[Milhouse]

Here is the dmesg | pasteinit: http://ix.io/1raX. #1016 plays fine. #1016x does not. 

iperf -c 10.0.1.55 -u -P 1 -i 1 -p 5001 -f G -b 100.0M -t 10 -T 1

Can you test TCP with iperf rather than UDP?

This doesn't look good:

text:


[   19.362489] pcieport 0000:00:1c.0: AER: Multiple Corrected error received: 0000:00:1c.0

[   19.362505] pcieport 0000:00:1c.0: PCIe Bus Error: severity=Corrected, type=Data Link Layer, (Transmitter ID)

[   19.362509] pcieport 0000:00:1c.0:   device [8086:9c14] error status/mask=00001000/00002000

[   19.362513] pcieport 0000:00:1c.0:    [12] Timeout               

[   22.571954] pcieport 0000:00:1c.0: AER: Corrected error received: 0000:00:1c.0

[   22.571976] pcieport 0000:00:1c.0: PCIe Bus Error: severity=Corrected, type=Data Link Layer, (Transmitter ID)

[   22.571983] pcieport 0000:00:1c.0:   device [8086:9c14] error status/mask=00001000/00002000

[   22.571991] pcieport 0000:00:1c.0:    [12] Timeout               

[   22.751504] pcieport 0000:00:1c.0: AER: Corrected error received: 0000:00:1c.0

No idea what is causing that at the moment. Unfortunately the log spam has resulted in the log buffer being truncated.

Can you add log_buf_len=16M to your kernel command line, then boot with #1022 (4.18.14, no problems) and run dmesg | pastebinit, then upgrade to #1106 (4.19.1, kernel errors) and again run dmesg | pastebinit, pasting both links here - hopefully this time we'll have complete kernel logs for both a working and non-working kernel. I'll then be able to do some digging. It looks like 4.19 introduced quite a number of r8169-related bugs.