On Thu, 2013-10-10 at 09:38 +0900, Tetsuo Handa wrote:
> Greg KH wrote:
> > > We should check whether this patch actually fixes the problem in 3.{0,4,10}.
> > > RHEL kernels would want to check for NULL at do_coredump().
> >
> > Can you do that?
>
> OK. I checked the status of this bug.
>
> v3.11.4 : This bug remains.
> v3.10.15 : Fix already backported.
> v3.9.11 : Fix already backported.
> v3.8.13 : This bug remains.
> v3.7.10 : This bug remains.
> v3.6.11 : This bug remains.
> v3.5.7 : This bug remains.
> v3.4.65 : Fix already backported.
> v3.3.8 : This bug remains.
> v3.2.51 : Fix already backported.
> v3.1.10 : This bug remains.
> v3.0.99 : Fix already backported.
> v2.6.32.61 : This bug remains.
>
> Regarding vanilla LTS kernels, only 3.11-stable wants this patch.
Thanks very much for sorting this out Tetsuo.
> Kamal, your v3.8.13.10 kernel already has the fix backported.
Ah yes, I see that now. I've now dropped this from the 3.8-stable
queue:
4c1c7be kernel/kmod.c: check for NULL in call_usermodehelper_exec()
-Kamal
Attachment:
signature.asc
Description: This is a digitally signed message part