lkml.org 
[lkml]   [2007]   [Apr]   [10]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
    /
    Date
    From
    SubjectRe: [BUG] scheduler: first timeslice of the exiting thread
    > > This comes at an awkward time, because we might well merge the
    > > staircase/deadline work into 2.6.22, and I think it rewrites the part of
    > > the scheduler which is causing the problems you're observing.
    > >
    > > Has anyone verified that SD fixes this problem and the one at
    > > http://lkml.org/lkml/2007/4/7/21 ?
    >
    > Not verified either way in testing, but I believe this should be a
    > problem for SD as well because timeslice fork/exit handling is identical
    > with mainline. Individual slices are much smaller than mainline, so
    > priority should drop rapidly, consuming bandwidth allotted for the
    > current rotation, sending the creator off to the expired array
    > prematurely.
    >
    > -Mike
    >

    Yes, this probelm also occurs on the SD. Here is the 2.6.21-rc6-mm1's
    trace.

    Process model:

    fork: pid = 12431, tgid = 12431, ppid = 2737, time_slice = 8000000
    exit: pid = 12431, tgid = 12431, ppid = 12430, time_slice = 3842675
    fork: pid = 12432, tgid = 12432, ppid = 2737, time_slice = 6175786
    exit: pid = 12432, tgid = 12432, ppid = 12430, time_slice = 2948968
    fork: pid = 12433, tgid = 12433, ppid = 2737, time_slice = 5862607
    exit: pid = 12433, tgid = 12433, ppid = 12430, time_slice = 2793696
    fork: pid = 12434, tgid = 12434, ppid = 2737, time_slice = 5515983
    exit: pid = 12434, tgid = 12434, ppid = 12430, time_slice = 2583218
    fork: pid = 12435, tgid = 12435, ppid = 2737, time_slice = 5174253
    exit: pid = 12435, tgid = 12435, ppid = 12430, time_slice = 2421119
    fork: pid = 12436, tgid = 12436, ppid = 2737, time_slice = 4433521
    exit: pid = 12436, tgid = 12436, ppid = 12430, time_slice = 2068271
    fork: pid = 12437, tgid = 12437, ppid = 2737, time_slice = 4115828
    exit: pid = 12437, tgid = 12437, ppid = 12430, time_slice = 1906364
    fork: pid = 12438, tgid = 12438, ppid = 2737, time_slice = 3817721
    exit: pid = 12438, tgid = 12438, ppid = 12430, time_slice = 1763586
    fork: pid = 12439, tgid = 12439, ppid = 2737, time_slice = 3523880
    exit: pid = 12439, tgid = 12439, ppid = 12430, time_slice = 1621289
    fork: pid = 12440, tgid = 12440, ppid = 2737, time_slice = 2856168
    exit: pid = 12440, tgid = 12440, ppid = 12430, time_slice = 1269502
    exit: pid = 12430, tgid = 12430, ppid = 2737, time_slice = 2434379

    Thread model:

    fork: pid = 12554, tgid = 12553, ppid = 2737, time_slice = 2841904
    exit: pid = 12554, tgid = 12553, ppid = 2737, time_slice = 1420952
    fork: pid = 12555, tgid = 12553, ppid = 2737, time_slice = 29977
    exit: pid = 12555, tgid = 12553, ppid = 2737, time_slice = 14988
    fork: pid = 12556, tgid = 12553, ppid = 2737, time_slice = 8000000
    exit: pid = 12556, tgid = 12553, ppid = 2737, time_slice = 4000000
    fork: pid = 12557, tgid = 12553, ppid = 2737, time_slice = 3976416
    exit: pid = 12557, tgid = 12553, ppid = 2737, time_slice = 1988208
    fork: pid = 12558, tgid = 12553, ppid = 2737, time_slice = 1965192
    exit: pid = 12558, tgid = 12553, ppid = 2737, time_slice = 982596
    fork: pid = 12559, tgid = 12553, ppid = 2737, time_slice = 955481
    exit: pid = 12559, tgid = 12553, ppid = 2737, time_slice = 477740
    fork: pid = 12560, tgid = 12553, ppid = 2737, time_slice = 453317
    exit: pid = 12560, tgid = 12553, ppid = 2737, time_slice = 226658
    fork: pid = 12561, tgid = 12553, ppid = 2737, time_slice = 203189
    exit: pid = 12561, tgid = 12553, ppid = 2737, time_slice = 101594
    fork: pid = 12562, tgid = 12553, ppid = 2737, time_slice = 76756
    exit: pid = 12562, tgid = 12553, ppid = 2737, time_slice = 38378
    fork: pid = 12563, tgid = 12553, ppid = 2737, time_slice = 15096
    exit: pid = 12563, tgid = 12553, ppid = 2737, time_slice = 7548
    exit: pid = 12553, tgid = 12553, ppid = 2737, time_slice = 7851311

    NOTE: Nanosecond is the unit of time_slice on SD, so the time_slice
    value is very differ from the current scheduler's one.

    Satoru
    -
    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/

    \
     
     \ /
      Last update: 2007-04-10 09:23    [W:0.026 / U:90.468 seconds]
    ©2003-2016 Jasper Spaans. hosted at Digital OceanAdvertise on this site