Re: RFA: Changing scheduler quantum (Was: REQUEST: OpenLDAP 2.3.7)

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Sun, 2005-09-18 at 23:53 +0200, Bernardo Innocenti wrote:
> Each and every call to sched_yield() makes the process
> sleep for over *50ms* while a "nice make bootstrap" is
> running in the background:

sched_yield() not only yields the processor and current timeslice, but
also moves the process to the expired queue.  It will remain there until
the next time schedule() runs to reassign priorities and timeslices.
So, you are having to wait for the make process to use it's timeslice
and for the kernel to run schedule().  Could very easily be 50ms.  I
don't really know.  Btw, the man page for sched_yield() is out of date
on my box.  It's from Linux 1.3.81.  Everything has changed with the
introduction of the O(1) scheduler.  Before it just moved your process
to the end of its queue, now it moves the process to the expired queue.
sched_yield() should only be used if you don't need the processor back
for a while.

Matthew E. Lauterbach

-- 
fedora-devel-list mailing list
fedora-devel-list@xxxxxxxxxx
https://www.redhat.com/mailman/listinfo/fedora-devel-list

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Index of Archives]     [Fedora Announce]     [Fedora Kernel]     [Fedora Testing]     [Fedora Formulas]     [Fedora PHP Devel]     [Kernel Development]     [Fedora Legacy]     [Fedora Maintainers]     [Fedora Desktop]     [PAM]     [Red Hat Development]     [Gimp]     [Yosemite News]
  Powered by Linux