Re: [RC][PATCH -mm] PM: Prevent frozen user mode helpers from failing the freezing of tasks

[Pavel Machek <pavel@ucw.cz>]

Hi!

> From: Rafael J. Wysocki <rjw@sisk.pl>
> 
> At present, if a user mode helper is running while usermodehelper_pm_callback()
> is executed, the helper may be frozen and the completion in
> call_usermodehelper_exec() won't be completed until user space processes are
> thawed.  As a result, the freezing of kernel threads may fail, which is not
> desirable.
> 
> Prevent this from happening by introducing a counter of running user mode
> helpers and allowing usermodehelper_pm_callback() to succeed for
> action = PM_HIBERNATION_PREPARE or action = PM_SUSPEND_PREPARE only if there
> are no helpers running.  [Namely, usermodehelper_pm_callback() waits for at most
> RUNNING_HELPERS_TIMEOUT for the number of running helpers to become zero and
> fails if that doesn't happen.]
> 
> Special thanks to Uli Luckas <u.luckas@road.de> for reviewing the previous
> versions of this patch and for very useful comments.
...
> Reviewed-by: Pavel Machek <pavel@ucw.cz>

Eh, not sure this header is worth anything. Sometimes I'm lazy and
stop when I see first problem.

>  	switch (action) {
>  	case PM_HIBERNATION_PREPARE:
>  	case PM_SUSPEND_PREPARE:
>  		usermodehelper_disabled = 1;
> -		return NOTIFY_OK;
> +		retval = wait_event_timeout(running_helpers_waitq,
> +					atomic_read(&running_helpers) == 0,

Are you sure here? What happens when atomic variable changes between
the atomic_read and the function call?
						Pavel
-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html