PCI ethernet device issue related to AER

[__name__]

On boot dmesg is full of the following message:

Code: Select all

[   72.199427] pcieport 0000:00:1c.0: AER: Multiple Corrected error received: 0000:01:00.0
[   72.199458] r8169 0000:01:00.0: PCIe Bus Error: severity=Corrected, type=Physical Layer, (Receiver ID)
[   72.199465] r8169 0000:01:00.0:   device [10ec:8136] error status/mask=00000001/00006000
[   72.199473] r8169 0000:01:00.0:    [ 0] RxErr                  (First)

This message repeats constantly making for a very large dmesg output.

I found this trick to "fix" the issue but it's not a real fix.
https://gist.github.com/flisboac/5a0711 ... 110bb383cd

Command I have used to "fix" issue:

Code: Select all

setpci -v -d 10ec:8136 CAP_EXP+0x8.w=0x0e

I don't actually use the ethernet on this laptop so it's not critical but the reporting in dmesg is excessive and I would like to eliminate the issue so I can better troubleshoot other issues while refining my kernel config.

Device:

Code: Select all

01:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL810xE PCI Express Fast Ethernet controller (rev 07)
        Subsystem: Dell RTL810xE PCI Express Fast Ethernet controller
        Kernel driver in use: r8169
        Kernel modules: r8169

Current firmware used for device:

Code: Select all

rtl_nic/rtl8106e-1.fw

I appreciate any insight on this issue.

[__name__]

I found the first instance of the error. It happens in the first second after boot. It is unrelated to the r8169 driver it seems. Here is the first report:

Code: Select all

[    0.828786] pcieport 0000:00:1c.0: AER: Multiple Corrected error received: 0000:01:00.0
[    0.830048] pci 0000:01:00.0: PCIe Bus Error: severity=Corrected, type=Physical Layer, (Receiver ID)
[    0.831305] pci 0000:01:00.0:   device [10ec:8136] error status/mask=00000001/00006000
[    0.832524] pci 0000:01:00.0:    [ 0] RxErr                  (First)

Does anyone now how to further troubleshoot this issue?

[__name__]

I have been searching for answers to this problem and it seems to be a common bug for the Skylake CPUs and Intel mobos with Sunrise Point host controller. I have found Ubuntu bugs from 2015 with exactly the same issue.

As for a solution, I have just found this webpage that describes the codebase responsible. I will try to dive into the gentoo-sources code to see if I can find the lines but it is a bit above my skill level to implement any sort of fix.

https://lore.kernel.org/all/20151229155 ... lhost/T/#u

[pietinger]

In the meantime you might disable the PCIE advanced error reporting with a kernel command line parameter: pci=noaer

[__name__]

I did that but I also have some new kernel configurations that are creating different error messages in dmesg from the same device. I will have to track those down and recompile my kernel. I'm using gentoo-sources-6.6.21 as of now.