MIME-Version: 1.0
In-Reply-To: <20161213184057.GA17672@htj.duckdns.org>
References: <1481593143-18756-1-git-send-email-john.stultz@linaro.org>
 <CAKgNAkhGSYmtqLnCyXH8-dbmBJreo9mbVGRMyd_9hZ5D1F7q4g@mail.gmail.com>
 <CALAqxLVz4WH5T7mjD1U5XG1pWQCJqsO2LWYcpJ+xDPjRb9QMig@mail.gmail.com> <20161213184057.GA17672@htj.duckdns.org>
From: John Stultz <john.stultz@linaro.org>
Date: Tue, 13 Dec 2016 10:47:19 -0800
Message-ID: <CALAqxLU=3CC3o6Z_-bzuvtOd6=GAZBfmgbEk9P1UP==PGyJ3bg@mail.gmail.com>
Subject: Re: [PATCH v5] cgroup: Add new capability to allow a process to
 migrate other tasks between cgroups
To: Tejun Heo <tj@kernel.org>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>,
        lkml <linux-kernel@vger.kernel.org>, Li Zefan <lizefan@huawei.com>,
        Jonathan Corbet <corbet@lwn.net>,
        "open list:CONTROL GROUP (CGROUP)" <cgroups@vger.kernel.org>,
        Android Kernel Team <kernel-team@android.com>,
        Rom Lemarchand <romlem@android.com>, Colin Cross <ccross@android.com>,
        Dmitry Shmidt <dimitrysh@google.com>, Todd Kjos <tkjos@google.com>,
        Christian Poetzsch <christian.potzsch@imgtec.com>,
        Amit Pundir <amit.pundir@linaro.org>,
        Dmitry Torokhov <dmitry.torokhov@gmail.com>,
        Kees Cook <keescook@chromium.org>,
        "Serge E . Hallyn" <serge@hallyn.com>,
        Andy Lutomirski <luto@amacapital.net>,
        Linux API <linux-api@vger.kernel.org>
Content-Type: text/plain; charset=UTF-8
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1312
Lines: 33

On Tue, Dec 13, 2016 at 10:40 AM, Tejun Heo <tj@kernel.org> wrote:
> Hello,
>
> On Tue, Dec 13, 2016 at 08:08:16AM -0800, John Stultz wrote:
>> On Tue, Dec 13, 2016 at 1:47 AM, Michael Kerrisk (man-pages)
>> <mtk.manpages@gmail.com> wrote:
>> > On 13 December 2016 at 02:39, John Stultz <john.stultz@linaro.org> wrote:
>> > So, back to the discussion of silos. I understand the argument for
>> > wanting a new silo. But, in that case can we at least try not to make
>> > it a single-use silo?
>> >
>> > How about CAP_CGROUP_CONTROL or some such, with the idea that this
>> > might be a capability that allows the holder to step outside usual
>> > cgroup rules? At the moment, that capability would allow only one such
>> > step, but maybe there would be others in the future.
>>
>> This sounds reasonable to me. Tejun/Andy: Objections?
>
> Control group control?  The word control has a specific meaning for
> cgroups and that second control doesn't make much sense to me.

But this would go against the long tradition of RAS syndrome and
things like "struct task_struct".  :)

>  Given
> how this is mostly to patch up a hole in v1's delegation model and how
> migration operations are different from others, I doubt that we will
> end up overloading it.  Maybe just CAP_CGROUP?

Sounds ok to me.

thanks
-john