lsusb inside guest shows one captured device twice, but no mention of other captured device

[HarryMan]

[I want to apologize to anyone who has been looking into this and gotten frustrated with all of the changes and my having moved this issue back and forth between forums. I've been at this for a few days and messed up those posts. So I will start all over again. I sincerely appreciate the help.]

My host here is Devuan Chimaera (debian 11).
The guest is openwrt, a custom linux using musl.
I'm running virtualbox 6.1.32.

I have a virtualbox VM that captures 2 usb ethernet devices. They are different brands with different chipsets.

Normally, I expect to see something like this when I call lsusb inside the guest:

Code: Select all

# lsusb 
Bus 001 Device 003: ID 2357:0601 Realtek USB 10/100/1000 LAN
Bus 001 Device 002: ID 0b95:1790 ASIX Elec. AX88179
Bus 001 Device 001: ID 1d6b:0002 Linux 5.4.188 ehci_hcd EHCI Host Controller
Bus 002 Device 002: ID 80ee:0021 VirtualBox USB Tablet
Bus 002 Device 001: ID 1d6b:0001 Linux 5.4.188 ohci_hcd OHCI PCI host controller

Intermittently--meaning from boot to reboot of this vm, warm or cold--and without changing the virtualbox usb configuration or replugging any devices, I will see a listing of lsusb inside the guest like this:

Code: Select all

# lsusb
Bus 001 Device 003: ID 2357:0601 Realtek USB 10/100/1000 LAN
Bus 001 Device 002: ID 2357:0601 Realtek USB 10/100/1000 LAN
Bus 001 Device 001: ID 1d6b:0002 Linux 5.4.188 ehci_hcd EHCI Host Controller
Bus 002 Device 002: ID 80ee:0021 VirtualBox USB Tablet
Bus 002 Device 001: ID 1d6b:0001 Linux 5.4.188 ohci_hcd OHCI PCI host controller

The two usb devices are actually USB 3, but I have not yet implemented that in virtualbox. The host kernel requires an upgrade, which I hope to do soon to take advantage of the higher speeds. There really is another realtek device on the host, and I hope that is not interfering somehow with virtualbox's capture management. I am NOT capturing that other realtek device to the vm in question. That device is already captured by another vm, where it is working correctly.

I will note that the serial numbers for one of the devices is a little funky, being something like "000000" which probably isn't helping matters. I have some new dongles on order that should arrive in the next few days. But given that on the host side, devuan is correctly handling these devices and I can use them like any other ethernet device from the host, it seems to only be a problem when the devices are captured by the VM. I have tried rebooting the host, even cold booting, with no change in behavior.

Attached is the log from one recent run of this vm with this problem.

Again, this is one of those intermittent problems, and in my experience with other software, these type are very hard to run down and fix. I'm more than happy to provide additional information. Thank you for your help.

[HarryMan]

More drama. I had forgotten I'd rolled back to 6.1.32 for some reason--I think it had to do with my host upgrade to devuan chimaera (debian 11). I had been having certain issues and thought that maybe 6.1.34 was the source of those issues. Given I am still having some problems even with 6.1.32, I should upgrade to 6.1.34.

[HarryMan]

I have re-installed/upgraded to 6.1.34 again. So far, I have not noticed this problem again, but will post back if it does happen.