Re: [PATCH] specific do_timer_cpu value for nohz off mode

[Mike Galbraith <efault@gmx.de>]

On Sun, 2012-01-15 at 14:46 +0100, Mike Galbraith wrote:
> On Tue, 2011-11-08 at 13:11 -0600, Dimitri Sivanich wrote:
> > Resending this.
> > 
> > 
> > Allow manual override of the tick_do_timer_cpu.
> 
> Bigger button below.


> +/* Called with cgroup_mutex held */
> +void cpuset_flags_set(int cpu, unsigned bits)
> +{
> +	struct rq *rq = cpu_rq(cpu);
> +	unsigned long flags;
> +	int nr;
> +
> +	raw_spin_lock_irqsave(&rq->lock, flags);
> +	/* Set blocker flags before taking any action */
> +	rq->cpuset_flags |= bits;
> +	for (nr = 0; bits; nr++) {
> +		if (!(bits & (1 << nr)))
> +			continue;
> +		switch (nr) {
> +		case RQ_TICK:
> +			break;
> +		case RQ_HPC:
> +			/* Ensure that jiffies doesn't go stale */
> +			if (!nr_hpc_cpus++) {
> +				tick_do_timer_cpu = 0;
> +				/* safe, CPU0 is modifier excluded */
> +				cpuset_flags_set(0, 1 << RQ_TICK);
> +				wake_up_idle_cpu(0);

Just in case someone was going to mention it, I'd already moved that
wakeup_idle_cpu() to RQ_TICK case, but the darn thing crawled back.

	-Mike