From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755957AbdJPWJR (ORCPT ); Mon, 16 Oct 2017 18:09:17 -0400 Received: from mail-qk0-f176.google.com ([209.85.220.176]:54294 "EHLO mail-qk0-f176.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753006AbdJPWJQ (ORCPT ); Mon, 16 Oct 2017 18:09:16 -0400 X-Google-Smtp-Source: ABhQp+SP+VdrXY851vTrUOBUxrocWogA2kIi5k2x9k55PxGNpwgYDtGs/gHwofzICq4YpAbCydRy+w== Subject: Re: [PATCH v5 2/2] staging: ion: create one device entry per heap To: Mark Brown Cc: Benjamin Gaignard , Sandeep Patil , driverdevel , Greg Kroah-Hartman , =?UTF-8?Q?Arve_Hj=c3=b8nnev=c3=a5g?= , "dri-devel@lists.freedesktop.org" , Linux Kernel Mailing List , Riley Andrews , linux-api@vger.kernel.org, Sumit Semwal , Dan Carpenter References: <1506518409-16887-3-git-send-email-benjamin.gaignard@linaro.org> <2e15edc2-a17f-3930-7d5b-4b5b7d2e0a4d@redhat.com> <20171003164849.rcdgez6lbpmq5llt@sirena.org.uk> <2417c969-357f-d5d9-153a-2180d09b0dc6@redhat.com> <20171003230830.GA132839@sspatil-desktop.mtv.corp.google.com> <20171004101720.ob5f467tro7agpcz@sirena.co.uk> <20171009220858.zwbguqkamzmswqcq@sirena.co.uk> <20171010091151.rfdsphstboubm7ps@sirena.co.uk> From: Laura Abbott Message-ID: <2a9246e0-eafa-7812-6cf4-2482ff1ed158@redhat.com> Date: Mon, 16 Oct 2017 15:09:11 -0700 User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:52.0) Gecko/20100101 Thunderbird/52.3.0 MIME-Version: 1.0 In-Reply-To: <20171010091151.rfdsphstboubm7ps@sirena.co.uk> Content-Type: text/plain; charset=windows-1252 Content-Language: en-US Content-Transfer-Encoding: 8bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On 10/10/2017 02:11 AM, Mark Brown wrote: > On Mon, Oct 09, 2017 at 05:10:37PM -0700, Laura Abbott wrote: >> On 10/09/2017 03:08 PM, Mark Brown wrote: >>> On Mon, Oct 09, 2017 at 02:25:47PM -0700, Laura Abbott wrote: > >>>> Anyway, to move this forward I think we need to see a proof of concept >>>> of using selinux to protect access to specific heaps. > >>> Aren't Unix permissions enough with separate files or am I >>> misunderstanding what you're looking to see a proof of concept for? > >> The goal is to be able to restrict heap access to certain services >> and selinux groups on Android so straight unix permissions aren't >> sufficient. > > Oh, there's Android users for this? The users I was aware of were > non-Android. Though even so I'd have thought that given that SELinux is > a superset of Unix file permissions it ought to be sufficient to be able > to use them. I'd been thinking people were suggesting SELinux as a > replacement for file permissions, using the single file and the greater > capabilities of SELinux. > Unix file permissions are necessary but not sufficient, they can be used separately. Mostly what I want to see before merging this is an example that splitting the Ion heaps provides more protection than just keeping /dev/ion. Thanks, Laura