Issue with OP Entry Delay and Failures in LinuxCNC + EtherCAT System

Hi everyone, I'm back. I've carefully reviewed your replies and compared the dmesg information from my two systems.

I noticed that the second system has started the EOE thread again, and afterward, it logged many EOE-related errors/warnings.

I'm thinking about trying to completely disable the EOE function. I remember that I manually compiled the EtherCAT driver on the second system. According to the documentation, I configured it with something like --enable-eoe=no.

Is there a way for me to verify whether the EtherCAT I compiled actually has the EOE function disabled?

Of course, I can also recompile it, but I'm not entirely sure how to completely disable the EOE function.

Thank you for your help!

I've carefully reviewed the configurations of my two previous systems. For the first system, I didn't intentionally disable EOE, so it should have been enabled by default.

For the second system, I'm somewhat uncertain whether EOE was actually disabled when checking dmesg, because I've seen similar posts suggesting that disabling EOE might be beneficial.

Another finding: I tried compiling EtherCAT version 1.5.2 on the second system (replacing version 1.6.1), and it did successfully connect quickly and enter OP state.

However, during operation, when I used watch -n 1 ethercat slaves to monitor, I noticed that it would occasionally flash "0x0000:0x0000" for all slaves. At these moments, it seemed like all my IO modules disconnected, but they recovered quickly.

This instability makes me hesitant to use this setup.

I would greatly appreciate any help or suggestions from the community. Thank you very much!

I've only connected two slaves this time, and I've confirmed that neither of them has EOE functionality.
I've recorded the dmesg output after the rescan.
Why is the connection process relatively slow?
In which direction should I focus my troubleshooting efforts?
Thank you in advance for your guidance!

[ 5140.632281] EtherCAT 0: Re-scanning now.
[ 5140.640312] EtherCAT 0: Scanning bus.
[ 5141.334336] EtherCAT WARNING 0: 5 datagrams TIMED OUT!
[ 5141.334343] EtherCAT WARNING 0: 5 datagrams UNMATCHED!
[ 5142.333525] EtherCAT WARNING 0: 13 datagrams TIMED OUT!
[ 5142.333531] EtherCAT WARNING 0: 13 datagrams UNMATCHED!
[ 5143.332840] EtherCAT WARNING 0: 7 datagrams TIMED OUT!
[ 5143.332847] EtherCAT WARNING 0: 7 datagrams UNMATCHED!
[ 5144.331944] EtherCAT WARNING 0: 4 datagrams TIMED OUT!
[ 5144.331950] EtherCAT WARNING 0: 4 datagrams UNMATCHED!
[ 5145.335015] EtherCAT WARNING 0: 1 datagram TIMED OUT!
[ 5145.335022] EtherCAT WARNING 0: 1 datagram UNMATCHED!
[ 5146.333852] EtherCAT WARNING 0: 4 datagrams TIMED OUT!
[ 5146.333858] EtherCAT WARNING 0: 4 datagrams UNMATCHED!
[ 5147.332551] EtherCAT WARNING 0: 3 datagrams TIMED OUT!
[ 5147.332558] EtherCAT WARNING 0: 3 datagrams UNMATCHED!
[ 5149.333802] EtherCAT WARNING 0: 4 datagrams TIMED OUT!
[ 5149.333808] EtherCAT WARNING 0: 4 datagrams UNMATCHED!
[ 5150.256544] EtherCAT 0: Bus scanning completed in 9624 ms.
[ 5150.256552] EtherCAT 0: Using slave 0 as DC reference clock.
[ 5150.276608] EtherCAT 0: Slave states on main device: PREOP.
[ 5150.332789] EtherCAT WARNING 0: 2 datagrams TIMED OUT!
[ 5150.332794] EtherCAT WARNING 0: 2 datagrams UNMATCHED!
[ 5150.625634] EtherCAT ERROR 0-0: Reception of CoE download response failed: No response.
[ 5150.625642] EtherCAT WARNING 0-0: Failed to set number of entries.
[ 5150.625643] EtherCAT WARNING 0-0: Currently mapped PDO entries: 0x2600:01/16 0x2600:03/16 0x2600:04/16 0x2601:02/16. Entries to map: 0x2600:01/16 0x2600:03/16 0x2600:04/16 0x2601:02/16
[ 5150.625657] EtherCAT WARNING 0-0: Failed to configure mapping of PDO 0x1602.
[ 5151.331810] EtherCAT WARNING 0: 7 datagrams TIMED OUT!
[ 5151.331816] EtherCAT WARNING 0: 7 datagrams UNMATCHED!
[ 5155.334342] EtherCAT WARNING 0: 7 datagrams TIMED OUT!
[ 5155.334348] EtherCAT WARNING 0: 7 datagrams UNMATCHED!
[ 5156.316919] EtherCAT WARNING 0-0: Slave did not sync after 5000 ms.
[ 5156.327444] EtherCAT 0: Domain 0: Working counter changed to 3/6.
[ 5156.332974] EtherCAT WARNING 0: 5 datagrams TIMED OUT!
[ 5156.332979] EtherCAT WARNING 0: 5 datagrams UNMATCHED!
[ 5157.667444] EtherCAT 0: Domain 0: Working counter changed to 6/6.
[ 5158.334745] EtherCAT WARNING 0: 5 datagrams TIMED OUT!
[ 5158.334751] EtherCAT WARNING 0: 5 datagrams UNMATCHED!
[ 5158.671444] EtherCAT 0: Domain 0: 4 working counter changes - now 6/6.
[ 5160.132443] EtherCAT 0: Domain 0: Working counter changed to 0/6.
[ 5160.140193] EtherCAT ERROR 0-0: Reception of CoE SDO description response failed: No response.
[ 5160.144210] EtherCAT 0: Slave states on main device: OP.

That looks like the situation I had yesterday. Traced it to a bad kernel (I think I traced it there).
I upgraded the kernel to latest 6.12-48 and all problems went away.

Thank you for your response!
I'd like to try your method and hope it can resolve my current issue.
I haven't worked on updating the preempt-rt kernel before.
Are there any tutorial posts you could recommend to me?
Also, where can I download the 6.12-48 version you mentioned?

Current Kernel: Linux debian 6.1.0-23-rt-amd64

It was just an "apt upgrade", I am using debian trixie on that computer.
It may not be the kernel for you. It may not have been for me either, I messed around with many things that day.
You can always try an "apt update" followed by "apt upgrade". That should install the latest kernel and many more upgraded apps.

The messages you get is what I got using Realtek network interface.
What network interface do you have? Find out with "lspci"
 

I'm updating to Debian 13 and I see it's downloading version 6.12.48 now.
I'm really excited—this issue has been bothering me for nearly half a year, and I hope it gets resolved.

# lspci
01:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168/8411 PCI Express Gigabit Ethernet Controller (rev 15)
02:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168/8411 PCI Express Gigabit Ethernet Controller (rev 15)

Then at least part of the problem is with the Realtek network adapter.
There are two things you can try
1. Install the realtek-dkms drivers and use them instead of the default realtek drivers. Search in the forum for info, many people has done that.
2. Install a PCIe Intel network adapter.
 

Hi, I'm back to provide feedback.
Following your instructions, I have solved this problem. It worked after I updated the r8168-dkms network card driver.
The datagrams UNMATCHED situation no longer occurs. The device can enter the PREOP state normally as soon as it starts, and it can enter the OP state 100% of the time. Previously, some slaves were in INIT+E or PREOP+E state.
Thank you so much!