A laptop loses power exactly every five days, and the reason is out of Linux's reach

A host with a healthy battery, healthy thermals and stable mains power should not be able to lose power on its own. This one has, three times since 20260906, and it is still doing it as of 20260920.

There is nothing to read afterwards. No kernel panic, no shutdown sequence, no last log line that means anything, no network link lights on the way down. The machine is simply off and needs the physical power button. Each outage on its own looks like an unrelated glitch.

What turned three glitches into a pattern was reading the boot history as a table rather than as separate events. Listed side by side, every failure had landed at almost exactly five days of uptime, regular enough to predict the next one to the day. That pattern had been sitting in already-collected data the whole time and nobody had put the rows next to each other.

From there it is elimination. Thermal readings, battery health, mains power, lid-suspend configuration, resource exhaustion and one suspicious kernel module were each checked against real measured data, and each came back clean. What is left is the explanation that leaves no evidence by construction. This model’s T2 security co-processor runs its own watchdog. It can cut power to the whole machine without the operating system ever being told.

Two decisions came out of it while the fault is still open. Kernel output is now streamed to another host over the network, so the next event arrives somewhere it can be read. And the real diagnostic, reading the T2’s own panic log, which needs a macOS boot, was deliberately left until the fault recurs, rather than spent on a guess.