[RFC PATCH 2/4] linsched: add check on invoke "tick_nohz_irq_exit" in "irq_exit"

From: Michael Wang <wangyun@linux.vnet.ibm.com>
To: LKML <linux-kernel@vger.kernel.org>
Cc: Paul Turner <pjt@google.com>, Dhaval Giani <dhaval.giani@gmail.com>
Subject: [RFC PATCH 2/4] linsched: add check on invoke "tick_nohz_irq_exit" in "irq_exit"
Date: Mon, 16 Apr 2012 11:37:53 +0800	[thread overview]
Message-ID: <4F8B9411.8090600@linux.vnet.ibm.com> (raw)

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

"tick_nohz_irq_exit" will make sure the tick timer reprogram correctly
after cpu enter idle.

With out this check, after the interrupt, tick timer will be enabled
even cpu is still in idle, this will cause inaccuracy if cpu do load
balance in next tick.

Signed-off-by: Michael Wang <wangyun@linux.vnet.ibm.com>
---
 arch/linsched/kernel/irq.c |    4 ++++
 1 files changed, 4 insertions(+), 0 deletions(-)

diff --git a/arch/linsched/kernel/irq.c b/arch/linsched/kernel/irq.c
index a800bbd..d5d16e5 100644
--- a/arch/linsched/kernel/irq.c
+++ b/arch/linsched/kernel/irq.c
@@ -1,6 +1,7 @@
 #include <linux/interrupt.h>
 #include <asm/hardirq.h>
 #include <linux/sched.h>
+#include <linux/tick.h>

 unsigned long linsched_irq_flags = ARCH_IRQ_ENABLED;

@@ -81,6 +82,9 @@ void irq_exit(void)
 	sub_preempt_count(IRQ_EXIT_OFFSET);
 	if (!in_interrupt() && local_softirq_pending())
 		do_softirq();
+
+	if (idle_cpu(smp_processor_id()) && !in_interrupt() && !need_resched())
+		tick_nohz_irq_exit();
 }

 void local_bh_enable_ip(unsigned long ip)
-- 
1.7.1

