Install medium: Sandisk USB (Baremetal no virtualization)
Partition format: ext4 (auto-formatted) + LUKS
BIOS: Coreboot
Hi everyone,
I’m a bit puzzled by the way emerg‑shutdown operates in both persistence and live modes when I’m using a USB thumb drive. I want to report that it does work in both modes for my installation. The question at hand is when I unplug the drive, the system simply powers off without displaying any on screen notification.
Is this the intended behavior? If not, could we consider adding a brief display message before the shutdown occurs?
Adding a display message (or any pause of any kind) would be a security hazard for the threat model emerg-shutdown is meant to defend against. Imagine you’re using Kicksecure from an external USB drive, and you’re using it to do something a physically proximate adversary doesn’t like. As you’re doing this work, the adversary suddenly breaks open the door to the room you’re in and starts walking toward you quickly. At this point one of your primary concerns is turning off the computer as quickly as possible so that all traces of what you were doing are erased. Every millisecond counts in this scenario, so emerg-shutdown is designed to power off the computer with all haste. It doesn’t make any attempt to save unsaved data, sync filesystems, warn the user of what’s going to happen, etc. it just* tells the kernel to cut power to the system as quickly as it can.
*well, almost; to work around kernel bugs that can result in a kernel panic during shutdown, emerg-shutdown switches to an unused TTY first. This is necessary to keep the system from hanging or rebooting instead of shutting down if you’re using Intel graphics, at least on some systems.
More about emerg-shutdown’s threat model, implementation, and usage is here:
good work!