Re: [PATCH v2 1/3] capabilities: Introduce CAP_CHECKPOINT_RESTORE

[Christian Brauner]

On Wed, Jun 10, 2020 at 08:41:29AM -0700, Casey Schaufler wrote:
> 
> On 6/10/2020 12:59 AM, Andrei Vagin wrote:
> > On Tue, Jun 09, 2020 at 06:14:27PM +0200, Christian Brauner wrote:
> >> On Tue, Jun 09, 2020 at 09:06:27AM -0700, Andrei Vagin wrote:
> >>> On Tue, Jun 09, 2020 at 09:44:22AM +0200, Christian Brauner wrote:
> >>>> On Mon, Jun 08, 2020 at 08:42:21PM -0700, Andrei Vagin wrote:
> > ...
> >>>>> PTRACE_O_SUSPEND_SECCOMP is needed for C/R and it is protected by
> >>>>> CAP_SYS_ADMIN too.
> >>>> This is currently capable(CAP_SYS_ADMIN) (init_ns capable) why is it
> >>>> safe to allow unprivileged users to suspend security policies? That
> >>>> sounds like a bad idea.
> > ...
> >>> I don't suggest to remove or
> >>> downgrade this capability check. The patch allows all c/r related
> >>> operations if the current has CAP_CHECKPOINT_RESTORE.
> >>>
> >>> So in this case the check:
> >>>      if (!capable(CAP_SYS_ADMIN))
> >>>              return -EPERM;
> >>>
> >>> will be converted in:
> >>>      if (!capable(CAP_SYS_ADMIN) && !capable(CAP_CHECKPOINT_RESTORE))
> >>>              return -EPERM;
> >> Yeah, I got that but what's the goal here? Isn't it that you want to
> >> make it safe to install the criu binary with the CAP_CHECKPOINT_RESTORE
> >> fscap set so that unprivileged users can restore their own processes
> >> without creating a new user namespace or am I missing something? The
> >> use-cases in the cover-letter make it sound like that's what this is
> >> leading up to:
> >>>>>> * Checkpoint/Restore in an HPC environment in combination with a resource
> >>>>>>   manager distributing jobs where users are always running as non-root.
> >>>>>>   There is a desire to provide a way to checkpoint and restore long running
> >>>>>>   jobs.
> >>>>>> * Container migration as non-root
> >>>>>> * We have been in contact with JVM developers who are integrating
> >>>>>>   CRIU into a Java VM to decrease the startup time. These checkpoint/restore
> >>>>>>   applications are not meant to be running with CAP_SYS_ADMIN.
> >> But maybe I'm just misunderstanding crucial bits (likely (TM)).
> > I think you understand this right. The goal is to make it possible to
> > use C/R functionality for unprivileged processes.
> 
> Y'all keep saying "unprivileged processes" when you mean
> "processes with less than root privilege". A process with
> CAP_CHECKPOINT_RESTORE *is* a privileged process. It would

That was me being imprecise. What I mean is "unprivileged user"
not "unprivileged process". It makes me a little uneasy that an
unprivileged _user_ can call the criu binary with the
CAP_CHECKPOINT_RESTORE fscap set and suspend seccomp of a process (Which
is what my original question here was about). Maybe this is paranoia but
shouldn't suspending _security_ mechanisms be kept either under
CAP_SYS_ADMIN or CAP_MAC_ADMIN?

Christian