On 05/09/2013 01:29 PM, Alex Shi wrote:
>
> I reviewed the cfs_rq->blocked_load_avg code path, no clear abnormal found.
> Seems the blocked load avg is fit current balance rules.
Sorry, I mean, the blocked load avg doesn't fit current balance rules.
The reason is blow, any comments on this?
> Som
On 05/09/2013 09:24 AM, Alex Shi wrote:
> Paul,
>
> I am wondering if the following patch needed.
> The wakeuped migrated task will __synchronize_entity_decay(se); in
> migrate_task_fair,
> then it needs to set `se->avg.last_runnable_update -= (-se->avg.decay_count)
> << 20'
> before update_enti
On 05/08/2013 09:39 AM, Alex Shi wrote:
> On 05/07/2013 01:17 PM, Alex Shi wrote:
>> > Sorry, what I meant to say here is:
>> > If we're going to be using a runnable average based load here the
>> > fraction we take (currently instantaneous) in tg_load_down should be
>> > consis
Paul,
I am wondering if the following patch needed.
The wakeuped migrated task will __synchronize_entity_decay(se); in
migrate_task_fair,
then it needs to set `se->avg.last_runnable_update -= (-se->avg.decay_count) <<
20'
before update_entity_load_avg, in order to avoid slept time is updated twi
On 05/07/2013 01:17 PM, Alex Shi wrote:
>> > Sorry, what I meant to say here is:
>> > If we're going to be using a runnable average based load here the
>> > fraction we take (currently instantaneous) in tg_load_down should be
>> > consistent.
> yes. I think so.
>
> So, here is the patch, could you
On 05/07/2013 04:59 AM, Paul Turner wrote:
>>> Similarly, I think you also want to at least include blocked_load_avg here.
>> >
>> > I'm puzzled, this is an entity weight. Entity's don't have
>> > blocked_load_avg.
>> >
>> > The purpose here is to compute the amount of weight that's being moved by
On Mon, May 6, 2013 at 8:04 AM, Peter Zijlstra wrote:
> On Mon, May 06, 2013 at 01:53:44AM -0700, Paul Turner wrote:
>> On Sun, May 5, 2013 at 6:45 PM, Alex Shi wrote:
>> > Except using runnable load average in background, move_tasks is also
>> > the key functions in load balance. We need conside
On 05/06/2013 04:53 PM, Paul Turner wrote:
>> > +static unsigned long task_h_load_avg(struct task_struct *p)
>> > +{
>> > + return div_u64(task_h_load(p) * (u64)p->se.avg.runnable_avg_sum,
>> > + p->se.avg.runnable_avg_period + 1);
> Similarly, I think you also want to a
On Mon, May 06, 2013 at 01:53:44AM -0700, Paul Turner wrote:
> On Sun, May 5, 2013 at 6:45 PM, Alex Shi wrote:
> > Except using runnable load average in background, move_tasks is also
> > the key functions in load balance. We need consider the runnable load
> > average in it in order to the apple
On Sun, May 5, 2013 at 6:45 PM, Alex Shi wrote:
> Except using runnable load average in background, move_tasks is also
> the key functions in load balance. We need consider the runnable load
> average in it in order to the apple to apple load comparison.
>
> Signed-off-by: Alex Shi
> ---
> kerne
Except using runnable load average in background, move_tasks is also
the key functions in load balance. We need consider the runnable load
average in it in order to the apple to apple load comparison.
Signed-off-by: Alex Shi
---
kernel/sched/fair.c | 8 +++-
1 file changed, 7 insertions(+),
11 matches
Mail list logo
