Merge git://git.kernel.org/pub/scm/linux/kernel/git/sam/kbuild
[sfrench/cifs-2.6.git] / kernel / softlockup.c
index 0131e296ffb416435f7c54a935657c624a09d0c5..11df812263c8fd3eed16f4f1a57d36a68f6eaedc 100644 (file)
 #include <linux/cpu.h>
 #include <linux/init.h>
 #include <linux/delay.h>
+#include <linux/freezer.h>
 #include <linux/kthread.h>
 #include <linux/notifier.h>
 #include <linux/module.h>
 
+#include <asm/irq_regs.h>
+
 static DEFINE_SPINLOCK(print_lock);
 
 static DEFINE_PER_CPU(unsigned long, touch_timestamp);
 static DEFINE_PER_CPU(unsigned long, print_timestamp);
 static DEFINE_PER_CPU(struct task_struct *, watchdog_task);
 
-static int did_panic = 0;
+static int did_panic;
+int softlockup_thresh = 10;
 
 static int
 softlock_panic(struct notifier_block *this, unsigned long event, void *ptr)
@@ -39,14 +43,16 @@ static struct notifier_block panic_block = {
  * resolution, and we don't need to waste time with a big divide when
  * 2^30ns == 1.074s.
  */
-static unsigned long get_timestamp(void)
+static unsigned long get_timestamp(int this_cpu)
 {
-       return sched_clock() >> 30;  /* 2^30 ~= 10^9 */
+       return cpu_clock(this_cpu) >> 30;  /* 2^30 ~= 10^9 */
 }
 
 void touch_softlockup_watchdog(void)
 {
-       __raw_get_cpu_var(touch_timestamp) = get_timestamp();
+       int this_cpu = raw_smp_processor_id();
+
+       __raw_get_cpu_var(touch_timestamp) = get_timestamp(this_cpu);
 }
 EXPORT_SYMBOL(touch_softlockup_watchdog);
 
@@ -69,6 +75,7 @@ void softlockup_tick(void)
        int this_cpu = smp_processor_id();
        unsigned long touch_timestamp = per_cpu(touch_timestamp, this_cpu);
        unsigned long print_timestamp;
+       struct pt_regs *regs = get_irq_regs();
        unsigned long now;
 
        if (touch_timestamp == 0) {
@@ -79,10 +86,11 @@ void softlockup_tick(void)
        print_timestamp = per_cpu(print_timestamp, this_cpu);
 
        /* report at most once a second */
-       if (print_timestamp < (touch_timestamp + 1) ||
-               did_panic ||
-                       !per_cpu(watchdog_task, this_cpu))
+       if ((print_timestamp >= touch_timestamp &&
+                       print_timestamp < (touch_timestamp + 1)) ||
+                       did_panic || !per_cpu(watchdog_task, this_cpu)) {
                return;
+       }
 
        /* do not print during early bootup: */
        if (unlikely(system_state != SYSTEM_RUNNING)) {
@@ -90,33 +98,37 @@ void softlockup_tick(void)
                return;
        }
 
-       now = get_timestamp();
+       now = get_timestamp(this_cpu);
 
        /* Wake up the high-prio watchdog task every second: */
        if (now > (touch_timestamp + 1))
                wake_up_process(per_cpu(watchdog_task, this_cpu));
 
        /* Warn about unreasonable 10+ seconds delays: */
-       if (now > (touch_timestamp + 10)) {
-               per_cpu(print_timestamp, this_cpu) = touch_timestamp;
+       if (now <= (touch_timestamp + softlockup_thresh))
+               return;
 
-               spin_lock(&print_lock);
-               printk(KERN_ERR "BUG: soft lockup detected on CPU#%d!\n",
-                       this_cpu);
+       per_cpu(print_timestamp, this_cpu) = touch_timestamp;
+
+       spin_lock(&print_lock);
+       printk(KERN_ERR "BUG: soft lockup - CPU#%d stuck for %lus! [%s:%d]\n",
+                       this_cpu, now - touch_timestamp,
+                       current->comm, task_pid_nr(current));
+       if (regs)
+               show_regs(regs);
+       else
                dump_stack();
-               spin_unlock(&print_lock);
-       }
+       spin_unlock(&print_lock);
 }
 
 /*
  * The watchdog thread - runs every second and touches the timestamp.
  */
-static int watchdog(void * __bind_cpu)
+static int watchdog(void *__bind_cpu)
 {
        struct sched_param param = { .sched_priority = MAX_RT_PRIO-1 };
 
        sched_setscheduler(current, SCHED_FIFO, &param);
-       current->flags |= PF_NOFREEZE;
 
        /* initialize timestamp */
        touch_softlockup_watchdog();
@@ -150,13 +162,13 @@ cpu_callback(struct notifier_block *nfb, unsigned long action, void *hcpu)
                BUG_ON(per_cpu(watchdog_task, hotcpu));
                p = kthread_create(watchdog, hcpu, "watchdog/%d", hotcpu);
                if (IS_ERR(p)) {
-                       printk("watchdog for %i failed\n", hotcpu);
+                       printk(KERN_ERR "watchdog for %i failed\n", hotcpu);
                        return NOTIFY_BAD;
                }
-               per_cpu(touch_timestamp, hotcpu) = 0;
-               per_cpu(watchdog_task, hotcpu) = p;
+               per_cpu(touch_timestamp, hotcpu) = 0;
+               per_cpu(watchdog_task, hotcpu) = p;
                kthread_bind(p, hotcpu);
-               break;
+               break;
        case CPU_ONLINE:
        case CPU_ONLINE_FROZEN:
                wake_up_process(per_cpu(watchdog_task, hotcpu));
@@ -176,7 +188,7 @@ cpu_callback(struct notifier_block *nfb, unsigned long action, void *hcpu)
                kthread_stop(p);
                break;
 #endif /* CONFIG_HOTPLUG_CPU */
-       }
+       }
        return NOTIFY_OK;
 }