From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755586Ab1KKEbr (ORCPT ); Thu, 10 Nov 2011 23:31:47 -0500 Received: from gate.crashing.org ([63.228.1.57]:53993 "EHLO gate.crashing.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751326Ab1KKEbq (ORCPT ); Thu, 10 Nov 2011 23:31:46 -0500 Message-ID: <1320985863.21206.40.camel@pasglop> Subject: Re: mm: convert vma->vm_flags to 64bit From: Benjamin Herrenschmidt To: Hugh Dickins Cc: KOSAKI Motohiro , Linus Torvalds , nai.xia@gmail.com, linux-kernel@vger.kernel.org, linux-mm@kvack.org, akpm@linux-foundation.org, dave@linux.vnet.ibm.com, kamezawa.hiroyu@jp.fujitsu.com, lethal@linux-sh.org, linux@arm.linux.org.uk Date: Fri, 11 Nov 2011 15:31:03 +1100 In-Reply-To: References: <20110412151116.B50D.A69D9226@jp.fujitsu.com> <4EBC085D.3060107@jp.fujitsu.com> <1320959579.21206.24.camel@pasglop> Content-Type: text/plain; charset="UTF-8" X-Mailer: Evolution 3.2.1- Content-Transfer-Encoding: 7bit Mime-Version: 1.0 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Thu, 2011-11-10 at 18:09 -0800, Hugh Dickins wrote: > It was in this mail below, when Andrew sent Linus the patch, and Linus > opposed my "argument" in support: that wasn't on lkml or linux-mm, > but I don't see that its privacy needs protecting. > > KOSAKI-san then sent instead a patch to correct some ints to longs, > which Linus did put in: but changing them to a new "vm_flags_t". > > He was, I think, hoping that one of us would change all the other uses > of unsigned long vm_flags to vm_flags_t; but in fact none of us has > stepped up yet - yeah, we're still sulking that we didn't get our > shiny new 64-bit vm_flags ;) > > I think Linus is not opposed to PowerPC and others defining a 64-bit > vm_flags_t if you need it, but wants not to bloat the x86_32 vma. > > I'm still wary of the contortions we go to in constraining flags, > and feel that the 32-bit case holds back the 64-bit, which would > not itself be bloated at all. > > The subject is likely to come up again, more pressingly, with page > flags. Right, tho the good first step is to convert everything to vm_flags_t so we can easily switch if we want to, even on a per-arch basis... Oh well, now all we need is a volunteer :-) Cheers, Ben.