Re: [PATCH v5 1/1] fs: Allow no_new_privs tasks to call chroot(2)

[Casey Schaufler]

On 3/30/2021 12:28 PM, Mickaël Salaün wrote:
> On 30/03/2021 20:40, Casey Schaufler wrote:
>> On 3/30/2021 11:11 AM, Mickaël Salaün wrote:
>>> On 30/03/2021 19:19, Casey Schaufler wrote:
>>>> On 3/30/2021 10:01 AM, Mickaël Salaün wrote:
>>>>> Hi,
>>>>>
>>>>> Is there new comments on this patch? Could we move forward?
>>>> I don't see that new comments are necessary when I don't see
>>>> that you've provided compelling counters to some of the old ones.
>>> Which ones? I don't buy your argument about the beauty of CAP_SYS_CHROOT.
>> CAP_SYS_CHROOT, namespaces. Bind mounts. The restrictions on
>> "unprivileged" chroot being sufficiently onerous to make it
>> unlikely to be usable.
> There is multiple use cases for these features.

All of which can be done using existing mechanism, best I can tell.


>>>> It's possible to use minimal privilege with CAP_SYS_CHROOT.
>>> CAP_SYS_CHROOT can lead to privilege escalation.
>> Not when used in conjunction with the same set of
>> restrictions you're requiring for "unprivileged" chroot. 
> I'm talking about security with the principle of least privilege:

<RANT>
That is what CAP_SYS_CHROOT is all about. Identify the thing that
you want to do that violates policy. Provide an explicit override
for that policy. Control the use of that override in an orderly way.
Document how it works.

That is the whole intent of the capability mechanism. Don't introduce
baroque conditional behaviors based on environmental nuances.
</RANT>

>  when
> we consider that a process may be(come) malicious but should still be
> able to drop (more) accesses, e.g. with prctl(set_no_new_privs) *then*
> chroot()

How is that better than chroot() *then* drop CAP_SYS_CHROOT?


>>>> It looks like namespaces provide alternatives for all your
>>>> use cases.
>>> I explained in the commit message why it is not the case. In a nutshell,
>>> namespaces bring complexity which may not be required.
>> So? I can use a Swiss Army Knife to cut a string even though it
>> has a corkscrew.
> Complexity leads to (security) issues.

Yes. Explict use of CAP_SYS_CHROOT is simple.
Implicit use of NO_NEW_PRIVS is difficult to verify and track.

>  In secure systems, we want to
> reduce the attack surfaces. There is some pointers here:
> https://lwn.net/Articles/673597/

You've identified a clever hack to justify expanding when
chroot() could be done "safely" without using privilege.
Why not learn how to use the existing mechanism properly?
And teach the next set of people how to do the same?
I am under no delusion that we can tweak here and fiddle
there and make security all rainbows and unicorns. Mature
mechanisms that are general are safer than tangled heaps of
special cases that make individual projects easier.

>>>  When designing a
>>> secure system, we want to avoid giving access to such complexity to
>>> untrusted processes (i.e. more complexity leads to more bugs).
>> If you're *really* designing a secure system you can design it to
>> use existing mechanisms, like CAP_SYS_CHROOT!
> Not always. For instance, in the case of a web browser, we don't want to
> give CAP_SYS_CHROOT to every users just because their browser could
> (legitimately) use it as a security sandbox mechanism.

We already have bunches of mechanisms, including userids, three flavors
of Mandatory Access Control, namespaces, seccomp and virtualization for
sandboxing. Do we really need to change a behavior that's been around for
over 40 years?

>  The same
> principle can be applied to a lot of use cases, e.g. network services,
> file parsers, etc.
>
>>>  An
>>> unprivileged chroot would enable to give just the minimum feature to
>>> drop some accesses. Of course it is not enough on its own, but it can 
be
>>> combined with existing (and future) security features.
>> Like NO_NEW_PRIVS, namespaces and capabilities!
>> You don't need anything new!
> If a process is compromised before chrooting itself and dropping
> CAP_SYS_CHROOT, then there is a bigger security issue than without
> CAP_SYS_CHROOT.

If a process is compromised before setting NO_NEW_PRIVS you
could say the same. 


>>>> The constraints required to make this work are quite
>>>> limiting. Where is the real value add?
>>> As explain in the commit message, it is useful when hardening
>>> applications (e.g. network services, browsers, parsers, etc.). We don't
>>> want an untrusted (or compromised) application to have CAP_SYS_CHROOT
>>> nor (complex) namespace access.
>> If you can ensure that an unprivileged application is
>> always run with NO_NEW_PRIVS you could also ensure that
>> it runs with only CAP_SYS_CHROOT or in an appropriate
>> namespace. I believe that it would be easier for your
>> particular use case. I don't believe that is sufficient.
> You can't always have this assertion, e.g. because a user may require to
> run (legitimate) SETUID binaries…
>
> For everyone following a defense in depth approach (i.e. multiple layers
> of security), an unprivileged chroot is valuable.

If you need to run legitimate SETUID (or file capability enabled) binaries
you can't use NO_NEW_PRIVS. You can use CAP_SYS_CHROOT, because capabilities
where designed to work with the UID mechanisms.

In any case, if you can get other people to endorse your change I'm not
all that opposed to it. I think it's gratuitous. It irks me that you're
unwilling to use the facilities that are available, and instead want to
complicate the security mechanisms and policy further. But, that hasn't
seemed to stop anyone before.
 

>>>>> Regards,
>>>>>  Mickaël
>>>>>
>>>>>
>>>>> On 16/03/2021 21:36, Mickaël Salaün wrote:
>>>>>> From: Mickaël Salaün <mic@xxxxxxxxxxxxxxxxxxx>
>>>>>>
>>>>>> Being able to easily change root directories enables to ease some
>>>>>> development workflow and can be used as a tool to strengthen
>>>>>> unprivileged security sandboxes.  chroot(2) is not an access-control
>>>>>> mechanism per se, but it can be used to limit the absolute view of 
the
>>>>>> filesystem, and then limit ways to access data and kernel interfaces
>>>>>> (e.g. /proc, /sys, /dev, etc.).
>>>>>>
>>>>>> Users may not wish to expose namespace complexity to potentially
>>>>>> malicious processes, or limit their use because of limited resources.
>>>>>> The chroot feature is much more simple (and limited) than the mount
>>>>>> namespace, but can still be useful.  As for containers, users of
>>>>>> chroot(2) should take care of file descriptors or data accessible by
>>>>>> other means (e.g. current working directory, leaked FDs, passed FDs,
>>>>>> devices, mount points, etc.).  There is a lot of literature that discuss
>>>>>> the limitations of chroot, and users of this feature should be aware 
>> of
>>>>>> the multiple ways to bypass it.  Using chroot(2) for security purposes
>>>>>> can make sense if it is combined with other features (e.g. dedicated
>>>>>> user, seccomp, LSM access-controls, etc.).
>>>>>>
>>>>>> 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).
>>>>>>
>>>>>> Allowing a task to change its own root directory is not a threat to the
>>>>>> system if we can prevent confused deputy attacks, which could be
>>>>>> performed through execution of SUID-like binaries.  This can be
>>>>>> prevented if the calling task sets PR_SET_NO_NEW_PRIVS on itself with
>>>>>> prctl(2).  To only affect this task, its filesystem information must 
>> not
>>>>>> be shared with other tasks, which can be achieved by not passing
>>>>>> CLONE_FS to clone(2).  A similar no_new_privs check is already used by
>>>>>> seccomp to avoid the same kind of security issues.  Furthermore, because
>>>>>> of its security use and to avoid giving a new way for attackers to 
get
>>>>>> out of a chroot (e.g. using /proc/<pid>/root, or chroot/chdir), an
>>>>>> unprivileged chroot is only allowed if the calling process is not
>>>>>> already chrooted.  This limitation is the same as for creating user
>>>>>> namespaces.
>>>>>>
>>>>>> This change may not impact systems relying on other permission models
>>>>>> than POSIX capabilities (e.g. Tomoyo).  Being able to use chroot(2) on
>>>>>> such systems may require to update their security policies.
>>>>>>
>>>>>> Only the chroot system call is relaxed with this no_new_privs check; 
>> the
>>>>>> init_chroot() helper doesn't require such change.
>>>>>>
>>>>>> Allowing unprivileged users to use chroot(2) is one of the initial
>>>>>> objectives of no_new_privs:
>>>>>> https://www.kernel.org/doc/html/latest/userspace-api/no_new_privs.html
>>>>>> This patch is a follow-up of a previous one sent by Andy Lutomirski:
>>>>>> https://lore.kernel.org/lkml/0e2f0f54e19bff53a3739ecfddb4ffa9a6dbde4d.1327858005.git.luto@xxxxxxxxxxxxxx/
>>>>>>
>>>>>> Cc: Al Viro <viro@xxxxxxxxxxxxxxxxxx>
>>>>>> Cc: Andy Lutomirski <luto@xxxxxxxxxxxxxx>
>>>>>> Cc: Christian Brauner <christian.brauner@xxxxxxxxxx>
>>>>>> Cc: Christoph Hellwig <hch@xxxxxx>
>>>>>> Cc: David Howells <dhowells@xxxxxxxxxx>
>>>>>> Cc: Dominik Brodowski <linux@xxxxxxxxxxxxxxxxxxxx>
>>>>>> Cc: Eric W. Biederman <ebiederm@xxxxxxxxxxxx>
>>>>>> Cc: James Morris <jmorris@xxxxxxxxx>
>>>>>> Cc: Jann Horn <jannh@xxxxxxxxxx>
>>>>>> Cc: John Johansen <john.johansen@xxxxxxxxxxxxx>
>>>>>> Cc: Kentaro Takeda <takedakn@xxxxxxxxxxxxx>
>>>>>> Cc: Serge Hallyn <serge@xxxxxxxxxx>
>>>>>> Cc: Tetsuo Handa <penguin-kernel@xxxxxxxxxxxxxxxxxxx>
>>>>>> Signed-off-by: Mickaël Salaün <mic@xxxxxxxxxxxxxxxxxxx>
>>>>>> Reviewed-by: Kees Cook <keescook@xxxxxxxxxxxx>
>>>>>> Link: https://lore.kernel.org/r/20210316203633.424794-2-mic@xxxxxxxxxxx
>>>>>> ---
>>>>>>
>>>>>> Changes since v4:
>>>>>> * Use READ_ONCE(current->fs->users) (found by Jann Horn).
>>>>>> * Remove ambiguous example in commit description.
>>>>>> * Add Reviewed-by Kees Cook.
>>>>>>
>>>>>> Changes since v3:
>>>>>> * Move the new permission checks to a dedicated helper
>>>>>>   current_chroot_allowed() to make the code easier to read and align
>>>>>>   with user_path_at(), path_permission() and security_path_chroot()
>>>>>>   calls (suggested by Kees Cook).
>>>>>> * Remove now useless included file.
>>>>>> * Extend commit description.
>>>>>> * Rebase on v5.12-rc3 .
>>>>>>
>>>>>> Changes since v2:
>>>>>> * Replace path_is_under() check with current_chrooted() to gain the same
>>>>>>   protection as create_user_ns() (suggested by Jann Horn). See commit
>>>>>>   3151527ee007 ("userns:  Don't allow creation if the user is chrooted")
>>>>>>
>>>>>> Changes since v1:
>>>>>> * Replace custom is_path_beneath() with existing path_is_under().
>>>>>> ---
>>>>>>  fs/open.c | 23 +++++++++++++++++++++--
>>>>>>  1 file changed, 21 insertions(+), 2 deletions(-)
>>>>>>
>>>>>> diff --git a/fs/open.c b/fs/open.c
>>>>>> index e53af13b5835..480010a551b2 100644
>>>>>> --- a/fs/open.c
>>>>>> +++ b/fs/open.c
>>>>>> @@ -532,6 +532,24 @@ SYSCALL_DEFINE1(fchdir, unsigned int, fd)
>>>>>>  	return error;
>>>>>>  }
>>>>>>  
>>>>>> +static inline int current_chroot_allowed(void)
>>>>>> +{
>>>>>> +	/*
>>>>>> +	 * Changing the root directory for the calling task (and its future
>>>>>> +	 * children) requires that this task has CAP_SYS_CHROOT in its
>>>>>> +	 * namespace, or be running with no_new_privs and not sharing its
>>>>>> +	 * fs_struct and not escaping its current root (cf. create_user_ns()).
>>>>>> +	 * As for seccomp, checking no_new_privs avoids scenarios where
>>>>>> +	 * unprivileged tasks can affect the behavior of privileged children.
>>>>>> +	 */
>>>>>> +	if (task_no_new_privs(current) && READ_ONCE(current->fs->users) == 
>>>> 1 &&
>>>>>> +			!current_chrooted())
>>>>>> +		return 0;
>>>>>> +	if (ns_capable(current_user_ns(), CAP_SYS_CHROOT))
>>>>>> +		return 0;
>>>>>> +	return -EPERM;
>>>>>> +}
>>>>>> +
>>>>>>  SYSCALL_DEFINE1(chroot, const char __user *, filename)
>>>>>>  {
>>>>>>  	struct path path;
>>>>>> @@ -546,9 +564,10 @@ SYSCALL_DEFINE1(chroot, const char __user *, filename)
>>>>>>  	if (error)
>>>>>>  		goto dput_and_out;
>>>>>>  
>>>>>> -	error = -EPERM;
>>>>>> -	if (!ns_capable(current_user_ns(), CAP_SYS_CHROOT))
>>>>>> +	error = current_chroot_allowed();
>>>>>> +	if (error)
>>>>>>  		goto dput_and_out;
>>>>>> +
>>>>>>  	error = security_path_chroot(&path);
>>>>>>  	if (error)
>>>>>>  		goto dput_and_out;
>>>>>>