I211 periodic disconnection w/ igb driver in Linux 5.10

[Arya]

I have an Asus Prime X570 Pro motherboard with an I211 controller onboard. Under Debian Linux (5.10.4-1) with the igb module (unsure what version) and default settings, the link drops and comes up frequently. Sometimes I see a "reset adapter" message. (See syslog snippet attached.)

• The problem does not occur under Windows.
• Changing cables and switch ports does not help.
• Other devices plugged into this switch (model Netgear GS316) do not exhibit this problem.
• However, using a different switch (old gigabit 3com thing), the connection is stable.
• I also do not observe link drops as frequently if I force 1000 duplex operation and turn off autonegotiation with ethtool. I cannot be sure that this eliminates the issue, though.

I suspect that something about how this switch does autonegotiation is broken and something about how the igb driver handles it is also peculiar.

How can I diagnose the real cause of the problem? I tried turning the "debug" info for the module to 16 but that just included a memory dump or some register file or something, no more helpful messages. I need to know if have to replace my switch, motherboard, driver, operating system...

(I have been unable to compile and install an up-to-date igb module on my own due to incompatibility with my toolchain.)

[Arya]

Actually I have now observed similar behaviour under Windows, too. The negotiation is probably failing in hardware (PHY or MAC?) so this is nothing to do with the driver. Is there a way to debug the negotiation there?

[Mike_Intel]

Hello Arya,

Thank you for posting in Intel Ethernet Communities. 

After reading your 2 post and all your troubleshooting steps and just like what you said. There is a big chance that this is not a software or driver issue. At this point, since the network card is onboard, it is best to contact Asus for further hardware debug and troubleshooting.

If you have questions, please let us know. In case we do not hear from you, we will make a follow up after 3 workings days. Thank you.

Best regards,

Michael L.

Intel® Customer Support

[Mike_Intel]

Hello Arya,

We hope you had a great weekend. We just want to check if you already tried contacting Asus regarding this issue.

If you have questions, please let us know. In case we do not hear from you, we will make a follow up after 3 workings days. Thank you.

Best regards,

Michael L.

Intel® Customer Support

[Arya]

I did not contact Asus. I replaced the switch and the problem has gone away.

I wanted help from this community, really from Intel, with introspection into what exactly the problem was so that I could better determine which device, or which part of the interaction, was creating the fault. But no chance of that if I contact Asus. And apparently no chance of that here.

[Mike_Intel]

Hello Arya,

I am sorry to hear that. We just thought you need further help in debugging the onboard network card.

However, we are glad that the issue is now fixed and the cause was the old switch.

We will close this thread now and we hope we can talk to you again soon.

Thank you.

Best regards,

Michael L.

Intel® Customer Support