Re: [PATCH] sched/deadline: dome some cleanup for push_dl_task()
On Fri, Jul 24, 2020 at 09:14:26AM +0200, Juri Lelli wrote:
> Hi,
>
> On 07/07/20 00:04, Peng Liu wrote:
> > 'commit 840d719604b0 ("sched/deadline: Update rq_clock of later_rq when
> > pushing a task")'
> > introduced the update_rq_clock() to fix the "used-before-update" bug.
> >
> > 'commit f4904815f97a ("sched/deadline: Fix double accounting of rq/running
> > bw in push & pull")'
> > took away the bug source(add_running_bw()).
> >
> > We no longer need to update rq_clock in advance, let activate_task()
> > worry about that.
> >
> > Signed-off-by: Peng Liu
> > ---
> > kernel/sched/deadline.c | 8 +---
> > 1 file changed, 1 insertion(+), 7 deletions(-)
> >
> > diff --git a/kernel/sched/deadline.c b/kernel/sched/deadline.c
> > index 504d2f51b0d6..c3fa11f84d93 100644
> > --- a/kernel/sched/deadline.c
> > +++ b/kernel/sched/deadline.c
> > @@ -2104,13 +2104,7 @@ static int push_dl_task(struct rq *rq)
> >
> > deactivate_task(rq, next_task, 0);
> > set_task_cpu(next_task, later_rq->cpu);
> > -
> > - /*
> > -* Update the later_rq clock here, because the clock is used
> > -* by the cpufreq_update_util() inside __add_running_bw().
> > -*/
> > - update_rq_clock(later_rq);
> > - activate_task(later_rq, next_task, ENQUEUE_NOCLOCK);
> > + activate_task(later_rq, next_task, 0);
> > ret = 1;
>
> The change looks good to me, since now add_running_bw() is called later
> by enqueue_task_dl(), but rq_clock has already been updated by core's
> enqueue_task().
>
Thanks for your time.
> Daniel, Dietmar, a second pair of eyes (since you authored the commits
> above)?
>
> I'd chage subject to something like "sched/deadline: Stop updating
> rq_clock before pushing a task".
>
> Thanks,
>
> Juri
>
Re: [PATCH] sched/deadline: dome some cleanup for push_dl_task()
On 24/07/2020 17:31, Daniel Bristot de Oliveira wrote:
> On 7/24/20 9:14 AM, Juri Lelli wrote:
>> Hi,
>>
>> On 07/07/20 00:04, Peng Liu wrote:
>>> 'commit 840d719604b0 ("sched/deadline: Update rq_clock of later_rq when
>>> pushing a task")'
>>> introduced the update_rq_clock() to fix the "used-before-update" bug.
>>>
>>> 'commit f4904815f97a ("sched/deadline: Fix double accounting of rq/running
>>> bw in push & pull")'
>>> took away the bug source(add_running_bw()).
>>>
>>> We no longer need to update rq_clock in advance, let activate_task()
>>> worry about that.
>>>
>>> Signed-off-by: Peng Liu
>>> ---
>>> kernel/sched/deadline.c | 8 +---
>>> 1 file changed, 1 insertion(+), 7 deletions(-)
>>>
>>> diff --git a/kernel/sched/deadline.c b/kernel/sched/deadline.c
>>> index 504d2f51b0d6..c3fa11f84d93 100644
>>> --- a/kernel/sched/deadline.c
>>> +++ b/kernel/sched/deadline.c
>>> @@ -2104,13 +2104,7 @@ static int push_dl_task(struct rq *rq)
>>>
>>> deactivate_task(rq, next_task, 0);
>>> set_task_cpu(next_task, later_rq->cpu);
>>> -
>>> - /*
>>> -* Update the later_rq clock here, because the clock is used
>>> -* by the cpufreq_update_util() inside __add_running_bw().
>>> -*/
>>> - update_rq_clock(later_rq);
>>> - activate_task(later_rq, next_task, ENQUEUE_NOCLOCK);
>>> + activate_task(later_rq, next_task, 0);
>>> ret = 1;
>> The change looks good to me, since now add_running_bw() is called later
>> by enqueue_task_dl(), but rq_clock has already been updated by core's
>> enqueue_task().
>>
>> Daniel, Dietmar, a second pair of eyes (since you authored the commits
>> above)?
>>
>> I'd chage subject to something like "sched/deadline: Stop updating
>> rq_clock before pushing a task".
>
> Looks good to me!
>
> Acked-by: Daniel Bristot de Oliveira
Yes, makes sense to me!
Reviewed-by: Dietmar Eggemann
Re: [PATCH] sched/deadline: dome some cleanup for push_dl_task()
On 7/24/20 9:14 AM, Juri Lelli wrote:
> Hi,
>
> On 07/07/20 00:04, Peng Liu wrote:
>> 'commit 840d719604b0 ("sched/deadline: Update rq_clock of later_rq when
>> pushing a task")'
>> introduced the update_rq_clock() to fix the "used-before-update" bug.
>>
>> 'commit f4904815f97a ("sched/deadline: Fix double accounting of rq/running
>> bw in push & pull")'
>> took away the bug source(add_running_bw()).
>>
>> We no longer need to update rq_clock in advance, let activate_task()
>> worry about that.
>>
>> Signed-off-by: Peng Liu
>> ---
>> kernel/sched/deadline.c | 8 +---
>> 1 file changed, 1 insertion(+), 7 deletions(-)
>>
>> diff --git a/kernel/sched/deadline.c b/kernel/sched/deadline.c
>> index 504d2f51b0d6..c3fa11f84d93 100644
>> --- a/kernel/sched/deadline.c
>> +++ b/kernel/sched/deadline.c
>> @@ -2104,13 +2104,7 @@ static int push_dl_task(struct rq *rq)
>>
>> deactivate_task(rq, next_task, 0);
>> set_task_cpu(next_task, later_rq->cpu);
>> -
>> -/*
>> - * Update the later_rq clock here, because the clock is used
>> - * by the cpufreq_update_util() inside __add_running_bw().
>> - */
>> -update_rq_clock(later_rq);
>> -activate_task(later_rq, next_task, ENQUEUE_NOCLOCK);
>> +activate_task(later_rq, next_task, 0);
>> ret = 1;
> The change looks good to me, since now add_running_bw() is called later
> by enqueue_task_dl(), but rq_clock has already been updated by core's
> enqueue_task().
>
> Daniel, Dietmar, a second pair of eyes (since you authored the commits
> above)?
>
> I'd chage subject to something like "sched/deadline: Stop updating
> rq_clock before pushing a task".
Looks good to me!
Acked-by: Daniel Bristot de Oliveira
Thanks
-- Daniel
Re: [PATCH] sched/deadline: dome some cleanup for push_dl_task()
Hi,
On 07/07/20 00:04, Peng Liu wrote:
> 'commit 840d719604b0 ("sched/deadline: Update rq_clock of later_rq when
> pushing a task")'
> introduced the update_rq_clock() to fix the "used-before-update" bug.
>
> 'commit f4904815f97a ("sched/deadline: Fix double accounting of rq/running bw
> in push & pull")'
> took away the bug source(add_running_bw()).
>
> We no longer need to update rq_clock in advance, let activate_task()
> worry about that.
>
> Signed-off-by: Peng Liu
> ---
> kernel/sched/deadline.c | 8 +---
> 1 file changed, 1 insertion(+), 7 deletions(-)
>
> diff --git a/kernel/sched/deadline.c b/kernel/sched/deadline.c
> index 504d2f51b0d6..c3fa11f84d93 100644
> --- a/kernel/sched/deadline.c
> +++ b/kernel/sched/deadline.c
> @@ -2104,13 +2104,7 @@ static int push_dl_task(struct rq *rq)
>
> deactivate_task(rq, next_task, 0);
> set_task_cpu(next_task, later_rq->cpu);
> -
> - /*
> - * Update the later_rq clock here, because the clock is used
> - * by the cpufreq_update_util() inside __add_running_bw().
> - */
> - update_rq_clock(later_rq);
> - activate_task(later_rq, next_task, ENQUEUE_NOCLOCK);
> + activate_task(later_rq, next_task, 0);
> ret = 1;
The change looks good to me, since now add_running_bw() is called later
by enqueue_task_dl(), but rq_clock has already been updated by core's
enqueue_task().
Daniel, Dietmar, a second pair of eyes (since you authored the commits
above)?
I'd chage subject to something like "sched/deadline: Stop updating
rq_clock before pushing a task".
Thanks,
Juri
[PATCH] sched/deadline: dome some cleanup for push_dl_task()
'commit 840d719604b0 ("sched/deadline: Update rq_clock of later_rq when pushing
a task")'
introduced the update_rq_clock() to fix the "used-before-update" bug.
'commit f4904815f97a ("sched/deadline: Fix double accounting of rq/running bw
in push & pull")'
took away the bug source(add_running_bw()).
We no longer need to update rq_clock in advance, let activate_task()
worry about that.
Signed-off-by: Peng Liu
---
kernel/sched/deadline.c | 8 +---
1 file changed, 1 insertion(+), 7 deletions(-)
diff --git a/kernel/sched/deadline.c b/kernel/sched/deadline.c
index 504d2f51b0d6..c3fa11f84d93 100644
--- a/kernel/sched/deadline.c
+++ b/kernel/sched/deadline.c
@@ -2104,13 +2104,7 @@ static int push_dl_task(struct rq *rq)
deactivate_task(rq, next_task, 0);
set_task_cpu(next_task, later_rq->cpu);
-
- /*
-* Update the later_rq clock here, because the clock is used
-* by the cpufreq_update_util() inside __add_running_bw().
-*/
- update_rq_clock(later_rq);
- activate_task(later_rq, next_task, ENQUEUE_NOCLOCK);
+ activate_task(later_rq, next_task, 0);
ret = 1;
resched_curr(later_rq);
--
2.20.1
