From: Scott Wood <scott@timesys.com>
To: Ingo Molnar <mingo@elte.hu>
Cc: Scott Wood <scott@timesys.com>, Lee Revell <rlrevell@joe-job.com>,
Andrew Morton <akpm@osdl.org>,
linux-audio-dev@music.columbia.edu, arjanv@redhat.com,
linux-kernel <linux-kernel@vger.kernel.org>,
"La Monte H.P. Yarroll" <piggy@timesys.com>
Subject: Re: [linux-audio-dev] Re: [announce] [patch] Voluntary Kernel Preemption Patch
Date: Thu, 22 Jul 2004 14:36:18 -0400 [thread overview]
Message-ID: <20040722183618.GB4774@yoda.timesys> (raw)
In-Reply-To: <20040722095111.GA13125@elte.hu>
On Thu, Jul 22, 2004 at 11:51:11AM +0200, Ingo Molnar wrote:
> * Scott Wood <scott@timesys.com> wrote:
> > I meant the current split between immediate-context softirqs (which
> > are repesented in the patch by the high-priority ksoftirqd) and the
> > low-priority thread which is used to avoid starvation while allowing
> > softirqs to continue running if the system's otherwise more or less
> > idle.
>
> ok, i understand what you are trying to do. I dont think it makes much
> sense to preserve the throttling property of the current
> immediate/ksoftirqd processing. It was really an ad-hoc way to keep
> softirqs from monopolizing the CPU. The simplest solution for softirq
> deferral is to push it all into ksoftirqd and then let users change the
> priority/policy of ksoftirqd.
Ideally, yes. However, there isn't currently a scheduling policy that
allows the softirq thread to run as a moderately high priority
realtime thread for a short period of time, and drop it to a low
priority non-realtime thread if it runs for longer than it "should".
Running it as one high priority non-realtime task would work if all
you want to run are non-realtime tasks and very high priority RT
tasks (which are intended to have higher priority than softirqs).
This is assuming that a high-priority non-RT task will always preempt
a lower priority task except when it has depleted its timeslice; I'm
not familiar enough with the current scheduler to know whether that
is the case.
> it might make sense to create separate threads for each softirq level
> that exists currently:
>
> HI_SOFTIRQ=0,
> TIMER_SOFTIRQ,
> NET_TX_SOFTIRQ,
> NET_RX_SOFTIRQ,
> SCSI_SOFTIRQ,
> TASKLET_SOFTIRQ
>
> but this doesnt solve the problem either, because the current softirq
> splitup is too opaque - there's no priority-based distinction between
> softirq processing.
Splitting it that way would be nice from the perspective of giving
the user the ability to place more importance on certain types of
softirqs, but it's not what I was trying to do with the high/low
threads.
> so since we cannot do it perfectly i'd go for the simplest approach for
> now: defer to a single ksoftirqd per CPU.
The dual-thread model is about as simple as the immediate/thread
split that is currently there, IMHO, and while not perfect, it is an
improvement over having to pick only one policy/priority given the
current choices of policy.
-Scott
next prev parent reply other threads:[~2004-07-22 18:36 UTC|newest]
Thread overview: 271+ messages / expand[flat|nested] mbox.gz Atom feed top
2004-07-09 18:26 Ingo Molnar
2004-07-09 19:51 ` Christoph Hellwig
2004-07-09 23:50 ` Andrea Arcangeli
2004-07-10 0:52 ` Andrea Arcangeli
2004-07-10 1:02 ` Dave Jones
2004-07-12 10:17 ` Takashi Iwai
2004-07-12 19:36 ` Andrea Arcangeli
2004-07-13 7:43 ` Takashi Iwai
2004-07-10 6:32 ` Arjan van de Ven
2004-07-10 7:57 ` Ingo Molnar
2004-07-10 8:10 ` ismail dönmez
2004-07-10 8:28 ` Ingo Molnar
2004-07-19 10:36 ` [patch] voluntary-preempt 2.6.8-rc2-H4 Ingo Molnar
2004-07-20 1:38 ` Ramón Rey Vicente
2004-07-20 6:02 ` Ingo Molnar
2004-07-10 15:36 ` [announce] [patch] Voluntary Kernel Preemption Patch Andrea Arcangeli
2004-07-10 12:48 ` Ingo Molnar
2004-07-10 15:12 ` Con Kolivas
2004-07-10 15:14 ` Arjan van de Ven
2004-07-10 15:18 ` Con Kolivas
2004-07-11 7:07 ` Zwane Mwaikambo
2004-07-11 7:09 ` Con Kolivas
2004-07-11 7:16 ` [announce] [patch] " Zwane Mwaikambo
2004-07-11 9:18 ` Ingo Molnar
2004-07-11 9:29 ` Con Kolivas
2004-07-10 16:18 ` [announce] [patch] " Con Kolivas
2004-07-11 6:47 ` [ck] " Ingo Molnar
2004-07-11 6:59 ` Con Kolivas
2004-07-11 11:18 ` R. J. Wysocki
2004-07-11 7:09 ` [ck] Re: [announce] [patch] " Zwane Mwaikambo
2004-07-11 14:27 ` Con Kolivas
2004-07-11 14:29 ` Con Kolivas
2004-07-11 14:38 ` Ingo Molnar
2004-07-11 17:29 ` André Goddard Rosa
2004-07-11 17:49 ` Grzegorz Kulewski
2004-07-11 21:20 ` Con Kolivas
2004-07-11 21:25 ` Grzegorz Kulewski
2004-07-11 20:17 ` Ingo Molnar
2004-07-12 8:23 ` Voluntary Preemption + concurent games Aivils
2004-07-12 9:51 ` Con Kolivas
2004-07-12 10:45 ` Lenar Lõhmus
2004-07-12 11:52 ` Aivils
2004-07-12 13:35 ` Nick Piggin
2004-07-12 14:15 ` [ck] Re: [announce] [patch] Voluntary Kernel Preemption Patch Chris Friesen
2004-07-12 14:19 ` Ingo Molnar
2004-07-10 15:58 ` Andrea Arcangeli
2004-07-09 21:13 ` Redeeman
2004-07-10 8:02 ` Ingo Molnar
2004-07-10 8:50 ` Ingo Molnar
2004-07-10 10:55 ` ismail dönmez
2004-07-10 11:26 ` Arjan van de Ven
2004-07-10 12:01 ` ismail dönmez
2004-07-10 12:05 ` Arjan van de Ven
2004-07-10 12:09 ` Nick Piggin
2004-07-10 12:35 ` Ingo Molnar
2004-07-10 12:58 ` ismail dönmez
2004-07-10 13:55 ` Ingo Molnar
2004-07-10 14:45 ` Nick Piggin
2004-07-10 14:57 ` ismail dönmez
2004-07-12 10:48 ` Takashi Iwai
2004-07-12 13:09 ` ismail dönmez
2004-07-12 13:14 ` Takashi Iwai
2004-07-09 21:36 ` Torrey Hoffman
2004-07-10 8:23 ` Ingo Molnar
2004-07-10 1:00 ` Con Kolivas
2004-07-10 1:04 ` Dave Jones
2004-07-10 12:55 ` Ingo Molnar
2004-07-11 5:25 ` Andrew Morton
2004-07-11 9:32 ` Ingo Molnar
2004-07-11 9:45 ` Andrew Morton
2004-07-11 9:50 ` Ingo Molnar
2004-07-11 9:58 ` Andrew Morton
2004-07-11 10:30 ` Ingo Molnar
2004-07-11 10:42 ` Andrew Morton
2004-07-11 10:59 ` Arjan van de Ven
2004-07-11 11:13 ` Andrew Morton
2004-07-11 18:07 ` Mikulas Patocka
2004-07-12 7:08 ` Zwane Mwaikambo
2004-07-19 10:20 ` Ingo Molnar
2004-07-11 23:12 ` Robert Love
2004-07-12 22:40 ` Bill Davidsen
2004-07-12 13:24 ` John Stoffel
2004-07-12 22:56 ` [linux-audio-dev] " Lee Revell
2004-07-12 23:31 ` Andrew Morton
2004-07-12 23:58 ` Paul Davis
2004-07-13 0:06 ` Andrew Morton
2004-07-13 0:21 ` Con Kolivas
2004-07-13 1:01 ` [linux-audio-dev] Re: [announce] [patch] " Lee Revell
2004-07-13 7:49 ` Takashi Iwai
2004-07-13 8:12 ` Lee Revell
2004-07-13 0:01 ` Paul Davis
2004-07-13 0:08 ` Andrew Morton
2004-07-13 16:25 ` Andrea Arcangeli
2004-07-13 18:42 ` Lee Revell
2004-07-13 21:29 ` Andrew Morton
2004-07-13 21:45 ` Lee Revell
2004-07-14 8:51 ` Takashi Iwai
2004-07-14 9:36 ` Lee Revell
2004-07-14 10:48 ` Takashi Iwai
2004-07-14 14:52 ` Lee Revell
2004-07-14 16:29 ` Lee Revell
2004-07-14 10:26 ` Benno Senoner
2004-07-13 18:48 ` Andrew Morton
2004-07-13 21:38 ` Andrea Arcangeli
2004-07-13 21:54 ` Andrew Morton
2004-07-13 22:01 ` Andrea Arcangeli
2004-07-13 22:25 ` Andrew Morton
2004-07-13 22:37 ` Andrea Arcangeli
2004-07-13 22:44 ` Andrew Morton
2004-07-13 22:53 ` Andrea Arcangeli
2004-07-13 23:06 ` Andrew Morton
2004-07-13 23:18 ` Andrea Arcangeli
2004-07-13 23:32 ` Andrew Morton
2004-07-14 0:40 ` Andrea Arcangeli
2004-07-14 0:45 ` Andrew Morton
2004-07-19 11:59 ` Ingo Molnar
2004-09-02 22:03 ` Andrea Arcangeli
2004-09-02 22:20 ` Andrew Morton
2004-09-02 23:03 ` Ingo Molnar
2004-09-02 23:21 ` Ingo Molnar
2004-07-13 7:58 ` Takashi Iwai
2004-07-13 0:17 ` Lee Revell
2004-07-13 0:46 ` Andrew Morton
2004-07-13 2:52 ` Lee Revell
2004-07-13 3:59 ` Andrew Morton
2004-07-13 7:57 ` Lee Revell
2004-07-19 10:48 ` Ingo Molnar
2004-07-19 11:34 ` Ingo Molnar
2004-07-20 5:38 ` Lee Revell
2004-07-20 6:12 ` Ingo Molnar
2004-07-20 6:59 ` Lee Revell
2004-07-20 7:11 ` Ingo Molnar
[not found] ` <200407202011.20558.musical_snake@gmx.de>
2004-07-20 19:56 ` Lee Revell
2004-07-20 23:49 ` Benno Senoner
2004-07-21 0:32 ` [linux-audio-dev] " Lee Revell
2004-07-21 1:25 ` Timothy Miller
2004-07-21 21:38 ` Another dumb question about " Timothy Miller
2004-07-24 0:25 ` Lee Revell
2004-07-25 20:19 ` Timothy Miller
2004-07-25 20:13 ` Lee Revell
2004-07-21 10:53 ` [linux-audio-dev] Re: [announce] [patch] " Florian Schmidt
2004-07-21 11:18 ` Thomas Charbonnel
2004-07-21 12:04 ` pci gfx card / jack xruns / pci latencies (was: Re: [linux-audio-dev] Re: [announce] [patch] Voluntary Kernel Preemption Patch) Florian Schmidt
2004-07-21 17:49 ` [linux-audio-dev] Re: [announce] [patch] Voluntary Kernel Preemption Patch Fernando Pablo Lopez-Lezcano
2004-07-22 22:25 ` Andrew McGregor
2004-07-22 23:16 ` Florian Schmidt
2004-07-21 14:27 ` Benno Senoner
2004-07-20 12:19 ` Jens Axboe
2004-07-20 21:32 ` Lee Revell
2004-07-24 4:07 ` Lee Revell
2004-07-24 5:46 ` Lee Revell
2004-07-24 6:33 ` Lee Revell
2004-07-24 6:43 ` Ingo Molnar
2004-07-24 22:49 ` Lee Revell
2004-07-24 23:30 ` Lee Revell
2004-07-24 11:27 ` max request size 1024KiB by default? Rudo Thomas
2004-07-24 22:58 ` Lee Revell
2004-07-26 19:29 ` Johannes Stezenbach
2004-07-27 5:15 ` Jens Axboe
2004-07-24 12:41 ` [linux-audio-dev] Re: [announce] [patch] Voluntary Kernel Preemption Patch Mark Lord
2004-07-24 4:58 ` Lee Revell
2004-07-13 8:31 ` Lee Revell
2004-07-13 8:43 ` Andrew Morton
2004-07-13 8:53 ` Lee Revell
2004-07-13 9:00 ` Andrew Morton
2004-07-13 9:21 ` Takashi Iwai
2004-07-13 9:25 ` Andrew Morton
2004-07-13 9:33 ` Takashi Iwai
2004-07-13 9:44 ` Andrew Morton
2004-07-13 9:23 ` Lee Revell
2004-07-13 9:29 ` Andrew Morton
2004-07-13 20:37 ` Lee Revell
2004-07-13 22:32 ` Lee Revell
2004-07-13 8:07 ` Takashi Iwai
2004-07-13 22:40 ` Bill Davidsen
2004-07-13 23:47 ` Randy.Dunlap
2004-07-19 10:29 ` Ingo Molnar
2004-07-21 3:27 ` Lee Revell
2004-07-21 7:03 ` Andrew Morton
2004-07-21 4:58 ` Lee Revell
2004-07-21 5:30 ` Ingo Molnar
2004-07-21 6:03 ` Lee Revell
2004-07-21 8:22 ` Ingo Molnar
2004-07-21 8:52 ` Ingo Molnar
2004-07-21 11:32 ` Nick Piggin
2004-07-21 15:44 ` Ingo Molnar
2004-07-22 4:56 ` Nick Piggin
2004-07-22 7:07 ` Ingo Molnar
2004-07-22 10:54 ` Nick Piggin
2004-07-22 16:23 ` Ingo Molnar
2004-07-22 22:11 ` Nick Piggin
2004-07-23 5:47 ` Ingo Molnar
2004-07-23 6:45 ` Nick Piggin
2004-07-23 6:55 ` Ingo Molnar
2004-07-23 7:11 ` Nick Piggin
2004-07-23 7:21 ` Ingo Molnar
2004-07-23 7:28 ` Nick Piggin
2004-07-21 18:34 ` Scott Wood
2004-07-21 18:46 ` Ingo Molnar
2004-07-21 19:56 ` Scott Wood
2004-07-21 21:08 ` Ingo Molnar
2004-07-22 2:47 ` Scott Wood
2004-07-21 21:45 ` Ingo Molnar
2004-07-22 2:28 ` Scott Wood
2004-07-22 7:40 ` Ingo Molnar
2004-07-22 18:53 ` Scott Wood
2004-07-22 19:45 ` Bill Huey
2004-07-28 6:45 ` Ingo Molnar
2004-07-28 20:52 ` Scott Wood
2004-07-29 18:21 ` Ingo Molnar
2004-07-29 18:36 ` Bill Huey
2004-07-29 19:17 ` Scott Wood
2004-07-29 19:25 ` Bill Huey
2004-07-30 2:23 ` Eric St-Laurent
2004-07-29 20:26 ` Pavel Machek
2004-07-29 20:57 ` Scott Wood
2004-07-29 21:00 ` Pavel Machek
2004-07-29 21:07 ` Ingo Molnar
2004-07-21 22:52 ` Nick Piggin
2004-07-21 18:32 ` Scott Wood
2004-07-21 20:40 ` Ingo Molnar
2004-07-21 18:30 ` Scott Wood
2004-07-21 18:43 ` Ingo Molnar
2004-07-22 2:32 ` Scott Wood
2004-07-22 9:51 ` Ingo Molnar
2004-07-22 18:36 ` Scott Wood [this message]
2004-07-22 19:11 ` Robert Love
2004-07-21 21:00 ` Scott Wood
2004-07-21 21:18 ` Ingo Molnar
2004-07-21 22:14 ` Bill Huey
2004-07-21 22:31 ` Bill Huey
2004-07-21 22:37 ` Scott Wood
2004-07-22 10:06 ` [patch] voluntary-preempt-2.6.8-rc2-H9 Ingo Molnar
2004-07-22 16:00 ` scheduling while atomic (Re: voluntary-preempt-2.6.8-rc2-H9) Rudo Thomas
2004-07-22 16:19 ` Ingo Molnar
2004-07-22 17:24 ` voluntary-preempt I0: sluggish feel Rudo Thomas
2004-07-22 17:52 ` Matt Heler
2004-07-22 17:54 ` Rudo Thomas
2004-07-22 18:01 ` Ingo Molnar
2004-07-22 18:08 ` Ingo Molnar
2004-07-22 18:14 ` Ingo Molnar
2004-07-22 18:37 ` Rudo Thomas
2004-07-23 10:42 ` [patch] voluntary-preempt-2.6.8-rc2-I3 Ingo Molnar
2004-07-23 10:58 ` Paolo Ciarrocchi
2004-07-23 11:04 ` Ingo Molnar
2004-07-23 11:28 ` [patch] voluntary-preempt-2.6.8-rc2-I4 Ingo Molnar
2004-07-24 0:16 ` Lee Revell
2004-07-23 11:42 ` [patch] voluntary-preempt-2.6.8-rc2-I3 Paolo Ciarrocchi
2004-07-23 12:00 ` Ingo Molnar
2004-07-23 15:36 ` Dang, Linh [CAR:2X23:EXCH]
2004-07-23 23:48 ` Thomas Charbonnel
2004-07-26 11:10 ` Ingo Molnar
2004-07-26 12:27 ` Thomas Charbonnel
2004-07-22 18:31 ` voluntary-preempt I0: sluggish feel Robert Love
2004-07-22 18:32 ` Rudo Thomas
2004-07-22 18:38 ` Ingo Molnar
2004-07-22 18:00 ` Ingo Molnar
2004-07-21 6:07 ` [linux-audio-dev] Re: [announce] [patch] Voluntary Kernel Preemption Patch Lee Revell
2004-07-21 10:14 ` Roger Luethi
2004-07-13 0:58 ` Jan Depner
2004-07-13 1:00 ` Andrew Morton
2004-07-13 3:31 ` Florin Andrei
2004-07-13 6:31 ` Paul Winkler
2004-07-13 13:26 ` Benno Senoner
2004-07-13 14:50 ` Paul Davis
2004-07-13 19:45 ` Lee Revell
2004-07-13 22:17 ` Bill Davidsen
2004-07-16 4:50 ` Florin Andrei
2004-07-16 12:49 ` Takashi Iwai
2004-07-13 21:53 ` Timothy Miller
2004-07-13 15:16 [linux-audio-dev] " Shane Shrybman
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=20040722183618.GB4774@yoda.timesys \
--to=scott@timesys.com \
--cc=akpm@osdl.org \
--cc=arjanv@redhat.com \
--cc=linux-audio-dev@music.columbia.edu \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@elte.hu \
--cc=piggy@timesys.com \
--cc=rlrevell@joe-job.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
all inboxes | Powered by JetHome®