mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Aubrey Li <aubrey.li@intel.com>
To: tglx@linutronix.de, peterz@infradead.org, rjw@rjwysocki.net,
	len.brown@intel.com, ak@linux.intel.com,
	tim.c.chen@linux.intel.com
Cc: x86@kernel.org, linux-kernel@vger.kernel.org,
	Aubrey Li <aubrey.li@intel.com>
Subject: [RFC PATCH v2 0/8] Introduct cpu idle prediction functionality
Date: Sat, 30 Sep 2017 15:20:26 +0800	[thread overview]
Message-ID: <1506756034-6340-1-git-send-email-aubrey.li@intel.com> (raw)

We found under some latency intensive workloads, short idle periods occurs
very common, then idle entry and exit path starts to dominate, so it's
important to optimize them. To determine the short idle pattern, we need
to figure out how long of the coming idle and the threshold of the short
idle interval.

A cpu idle prediction functionality is introduced in this proposal to catch
the short idle pattern.

Firstly, we check the IRQ timings subsystem, if there is an event
coming soon.
-- https://lwn.net/Articles/691297/

Secondly, we check the idle statistics of scheduler, if it's likely we'll
go into a short idle.
-- https://patchwork.kernel.org/patch/2839221/

Thirdly, we predict the next idle interval by using the prediction
fucntionality in the idle governor if it has.

For the threshold of the short idle interval, we record the timestamps of
the idle entry, and multiply by a tunable parameter at here:
-- /proc/sys/kernel/fast_idle_ratio

We use the output of the idle prediction to skip turning tick off if a
short idle is determined in this proposal. Reprogramming hardware timer
twice(off and on) is expensive for a very short idle. There are some
potential optimizations can be done according to the same indicator.

I observed when system is idle, the idle predictor reports 20/s long idle
and ZERO fast idle on one CPU. And when the workload is running, the idle
predictor reports 72899/s fast idle and ZERO long idle on the same CPU.

Aubrey Li (8):
  cpuidle: menu: extract prediction functionality
  cpuidle: record the overhead of idle entry
  cpuidle: add a new predict interface
  tick/nohz: keep tick on for a fast idle
  timers: keep sleep length updated as needed
  cpuidle: make fast idle threshold tunable
  cpuidle: introduce irq timing to make idle prediction
  cpuidle: introduce run queue average idle to make idle prediction

 drivers/cpuidle/Kconfig          |   1 +
 drivers/cpuidle/cpuidle.c        | 109 +++++++++++++++++++++++++++++++++++++++
 drivers/cpuidle/governors/menu.c |  69 ++++++++++++++++---------
 include/linux/cpuidle.h          |  21 ++++++++
 kernel/sched/idle.c              |  14 ++++-
 kernel/sysctl.c                  |  12 +++++
 kernel/time/tick-sched.c         |   7 +++
 7 files changed, 209 insertions(+), 24 deletions(-)

-- 
2.7.4

             reply	other threads:[~2017-09-30  7:21 UTC|newest]

Thread overview: 42+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-09-30  7:20 Aubrey Li [this message]
2017-09-30  7:20 ` [RFC PATCH v2 1/8] cpuidle: menu: extract " Aubrey Li
2017-10-14  0:26   ` Rafael J. Wysocki
2017-10-16  2:46     ` Li, Aubrey
2017-09-30  7:20 ` [RFC PATCH v2 2/8] cpuidle: record the overhead of idle entry Aubrey Li
2017-10-14  0:35   ` Rafael J. Wysocki
2017-10-16  3:11     ` Li, Aubrey
2017-10-17  0:05       ` Rafael J. Wysocki
2017-10-17  7:04         ` Li, Aubrey
2017-09-30  7:20 ` [RFC PATCH v2 3/8] cpuidle: add a new predict interface Aubrey Li
2017-10-14  0:45   ` Rafael J. Wysocki
2017-10-16  8:04     ` Li, Aubrey
2017-10-14  1:27   ` Rafael J. Wysocki
2017-10-16  9:52     ` Li, Aubrey
2017-09-30  7:20 ` [RFC PATCH v2 4/8] tick/nohz: keep tick on for a fast idle Aubrey Li
2017-10-14  0:51   ` Rafael J. Wysocki
2017-10-16  3:26     ` Li, Aubrey
2017-10-16  4:45       ` Mike Galbraith
2017-10-16  5:34         ` Li, Aubrey
2017-10-16  6:25           ` Mike Galbraith
2017-10-16  6:31             ` Li, Aubrey
2017-09-30  7:20 ` [RFC PATCH v2 5/8] timers: keep sleep length updated as needed Aubrey Li
2017-10-14  0:56   ` Rafael J. Wysocki
2017-10-16  6:46     ` Li, Aubrey
2017-10-16 23:58       ` Rafael J. Wysocki
2017-10-17  6:10         ` Li, Aubrey
2017-09-30  7:20 ` [RFC PATCH v2 6/8] cpuidle: make fast idle threshold tunable Aubrey Li
2017-10-14  0:59   ` Rafael J. Wysocki
2017-10-16  6:00     ` Li, Aubrey
2017-10-17  0:01       ` Rafael J. Wysocki
2017-10-17  6:12         ` Li, Aubrey
2017-09-30  7:20 ` [RFC PATCH v2 7/8] cpuidle: introduce irq timing to make idle prediction Aubrey Li
2017-10-14  1:01   ` Rafael J. Wysocki
2017-10-16  6:03     ` Li, Aubrey
2017-09-30  7:20 ` [RFC PATCH v2 8/8] cpuidle: introduce run queue average idle " Aubrey Li
2017-10-14  1:02   ` Rafael J. Wysocki
2017-10-14  1:14 ` [RFC PATCH v2 0/8] Introduct cpu idle prediction functionality Rafael J. Wysocki
2017-10-16  7:44   ` Li, Aubrey
2017-10-17  0:07     ` Rafael J. Wysocki
2017-10-17  7:32       ` Li, Aubrey
2017-11-30  1:00 ` Li, Aubrey
2017-11-30  1:37   ` Rafael J. Wysocki

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1506756034-6340-1-git-send-email-aubrey.li@intel.com \
    --to=aubrey.li@intel.com \
    --cc=ak@linux.intel.com \
    --cc=len.brown@intel.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=peterz@infradead.org \
    --cc=rjw@rjwysocki.net \
    --cc=tglx@linutronix.de \
    --cc=tim.c.chen@linux.intel.com \
    --cc=x86@kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
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®