[PATCH v2 bpf-next 1/4] bpf: unprivileged BPF access via /dev/bpf

[Song Liu]

> On Jul 30, 2019, at 1:24 PM, Andy Lutomirski <luto at kernel.org> wrote:
> 
> On Mon, Jul 29, 2019 at 10:07 PM Song Liu <songliubraving at fb.com> wrote:
>> 
>> Hi Andy,
>> 
>>> On Jul 27, 2019, at 11:20 AM, Song Liu <songliubraving at fb.com> wrote:
>>> 
>>> Hi Andy,
>>> 
>>> 

[...]

>>> 
>> 
>> I would like more comments on this.
>> 
>> Currently, bpf permission is more or less "root or nothing", which we
>> would like to change.
>> 
>> The short term goal is to separate bpf from root, in other words, it is
>> "all or nothing". Special user space utilities, such as systemd, would
>> benefit from this. Once this is implemented, systemd can call sys_bpf()
>> when it is not running as root.
> 
> As generally nasty as Linux capabilities are, this sounds like a good
> use for CAP_BPF_ADMIN.

I actually agree CAP_BPF_ADMIN makes sense. The hard part is to make 
existing tools (setcap, getcap, etc.) and libraries aware of the new CAP.

> 
> But what do you have in mind?  Isn't non-root systemd mostly just the
> user systemd session?  That should *not* have bpf() privileges until
> bpf() is improved such that you can't use it to compromise the system.

cgroup bpf is the major use case here. A less important use case is to 
run bpf selftests without being root. 

> 
>> 
>> In longer term, it may be useful to provide finer grain permission of
>> sys_bpf(). For example, sys_bpf() should be aware of containers; and
>> user may only have access to certain bpf maps. Let's call this
>> "fine grain" capability.
>> 
>> 
>> Since we are seeing new use cases every year, we will need many
>> iterations to implement the fine grain permission. I think we need an
>> API that is flexible enough to cover different types of permission
>> control.
>> 
>> For example, bpf_with_cap() can be flexible:
>> 
>>        bpf_with_cap(cmd, attr, size, perm_fd);
>> 
>> We can get different types of permission via different combinations of
>> arguments:
>> 
>>    A perm_fd to /dev/bpf gives access to all sys_bpf() commands, so
>>    this is "all or nothing" permission.
>> 
>>    A perm_fd to /sys/fs/cgroup/.../bpf.xxx would only allow some
>>    commands to this specific cgroup.
>> 
> 
> I don't see why you need to invent a whole new mechanism for this.
> The entire cgroup ecosystem outside bpf() does just fine using the
> write permission on files in cgroupfs to control access.  Why can't
> bpf() do the same thing?

It is easier to use write permission for BPF_PROG_ATTACH. But it is 
not easy to do the same for other bpf commands: BPF_PROG_LOAD and 
BPF_MAP_*. A lot of these commands don't have target concept. Maybe 
we should have target concept for all these commands. But that is a 
much bigger project. OTOH, "all or nothing" model allows all these 
commands at once.

Well, that being said, I will look more into using write permission 
in cgroupfs. 

Thanks again for all these comments and suggestions. Please let us 
know your future thoughts and insights. 

Song