On Wed, Jan 5, 2011 at 12:44 AM, Rik van Riel <riel@redhat.com> wrote:
quoted text
> On 01/04/2011 11:41 AM, Hillf Danton wrote:
>
>> /*                       !curr->sched_class->yield_to_task ||        */
>>
>>> +                       curr->sched_class != p->sched_class) {
>>> +               goto out;
>>> +       }
>>> +
>>
>>        /*
>>          * ask scheduler to compute the next for successfully kicking
>> @p onto its CPU
>>          * what if p_rq is rt_class to do?
>>          */
>>        next = pick_next_task(p_rq);
>>        if (next != p)
>>                p->se.vruntime = next->se.vruntime -1;
>>        deactivate_task(p_rq, p, 0);
>>        activate_task(p_rq, p, 0);
>>        if (rq == p_rq)
>>                schedule();
>>        else
>>                resched_task(p_rq->curr);
>>        yield = 0;
>
> Wouldn't that break for FIFO and RR tasks?
>
> There's a reason all the scheduler folks wanted a
> per-class yield_to_task function :)
>

Where is the yield_to callback in the patch for RT schedule class?
If @p is RT, what could you do?

Hillf