Closed 652436962 closed 6 months ago
You mean ts2phc.pin_index
, right? It defaults to zero, which is often right, but I agree it would be helpful to put it in the configuration file. Thanks for the suggestion!
”> You mean ts2phc.pin_index
, right? It defaults to zero, which is often right, but I agree it would be helpful to put it in the configuration file. Thanks for the suggestion!”
You mean ts2phc.pin_index
, right? yes
I started the task based on your documentation and official documentation, but after a while there will be a value of a few microseconds. Is this normal?
Is there a switch between the GM and the client?
Is there a switch between the GM and the client?
no
That doesn't sound normal.
I don't understand what ts2phc_slave is doing,
That doesn't sound normal.
- What is your hardware configuration (GM, client, GPS)?
- What are you measuring? What is showing a value of a few microseconds?
I don't understand what ts2phc_slave is doing,
os : buildroot
gm is rockchip3568 inter i210 gprmc + pps .
ts2phc_slave output 1pps
between gm 1pps and ptp client have a value of a few microseconds
Sorry, I don't understand. Please explain your setup in much more detail.
I still don't understand why you are using ts2phc_slave. Why aren't you getting the PPS from the GM using PHC periodic output on the SDP of the i210?
Sorry, I don't understand. Please explain your setup in much more detail.
I still don't understand why you are using ts2phc_slave. Why aren't you getting the PPS from the GM using PHC periodic output on the SDP of the i210?
The two devices are synchronized using ptp, and the slave device outputs 1pps through ts2phc
I am not going to be able to help unless you give me much, more detail.
I am not going to be able to help unless you give me much, more detail.
- config files
- command lines
- logs
- hardware used for the client
- hardware used for the GM in detail; I don't know what "rockchip3568 inter i210 gprmc + pps" means
Why aren't you getting the PPS from the GM using PHC periodic output on the SDP of the i210? How to set this up specifically?
arm64 chip : rk3568
ts2phc_slave [global] use_syslog 0 verbose 1 logging_level 6 ts2phc.pulsewidth 100000000 [enp1s0] ts2phc.channel 0
ts2phc.extts_polarity both ts2phc.pin_index 1
You set it up like this:
https://github.com/jclark/rpi-cm4-ptp-guide/blob/main/measure.md#enable-pps-output
However with the i210, you cannot specify the -w
argument. It always defaults to 50% duty cycle.
You set it up like this:
https://github.com/jclark/rpi-cm4-ptp-guide/blob/main/measure.md#enable-pps-output
However with the i210, you cannot specify the
-w
argument. It always defaults to 50% duty cycle.
I tested it, testptp cannot output on my platform
You're saying:
That seems very strange. I can't think of any reason why that could be the case.
What does phc_ctl enp1s0
say?
You're saying:
- you have an i210
- you have SDPs
- external timestamping works
- periodic output doesn't work
That seems very strange. I can't think of any reason why that could be the case.
What does
phc_ctl enp1s0
say?
I think it's a switch problem. I removed the switch and it doesn't seem to have this problem. I still need to test it.
I tested and found that after about 26393 seconds, the difference between the pps of his server and the client became larger and larger.
What happens after that?
My guess is that ts2phc got confused: it's not very robust. I suggest you start to investigate the problem by looking at the logs for ts2phc.
What happens after that?
My guess is that ts2phc got confused: it's not very robust. I suggest you start to investigate the problem by looking at the logs for ts2phc.
It shows this, there is no pps deviation data anymore
"poll returns zero" means that ts2phc isn't receiving PPS events.
"clearing Tx timestamp bug" suggests you have a kernel/driver problem (I guess a problem with igb driver on your ARM system).
What kernel are you using? Is the i210 on a separate PCIe card or is it on the board?
What kernel are you using? Is the i210 on a separate PCIe card or is it on the board?
kernel 5.10 pcie2.0 rockchip rk3568(arm64)
This isn't going to be straightforward to fix.
I suggest first trying a newer kernel. I believe 6.1 is available or coming soon for the rk3568. Another thing to try is to figure out whether this a arm64 problem or whether it's specific to rk3568.
What board do you have?
This isn't going to be straightforward to fix.
I suggest first trying a newer kernel. I believe 6.1 is available or coming soon for the rk3568. Another thing to try is to figure out whether this a arm64 problem or whether it's specific to rk3568.
What board do you have?
itop-3568, a board by a small company
ts2phc[481.201]: config item (null).free_running is 0 ts2phc[481.201]: PPS sink enp1s0 has ptp index 0 ts2phc[481.201]: UTC-TAI offset not set in system! Trying to revert to leapfile ts2phc[481.201]: config item (null).leapfile is '/usr/share/zoneinfo/leap-seconds.list' ts2phc[481.823]: UTC time is past leap second table expiry date ts2phc[481.823]: source ts not valid ts2phc[481.823]: UTC time is past leap second table expiry date ts2phc[481.823]: source ts not valid ts2phc[481.923]: UTC time is past leap second table expiry date ts2phc[481.923]: source ts not valid ts2phc[481.923]: UTC time is past leap second table expiry date ts2phc[481.923]: source ts not valid ts2phc[482.823]: UTC time is past leap second table expiry date ts2phc[482.823]: source ts not valid ts2phc[482.824]: UTC time is past leap second table expiry datets2phc[481.201]: config item (null).free_running is 0 ts2phc[481.201]: PPS sink enp1s0 has ptp index 0 ts2phc[481.201]: UTC-TAI offset not set in system! Trying to revert to leapfile ts2phc[481.201]: config item (null).leapfile is '/usr/share/zoneinfo/leap-seconds.list' ts2phc[481.823]: UTC time is past leap second table expiry date ts2phc[481.823]: source ts not valid ts2phc[481.823]: UTC time is past leap second table expiry date ts2phc[481.823]: source ts not valid ts2phc[481.923]: UTC time is past leap second table expiry date ts2phc[481.923]: source ts not valid ts2phc[481.923]: UTC time is past leap second table expiry date ts2phc[481.923]: source ts not valid ts2phc[482.823]: UTC time is past leap second table expiry date ts2phc[482.823]: source ts not valid ts2phc[482.824]: UTC time is past leap second table expiry date
hello , why is that? I ran ts2phc according to the previous settings, and this appeared:
This means /usr/share/zoneinfo/leap-seconds.list
is out of date. You can update the relevant distro package. Or download an updated one from
This means is out of date. You can update the relevant distro package. Or download an updated one from
/usr/share/zoneinfo/leap-seconds.list
ok thanks
This means is out of date. You can update the relevant distro package. Or download an updated one from
/usr/share/zoneinfo/leap-seconds.list
In which directory are the logs generated by ts2phc located? Can I set the path?
ts2phc logs using syslog.
ts2phc logs using syslog.
thank you
Why don’t you need to specify sdp?
cat /etc/linuxptp/ts2phc-generic.cfg
[global] ts2phc.extts_polarity both ts2phc.pulsewidth 100000000 leapfile /usr/share/zoneinfo/leap-seconds.list step_threshold 0.1
ts2phc.nmea_serialport /dev/ttyS9
ts2phc.nmea_baudrate 9600
logging_level 7
[enp1s0]