[PATCH 0/2] sched/core: Cleanup leftovers in __schedule()

[PATCH 0/2] sched/core: Cleanup leftovers in __schedule()

[Alexander Gordeev]

*** BLURB HERE ***

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

 kernel/sched/core.c         | 7 ++-----
 kernel/workqueue.c          | 4 ++--
 kernel/workqueue_internal.h | 2 +-
 3 files changed, 5 insertions(+), 8 deletions(-)

-- 
1.8.3.1

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

[Alexander Gordeev]

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.

Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Alexander Gordeev <agordeev@redhat.com>
---
 kernel/sched/core.c         | 2 +-
 kernel/workqueue.c          | 4 ++--
 kernel/workqueue_internal.h | 2 +-
 3 files changed, 4 insertions(+), 4 deletions(-)

diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index 9503d59..4e56a4d 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -3257,7 +3257,7 @@ 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..967a1ab 100644
--- a/kernel/workqueue.c
+++ b/kernel/workqueue.c
@@ -870,7 +870,7 @@ 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 +886,7 @@ 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 != 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

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

[kbuild test robot]

[-- Attachment #1: Type: text/plain, Size: 3368 bytes --]

Hi Alexander,

[auto build test WARNING on wq/for-next]
[also build test WARNING on v4.5-rc6 next-20160226]
[cannot apply to tip/sched/core]
[if your patch is applied to the wrong git tree, please drop us a note to help improving the system]

url:    https://github.com/0day-ci/linux/commits/Alexander-Gordeev/sched-core-Cleanup-leftovers-in-__schedule/20160229-004522
base:   https://git.kernel.org/pub/scm/linux/kernel/git/tj/wq.git for-next
reproduce: make htmldocs

All warnings (new ones prefixed by >>):

   include/linux/init.h:1: warning: no structured comments found
>> kernel/workqueue.c:874: warning: Excess function parameter 'cpu' description in 'wq_worker_sleeping'
   kernel/sys.c:1: warning: no structured comments found
   drivers/dma-buf/seqno-fence.c:1: warning: no structured comments found
   drivers/dma-buf/reservation.c:1: warning: no structured comments found
   include/linux/reservation.h:1: warning: no structured comments found
   include/linux/spi/spi.h:540: warning: No description found for parameter 'max_transfer_size'

vim +874 kernel/workqueue.c

0fcb78c22f Rolf Eike Beer    2006-07-30  858  /**
e22bee782b Tejun Heo         2010-06-29  859   * wq_worker_sleeping - a worker is going to sleep
e22bee782b Tejun Heo         2010-06-29  860   * @task: task going to sleep
e22bee782b Tejun Heo         2010-06-29  861   * @cpu: CPU in question, must be the current CPU number
^1da177e4c Linus Torvalds    2005-04-16  862   *
e22bee782b Tejun Heo         2010-06-29  863   * This function is called during schedule() when a busy worker is
e22bee782b Tejun Heo         2010-06-29  864   * going to sleep.  Worker on the same cpu can be woken up by
e22bee782b Tejun Heo         2010-06-29  865   * returning pointer to its task.
^1da177e4c Linus Torvalds    2005-04-16  866   *
e22bee782b Tejun Heo         2010-06-29  867   * CONTEXT:
e22bee782b Tejun Heo         2010-06-29  868   * spin_lock_irq(rq->lock)
^1da177e4c Linus Torvalds    2005-04-16  869   *
d185af300f Yacine Belkadi    2013-07-31  870   * Return:
e22bee782b Tejun Heo         2010-06-29  871   * Worker task on @cpu to wake up, %NULL if none.
^1da177e4c Linus Torvalds    2005-04-16  872   */
479b9d020d Alexander Gordeev 2016-02-28  873  struct task_struct *wq_worker_sleeping(struct task_struct *task)
^1da177e4c Linus Torvalds    2005-04-16 @874  {
e22bee782b Tejun Heo         2010-06-29  875  	struct worker *worker = kthread_data(task), *to_wakeup = NULL;
111c225a5f Tejun Heo         2013-01-17  876  	struct worker_pool *pool;
f293ea9200 Oleg Nesterov     2007-05-09  877  
111c225a5f Tejun Heo         2013-01-17  878  	/*
111c225a5f Tejun Heo         2013-01-17  879  	 * Rescuers, which may not have all the fields set up like normal
111c225a5f Tejun Heo         2013-01-17  880  	 * workers, also reach here, let's not access anything before
111c225a5f Tejun Heo         2013-01-17  881  	 * checking NOT_RUNNING.
111c225a5f Tejun Heo         2013-01-17  882  	 */

:::::: The code at line 874 was first introduced by commit
:::::: 1da177e4c3f41524e886b7f1b8a0c1fc7321cac2 Linux-2.6.12-rc2

:::::: TO: Linus Torvalds <torvalds@ppc970.osdl.org>
:::::: CC: Linus Torvalds <torvalds@ppc970.osdl.org>

---
0-DAY kernel test infrastructure                Open Source Technology Center
https://lists.01.org/pipermail/kbuild-all                   Intel Corporation

[-- Attachment #2: .config.gz --]
[-- Type: application/octet-stream, Size: 6238 bytes --]

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

[Oleg Nesterov]

On 02/28, Alexander Gordeev wrote:
>
> -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 +886,7 @@ 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 != smp_processor_id()))
>  		return NULL;

I think both changes are fine, but perhaps it would be better to keep
raw_smp_processor_id(). Unless you did this on purpose, but then it makes
sense to add a note into the changelog.

Oleg.

[PATCH 2/2] sched/core: Get rid of 'cpu' variable in __schedule()

[Alexander Gordeev]

'cpu' variable is only used to acquire the current runqueue.
By using this_rq() instead, we can get rid of that variable.

Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Alexander Gordeev <agordeev@redhat.com>
---
 kernel/sched/core.c | 5 +----
 1 file changed, 1 insertion(+), 4 deletions(-)

diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index 4e56a4d..82206f0 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -3205,10 +3205,8 @@ 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;
 
 	/*
@@ -3280,7 +3278,6 @@ static void __sched notrace __schedule(bool preempt)
 
 		trace_sched_switch(preempt, prev, next);
 		rq = context_switch(rq, prev, next); /* unlocks the rq */
-		cpu = cpu_of(rq);
 	} else {
 		lockdep_unpin_lock(&rq->lock);
 		raw_spin_unlock_irq(&rq->lock);
-- 
1.8.3.1