MIME-Version: 1.0
In-Reply-To: <20180108182510.3c6dda34@alans-desktop>
References: <1515427939-10999-1-git-send-email-w@1wt.eu> <1515427939-10999-4-git-send-email-w@1wt.eu>
 <alpine.DEB.2.20.1801081803290.1735@nanos> <760b7264-1ae7-bcaa-6d20-f47cc7c7fce1@intel.com>
 <20180108175028.acwe3glhw4rsvdsx@gmail.com> <20180108182510.3c6dda34@alans-desktop>
From: Linus Torvalds <torvalds@linux-foundation.org>
Date: Mon, 8 Jan 2018 10:35:32 -0800
Message-ID: <CA+55aFyCGw3sq4s_5qtg5rJGU_F6efWanOutYo4E-eiZSs05nQ@mail.gmail.com>
Subject: Re: [PATCH RFC 3/4] x86/pti: don't mark the user PGD with _PAGE_NX.
To: Alan Cox <gnomes@lxorguk.ukuu.org.uk>
Cc: Ingo Molnar <mingo@kernel.org>,
        Dave Hansen <dave.hansen@intel.com>,
        Thomas Gleixner <tglx@linutronix.de>, Willy Tarreau <w@1wt.eu>,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
        "the arch/x86 maintainers" <x86@kernel.org>
Content-Type: text/plain; charset="UTF-8"
Sender: linux-kernel-owner@vger.kernel.org

On Mon, Jan 8, 2018 at 10:25 AM, Alan Cox <gnomes@lxorguk.ukuu.org.uk> wrote:
>
> I still think cgroups are the best model for this. In particular it
> naturally fits things like containers, or network facing apps that fork
> helpers.
>
> Secondly when you are looking at barrier semantics between client/client
> a cgroup is much more natural as a way to group processes together who
> don't need to be protected from each other as they are trusting each
> other. (Or we could just harcode this based upon ptraceability ?)

I agree that cgroups would be fairly natural, but I do think we could
look at things like simply trusted users too ("running as root? Yeah,
we're not going to try to protect the kernel from you") and/or trusted
binaries.

But all of those things are likely things that can easily be
determined at execve() time.

                      Linus