[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [gpsd-users] NTP for more than two PPS-capable devices
From: |
Eric S. Raymond |
Subject: |
Re: [gpsd-users] NTP for more than two PPS-capable devices |
Date: |
Wed, 14 Jan 2015 16:32:05 -0500 |
User-agent: |
Mutt/1.5.23 (2014-03-12) |
Gary E. Miller <address@hidden>:
> > If it is not real, how do you know how many segments NTP will
> > support. Is there a way to ask it?
>
> Here is some doc:
>
> "SYSV IPC creates a shared memory segment with a key value of 0x4E545030
> + u, where u is again the clock unit. (This value could be hex-decoded
> as 'NTP0', 'NTP1',..., with funny characters for units > 9.) "
>
> ntp_refclock.c line 193 does truncate unit to a u_char, so 256 would be
> the upper limit. I see some overflow issues, in refclock_shm.c line
> 126, so 207 is probably the max useable.
>
> If you wanted to stick to ASCII (no good reason to) you would have about
> 80.
OK, that's interesting, but it doesn't completely answer my question. Say
I'm a running instance of GPSD. Can I just create segments ad-hoc up to
a limit up to (say) 128 and expect ntpd to pick up on all of them
automatically? If not, how can I find out what its upper limit is from
a running instance of ntpd?
--
<a href="http://www.catb.org/~esr/">Eric S. Raymond</a>