Re: Exit ACPI drivers immediately if acpi is disabled

[Thomas Renninger <trenn@suse.de>]

On Mon, 2006-04-24 at 08:59 -0600, Bjorn Helgaas wrote:
> On Monday 24 April 2006 04:52, Thomas Renninger wrote:
> > Rethinking about this: driver_registered is a bit too general variable?
> > 
> > Is this OK, now?
> 
> It's static, so I don't care either way.
> 
> The powernow/speedstep/acpi code sharing seems a little strange, but
> I don't know anything about it.  And you're not changing that part
> of things anyway.
> 
> > @@ -943,12 +947,13 @@ static void __exit acpi_processor_exit(v
> >  
> >  	acpi_thermal_cpufreq_exit();
> >  
> > +	if (processor_driver_registered){
> >  	acpi_processor_uninstall_hotplug_notify();
> >  
> >  	acpi_bus_unregister_driver(&acpi_processor_driver);
> >  
> >  	remove_proc_entry(ACPI_PROCESSOR_CLASS, acpi_root_dir);
> > -
> > +	}
> 
> Shouldn't the remove_proc_entry() be outside the "if (processor_driver_registered)"
> block?  The directory was created unconditionally, before the
> acpi_bus_register_driver() call.
That should be OK as it is conditionally created in the way that init
func returns,
if proc_mkdir returns NULL:

init():
	acpi_processor_dir = proc_mkdir(ACPI_PROCESSOR_CLASS, acpi_root_dir);
	if (!acpi_processor_dir)
/* -> proc_mkdir failed -> remove_proc_entry not needed */
		return_VALUE(0);

	acpi_processor_dir->owner = THIS_MODULE;

	result = acpi_bus_register_driver(&acpi_processor_driver);
	if (result < 0) {
		remove_proc_entry(ACPI_PROCESSOR_CLASS, acpi_root_dir);
/* here it is already removed ... */
		return_VALUE(0);
	}
/*        else ... here it must be removed... */
	processor_driver_registered = 1;

Thanks,

         Thomas