Hi linuxptp community!
Hope someone can give me an idea of what could be probably wrong with the
following...
I'm trying to run ptp4l(v2.0) in a custom hardware(PWERPC e500) running a
custom linux image( Linux version: 4.1.38 )with the following command:
$ptp4l -m -i eth0 -f default.cfg -S
The issue that we're facing is that when ptp4l call recvmsg() inside sk.c
i'm getting just one of the two ancillary messages that i'm supposed to
receive
i.e. i'm only receiving cm_level SOL_IP with cm_type IP_RECVERR message
within the function sk_receive() there is a routine that looks for the
options received in a ancillary buffer
i.e. for (cm = CMSG_FIRSTHDR(&msg); cm != NULL; cm = CMSG_NXTHDR(&msg, cm))
{ ... }
And i can verify that it is doing this just once and is in fact, the
RECVERR message with the ee_errno field equals to ENOMSG
so according to section 2.1.1 in [1] i'm missing cm_level SOL_SOCKET and
cm_type SCM_TIMESTAMPING message and because of this
ptp4l shows the following log:
-----------------------------------------------------------------------------------------------------------
#[/root]:ptp4l -m -i eth0 -f default.cfg -S
ptp4l[273.015]: port 1: INITIALIZING to LISTENING on INIT_COMPLETE
ptp4l[273.015]: port 0: INITIALIZING to LISTENING on INIT_COMPLETE
ptp4l[280.672]: port 1: LISTENING to MASTER on
ANNOUNCE_RECEIPT_TIMEOUT_EXPIRES
ptp4l[280.672]: selected local clock 00302f.fffe.023379 as best master
ptp4l[280.672]: assuming the grand master role
ptp4l[281.677]: missing timestamp on transmitted sync <<<<---------!
ptp4l[281.677]: port 1: MASTER to FAULTY on FAULT_DETECTED (FT_UNSPECIFIED)
ptp4l[297.678]: port 1: FAULTY to LISTENING on INIT_COMPLETE
ptp4l[304.170]: port 1: LISTENING to MASTER on
ANNOUNCE_RECEIPT_TIMEOUT_EXPIRES
-----------------------------------------------------------------------------------------------------------
ethtool -T eth0 shows the following:
-----------------------------------------------------------------------------------------------------------
Time stamping parameters for eth0:
Capabilities:
hardware-transmit (SOF_TIMESTAMPING_TX_HARDWARE)
software-transmit (SOF_TIMESTAMPING_TX_SOFTWARE)
hardware-receive (SOF_TIMESTAMPING_RX_HARDWARE)
software-receive (SOF_TIMESTAMPING_RX_SOFTWARE)
software-system-clock (SOF_TIMESTAMPING_SOFTWARE)
hardware-raw-clock (SOF_TIMESTAMPING_RAW_HARDWARE)
PTP Hardware Clock: 0
Hardware Transmit Timestamp Modes:
off (HWTSTAMP_TX_OFF)
on (HWTSTAMP_TX_ON)
one-step-sync (HWTSTAMP_TX_ONESTEP_SYNC)
Hardware Receive Filter Modes:
none (HWTSTAMP_FILTER_NONE)
all (HWTSTAMP_FILTER_ALL)
-----------------------------------------------------------------------------------------------------------
Linux Config file includes
-----------------------------------------------------------------------------------------------------------
- CONFIG_PPS
- CONFIG_NETWORK_PHY_TIMESTAMPING
- PTP_1588_CLOCKPTP
-----------------------------------------------------------------------------------------------------------
Following the skb_tstamp_tx() i can see when sk_error_queue is filled with
ENOMSG but where does the second message(cm_level SOL_SOCKET and cm_type
SCM_TIMESTAMPING) comes from?
Please let me know how to debug this issue; what information i can provide
in order to give a better context of what is happening? at which level i
can confirm(i.e. print) that socket's error_queue was filled properly? does
this seems to be the real cause of the problem or should i look another
path?
Thank you very much for your support!
[1] https://www.kernel.org/doc/Documentation/networking/timestamping.txt
_______________________________________________
Linuxptp-users mailing list
Linuxptp-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linuxptp-users
