All of lore.kernel.org
 help / color / mirror / Atom feed
From: Guy <fsos_guy@earthlink.net>
To: linux-kernel@vger.kernel.org
Subject: 2.6 scheduler and "fast user switching"
Date: Thu, 13 Nov 2003 04:30:02 -0500	[thread overview]
Message-ID: <200311130430.06882.fsos_guy@earthlink.net> (raw)

Scenario:

I typically log in as 'root' on the first console. I then invoke 
fluxbox as the GUI.

# XSESSION=fluxbox startx -- :0

I then ctl-alt-F2 another console and login as 'user1'. I then 
invoke KDE as the GUI.

$ XSESSION=kde-3.1.4 startx -- :1

I may or may not ctl-alt-Fn and login as 'usern' and repeat the 
process.

Several thoughts:

1} I've seen Nick Piggin's suggestion of nicing X server to -10. 
At the moment, the only way I know to do this is something like

# XSESSION=fluxbox nice --adjustment=-10 startx -- :N

A} My default security is that only 'root' can perform nice with 
negative values. I am reluctant to play with security for such a 
crticial command.

B} All child threads inherit the new nice value. So in the example 
just above, this means all applications started from the GUI 
desktop run at a nice value of -10. I believe enhancing the X 
server nice value this way defeats the purpose of nicing it to 
begin with. Obviously, despite my readings and attempts at 
research, I'm must be missing something here.

2} I expect to travel down to Florida for Xmass to visit family. 
One of the things I had hoped to do was to set up my mother's 
computer as an X server and hang a thin client terminal {read: 
older PC} off of it. This would allowed my mother and brother to 
share a reasonably modern system at the same time.

This is not me just being cheap. I'm interested in setting up 
diskless workstations aound a good central X server. I see such 
setups as appropriate for a number of situations. If the X server 
requires 'nicing' in a single user environment, what happens in 
an LTSP environment?

My base reference environment is 2.4.20. I still actively use it 
for everything I do as everything works as expected. 

Despite my enthusiasm for 2.6, I find it difficult to get 
everything to 'just work'. I still see problems in the area of 
nForce based mobos {stupid proprietary nVidia!}, broken BIOSes, 
and scheduler issues like the above.

Flames, instruction, suggestions, thoughts would be appreciated.

FWIW, I'm not a C/C++ programmer. I'm computer literate and am not 
afraid to run bleeding edge.

Guy

-- 
Recyle computers. Install Gentoo GNU/Linux.


             reply	other threads:[~2003-11-13 10:59 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2003-11-13  9:30 Guy [this message]
2003-11-13 11:11 ` 2.6 scheduler and "fast user switching" Nick Piggin
2003-11-13 21:11   ` Guy
2003-11-14  0:58     ` Guy

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=200311130430.06882.fsos_guy@earthlink.net \
    --to=fsos_guy@earthlink.net \
    --cc=linux-kernel@vger.kernel.org \
    /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.