DS414 with 2xClub 3D CAC-1420 hangs driver

[endorianix]

Description of the problem

As part of testing the possibility of using 2x Club 3D CAC-1420 adapters with my old DS414 (just for fun).
The configuration works correctly with a single adapter (tested on both rear USB ports). However, when connecting the second adapter (same Club 3D CAC-1420) it isnt recognized as network interface.

I used:
sudo bash /var/packages/r8152/scripts/install-udev-rules
However, this does not produce the desired effect.

Description of your products

DS414
Club 3D CAC-1420 (x2)

Linux Syn_DS 3.2.101 #42962 SMP Mon Aug 19 15:11:03 CST 2024 armv7l GNU/Linux synology_armadaxp_ds414
DSM 7.1.1-42962 Update 7

Driver used: r8152-armadaxp-2.17.1-2_7.1 (the last one with support for armadaxp)

Output of dmesg command

[ 6468.403172] r8152 4-1:1.0: eth3: chip rev 14
[ 6468.407650] usb 4-1: USB disconnect, device number 122
[ 6468.509702] etxhci_hcd-161118 0000:01:00.0: xHCI etxhci_drop_endpoint called with disabled ep ba5348e0
[ 6468.519042] etxhci_hcd-161118 0000:01:00.0: xHCI etxhci_drop_endpoint called with disabled ep ba53490c
[ 6468.528417] etxhci_hcd-161118 0000:01:00.0: xHCI etxhci_drop_endpoint called with disabled ep ba534938
[ 6468.659926] usb 4-1: new SuperSpeed USB device number 123 using etxhci_hcd-161118
[ 6468.687411] drivers/usb/core/ethub.c (1935) Same device found. Change serial to 401000000
[ 6468.819917] usb 4-1: reset SuperSpeed USB device number 123 using etxhci_hcd-161118
[ 6468.843527] usb 4-1: device firmware changed
[ 6468.847876] r8152 4-1:1.0: (unregistered net_device): get_registers -19
[ 6468.854603] r8152 4-1:1.0: (unregistered net_device): Get ether addr fail
[ 6468.861909] r8152 4-1:1.0: eth3: v2.17.1 (2023/06/13)
[ 6468.866994] r8152 4-1:1.0: eth3: This product is covered by one or more of the following patents:
[ 6468.867002] US6,570,884, US6,115,776, and US6,327,625.

[ 6468.882753] r8152 4-1:1.0: eth3: chip rev 14
[ 6468.887254] usb 4-1: USB disconnect, device number 123

Output of lsusb command

|__usb1 1d6b:0002:0302 09 2.00 480MBit/s 0mA 1IF (Linux 3.2.101 ehci_hcd Marvell Orion EHCI ehci_marvell.0) hub
|__usb2 1d6b:0002:0302 09 2.00 480MBit/s 0mA 1IF (Linux 3.2.101 ehci_hcd Marvell Orion EHCI ehci_marvell.1) hub
|__usb3 1d6b:0002:0302 09 2.00 480MBit/s 0mA 1IF (Linux 3.2.101 etxhci_hcd-161118 Etron xHCI Host Controller 0000:01:00.0) hub
|__usb4 1d6b:0003:0302 09 3.00 5000MBit/s 0mA 1IF (Linux 3.2.101 etxhci_hcd-161118 Etron xHCI Host Controller 0000:01:00.0) hub
|__4-2 0bda:8156:3104 00 3.20 5000MBit/s 64mA 1IF (Realtek USB 10/100/1G/2.5G LAN 401000001)
|__4-1 0bda:8156:3104 00 3.20 5000MBit/s 0IFs (Realtek USB 10/100/1G/2.5G LAN 401000000)

Output of ifconfig -a command

(no eth3)

eth2 Link encap:Ethernet HWaddr A0:CE:C8:AF:C7:40
inet addr:192.168.0.248 Bcast:192.168.0.255 Mask:255.255.255.0
inet6 addr: fe80::a2ce:c8ff:feaf:c740/64 Scope:Link
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:3737 errors:0 dropped:0 overruns:0 frame:0
TX packets:561 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:1041280 (1016.8 KiB) TX bytes:65183 (63.6 KiB)

[endorianix]

Small update:

Changed one CAC-1420 adapter to Asus USB-C2500 (0b05:1976:3105) and now I got 2 working ETHs.

|__usb1 1d6b:0002:0302 09 2.00 480MBit/s 0mA 1IF (Linux 3.2.101 ehci_hcd Marvell Orion EHCI ehci_marvell.0) hub
|__usb2 1d6b:0002:0302 09 2.00 480MBit/s 0mA 1IF (Linux 3.2.101 ehci_hcd Marvell Orion EHCI ehci_marvell.1) hub
|__usb3 1d6b:0002:0302 09 2.00 480MBit/s 0mA 1IF (Linux 3.2.101 etxhci_hcd-161118 Etron xHCI Host Controller 0000:01:00.0) hub
|__usb4 1d6b:0003:0302 09 3.00 5000MBit/s 0mA 1IF (Linux 3.2.101 etxhci_hcd-161118 Etron xHCI Host Controller 0000:01:00.0) hub
|__4-1 0b05:1976:3105 00 3.20 5000MBit/s 64mA 1IF (Realtek USB 10/100/1G/2.5G LAN 5013000001)
|__4-2 0bda:8156:3104 00 3.20 5000MBit/s 64mA 1IF (Realtek USB 10/100/1G/2.5G LAN 401000001)

[bb-qq]

|__4-2 0bda:8156:3104 00 3.20 5000MBit/s 64mA 1IF (Realtek USB 10/100/1G/2.5G LAN 401000001)
|__4-1 0bda:8156:3104 00 3.20 5000MBit/s 0IF (Realtek USB 10/100/1G/2.5G LAN 401000000)

From this output, it appears that the port is not getting enough power.
The USB-C2500 has a more power-efficient 8156B chip, which may be the reason it worked so well.

By the way what is your motivation for adding two NICs?

[endorianix]

|__4-2 0bda:8156:3104 00 3.20 5000MBit/s 64mA 1IF (Realtek USB 10/100/1G/2.5G LAN 401000001)
|__4-1 0bda:8156:3104 00 3.20 5000MBit/s 0IF (Realtek USB 10/100/1G/2.5G LAN 401000000)

From this output, it appears that the port is not getting enough power. The USB-C2500 has a more power-efficient 8156B chip, which may be the reason it worked so well.

I was able to change serial number of one CAC-1420 and it works - so it looks that on old kernel hot plug fix doesnt work (maybe because of old udev?)

By the way what is your motivation for adding two NICs?

Just for fun (but at the end I see some performance gains, notably with SMB multichannel - which by definition should not work on this device... but works)

[bb-qq]

so it looks that on old kernel hot plug fix doesnt work (maybe because of old udev?)

I don't know the details, which is a bit odd since there is no intervening class driver for this output of lsusb.

Just for fun (but at the end I see some performance gains, notably with SMB multichannel - which by definition should not work on this device... but works)

Understand. If the goal is to increase throughput, RTL8157 would be an option, just to let you know.