On Mon, Oct 16, 2017 at 05:50:38PM +0200, Stanislav Kinsburskiy wrote:
> Well, patch looks ok.
> But shouldn't all the ve_init_user_ns() replaced by the par->net?
This patch does this.
>
> 14.10.2017 01:20, Andrei Vagin пишет:
> > From: "Eric W. Biederman"
> >
> > ML: 9847371a84b0be330f4bc4aaa
On Mon, Oct 16, 2017 at 01:33:12PM +0300, Konstantin Khorenko wrote:
> Stas, please review the patch.
>
> Andrey, why do we need to support deeper user namespaces at all?
> Someone app tries to create a new userns inside a vz7 CT and use ipt_owner
> inside it?
The kernel grabs userns when we sen
Well, patch looks ok.
But shouldn't all the ve_init_user_ns() replaced by the par->net?
14.10.2017 01:20, Andrei Vagin пишет:
> From: "Eric W. Biederman"
>
> ML: 9847371a84b0be330f4bc4aaa98904101ee8573d
> https://jira.sw.ru/browse/PSBM-69409?
>
> Making this work is a little tricky as it really
Stas, please review the patch.
Andrey, why do we need to support deeper user namespaces at all?
Someone app tries to create a new userns inside a vz7 CT and use ipt_owner
inside it?
--
Best regards,
Konstantin Khorenko,
Virtuozzo Linux Kernel Team
On 10/14/2017 02:20 AM, Andrei Vagin wrote: