[ntp:questions] Re: refclock WWV on freeBSD 5.3
Ronan Flood
ronan at noc.ulcc.ac.uk
Mon Dec 20 15:13:36 UTC 2004
david.margrave at gmail.com wrote:
> A while back I set up WWV on a linux box, and decided to give it a try
> on free BSD next. I have just installed freeBSD 5.3 on x86 hardware,
> and built the 'port' package for ntp. I got this in the logs:
>
> Dec 19 17:00:47 scooby ntpd[25416]: mlockall(): Resource temporarily
> unavailable
> Dec 19 17:00:47 scooby ntpd[25416]: internal error: refclockio
> structure not found
> Dec 19 17:00:48 scooby kernel: pid 25416 (ntpd), uid 0: exited on
> signal 11 (core dumped)
>
> I found another post on the mlockall and changed that in my config.h.
>
> It appears that the WWV refclock is being built (there is a .o module
> for it in the work dir, and nm on the ntpd binary shows symbols like
> wwv_epoch defined). I looked at the code and this 'refclockio
> structure not found' seems to happen when it walks through a linked
> list of refclockio structs searching for the matching one, and can't
> find it.
Yes, when it is trying to close the file-descriptor of the refclock.
> I went ahead and added this to my top-level port Makefile:
> CONFIGURE_ARGS+= --enable-WWV
>
> and rebuilt, and still got the same syslog entries. So I'm basically
> stuck at this point. Here are the relevant bits of my ntp.conf
>
> server 127.127.36.0
> fudge 127.127.36.0 time1 0.005200 time2 0.014657 flag2 1 flag3 1 flag4
> 1
I'd guess there's some trouble with the audio device, which is causing
the refclock initialization to fail silently; at which point it calls
the shutdown routine which fails with the error you see because the
init didn't work.
Reconfiguring and running with debugging on might give more info.
--
Ronan Flood <R.Flood at noc.ulcc.ac.uk>
working for but not speaking for
Network Services, University of London Computer Centre
(which means: don't bother ULCC if I've said something you don't like)
More information about the questions
mailing list