linux-kernel - [RFC PATCH v1 11/11] sched/idle: Add a tuning knob to allow changing fast idle threshold

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-Id: <1499650721-5928-12-git-send-email-aubrey.li@intel.com>
Date:   Mon, 10 Jul 2017 09:38:41 +0800
From:   Aubrey Li <aubrey.li@...el.com>
To:     tglx@...utronix.de, peterz@...radead.org, len.brown@...el.com,
        rjw@...ysocki.net, ak@...ux.intel.com, tim.c.chen@...ux.intel.com,
        arjan@...ux.intel.com, paulmck@...ux.vnet.ibm.com,
        yang.zhang.wz@...il.com
Cc:     x86@...nel.org, linux-kernel@...r.kernel.org,
        Aubrey Li <aubrey.li@...ux.intel.com>
Subject: [RFC PATCH v1 11/11] sched/idle: Add a tuning knob to allow changing fast idle threshold

From: Aubrey Li <aubrey.li@...ux.intel.com>

Short idle periods varies from different workload, make the switch
tunable
---
 include/linux/sched/sysctl.h |  1 +
 kernel/sched/idle.c          |  5 +++--
 kernel/sysctl.c              | 10 ++++++++++
 3 files changed, 14 insertions(+), 2 deletions(-)

diff --git a/include/linux/sched/sysctl.h b/include/linux/sched/sysctl.h
index 0f5ecd4..9771506 100644
--- a/include/linux/sched/sysctl.h
+++ b/include/linux/sched/sysctl.h
@@ -65,6 +65,7 @@ extern unsigned int sysctl_sched_autogroup_enabled;
 
 extern int sysctl_sched_rr_timeslice;
 extern int sched_rr_timeslice;
+extern unsigned int sysctl_fast_idle_threshold;
 
 extern int sched_rr_handler(struct ctl_table *table, int write,
 		void __user *buffer, size_t *lenp,
diff --git a/kernel/sched/idle.c b/kernel/sched/idle.c
index 3358db2..555b02d 100644
--- a/kernel/sched/idle.c
+++ b/kernel/sched/idle.c
@@ -287,6 +287,8 @@ static void cpuidle_generic(void)
 	tick_nohz_idle_exit();
 }
 
+unsigned int sysctl_fast_idle_threshold = USEC_PER_SEC / HZ / 2;
+
 /*
  * Generic idle loop implementation
  *
@@ -295,7 +297,6 @@ static void cpuidle_generic(void)
 static void do_idle(void)
 {
 	unsigned int predicted_idle_us;
-	unsigned int short_idle_threshold = jiffies_to_usecs(1) / 2;
 	/*
 	 * If the arch has a polling bit, we maintain an invariant:
 	 *
@@ -309,7 +310,7 @@ static void do_idle(void)
 
 	predicted_idle_us = cpuidle_predict();
 
-	if (likely(predicted_idle_us < short_idle_threshold))
+	if (likely(predicted_idle_us < sysctl_fast_idle_threshold))
 		cpuidle_fast();
 	else
 		cpuidle_generic();
diff --git a/kernel/sysctl.c b/kernel/sysctl.c
index 8c8714f..0acf81d 100644
--- a/kernel/sysctl.c
+++ b/kernel/sysctl.c
@@ -128,6 +128,7 @@ static int __maybe_unused four = 4;
 static unsigned long one_ul = 1;
 static int one_hundred = 100;
 static int one_thousand = 1000;
+static int one_tick_to_us = USEC_PER_SEC / HZ;
 #ifdef CONFIG_PRINTK
 static int ten_thousand = 10000;
 #endif
@@ -1201,6 +1202,15 @@ static struct ctl_table kern_table[] = {
 		.extra2		= &one,
 	},
 #endif
+	{
+		.procname       = "sched_fast_idle_threshold",
+		.data           = &sysctl_fast_idle_threshold,
+		.maxlen         = sizeof(unsigned int),
+		.mode           = 0644,
+		.proc_handler   = proc_dointvec_minmax,
+		.extra1         = &zero,
+		.extra2         = &one_tick_to_us,
+	},
 	{ }
 };
 
-- 
2.7.4