From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <linux-kernel-owner@vger.kernel.org>
Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand
        id S936130AbdCXQwV (ORCPT <rfc822;w@1wt.eu>);
        Fri, 24 Mar 2017 12:52:21 -0400
Received: from mga09.intel.com ([134.134.136.24]:35049 "EHLO mga09.intel.com"
        rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP
        id S1757454AbdCXQwO (ORCPT <rfc822;linux-kernel@vger.kernel.org>);
        Fri, 24 Mar 2017 12:52:14 -0400
X-ExtLoop1: 1
X-IronPort-AV: E=Sophos;i="5.36,215,1486454400"; 
   d="scan'208";a="1111827180"
Message-ID: <1490374331.2733.130.camel@linux.intel.com>
Subject: Re: [PATCH -v2 1/2] mm, swap: Use kvzalloc to allocate some swap
 data structure
From: Tim Chen <tim.c.chen@linux.intel.com>
To: Dave Hansen <dave.hansen@linux.intel.com>,
        John Hubbard <jhubbard@nvidia.com>,
        "Huang, Ying" <ying.huang@intel.com>
Cc: David Rientjes <rientjes@google.com>,
        Andrew Morton <akpm@linux-foundation.org>,
        Andi Kleen <ak@linux.intel.com>, Shaohua Li <shli@kernel.org>,
        Rik van Riel <riel@redhat.com>, Michal Hocko <mhocko@suse.com>,
        Mel Gorman <mgorman@techsingularity.net>,
        Aaron Lu <aaron.lu@intel.com>,
        Gerald Schaefer <gerald.schaefer@de.ibm.com>,
        "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>,
        Hugh Dickins <hughd@google.com>, Ingo Molnar <mingo@kernel.org>,
        Vegard Nossum <vegard.nossum@oracle.com>, linux-mm@kvack.org,
        linux-kernel@vger.kernel.org
Date: Fri, 24 Mar 2017 09:52:11 -0700
In-Reply-To: <e79064f1-8594-bef2-fbd8-1579afb4aac3@linux.intel.com>
References: <20170320084732.3375-1-ying.huang@intel.com>
         <alpine.DEB.2.10.1703201430550.24991@chino.kir.corp.google.com>
         <8737e3z992.fsf@yhuang-dev.intel.com>
         <f17cb7e4-4d47-4aed-6fdb-cda5c5d47fa4@nvidia.com>
         <87poh7xoms.fsf@yhuang-dev.intel.com>
         <2d55e06d-a0b6-771a-bba0-f9517d422789@nvidia.com>
         <87d1d7uoti.fsf@yhuang-dev.intel.com>
         <624b8e59-34e5-3538-0a93-d33d9e4ac555@nvidia.com>
         <e79064f1-8594-bef2-fbd8-1579afb4aac3@linux.intel.com>
Content-Type: text/plain; charset="UTF-8"
X-Mailer: Evolution 3.18.5.2 (3.18.5.2-1.fc23) 
Mime-Version: 1.0
Content-Transfer-Encoding: 8bit
Sender: linux-kernel-owner@vger.kernel.org
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org

On Fri, 2017-03-24 at 06:56 -0700, Dave Hansen wrote:
> On 03/24/2017 12:33 AM, John Hubbard wrote:
> > 
> > There might be some additional information you are using to come up with
> > that conclusion, that is not obvious to me. Any thoughts there? These
> > calls use the same underlying page allocator (and I thought that both
> > were subject to the same constraints on defragmentation, as a result of
> > that). So I am not seeing any way that kmalloc could possibly be a
> > less-fragmenting call than vmalloc.
> You guys are having quite a discussion over a very small point.
> 
> But, Ying is right.
> 
> Let's say we have a two-page data structure.  vmalloc() takes two
> effectively random order-0 pages, probably from two different 2M pages
> and pins them.  That "kills" two 2M pages.
> 
> kmalloc(), allocating two *contiguous* pages, is very unlikely to cross
> a 2M boundary (it theoretically could).  That means it will only "kill"
> the possibility of a single 2M page.  More 2M pages == less fragmentation.

In vmalloc, it eventually calls __vmalloc_area_node that allocates the
page one at a time.  There's no attempt there to make the pages contiguous
if I am reading the code correctly.  So that will increase the memory
fragmentation as we will be piecing together pages from all over the places.  

Tim