[lkml]   [2004]   [Jan]   [21]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
    SubjectRe: 2.6.1 "clock preempt"?
    On Tue, 2004-01-20 at 16:26, john stultz wrote:
    > On Tue, 2004-01-20 at 12:36, Steinar Hauan wrote:
    > > i've started to test the 2.6 series of kernels and observed a
    > > strange thing: with moderate background load, the system clock
    > > (i.e. time) seems to slow down to about 60% of normal speed
    > > and the normally reliable ntp process (v4.2.0)

    omission in original post:
    [...] ntp process "was running", but failed to update time.

    > > the kernel logs show messages on the form:
    > >
    > > localhost kernel: Losing too many ticks!
    > > localhost kernel: TSC cannot be used as a timesource.
    > > (Are you running with SpeedStep?)
    > How quickly do you see this message? Does it happen right at boot time,
    > or during load?

    Jan 19 08:59:24 localhost kernel: Linux version 2.6.1
    (root@offa) (gcc version 3.3.2 20031218
    (Red Hat Linux 3.3.2-5)) #2
    SMP Sat Jan 17 18:05:09 EST 2004

    (yes, i have SMP enabled even for a single cpu machine; the kernel
    will be run on a multitude of machines of which most are SMP)

    log messages prior to time issues:

    Jan 20 04:02:01 localhost anacron[6057]: Updated timestamp for
    job `cron.daily' to
    Jan 20 04:05:38 localhost kernel: hdb: dma_timer_expiry:
    dma status == 0x61
    Jan 20 04:05:48 localhost kernel: hdb: DMA timeout error
    Jan 20 04:05:48 localhost kernel: hdb: dma timeout error:
    status=0xd0 {Busy }
    Jan 20 04:05:48 localhost kernel: hda: DMA disabled
    Jan 20 04:05:48 localhost kernel: hdb: DMA disabled
    Jan 20 04:05:48 localhost kernel: ide0: reset: success

    (hda and hdb are on the same controller; the only active disk at
    this time should be hda ... hdb should be essentially idle
    except for cron.daily scripts running around then)

    timing error starts here; interactive work

    Jan 20 07:52:40 localhost kernel: Losing too many ticks!

    > You might want to try the attached patch to see if we're overreacting
    > Also, do you see the problem when preempt is disabled

    testing overnight failed to reproduce the problem. note that the
    original event only occurred after about 12 hrs of testing.
    i'll install lm_sensors & smarttools and run cpuburn for a while
    to ensure this is not related to any hardware issues.

    --> other suggestions also appreciated.

    Steinar Hauan, dept of ChemE --
    Carnegie Mellon University, Pittsburgh PA, USA

    To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
    the body of a message to
    More majordomo info at
    Please read the FAQ at

     \ /
      Last update: 2005-03-22 14:00    [W:0.023 / U:19.508 seconds]
    ©2003-2017 Jasper Spaans. hosted at Digital OceanAdvertise on this site