mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
* [PATCH]kernel/sched_fair.c: task_new_fair() have call update_curr() twice
@ 2009-04-08  2:10 figo.zhang
  2009-04-08  9:54 ` Peter Zijlstra
  0 siblings, 1 reply; 2+ messages in thread
From: figo.zhang @ 2009-04-08  2:10 UTC (permalink / raw)
  To: mingo; +Cc: linux-kernel

The task_new_fair() will call update_curr() twice, one is in
task_new_fair(), and other one in
enqueue_task_fair()->enqueue_entity(), they do the same,  so  it
should be need one time.

Signed-off-by: figo.zhang <figo.zhang@kolorific.com>


diff --git a/kernel/sched_fair.c b/kernel/sched_fair.c
index 3816f21..d5913bb 100644
--- a/kernel/sched_fair.c
+++ b/kernel/sched_fair.c
@@ -1713,7 +1713,7 @@ static void task_new_fair(struct rq *rq, struct
task_struct *p) 
 	sched_info_queued(p);
 
-	update_curr(cfs_rq);
+
 	place_entity(cfs_rq, se, 1);
 
 	/* 'curr' will be NULL if the child belongs to a different
group */

^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: [PATCH]kernel/sched_fair.c: task_new_fair() have call update_curr() twice
  2009-04-08  2:10 [PATCH]kernel/sched_fair.c: task_new_fair() have call update_curr() twice figo.zhang
@ 2009-04-08  9:54 ` Peter Zijlstra
  0 siblings, 0 replies; 2+ messages in thread
From: Peter Zijlstra @ 2009-04-08  9:54 UTC (permalink / raw)
  To: figo.zhang; +Cc: mingo, linux-kernel

On Wed, 2009-04-08 at 10:10 +0800, figo.zhang wrote:
> The task_new_fair() will call update_curr() twice, one is in
> task_new_fair(), and other one in
> enqueue_task_fair()->enqueue_entity(), they do the same,  so  it
> should be need one time.
> 
> Signed-off-by: figo.zhang <figo.zhang@kolorific.com>
> 
> 
> diff --git a/kernel/sched_fair.c b/kernel/sched_fair.c
> index 3816f21..d5913bb 100644
> --- a/kernel/sched_fair.c
> +++ b/kernel/sched_fair.c
> @@ -1713,7 +1713,7 @@ static void task_new_fair(struct rq *rq, struct
> task_struct *p) 
>  	sched_info_queued(p);
>  
> -	update_curr(cfs_rq);
> +
>  	place_entity(cfs_rq, se, 1);
>  
>  	/* 'curr' will be NULL if the child belongs to a different
> group */

Except that update_curr() changes ->vruntime, and place_entity() uses
->vruntime. Therefore the above is not a NOP.



^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2009-04-08  9:53 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2009-04-08  2:10 [PATCH]kernel/sched_fair.c: task_new_fair() have call update_curr() twice figo.zhang
2009-04-08  9:54 ` Peter Zijlstra

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

all inboxes | Powered by JetHome®