All of lore.kernel.org
 help / color / mirror / Atom feed
From: Peter Zijlstra <peterz@infradead.org>
To: svaidy@linux.vnet.ibm.com
Cc: Youquan Song <youquan.song@intel.com>,
	linux-kernel@vger.kernel.org, mingo@elte.hu, tglx@linutronix.de,
	hpa@zytor.com, akpm@linux-foundation.org, stable@vger.kernel.org,
	suresh.b.siddha@intel.com, arjan@linux.intel.com,
	len.brown@intel.com, anhua.xu@intel.com, chaohong.guo@intel.com,
	Youquan Song <youquan.song@linux.intel.com>,
	Paul Turner <pjt@google.com>
Subject: Re: [PATCH] x86,sched: Fix sched_smt_power_savings totally broken
Date: Mon, 09 Jan 2012 17:13:17 +0100	[thread overview]
Message-ID: <1326125597.2442.90.camel@twins> (raw)
In-Reply-To: <20120109160317.GA29142@dirshya.in.ibm.com>

On Mon, 2012-01-09 at 21:33 +0530, Vaidyanathan Srinivasan wrote:

> Yes, based on the architecture and topology, we do have two sweet
> spots for power vs performance trade offs.  The first level should be
> to reduce power savings with marginal performance impact and second
> one will be to go for the most aggressive power savings.

Colour me unconvinced, cache heavy workloads will suffer greatly from
your 1.

> The first one should generally be recommended as default to have
> a right balance between performance and power savings, while the
> second one should be used for reducing power consumption on
> unimportant workloads or under certain constraints.
> 
> Some example policies:
> 
> sched_powersavings=1:
> 
>         Enable consolidation at MC level
> 
> sched_powersavings=2:
> 
>         Enable aggressive consolidation at MC level and SMT level if
>         available. In case arch can benefit from cross node
>         consolidation, then enable it.

You fail for mentioning MC/SMT..

> Having the above simple split in policy will enable wide adoption
> where the first level can be a recommended default.  Having just
> a boolean enable/disable will mean the end-user will have to decide
> when to turn on and later off for best workload experience.

Picking one of two states is too hard, hence we given them one of three
states to pick from.. How does that make sense?

> Just similar to cpufreq policy of performance, ondemand and powersave.
> They have their unique use cases and this design choice helps us ship
> ondemand as default.

You fail for thinking having multiple cpufreq governors is a good thing.
The result is that they all suck.

  reply	other threads:[~2012-01-09 16:13 UTC|newest]

Thread overview: 39+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-01-09  8:56 [PATCH] x86,sched: Fix sched_smt_power_savings totally broken Youquan Song
2012-01-09 10:06 ` Peter Zijlstra
2012-01-09 10:28   ` Peter Zijlstra
2012-01-09 10:30     ` Peter Zijlstra
2012-01-09 11:00     ` Vaidyanathan Srinivasan
2012-01-09 14:35       ` Peter Zijlstra
2012-01-09 16:03         ` Vaidyanathan Srinivasan
2012-01-09 16:13           ` Peter Zijlstra [this message]
2012-01-09 17:05             ` Vaidyanathan Srinivasan
2012-01-09 14:13     ` Arjan van de Ven
2012-05-18 10:19     ` [tip:sched/core] sched: Remove stale power aware scheduling remnants and dysfunctional knobs tip-bot for Peter Zijlstra
2012-01-10  0:14   ` [PATCH] x86,sched: Fix sched_smt_power_savings totally broken Youquan Song
2012-01-09 11:05     ` Peter Zijlstra
2012-01-10  5:58       ` Youquan Song
2012-01-09 23:52         ` Suresh Siddha
2012-01-10  9:18           ` Ingo Molnar
2012-01-10 14:32             ` Arjan van de Ven
2012-01-10 14:41               ` Peter Zijlstra
2012-01-10 14:54                 ` Arjan van de Ven
2012-01-10 15:32               ` Vincent Guittot
2012-01-10 15:32                 ` Vincent Guittot
2012-01-10 16:49               ` Vaidyanathan Srinivasan
2012-01-10 19:41               ` Ingo Molnar
2012-01-10 19:44                 ` Ingo Molnar
2012-01-10 16:54           ` Youquan Song
2012-01-10 16:51             ` Vaidyanathan Srinivasan
2012-01-10 19:01               ` Suresh Siddha
2012-01-11  3:52                 ` Vaidyanathan Srinivasan
2012-01-11 17:37                   ` Youquan Song
2012-01-10 16:44       ` Vaidyanathan Srinivasan
2012-01-09 11:12     ` Peter Zijlstra
2012-01-09 14:29       ` Vincent Guittot
2012-01-09 14:29         ` Vincent Guittot
2012-01-09 14:46         ` Peter Zijlstra
2012-01-10  2:12           ` Indan Zupancic
2012-01-10  9:26             ` Peter Zijlstra
2012-01-10  1:54         ` Suresh Siddha
2012-01-10  8:08           ` Vincent Guittot
2012-01-09 15:37 ` Greg KH

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=1326125597.2442.90.camel@twins \
    --to=peterz@infradead.org \
    --cc=akpm@linux-foundation.org \
    --cc=anhua.xu@intel.com \
    --cc=arjan@linux.intel.com \
    --cc=chaohong.guo@intel.com \
    --cc=hpa@zytor.com \
    --cc=len.brown@intel.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@elte.hu \
    --cc=pjt@google.com \
    --cc=stable@vger.kernel.org \
    --cc=suresh.b.siddha@intel.com \
    --cc=svaidy@linux.vnet.ibm.com \
    --cc=tglx@linutronix.de \
    --cc=youquan.song@intel.com \
    --cc=youquan.song@linux.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.