[lkml]   [2002]   [Nov]   [2]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
    Subjecttime() glitch on 2.4.18 at 177 days uptime?
    I'm running Linux 2.4.18 on an Athlon (i386).

    My uptime is 182 days. About five days ago, I started noticing
    strange date effects. It may be hardware related, but I'm also
    suspecting that it may be due to the long uptime, and so I'm hesitant
    to reboot, in case there is a bug here that needs to get fixed.

    The problem is with time(). Every second, for approximately 1.1ms,
    time() reports a value that is about 2^32 microseconds (4295 seconds,
    or about an hour and a quarter) in the future. The glitches always
    occur between a change of seconds. Look at this:

    $ while true; do paste <( cat /proc/uptime ) <( date ) ; done | grep -A 1 -B 1 ' 02:' | head -11
    15765463.31 15455589.84 Sat Nov 2 00:59:16 EST 2002
    15765463.32 15455589.84 Sat Nov 2 02:10:51 EST 2002
    15765463.33 15455589.84 Sat Nov 2 00:59:17 EST 2002
    15765465.31 15455589.84 Sat Nov 2 00:59:18 EST 2002
    15765465.32 15455589.84 Sat Nov 2 02:10:53 EST 2002
    15765465.34 15455589.84 Sat Nov 2 00:59:19 EST 2002
    15765466.30 15455589.84 Sat Nov 2 00:59:19 EST 2002
    15765466.32 15455589.84 Sat Nov 2 02:10:54 EST 2002
    15765466.33 15455589.84 Sat Nov 2 00:59:20 EST 2002

    The first two values on each line are from /proc/uptime, and the rest
    is of course from "date". This bash script runs a bit slow, which is
    why it missed the glitch between :17 and :18, but by watching time()s
    more frequently, I can definitely see that they're there. I don't
    know for certain whether this happens at all times during the day, but
    I believe it does. I can do some more logging to figure that out if

    I've been running ntpdate (and therefore calling do_adjtime()) every
    half hour since the system booted, in case that might affect

    This glitch does _not_ occur on a 2.2.? system on some x86 processor
    with about 256 days uptime (I unfortunately don't know any more
    details than that; I only have very limited access to that machine).
    I also don't have access to any other 2.4 systems with an uptime of
    greater than about 80 days.

    This is not a glibc bug; the erroneous values are being returned by
    the kernel. Perhaps someone can help me track this down. I've put my
    kernel config and relevant info from /proc at

    (and would appreciate a CC of any replies)

    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 13:30    [W:0.021 / U:13.844 seconds]
    ©2003-2017 Jasper Spaans. hosted at Digital OceanAdvertise on this site