linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Lee Revell <rlrevell@joe-job.com>
To: sylvanino b <sylvanino@gmail.com>
Cc: linux-kernel@vger.kernel.org
Subject: Re: I wrote a kernel tool for monitoring / web page
Date: Fri, 18 Feb 2005 21:52:14 -0500	[thread overview]
Message-ID: <1108781534.6040.36.camel@krustophenia.net> (raw)
In-Reply-To: <d14685de05021817333c563cc9@mail.gmail.com>

On Sat, 2005-02-19 at 02:33 +0100, sylvanino b wrote:
> Sorry, it's meant to run on linux.
> Actually, patch provided is for linux 2.6.9 + kdb 4.4
> 

Cool program.  It has an annoying bug where every time you go to "Open
Log File", it starts you in your home directory again.  Otherwise it's a
nice utility.

I actually have a problem that this might help with.  The issue is that
the scheduler seems to treat Evolution as a CPU bound rather than an
event driven, I/O bound process.  The most obvious symptom is that a
real CPU bound activity like a kernel compile will cause navigating the
message list in Evolution to slow to a crawl.  Evolution is perfectly
usable when no other CPU hogs are running, or when the CPU hogs are
niced, so it's definitely a scheduler issue.

My understanding of Unix schedulers is that the basic idea is to
penalize CPU bound and reward I/O bound processes by giving the former
lower dynamic priority with longer timeslice and the latter high
priority with shorter timeslice.  I suspect the scheduler does not
handle interactive, event driven apps that also consume a lot of CPU due
to bloat very well.  These would seem to need high priority and long
timeslices, which would require the scheduler to distinguish a process
like a kernel compile that will continually exhaust its timeslice no
matter how long, and a process like evolution that if given a long
enough timeslice will finish rendering the message and go back to sleep.

Anyway, that's my hypothesis, I'll let you know what I find out.

Lee



  reply	other threads:[~2005-02-19  2:52 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2005-02-19  0:41 I wrote a kernel tool for monitoring / web page sylvanino b
2005-02-19  1:01 ` Lee Revell
2005-02-19  1:33   ` sylvanino b
2005-02-19  2:52     ` Lee Revell [this message]
2005-02-19 23:50     ` Florian Schmidt
2005-02-20  0:04       ` Randy.Dunlap
2005-02-20  0:17         ` Lee Revell
2005-02-20  0:23           ` sylvanino b
2005-02-19 10:10 ` Matthias-Christian Ott
2005-02-23 12:24 ` sylvanino b

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=1108781534.6040.36.camel@krustophenia.net \
    --to=rlrevell@joe-job.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=sylvanino@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;
as well as URLs for NNTP newsgroup(s).