Re: [PATCH 3/3] RT: CPU priority management

[Steven Rostedt <rostedt@goodmis.org>]

--

> +
> +struct cpu_priority {
> +	raw_spinlock_t lock;
> +	cpumask_t      pri_to_cpu[CPUPRI_NR_PRIORITIES];
> +	long           pri_active[CPUPRI_NR_PRIORITIES/BITS_PER_LONG];
> +	int            cpu_to_pri[NR_CPUS];
> +};
> +
> +static __cacheline_aligned_in_smp struct cpu_priority cpu_priority;
> +

[...]

> +int cpupri_find(int def_cpu, struct task_struct *p)
> +{
> +	int                  idx      = 0;
> +	struct cpu_priority *cp       = &cpu_priority;
> +	int                  this_cpu = smp_processor_id();
> +	int                  cpu      = def_cpu;
> +	int                  task_pri = convert_prio(p->prio);
> +
> +	for_each_cpupri_active(cp->pri_active, idx) {

[...]

> +void cpupri_set(int cpu, int newpri)
> +{
> +	struct cpu_priority *cp      = &cpu_priority;
> +	int                 *currpri = &cp->cpu_to_pri[cpu];
> +	int                  oldpri  = *currpri;
> +	unsigned long        flags;
> +
> +	newpri = convert_prio(newpri);
> +
> +	if (newpri == oldpri)
> +		return;
> +
> +	spin_lock_irqsave(&cp->lock, flags);

The cpu_priority and the cp->lock will be aboslutely horrible for
cacheline bouncing.  Ironically, this will kill performance for the very
machines this code is to help with.  The larger the number of CPUs you
have the more cacheline bouncing this code will create.

I still don't see the benefit from the cpupri code.

-- Steve

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/