[fairydust / GNOME] M1 MacBook Pro: USB-C display black after login until dock reconnect #646

Description

Disclosure: prepared with AI assistance from my firsthand observations and local journal records. The root cause has not been isolated.

The external display works during boot and in GDM, but turns black immediately after I log in. Unplugging the dock, waiting a second and reconnecting restores it. The same dock/display combination works in macOS. Once recovered, the display can work with the lid closed.

Environment:

• Apple MacBook Pro 13-inch M1 (2020), not M1 Pro/Max
• Fedora Asahi Remix 44, aarch64, GNOME Wayland
• COPR kernel 7.1.13-911.fairydust.fc44.aarch64+16k; matching /boot/dtb target
• GDM 50.3-1.fc44; Mutter/GNOME Shell 50.5-1.fc44; Mesa DRI 26.2.4-1.fc44
• AV Access iDock C20 over USB-C; Samsung LS32DG802SUXEN / Odyssey G80SD on DP-1
• External 3840x2160 at 59.997 Hz, scale 1.5; internal 2560x1600/60, scale approximately 1.667, primary
• Desktop extensions enabled; an extension-free session and an alternate dock/direct cable have not been tested

Journal timeline, 11 October 2026, CEST (selected excerpts; hostname and username omitted):

15:18:03 External DCP powers on, selects 3840x2160/60
15:18:08 GDM GNOME Shell starts as a Wayland display server
15:18:13 apple-dcp 271c00000.dcp: cb_hotplug() connected:0, valid_mode:1
15:18:15 apple-dcp 271c00000.dcp: dcp_hotplug() connected:1 valid_mode:0 nr_modes:26
15:18:18 User session opens
15:18:18 swap_submit_dcp: swallowed swap ID 8 as fControllerPowerState is 0 for External Display
15:18:18 swap_submit_dcp: swallowed swap ID 8 as timinsg are not enabled for External Display
15:18:19 User GNOME Shell starts
15:18:29 phy-apple-atc 503000000.phy: Pipehandler lock not acked.
15:18:29 phy-apple-atc 503000000.phy: Failed to lock pipehandler
15:18:33 apple-dcp 271c00000.dcp: dcp_poweroff() done
15:18:37 apple-dcp 271c00000.dcp: dcp_poweron() starting
15:18:37 External DCP selects the same 3840x2160/60 mode
15:19:15 Lid closes, after recovery

The physical unplug/replug timestamps were not separately recorded; the 15:18:29–37 interval corresponds to recovery. Between the 15:18:15 reconnect and that interval, no fresh external dcp_poweron() or set_digital_out_mode() line was found. Both appear after replug. No flip_done or commit-wait timeout was found in the inspected 15:18:00–15:18:45 window. The getClockFrequency(0)=0 warning also appears after successful recovery, so it does not identify the cause by itself.

The initial HPD drop precedes user-session startup, so login is not proven to cause the first disconnect. Driver recovery, dock/monitor behavior and Mutter handling remain possible contributors. This is a boot/login failure without suspend; USB wake is a separate question.

Related reports: #430, #571, and the M1 USB-C follow-up at #610 (comment) . Different hardware/compositors; not claimed to be duplicates. No speculative patch or startup delay has been installed. Is there a known fix or a specific recovery issue matching this sequence?

Repro Steps

1. Connect the AV Access iDock C20 and external Samsung display before booting.
2. Boot Fedora Asahi Remix with the fairydust kernel.
3. Observe boot output and the GDM login screen on the external display; both screens work.
4. Enter credentials and log in to GNOME Wayland.
5. The external screen goes black/loses output.
6. Unplug the dock, wait a second, and reconnect it. External output returns.

Expected Behaviour

The external display should continue working when the GNOME desktop starts, without requiring a physical dock reconnect.

Upload Files

No response

Please agree to the following

• I have reproduced this on Fedora Asahi Remix
• I have isolated this issue to software vendored by Asahi Linux
• I am not acting as a meat proxy for an LLM