Use get_online_cpus to avoid races involving CPU hotplug

PATCH Applied Silas Boyd-Wickizer 2 participants 2 messages Aug 3, 2012 – Aug 4, 2012

Subsystems (2)

Hardware monitoring Coretemp hardware monitoring driver

lore

Share & tools

lore

Silas Boyd-Wickizer

author

2012-08-03 19:40 UTC #0 lore raw

PATCH v2 4/4 Use get_online_cpus to avoid races involving CPU hotplug

coretemp_init in drivers/hwmon/coretemp.c loops with for_each_online_cpu, adding platform_devices and sysfs interfaces, then calls register_hotcpu_notifier. There is a race if a CPU is offlined or onlined after the loop, but before register_hotcpu_notifier. The race might result in the absence of a platform_device+sysfs interface for an online CPU, or the presence of a platform_device+sysfs interface for an offline CPU. A similar race occurs during coretemp_exit, after the module calls unregister_hotcpu_notifier, but before it unregisters all devices, a CPU might offline and a device for an offline CPU will exist for a short while.

This fix surrounds for_each_online_cpu and register_hotcpu_notifier with get_online_cpus+put_online_cpus; and surrounds unregister_hotcpu_notifier and device unregistering with get_online_cpus+put_online_cpus.

Build tested.

Signed-off-by Silas Boyd-Wickizer <[email protected]>

---
 drivers/hwmon/coretemp.c |    5 +++++
 1 file changed, 5 insertions(+)

drivers/hwmon/coretemp.c +5 -0

index 637c51c..9483e90 100644
@@ -815,17 +815,20 @@ static int __init coretemp_init(void)
 	if (err)
 		goto exit;
 
+	get_online_cpus();
 	for_each_online_cpu(i)
 		get_core_online(i);
 
 #ifndef CONFIG_HOTPLUG_CPU
 	if (list_empty(&pdev_list)) {
+		put_online_cpus();
 		err = -ENODEV;
 		goto exit_driver_unreg;
 	}
 #endif
 
 	register_hotcpu_notifier(&coretemp_cpu_notifier);
+	put_online_cpus();
 	return 0;
 
 #ifndef CONFIG_HOTPLUG_CPU

@@ -840,6 +843,7 @@ static void __exit coretemp_exit(void)
 {
 	struct pdev_entry *p, *n;
 
+	get_online_cpus();
 	unregister_hotcpu_notifier(&coretemp_cpu_notifier);
 	mutex_lock(&pdev_list_mutex);
 	list_for_each_entry_safe(p, n, &pdev_list, list) {

@@ -848,6 +852,7 @@ static void __exit coretemp_exit(void)
 		kfree(p);
 	}
 	mutex_unlock(&pdev_list_mutex);
+	put_online_cpus();
 	platform_driver_unregister(&coretemp_driver);
 }

Signature · 1 line

1.7.10.4

Jean Delvare

Applied

2012-08-04 08:52 UTC #1 lore raw

Quoted 17 lines · Silas Boyd-Wickizer

coretemp_init in drivers/hwmon/coretemp.c loops with for_each_online_cpu, adding platform_devices and sysfs interfaces, then calls register_hotcpu_notifier. There is a race if a CPU is offlined or onlined after the loop, but before register_hotcpu_notifier. The race might result in the absence of a platform_device+sysfs interface for an online CPU, or the presence of a platform_device+sysfs interface for an offline CPU. A similar race occurs during coretemp_exit, after the module calls unregister_hotcpu_notifier, but before it unregisters all devices, a CPU might offline and a device for an offline CPU will exist for a short while.
This fix surrounds for_each_online_cpu and register_hotcpu_notifier with get_online_cpus+put_online_cpus; and surrounds unregister_hotcpu_notifier and device unregistering with get_online_cpus+put_online_cpus.

Quoted 7 lines

Build tested.
Signed-off-by Silas Boyd-Wickizer <[email protected]>
---
 drivers/hwmon/coretemp.c |    5 +++++
 1 file changed, 5 insertions(+)
(...)

Applied, thanks.

Signature · 1 line

Jean Delvare