Results 1 to 7 of 7

Thread: After BFS, Now Comes The BLD Linux Scheduler

  1. #1
    Join Date
    Jan 2007
    Posts
    14,539

    Default After BFS, Now Comes The BLD Linux Scheduler

    Phoronix: After BFS, Now Comes The BLD Linux Scheduler

    Adding to the already existing Linux schedulers, including the out-of-tree Brain Fuck Scheduler (BFS), is now a proposal to integrate a new scheduler... The Barbershop Load Distribution scheduler...

    http://www.phoronix.com/vr.php?view=MTA1NzE

  2. #2
    Join Date
    May 2010
    Posts
    190

    Default

    Quote Originally Posted by TFA
    Please be patient if you
    started to feel disgusting.
    I laughed.

  3. #3
    Join Date
    Jul 2008
    Location
    Greece
    Posts
    3,788

    Default

    This isn't a scheduler. Why is the article claiming that it is? Please correct the article. It's a load balancing algorithm for the existing CPU scheduler.

  4. #4
    Join Date
    Feb 2009
    Location
    Province of Québec, Canada
    Posts
    52

    Default

    I like dev's imagination when it come to name their project

  5. #5
    Join Date
    Jan 2008
    Posts
    5

    Default

    Wow. The credibility of Phoronix hits a new low?

  6. #6
    Join Date
    Jan 2012
    Posts
    179

    Default

    Could you elaborate on this new algorithm vs. new scheduler topic?

  7. #7
    Join Date
    Oct 2009
    Posts
    2,081

    Default

    Quote Originally Posted by varikonniemi View Post
    Could you elaborate on this new algorithm vs. new scheduler topic?
    The load distribution algorithm is *part of* the scheduler. The part that determines how the current workload is distributed across multiple processing cores. It has no context on single-core systems, which STILL need a process scheduler in order to provide every process with running time... i.e., you run this process for a bit, then you switch to this process, then another process, now back to the first one, etc. With multiple processing cores, you need to distribute your work across all the cores... when the system is fairly loaded. One of the things that this is doing is (I've only skimmed everything, so I might be off a bit) apparenty cutting out the part about even distribution across all cores when the work load is low. If you have an 8 core system and only enough work is being done to load 1 core by 5%, what's the use in distributing the workload evenly across all 8 cores?

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •