[HEADSUP] retire TIOCDCDTIMESTAMP
freebsd-misuser at remove-NOSPAM-to-reply.NOSPAM.dyndns.dk
Mon Aug 2 06:17:00 PDT 2004
[keep replies on the list, not to me, to avoid bounces, thanks]
[sorry for the lateness of this -- that's what I get for being offline]
"M. Warner Losh" <imp at bsdimp.com> wrote:
> In message: <23206.1087846355 at critter.freebsd.dk>
> Poul-Henning Kamp <phk at phk.freebsd.dk> writes:
> : The RFC 2783 PPS-API has been standardized to do everything
> : TIOCDCDTIMESTAMP did and quite a bit more.
> I think this is reasonable on its face. Unless this breaks a bunch of
> ports I don't see a big problem with it. I have some problems using
> pps api at times, but those aren't relevant for this change...
I have expanded TIOC*TIMESTAMP in my own personal hive of hackery
to also handle signals on the other modem status lines as well.
Therefore I'm curious.
As far as I know, I can simply revert this retirement in my own
personalized codebase without difficulty if I choose. So far.
Supposing I want to use the PPS-API. It seems that it's possible
for me to select whether I want it to work on the DCD status line,
or (hmm, one other, must grep code) status line.
But is it possible for the PPS-API to work simultaneously on more
than a single status line? A preliminary perusal of the code would
suggest not, but I can't read code to save my life.
Basically, my TIOC*TIMESTAMP hacks allowed me to gather timestamps
from the four status lines, to be used by programs that could
handle four sources of PPS data on a single port. I'd like to
be able to do that with PPS-API as well, if possible.
(The other wishlist item I have that might be applicable to PPS-API
is to be able to query system-uptime as well as system-time at a
status change, for cases where the system time has been stepped.
This may be something irrelevant to or outside the scope of PPS-API.
Clearly I need to figure out how to use that first, and maybe then
I won't have such stupid questions.)
More information about the freebsd-current