2016-03-02 11:53:44

by Alexander Gordeev

[permalink] [raw]
Subject: [PATCH v3 0/2] sched/core: Cleanup leftovers in __schedule()

Changes since v2:
- rebased on top of tip/sched/core

Changes since v1:
- cover letter BLURB added :)
- forgotten Suggested-by Oleg field added to patch 1
- raw_smp_processor_id() returned instead of mistaken smp_processor_id()
- kbuild excess function parameter 'cpu' description warning fixed


Cc: Oleg Nesterov <[email protected]>
Cc: Peter Zijlstra <[email protected]>

Alexander Gordeev (2):
sched/core: Get rid of 'cpu' argument in wq_worker_sleeping()
sched/core: Get rid of 'cpu' local variable in __schedule()

kernel/sched/core.c | 6 ++----
kernel/workqueue.c | 5 ++---
kernel/workqueue_internal.h | 2 +-
3 files changed, 5 insertions(+), 8 deletions(-)

--
1.8.3.1


2016-03-02 11:53:47

by Alexander Gordeev

[permalink] [raw]
Subject: [PATCH v3 2/2] sched/core: Get rid of 'cpu' local variable in __schedule()

This update is a continuation of commit 3223d05 ("sched/core:
Remove dead statement in __schedule()") and completes the
removal of a dead code with a removal of an unnecessary 'cpu'
local variable - as it is only used to acquire the current
runqueue.

Cc: Oleg Nesterov <[email protected]>
Cc: Peter Zijlstra <[email protected]>
Signed-off-by: Alexander Gordeev <[email protected]>
---
kernel/sched/core.c | 4 +---
1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index 735e87c..75e62ef 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -3131,10 +3131,8 @@ static void __sched notrace __schedule(bool preempt)
struct task_struct *prev, *next;
unsigned long *switch_count;
struct rq *rq;
- int cpu;

- cpu = smp_processor_id();
- rq = cpu_rq(cpu);
+ rq = this_rq();
prev = rq->curr;

/*
--
1.8.3.1

2016-03-02 11:54:06

by Alexander Gordeev

[permalink] [raw]
Subject: [PATCH v3 1/2] sched/core: Get rid of 'cpu' argument in wq_worker_sleeping()

Given that wq_worker_sleeping() could only be called for a
CPU it is running on, we do not need passing a CPU ID as an
argument.

Suggested-by: Oleg Nesterov <[email protected]>
Cc: Oleg Nesterov <[email protected]>
Cc: Peter Zijlstra <[email protected]>
Signed-off-by: Alexander Gordeev <[email protected]>
---
kernel/sched/core.c | 2 +-
kernel/workqueue.c | 5 ++---
kernel/workqueue_internal.h | 2 +-
3 files changed, 4 insertions(+), 5 deletions(-)

diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index 423452c..735e87c 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -3183,7 +3183,7 @@ static void __sched notrace __schedule(bool preempt)
if (prev->flags & PF_WQ_WORKER) {
struct task_struct *to_wakeup;

- to_wakeup = wq_worker_sleeping(prev, cpu);
+ to_wakeup = wq_worker_sleeping(prev);
if (to_wakeup)
try_to_wake_up_local(to_wakeup);
}
diff --git a/kernel/workqueue.c b/kernel/workqueue.c
index 7ff5dc7..2fa8d26 100644
--- a/kernel/workqueue.c
+++ b/kernel/workqueue.c
@@ -858,7 +858,6 @@ void wq_worker_waking_up(struct task_struct *task, int cpu)
/**
* wq_worker_sleeping - a worker is going to sleep
* @task: task going to sleep
- * @cpu: CPU in question, must be the current CPU number
*
* This function is called during schedule() when a busy worker is
* going to sleep. Worker on the same cpu can be woken up by
@@ -870,7 +869,7 @@ void wq_worker_waking_up(struct task_struct *task, int cpu)
* Return:
* Worker task on @cpu to wake up, %NULL if none.
*/
-struct task_struct *wq_worker_sleeping(struct task_struct *task, int cpu)
+struct task_struct *wq_worker_sleeping(struct task_struct *task)
{
struct worker *worker = kthread_data(task), *to_wakeup = NULL;
struct worker_pool *pool;
@@ -886,7 +885,7 @@ struct task_struct *wq_worker_sleeping(struct task_struct *task, int cpu)
pool = worker->pool;

/* this can only happen on the local cpu */
- if (WARN_ON_ONCE(cpu != raw_smp_processor_id() || pool->cpu != cpu))
+ if (WARN_ON_ONCE(pool->cpu != raw_smp_processor_id()))
return NULL;

/*
diff --git a/kernel/workqueue_internal.h b/kernel/workqueue_internal.h
index 4521587..8635417 100644
--- a/kernel/workqueue_internal.h
+++ b/kernel/workqueue_internal.h
@@ -69,6 +69,6 @@ static inline struct worker *current_wq_worker(void)
* sched/core.c and workqueue.c.
*/
void wq_worker_waking_up(struct task_struct *task, int cpu);
-struct task_struct *wq_worker_sleeping(struct task_struct *task, int cpu);
+struct task_struct *wq_worker_sleeping(struct task_struct *task);

#endif /* _KERNEL_WORKQUEUE_INTERNAL_H */
--
1.8.3.1

2016-03-02 12:03:21

by Peter Zijlstra

[permalink] [raw]
Subject: Re: [PATCH v3 1/2] sched/core: Get rid of 'cpu' argument in wq_worker_sleeping()


This really should also have been Cc'ed to TJ, doing so now.

On Wed, Mar 02, 2016 at 12:53:31PM +0100, Alexander Gordeev wrote:
> Given that wq_worker_sleeping() could only be called for a
> CPU it is running on, we do not need passing a CPU ID as an
> argument.
>
> Suggested-by: Oleg Nesterov <[email protected]>
> Cc: Oleg Nesterov <[email protected]>
> Cc: Peter Zijlstra <[email protected]>
> Signed-off-by: Alexander Gordeev <[email protected]>
> ---
> kernel/sched/core.c | 2 +-
> kernel/workqueue.c | 5 ++---
> kernel/workqueue_internal.h | 2 +-
> 3 files changed, 4 insertions(+), 5 deletions(-)
>
> diff --git a/kernel/sched/core.c b/kernel/sched/core.c
> index 423452c..735e87c 100644
> --- a/kernel/sched/core.c
> +++ b/kernel/sched/core.c
> @@ -3183,7 +3183,7 @@ static void __sched notrace __schedule(bool preempt)
> if (prev->flags & PF_WQ_WORKER) {
> struct task_struct *to_wakeup;
>
> - to_wakeup = wq_worker_sleeping(prev, cpu);
> + to_wakeup = wq_worker_sleeping(prev);
> if (to_wakeup)
> try_to_wake_up_local(to_wakeup);
> }
> diff --git a/kernel/workqueue.c b/kernel/workqueue.c
> index 7ff5dc7..2fa8d26 100644
> --- a/kernel/workqueue.c
> +++ b/kernel/workqueue.c
> @@ -858,7 +858,6 @@ void wq_worker_waking_up(struct task_struct *task, int cpu)
> /**
> * wq_worker_sleeping - a worker is going to sleep
> * @task: task going to sleep
> - * @cpu: CPU in question, must be the current CPU number
> *
> * This function is called during schedule() when a busy worker is
> * going to sleep. Worker on the same cpu can be woken up by
> @@ -870,7 +869,7 @@ void wq_worker_waking_up(struct task_struct *task, int cpu)
> * Return:
> * Worker task on @cpu to wake up, %NULL if none.
> */
> -struct task_struct *wq_worker_sleeping(struct task_struct *task, int cpu)
> +struct task_struct *wq_worker_sleeping(struct task_struct *task)
> {
> struct worker *worker = kthread_data(task), *to_wakeup = NULL;
> struct worker_pool *pool;
> @@ -886,7 +885,7 @@ struct task_struct *wq_worker_sleeping(struct task_struct *task, int cpu)
> pool = worker->pool;
>
> /* this can only happen on the local cpu */
> - if (WARN_ON_ONCE(cpu != raw_smp_processor_id() || pool->cpu != cpu))
> + if (WARN_ON_ONCE(pool->cpu != raw_smp_processor_id()))
> return NULL;
>
> /*
> diff --git a/kernel/workqueue_internal.h b/kernel/workqueue_internal.h
> index 4521587..8635417 100644
> --- a/kernel/workqueue_internal.h
> +++ b/kernel/workqueue_internal.h
> @@ -69,6 +69,6 @@ static inline struct worker *current_wq_worker(void)
> * sched/core.c and workqueue.c.
> */
> void wq_worker_waking_up(struct task_struct *task, int cpu);
> -struct task_struct *wq_worker_sleeping(struct task_struct *task, int cpu);
> +struct task_struct *wq_worker_sleeping(struct task_struct *task);
>
> #endif /* _KERNEL_WORKQUEUE_INTERNAL_H */
> --
> 1.8.3.1
>

2016-03-02 12:05:05

by Peter Zijlstra

[permalink] [raw]
Subject: Re: [PATCH v3 2/2] sched/core: Get rid of 'cpu' local variable in __schedule()

On Wed, Mar 02, 2016 at 12:53:32PM +0100, Alexander Gordeev wrote:
> This update is a continuation of commit 3223d05 ("sched/core:

Please put

[core]
abbrev = 12

Into your .gitconfig, 7 char SHA1s are no good.

If TJ acks the first, I can fix up this time, so no need to repost just
yet.

2016-03-02 15:28:27

by Tejun Heo

[permalink] [raw]
Subject: Re: [PATCH v3 1/2] sched/core: Get rid of 'cpu' argument in wq_worker_sleeping()

On Wed, Mar 02, 2016 at 01:03:16PM +0100, Peter Zijlstra wrote:
>
> This really should also have been Cc'ed to TJ, doing so now.
>
> On Wed, Mar 02, 2016 at 12:53:31PM +0100, Alexander Gordeev wrote:
> > Given that wq_worker_sleeping() could only be called for a
> > CPU it is running on, we do not need passing a CPU ID as an
> > argument.
> >
> > Suggested-by: Oleg Nesterov <[email protected]>
> > Cc: Oleg Nesterov <[email protected]>
> > Cc: Peter Zijlstra <[email protected]>
> > Signed-off-by: Alexander Gordeev <[email protected]>

Acked-by: Tejun Heo <[email protected]>

I'll route this through wq/for-4.6.

Thanks.

--
tejun