alarmtimer: Prevent overflow of relative timers

Andrey reported a alartimer related RCU stall while fuzzing the kernel with
syzkaller.

The reason for this is an overflow in ktime_add() which brings the
resulting time into negative space and causes immediate expiry of the
timer. The following rearm with a small interval does not bring the timer
back into positive space due to the same issue.

This results in a permanent firing alarmtimer which hogs the CPU.

Use ktime_add_safe() instead which detects the overflow and clamps the
result to KTIME_SEC_MAX.

Reported-by: Andrey Konovalov <andreyknvl@google.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Kostya Serebryany <kcc@google.com>
Cc: syzkaller <syzkaller@googlegroups.com>
Cc: John Stultz <john.stultz@linaro.org>
Cc: Dmitry Vyukov <dvyukov@google.com>
Cc: stable@vger.kernel.org
Link: http://lkml.kernel.org/r/20170530211655.802921648@linutronix.de

kernel/time/alarmtimer.c

@@ -387,7 +387,7 @@ void alarm_start_relative(struct alarm *alarm, ktime_t start)
 {
 	struct alarm_base *base = &alarm_bases[alarm->type];
 
-	start = ktime_add(start, base->gettime());
+	start = ktime_add_safe(start, base->gettime());
 	alarm_start(alarm, start);
 }
 EXPORT_SYMBOL_GPL(alarm_start_relative);
@@ -475,7 +475,7 @@ u64 alarm_forward(struct alarm *alarm, ktime_t now, ktime_t interval)
 		overrun++;
 	}
 
-	alarm->node.expires = ktime_add(alarm->node.expires, interval);
+	alarm->node.expires = ktime_add_safe(alarm->node.expires, interval);
 	return overrun;
 }
 EXPORT_SYMBOL_GPL(alarm_forward);
@@ -666,7 +666,7 @@ static int alarm_timer_set(struct k_itimer *timr, int flags,
 		ktime_t now;
 
 		now = alarm_bases[timr->it.alarm.alarmtimer.type].gettime();
-		exp = ktime_add(now, exp);
+		exp = ktime_add_safe(now, exp);
 	}
 
 	alarm_start(&timr->it.alarm.alarmtimer, exp);