In message <[EMAIL PROTECTED]>, Dag-Erling Smorgrav writes:
>I regularly get "microuptime() went backwards" warnings on my desktop
>box. The funny thing about them is that the reported timevals have the
>same seconds part, but the microseconds part of the second timeval is
>so large that it's wrapped around to a negative number (causing the
>signed comparison to report that it went backwards). This suggests
>that the current process has been running uninterrupted for several
>seconds, which seems unlikely - or that the timecounter was adjusted
>upwards while the process was running (could ntp cause that?)

No, this is either a problem reading the i8254 timecounter reliably
or an interrupt latency problem.

If you sysctl's indicate that you are running on the TSC timecounter
and you can reproduce this there is some chance we can create a workaround.

Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
[EMAIL PROTECTED]         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.

To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message

Reply via email to