Re: [PATCH] sched: Do not stop cpu in set_cpus_allowed_ptr() if task is not running

[Preeti U Murthy <preeti@linux.vnet.ibm.com>]

On 09/15/2014 03:32 PM, Kirill Tkhai wrote:
> Hi, Preeti,
> 
> В Пн, 15/09/2014 в 15:18 +0530, Preeti Murthy пишет:
>> Hi Kirill,
>>
>> Which tree is this patch based on? __migrate_task() does a
>> double_rq_lock/unlock() today in mainline, doesn't it? I don't
>> however see that in your patch.
> 
> it's based on recent tip tree:
> https://git.kernel.org/cgit/linux/kernel/git/tip/tip.git/
> 
> Also, once I received a build-robot message about my patch
> makes warning on allyesconfig, and since that time I learned
> about Peter's tree :) 
> 
> https://git.kernel.org/cgit/linux/kernel/git/peterz/queue.git
> 
> (This is for info, I don't know if it is right to make patches
> based on in. But really good if you was away for some time
> and you're interested in recent news without lkml archive reading.
> I use it :)

Ok I checked the patch against tip tree and the patch looks good to me.

Reviewed-by; Preeti U Murthy <preeti@linux.vnet.ibm.com>

Thanks!

Regards
Preeti U Murthy
> 
> Regards,
> Kirill
> 
>> On Fri, Sep 12, 2014 at 4:33 PM, Kirill Tkhai <ktkhai@parallels.com> wrote:
>>>
>>> If a task is queued but not running on it rq, we can simply migrate
>>> it without migration thread and switching of context.
>>>
>>> Signed-off-by: Kirill Tkhai <ktkhai@parallels.com>
>>> ---
>>>  kernel/sched/core.c |   47 ++++++++++++++++++++++++++++++++---------------
>>>  1 file changed, 32 insertions(+), 15 deletions(-)
>>>
>>> diff --git a/kernel/sched/core.c b/kernel/sched/core.c
>>> index d4399b4..dbbba26 100644
>>> --- a/kernel/sched/core.c
>>> +++ b/kernel/sched/core.c
>>> @@ -4594,6 +4594,33 @@ void init_idle(struct task_struct *idle, int cpu)
>>>  }
>>>
>>>  #ifdef CONFIG_SMP
>>> +/*
>>> + * move_queued_task - move a queued task to new rq.
>>> + *
>>> + * Returns (locked) new rq. Old rq's lock is released.
>>> + */
>>> +static struct rq *move_queued_task(struct task_struct *p, int new_cpu)
>>> +{
>>> +       struct rq *rq = task_rq(p);
>>> +
>>> +       lockdep_assert_held(&rq->lock);
>>> +
>>> +       dequeue_task(rq, p, 0);
>>> +       p->on_rq = TASK_ON_RQ_MIGRATING;
>>> +       set_task_cpu(p, new_cpu);
>>> +       raw_spin_unlock(&rq->lock);
>>> +
>>> +       rq = cpu_rq(new_cpu);
>>> +
>>> +       raw_spin_lock(&rq->lock);
>>> +       BUG_ON(task_cpu(p) != new_cpu);
>>> +       p->on_rq = TASK_ON_RQ_QUEUED;
>>> +       enqueue_task(rq, p, 0);
>>> +       check_preempt_curr(rq, p, 0);
>>> +
>>> +       return rq;
>>> +}
>>> +
>>>  void do_set_cpus_allowed(struct task_struct *p, const struct cpumask *new_mask)
>>>  {
>>>         if (p->sched_class && p->sched_class->set_cpus_allowed)
>>> @@ -4650,14 +4677,15 @@ int set_cpus_allowed_ptr(struct task_struct *p, const struct cpumask *new_mask)
>>>                 goto out;
>>>
>>>         dest_cpu = cpumask_any_and(cpu_active_mask, new_mask);
>>> -       if (task_on_rq_queued(p) || p->state == TASK_WAKING) {
>>> +       if (task_running(rq, p) || p->state == TASK_WAKING) {
>>>                 struct migration_arg arg = { p, dest_cpu };
>>>                 /* Need help from migration thread: drop lock and wait. */
>>>                 task_rq_unlock(rq, p, &flags);
>>>                 stop_one_cpu(cpu_of(rq), migration_cpu_stop, &arg);
>>>                 tlb_migrate_finish(p->mm);
>>>                 return 0;
>>> -       }
>>> +       } else if (task_on_rq_queued(p))
>>> +               rq = move_queued_task(p, dest_cpu);
>>>  out:
>>>         task_rq_unlock(rq, p, &flags);
>>>
>>> @@ -4700,19 +4728,8 @@ static int __migrate_task(struct task_struct *p, int src_cpu, int dest_cpu)
>>>          * If we're not on a rq, the next wake-up will ensure we're
>>>          * placed properly.
>>>          */
>>> -       if (task_on_rq_queued(p)) {
>>> -               dequeue_task(rq, p, 0);
>>> -               p->on_rq = TASK_ON_RQ_MIGRATING;
>>> -               set_task_cpu(p, dest_cpu);
>>> -               raw_spin_unlock(&rq->lock);
>>> -
>>> -               rq = cpu_rq(dest_cpu);
>>> -               raw_spin_lock(&rq->lock);
>>> -               BUG_ON(task_rq(p) != rq);
>>> -               p->on_rq = TASK_ON_RQ_QUEUED;
>>> -               enqueue_task(rq, p, 0);
>>> -               check_preempt_curr(rq, p, 0);
>>> -       }
>>> +       if (task_on_rq_queued(p))
>>> +               rq = move_queued_task(p, dest_cpu);
>>>  done:
>>>         ret = 1;
>>>  fail:
>>>
>>>
>>>
>>> --
>>> 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/
> 
>