From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932503Ab2CNNGH (ORCPT ); Wed, 14 Mar 2012 09:06:07 -0400 Received: from e28smtp06.in.ibm.com ([122.248.162.6]:33732 "EHLO e28smtp06.in.ibm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S932360Ab2CNNGF (ORCPT ); Wed, 14 Mar 2012 09:06:05 -0400 From: "Aneesh Kumar K.V" To: Andrew Morton Cc: linux-mm@kvack.org, mgorman@suse.de, kamezawa.hiroyu@jp.fujitsu.com, dhillf@gmail.com, aarcange@redhat.com, mhocko@suse.cz, hannes@cmpxchg.org, linux-kernel@vger.kernel.org, cgroups@vger.kernel.org Subject: Re: [PATCH -V3 0/8] memcg: Add memcg extension to control HugeTLB allocation In-Reply-To: <20120313144930.284228c4.akpm@linux-foundation.org> References: <1331622432-24683-1-git-send-email-aneesh.kumar@linux.vnet.ibm.com> <20120313144930.284228c4.akpm@linux-foundation.org> User-Agent: Notmuch/0.11.1+190~g31a336a (http://notmuchmail.org) Emacs/23.3.1 (x86_64-pc-linux-gnu) Date: Wed, 14 Mar 2012 18:31:09 +0530 Message-ID: <87fwdb8hgq.fsf@linux.vnet.ibm.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii x-cbid: 12031413-9574-0000-0000-000001CC1D44 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Tue, 13 Mar 2012 14:49:30 -0700, Andrew Morton wrote: > On Tue, 13 Mar 2012 12:37:04 +0530 > "Aneesh Kumar K.V" wrote: > > > This patchset implements a memory controller extension to control > > HugeTLB allocations. > > Well, why? What are the use cases? Who is asking for this? Why do > they need it and how will they use it? etcetera. > > Please explain, with some care, why you think we should add this > feature to the kernel. So that others can assess whether the value it > adds is worth the cost of adding and maintaining it. > The goal is to control how many HugeTLB pages a group of task can allocate. It can be looked at as an extension of the existing quota interface which limits the number of HugeTLB pages per hugetlbfs superblock. HPC job scheduler requires jobs to specify their resource requirements in the job file. Once their requirements can be met, job schedulers like (SLURM) will schedule the job. We need to make sure that the jobs won't consume more resources than requested. If they do we should error out or kill the application. -aneesh