Re: [PATCH v4 1/1] fs: Allow no_new_privs tasks to call chroot(2)

From: Mickaël Salaün
Date: Tue Mar 16 2021 - 16:07:12 EST

On 16/03/2021 20:31, Jann Horn wrote:
> On Tue, Mar 16, 2021 at 8:26 PM Mickaël Salaün <mic@xxxxxxxxxxx> wrote:
>> On 16/03/2021 20:04, Jann Horn wrote:
>>> On Tue, Mar 16, 2021 at 6:02 PM Mickaël Salaün <mic@xxxxxxxxxxx> wrote:
>>>> One could argue that chroot(2) is useless without a properly populated
>>>> root hierarchy (i.e. without /dev and /proc). However, there are
>>>> multiple use cases that don't require the chrooting process to create
>>>> file hierarchies with special files nor mount points, e.g.:
>>>> * A process sandboxing itself, once all its libraries are loaded, may
>>>> not need files other than regular files, or even no file at all.
>>>> * Some pre-populated root hierarchies could be used to chroot into,
>>>> provided for instance by development environments or tailored
>>>> distributions.
>>>> * Processes executed in a chroot may not require access to these special
>>>> files (e.g. with minimal runtimes, or by emulating some special files
>>>> with a LD_PRELOADed library or seccomp).
>>>> Unprivileged chroot is especially interesting for userspace developers
>>>> wishing to harden their applications. For instance, chroot(2) and Yama
>>>> enable to build a capability-based security (i.e. remove filesystem
>>>> ambient accesses) by calling chroot/chdir with an empty directory and
>>>> accessing data through dedicated file descriptors obtained with
>>> I don't entirely understand. Are you writing this with the assumption
>>> that a future change will make it possible to set these RESOLVE flags
>>> process-wide, or something like that?
>> No, this scenario is for applications willing to sandbox themselves and
>> only use the FDs to access legitimate data.
> But if you're chrooted to /proc/self/fdinfo and have an fd to some
> directory - let's say /home/user/Downloads - there is nothing that
> ensures that you only use that fd with RESOLVE_BENEATH, right? If the
> application is compromised, it can do something like openat(fd,
> "../.bashrc", O_RDWR), right? Or am I missing something?

You're totally right, I was mistaken, this simple use case doesn't work
without a broker. Perhaps when seccomp will be able to check referenced
structs, or with a new FD limitation…

>>> As long as that doesn't exist, I think that to make this safe, you'd
>>> have to do something like the following - let a child process set up a
>>> new mount namespace for you, and then chroot() into that namespace's
>>> root:
>>> struct shared_data {
>>> int root_fd;
>>> };
>>> int helper_fn(void *args) {
>>> struct shared_data *shared = args;
>>> mount("none", "/tmp", "tmpfs", MS_NOSUID|MS_NODEV, "");
>>> mkdir("/tmp/old_root", 0700);
>>> pivot_root("/tmp", "/tmp/old_root");
>>> umount("/tmp/old_root", "");
>>> shared->root_fd = open("/", O_PATH);
>>> }
>>> void setup_chroot() {
>>> struct shared_data shared = {};
>>> prctl(PR_SET_NO_NEW_PRIVS, 1, 0, 0, 0);
>>> clone(helper_fn, my_stack,
>>> NULL);
>>> fchdir(shared.root_fd);
>>> chroot(".");
>>> }
>> What about this?
>> chdir("/proc/self/fdinfo");
>> chroot(".");
>> close(all unnecessary FDs);
> That breaks down if you can e.g. get a unix domain socket connected to
> a process in a different chroot, right? Isn't that a bit too fragile?

This relies on other (trusted) components, and yes it is fragile if the
process communicates with a service able send FDs.