchdog);
 
 static void die_nmi(const char *str, struct pt_regs *regs, int do_panic)
 {

index 5e2e57536d98..bd387ef8bccd 100644
@@ -6,6 +6,9 @@
 
 #include <linux/sched.h>
 #include <asm/irq.h>
+#if defined(CONFIG_HAVE_NMI_WATCHDOG)
+#include <asm/nmi.h>
+#endif
 
 #ifdef CONFIG_LOCKUP_DETECTOR
 extern void touch_softlockup_watchdog_sched(void);

@@ -58,6 +61,18 @@ static inline void reset_hung_task_detector(void)
 #define NMI_WATCHDOG_ENABLED      (1 << NMI_WATCHDOG_ENABLED_BIT)
 #define SOFT_WATCHDOG_ENABLED     (1 << SOFT_WATCHDOG_ENABLED_BIT)
 
+#if defined(CONFIG_HARDLOCKUP_DETECTOR)
+extern void hardlockup_detector_disable(void);
+#else
+static inline void hardlockup_detector_disable(void) {}
+#endif
+
+#if defined(CONFIG_HARDLOCKUP_DETECTOR) || defined(CONFIG_HAVE_NMI_WATCHDOG)
+extern void arch_touch_nmi_watchdog(void);
+#else
+static inline void arch_touch_nmi_watchdog(void) {}
+#endif
+
 /**
  * touch_nmi_watchdog - restart NMI watchdog timeout.
  * 

@@ -65,21 +80,11 @@ static inline void reset_hung_task_detector(void)
  * may be used to reset the timeout - for code which intentionally
  * disables interrupts for a long time. This call is stateless.
  */
-#if defined(CONFIG_HAVE_NMI_WATCHDOG) || defined(CONFIG_HARDLOCKUP_DETECTOR)
-#include <asm/nmi.h>
-extern void touch_nmi_watchdog(void);
-#else
 static inline void touch_nmi_watchdog(void)
 {
+	arch_touch_nmi_watchdog();
 	touch_softlockup_watchdog();
 }
-#endif
-
-#if defined(CONFIG_HARDLOCKUP_DETECTOR)
-extern void hardlockup_detector_disable(void);
-#else
-static inline void hardlockup_detector_disable(void) {}
-#endif
 
 /*
  * Create trigger_all_cpu_backtrace() out of the arch-provided

index 54a427d1f344..90d688df6ce1 100644
@@ -56,7 +56,7 @@ static int __init hardlockup_panic_setup(char *str)
 }
 __setup("nmi_watchdog=", hardlockup_panic_setup);
 
-void touch_nmi_watchdog(void)
+void arch_touch_nmi_watchdog(void)
 {
 	/*
 	 * Using __raw here because some code paths have

@@ -66,9 +66,8 @@ void touch_nmi_watchdog(void)
 	 * going off.
 	 */
 	raw_cpu_write(watchdog_nmi_touch, true);
-	touch_softlockup_watchdog();
 }
-EXPORT_SYMBOL(touch_nmi_watchdog);
+EXPORT_SYMBOL(arch_touch_nmi_watchdog);
 
 static struct perf_event_attr wd_hw_attr = {
 	.type		= PERF_TYPE_HARDWARE,

index f35ff9dea4fb..ab650905f75a 100644
@@ -727,7 +727,7 @@ struct ppc_pci_io ppc_pci_io;
 EXPORT_SYMBOL(ppc_pci_io);
 #endif
 
-#ifdef CONFIG_HARDLOCKUP_DETECTOR
+#ifdef CONFIG_HARDLOCKUP_DETECTOR_PERF
 u64 hw_nmi_get_sample_period(int watchdog_thresh)
 {
 	return ppc_proc_freq * watchdog_thresh;

index c73c9fb281e1..d6f387780849 100644
@@ -19,7 +19,7 @@
 #include <linux/init.h>
 #include <linux/delay.h>
 
-#ifdef CONFIG_HARDLOCKUP_DETECTOR
+#ifdef CONFIG_HARDLOCKUP_DETECTOR_PERF
 u64 hw_nmi_get_sample_period(int watchdog_thresh)
 {
 	return (u64)(cpu_khz) * 1000 * watchdog_thresh;

index bd387ef8bccd..8aa01fd859fb 100644
@@ -11,13 +11,21 @@
 #endif
 
 #ifdef CONFIG_LOCKUP_DETECTOR
+void lockup_detector_init(void);
+#else
+static inline void lockup_detector_init(void)
+{
+}
+#endif
+
+#ifdef CONFIG_SOFTLOCKUP_DETECTOR
 extern void touch_softlockup_watchdog_sched(void);
 extern void touch_softlockup_watchdog(void);
 extern void touch_softlockup_watchdog_sync(void);
 extern void touch_all_softlockup_watchdogs(void);
 extern unsigned int  softlockup_panic;
-extern unsigned int  hardlockup_panic;
-void lockup_detector_init(void);
+extern int soft_watchdog_enabled;
+extern atomic_t watchdog_park_in_progress;
 #else
 static inline void touch_softlockup_watchdog_sched(void)
 {

@@ -31,9 +39,6 @@ static inline void touch_softlockup_watchdog_sync(void)
 static inline void touch_all_softlockup_watchdogs(void)
 {
 }
-static inline void lockup_detector_init(void)
-{
-}
 #endif
 
 #ifdef CONFIG_DETECT_HUNG_TASK

@@ -63,15 +68,18 @@ static inline void reset_hung_task_detector(void)
 
 #if defined(CONFIG_HARDLOCKUP_DETECTOR)
 extern void hardlockup_detector_disable(void);
+extern unsigned int hardlockup_panic;
 #else
 static inline void hardlockup_detector_disable(void) {}
 #endif
 
-#if defined(CONFIG_HARDLOCKUP_DETECTOR) || defined(CONFIG_HAVE_NMI_WATCHDOG)
+#if defined(CONFIG_HARDLOCKUP_DETECTOR_PERF)
 extern void arch_touch_nmi_watchdog(void);
 #else
+#if !defined(CONFIG_HAVE_NMI_WATCHDOG)
 static inline void arch_touch_nmi_watchdog(void) {}
 #endif
+#endif
 
 /**
  * touch_nmi_watchdog - restart NMI watchdog timeout.

@@ -141,15 +149,18 @@ static inline bool trigger_single_cpu_backtrace(int cpu)
 }
 #endif
 
-#ifdef CONFIG_LOCKUP_DETECTOR
+#ifdef CONFIG_HARDLOCKUP_DETECTOR_PERF
 u64 hw_nmi_get_sample_period(int watchdog_thresh);
+#endif
+
+#ifdef CONFIG_LOCKUP_DETECTOR
 extern int nmi_watchdog_enabled;
-extern int soft_watchdog_enabled;
 extern int watchdog_user_enabled;
 extern int watchdog_thresh;
 extern unsigned long watchdog_enabled;
+extern struct cpumask watchdog_cpumask;
 extern unsigned long *watchdog_cpumask_bits;
-extern atomic_t watchdog_park_in_progress;
+extern int __read_mostly watchdog_suspended;
 #ifdef CONFIG_SMP
 extern int sysctl_softlockup_all_cpu_backtrace;
 extern int sysctl_hardlockup_all_cpu_backtrace;

index 4dfba1a76cc3..655686d546cb 100644
@@ -880,6 +880,14 @@ static struct ctl_table kern_table[] = {
 #endif
 	},
 	{
+		.procname	= "watchdog_cpumask",
+		.data		= &watchdog_cpumask_bits,
+		.maxlen		= NR_CPUS,
+		.mode		= 0644,
+		.proc_handler	= proc_watchdog_cpumask,
+	},
+#ifdef CONFIG_SOFTLOCKUP_DETECTOR
+	{
 		.procname       = "soft_watchdog",
 		.data           = &soft_watchdog_enabled,
 		.maxlen         = sizeof (int),

@@ -889,13 +897,6 @@ static struct ctl_table kern_table[] = {
 		.extra2		= &one,
 	},
 	{
-		.procname	= "watchdog_cpumask",
-		.data		= &watchdog_cpumask_bits,
-		.maxlen		= NR_CPUS,
-		.mode		= 0644,
-		.proc_handler	= proc_watchdog_cpumask,
-	},
-	{
 		.procname	= "softlockup_panic",
 		.data		= &softlockup_panic,
 		.maxlen		= sizeof(int),

@@ -904,27 +905,29 @@ static struct ctl_table kern_table[] = {
 		.extra1		= &zero,
 		.extra2		= &one,
 	},
-#ifdef CONFIG_HARDLOCKUP_DETECTOR
+#ifdef CONFIG_SMP
 	{
-		.procname	= "hardlockup_panic",
-		.data		= &hardlockup_panic,
+		.procname	= "softlockup_all_cpu_backtrace",
+		.data		= &sysctl_softlockup_all_cpu_backtrace,
 		.maxlen		= sizeof(int),
 		.mode		= 0644,
 		.proc_handler	= proc_dointvec_minmax,
 		.extra1		= &zero,
 		.extra2		= &one,
 	},
+#endif /* CONFIG_SMP */
 #endif
-#ifdef CONFIG_SMP
+#ifdef CONFIG_HARDLOCKUP_DETECTOR
 	{
-		.procname	= "softlockup_all_cpu_backtrace",
-		.data		= &sysctl_softlockup_all_cpu_backtrace,
+		.procname	= "hardlockup_panic",
+		.data		= &hardlockup_panic,
 		.maxlen		= sizeof(int),
 		.mode		= 0644,
 		.proc_handler	= proc_dointvec_minmax,
 		.extra1		= &zero,
 		.extra2		= &one,
 	},
+#ifdef CONFIG_SMP
 	{
 		.procname	= "hardlockup_all_cpu_backtrace",
 		.data		= &sysctl_hardlockup_all_cpu_backtrace,

@@ -936,6 +939,8 @@ static struct ctl_table kern_table[] = {
 	},
 #endif /* CONFIG_SMP */
 #endif
+#endif
+
 #if defined(CONFIG_X86_LOCAL_APIC) && defined(CONFIG_X86)
 	{
 		.procname       = "unknown_nmi_panic",

318 lines not on this page · raw message

index 90d688df6ce1..295a0d84934c 100644
@@ -22,39 +22,7 @@ static DEFINE_PER_CPU(bool, hard_watchdog_warn);
 static DEFINE_PER_CPU(bool, watchdog_nmi_touch);
 static DEFINE_PER_CPU(struct perf_event *, watchdog_ev);
 
-/* boot commands */
-/*
- * Should we panic when a soft-lockup or hard-lockup occurs:
- */
-unsigned int __read_mostly hardlockup_panic =
-			CONFIG_BOOTPARAM_HARDLOCKUP_PANIC_VALUE;
 static unsigned long hardlockup_allcpu_dumped;
-/*
- * 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().
- */
-void hardlockup_detector_disable(void)
-{
-	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);
 
 void arch_touch_nmi_watchdog(void)
 {

index 1fba9c3d66dc..06cd965f64d2 100644
@@ -123,6 +123,11 @@ void __weak watchdog_nmi_disable(unsigned int cpu)
 {
 }
 
+void __weak watchdog_nmi_reconfigure(void)
+{
+}
+
+
 #ifdef CONFIG_SOFTLOCKUP_DETECTOR
 
 /* Helper for online, unparked cpus. */

@@ -600,6 +605,12 @@ static void watchdog_disable_all_cpus(void)
 	}
 }
 
+static int watchdog_update_cpus(void)
+{
+	return smpboot_update_cpumask_percpu_thread(
+		    &watchdog_threads, &watchdog_cpumask);
+}
+
 #else /* SOFTLOCKUP */
 static int watchdog_park_threads(void)
 {

@@ -619,6 +630,11 @@ static void watchdog_disable_all_cpus(void)
 {
 }
 
+static int watchdog_update_cpus(void)
+{
+	return 0;
+}
+
 static void set_sample_period(void)
 {
 }

@@ -651,6 +667,8 @@ int lockup_detector_suspend(void)
 		watchdog_enabled = 0;
 	}
 
+	watchdog_nmi_reconfigure();
+
 	mutex_unlock(&watchdog_proc_mutex);
 
 	return ret;

@@ -671,6 +689,8 @@ void lockup_detector_resume(void)
 	if (watchdog_running && !watchdog_suspended)
 		watchdog_unpark_threads();
 
+	watchdog_nmi_reconfigure();
+
 	mutex_unlock(&watchdog_proc_mutex);
 	put_online_cpus();
 }

@@ -696,6 +716,8 @@ static int proc_watchdog_update(void)
 	else
 		watchdog_disable_all_cpus();
 
+	watchdog_nmi_reconfigure();
+
 	return err;
 
 }

@@ -881,12 +903,11 @@ int proc_watchdog_cpumask(struct ctl_table *table, int write,
 			 * a temporary cpumask, so we are likely not in a
 			 * position to do much else to make things better.
 			 */
-#ifdef CONFIG_SOFTLOCKUP_DETECTOR
-			if (smpboot_update_cpumask_percpu_thread(
-				    &watchdog_threads, &watchdog_cpumask) != 0)
+			if (watchdog_update_cpus() != 0)
 				pr_err("cpumask update failed\n");
-#endif
 		}
+
+		watchdog_nmi_reconfigure();
 	}
 out:
 	mutex_unlock(&watchdog_proc_mutex);

index 06cd965f64d2..36531025496f 100644
@@ -114,6 +114,10 @@ int __read_mostly watchdog_suspended;
 /*
  * These functions can be overridden if an architecture implements its
  * own hardlockup detector.
+ *
+ * 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.
  */
 int __weak watchdog_nmi_enable(unsigned int cpu)
 {

@@ -123,6 +127,17 @@ void __weak watchdog_nmi_disable(unsigned int cpu)
 {
 }
 
+/*
+ * 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
+ * - watchdog_suspended
+ */
 void __weak watchdog_nmi_reconfigure(void)
 {
 }

index ff1ccb375e60..6f8e79cd35d8 100644
@@ -1,4 +1,15 @@
 #ifndef _ASM_NMI_H
 #define _ASM_NMI_H
 
+#ifdef CONFIG_HARDLOCKUP_DETECTOR
+extern void arch_touch_nmi_watchdog(void);
+
+extern void arch_trigger_cpumask_backtrace(const cpumask_t *mask,
+					   bool exclude_self);
+#define arch_trigger_cpumask_backtrace arch_trigger_cpumask_backtrace
+
+#else
+static inline void arch_touch_nmi_watchdog(void) {}
+#endif
+
 #endif /* _ASM_NMI_H */

index ebddb2111d87..8ea98504f900 100644
@@ -55,6 +55,8 @@ struct smp_ops_t {
 	int   (*cpu_bootable)(unsigned int nr);
 };
 
+extern void smp_flush_nmi_ipi(u64 delay_us);
+extern int smp_send_nmi_ipi(int cpu, void (*fn)(struct pt_regs *), u64 delay_us);
 extern void smp_send_debugger_break(void);
 extern void start_secondary_resume(void);
 extern void smp_generic_give_timebase(void);

index 9ad37f827a97..1086ea37c832 100644
@@ -25,6 +25,7 @@
 #include <linux/kvm_para.h>
 #include <linux/slab.h>
 #include <linux/of.h>
+#include <linux/nmi.h> /* hardlockup_detector_disable() */
 
 #include <asm/reg.h>
 #include <asm/sections.h>

@@ -718,6 +719,12 @@ static __init void kvm_free_tmp(void)
 
 static int __init kvm_guest_init(void)
 {
+	/*
+	 * The hardlockup detector is likely to get false positives in
+	 * KVM guests, so disable it by default.
+	 */
+	hardlockup_detector_disable();
+
 	if (!kvm_para_available())
 		goto free_tmp;
 

index ab650905f75a..9c15b3b8ba4d 100644
@@ -726,22 +726,3 @@ unsigned long memory_block_size_bytes(void)
 struct ppc_pci_io ppc_pci_io;
 EXPORT_SYMBOL(ppc_pci_io);
 #endif
-
-#ifdef CONFIG_HARDLOCKUP_DETECTOR_PERF
-u64 hw_nmi_get_sample_period(int watchdog_thresh)
-{
-	return ppc_proc_freq * watchdog_thresh;
-}
-
-/*
- * The hardlockup detector breaks PMU event based branches and is likely
- * to get false positives in KVM guests, so disable it by default.
- */
-static int __init disable_hardlockup_detector(void)
-{
-	hardlockup_detector_disable();
-
-	return 0;
-}
-early_initcall(disable_hardlockup_detector);
-#endif

index df2a41647d8e..5b602a284084 100644
@@ -433,13 +433,31 @@ static void do_smp_send_nmi_ipi(int cpu)
 	}
 }
 
+void smp_flush_nmi_ipi(u64 delay_us)
+{
+	unsigned long flags;
+
+	nmi_ipi_lock_start(&flags);
+	while (nmi_ipi_busy_count) {
+		nmi_ipi_unlock_end(&flags);
+		udelay(1);
+		if (delay_us) {
+			delay_us--;
+			if (!delay_us)
+				return;
+		}
+		nmi_ipi_lock_start(&flags);
+	}
+	nmi_ipi_unlock_end(&flags);
+}
+
 /*
  * - cpu is the target CPU (must not be this CPU), or NMI_IPI_ALL_OTHERS.
  * - fn is the target callback function.
  * - delay_us > 0 is the delay before giving up waiting for targets to
  *   enter the handler, == 0 specifies indefinite delay.
  */
-static int smp_send_nmi_ipi(int cpu, void (*fn)(struct pt_regs *), u64 delay_us)
+int smp_send_nmi_ipi(int cpu, void (*fn)(struct pt_regs *), u64 delay_us)
 {
 	unsigned long flags;
 	int me = raw_smp_processor_id();

388 lines not on this page · raw message