Arch Linux

Please read this before reporting a bug:
https://wiki.archlinux.org/title/Bug_reporting_guidelines

Do NOT report bugs when a package is just outdated, or it is in the AUR. Use the 'flag out of date' link on the package page, or the Mailing List.

REPEAT: Do NOT report bugs for outdated packages!
Tasklist

FS#66969 - [kernel] Thunderbolt dock failure and Intel display driver corrruption

Attached to Project: Arch Linux
Opened by Donald Sigent (DonSig) - Wednesday, 10 June 2020, 16:53 GMT
Last edited by freswa (frederik) - Friday, 12 June 2020, 04:31 GMT
Task Type Bug Report
Category Packages: Extra
Status Closed
Assigned To No-one
Architecture All
Severity Low
Priority Normal
Reported Version
Due in Version Undecided
Due Date Undecided
Percent Complete 100%
Votes 0
Private No

Details

Description:

Ever since I first installed Arch on my device (Dell XPS 7390 2-in-1 with Intel Ice Lake i7-1065G7 CPU) and the corresponding Thunderbolt dock (Dell WD19TB) I observed random "disconnect events" after anything between a few minutes and a few hours in which all attached Thunderbolt devices will briefly fail:

* External hard drives disconnect, with buffer I/O errors, and LUKS drives don't recover at all
* other USB devices briefly go out
* All external monitors go dark, and might or might not recover as the Intel display driver enters a corrupted state in which it sees the monitors, but does not push out a signal to them - turning them off and on again fixes it for a few seconds before they go out

Only a cold reboot fixes the issue for a time.

Linux versions: ever since 5.4 (first kernel) all the way to the latest 5.7 in testing, including all lte and mainline kernels.

I filed a bugreport upstream to Intel drm since I assumed a diplay driver issue at first, but the external drive disconnects suggest that the entire dock is affected, and the Intel display driver can't recover from the failure. So it is possibly an issue with the Thunderbolt driver?

So I am filing this here in the hopes that someone knows a solution, or can please kindly direct me to the correct place to file a report.

I attached an extensive dmesg, the disconnect happens at 29337.443777 (xHCI host controller not responding, assume dead), other errors that sometimes appear are CPU FIFO underrun errors.

Steps to reproduce:
Use the device connected to the dock, the issue will appear within a few hours of use.

This task depends upon

Closed by  freswa (frederik)
Friday, 12 June 2020, 04:31 GMT
Reason for closing:  Not a bug
Additional comments about closing:  This bugtracker is for Arch packaging bugs. For support please ask on the ML, the forums or in the IRC. Thank you

Loading...