octeon, watchdog: Fix CPU hotplug callback registration
authorSrivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>
Mon, 10 Mar 2014 20:40:43 +0000 (02:10 +0530)
committerRafael J. Wysocki <rafael.j.wysocki@intel.com>
Thu, 20 Mar 2014 12:43:46 +0000 (13:43 +0100)
commit99c3bf361a4afd2ad331e767b3e5e6fa3488a096
tree5feb67adb0a1cda97bd5f09d3ca6d2af20b412aa
parent180d86463257812dc17e5df912f3bddcc96abb00
octeon, watchdog: Fix CPU hotplug callback registration

Subsystems that want to register CPU hotplug callbacks, as well as perform
initialization for the CPUs that are already online, often do it as shown
below:

get_online_cpus();

for_each_online_cpu(cpu)
init_cpu(cpu);

register_cpu_notifier(&foobar_cpu_notifier);

put_online_cpus();

This is wrong, since it is prone to ABBA deadlocks involving the
cpu_add_remove_lock and the cpu_hotplug.lock (when running concurrently
with CPU hotplug operations).

Instead, the correct and race-free way of performing the callback
registration is:

cpu_notifier_register_begin();

for_each_online_cpu(cpu)
init_cpu(cpu);

/* Note the use of the double underscored version of the API */
__register_cpu_notifier(&foobar_cpu_notifier);

cpu_notifier_register_done();

Fix the watchdog code in octeon by using this latter form of callback
registration.

Cc: Wim Van Sebroeck <wim@iguana.be>
Cc: Ingo Molnar <mingo@kernel.org>
Signed-off-by: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
drivers/watchdog/octeon-wdt-main.c
This page took 0.025576 seconds and 5 git commands to generate.