[PATCH] Staircase scheduler v7.8

From: Con Kolivas (kernel_at_kolivas.org)
Date: 06/30/04

  • Next message: Jacky Malcles: "A question about extended attributes of filesystem objects (setfattr command)"
    Date:	Wed, 30 Jun 2004 22:40:48 +1000
    To: linux kernel mailing list <linux-kernel@vger.kernel.org>
    
    

    -----BEGIN PGP SIGNED MESSAGE-----
    Hash: SHA1

    This is a scheduler policy rewrite designed to be interactive by design
    without tweaks or tuning and be lean and extensible for all sorts of
    settings. (see previous announcements for more detail).

    Patches (including incrementals from previous versions) against 2.6.7
    can be downloaded from:
    http://ck.kolivas.org/patches/2.6/2.6.7

    For those with -ck kernels, the ck patchset was updated to 2.6.7-ck4
    with no other changes to remain in sync with the staircase scheduler:
    http://kernel.kolivas.org

    Version 7.7 proved to be very stable so this version introduces some
    planned improvements. So far no issues have shown up in testing, and
    performance appears better.

    Changes:
    - - Yield logic made robust. Tasks that yield go after everything else,
    but once scheduled are seen as their normal priority - lots of
    applications use yield and this makes them behave a lot better.
    - - Uninterruptible sleep has no effect on burst during interactive mode -
    this improves the responsiveness under I/O load
    - - The 'non-interactive' and 'compute' mode is now much stricter about
    cpu distribution
    - - Code cleanups

    Patch not attached for brevity of email size.
    7 files changed, 283 insertions(+), 610 deletions(-)
    Signed-off-by: Con Kolivas <kernel@kolivas.org>

    Comments, questions, patches and testing welcome,
    Con

    -----BEGIN PGP SIGNATURE-----
    Version: GnuPG v1.2.4 (GNU/Linux)
    Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org

    iD8DBQFA4rTQZUg7+tp6mRURAi+tAJ9ZvacG1YlZPqLZP2qkwx1L3lTGGgCgkvkE
    ekatU5O6OGH7r7Y8ID42SUE=
    =HVc4
    -----END PGP SIGNATURE-----
    -
    To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
    the body of a message to majordomo@vger.kernel.org
    More majordomo info at http://vger.kernel.org/majordomo-info.html
    Please read the FAQ at http://www.tux.org/lkml/


  • Next message: Jacky Malcles: "A question about extended attributes of filesystem objects (setfattr command)"

    Relevant Pages

    • RE: Network slowdown due to CFS
      ... it calls yield before re-acquiring them to go back to work. ... You are assuming that random latencies are necessarily bad. ... scheduler implementation details in random ways. ... I think CFS' semantics are pretty bad. ...
      (Linux-Kernel)
    • Re: Network slowdown due to CFS
      ... cannot afford to wait for the lock to be released. ... it calls yield before re-acquiring them to go back to work. ... and it will introduce random latencies into apps using ... used scheduler functionalities, so it does not really matter. ...
      (Linux-Kernel)
    • Re: 2.6.0-test4-mm2
      ... I havent tried it myself yet but Ive heard of colliding code in scheduler. ... Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org ... send the line "unsubscribe linux-kernel" in ...
      (Linux-Kernel)
    • Re: [patch] CFS scheduler, -v12
      ... any scheduler that implements sys_sched_yieldin a sane manner. ... interaction of CFS's yield implementation with this user-space bug could ... every frame specifically for SGI IRIX frame scheduler (driven ...
      (Linux-Kernel)
    • Re: [patch] CFS scheduler, -v12
      ... any scheduler that implements sys_sched_yieldin a sane manner. ... interaction of CFS's yield implementation with this user-space bug could ... every frame specifically for SGI IRIX frame scheduler (driven ...
      (Linux-Kernel)