On 08/17, Kay Sievers wrote:
>
> On Wed, Aug 17, 2011 at 15:05, Oleg Nesterov <oleg@redhat.com> wrote:
> >
> > But, I seem to remember, that patch cleared ->child_reaper on exec,
>
> I don't think he original patch did.
>
> > I think this makes sense.
>
> Why would it? Systemd can serialize its state and properly re-exec
> itself as many times as needed during its lifetime. Why would the
> kernel take something away from a process, which it explicitly asked
> for?
>
> > And I am not sure about security. No, I do not see any problems, just
> > I don't know. Say, should we check the creds during reparenting? I
> > dunno.
>
> Hmm, I don't see why that would be necessary. It's just one of our
> parents that aks for our signals.

Oh, I do not know. I do not pretend I understand the security ;)

For example. I simply can't understand why do we have security_task_wait().
Why waitpid(my_natural_child) can fail for security reasons? But we have
selinux_task_wait().



So, once again. I am not arguing. I am only asking the questions.
I didn't mean I see any problem here.

Oleg.