Re: [patch V2 08/10] timer: Implement the hierarchical pull model

[Thomas Gleixner]

On Wed, 19 Apr 2017, Peter Zijlstra wrote:
> On Tue, Apr 18, 2017 at 01:11:10PM +0200, Thomas Gleixner wrote:
> > +static struct tmigr_group *tmigr_get_group(unsigned int node, unsigned int lvl)
> > +{
> > +	struct tmigr_group *group;
> > +
> > +	/* Try to attach to an exisiting group first */
> > +	list_for_each_entry(group, &tmigr_level_list[lvl], list) {
> > +		/*
> > +		 * If @lvl is below the cross numa node level, check
> > +		 * whether this group belongs to the same numa node.
> > +		 */
> > +		if (lvl < tmigr_crossnode_level && group->numa_node != node)
> > +			continue;
> > +		/* If the group has capacity, use it */
> > +		if (group->num_childs < tmigr_childs_per_group) {
> > +			group->num_childs++;
> > +			return group;
> > +		}
> 
> This would result in SMT siblings not sharing groups on regular Intel
> systems, right? Since they get enumerated last.

Indeed. Will fix.

> > +	}
> > +	/* Allocate and set up a new group */
> > +	group = kzalloc_node(sizeof(*group), GFP_KERNEL, node);
> > +	if (!group)
> > +		return ERR_PTR(-ENOMEM);
> > +
> > +	if (!zalloc_cpumask_var_node(&group->cpus, GFP_KERNEL, node)) {
> > +		kfree(group);
> > +		return ERR_PTR(-ENOMEM);
> > +	}
> 
> So if you place that cpumask last, you can do:
> 
> 	group = kzalloc_node(sizeof(*group) + cpumask_size(),
> 	                     GFP_KERNEL, node);

Hrm, that would allocate extra space for CPUMASK_OFF_STACK=n. I'll have a
look.

Thanks,

	tglx