Failed to create EtherCAT Master

MattiaDeiRossi

Good morning everyone,

We're building a new robot, and when I power it on, the following error appears:

2026-06-05 08:22:44.402 +0000 [ERROR] Failed to create EtherCAT master, id: 0, driver name: EtherLAB

However, after restarting the PC or the Motorcortex service, the system returns to working order.

Do you have any suggestions?

Philippe

Are you using MCX-RTOS or something else (Ubuntu)? What kind of hardware are you using? Are you using our reference design IPC (MCX-A4+)?
What does dmesg show? Maybe the network card is not yet online. Did you change the order of loading the kernel modules or services? The Ethercat driver should be loaded after networking is up and before motorcortex is started.
On our MCX-RTOS everything should be configured correctly.

MattiaDeiRossi

Hi Philippe,

I followed the same Clonezilla-based OS installation procedure that I used for the other two robots on the same PC, and it worked correctly in those cases.

I didn't change anything about kernel modules or services.

So in summary:

same hardware pc
same .bin of the project ( a cloned version of the older one )
same OS

I have no evidence from the dmesg that something happens, in fact this issue appears sometimes, not always.

Philippe

What brand and type of computer are you using in all cases?
I can only conclude that it must be some hardware issue (network card defective), or maybe a setting in the BIOS.

MattiaDeiRossi

May this dmesg can help a little bit:


[    8.259695] EtherCAT 0: Starting EtherCAT-IDLE thread.
[    8.260726] EtherCAT 0: Link state of ecm0 changed to UP.
[    8.261761] EtherCAT 0: 6 slave(s) responding on main device.
[    8.261766] EtherCAT 0: Slave states on main device: INIT.
[    8.263766] EtherCAT 0: Scanning bus.
[    8.813547] EtherCAT: Requesting master 0...
[    9.390248] EtherCAT 0: Waiting for slave scan interrupted by signal.
[    9.390252] EtherCAT ERROR 0: Failed to enter OPERATION phase!
[    9.411390] acpi device:0f: registered as cooling_device9
[    9.413281] acpi device:10: registered as cooling_device10
[    9.883479] EtherCAT 0: Bus scanning completed in 1621 ms.
[    9.883486] EtherCAT 0: Using slave main-0 as DC reference clock.
[    9.884508] EtherCAT 0: Slave states on main device: PREOP.
[    9.891594] EtherCAT 0: Slave states on main device: INIT, PREOP.
[    9.905745] EtherCAT 0: Slave states on main device: PREOP.
[  464.005742] EtherCAT: Requesting master 0...
[  464.005748] EtherCAT: Successfully requested master 0.
[  466.659032] EtherCAT 0: Domain0: Logical address 0x00000000, 492 byte, expected working counter 18.
[  466.659037] EtherCAT 0:   Datagram domain0-0-main: Logical offset 0x00000000, 492 byte, type LRW at 00000000d1016f91.
[  466.659651] EtherCAT 0: Master thread exited.
[  466.659653] EtherCAT 0: Starting EtherCAT-OP thread.
[  466.717692] EtherCAT 0: Slave states on main device: INIT.
[  466.731691] EtherCAT 0: Slave states on main device: PREOP.
[  467.053681] EtherCAT 0: Slave states on main device: PREOP, OP.
[  468.369640] EtherCAT 0: Slave states on main device: OP.

So it seems that there was a trigger of a signal like: SIGINT, SIGTERM

Philippe

Hi Mattia,

we cannot reproduce this on our side. Did you try a different IPC hardware? Did you also try it with our latest RTOS image that you can download from our store?

MattiaDeiRossi

Philippe

Ok, thanks for the reply, I can't change the IPC but I'm trying changing the RTOS image

Actually with

same pc
mcx-image-mcx-intel.rootfs-2026.05.1-0-g34c0e28-20260508091432.wic
same project

it seems working, but I keep it monitored to validate this version of RTOS