Skip to content

Why chroot is not a security boundary

A process is confined with chroot to a directory, so it sees that directory as the root of the file system.

Why is this not sufficient isolation for untrusted code?