Re: [Qemu-devel] [PATCH for-1.7] seccomp: setting "-sandbox on" by default

[Eduardo Otubo <otubo@linux.vnet.ibm.com>]

On 11/21/2013 01:48 PM, Paul Moore wrote:
> On Thursday, November 21, 2013 04:14:11 PM Paolo Bonzini wrote:
>> Il 30/10/2013 11:04, Stefan Hajnoczi ha scritto:
>>> On Wed, Oct 23, 2013 at 12:42:34PM -0200, Eduardo Otubo wrote:
>>>> On 10/22/2013 11:00 AM, Anthony Liguori wrote:
>>>>> On Tue, Oct 22, 2013 at 12:21 PM, Eduardo Otubo
>>>>>
>>>>> <otubo@linux.vnet.ibm.com> wrote:
>>>>>> Inverting the way sandbox handles arguments, making possible to have no
>>>>>> argument and still have '-sandbox on' enabled.
>>>>>>
>>>>>> Signed-off-by: Eduardo Otubo <otubo@linux.vnet.ibm.com>
>>>>>> ---
>>>>>>
>>>>>> The option '-sandbox on' is now used by default by virt-test[0] -- it
>>>>>> has been merged into the 'next' branch and will be available in the
>>>>>> next release, meaning we have a back support for regression tests if
>>>>>> anything breaks because of some missing system call not listed in the
>>>>>> whitelist.
>>>>>>
>>>>>> This being said, I think it makes sense to have this option set to 'on'
>>>>>> by
>>>>>> default in the next Qemu version. It's been a while since no missing
>>>>>> syscall is reported and at this point the whitelist seems to be pretty
>>>>>> mature.
>>>>>>
>>>>>> [0] -
>>>>>> https://github.com/autotest/virt-test/commit/50e1f7d47a94f4c770880cd8e
>>>>>> c0f18365dcba714>>>
>>>>> This breaks hot_add of a network device that uses a script= argument,
>>>>> correct?
>>>>>
>>>>> If so, this cannot be made default.
>>>>
>>>> Anthony, I believe you're talking about the blacklist feature. This
>>>> is the old whitelist that is already upstream and it does not block
>>>> any network device to be hot plugged.
>>>
>>> The following fails to start here (the shell hangs and ps shows QEMU is
>>> a <defunct> process):
>>>
>>> qemu-system-x86_64 -sandbox on -enable-kvm -m 1024 -cpu host \
>>>
>>>                     -drive if=virtio,cache=none,file=test.img
>>
>> Easier-to-debug failures are another prerequisite for enabling the
>> sandbox by default, I think.

I've already sent a patch for a "debug mode" in the past. It was denied 
because of two main points: (1) Anthony was looking for a more solid and 
closed solution for sandboxing. A "debug" or "learning" mode would be 
too much exposure of the attack surface. (2) Debug mode was changing the 
Qemu's sig mask in a way that it was breaking it. And besides, at that 
time, there were too many linked libraries that would interfere on this 
sig handling, so we gave up on this feature.

>
> I believe I've posted this information before, but just in case ...
>
> IMHO, it is really not that hard to debug a seccomp failure; the first step is
> to look for the failure in the audit log or syslog.  If you are on a
> Fedora/RHEL based system you are most likely running audit, so finding the
> seccomp failures are quite simple with the 'ausearch' command:
>
>   # ausearch -m SECCOMP
>   ----
>   time->Wed Nov 20 09:52:08 2013
>   type=SECCOMP msg=audit(1384912328.482:6656): auid=0 uid=0 gid=0 ses=854
>    subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 pid=12087
>    comm="qemu-kvm" sig=31 syscall=62 compat=0 ip=0x7f7a1d2abc67 code=0x0
>
> ... if you are using syslog, feel free to use whatever tool you prefer, e.g.
> grep, less, etc.
>
> Once you have the syscall number, "syscall=62", in the audit message above,
> you can use the 'scmp_sys_resolver' to resolve the number into a name:
>
>   # scmp_sys_resolver 62
>   kill
>
> The 'scmp_sys_resolver' tool is part of the libseccomp-devel package on
> Fedora/RHEL based systems, it may be packaged differently on other
> distributions.
>
> I'm always open to suggestions on how to improve the development/debugging
> process, so if you have any ideas please let me know.

Also, I've been working on some improvements on virt-test side. I'm 
trying to make it report the illegal system call using audit log and 
libseccomp as source of information. This way a simple run could 
identify missing system calls for the whitelist.

-- 
Eduardo Otubo
IBM Linux Technology Center