"Maybe the USB cable is bad?" error at boot

Software & Operating Systems Linux
1

I’ve recently upgraded to AM5 and with my new setup, I see a handful of these error messages during boot.
By unplugging an USB device at a time, I’ve narrowed it down to to only happen if I use any USB type C connectors on the PC.
The device in question works once Linux is booted, however the errors cause quite a significant delay to boot time.

I’m running Opensuse Tumbleweed with kernel 6.8.4-rc1-1-default at the time of writing, though it happened with older kernels as well.

Does anyone know this error and better yet a fix for it?

2

What is it, external storage? Webcam?

What exactly are the errors? Dmesg, journalctl or the like should turn something up.

3

Sorry about that that, it does apparently not matter what I stick in a Type C port as well as which port.
If something is placed in a Type C port at boot I get this in dmesg:

usb usb2-port6: cannot enable. Maybe the USB cable is bad?

I’ve tested it using an ethernet adapter as well as a sound card (the only 2 Type C devices I have)
And I’ve tested with both the Type C port on the back of the motherboard, as well as the one in front of my case, which is connected with cable to the motherboard.
I do admittedly find it a bit odd the usb2-port6 part doesn’t change, even though I use a different port.

4

Did a bit more testing, and oddly enough the problem goes away both if I remove anything from Type C ports AND if move the Type A ports I’m using but keep using the Type C. Guess it must be some sort of clash between the particular ports in question.

But for now it appears to have been solved by simply using different Type A ports.

5

You could also try adding loglevel=3 to your grub boot flags.

The default log level is 4, which will cause notices to display on boot. 3 will still give you critical information but might help prevent these boot delays. Since the type c works after the machine is up, its probably level 4. I ran into a similar problem recently and I believe this is how I fixed it.

6

Thanks for the tip compy386 - I’ll give that a go as well :slight_smile:

7

Cables aren’t so passive a component as they might seem. As far as USB-C is concerned I might even go so far as to say they’re a disaster.

Just because it might work after boot, also doesn’t mean there’s not a real underlying issue (though, I do favor this being in the board, from what was said so far).

monitor dmesg -wH during connection, and check lsusb -t to see where it shows in the usb tree (and what speed).

Could also be interesting to see if another distro’s live USB acts the same way.

2 Likes
8

Which motherboard, and is the bios up to date?

You could try playing with the usb settings in the bios, like xhci handoff, etc. to see if it improves anything?

Are these USB4 ports? Is usb4/pci-e tunneling over usb4 enabled in the bios?

9

In my case, it was likely EMI. I had both your error message and an explicit one mentioning EMI. My issue got fixed after I gently shifted some cables inside the case. I suspect my unshielded SATA cables, because I also had SATA errors at the tiime.

10

@exovert Finally got around to testing a bit more, dmesg -wH shows something interesting: The USB-C port is not the one boot complains about:

apr16 21:01] usb 3-1: USB disconnect, device number 2
[  +3,133389] usb 3-1: new full-speed USB device number 3 using xhci_hcd
[  +0,709728] usb 3-1: New USB device found, idVendor=041e, idProduct=3276, bcdDevice= 0.11
[  +0,000008] usb 3-1: New USB device strings: Mfr=3, Product=1, SerialNumber=2
[  +0,000002] usb 3-1: Product: Sound Blaster Play! 4
[  +0,000002] usb 3-1: Manufacturer: Generic
[  +0,000002] usb 3-1: SerialNumber: WWSB1860113007184K
[  +0,904225] input: Generic Sound Blaster Play! 4 Consumer Control as /devices/pci0000:00/0000:00:08.1/0000:15:00.3/usb3/3-1/3-1:1.3/0003:041E:3276.0007/input/input26
[  +0,056224] input: Generic Sound Blaster Play! 4 as /devices/pci0000:00/0000:00:08.1/0000:15:00.3/usb3/3-1/3-1:1.3/0003:041E:3276.0007/input/input27
[  +0,000109] input: Generic Sound Blaster Play! 4 as /devices/pci0000:00/0000:00:08.1/0000:15:00.3/usb3/3-1/3-1:1.3/0003:041E:3276.0007/input/input28
[  +0,000179] hid-generic 0003:041E:3276.0007: input,hiddev97,hidraw1: USB HID v1.11 Device [Generic Sound Blaster Play! 4] on usb-0000:15:00.3-1/input3

lsusb -t output shows it running at USB 2 speed though, not too sure what’s up with that as I had expected all USB-C were USB 3 speeds.

/:  Bus 003.Port 001: Dev 001, Class=root_hub, Driver=xhci_hcd/2p, 480M
    |__ Port 001: Dev 003, If 0, Class=Audio, Driver=snd-usb-audio, 12M
    |__ Port 001: Dev 003, If 1, Class=Audio, Driver=snd-usb-audio, 12M
    |__ Port 001: Dev 003, If 2, Class=Audio, Driver=snd-usb-audio, 12M
    |__ Port 001: Dev 003, If 3, Class=Human Interface Device, Driver=usbhid, 12M

My conclusion for now is that the USB-C attached device somehow conflicts/interacts with the USB-A ports I used (maybe they’re same root bus or something)
Since moving keyboard + mouse to different USB-A ports I’ve had no further dmesg warnings about bad cabling.

@quilt As for the motherboard, it’s a Gigabyte B650M DS3H running the latest BIOS revision as of writing. All USB settings are stock, as I’m not knowledgeable enough about what impact they have to touch them :smiley:

Finally there are no USB4 ports on the board to the best of my knowledge.

@LiKenun I guess that could be the culprit, though I have little idea of how to solve it if that’s the case. I have only the power cables as well as the USB in front cable attached to my board - no SATA devices at all.