From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-wr0-f197.google.com (mail-wr0-f197.google.com [209.85.128.197]) by kanga.kvack.org (Postfix) with ESMTP id 993E76B02B4 for ; Thu, 27 Jul 2017 04:12:40 -0400 (EDT) Received: by mail-wr0-f197.google.com with SMTP id x43so34549834wrb.9 for ; Thu, 27 Jul 2017 01:12:40 -0700 (PDT) Received: from mx1.suse.de (mx2.suse.de. [195.135.220.15]) by mx.google.com with ESMTPS id z62si1342890wme.145.2017.07.27.01.12.39 for (version=TLS1 cipher=AES128-SHA bits=128/128); Thu, 27 Jul 2017 01:12:39 -0700 (PDT) Date: Thu, 27 Jul 2017 10:12:36 +0200 From: Michal Hocko Subject: Re: gigantic hugepages vs. movable zones Message-ID: <20170727081236.GK20970@dhcp22.suse.cz> References: <20170726105004.GI2981@dhcp22.suse.cz> <87inie1uwf.fsf@linux.vnet.ibm.com> <20170727072857.GI20970@dhcp22.suse.cz> <1529e986-5f28-35dd-c82e-a4b5801b4afd@linux.vnet.ibm.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <1529e986-5f28-35dd-c82e-a4b5801b4afd@linux.vnet.ibm.com> Sender: owner-linux-mm@kvack.org List-ID: To: "Aneesh Kumar K.V" Cc: Luiz Capitulino , Mike Kravetz , linux-mm@kvack.org, LKML On Thu 27-07-17 13:30:31, Aneesh Kumar K.V wrote: > > > On 07/27/2017 12:58 PM, Michal Hocko wrote: > >On Thu 27-07-17 07:52:08, Aneesh Kumar K.V wrote: > >>Michal Hocko writes: > >> > >>>Hi, > >>>I've just noticed that alloc_gigantic_page ignores movability of the > >>>gigantic page and it uses any existing zone. Considering that > >>>hugepage_migration_supported only supports 2MB and pgd level hugepages > >>>then 1GB pages are not migratable and as such allocating them from a > >>>movable zone will break the basic expectation of this zone. Standard > >>>hugetlb allocations try to avoid that by using htlb_alloc_mask and I > >>>believe we should do the same for gigantic pages as well. > >>> > >>>I suspect this behavior is not intentional. What do you think about the > >>>following untested patch? > >> > >> > >>I also noticed an unrelated issue with the usage of > >>start_isolate_page_range. On error we set the migrate type to > >>MIGRATE_MOVABLE. > > > >Why that should be a problem? I think it is perfectly OK to have > >MIGRATE_MOVABLE pageblocks inside kernel zones. > > > > we can pick pages with migrate type movable and if we fail to isolate won't > we set the migrate type of that pages to MOVABLE ? I do not see an immediate problem. GFP_KERNEL allocations can fallback to movable migrate pageblocks AFAIR. But I am not very much familiar with migratetypes. Vlastimil, could you have a look please? -- Michal Hocko SUSE Labs -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@kvack.org. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: email@kvack.org