From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751837AbdAMOpj (ORCPT ); Fri, 13 Jan 2017 09:45:39 -0500 Received: from mx1.redhat.com ([209.132.183.28]:32858 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751721AbdAMOph (ORCPT ); Fri, 13 Jan 2017 09:45:37 -0500 Date: Fri, 13 Jan 2017 09:45:08 -0500 From: Jerome Glisse To: Balbir Singh Cc: akpm@linux-foundation.org, linux-kernel@vger.kernel.org, linux-mm@kvack.org, John Hubbard , Russell King , Benjamin Herrenschmidt , Paul Mackerras , Michael Ellerman , Martin Schwidefsky , Heiko Carstens , Yoshinori Sato , Rich Felker , Chris Metcalf , Thomas Gleixner , Ingo Molnar , "H. Peter Anvin" Subject: Re: [HMM v16 01/15] mm/memory/hotplug: convert device bool to int to allow for more flags v2 Message-ID: <20170113144508.GA3758@redhat.com> References: <1484238642-10674-1-git-send-email-jglisse@redhat.com> <1484238642-10674-2-git-send-email-jglisse@redhat.com> <20170113135741.GA26827@localhost.localdomain> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Disposition: inline Content-Transfer-Encoding: 8bit In-Reply-To: <20170113135741.GA26827@localhost.localdomain> User-Agent: Mutt/1.7.1 (2016-10-04) X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.5.16 (mx1.redhat.com [10.5.110.29]); Fri, 13 Jan 2017 14:45:12 +0000 (UTC) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Fri, Jan 13, 2017 at 07:27:41PM +0530, Balbir Singh wrote: > On Thu, Jan 12, 2017 at 11:30:28AM -0500, Jérôme Glisse wrote: > > When hotpluging memory we want more informations on the type of memory and > > its properties. Replace the device boolean flag by an int and define a set > > of flags. > > > > New property for device memory is an opt-in flag to allow page migration > > from and to a ZONE_DEVICE. Existing user of ZONE_DEVICE are not expecting > > page migration to work for their pages. New changes to page migration i > > changing that and we now need a flag to explicitly opt-in page migration. > > Given that ZONE_DEVICE is dependent on X86_64, do we need to touch all > architectures? I guess we could selectively enable things as we enable > ZONE_DEVICE for other architectures? Yes i need to change all architecture because the function prototype changes. I add the bug stuff to be bullet proof from new feature added that might cause trouble if arch does not handle them explicitly. If the bug stuff scares people i can remove it. Cheers, Jérôme