Re: CFS and suspend2: hang in atomic copy

[Christian Hesse <mail@earthworm.de>]

On Thursday 19 April 2007, Ingo Molnar wrote:
> * Christian Hesse <mail@earthworm.de> wrote:
> > I now got some error message from my system:
> >
> > http://www.eworm.de/tmp/cfs-suspend.jpg
>
> ah, this pinpoints a bug: for performance reasons pick_next_task()
> assumes that the runqueue is not empty - which is true for schedule(),
> but not in migrate_dead_tasks(). Does the patch below fix the crash for
> you?
>
>  kernel/sched.c |    2 ++
>  1 file changed, 2 insertions(+)
>
> Index: linux/kernel/sched.c
> ===================================================================
> --- linux.orig/kernel/sched.c
> +++ linux/kernel/sched.c
> @@ -4425,6 +4425,8 @@ static void migrate_dead_tasks(unsigned
>  	struct task_struct *next;
>
>  	for (;;) {
> +		if (!rq->nr_running)
> +			break;
>  		next = pick_next_task(rq, rq->curr);
>  		if (!next)
>  			break;

Suspend works perfectly with this patch. Thanks a lot and keep up the good 
work!
-- 
Regards,
Chris

Attachment: signature.asc
Description: This is a digitally signed message part.