Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > linux.kernel > #1722393 > unrolled thread

[RFC PATCH v2 6/7] KVM guest: introduce smart idle poll algorithm

Started byYang Zhang <yang.zhang.wz@gmail.com>
First post2017-08-29 13:50 +0200
Last post2017-08-29 13:50 +0200
Articles 1 — 1 participant

Back to article view | Back to linux.kernel

This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by below is the oldest one visible, not the original post.


Contents

  [RFC PATCH v2 6/7] KVM guest: introduce smart idle poll algorithm Yang Zhang <yang.zhang.wz@gmail.com> - 2017-08-29 13:50 +0200

#1722393 — [RFC PATCH v2 6/7] KVM guest: introduce smart idle poll algorithm

FromYang Zhang <yang.zhang.wz@gmail.com>
Date2017-08-29 13:50 +0200
Subject[RFC PATCH v2 6/7] KVM guest: introduce smart idle poll algorithm
Message-ID<ujNbP-25K-3@gated-at.bofh.it>
using smart idle poll to reduce the useless poll when system is idle.

Signed-off-by: Yang Zhang <yang.zhang.wz@gmail.com>
Signed-off-by: Quan Xu <quan.xu0@gmail.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: x86@kernel.org
Cc: kvm@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
---
 arch/x86/kernel/kvm.c | 41 +++++++++++++++++++++++++++++++++++++++++
 1 file changed, 41 insertions(+)

diff --git a/arch/x86/kernel/kvm.c b/arch/x86/kernel/kvm.c
index 7d84a02..ffc8307 100644
--- a/arch/x86/kernel/kvm.c
+++ b/arch/x86/kernel/kvm.c
@@ -373,12 +373,53 @@ static void kvm_idle_poll(void)
 	} while (ktime_before(cur, stop));
 }
 
+static unsigned int grow_poll_ns(unsigned int old, unsigned int grow,
+				 unsigned int max)
+{
+	unsigned int val;
+
+	/* set base poll time to 10000ns */
+	if (old == 0 && grow)
+		return 10000;
+
+	val = old * grow;
+	if (val > max)
+		val = max;
+
+	return val;
+}
+
+static unsigned int shrink_poll_ns(unsigned int old, unsigned int shrink)
+{
+	if (shrink == 0)
+		return 0;
+
+	return old / shrink;
+}
+
+static int kvm_idle_update_poll_duration(unsigned long idle_duration)
+{
+	unsigned long poll_duration = this_cpu_read(poll_duration_ns);
+
+	if (poll_duration && idle_duration > poll_threshold_ns)
+		poll_duration = shrink_poll_ns(poll_duration, poll_shrink);
+	else if (poll_duration < poll_threshold_ns &&
+		 idle_duration < poll_threshold_ns)
+		poll_duration = grow_poll_ns(poll_duration, poll_grow,
+					     poll_threshold_ns);
+
+	this_cpu_write(poll_duration_ns, poll_duration);
+
+	return 0;
+}
+
 static void kvm_guest_idle_init(void)
 {
 	if (!kvm_para_available())
 		return;
 
 	pv_idle_ops.poll = kvm_idle_poll;
+	pv_idle_ops.update = kvm_idle_update_poll_duration;
 }
 
 static void kvm_pv_disable_apf(void)
-- 
1.8.3.1

[toc] | [standalone]


Back to top | Article view | linux.kernel


csiph-web