If the user does not have root access (or any way to gain it, such as exploiting an insecure setuid program), escaping a chroot jail should be impossible.
With root access, escaping a chroot jail is trivial. In fact, the `chroot(2)` manpage even gives instructions:
> This call does not change the current working directory, so that after the call '.' can be outside the tree rooted at '/'. In particular, the superuser can escape from a "chroot jail" by doing:
>
>
> mkdir foo; chroot foo; cd ..
>