Bug 216950
| Summary: | DisplayPort USB-C hub issue | ||
|---|---|---|---|
| Product: | Drivers | Reporter: | Marcin (martii) |
| Component: | PCI | Assignee: | drivers_pci (drivers_pci) |
| Status: | RESOLVED CODE_FIX | ||
| Severity: | normal | CC: | alexdeucher, bjorn, kw, mario.limonciello |
| Priority: | P1 | ||
| Hardware: | AMD | ||
| OS: | Linux | ||
| Kernel Version: | 6.2.0-rc3 | Subsystem: | |
| Regression: | Yes | Bisected commit-id: | |
| Attachments: |
dmidecode + kern.log
lshw + lspci |
||
Marking as a regression since 5.16.15-76051615-generic works and 6.2.0-rc3 does not. If you know any newer working kernels (or older broken ones), we can tighten the bounds. I see what looks like lots of USB hot-adds and hot-removes, but I don't see any PCI connection there. I think I would raise this with the graphics folks who might have ideas of how to debug this. If this is a DP MST hub, this sounds like a duplicate of https://gitlab.freedesktop.org/drm/amd/-/issues/2171 kern.log from initial attachment contains this, but I don't know whether this means a DP MST hub: [43773.390727] [drm] DP Alt mode state on HPD: 1 [43773.483585] [drm] DM_MST: starting TM on aconnector: 00000000b17636fe [id: 94] [43773.735242] [drm] Downstream port present 1, type 0 [43773.858532] [drm] crtc[1] needs mode_changed [44592.056428] [drm] DM_MST: stopping TM on aconnector: 00000000b17636fe [id: 94] Here is the docking station: https://www.hp.com/us-en/shop/pdp/hp-usb-c-dock-g5-p-26d32aa-abl-1 Laptop was not put to sleep - all I did was disconnect/connect USB-C cable provided with docking station. Will lspci or lsusb be of any use with docking station connected? Looks like a duplicate. Please try the patches here: https://gitlab.freedesktop.org/drm/amd/-/issues/2171#note_1720234 Created attachment 303626 [details]
lshw + lspci
The patches to fix the core of the MST issues are landed in 6.2-rc as well as 6.1.y. There might be other cases with MST problems still, so if you reproduce with the latest 6.2 RC or latest 6.1.y please open a separate issue, preferably on the AMD Gitlab bug tracker. |
Created attachment 303624 [details] dmidecode + kern.log I have HP 845 EliteBook G7 (Ryzen) and before 6.2.0-rc3 installation there was no issue with my laptop and HP docking station via USB-C. This kernel display behaves weirdly. LCD monitor (connected to docking station via DisplayPort) wakes up only once after reboot. Second docking connection ends up with blank external screen and only restart helps. Display manager detects extra screen and even newly launched apps land there. I didn't have such issues with original distro kernel 5.16.15-76051615-generic