linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Daniel Bristot de Oliveira <daniel@bristot.me>
To: Steven Rostedt <rostedt@goodmis.org>,
	Clark Williams <williams@redhat.com>
Cc: Daniel Bristot de Oliveira <bristot@redhat.com>,
	Ingo Molnar <mingo@redhat.com>,
	Peter Zijlstra <peterz@infradead.org>,
	Christoph Lameter <cl@linux.com>,
	linux-rt-users <linux-rt-users@vger.kernel.org>,
	LKML <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH] sched/rt: RT_RUNTIME_GREED sched feature
Date: Mon, 7 Nov 2016 19:38:40 +0100	[thread overview]
Message-ID: <5d1b0eb1-6bf5-7148-a7b0-58ef1e0fd37f@bristot.me> (raw)
In-Reply-To: <20161107133046.7e4458c1@gandalf.local.home>



On 11/07/2016 07:30 PM, Steven Rostedt wrote:
>> I'm still reviewing the patch, but I have to wonder why bother with making it a scheduler feature?
>> > 
>> > The SCHED_FIFO definition allows a fifo thread to starve others
>> > because a fifo task will run until it yields. Throttling was added as
>> > a safety valve to allow starved SCHED_OTHER tasks to get some cpu
>> > time.  Adding this unconditionally gets us a safety valve for
>> > throttling a badly written fifo task, but allows the fifo task to
>> > continue to consume cpu cycles if it's not starving anyone. 
>> > 
>> > Or am I missing something that's blazingly obvious?
> Or I say make it the default. If people want the old behavior, they can
> modify SCHED_FEATURES to do so.

I added it as a feature to keep the current behavior by default.
Currently, we have two throttling modes:

With RT_RUNTIME_SHARING (default):
  before throttle, try to borrow some runtime from other CPU.

Without RT_RUNTIME_SHARING:
  throttle the RT task, even if there is nothing else to do.

The problem of the first is that an CPU easily borrow enough runtime to
make the spin-rt-task to run forever, allowing the starvation of the
non-rt-tasks, hence invalidating the mechanism.

The problem of the second is that (with the default values) the CPU will
be idle 5% of the time.

IMHO, the balanced behavior is using GREED option + without
RT_RUNTIME_SHARING: the non-rt tasks will be able to run, while avoiding
CPU going idle.

We can turn it by default setting default options.

Moreover, AFAICS, these sched options are static keys, so they are very
very low overhead conditions.

-- Daniel

  reply	other threads:[~2016-11-07 18:44 UTC|newest]

Thread overview: 37+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-11-07  8:17 [PATCH] sched/rt: RT_RUNTIME_GREED sched feature Daniel Bristot de Oliveira
2016-11-07 10:31 ` Tommaso Cucinotta
2016-11-07 13:51   ` Daniel Bristot de Oliveira
2016-11-07 18:03     ` Tommaso Cucinotta
2016-11-07 18:06       ` Luca Abeni
2016-11-08  7:55     ` luca abeni
2016-11-08 10:30     ` Juri Lelli
2016-11-07 16:55 ` Christoph Lameter
2016-11-07 18:32   ` Steven Rostedt
2016-11-07 18:49     ` Daniel Bristot de Oliveira
2016-11-07 19:16       ` Steven Rostedt
2016-11-07 19:30     ` Christoph Lameter
2016-11-07 19:47       ` Steven Rostedt
2016-11-07 19:54         ` Christoph Lameter
2016-11-07 20:00           ` Steven Rostedt
2016-11-07 20:06             ` Daniel Bristot de Oliveira
2016-11-07 20:16               ` Steven Rostedt
2016-11-07 20:33                 ` Daniel Bristot de Oliveira
2016-11-07 20:44                   ` Steven Rostedt
2016-11-08  9:22                     ` [PATCH] sched/rt: Change default setup for RT THROTTLING Daniel Bristot de Oliveira
2016-11-08 23:42                 ` [PATCH] sched/rt: RT_RUNTIME_GREED sched feature Christoph Lameter
2016-11-07 18:22 ` Clark Williams
2016-11-07 18:30   ` Steven Rostedt
2016-11-07 18:38     ` Daniel Bristot de Oliveira [this message]
2016-11-07 18:39     ` Clark Williams
2016-11-08 11:59 ` Peter Zijlstra
2016-11-08 14:07   ` Steven Rostedt
2016-11-08 16:51     ` Peter Zijlstra
2016-11-08 17:17       ` Steven Rostedt
2016-11-08 18:05         ` Peter Zijlstra
2016-11-08 19:29           ` Daniel Bristot de Oliveira
2016-11-08 19:50             ` Peter Zijlstra
2016-11-09 13:33               ` Daniel Bristot de Oliveira
     [not found]                 ` <CAA7rmPF0nQb9721MQWurRCy7E3X46hAy2qV=joK=z5U-t70NOg@mail.gmail.com>
2016-11-11 18:46                   ` Christoph Lameter
2016-11-11 22:53                     ` Peter Zijlstra
2016-11-13 18:53                       ` Christoph Lameter
2016-11-14  9:20                         ` Peter Zijlstra

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=5d1b0eb1-6bf5-7148-a7b0-58ef1e0fd37f@bristot.me \
    --to=daniel@bristot.me \
    --cc=bristot@redhat.com \
    --cc=cl@linux.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-rt-users@vger.kernel.org \
    --cc=mingo@redhat.com \
    --cc=peterz@infradead.org \
    --cc=rostedt@goodmis.org \
    --cc=williams@redhat.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).