I'm investigating repeated WindowServer watchdog kernel panics on a Mac mini M6 running macOS 27.0.1 (build 26A434), and would appreciate technical guidance on identifying the initiating failure.
System configuration
- Mac mini M6, 16 GB unified memory
- macOS 27.0.1 (26A434)
- Redmi G Pro 27U external monitor
- Physical output: 3840 × 2160 @ 160 Hz, RGB 10-bit HDR10
- USB-C-to-DisplayPort 1.4 direct connection, HBR3 ×4, DSC enabled
- Similar system-level crashes have occurred over both HDMI and DisplayPort
The system has experienced repeated freezes and forced restarts over several days.
Kernel panic signature
userspace watchdog timeout:
no successful checkins from WindowServer
in 120 seconds
WindowServer initialization not complete
(post IOKitWaitQuiet)
Some WindowServer recovery diagnostics involve IOMobileFramebuffer and IOMobileGraphicsFamily-DCP.
Latest incident: October 24, 2026 (UTC+08:00)
At 01:33:25.578, the unified log began recording repeated messages:
DCPEXT1:GMailboxHandler.cpp:232
apt firmware: immediate_swap_denied landed
frame_id = 1, reason = 1
At 01:33:35.612, WindowServer reported:
[Watchdog] display 1 has become stuck
after 10.00154s of unreadiness
(2463 failures)
Approximately 14,264 immediate_swap_denied messages were recorded in the retrieved interval ending at 01:34:55, followed by a system restart around 01:35.
The corresponding panic report is:
panic-full-2026-10-24-013618.0002.panic
What I have checked
- The failures are not limited to one video connection type.
- The physical display signal is 4K 160 Hz, not 5K 160 Hz. The 5120 × 2880 value reported by HiDPI utilities represents the internal scaling buffer.
- Crashes occurred before BetterDisplay was first used.
- The latest incident occurred while viewing a remote desktop stream, but available process-count samples did not reveal proliferation of the remote-control processes.
- Memory/resource pressure remains under investigation; there is not enough evidence to attribute the panics to memory exhaustion.
- It is unclear whether the initial WindowServer stall precedes or follows the DCP frame-swap errors.
Questions
- What does
immediate_swap_deniedwithframe_id=1, reason=1indicate in the Apple DCP display pipeline? - Can this error be a downstream symptom of an already-stalled WindowServer, or does it suggest a DCP/display frame-submission failure?
- Is there a known macOS 27 issue involving WindowServer watchdog failures or recovery stalls in IOMobileGraphicsFamily-DCP?
- What additional diagnostics would help distinguish a macOS display firmware/driver problem from external-display compatibility, memory pressure, or defective hardware?
I can provide the panic report, WindowServer diagnostic samples, relevant unified-log extracts, display configuration information, and a sysdiagnose through Feedback Assistant.
Feedback Assistant report: FB25300098
Any guidance from Apple DTS engineers or developers who have encountered a similar DCP/WindowServer failure would be appreciated.