linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [RFC PATCH 1/4] linsched: remove process_all_softirqs() in main loop for accuracy
@ 2012-09-03  3:58 Michael Wang
  0 siblings, 0 replies; 2+ messages in thread
From: Michael Wang @ 2012-09-03  3:58 UTC (permalink / raw)
  To: LKML; +Cc: Paul Turner, Dhaval Giani, Peter Zijlstra

From: Michael Wang <wangyun@linux.vnet.ibm.com>

process_all_softirqs() will handle softirq for all the cpu even
it's not the right timing for them, this will cause inaccuracy.

This patch stop invoke process_all_softirqs(), so softirq will
only be handled after timer interrupt arrived.

Signed-off-by: Michael Wang <wangyun@linux.vnet.ibm.com>
---
 tools/linsched/hrtimer.c |    1 -
 1 files changed, 0 insertions(+), 1 deletions(-)

diff --git a/tools/linsched/hrtimer.c b/tools/linsched/hrtimer.c
index 26be1d8..de88b25 100644
--- a/tools/linsched/hrtimer.c
+++ b/tools/linsched/hrtimer.c
@@ -206,7 +206,6 @@ void linsched_run_sim(int sim_ticks)
                        /* a handler should never leave this state changed */
                        BUG_ON(smp_processor_id() != active_cpu);

-                       process_all_softirqs();
                        linsched_rcu_invoke();

                        BUG_ON(irqs_disabled());


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

* [RFC PATCH 1/4] linsched: remove "process_all_softirqs" in main loop for accuracy
@ 2012-04-16  3:37 Michael Wang
  0 siblings, 0 replies; 2+ messages in thread
From: Michael Wang @ 2012-04-16  3:37 UTC (permalink / raw)
  To: LKML; +Cc: Paul Turner, Dhaval Giani

From: Michael Wang <wangyun@linux.vnet.ibm.com>

"process_all_softirqs" will simulate an extra interrupt which will
cause cpu do extra load balance work.

This will make some cpu more active then others, the unfair will
cause inaccuracy.

Signed-off-by: Michael Wang <wangyun@linux.vnet.ibm.com>
---
 tools/linsched/hrtimer.c |    1 -
 1 files changed, 0 insertions(+), 1 deletions(-)

diff --git a/tools/linsched/hrtimer.c b/tools/linsched/hrtimer.c
index 26be1d8..de88b25 100644
--- a/tools/linsched/hrtimer.c
+++ b/tools/linsched/hrtimer.c
@@ -206,7 +206,6 @@ void linsched_run_sim(int sim_ticks)
 			/* a handler should never leave this state changed */
 			BUG_ON(smp_processor_id() != active_cpu);

-			process_all_softirqs();
 			linsched_rcu_invoke();

 			BUG_ON(irqs_disabled());
-- 
1.7.1


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

end of thread, other threads:[~2012-09-03  3:58 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2012-09-03  3:58 [RFC PATCH 1/4] linsched: remove process_all_softirqs() in main loop for accuracy Michael Wang
  -- strict thread matches above, loose matches on Subject: below --
2012-04-16  3:37 [RFC PATCH 1/4] linsched: remove "process_all_softirqs" " Michael Wang

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).