I225-IT Ethernet (8086:0d9f) does not work on AAEON's EHL Board

[Kent Lin]

[Summary]
I225-IT Ethernet (8086:0d9f) does not work on AAEON's EHL Board
[Steps to reproduce]
1. Install https://cdimage.ubuntu.com/focal/daily-preinstalled/20210929/ to the system
2. Boot to desktop
3. Check if all Ethernet get IP address
[Expected result]
All Ethernet Port should up and runing and get the IP address.
[Actual result]
I225-IT Ethernet (8086:0d9f) does not work and did not get IP address
[Failure rate]
100%
[Additional information]
CID: 202109-29492
SKU: N/A
system-manufacturer: AAEON
system-product-name: UPN-EHL01
bios-version: UNEHAM0D
CPU: Intel Atom(R) x6425RE Processor @ 1.90GHz (4x)
GPU: 00:02.0 VGA compatible controller [0300]: Intel Corporation Device [8086:4571] (rev 01)
kernel-version: 5.13.0-1004-intel

[Stage]
Issue reported and logs collected right after it happened

[Kent Lin (kent-jclin)]

Automatically attached

[Kent Lin (kent-jclin)]

Automatically attached

[Kent Lin (kent-jclin)]

Automatically attached

[Kent Lin (kent-jclin)]

AAEON's feedback from Intel for this issue:

cid:image001.jpg@01D7AA33.A0E5D5B0

Since you are using NVM 1.79, please manually add two patches below. IGC driver should work. These patches will be in Kernel 5.15.
https://git.kernel.org/pub/scm/linux/kernel/git/tnguy/next-queue.git/commit/drivers/net/ethernet/intel/igc?h=dev-queue&id=7c496de538eebd8212dc2a3c9a468386b264d0d4
https://git.kernel.org/pub/scm/linux/kernel/git/tnguy/next-queue.git/commit/drivers/net/ethernet/intel/igc?h=dev-queue&id=47bca7de6a4fb8dcb564c7ca14d885c91ed19e03

Following this approach, we added these two kernel patches to Yocto, and the result was fine.

As above mentioned by Intel, these patches will be included in Kernel 5.15. May I know if there is any possible to implement the patches earlier such as 5.13 or 5.14? Currently we saw that TGL with 5.11 already has this problem and for EHL platform, shall we assume Ubuntu 20.04.30 which is going be released to support EHL on Oct.14 (correct me if wrong) can have the patches included?

[Doug Jacobs (djacobs98)]

On my EHL-Aaeon, Ubuntu only sees 1 Ethernet interface, which corresponds to the bottom Ethernet port. This one works as normal. The top port is not recognized by Ubuntu, although it lights up when you plug in a cable.

[Kent Lin (kent-jclin)]

@Anthony,
Test Kernel fix the issue: https://people.canonical.com/~ypwong/linux-intel-sprint2/

u@linux:~$ ip a
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
       valid_lft forever preferred_lft forever
    inet6 ::1/128 scope host
       valid_lft forever preferred_lft forever
2: enp1s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP group default qlen 1000
    link/ether 00:07:32:61:5b:39 brd ff:ff:ff:ff:ff:ff
    inet 10.102.159.126/22 brd 10.102.159.255 scope global dynamic noprefixroute enp1s0
       valid_lft 6877sec preferred_lft 6877sec
    inet6 fe80::f554:7623:6398:b8d8/64 scope link noprefixroute
       valid_lft forever preferred_lft forever
3: enp2s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP group default qlen 1000
    link/ether 00:07:32:61:5b:3a brd ff:ff:ff:ff:ff:ff
    inet 10.102.159.135/22 brd 10.102.159.255 scope global dynamic noprefixroute enp2s0
       valid_lft 6880sec preferred_lft 6880sec
    inet6 fe80::2e77:71f2:d10d:962/64 scope link noprefixroute
       valid_lft forever preferred_lft forever
u@linux:~$

01:00.0 Ethernet controller [0200]: Intel Corporation I210 Gigabit Network Connection [8086:157b] (rev 03)
 Subsystem: Intel Corporation I210 Gigabit Network Connection [8086:0000]
 Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx+
 Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
 Latency: 0
 Interrupt: pin A routed to IRQ 18
 Region 0: Memory at 7fe00000 (32-bit, non-prefetchable) [size=128K]
 Region 2: I/O ports at 4000 [size=32]
 Region 3: Memory at 7fe20000 (32-bit, non-prefetchable) [size=16K]
 Capabilities: <access denied>
 Kernel driver in use: igb
 Kernel modules: igb
00: 86 80 7b 15 07 04 10 00 03 00 00 02 00 00 00 00
10: 00 00 e0 7f 00 00 00 00 01 40 00 00 00 00 e2 7f
20: 00 00 00 00 00 00 00 00 00 00 00 00 86 80 00 00
30: 00 00 00 00 40 00 00 00 00 00 00 00 ff 01 00 00

02:00.0 Ethernet controller [0200]: Intel Corporation Device [8086:0d9f] (rev 03)
 Subsystem: Intel Corporation Device [8086:0000]
 Control: I/O- Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx+
 Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
 Latency: 0
 Interrupt: pin A routed to IRQ 16
 Region 0: Memory at 7fc00000 (32-bit, non-prefetchable) [size=1M]
 Region 3: Memory at 7fd00000 (32-bit, non-prefetchable) [size=16K]
 Capabilities: <access denied>
 Kernel driver in use: igc
 Kernel modules: igc
00: 86 80 9f 0d 06 04 10 00 03 00 00 02 00 00 00 00
10: 00 00 c0 7f 00 00 00 00 00 00 00 00 00 00 d0 7f
20: 00 00 00 00 00 00 00 00 00 00 00 00 86 80 00 00
30: 00 00 00 00 40 00 00 00 00 00 00 00 ff 01 00 00

[Ubuntu Kernel Bot (ubuntu-kernel-bot)]

This bug is awaiting verification that the linux-intel-5.13/5.13.0-1006.6 kernel in -proposed solves the problem. Please test the kernel and update this bug with the results. If the problem is solved, change the tag 'verification-needed-focal' to 'verification-done-focal'. If the problem still exists, change the tag 'verification-needed-focal' to 'verification-failed-focal'.

If verification is not done by 5 working days from today, this fix will be dropped from the source code, and this bug will be closed.

See https://wiki.ubuntu.com/Testing/EnableProposed for documentation how to enable and use -proposed. Thank you!

[Pierre Equoy (pieq)]

Trying this with

CID: 202109-29496
SKU: AAEON UPN-EHL01
Image used: ubuntu-core-20-amd64+intel-iot.img.xz (20211014.2)
kernel: 5.13.0-1007-intel

When plugging in both Ethernet ports, only one of them retrieves an IP:

$ ip a
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
       valid_lft forever preferred_lft forever
    inet6 ::1/128 scope host
       valid_lft forever preferred_lft forever
2: enp2s0: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN group default qlen 1000
    link/ether 00:07:32:61:5b:08 brd ff:ff:ff:ff:ff:ff
3: enp1s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP group default qlen 1000
    link/ether 00:07:32:61:5b:07 brd ff:ff:ff:ff:ff:ff
    inet 192.168.1.126/24 brd 192.168.1.255 scope global dynamic enp1s0
       valid_lft 86268sec preferred_lft 86268sec
    inet6 fe80::207:32ff:fe61:5b07/64 scope link
       valid_lft forever preferred_lft forever

Nothing happens in `dmesg -w` when I unplug/replug an Ethernet cable in enp2s0.

[Doug Jacobs (djacobs98)]

I see this same behavior on the TGL-Aaeon running Core.

I even tried to manually edit /etc/netplan/00-snapd-config.yaml but I could not make it bring the interface up.

When the cable is plugged in the link lights are on but the system doesn't seem to acknowledge the connection.

The other port works just fine.

[Doug Jacobs (djacobs98)]

TGL-Aaeon:

20.04LTS Desktop: Detects both interfaces and allows them to get an IP# when you plug in a cable.

IOTG Desktop (Oct. 13 build): Detects both interfaces and allows them to get an IP# when you plug in a cable. (Desktop Manual network tests pass.)

IOTG Core (Oct. 14 build): If you perform initial configuration while both ports and plugged in, both ports will get an IP#.

[Pierre Equoy (pieq)]

After additional discussion with other members, I retried with the same config as in comment #9:

CID: 202109-29496
SKU: AAEON UPN-EHL01
Image used: ubuntu-core-20-amd64+intel-iot.img.xz (20211014.2)
kernel: 5.13.0-1007-intel

Turns out, as Doug mentioned in comment #11, the UC20 configuration is based on the initial setup wizard when installing UC20. In my case, I had only plugged one of the two network interfaces (enp1s0), so only this one was configured. The other was marked as `disabled` and never used, even when plugging a cable afterwards.

The netplan configuration looked like this:

$ cat /etc/netplan/00-snapd-config.yaml
# This is the network config written by 'console-conf'
network:
  ethernets:
    enp1s0:
      dhcp4: true
  version: 2

In order to change this, I edited it like so:

$ cat /etc/netplan/00-snapd-config.yaml
# This is the network config written by 'console-conf'
network:
  ethernets:
    enp1s0:
      dhcp4: true
    enp2s0:
      dhcp4: true
  version: 2

then ran `sudo netplay apply`, and the second interface (enp2s0) got an IP shortly afterwards.

So, at least on my side (Aaeon EHL board), it's a non-issue.

[Sachin Mokashi (sachinmokashi)]

Can you please also check the BIOS settings -

In EHL CRB, PCH and PSE0 already enabled by default . In order to enable PSE1, need to change from ‘None’ to ‘Host owned with pin muxed’ in BIOS

BIOS configuration for setting both PSE GBE0 and PSE GBE1 to host owned:
• IntelAdvanced Menu→PCH-IO Configuration→PSE configuration→GBE0→Host owned
• IntelAdvanced Menu→PCH-IO Configuration→PSE configuration→GBE1→Host owned