Comment # 13 on bug 1028966 from Richard Palethorpe

The following errors seen when plugging the port replicator in, firstly

[  105.425932] xhci_hcd 0000:3e:00.0: xHCI Host Controller
[  105.425934] xhci_hcd 0000:3e:00.0: new USB bus registered, assigned bus
number 4
[  105.425988] usb usb4: New USB device found, idVendor=1d6b, idProduct=0003
[  105.425989] usb usb4: New USB device strings: Mfr=3, Product=2,
SerialNumber=1
[  105.425990] usb usb4: Product: xHCI Host Controller
[  105.425991] usb usb4: Manufacturer: Linux 4.10.5-1-default xhci-hcd
[  105.425992] usb usb4: SerialNumber: 0000:3e:00.0
[  105.426158] hub 4-0:1.0: USB hub found
[  105.426174] hub 4-0:1.0: 2 ports detected
[  105.783147] usb 4-1: Device not responding to setup address.
[  110.448469] xhci_hcd 0000:3e:00.0: remove, state 1
[  110.448476] usb usb4: USB disconnect, device number 1
[  111.081323] xhci_hcd 0000:3e:00.0: Stopped the command ring failed, maybe
the host is dead
[  111.081335] xhci_hcd 0000:3e:00.0: Host halt failed, -19
[  111.081336] xhci_hcd 0000:3e:00.0: Abort command ring failed
[  111.081340] xhci_hcd 0000:3e:00.0: Timeout while waiting for setup device
command
[  111.081343] xhci_hcd 0000:3e:00.0: HC died; cleaning up
[  111.288351] usb 4-1: device not accepting address 2, error -62
[  111.288370] usb usb4-port1: couldn't allocate usb_device
[  111.288562] xhci_hcd 0000:3e:00.0: Host halt failed, -19
[  111.288571] xhci_hcd 0000:3e:00.0: Host not accessible, reset failed.
[  111.288573] xhci_hcd 0000:3e:00.0: USB bus 4 deregistered
[  111.288579] xhci_hcd 0000:3e:00.0: remove, state 4

This says to me that the host controller is taking too long to respond. I would
guess this is due to a stall rather than it being put in the wrong state. So
maybe increasing the timeout might fix it? I am guessing from the number of
ports and the USB number it gets that this is the controller the ethernet
adapter is connected to.

To see if this error always happens I reconnected the port replicator again

[  +0.000065] usb usb4: New USB device found, idVendor=1d6b, idProduct=0003
[  +0.000004] usb usb4: New USB device strings: Mfr=3, Product=2,
SerialNumber=1
[  +0.000003] usb usb4: Product: xHCI Host Controller
[  +0.000009] usb usb4: Manufacturer: Linux 4.10.5-1-default xhci-hcd
[  +0.000003] usb usb4: SerialNumber: 0000:3e:00.0
[  +0.000253] hub 4-0:1.0: USB hub found
[  +0.000022] hub 4-0:1.0: 2 ports detected
[  +0.328565] usb 4-1: new SuperSpeed USB device number 2 using xhci_hcd
[  +0.040609] usb 4-1: device descriptor read/8, error -71
[  +0.107134] usb 4-1: new SuperSpeed USB device number 2 using xhci_hcd
[  +4.536137] xhci_hcd 0000:3e:00.0: remove, state 1
[  +0.000008] usb usb4: USB disconnect, device number 1
[  +0.727842] usb 4-1: device descriptor read/8, error -110
[  +5.119974] xhci_hcd 0000:3e:00.0: Error while assigning device slot ID
[  +0.000015] xhci_hcd 0000:3e:00.0: Max number of devices this xHCI host
supports is 64.
[  +0.000008] usb usb4-port1: couldn't allocate usb_device
[  +0.001545] xhci_hcd 0000:3e:00.0: USB bus 4 deregistered
[  +0.000023] xhci_hcd 0000:3e:00.0: remove, state 4

Different error, but looking at the code and guessing a bit I think it could
also be caused by a stall. It seems that the system retries and the device
begins working. There don't seem to be any similar looking errors in rbrown's
logs.

Possibly we have different firmware installed on laptop and port replicator
because rbrown's is at least half a year newer. There are firmware updates
available, but currently there is only an update utility which runs on Windows.
Allegedly DELL are working on a Linux version.

The port is Thunderbolt 3 which uses USB-C.