On Thu, Oct 05, 2000 at 01:58:27PM +1100, William J. Tango wrote:
> Michael Barabanov wrote:
> >
> > RTLinux3.0-pre8 is available at ftp://fsmlabs.com/pub/rtlinux/v3.
> > The highlights are: HTML documentation, debugger for the PowerPC,
> > some important bug fixes, installation procedure changes.
> > RPMs to be available shortly.
> >
> > All RTLinux 3.0-pre7 users are strongly encouraged to upgrade
> > to this release.
> >
> > Michael.
>
>
> In RTL3.0-pre7 clock_gethrtime() works per manual page: ie it
> returns the time elapsed since the Epoch.
>
> In RTL3.0-pre8 clock_gethrtime() has reverted to its previous
> and incorrect behaviour: it returns the time elapsed since last
> boot.
We'll check. Thanks for the tip.
>
> Also, I use xntpd. When using RTL3.0-pre7 I find that the time reported
> by clock_gethrtime() in my RT process drifts wrt the time reported
> by ntp_gettime() in my non-RT process. It appears that CLOCK_REALTIME
> is *not* the clock disciplined by xntpd. Is it possible to access the
Yes.
> disciplined clock from RT processes? The control system is intended
> for an astronomical application where it is essential to have UTC with
> millisec accuracy.
We have not yet figured out how to do this. there is a new draft POSIX standard, but
it is rather complex. But you are actually asking for something that seems to me to
be even more difficult: xntpd itself must run with no
no millisecond delays and sync messages to xntpd must also be delivered with no
delays/
--
---------------------------------------------------------
Victor Yodaiken
Finite State Machine Labs: The RTLinux Company.
www.fsmlabs.com www.rtlinux.com
-- [rtl] ---
To unsubscribe:
echo "unsubscribe rtl" | mail [EMAIL PROTECTED] OR
echo "unsubscribe rtl <Your_email>" | mail [EMAIL PROTECTED]
---
For more information on Real-Time Linux see:
http://www.rtlinux.org/rtlinux/
