[etherlab-users] e1000e driver freeze on kernel 4.19.59 Xenomai 3.1

[gbcwbz]

Hi all,
  I'm using Linux kernel 4.19.59 with Xenomai 3.1 , and with Cobal posix API.

My NIC is a standalone PCI-E card:
>
>   01:00.0 Ethernet controller: Intel Corporation 82571EB/82571GB Gigabit Ethernet Controller D0/D1 (copper applications) (rev 06)
>   01:00.1 Ethernet controller: Intel Corporation 82571EB/82571GB Gigabit Ethernet Controller D0/D1 (copper applications) (rev 06)

ethercat master version 1.5.2 with gavin-20190904 patchset from
https://github.com/ribalda/ethercat

It ran smoothly in IDLE mode when I start ethercat with:
/etc/init.d/ethercat start
>
> 0  0:0  OP  +  EK1100 EtherCAT Coupler (2A E-Bus)
> 1  3:0  OP  +  EL2624 4K. Relais Ausgang, Schlie�er (125V AC / 30V DC)


I got an example from
https://hg.beremiz.org/etherlabmaster/file/b5e3de0467a5/examples/xenomai_posix/main.c
system freeze after about 2000 cycles.
If I delete
>
>         ecrt_master_receive(master);
>         ecrt_domain_process(domain1);

and
>
>         ecrt_domain_queue(domain1);
>         ecrt_master_send(master);
system never freeze.

Then I applied the patch from
http://lists.etherlab.org/pipermail/etherlab-users/2018/003435.html,
and now after 24 hours,
It's still working. Only one warning:
>
> [ 6525.835070] EtherCAT 0: Starting EtherCAT-OP thread.
> [ 6525.844168] EtherCAT WARNING 0: No app_time received up to now, abort DC time offset calculation.
> [ 6525.854170] EtherCAT 0: Slave states on main device: INIT.
> [ 6525.862170] EtherCAT 0: Slave states on main device: INIT, PREOP.
> [ 6525.868186] EtherCAT 0: Slave states on main device: PREOP.
> [ 6525.868278] EtherCAT 0: Domain 0: Working counter changed to 1/1.
> [ 6525.874168] EtherCAT 0: Slave states on main device: SAFEOP.
> [ 6525.880166] EtherCAT 0: Slave states on main device: OP.

But I don't know why this patch fixed the problem, do you have this
problem with e1000e and xenomai ?

Bryan Cheng