>>> In article <[email protected]>, Juergen Kosel 
>>> <[email protected]> writes:

Juergen> Those driver give the time in the format month, day, hour,
Juergen> minute,...  Function refclock_process() converts this time into
Juergen> seconds and nanosecond. But my reference clock already delivers the
Juergen> time in this format.

OK, and the SHM refclock is probably very similar then.

Juergen> After the conversation refclock_process()
Juergen> calls internally the SAMPLE() macro. Since computing time is a
Juergen> concern, I believe it is the better way to use directly the
Juergen> SAMPLE() macro.

Your choice, and if it was me I'd want to see a very clear benefit to this
choice, as it looks to me like you are violating the API layering.

And I also don't understand what you mean by "computing time is a concern".
Is the overhead of a subroutine call that significant in your application?
-- 
Harlan Stenn <[email protected]>
http://ntpforum.isc.org  - be a member!

_______________________________________________
questions mailing list
[email protected]
https://lists.ntp.org/mailman/listinfo/questions

Reply via email to