Re: [PATCH v3 2/2] modules:capabilities: add a per-task modules autoload restriction

[Andy Lutomirski <luto@kernel.org>]

On Fri, Apr 21, 2017 at 4:52 PM, Casey Schaufler <casey@schaufler-ca.com> wrote:
> On 4/21/2017 4:28 PM, Andy Lutomirski wrote:
>> On Fri, Apr 21, 2017 at 4:19 PM, Kees Cook <keescook@chromium.org> wrote:
>>> On Wed, Apr 19, 2017 at 7:41 PM, Andy Lutomirski <luto@kernel.org> wrote:
>>>> On Wed, Apr 19, 2017 at 4:43 PM, Kees Cook <keescook@chromium.org> wrote:
>>>>> On Wed, Apr 19, 2017 at 4:15 PM, Andy Lutomirski <luto@kernel.org> wrote:
>>>>>> On Wed, Apr 19, 2017 at 3:20 PM, Djalal Harouni <tixxdz@gmail.com> wrote:
>>>>>>> +/* Sets task's modules_autoload */
>>>>>>> +static inline int task_set_modules_autoload(struct task_struct *task,
>>>>>>> +                                           unsigned long value)
>>>>>>> +{
>>>>>>> +       if (value > MODULES_AUTOLOAD_DISABLED)
>>>>>>> +               return -EINVAL;
>>>>>>> +       else if (task->modules_autoload > value)
>>>>>>> +               return -EPERM;
>>>>>>> +       else if (task->modules_autoload < value)
>>>>>>> +               task->modules_autoload = value;
>>>>>>> +
>>>>>>> +       return 0;
>>>>>>> +}
>>>>>> This needs to be more locked down.  Otherwise someone could set this
>>>>>> and then run a setuid program.  Admittedly, it would be quite odd if
>>>>>> this particular thing causes a problem, but the issue exists
>>>>>> nonetheless.
>>>>> Eeeh, I don't agree this needs to be changed. APIs provided by modules
>>>>> are different than the existing privilege-manipulation syscalls this
>>>>> concern stems from. Applications are already forced to deal with
>>>>> things being missing like this in the face of it simply not being
>>>>> built into the kernel.
>>>>>
>>>>> Having to hide this behind nnp seems like it'd reduce its utility...
>>>>>
>>>> I think that adding an inherited boolean to task_struct that can be
>>>> set by unprivileged tasks and passed to privileged tasks is a terrible
>>>> precedent.  Ideally someone would try to find all the existing things
>>>> like this and kill them off.
>>> (Tristate, not boolean, but yeah.)
>>>
>>> I see two others besides seccomp and nnp:
>>>
>>> PR_MCE_KILL
>> Well, that's interesting.  That should presumably be reset on setuid
>> exec or something.
>>
>>> PR_SET_THP_DISABLE
>> Um.  At least that's just a performance issue.
>>
>>> I really don't think this needs nnp protection.
>>>
>>>> I agree that I don't see how one would exploit this particular
>>>> feature, but I still think I dislike the approach.  This is a slippery
>>>> slope to adding a boolean for perf_event_open(), unshare(), etc, and
>>>> we should solve these for real rather than half-arsing them IMO.
>>> I disagree (obviously); this would be protecting the entire module
>>> autoload attack surface. That's hardly a specific control, and it's a
>>> demonstrably needed flag.
>>>
>> The list is just going to get longer.  We should probably have controls for:
>>
>>  - Use of perf.  Unclear how fine grained they should be.
>>
>>  - Creation of new user namespaces.  Possibly also use of things like
>> iptables without global privilege.
>>
>>  - Ability to look up tasks owned by different uids (or maybe other
>> tasks *at all*) by pid/tid.  Conceptually, this is easy.  The API is
>> the only hard part, I think.
>>
>>  - Ability to bind ports, maybe?
>
> One of my longer term (i.e. after stacking) projects
> is to create sensible access control on ports. Why shouldn't
> they have owners and mode bits (or ACLs, if you prefer)
> or real names. I kind of think we should be able to eliminate
> the need for dbus without resorting to kdbus.

My implicit_rights concept gives any type of access control you can
use on inodes because they *are* inodes.  So you get ACLs, etc.

Brief summary for those who didn't read my old email: We add a new
kind of filesystem object called a "right".  It's a special kind of
socket inode that can't be bound or connected but is instead created
by a new syscall.  It has a name, so "port:1234" might be a name of a
right.

To use an implicit right, you do whatever syscall you would do
normally.  The kernel looks for a right object at
/dev/implicit_rights/<name>.  If that object exists, is a right of the
correct type (i.e. the right's name matches <name>) and you have
execute access, you win.  Otherwise you lose.

To avoid breaking existing distros, for things like modules_autoload,
you would set a sysctl
/proc/sys/kernel/required_implicit_rights/modules_autoload=1.  With
that set, to autoload a module without CAP_SYS_MODULE, you need the
/dev/implicit_rights/modules_autoload.

>
> So I don't like the idea of treating that as a special case.
> I'd rather see ports controlled properly. (Of course, the
> SELinux crowd will point out they have this handled, but I
> remain unconvinced of the overall solution)

Agreed.  But I think we should address all of these things together.