On 29.09.2015 20:29, Mike Galbraith wrote:
> On Tue, 2015-09-29 at 19:00 +0300, Kirill Tkhai wrote:
>>
>> On 29.09.2015 17:55, Mike Galbraith wrote:
>>> On Mon, 2015-09-28 at 18:36 +0300, Kirill Tkhai wrote:
>>>
---
diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
index 4df37
On Tue, 2015-09-29 at 19:00 +0300, Kirill Tkhai wrote:
>
> On 29.09.2015 17:55, Mike Galbraith wrote:
> > On Mon, 2015-09-28 at 18:36 +0300, Kirill Tkhai wrote:
> >
> >> ---
> >> diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
> >> index 4df37a4..dfbe06b 100644
> >> --- a/kernel/sched/fair
On 29.09.2015 19:00, Kirill Tkhai wrote:
>
>
> On 29.09.2015 17:55, Mike Galbraith wrote:
>> On Mon, 2015-09-28 at 18:36 +0300, Kirill Tkhai wrote:
>>
>>> ---
>>> diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
>>> index 4df37a4..dfbe06b 100644
>>> --- a/kernel/sched/fair.c
>>> +++ b/ker
On 29.09.2015 17:55, Mike Galbraith wrote:
> On Mon, 2015-09-28 at 18:36 +0300, Kirill Tkhai wrote:
>
>> ---
>> diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
>> index 4df37a4..dfbe06b 100644
>> --- a/kernel/sched/fair.c
>> +++ b/kernel/sched/fair.c
>> @@ -4930,8 +4930,13 @@ select_task_
On Mon, 2015-09-28 at 18:36 +0300, Kirill Tkhai wrote:
> ---
> diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
> index 4df37a4..dfbe06b 100644
> --- a/kernel/sched/fair.c
> +++ b/kernel/sched/fair.c
> @@ -4930,8 +4930,13 @@ select_task_rq_fair(struct task_struct *p, int
> prev_cpu, int sd_
On Mon, 2015-09-28 at 22:19 +0300, Kirill Tkhai wrote:
> >> Imagine a situation, when we share a mutex
> >> with a task on another NUMA node. When the task is realising the mutex
> >> it is waking us, but we definitelly won't use affine logic in this case.
> >
> > Why not? A wakeup is a wakeup i
On 28.09.2015 21:22, Mike Galbraith wrote:
> On Mon, 2015-09-28 at 18:36 +0300, Kirill Tkhai wrote:
>
>> Mike, one more moment. wake_wide() and current logic confuses me a bit.
>> It makes us to decide if we want affine wakeup or not, but
>> select_idle_sibling()
>> if a function is not for cho
On Mon, 2015-09-28 at 18:36 +0300, Kirill Tkhai wrote:
> Mike, one more moment. wake_wide() and current logic confuses me a bit.
> It makes us to decide if we want affine wakeup or not, but
> select_idle_sibling()
> if a function is not for choosing this_cpu's llc domain only. We use it
> for sea
On 28.09.2015 18:36, Kirill Tkhai wrote:
> On 28.09.2015 16:12, Mike Galbraith wrote:
>> On Mon, 2015-09-28 at 13:28 +0300, Kirill Tkhai wrote:
>>
>>> Looks like, NAK may be better, because it saves L1 cache, while the patch
>>> always invalidates it.
>>
>> Yeah, bounce hurts more when there's n
On 28.09.2015 16:12, Mike Galbraith wrote:
> On Mon, 2015-09-28 at 13:28 +0300, Kirill Tkhai wrote:
>
>> Looks like, NAK may be better, because it saves L1 cache, while the patch
>> always invalidates it.
>
> Yeah, bounce hurts more when there's no concurrency win waiting to be
> collected. Thi
On Mon, 2015-09-28 at 13:28 +0300, Kirill Tkhai wrote:
> Looks like, NAK may be better, because it saves L1 cache, while the patch
> always invalidates it.
Yeah, bounce hurts more when there's no concurrency win waiting to be
collected. This mixed load wasn't a great choice, but it turned out t
On 26.09.2015 18:25, Mike Galbraith wrote:
> On Fri, 2015-09-25 at 20:54 +0300, Kirill Tkhai wrote:
>> We are not interested in actual target if both prev
>> and curr cpus share CPU cache. select_idle_sibling()
>> searches in top-down order; top level is the same
>> for both of them, and the resul
On Fri, 2015-09-25 at 20:54 +0300, Kirill Tkhai wrote:
> We are not interested in actual target if both prev
> and curr cpus share CPU cache. select_idle_sibling()
> searches in top-down order; top level is the same
> for both of them, and the result will be the same.
> So, we can save a little CPU
13 matches
Mail list logo
