Results 1 to 6 of 6

Thread: Linux Dynticks Being Extended For Performance Wins

  1. #1
    Join Date
    Jan 2007
    Posts
    14,531

    Default Linux Dynticks Being Extended For Performance Wins

    Phoronix: Linux Dynticks Being Extended For Performance Wins

    Dynticks, the Dynamic Tick Timer for allowing the Linux kernel to skip ticks while idling and resume to running at full HZ when encountering load, is in the process of being extended. Developers are working on making Dynticks work even under select workloads in order to enhance the performance of CPU-intensive tasks...

    http://www.phoronix.com/vr.php?view=MTI1NTg

  2. #2
    Join Date
    Nov 2008
    Location
    Madison, WI, USA
    Posts
    862

    Default

    Sounds like a good idea in general. We'll see how it turns out.

  3. #3
    Join Date
    Apr 2008
    Location
    NJ
    Posts
    78

  4. #4

    Default I'm confused

    I thought the Linux kernel had gone "tick-less" a while back?

    And in any case, how would this be different?

  5. #5
    Join Date
    Oct 2008
    Posts
    3,070

    Default

    Quote Originally Posted by Bernard Swiss View Post
    I thought the Linux kernel had gone "tick-less" a while back?

    And in any case, how would this be different?
    It's "tickless" in the way it sleeps while idle, no longer polling itself when there is nothing to do.

    This new works extends that to keep being tickless even while the cpu is busy working.

    Not waking while idle was a big improvement for power usage. The new work should help improve performance, since the kernel is no longer interrupting tasks and flushing hardware caches so often.

  6. #6

    Default

    Quote Originally Posted by smitty3268 View Post
    It's "tickless" in the way it sleeps while idle, no longer polling itself when there is nothing to do.

    This new works extends that to keep being tickless even while the cpu is busy working.

    Not waking while idle was a big improvement for power usage. The new work should help improve performance, since the kernel is no longer interrupting tasks and flushing hardware caches so often.
    Thanks, that's enough to put it into perspective for me.
    (That was roughly what I was understanding, but I wasn't at all sure I got it right).

Posting Permissions

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