From: Randy Dunlap <rdunlap@xenotime.net>
To: arun@linux.vnet.ibm.com
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>,
Benjamin Herrenschmidt <benh@kernel.crashing.org>,
Ingo Molnar <mingo@elte.hu>,
Vaidyanathan Srinivasan <svaidy@linux.vnet.ibm.com>,
Dipankar Sarma <dipankar@in.ibm.com>,
Balbir Singh <balbir@in.ibm.com>,
Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>,
linux-kernel@vger.kernel.org
Subject: Re: [v11 PATCH 1/9]: cpuidle: Design documentation patch
Date: Tue, 09 Feb 2010 09:00:54 -0800 [thread overview]
Message-ID: <4B7194C6.8010306@xenotime.net> (raw)
In-Reply-To: <20100209082914.GB29988@linux.vnet.ibm.com>
On 02/09/10 00:29, Arun R Bharadwaj wrote:
> * Arun R Bharadwaj <arun@linux.vnet.ibm.com> [2010-02-09 13:58:16]:
>
> This patch adds a little information about the redesigned cpuidle
> infrastructure in Documentation/cpuidle/core.txt
>
> Signed-off-by: Arun R Bharadwaj <arun@linux.vnet.ibm.com>
> ---
> Documentation/cpuidle/core.txt | 35 +++++++++++++++++++++++++++++++++++
> 1 file changed, 35 insertions(+)
>
> Index: linux.trees.git/Documentation/cpuidle/core.txt
> ===================================================================
> --- linux.trees.git.orig/Documentation/cpuidle/core.txt
> +++ linux.trees.git/Documentation/cpuidle/core.txt
> @@ -21,3 +21,38 @@ which can be used to switch governors at
> is meant for developer testing only. In normal usage, kernel picks the
> best governor based on governor ratings.
> SEE ALSO: sysfs.txt in this directory.
> +
> +Design:
> +
> +Cpuidle allows for registration of multiple sets of idle routines.
> +The latest registered set is used by cpuidle governors as the current
> +active set to choose the right idle state. This set is managed as a
> +list and each time the newly registered set is added to the head of the
> +list and made the current active set.
> +
> +An example of how this would work on x86 is shown below.
> +
+----------------- -----------------
+| | | |
+| choose b/w | mwait is chosen | mwait |
+| mwait, poll, |-------------------------------------> |(current active|
+| default, c1e | register to cpuidle | set) |
+| | with mwait as the idle routine | |
+----------------- -----------------
What is "b/w" above? "between"? I've see "btw" used a few times as an
abbreviation for between, but I don't recall ever seeing b/w.
OTOH, I have seen b/w used for bandwidth.
> +
> +
> +----------------- -----------------
> +| | | c1, c2, c3 |
> +| ACPI | register to cpuidle | (current) |
> +| discovery |-------------------------------------> |---------------|
> +| | with c1, c2, c3 | mwait |
> +| | as set of idle routines | |
> +----------------- -----------------
> +
> +With this mechanism, a module can register and unregister its set of
> +idle routines at run time in a clean manner.
> +
> +The main idle routine called inside cpu_idle() of every arch is defined in
> +driver/cpuidle/cpuidle.c which would in turn call the idle routine selected
drivers/
> +by the governor. If the CONFIG_CPU_IDLE is disabled, the arch needs to
> +provide an alternate definition for cpuidle_idle_call().
--
~Randy
next prev parent reply other threads:[~2010-02-09 17:01 UTC|newest]
Thread overview: 12+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-02-09 8:28 [v11 PATCH 0/9]: cpuidle: cleanup cpuidle/ introduce cpuidle to POWER Arun R Bharadwaj
2010-02-09 8:29 ` [v11 PATCH 1/9]: cpuidle: Design documentation patch Arun R Bharadwaj
2010-02-09 17:00 ` Randy Dunlap [this message]
2010-02-10 2:13 ` Arun R Bharadwaj
2010-02-09 8:30 ` [v11 PATCH 2/9]: cpuidle: cleanup drivers/cpuidle/cpuidle.c Arun R Bharadwaj
2010-02-09 8:31 ` [v11 PATCH 3/9]: cpuidle: implement a list based approach to register a set of idle routines Arun R Bharadwaj
2010-02-09 8:32 ` [v11 PATCH 4/9]: x86: refactor x86 idle power management code, remove all instances of pm_idle Arun R Bharadwaj
2010-02-09 8:33 ` [v11 PATCH 5/9]: POWER: enable cpuidle for POWER Arun R Bharadwaj
2010-02-09 8:34 ` [v11 PATCH 6/9]: pSeries/cpuidle: refactor pseries idle loops Arun R Bharadwaj
2010-02-09 8:34 ` [v11 PATCH 7/9]: POWER: add a default_idle idle loop for POWER Arun R Bharadwaj
2010-02-09 8:35 ` [v11 PATCH 8/9]: pSeries: implement pSeries processor idle module Arun R Bharadwaj
2010-02-09 8:36 ` [v11 PATCH 9/9]: POWER: Enable default_idle when power_save=off Arun R Bharadwaj
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=4B7194C6.8010306@xenotime.net \
--to=rdunlap@xenotime.net \
--cc=a.p.zijlstra@chello.nl \
--cc=arun@linux.vnet.ibm.com \
--cc=balbir@in.ibm.com \
--cc=benh@kernel.crashing.org \
--cc=dipankar@in.ibm.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@elte.hu \
--cc=svaidy@linux.vnet.ibm.com \
--cc=venkatesh.pallipadi@intel.com \
/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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.