mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Ingo Molnar <mingo@elte.hu>
To: Rene Herman <rene.herman@gmail.com>
Cc: Mike Galbraith <efault@gmx.de>,
	Gene Heskett <gene.heskett@gmail.com>,
	linux-kernel@vger.kernel.org, Con Kolivas <kernel@kolivas.org>,
	Andrew Morton <akpm@linux-foundation.org>
Subject: Re: Ten percent test
Date: Mon, 9 Apr 2007 16:15:16 +0200	[thread overview]
Message-ID: <20070409141516.GB11244@elte.hu> (raw)
In-Reply-To: <461A2E39.8060106@gmail.com>


* Rene Herman <rene.herman@gmail.com> wrote:

> To me, the example rather serves as confirmation of what Kolivas has 
> been saying; endlessly tweaking the tweaks isn't going anywhere.

but ... SD clearly regresses in some areas, so by that logic SD isnt 
going anywhere either?

note that i still like the basic idea about SD, that it is an experiment 
that if the only conceptual focus is on "scheduling fairness", we'll get 
a better scheduler. But for that to work out two things have to be done 
i think:

 - the code actually has to match that stated goal. Right now it
   diverges from it (it is not a "fair" scheduler), and it's not clear
   why.

note that SD at the moment produces ~10% more code in sched.o, and the 
reason is that SD is more complex than the vanilla scheduler. People 
tend to get the impression that SD is simpler, partly because it is a 
net linecount win in sched.c, but many of the removed lines are 
comments.

this "provide fairness" goal is quite important, because if SD's code is 
not only about providing fairness, what is the rest of the logic doing? 
Are they "tweaks", to achieve interactivity? If yes, why are they not 
marked as such? I.e. will we go down the _same_ road again, but this 
time with a much less clearly defined rule for what a "tweak" is?

note that under the interactivity estimator it is not that hard to 
achieve forced "fairness".

So _if_ we accept that scheduling must include a fair dose of heuristics 
(which i tend to think it has to), we are perhaps better off with an 
interactivity design that _accepts_ this fundamental fact and separates 
heuristics from core scheduling. Right now i dont see the SD proponents 
even _accepting_ that even the current SD code does include heuristics.

the other one is:

 - the code has to demonstrate that it can flexibly react to various 
   complaints of regressions.

(I identified a few problem workloads that we tend to care about and i 
havent seen much progress with them - but i really reserve judgement 
about that, given Con's medical condition.)

	Ingo

  parent reply	other threads:[~2007-04-09 14:15 UTC|newest]

Thread overview: 92+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2007-03-28 16:37 [PATCH] sched: staircase deadline misc fixes Con Kolivas
2007-03-28 17:34 ` [ck] " Prakash Punnoor
2007-04-01  6:40   ` Prakash Punnoor
     [not found]     ` <b14e81f00704010724i3155a16en91074ab789416f3d@mail.gmail.com>
2007-04-01 20:03       ` Prakash Punnoor
2007-03-28 18:48 ` Ingo Molnar
2007-03-28 23:44   ` Con Kolivas
2007-03-29  5:50     ` Mike Galbraith
2007-03-29  6:29       ` Mike Galbraith
2007-03-29  6:54         ` Mike Galbraith
2007-03-29  8:18       ` Mike Galbraith
2007-03-29 12:55         ` [ck] " michael chang
2007-04-03  2:35         ` Con Kolivas
2007-04-03  2:37       ` Con Kolivas
2007-04-03  5:31         ` Mike Galbraith
2007-04-03  6:00           ` Mike Galbraith
2007-04-03  6:01           ` Ingo Molnar
2007-04-03  6:11             ` Mike Galbraith
2007-04-05 11:02             ` Mike Galbraith
2007-04-05 11:09               ` Ingo Molnar
2007-04-05 11:12                 ` Mike Galbraith
2007-04-05 11:15                   ` Ingo Molnar
2007-04-05 13:18                   ` Johannes Stezenbach
2007-04-05 15:28                     ` Mike Galbraith
2007-04-05 11:54               ` [test] sched: SD-latest versus Mike's latest Ingo Molnar
2007-04-05 12:10                 ` Mike Galbraith
2007-04-05 12:12                   ` Ingo Molnar
2007-04-05 12:24                     ` Mike Galbraith
2007-04-05 16:08                 ` Con Kolivas
2007-04-05 19:05                   ` Ingo Molnar
2007-04-05 20:29                   ` Mike Galbraith
2007-04-06  1:03                 ` Ten percent test Con Kolivas
2007-04-06  9:07                   ` Mike Galbraith
2007-04-06  9:28                     ` Con Kolivas
2007-04-06 10:03                       ` Ingo Molnar
2007-04-06 10:40                         ` Mike Galbraith
2007-04-07  6:50                         ` Con Kolivas
2007-04-07 16:12                           ` Gene Heskett
2007-04-07 18:08                             ` Ingo Molnar
2007-04-07 18:23                               ` Gene Heskett
2007-04-07 18:52                                 ` Ingo Molnar
2007-04-07 20:30                                   ` Gene Heskett
2007-04-08 10:41                                     ` Ingo Molnar
2007-04-08 10:58                                       ` Ingo Molnar
2007-04-08 17:04                                         ` Gene Heskett
2007-04-09  4:03                                           ` Mike Galbraith
2007-04-09  4:08                                             ` Gene Heskett
2007-04-09  5:59                                               ` Mike Galbraith
2007-04-09 13:01                                                 ` Gene Heskett
2007-04-08 11:33                                       ` Gene Heskett
2007-04-08 11:40                                         ` Mike Galbraith
2007-04-08 12:02                                           ` Mike Galbraith
2007-04-08 17:57                                             ` Gene Heskett
2007-04-09  4:19                                               ` Mike Galbraith
2007-04-09  5:23                                                 ` Gene Heskett
2007-04-09  6:09                                                   ` Mike Galbraith
2007-04-08 17:56                                           ` Gene Heskett
2007-04-09  4:17                                             ` Mike Galbraith
2007-04-09  5:16                                               ` Gene Heskett
2007-04-09  6:06                                                 ` Mike Galbraith
2007-04-09  8:24                                                 ` Mike Galbraith
2007-04-08 18:51                                       ` Rene Herman
2007-04-09  4:23                                         ` Mike Galbraith
2007-04-09 12:14                                           ` Rene Herman
2007-04-09 13:27                                             ` Andreas Mohr
2007-04-09 19:54                                               ` Rene Herman
2007-04-09 14:15                                             ` Ingo Molnar [this message]
2007-04-09 17:05                                               ` Rene Herman
2007-04-09 17:48                                                 ` Ingo Molnar
2007-04-09 19:09                                                   ` Rene Herman
2007-04-09 19:56                                                   ` Gene Heskett
2007-04-09 17:10                                             ` Mike Galbraith
2007-04-09 13:53                                         ` Ingo Molnar
2007-04-09 15:37                                           ` Rene Herman
2007-04-07 19:14                               ` Mike Galbraith
2007-04-07 20:31                                 ` Gene Heskett
2007-04-09 17:51                                 ` William Lee Irwin III
2007-04-09 18:03                                   ` Ingo Molnar
2007-04-09 18:44                                     ` William Lee Irwin III
2007-04-07 16:32                           ` Mike Galbraith
2007-04-08 13:08                           ` Ed Tomlinson
2007-04-09  5:38                             ` Mike Galbraith
2007-04-09 11:26                               ` Ed Tomlinson
2007-04-09 16:50                                 ` Mike Galbraith
2007-04-22 10:48                                   ` [ck] " Martin Steigerwald
2007-04-22 11:15                                     ` Con Kolivas
2007-04-10  2:39                               ` Mike Galbraith
2007-04-10 11:23                                 ` Ed Tomlinson
2007-04-10 12:04                                   ` Mike Galbraith
2007-04-06 10:48                       ` Mike Galbraith
2007-04-03 10:57           ` [PATCH] sched: staircase deadline misc fixes Mike Galbraith
2007-03-29  6:36 ` Con Kolivas
2007-04-23  8:58 ` Andrew Morton

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=20070409141516.GB11244@elte.hu \
    --to=mingo@elte.hu \
    --cc=akpm@linux-foundation.org \
    --cc=efault@gmx.de \
    --cc=gene.heskett@gmail.com \
    --cc=kernel@kolivas.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=rene.herman@gmail.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®