From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S933989Ab0EZJpF (ORCPT ); Wed, 26 May 2010 05:45:05 -0400 Received: from gir.skynet.ie ([193.1.99.77]:54088 "EHLO gir.skynet.ie" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S932591Ab0EZJpB (ORCPT ); Wed, 26 May 2010 05:45:01 -0400 Date: Wed, 26 May 2010 10:44:43 +0100 From: Mel Gorman To: Andi Kleen Cc: Naoya Horiguchi , linux-kernel@vger.kernel.org, linux-mm@kvack.org, Andrew Morton , Wu Fengguang , Andrea Arcangeli , Larry Woodman , Lee Schermerhorn Subject: Re: [PATCH 1/7] hugetlb, rmap: add reverse mapping for hugepage Message-ID: <20100526094442.GK29038@csn.ul.ie> References: <1273737326-21211-1-git-send-email-n-horiguchi@ah.jp.nec.com> <1273737326-21211-2-git-send-email-n-horiguchi@ah.jp.nec.com> <20100513152737.GE27949@csn.ul.ie> <20100514074641.GD10000@spritzerA.linux.bs1.fc.nec.co.jp> <20100514095449.GB21481@csn.ul.ie> <20100524071516.GC11008@spritzerA.linux.bs1.fc.nec.co.jp> <20100525105957.GD29038@csn.ul.ie> <20100526065156.GC7128@spritzerA.linux.bs1.fc.nec.co.jp> <20100526091958.GA24615@basil.fritz.box> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-15 Content-Disposition: inline In-Reply-To: <20100526091958.GA24615@basil.fritz.box> User-Agent: Mutt/1.5.17+20080114 (2008-01-14) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Wed, May 26, 2010 at 11:19:58AM +0200, Andi Kleen wrote: > > Mel, other than this nit are you happy with these changes now? > Pretty much but I also want to test the series myself to be sure I haven't missed something in review. > > > It adds another header dependency which is bad but moving hugetlb stuff > > > into mm.h seems bad too. > > > > I have another choice to move the definition of is_vm_hugetlb_page() into > > mm/hugetlb.c and introduce declaration of this function to pagemap.h, > > but this needed a bit ugly #ifdefs and I didn't like it. > > If putting hugetlb code in mm.h is worse, I'll take the second choice > > in the next post. > > You could always create a new include file hugetlb-inlines.h > That would be another option. It'd need to be figured out what should move from hugetlb.h to hugetlb-inlines.h in the future but ultimately it would still be tidier than moving hugetlb stuff to mm.h (at least to me). -- Mel Gorman Part-time Phd Student Linux Technology Center University of Limerick IBM Dublin Software Lab