hrtimer: migration: do not check expiry time on current CPU

The timer migration code needs to check whether the expiry time of the timer is before the programmed clock event expiry time when the timer is enqueued on another CPU because we can not reprogram the timer device on the other CPU. The current logic checks the expiry time even if we enqueue on the current CPU when nohz_get_load_balancer() returns current CPU. This might lead to an endless loop in the expiry check code when the expiry time of the timer is before the current programmed next event. Check whether nohz_get_load_balancer() returns current CPU and skip the expiry check if this is the case. The bug was triggered from the networking code. The patch fixes the regression http://bugzilla.kernel.org/show_bug.cgi?id=13738 (Soft-Lockup/Race in networking in 2.6.31-rc1+195) Cc: Arun Bharadwaj <[email protected] Tested-by: Joao Correia <[email protected]> Tested-by: Andres Freund <[email protected]> Signed-off-by: Thomas Gleixner <[email protected]>
mna · Jul 10, 2009 · 7e0c508 · 7e0c508
1 parent 507e123
commit 7e0c508
Showing 1 changed file with 13 additions and 2 deletions.
diff --git a/kernel/hrtimer.c b/kernel/hrtimer.c
@@ -206,8 +206,19 @@ switch_hrtimer_base(struct hrtimer *timer, struct hrtimer_clock_base *base,
 #if defined(CONFIG_NO_HZ) && defined(CONFIG_SMP)
 	if (!pinned && get_sysctl_timer_migration() && idle_cpu(cpu)) {
 		preferred_cpu = get_nohz_load_balancer();
-		if (preferred_cpu >= 0)
-			cpu = preferred_cpu;
+		if (preferred_cpu >= 0) {
+			/*
+			 * We must not check the expiry value when
+			 * preferred_cpu is the current cpu. If base
+			 * != new_base we would loop forever when the
+			 * timer expires before the current programmed
+			 * next timer event.
+			 */
+			if (preferred_cpu != cpu)
+				cpu = preferred_cpu;
+			else
+				preferred_cpu = -1;
+		}
 	}
 #endif