On Thu, 2016-04-07 at 16:04 +0100, George Dunlap wrote:
> On 07/04/16 06:04, Juergen Gross wrote:
> > On 06/04/16 19:23, Dario Faggioli wrote:
> > > @@ -2170,6 +2234,8 @@ csched2_init(struct scheduler *ops)
> > > printk(" load_window_shift: %d\n", opt_load_window_shift);
> > > printk(" un
On 07/04/16 06:04, Juergen Gross wrote:
> On 06/04/16 19:23, Dario Faggioli wrote:
>> In fact, credit2 uses CPU topology to decide how to arrange
>> its internal runqueues. Before this change, only 'one runqueue
>> per socket' was allowed. However, experiments have shown that,
>> for instance, havi
On 06/04/16 19:23, Dario Faggioli wrote:
> In fact, credit2 uses CPU topology to decide how to arrange
> its internal runqueues. Before this change, only 'one runqueue
> per socket' was allowed. However, experiments have shown that,
> for instance, having one runqueue per physical core improves
> p
In fact, credit2 uses CPU topology to decide how to arrange
its internal runqueues. Before this change, only 'one runqueue
per socket' was allowed. However, experiments have shown that,
for instance, having one runqueue per physical core improves
performance, especially in case hyperthreading is av
