* Re: [PATCH V2] sched/fair: Fix that tasks are not constrained by cfs_b->quota on hotplug core, when hotplug core is offline and then online.
@ 2016-09-22 7:59 Jeehong Kim
2016-09-22 16:53 ` bsegall
0 siblings, 1 reply; 9+ messages in thread
From: Jeehong Kim @ 2016-09-22 7:59 UTC (permalink / raw)
To: Ben Segall; +Cc: mingo, linux-kernel, ezjjilong, Peter Zijlstra
>Peter Zijlstra <peterz@xxxxxxxxxxxxx> writes:
>
>> You forgot to Cc Ben, who gave you feedback on v1, which is rather poor
>> style. Also, I don't see how kernel-janitors is relevant to this patch.
>> This is very much not a janitorial thing.
>>
>> (also, why send it twice?)
>>
>> On Tue, Aug 30, 2016 at 10:12:40PM +0900, Jeehong Kim wrote:
>>> In case that CONFIG_HOTPLUG_CPU and CONFIG_CFS_BANDWIDTH is turned on
>>> and tasks in bandwidth controlled task group run on hotplug core,
>>> the tasks are not controlled by cfs_b->quota when hotplug core is offline
>>> and then online. The remaining tasks in task group consume all of
>>> cfs_b->quota on other cores.
>>>
>>> The cause of this problem is described as below:
>>>
>>> 1. When hotplug core is offline while tasks in task group run
>>> on hotplug core, unregister_fair_sched_group() deletes
>>> leaf_cfs_rq_list of tg->cfs_rq[cpu] from &rq_of(cfs_rq)->leaf_cfs_rq_list.
>>>
>>> 2. Then, when hotplug core is online, update_runtime_enabled()
>>Peter Zijlstra <peterz@xxxxxxxxxxxxx> writes:
>
>> You forgot to Cc Ben, who gave you feedback on v1, which is rather poor
>> style. Also, I don't see how kernel-janitors is relevant to this patch.
>> This is very much not a janitorial thing.
>>
>> (also, why send it twice?)
>>
>> On Tue, Aug 30, 2016 at 10:12:40PM +0900, Jeehong Kim wrote:
>>> In case that CONFIG_HOTPLUG_CPU and CONFIG_CFS_BANDWIDTH is turned on
>>> and tasks in bandwidth controlled task group run on hotplug core,
>>> the tasks are not controlled by cfs_b->quota when hotplug core is offline
>>> and then online. The remaining tasks in task group consume all of
>>> cfs_b->quota on other cores.
>>>
>>> The cause of this problem is described as below:
>>>
>>> 1. When hotplug core is offline while tasks in task group run
>>> on hotplug core, unregister_fair_sched_group() deletes
>>> leaf_cfs_rq_list of tg->cfs_rq[cpu] from &rq_of(cfs_rq)->leaf_cfs_rq_list.
>>>
>>> 2. Then, when hotplug core is online, update_runtime_enabled()
>>> registers cfs_b->quota on cfs_rq->runtime_enabled of all leaf cfs_rq
>>> on runqueue. However, because this is before enqueue_entity() adds
>>> &cfs_rq->leaf_cfs_rq_list on &rq_of(cfs_rq)->leaf_cfs_rq_list,
>>> cfs->quota is not register on cfs_rq->runtime_enabled.
>>>
>>> To resolve this problem, this patch makes update_runtime_enabled()
>>> registers cfs_b->quota by using walk_tg_tree_from().
>>
>>
>>> +static int __maybe_unused __update_runtime_enabled(struct task_group *tg, void *data)
>>> {
>>> + struct rq *rq = data;
>>> + struct cfs_rq *cfs_rq = tg->cfs_rq[cpu_of(rq)];
>>> + struct cfs_bandwidth *cfs_b = &cfs_rq->tg->cfs_bandwidth;
>>>
>>> + raw_spin_lock(&cfs_b->lock);
>>> + raw_spin_unlock(&cfs_b->lock);
>>>
>>> + return 0;
>>> +}
>>> +
>>> +static void __maybe_unused update_runtime_enabled(struct rq *rq)
>>> +{
>>> + struct cfs_rq *cfs_rq = &rq->cfs;
>>> +
>>> + /* register cfs_b->quota on the whole tg tree */
>>> + rcu_read_lock();
>>> + walk_tg_tree_from(cfs_rq->tg, __update_runtime_enabled, tg_nop, (void *)rq);
>>> + rcu_read_unlock();
>>> }
>>
>> Looks ok, performance on hotplug doesn't really matter. Ben, you happy
>> with this?
>
> I'm not 100% sure about the exact timings and mechanics of hotplug, but
> cfs-bandwidth wise this is ok. We may still have runtime_remaining = 1,
> or we may have < 0 and yet be unthrottled, but either case is ok, even
> if hotplug allows tasks to have migrated here already (I'm not sure,
> looking at the code).
>
> Now that I check again you can just loop over the list of tgs rather
> than the hierarchical walk_tg_tree_from, but there's certainly no harm
> in it.
Ben,
Is there additional revision which I have to do?
If so, could you let me know about that?
Regards,
Jeehong Kim
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: [PATCH V2] sched/fair: Fix that tasks are not constrained by cfs_b->quota on hotplug core, when hotplug core is offline and then online.
2016-09-22 7:59 [PATCH V2] sched/fair: Fix that tasks are not constrained by cfs_b->quota on hotplug core, when hotplug core is offline and then online Jeehong Kim
@ 2016-09-22 16:53 ` bsegall
2016-09-27 3:41 ` Jeehong Kim
0 siblings, 1 reply; 9+ messages in thread
From: bsegall @ 2016-09-22 16:53 UTC (permalink / raw)
To: Jeehong Kim; +Cc: mingo, linux-kernel, ezjjilong, Peter Zijlstra
Jeehong Kim <jhez.kim@samsung.com> writes:
>>Peter Zijlstra <peterz@xxxxxxxxxxxxx> writes:
>>
>>> You forgot to Cc Ben, who gave you feedback on v1, which is rather poor
>>> style. Also, I don't see how kernel-janitors is relevant to this patch.
>>> This is very much not a janitorial thing.
>>>
>>> (also, why send it twice?)
>>>
>>> On Tue, Aug 30, 2016 at 10:12:40PM +0900, Jeehong Kim wrote:
>>>> In case that CONFIG_HOTPLUG_CPU and CONFIG_CFS_BANDWIDTH is turned on
>>>> and tasks in bandwidth controlled task group run on hotplug core,
>>>> the tasks are not controlled by cfs_b->quota when hotplug core is offline
>>>> and then online. The remaining tasks in task group consume all of
>>>> cfs_b->quota on other cores.
>>>>
>>>> The cause of this problem is described as below:
>>>>
>>>> 1. When hotplug core is offline while tasks in task group run
>>>> on hotplug core, unregister_fair_sched_group() deletes
>>>> leaf_cfs_rq_list of tg->cfs_rq[cpu] from &rq_of(cfs_rq)->leaf_cfs_rq_list.
>>>>
>>>> 2. Then, when hotplug core is online, update_runtime_enabled()
>>>Peter Zijlstra <peterz@xxxxxxxxxxxxx> writes:
>>
>>> You forgot to Cc Ben, who gave you feedback on v1, which is rather poor
>>> style. Also, I don't see how kernel-janitors is relevant to this patch.
>>> This is very much not a janitorial thing.
>>>
>>> (also, why send it twice?)
>>>
>>> On Tue, Aug 30, 2016 at 10:12:40PM +0900, Jeehong Kim wrote:
>>>> In case that CONFIG_HOTPLUG_CPU and CONFIG_CFS_BANDWIDTH is turned on
>>>> and tasks in bandwidth controlled task group run on hotplug core,
>>>> the tasks are not controlled by cfs_b->quota when hotplug core is offline
>>>> and then online. The remaining tasks in task group consume all of
>>>> cfs_b->quota on other cores.
>>>>
>>>> The cause of this problem is described as below:
>>>>
>>>> 1. When hotplug core is offline while tasks in task group run
>>>> on hotplug core, unregister_fair_sched_group() deletes
>>>> leaf_cfs_rq_list of tg->cfs_rq[cpu] from &rq_of(cfs_rq)->leaf_cfs_rq_list.
>>>>
>>>> 2. Then, when hotplug core is online, update_runtime_enabled()
>>>> registers cfs_b->quota on cfs_rq->runtime_enabled of all leaf cfs_rq
>>>> on runqueue. However, because this is before enqueue_entity() adds
>>>> &cfs_rq->leaf_cfs_rq_list on &rq_of(cfs_rq)->leaf_cfs_rq_list,
>>>> cfs->quota is not register on cfs_rq->runtime_enabled.
>>>>
>>>> To resolve this problem, this patch makes update_runtime_enabled()
>>>> registers cfs_b->quota by using walk_tg_tree_from().
>>>
>>>
>>>> +static int __maybe_unused __update_runtime_enabled(struct task_group *tg, void *data)
>>>> {
>>>> + struct rq *rq = data;
>>>> + struct cfs_rq *cfs_rq = tg->cfs_rq[cpu_of(rq)];
>>>> + struct cfs_bandwidth *cfs_b = &cfs_rq->tg->cfs_bandwidth;
>>>>
>>>> + raw_spin_lock(&cfs_b->lock);
>>>> + raw_spin_unlock(&cfs_b->lock);
>>>>
>>>> + return 0;
>>>> +}
>>>> +
>>>> +static void __maybe_unused update_runtime_enabled(struct rq *rq)
>>>> +{
>>>> + struct cfs_rq *cfs_rq = &rq->cfs;
>>>> +
>>>> + /* register cfs_b->quota on the whole tg tree */
>>>> + rcu_read_lock();
>>>> + walk_tg_tree_from(cfs_rq->tg, __update_runtime_enabled, tg_nop, (void *)rq);
>>>> + rcu_read_unlock();
>>>> }
>>>
>>> Looks ok, performance on hotplug doesn't really matter. Ben, you happy
>>> with this?
>>
>> I'm not 100% sure about the exact timings and mechanics of hotplug, but
>> cfs-bandwidth wise this is ok. We may still have runtime_remaining = 1,
>> or we may have < 0 and yet be unthrottled, but either case is ok, even
>> if hotplug allows tasks to have migrated here already (I'm not sure,
>> looking at the code).
>>
>> Now that I check again you can just loop over the list of tgs rather
>> than the hierarchical walk_tg_tree_from, but there's certainly no harm
>> in it.
>
> Ben,
>
> Is there additional revision which I have to do?
> If so, could you let me know about that?
>
> Regards,
> Jeehong Kim
Oh, no, this is fine by me.
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: [PATCH V2] sched/fair: Fix that tasks are not constrained by cfs_b->quota on hotplug core, when hotplug core is offline and then online.
2016-09-22 16:53 ` bsegall
@ 2016-09-27 3:41 ` Jeehong Kim
2016-09-27 17:13 ` bsegall
0 siblings, 1 reply; 9+ messages in thread
From: Jeehong Kim @ 2016-09-27 3:41 UTC (permalink / raw)
To: bsegall; +Cc: mingo, linux-kernel, ezjjilong, Peter Zijlstra
On 2016년 09월 23일 01:53, bsegall@google.com wrote:
> Jeehong Kim <jhez.kim@samsung.com> writes:
>
>>> Peter Zijlstra <peterz@xxxxxxxxxxxxx> writes:
>>>
>>>> You forgot to Cc Ben, who gave you feedback on v1, which is rather poor
>>>> style. Also, I don't see how kernel-janitors is relevant to this patch.
>>>> This is very much not a janitorial thing.
>>>>
>>>> (also, why send it twice?)
>>>>
>>>> On Tue, Aug 30, 2016 at 10:12:40PM +0900, Jeehong Kim wrote:
>>>>> In case that CONFIG_HOTPLUG_CPU and CONFIG_CFS_BANDWIDTH is turned on
>>>>> and tasks in bandwidth controlled task group run on hotplug core,
>>>>> the tasks are not controlled by cfs_b->quota when hotplug core is offline
>>>>> and then online. The remaining tasks in task group consume all of
>>>>> cfs_b->quota on other cores.
>>>>>
>>>>> The cause of this problem is described as below:
>>>>>
>>>>> 1. When hotplug core is offline while tasks in task group run
>>>>> on hotplug core, unregister_fair_sched_group() deletes
>>>>> leaf_cfs_rq_list of tg->cfs_rq[cpu] from &rq_of(cfs_rq)->leaf_cfs_rq_list.
>>>>>
>>>>> 2. Then, when hotplug core is online, update_runtime_enabled()
>>>> Peter Zijlstra <peterz@xxxxxxxxxxxxx> writes:
>>>> You forgot to Cc Ben, who gave you feedback on v1, which is rather poor
>>>> style. Also, I don't see how kernel-janitors is relevant to this patch.
>>>> This is very much not a janitorial thing.
>>>>
>>>> (also, why send it twice?)
>>>>
>>>> On Tue, Aug 30, 2016 at 10:12:40PM +0900, Jeehong Kim wrote:
>>>>> In case that CONFIG_HOTPLUG_CPU and CONFIG_CFS_BANDWIDTH is turned on
>>>>> and tasks in bandwidth controlled task group run on hotplug core,
>>>>> the tasks are not controlled by cfs_b->quota when hotplug core is offline
>>>>> and then online. The remaining tasks in task group consume all of
>>>>> cfs_b->quota on other cores.
>>>>>
>>>>> The cause of this problem is described as below:
>>>>>
>>>>> 1. When hotplug core is offline while tasks in task group run
>>>>> on hotplug core, unregister_fair_sched_group() deletes
>>>>> leaf_cfs_rq_list of tg->cfs_rq[cpu] from &rq_of(cfs_rq)->leaf_cfs_rq_list.
>>>>>
>>>>> 2. Then, when hotplug core is online, update_runtime_enabled()
>>>>> registers cfs_b->quota on cfs_rq->runtime_enabled of all leaf cfs_rq
>>>>> on runqueue. However, because this is before enqueue_entity() adds
>>>>> &cfs_rq->leaf_cfs_rq_list on &rq_of(cfs_rq)->leaf_cfs_rq_list,
>>>>> cfs->quota is not register on cfs_rq->runtime_enabled.
>>>>>
>>>>> To resolve this problem, this patch makes update_runtime_enabled()
>>>>> registers cfs_b->quota by using walk_tg_tree_from().
>>>>
>>>>> +static int __maybe_unused __update_runtime_enabled(struct task_group *tg, void *data)
>>>>> {
>>>>> + struct rq *rq = data;
>>>>> + struct cfs_rq *cfs_rq = tg->cfs_rq[cpu_of(rq)];
>>>>> + struct cfs_bandwidth *cfs_b = &cfs_rq->tg->cfs_bandwidth;
>>>>>
>>>>> + raw_spin_lock(&cfs_b->lock);
>>>>> + raw_spin_unlock(&cfs_b->lock);
>>>>>
>>>>> + return 0;
>>>>> +}
>>>>> +
>>>>> +static void __maybe_unused update_runtime_enabled(struct rq *rq)
>>>>> +{
>>>>> + struct cfs_rq *cfs_rq = &rq->cfs;
>>>>> +
>>>>> + /* register cfs_b->quota on the whole tg tree */
>>>>> + rcu_read_lock();
>>>>> + walk_tg_tree_from(cfs_rq->tg, __update_runtime_enabled, tg_nop, (void *)rq);
>>>>> + rcu_read_unlock();
>>>>> }
>>>> Looks ok, performance on hotplug doesn't really matter. Ben, you happy
>>>> with this?
>>> I'm not 100% sure about the exact timings and mechanics of hotplug, but
>>> cfs-bandwidth wise this is ok. We may still have runtime_remaining = 1,
>>> or we may have < 0 and yet be unthrottled, but either case is ok, even
>>> if hotplug allows tasks to have migrated here already (I'm not sure,
>>> looking at the code).
>>>
>>> Now that I check again you can just loop over the list of tgs rather
>>> than the hierarchical walk_tg_tree_from, but there's certainly no harm
>>> in it.
>> Ben,
>>
>> Is there additional revision which I have to do?
>> If so, could you let me know about that?
>>
>> Regards,
>> Jeehong Kim
> Oh, no, this is fine by me.
>
>
>
Ben,
If this is fine to you, could you sign off on this patch?
Regards,
Jeehong Kim.
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: [PATCH V2] sched/fair: Fix that tasks are not constrained by cfs_b->quota on hotplug core, when hotplug core is offline and then online.
2016-09-27 3:41 ` Jeehong Kim
@ 2016-09-27 17:13 ` bsegall
0 siblings, 0 replies; 9+ messages in thread
From: bsegall @ 2016-09-27 17:13 UTC (permalink / raw)
To: Jeehong Kim; +Cc: mingo, linux-kernel, ezjjilong, Peter Zijlstra
Jeehong Kim <jhez.kim@samsung.com> writes:
> On 2016년 09월 23일 01:53, bsegall@google.com wrote:
>> Jeehong Kim <jhez.kim@samsung.com> writes:
>>
>>>> Peter Zijlstra <peterz@xxxxxxxxxxxxx> writes:
>>>>
>>>>> You forgot to Cc Ben, who gave you feedback on v1, which is rather poor
>>>>> style. Also, I don't see how kernel-janitors is relevant to this patch.
>>>>> This is very much not a janitorial thing.
>>>>>
>>>>> (also, why send it twice?)
>>>>>
>>>>> On Tue, Aug 30, 2016 at 10:12:40PM +0900, Jeehong Kim wrote:
>>>>>> In case that CONFIG_HOTPLUG_CPU and CONFIG_CFS_BANDWIDTH is turned on
>>>>>> and tasks in bandwidth controlled task group run on hotplug core,
>>>>>> the tasks are not controlled by cfs_b->quota when hotplug core is offline
>>>>>> and then online. The remaining tasks in task group consume all of
>>>>>> cfs_b->quota on other cores.
>>>>>>
>>>>>> The cause of this problem is described as below:
>>>>>>
>>>>>> 1. When hotplug core is offline while tasks in task group run
>>>>>> on hotplug core, unregister_fair_sched_group() deletes
>>>>>> leaf_cfs_rq_list of tg->cfs_rq[cpu] from &rq_of(cfs_rq)->leaf_cfs_rq_list.
>>>>>>
>>>>>> 2. Then, when hotplug core is online, update_runtime_enabled()
>>>>> Peter Zijlstra <peterz@xxxxxxxxxxxxx> writes:
>>>>> You forgot to Cc Ben, who gave you feedback on v1, which is rather poor
>>>>> style. Also, I don't see how kernel-janitors is relevant to this patch.
>>>>> This is very much not a janitorial thing.
>>>>>
>>>>> (also, why send it twice?)
>>>>>
>>>>> On Tue, Aug 30, 2016 at 10:12:40PM +0900, Jeehong Kim wrote:
>>>>>> In case that CONFIG_HOTPLUG_CPU and CONFIG_CFS_BANDWIDTH is turned on
>>>>>> and tasks in bandwidth controlled task group run on hotplug core,
>>>>>> the tasks are not controlled by cfs_b->quota when hotplug core is offline
>>>>>> and then online. The remaining tasks in task group consume all of
>>>>>> cfs_b->quota on other cores.
>>>>>>
>>>>>> The cause of this problem is described as below:
>>>>>>
>>>>>> 1. When hotplug core is offline while tasks in task group run
>>>>>> on hotplug core, unregister_fair_sched_group() deletes
>>>>>> leaf_cfs_rq_list of tg->cfs_rq[cpu] from &rq_of(cfs_rq)->leaf_cfs_rq_list.
>>>>>>
>>>>>> 2. Then, when hotplug core is online, update_runtime_enabled()
>>>>>> registers cfs_b->quota on cfs_rq->runtime_enabled of all leaf cfs_rq
>>>>>> on runqueue. However, because this is before enqueue_entity() adds
>>>>>> &cfs_rq->leaf_cfs_rq_list on &rq_of(cfs_rq)->leaf_cfs_rq_list,
>>>>>> cfs->quota is not register on cfs_rq->runtime_enabled.
>>>>>>
>>>>>> To resolve this problem, this patch makes update_runtime_enabled()
>>>>>> registers cfs_b->quota by using walk_tg_tree_from().
>>>>>
>>>>>> +static int __maybe_unused __update_runtime_enabled(struct task_group *tg, void *data)
>>>>>> {
>>>>>> + struct rq *rq = data;
>>>>>> + struct cfs_rq *cfs_rq = tg->cfs_rq[cpu_of(rq)];
>>>>>> + struct cfs_bandwidth *cfs_b = &cfs_rq->tg->cfs_bandwidth;
>>>>>>
>>>>>> + raw_spin_lock(&cfs_b->lock);
>>>>>> + raw_spin_unlock(&cfs_b->lock);
>>>>>>
>>>>>> + return 0;
>>>>>> +}
>>>>>> +
>>>>>> +static void __maybe_unused update_runtime_enabled(struct rq *rq)
>>>>>> +{
>>>>>> + struct cfs_rq *cfs_rq = &rq->cfs;
>>>>>> +
>>>>>> + /* register cfs_b->quota on the whole tg tree */
>>>>>> + rcu_read_lock();
>>>>>> + walk_tg_tree_from(cfs_rq->tg, __update_runtime_enabled, tg_nop, (void *)rq);
>>>>>> + rcu_read_unlock();
>>>>>> }
>>>>> Looks ok, performance on hotplug doesn't really matter. Ben, you happy
>>>>> with this?
>>>> I'm not 100% sure about the exact timings and mechanics of hotplug, but
>>>> cfs-bandwidth wise this is ok. We may still have runtime_remaining = 1,
>>>> or we may have < 0 and yet be unthrottled, but either case is ok, even
>>>> if hotplug allows tasks to have migrated here already (I'm not sure,
>>>> looking at the code).
>>>>
>>>> Now that I check again you can just loop over the list of tgs rather
>>>> than the hierarchical walk_tg_tree_from, but there's certainly no harm
>>>> in it.
>>> Ben,
>>>
>>> Is there additional revision which I have to do?
>>> If so, could you let me know about that?
>>>
>>> Regards,
>>> Jeehong Kim
>> Oh, no, this is fine by me.
>>
>>
>>
>
> Ben,
>
> If this is fine to you, could you sign off on this patch?
>
> Regards,
> Jeehong Kim.
Reviewed-By: Ben Segall <bsegall@google.com>
(My understanding of signed-off-by is that I don't do that in this case)
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: [PATCH V2] sched/fair: Fix that tasks are not constrained by cfs_b->quota on hotplug core, when hotplug core is offline and then online.
2016-09-01 11:01 ` Peter Zijlstra
2016-09-01 11:52 ` Jeehong Kim
@ 2016-09-01 17:23 ` bsegall
1 sibling, 0 replies; 9+ messages in thread
From: bsegall @ 2016-09-01 17:23 UTC (permalink / raw)
To: Peter Zijlstra
Cc: Jeehong Kim, mingo, linux-kernel, kernel-janitors, ezjjilong
Peter Zijlstra <peterz@infradead.org> writes:
> You forgot to Cc Ben, who gave you feedback on v1, which is rather poor
> style. Also, I don't see how kernel-janitors is relevant to this patch.
> This is very much not a janitorial thing.
>
> (also, why send it twice?)
>
> On Tue, Aug 30, 2016 at 10:12:40PM +0900, Jeehong Kim wrote:
>> In case that CONFIG_HOTPLUG_CPU and CONFIG_CFS_BANDWIDTH is turned on
>> and tasks in bandwidth controlled task group run on hotplug core,
>> the tasks are not controlled by cfs_b->quota when hotplug core is offline
>> and then online. The remaining tasks in task group consume all of
>> cfs_b->quota on other cores.
>>
>> The cause of this problem is described as below:
>>
>> 1. When hotplug core is offline while tasks in task group run
>> on hotplug core, unregister_fair_sched_group() deletes
>> leaf_cfs_rq_list of tg->cfs_rq[cpu] from &rq_of(cfs_rq)->leaf_cfs_rq_list.
>>
>> 2. Then, when hotplug core is online, update_runtime_enabled()
>> registers cfs_b->quota on cfs_rq->runtime_enabled of all leaf cfs_rq
>> on runqueue. However, because this is before enqueue_entity() adds
>> &cfs_rq->leaf_cfs_rq_list on &rq_of(cfs_rq)->leaf_cfs_rq_list,
>> cfs->quota is not register on cfs_rq->runtime_enabled.
>>
>> To resolve this problem, this patch makes update_runtime_enabled()
>> registers cfs_b->quota by using walk_tg_tree_from().
>
>
>> +static int __maybe_unused __update_runtime_enabled(struct task_group *tg, void *data)
>> {
>> + struct rq *rq = data;
>> + struct cfs_rq *cfs_rq = tg->cfs_rq[cpu_of(rq)];
>> + struct cfs_bandwidth *cfs_b = &cfs_rq->tg->cfs_bandwidth;
>>
>> + raw_spin_lock(&cfs_b->lock);
>> + cfs_rq->runtime_enabled = cfs_b->quota != RUNTIME_INF;
>> + raw_spin_unlock(&cfs_b->lock);
>>
>> + return 0;
>> +}
>> +
>> +static void __maybe_unused update_runtime_enabled(struct rq *rq)
>> +{
>> + struct cfs_rq *cfs_rq = &rq->cfs;
>> +
>> + /* register cfs_b->quota on the whole tg tree */
>> + rcu_read_lock();
>> + walk_tg_tree_from(cfs_rq->tg, __update_runtime_enabled, tg_nop, (void *)rq);
>> + rcu_read_unlock();
>> }
>
> Looks ok, performance on hotplug doesn't really matter. Ben, you happy
> with this?
I'm not 100% sure about the exact timings and mechanics of hotplug, but
cfs-bandwidth wise this is ok. We may still have runtime_remaining = 1,
or we may have < 0 and yet be unthrottled, but either case is ok, even
if hotplug allows tasks to have migrated here already (I'm not sure,
looking at the code).
Now that I check again you can just loop over the list of tgs rather
than the hierarchical walk_tg_tree_from, but there's certainly no harm
in it.
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: [PATCH V2] sched/fair: Fix that tasks are not constrained by cfs_b->quota on hotplug core, when hotplug core is offline and then online.
2016-09-01 11:01 ` Peter Zijlstra
@ 2016-09-01 11:52 ` Jeehong Kim
2016-09-01 17:23 ` bsegall
1 sibling, 0 replies; 9+ messages in thread
From: Jeehong Kim @ 2016-09-01 11:52 UTC (permalink / raw)
To: Peter Zijlstra
Cc: mingo, linux-kernel, kernel-janitors, ezjjilong, Ben Segall
Hello,
First of all, thanks for review. And I am sorry that I forgot to CC Ben.
The reason why I sent last mail twice is that I forgot including the comments for changes from PATCH V1.
Thanks and BR.
On 2016년 09월 01일 20:01, Peter Zijlstra wrote:
> You forgot to Cc Ben, who gave you feedback on v1, which is rather poor
> style. Also, I don't see how kernel-janitors is relevant to this patch.
> This is very much not a janitorial thing.
>
> (also, why send it twice?)
>
> On Tue, Aug 30, 2016 at 10:12:40PM +0900, Jeehong Kim wrote:
>> In case that CONFIG_HOTPLUG_CPU and CONFIG_CFS_BANDWIDTH is turned on
>> and tasks in bandwidth controlled task group run on hotplug core,
>> the tasks are not controlled by cfs_b->quota when hotplug core is offline
>> and then online. The remaining tasks in task group consume all of
>> cfs_b->quota on other cores.
>>
>> The cause of this problem is described as below:
>>
>> 1. When hotplug core is offline while tasks in task group run
>> on hotplug core, unregister_fair_sched_group() deletes
>> leaf_cfs_rq_list of tg->cfs_rq[cpu] from &rq_of(cfs_rq)->leaf_cfs_rq_list.
>>
>> 2. Then, when hotplug core is online, update_runtime_enabled()
>> registers cfs_b->quota on cfs_rq->runtime_enabled of all leaf cfs_rq
>> on runqueue. However, because this is before enqueue_entity() adds
>> &cfs_rq->leaf_cfs_rq_list on &rq_of(cfs_rq)->leaf_cfs_rq_list,
>> cfs->quota is not register on cfs_rq->runtime_enabled.
>>
>> To resolve this problem, this patch makes update_runtime_enabled()
>> registers cfs_b->quota by using walk_tg_tree_from().
>
>> +static int __maybe_unused __update_runtime_enabled(struct task_group *tg, void *data)
>> {
>> + struct rq *rq = data;
>> + struct cfs_rq *cfs_rq = tg->cfs_rq[cpu_of(rq)];
>> + struct cfs_bandwidth *cfs_b = &cfs_rq->tg->cfs_bandwidth;
>>
>> + raw_spin_lock(&cfs_b->lock);
>> + cfs_rq->runtime_enabled = cfs_b->quota != RUNTIME_INF;
>> + raw_spin_unlock(&cfs_b->lock);
>>
>> + return 0;
>> +}
>> +
>> +static void __maybe_unused update_runtime_enabled(struct rq *rq)
>> +{
>> + struct cfs_rq *cfs_rq = &rq->cfs;
>> +
>> + /* register cfs_b->quota on the whole tg tree */
>> + rcu_read_lock();
>> + walk_tg_tree_from(cfs_rq->tg, __update_runtime_enabled, tg_nop, (void *)rq);
>> + rcu_read_unlock();
>> }
> Looks ok, performance on hotplug doesn't really matter. Ben, you happy
> with this?
>
>
>
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: [PATCH V2] sched/fair: Fix that tasks are not constrained by cfs_b->quota on hotplug core, when hotplug core is offline and then online.
2016-08-30 13:12 Jeehong Kim
@ 2016-09-01 11:01 ` Peter Zijlstra
2016-09-01 11:52 ` Jeehong Kim
2016-09-01 17:23 ` bsegall
0 siblings, 2 replies; 9+ messages in thread
From: Peter Zijlstra @ 2016-09-01 11:01 UTC (permalink / raw)
To: Jeehong Kim; +Cc: mingo, linux-kernel, kernel-janitors, ezjjilong, Ben Segall
You forgot to Cc Ben, who gave you feedback on v1, which is rather poor
style. Also, I don't see how kernel-janitors is relevant to this patch.
This is very much not a janitorial thing.
(also, why send it twice?)
On Tue, Aug 30, 2016 at 10:12:40PM +0900, Jeehong Kim wrote:
> In case that CONFIG_HOTPLUG_CPU and CONFIG_CFS_BANDWIDTH is turned on
> and tasks in bandwidth controlled task group run on hotplug core,
> the tasks are not controlled by cfs_b->quota when hotplug core is offline
> and then online. The remaining tasks in task group consume all of
> cfs_b->quota on other cores.
>
> The cause of this problem is described as below:
>
> 1. When hotplug core is offline while tasks in task group run
> on hotplug core, unregister_fair_sched_group() deletes
> leaf_cfs_rq_list of tg->cfs_rq[cpu] from &rq_of(cfs_rq)->leaf_cfs_rq_list.
>
> 2. Then, when hotplug core is online, update_runtime_enabled()
> registers cfs_b->quota on cfs_rq->runtime_enabled of all leaf cfs_rq
> on runqueue. However, because this is before enqueue_entity() adds
> &cfs_rq->leaf_cfs_rq_list on &rq_of(cfs_rq)->leaf_cfs_rq_list,
> cfs->quota is not register on cfs_rq->runtime_enabled.
>
> To resolve this problem, this patch makes update_runtime_enabled()
> registers cfs_b->quota by using walk_tg_tree_from().
> +static int __maybe_unused __update_runtime_enabled(struct task_group *tg, void *data)
> {
> + struct rq *rq = data;
> + struct cfs_rq *cfs_rq = tg->cfs_rq[cpu_of(rq)];
> + struct cfs_bandwidth *cfs_b = &cfs_rq->tg->cfs_bandwidth;
>
> + raw_spin_lock(&cfs_b->lock);
> + cfs_rq->runtime_enabled = cfs_b->quota != RUNTIME_INF;
> + raw_spin_unlock(&cfs_b->lock);
>
> + return 0;
> +}
> +
> +static void __maybe_unused update_runtime_enabled(struct rq *rq)
> +{
> + struct cfs_rq *cfs_rq = &rq->cfs;
> +
> + /* register cfs_b->quota on the whole tg tree */
> + rcu_read_lock();
> + walk_tg_tree_from(cfs_rq->tg, __update_runtime_enabled, tg_nop, (void *)rq);
> + rcu_read_unlock();
> }
Looks ok, performance on hotplug doesn't really matter. Ben, you happy
with this?
^ permalink raw reply [flat|nested] 9+ messages in thread
* [PATCH V2] sched/fair: Fix that tasks are not constrained by cfs_b->quota on hotplug core, when hotplug core is offline and then online.
@ 2016-08-30 13:12 Jeehong Kim
2016-09-01 11:01 ` Peter Zijlstra
0 siblings, 1 reply; 9+ messages in thread
From: Jeehong Kim @ 2016-08-30 13:12 UTC (permalink / raw)
To: mingo; +Cc: peterz, linux-kernel, kernel-janitors, ezjjilong, Jeehong Kim
In case that CONFIG_HOTPLUG_CPU and CONFIG_CFS_BANDWIDTH is turned on
and tasks in bandwidth controlled task group run on hotplug core,
the tasks are not controlled by cfs_b->quota when hotplug core is offline
and then online. The remaining tasks in task group consume all of
cfs_b->quota on other cores.
The cause of this problem is described as below:
1. When hotplug core is offline while tasks in task group run
on hotplug core, unregister_fair_sched_group() deletes
leaf_cfs_rq_list of tg->cfs_rq[cpu] from &rq_of(cfs_rq)->leaf_cfs_rq_list.
2. Then, when hotplug core is online, update_runtime_enabled()
registers cfs_b->quota on cfs_rq->runtime_enabled of all leaf cfs_rq
on runqueue. However, because this is before enqueue_entity() adds
&cfs_rq->leaf_cfs_rq_list on &rq_of(cfs_rq)->leaf_cfs_rq_list,
cfs->quota is not register on cfs_rq->runtime_enabled.
To resolve this problem, this patch makes update_runtime_enabled()
registers cfs_b->quota by using walk_tg_tree_from().
Signed-off-by: Jeehong Kim <jhez.kim@samsung.com>
---
V2: update_runtime_enabled() walks the whole tg tree,
instead of register quota on every enqueue.
kernel/sched/fair.c | 32 +++++++++++++++++++++-----------
1 file changed, 21 insertions(+), 11 deletions(-)
diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
index 6488815..63b376d 100644
--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -4373,17 +4373,27 @@ static void destroy_cfs_bandwidth(struct cfs_bandwidth *cfs_b)
hrtimer_cancel(&cfs_b->slack_timer);
}
-static void __maybe_unused update_runtime_enabled(struct rq *rq)
+static int __maybe_unused __update_runtime_enabled(struct task_group *tg, void *data)
{
- struct cfs_rq *cfs_rq;
+ struct rq *rq = data;
+ struct cfs_rq *cfs_rq = tg->cfs_rq[cpu_of(rq)];
+ struct cfs_bandwidth *cfs_b = &cfs_rq->tg->cfs_bandwidth;
- for_each_leaf_cfs_rq(rq, cfs_rq) {
- struct cfs_bandwidth *cfs_b = &cfs_rq->tg->cfs_bandwidth;
+ raw_spin_lock(&cfs_b->lock);
+ cfs_rq->runtime_enabled = cfs_b->quota != RUNTIME_INF;
+ raw_spin_unlock(&cfs_b->lock);
- raw_spin_lock(&cfs_b->lock);
- cfs_rq->runtime_enabled = cfs_b->quota != RUNTIME_INF;
- raw_spin_unlock(&cfs_b->lock);
- }
+ return 0;
+}
+
+static void __maybe_unused update_runtime_enabled(struct rq *rq)
+{
+ struct cfs_rq *cfs_rq = &rq->cfs;
+
+ /* register cfs_b->quota on the whole tg tree */
+ rcu_read_lock();
+ walk_tg_tree_from(cfs_rq->tg, __update_runtime_enabled, tg_nop, (void *)rq);
+ rcu_read_unlock();
}
--
^ permalink raw reply related [flat|nested] 9+ messages in thread
* [PATCH V2] sched/fair: Fix that tasks are not constrained by cfs_b->quota on hotplug core, when hotplug core is offline and then online.
@ 2016-08-30 13:09 Jeehong Kim
0 siblings, 0 replies; 9+ messages in thread
From: Jeehong Kim @ 2016-08-30 13:09 UTC (permalink / raw)
To: mingo; +Cc: peterz, linux-kernel, kernel-janitors, ezjjilong, Jeehong Kim
In case that CONFIG_HOTPLUG_CPU and CONFIG_CFS_BANDWIDTH is turned on
and tasks in bandwidth controlled task group run on hotplug core,
the tasks are not controlled by cfs_b->quota when hotplug core is offline
and then online. The remaining tasks in task group consume all of
cfs_b->quota on other cores.
The cause of this problem is described as below:
1. When hotplug core is offline while tasks in task group run
on hotplug core, unregister_fair_sched_group() deletes
leaf_cfs_rq_list of tg->cfs_rq[cpu] from &rq_of(cfs_rq)->leaf_cfs_rq_list.
2. Then, when hotplug core is online, update_runtime_enabled()
registers cfs_b->quota on cfs_rq->runtime_enabled of all leaf cfs_rq
on runqueue. However, because this is before enqueue_entity() adds
&cfs_rq->leaf_cfs_rq_list on &rq_of(cfs_rq)->leaf_cfs_rq_list,
cfs->quota is not register on cfs_rq->runtime_enabled.
To resolve this problem, this patch makes update_runtime_enabled()
registers cfs_b->quota by using walk_tg_tree_from().
Signed-off-by: Jeehong Kim <jhez.kim@samsung.com>
---
kernel/sched/fair.c | 32 +++++++++++++++++++++-----------
1 file changed, 21 insertions(+), 11 deletions(-)
diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
index 6488815..63b376d 100644
--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -4373,17 +4373,27 @@ static void destroy_cfs_bandwidth(struct cfs_bandwidth *cfs_b)
hrtimer_cancel(&cfs_b->slack_timer);
}
-static void __maybe_unused update_runtime_enabled(struct rq *rq)
+static int __maybe_unused __update_runtime_enabled(struct task_group *tg, void *data)
{
- struct cfs_rq *cfs_rq;
+ struct rq *rq = data;
+ struct cfs_rq *cfs_rq = tg->cfs_rq[cpu_of(rq)];
+ struct cfs_bandwidth *cfs_b = &cfs_rq->tg->cfs_bandwidth;
- for_each_leaf_cfs_rq(rq, cfs_rq) {
- struct cfs_bandwidth *cfs_b = &cfs_rq->tg->cfs_bandwidth;
+ raw_spin_lock(&cfs_b->lock);
+ cfs_rq->runtime_enabled = cfs_b->quota != RUNTIME_INF;
+ raw_spin_unlock(&cfs_b->lock);
- raw_spin_lock(&cfs_b->lock);
- cfs_rq->runtime_enabled = cfs_b->quota != RUNTIME_INF;
- raw_spin_unlock(&cfs_b->lock);
- }
+ return 0;
+}
+
+static void __maybe_unused update_runtime_enabled(struct rq *rq)
+{
+ struct cfs_rq *cfs_rq = &rq->cfs;
+
+ /* register cfs_b->quota on the whole tg tree */
+ rcu_read_lock();
+ walk_tg_tree_from(cfs_rq->tg, __update_runtime_enabled, tg_nop, (void *)rq);
+ rcu_read_unlock();
}
--
^ permalink raw reply related [flat|nested] 9+ messages in thread
end of thread, other threads:[~2016-09-27 17:13 UTC | newest]
Thread overview: 9+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2016-09-22 7:59 [PATCH V2] sched/fair: Fix that tasks are not constrained by cfs_b->quota on hotplug core, when hotplug core is offline and then online Jeehong Kim
2016-09-22 16:53 ` bsegall
2016-09-27 3:41 ` Jeehong Kim
2016-09-27 17:13 ` bsegall
-- strict thread matches above, loose matches on Subject: below --
2016-08-30 13:12 Jeehong Kim
2016-09-01 11:01 ` Peter Zijlstra
2016-09-01 11:52 ` Jeehong Kim
2016-09-01 17:23 ` bsegall
2016-08-30 13:09 Jeehong Kim
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.