From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752530AbcBLMXN (ORCPT ); Fri, 12 Feb 2016 07:23:13 -0500 Received: from mx2.parallels.com ([199.115.105.18]:60229 "EHLO mx2.parallels.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752447AbcBLMXM (ORCPT ); Fri, 12 Feb 2016 07:23:12 -0500 Subject: Re: [RFC PATCH 3/3] mm: increase scalability of global memory commitment accounting To: Andrew Morton , Tim Chen References: <1455115941-8261-1-git-send-email-aryabinin@virtuozzo.com> <1455115941-8261-3-git-send-email-aryabinin@virtuozzo.com> <1455127253.715.36.camel@schen9-desk2.jf.intel.com> <20160210132818.589451dbb5eafae3fdb4a7ec@linux-foundation.org> <1455150256.715.60.camel@schen9-desk2.jf.intel.com> <20160211125103.8a4fb0ffed593938321755d2@linux-foundation.org> CC: , , Andi Kleen , Mel Gorman , Vladimir Davydov , Konstantin Khlebnikov From: Andrey Ryabinin Message-ID: <56BDCEF2.3030409@virtuozzo.com> Date: Fri, 12 Feb 2016 15:24:18 +0300 User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:38.0) Gecko/20100101 Thunderbird/38.5.0 MIME-Version: 1.0 In-Reply-To: <20160211125103.8a4fb0ffed593938321755d2@linux-foundation.org> Content-Type: text/plain; charset="windows-1252" Content-Transfer-Encoding: 7bit X-ClientProxiedBy: US-EXCH2.sw.swsoft.com (10.255.249.46) To US-EXCH.sw.swsoft.com (10.255.249.47) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On 02/11/2016 11:51 PM, Andrew Morton wrote: > On Wed, 10 Feb 2016 16:24:16 -0800 Tim Chen wrote: > >> On Wed, 2016-02-10 at 13:28 -0800, Andrew Morton wrote: >> >>> >>> If a process is unmapping 4MB then it's pretty crazy for us to be >>> hitting the percpu_counter 32 separate times for that single operation. >>> >>> Is there some way in which we can batch up the modifications within the >>> caller and update the counter less frequently? Perhaps even in a >>> single hit? >> >> I think the problem is the batch size is too small and we overflow >> the local counter into the global counter for 4M allocations. > > That's one way of looking at the issue. The other way (which I point > out above) is that we're calling vm_[un]_acct_memory too frequently > when mapping/unmapping 4M segments. > We call it only once per mmap() or munmap(), so there is nothing to improve. > Exactly which mmap.c callsite is causing this issue? > mmap_region() (or do_brk()) -> security_vm_enough_memory() -> __vm_enough_memory() -> vm_acct_memory()