Re: Relation between hwtimestamp and PTP clocks
Miroslav Lichvar <[email protected]>
| Newsgroups | gmane.comp.time.chrony.user |
|---|---|
| Message-ID | <ZJk6HsJnfng0OPGj@localhost> |
On Fri, Jun 23, 2023 at 11:11:21AM +0000, Linas Zvirblis wrote: > Does that mean that chrony is adjusting the PHC by itself? Is > 'hwtimestamp' for PHC equivalent to 'rtcsync' for RTC? My experience > says no. Functionaly it's closer to rtcdrift. > My question is, would running with 'hwtimestamp eth0' in chrony and > 'phc2sys -O 37 -s CLOCK_REALTIME -c eth0' interfere with each other? Yes, that would lead to instability. > Can I run multiple processes (chrony and ptp4l) using hardware > timestamping on the same network interface, or does one of them have to > use software timestamping? They can both use hardware timestamping, but there needs to be a virtual PHC created on top of the real PHC for ptp4l, so it doesn't disrupt chrony. It's a new feature added in linux-5.14 and now supported in linuxptp-4.0. A vclock can be created like this echo 1 > /sys/class/ptp/ptp0/n_vclocks If if created /dev/ptp1, you would start ptp4l with --phc_index 1. -- Miroslav Lichvar -- To unsubscribe email chrony-users-request-kWFZVVI9zxvPqho9SqqRMmD2FQJk+8+b@public.gmane.org with "unsubscribe" in the subject. For help email chrony-users-request-kWFZVVI9zxvPqho9SqqRMmD2FQJk+8+b@public.gmane.org with "help" in the subject. Trouble? Email [email protected]