mm: clear N_CPU from node_states at CPU offline
authorToshi Kani <toshi.kani@hp.com>
Tue, 12 Nov 2013 23:08:13 +0000 (15:08 -0800)
committerLinus Torvalds <torvalds@linux-foundation.org>
Wed, 13 Nov 2013 03:09:09 +0000 (12:09 +0900)
vmstat_cpuup_callback() is a CPU notifier callback, which marks N_CPU to a
node at CPU online event.  However, it does not update this N_CPU info at
CPU offline event.

Changed vmstat_cpuup_callback() to clear N_CPU when the last CPU in the
node is put into offline, i.e.  the node no longer has any online CPU.

Signed-off-by: Toshi Kani <toshi.kani@hp.com>
Acked-by: Christoph Lameter <cl@linux.com>
Reviewed-by: Yasuaki Ishimatsu <isimatu.yasuaki@jp.fujitsu.com>
Tested-by: Yasuaki Ishimatsu <isimatu.yasuaki@jp.fujitsu.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
mm/vmstat.c

index 0a1f7de..b6d17ed 100644 (file)
@@ -1229,6 +1229,20 @@ static void start_cpu_timer(int cpu)
        schedule_delayed_work_on(cpu, work, __round_jiffies_relative(HZ, cpu));
 }
 
+static void vmstat_cpu_dead(int node)
+{
+       int cpu;
+
+       get_online_cpus();
+       for_each_online_cpu(cpu)
+               if (cpu_to_node(cpu) == node)
+                       goto end;
+
+       node_clear_state(node, N_CPU);
+end:
+       put_online_cpus();
+}
+
 /*
  * Use the cpu notifier to insure that the thresholds are recalculated
  * when necessary.
@@ -1258,6 +1272,7 @@ static int vmstat_cpuup_callback(struct notifier_block *nfb,
        case CPU_DEAD:
        case CPU_DEAD_FROZEN:
                refresh_zone_stat_thresholds();
+               vmstat_cpu_dead(cpu_to_node(cpu));
                break;
        default:
                break;