Hi again,
>From doing some further digging, the flagField[1] value only gets populated
in port_tx_announce(), which is called in the FD_MANNO_TIMER state. From
https://sourceforge.net/p/linuxptp/mailman/linuxptp-users/thread/CAEk6gTCnLGsdppUGCXHEc9Oh5%2B0%2BHt0zB0kW1MsZVSkazgGkZg%40mail.gmail.com/#msg30920783
Richard has explained that this is the Announce TX timer, only going into
this state when its time to send an announce message. Since I'm using the
Automotive profile, we inhibit the announce message, which explains why the
ptpTimescale value never gets actually gets set, regardless of what
tds->flags is.
So, does this mean that when using the Automotive Profile, we can never
have ptpTimescale set to 1? If so, is there a reason for this?
Regards,
On Wed, Jan 18, 2023 at 7:06 PM First Last wrote:
> Hello,
>
> I'm having some issues with the Automotive profile, namely that it doesn't
> recognize the presence of a GM, even if we have one port in MASTER state,
> and the only other port in the network in SLAVE state. I am verifying this
> by using pmc:
> gptp@gptp-desktop:~/linuxptp$ sudo pmc -u -b 1 -t 1 'GET PARENT_DATA_SET'
> sending: GET PARENT_DATA_SET
> 202564.fffe.7440f2-0 seq 0 RESPONSE MANAGEMENT PARENT_DATA_SET
> parentPortIdentity202564.fffe.7440f2-0
> parentStats 0
> observedParentOffsetScaledLogVariance 0x
> observedParentClockPhaseChangeRate0x7fff
> grandmasterPriority1 248
> gm.ClockClass 248
> gm.ClockAccuracy 0xfe
> gm.OffsetScaledLogVariance0x
> grandmasterPriority2 248
> grandmasterIdentity 202564.fffe.7440f2
> 489ebd.fffe.75b716-1 seq 0 RESPONSE MANAGEMENT PARENT_DATA_SET
> parentPortIdentity489ebd.fffe.75b716-0
> parentStats 0
> observedParentOffsetScaledLogVariance 0x
> observedParentClockPhaseChangeRate0x7fff
> grandmasterPriority1 250
> gm.ClockClass 255
> gm.ClockAccuracy 0xfe
> gm.OffsetScaledLogVariance0x
> grandmasterPriority2 248
> grandmasterIdentity 489ebd.fffe.75b716
>
> Above, we can see that both ports in the network show
> their parentPortIdentity to be themselves. Because of this, the
> TIME_STATUS_NP also shows that the gmIdentity of each port within the
> network is itself. I won't show it unless it's needed, but viewing
> PORT_DATA_SET shows the two ports in the network and their port state being
> MASTER and SLAVE.
>
> Background: I am using LinuxPTP in a simple 2 node network, using P2P
> because I plan on introducing a switch + more slaves soon. I'm using the
> automotive profile.
> On the GM machine, I am using the following command to start the ptp4l
> daemon: 'sudo ptp4l -i eno1 -f configs/automotive-master.cfg
> --step_threshold=1 -m' with the default automotive-master.cfg.
> On the Slave machine, I am using the following command to start the ptp4l
> daemon: 'sudo ptp4l -i eno1 -f configs/automotive-slave.cfg
> --step_threshold=1 -m' with the default automotive-slave.cfg. Both machines
> are running phc2sys as well.
>
> More importantly, I am also unable to use pmc to change ptpTimescale in
> the Sync/Fup/PDelayResp/PDelayRespFup messages sent to the slave (viewed on
> Wireshark). I see the correct ptpTimescale value of 1 when using 'GET
> GRANDMASTER_SETTINGS_NP' however this is not reflected in the flags portion
> of the PTP packet. How can I change ptpTimescale in PTP packets sent to the
> slave?
>
> Thanks in advance.
>
> Regards,
>
>
>
___
Linuxptp-users mailing list
Linuxptp-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linuxptp-users
