PMU power key events on the mainline image (task 1.3), 2026-10-14

Board-read over SSH on the landscape image (k230-mainline-drm-shell-landscape, mainline 7.3.0-rc5 with the ported k230-pmu-pwrkey driver). Limit: the task names a serial tools/console.py /dev/ttyACM0 read. On this bench /dev/ttyACM0 is now a different device, not the K230, so the same reads were taken over SSH instead.

/proc/bus/input/devices:

N: Name="K230 PMU Power Key"
P: Phys=k230-pmu-pwrkey/input0
S: Sysfs=/devices/platform/soc/91000000.pmu-pwrkey/input/input0
H: Handlers=kbd event0

shell-power-key.service was stopped for the window (it grabs the device), and evemu-record /dev/input/event0 ran for 180 s while the operator tapped the key twice:

E: 0.000001 0001 0074 0001	# EV_KEY / KEY_POWER            1
E: 0.000001 0000 0000 0000	# ------------ SYN_REPORT (0) ---------- +0ms
E: 0.195543 0001 0074 0000	# EV_KEY / KEY_POWER            0
E: 0.195543 0000 0000 0000	# ------------ SYN_REPORT (0) ---------- +195ms
E: 0.973287 0001 0074 0001	# EV_KEY / KEY_POWER            1
E: 0.973287 0000 0000 0000	# ------------ SYN_REPORT (0) ---------- +778ms
E: 1.133189 0001 0074 0000	# EV_KEY / KEY_POWER            0
E: 1.133189 0000 0000 0000	# ------------ SYN_REPORT (0) ---------- +160ms

Window 2026-10-14T16:24:49-05:00 (uptime 7739.26) to 16:27:49 (uptime 7919.31), boot ID ccd3f704-478b-4eff-b187-1c7bde3ae50f throughout: each press has its release, and the system kept running. The service was started again afterwards.