Re: [PATCH 3/5] v2 seccomp_filters: Enable ftrace-based system call filtering

[Will Drewry]

On Thu, May 26, 2011 at 12:02 PM, Will Drewry <wad@xxxxxxxxxxxx> wrote:
> On Thu, May 26, 2011 at 11:46 AM, Linus Torvalds
> <torvalds@xxxxxxxxxxxxxxxxxxxx> wrote:
>> On Thu, May 26, 2011 at 9:33 AM, Will Drewry <wad@xxxxxxxxxxxx> wrote:
>>>
>>> FWIW, none of the patches deal with privilege escalation via setuid
>>> files or file capabilities.
>>
>> That is NOT AT ALL what I'm talking about.
>>
>> I'm talking about the "setuid()" system call (and all its cousins:
>> setgit/setreuid etc). And the whole thread has been about filtering
>> system calls, no?
>>
>> Do a google code search for setuid.
>>
>> In good code, it will look something like
>>
>>  uid = getuid();
>>
>>  if (setuid(uid)) {
>>    fprintf(stderr, "Unable to drop provileges\n");
>>    exit(1);
>>  }
>>
>> but I guarantee you that there are cases where people just blindly
>> drop privileges. google code search found me at least the "heirloom"
>> source code doing exactly that.
>>
>> And if you filter system calls, it's entirely possible that you can
>> attack suid executables through such a vector. Your "limit system
>> calls for security" security suddenly turned into "avoid the system
>> call that made things secure"!
>>
>> See what I'm saying?
>
> Absolutely - that was what I meant :/  The patches do not currently
> check creds at creation or again at use, which would lead to
> unprivileged filters being used in a privileged context.  Right now,
> though, if setuid() is not allowed by the seccomp-filter, the process
> will be immediately killed with do_exit(SIGKILL) on call -- thus
> avoiding a silent failure. I mentioned file capabilities because they
> can have setuid-like side effects, too.  As long as system call
> rejection results in a process death, I *think* it helps with some of
> this complexity, but I haven't fully vetted the patches for these
> scenarios to be 100% confident.

Bah - by "setuid-like side effects", I meant suid executable-like side
effects.  And I blocking even outside of those scenarios, I think
immediate process-death helps resolves coding mistakes leading to
filtering setuid() calls prior to use.

cheers,
will
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/