|
@@ -34,7 +34,6 @@ unsigned long __read_mostly sysctl_hung_task_check_count = PID_MAX_LIMIT;
|
|
* Zero means infinite timeout - no checking done:
|
|
* Zero means infinite timeout - no checking done:
|
|
*/
|
|
*/
|
|
unsigned long __read_mostly sysctl_hung_task_timeout_secs = 120;
|
|
unsigned long __read_mostly sysctl_hung_task_timeout_secs = 120;
|
|
-static unsigned long __read_mostly hung_task_poll_jiffies;
|
|
|
|
|
|
|
|
unsigned long __read_mostly sysctl_hung_task_warnings = 10;
|
|
unsigned long __read_mostly sysctl_hung_task_warnings = 10;
|
|
|
|
|
|
@@ -69,33 +68,17 @@ static struct notifier_block panic_block = {
|
|
.notifier_call = hung_task_panic,
|
|
.notifier_call = hung_task_panic,
|
|
};
|
|
};
|
|
|
|
|
|
-/*
|
|
|
|
- * 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.
|
|
|
|
- */
|
|
|
|
-static unsigned long get_timestamp(void)
|
|
|
|
-{
|
|
|
|
- int this_cpu = raw_smp_processor_id();
|
|
|
|
-
|
|
|
|
- return cpu_clock(this_cpu) >> 30LL; /* 2^30 ~= 10^9 */
|
|
|
|
-}
|
|
|
|
-
|
|
|
|
-static void check_hung_task(struct task_struct *t, unsigned long now,
|
|
|
|
- unsigned long timeout)
|
|
|
|
|
|
+static void check_hung_task(struct task_struct *t, unsigned long timeout)
|
|
{
|
|
{
|
|
unsigned long switch_count = t->nvcsw + t->nivcsw;
|
|
unsigned long switch_count = t->nvcsw + t->nivcsw;
|
|
|
|
|
|
if (t->flags & PF_FROZEN)
|
|
if (t->flags & PF_FROZEN)
|
|
return;
|
|
return;
|
|
|
|
|
|
- if (switch_count != t->last_switch_count || !t->last_switch_timestamp) {
|
|
|
|
|
|
+ if (switch_count != t->last_switch_count) {
|
|
t->last_switch_count = switch_count;
|
|
t->last_switch_count = switch_count;
|
|
- t->last_switch_timestamp = now;
|
|
|
|
return;
|
|
return;
|
|
}
|
|
}
|
|
- if ((long)(now - t->last_switch_timestamp) < timeout)
|
|
|
|
- return;
|
|
|
|
if (!sysctl_hung_task_warnings)
|
|
if (!sysctl_hung_task_warnings)
|
|
return;
|
|
return;
|
|
sysctl_hung_task_warnings--;
|
|
sysctl_hung_task_warnings--;
|
|
@@ -111,7 +94,6 @@ static void check_hung_task(struct task_struct *t, unsigned long now,
|
|
sched_show_task(t);
|
|
sched_show_task(t);
|
|
__debug_show_held_locks(t);
|
|
__debug_show_held_locks(t);
|
|
|
|
|
|
- t->last_switch_timestamp = now;
|
|
|
|
touch_nmi_watchdog();
|
|
touch_nmi_watchdog();
|
|
|
|
|
|
if (sysctl_hung_task_panic)
|
|
if (sysctl_hung_task_panic)
|
|
@@ -145,7 +127,6 @@ static void check_hung_uninterruptible_tasks(unsigned long timeout)
|
|
{
|
|
{
|
|
int max_count = sysctl_hung_task_check_count;
|
|
int max_count = sysctl_hung_task_check_count;
|
|
int batch_count = HUNG_TASK_BATCHING;
|
|
int batch_count = HUNG_TASK_BATCHING;
|
|
- unsigned long now = get_timestamp();
|
|
|
|
struct task_struct *g, *t;
|
|
struct task_struct *g, *t;
|
|
|
|
|
|
/*
|
|
/*
|
|
@@ -168,19 +149,16 @@ static void check_hung_uninterruptible_tasks(unsigned long timeout)
|
|
}
|
|
}
|
|
/* use "==" to skip the TASK_KILLABLE tasks waiting on NFS */
|
|
/* use "==" to skip the TASK_KILLABLE tasks waiting on NFS */
|
|
if (t->state == TASK_UNINTERRUPTIBLE)
|
|
if (t->state == TASK_UNINTERRUPTIBLE)
|
|
- check_hung_task(t, now, timeout);
|
|
|
|
|
|
+ check_hung_task(t, timeout);
|
|
} while_each_thread(g, t);
|
|
} while_each_thread(g, t);
|
|
unlock:
|
|
unlock:
|
|
rcu_read_unlock();
|
|
rcu_read_unlock();
|
|
}
|
|
}
|
|
|
|
|
|
-static void update_poll_jiffies(void)
|
|
|
|
|
|
+static unsigned long timeout_jiffies(unsigned long timeout)
|
|
{
|
|
{
|
|
/* timeout of 0 will disable the watchdog */
|
|
/* timeout of 0 will disable the watchdog */
|
|
- if (sysctl_hung_task_timeout_secs == 0)
|
|
|
|
- hung_task_poll_jiffies = MAX_SCHEDULE_TIMEOUT;
|
|
|
|
- else
|
|
|
|
- hung_task_poll_jiffies = sysctl_hung_task_timeout_secs * HZ / 2;
|
|
|
|
|
|
+ return timeout ? timeout * HZ : MAX_SCHEDULE_TIMEOUT;
|
|
}
|
|
}
|
|
|
|
|
|
/*
|
|
/*
|
|
@@ -197,8 +175,6 @@ int proc_dohung_task_timeout_secs(struct ctl_table *table, int write,
|
|
if (ret || !write)
|
|
if (ret || !write)
|
|
goto out;
|
|
goto out;
|
|
|
|
|
|
- update_poll_jiffies();
|
|
|
|
-
|
|
|
|
wake_up_process(watchdog_task);
|
|
wake_up_process(watchdog_task);
|
|
|
|
|
|
out:
|
|
out:
|
|
@@ -211,20 +187,14 @@ int proc_dohung_task_timeout_secs(struct ctl_table *table, int write,
|
|
static int watchdog(void *dummy)
|
|
static int watchdog(void *dummy)
|
|
{
|
|
{
|
|
set_user_nice(current, 0);
|
|
set_user_nice(current, 0);
|
|
- update_poll_jiffies();
|
|
|
|
|
|
|
|
for ( ; ; ) {
|
|
for ( ; ; ) {
|
|
- unsigned long timeout;
|
|
|
|
|
|
+ unsigned long timeout = sysctl_hung_task_timeout_secs;
|
|
|
|
|
|
- while (schedule_timeout_interruptible(hung_task_poll_jiffies));
|
|
|
|
|
|
+ while (schedule_timeout_interruptible(timeout_jiffies(timeout)))
|
|
|
|
+ timeout = sysctl_hung_task_timeout_secs;
|
|
|
|
|
|
- /*
|
|
|
|
- * Need to cache timeout here to avoid timeout being set
|
|
|
|
- * to 0 via sysctl while inside check_hung_*_tasks().
|
|
|
|
- */
|
|
|
|
- timeout = sysctl_hung_task_timeout_secs;
|
|
|
|
- if (timeout)
|
|
|
|
- check_hung_uninterruptible_tasks(timeout);
|
|
|
|
|
|
+ check_hung_uninterruptible_tasks(timeout);
|
|
}
|
|
}
|
|
|
|
|
|
return 0;
|
|
return 0;
|