From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-5.3 required=3.0 tests=BAYES_00, HEADER_FROM_DIFFERENT_DOMAINS,MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS, USER_AGENT_SANE_1 autolearn=no autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 89A0DC433E6 for ; Fri, 5 Mar 2021 02:21:36 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id 541226500D for ; Fri, 5 Mar 2021 02:21:36 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S229642AbhCECVf (ORCPT ); Thu, 4 Mar 2021 21:21:35 -0500 Received: from mga17.intel.com ([192.55.52.151]:15710 "EHLO mga17.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S229523AbhCECVe (ORCPT ); Thu, 4 Mar 2021 21:21:34 -0500 IronPort-SDR: 0SwaPSXD4v3u8I/MbhJxdduLSEhUHl1TkNQ84h0+yceLHZnTHm3qhmjyHoObd5dfsOoBQUUVwW 24Lhl/vRidOg== X-IronPort-AV: E=McAfee;i="6000,8403,9913"; a="167446370" X-IronPort-AV: E=Sophos;i="5.81,224,1610438400"; d="scan'208";a="167446370" Received: from fmsmga008.fm.intel.com ([10.253.24.58]) by fmsmga107.fm.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 04 Mar 2021 18:21:27 -0800 IronPort-SDR: mvh8AuzZGQ8jGtK6aR008QyOmhqP4HI5sVJHzqbx60ezQUE4aI7QPtdUXpYE2HwcyOTcR/uJzm cNsxrrQQ+slw== X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.81,224,1610438400"; d="scan'208";a="401047505" Received: from shbuild999.sh.intel.com (HELO localhost) ([10.239.146.165]) by fmsmga008.fm.intel.com with ESMTP; 04 Mar 2021 18:21:23 -0800 Date: Fri, 5 Mar 2021 10:21:22 +0800 From: Feng Tang To: Michal Hocko Cc: "linux-mm@kvack.org" , "linux-kernel@vger.kernel.org" , Andrew Morton , Andrea Arcangeli , David Rientjes , Mel Gorman , Mike Kravetz , Randy Dunlap , Vlastimil Babka , "Hansen, Dave" , Andi leen , "Williams, Dan J" Subject: Re: [PATCH v3 RFC 14/14] mm: speedup page alloc for MPOL_PREFERRED_MANY by adding a NO_SLOWPATH gfp bit Message-ID: <20210305022122.GA17707@shbuild999.sh.intel.com> References: <20210303121833.GB16736@shbuild999.sh.intel.com> <20210303131832.GB78458@shbuild999.sh.intel.com> <20210303134644.GC78458@shbuild999.sh.intel.com> <20210303163141.v5wu2sfo2zj2qqsw@intel.com> <20210303172250.wbp47skyuf6r37wi@intel.com> <20210304081414.GC43191@shbuild999.sh.intel.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.5.24 (2015-08-30) Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Thu, Mar 04, 2021 at 01:59:40PM +0100, Michal Hocko wrote: > On Thu 04-03-21 16:14:14, Feng Tang wrote: > > On Wed, Mar 03, 2021 at 09:22:50AM -0800, Ben Widawsky wrote: > > > On 21-03-03 18:14:30, Michal Hocko wrote: > > > > On Wed 03-03-21 08:31:41, Ben Widawsky wrote: > > > > > On 21-03-03 14:59:35, Michal Hocko wrote: > > > > > > On Wed 03-03-21 21:46:44, Feng Tang wrote: > > > > > > > On Wed, Mar 03, 2021 at 09:18:32PM +0800, Tang, Feng wrote: > > > > > > > > On Wed, Mar 03, 2021 at 01:32:11PM +0100, Michal Hocko wrote: > > > > > > > > > On Wed 03-03-21 20:18:33, Feng Tang wrote: > > > > > > [...] > > > > > > > > > > One thing I tried which can fix the slowness is: > > > > > > > > > > > > > > > > > > > > + gfp_mask &= ~(__GFP_DIRECT_RECLAIM | __GFP_KSWAPD_RECLAIM); > > > > > > > > > > > > > > > > > > > > which explicitly clears the 2 kinds of reclaim. And I thought it's too > > > > > > > > > > hacky and didn't mention it in the commit log. > > > > > > > > > > > > > > > > > > Clearing __GFP_DIRECT_RECLAIM would be the right way to achieve > > > > > > > > > GFP_NOWAIT semantic. Why would you want to exclude kswapd as well? > > > > > > > > > > > > > > > > When I tried gfp_mask &= ~__GFP_DIRECT_RECLAIM, the slowness couldn't > > > > > > > > be fixed. > > > > > > > > > > > > > > I just double checked by rerun the test, 'gfp_mask &= ~__GFP_DIRECT_RECLAIM' > > > > > > > can also accelerate the allocation much! though is still a little slower than > > > > > > > this patch. Seems I've messed some of the tries, and sorry for the confusion! > > > > > > > > > > > > > > Could this be used as the solution? or the adding another fallback_nodemask way? > > > > > > > but the latter will change the current API quite a bit. > > > > > > > > > > > > I haven't got to the whole series yet. The real question is whether the > > > > > > first attempt to enforce the preferred mask is a general win. I would > > > > > > argue that it resembles the existing single node preferred memory policy > > > > > > because that one doesn't push heavily on the preferred node either. So > > > > > > dropping just the direct reclaim mode makes some sense to me. > > > > > > > > > > > > IIRC this is something I was recommending in an early proposal of the > > > > > > feature. > > > > > > > > > > My assumption [FWIW] is that the usecases we've outlined for multi-preferred > > > > > would want more heavy pushing on the preference mask. However, maybe the uapi > > > > > could dictate how hard to try/not try. > > > > > > > > What does that mean and what is the expectation from the kernel to be > > > > more or less cast in stone? > > > > > > > > > > (I'm not positive I've understood your question, so correct me if I > > > misunderstood) > > > > > > I'm not sure there is a stone-cast way to define it nor should we. At the very > > > least though, something in uapi that has a general mapping to GFP flags > > > (specifically around reclaim) for the first round of allocation could make > > > sense. > > > > > > In my head there are 3 levels of request possible for multiple nodes: > > > 1. BIND: Those nodes or die. > > > 2. Preferred hard: Those nodes and I'm willing to wait. Fallback if impossible. > > > 3. Preferred soft: Those nodes but I don't want to wait. > > > > > > Current UAPI in the series doesn't define a distinction between 2, and 3. As I > > > understand the change, Feng is defining the behavior to be #3, which makes #2 > > > not an option. I sort of punted on defining it entirely, in the beginning. > > > > As discussed earlier in the thread, one less hacky solution is to clear > > __GFP_DIRECT_RECLAIM bit so that it won't go into direct reclaim, but still > > wakeup the kswapd of target nodes and retry, which sits now between 'Preferred hard' > > and 'Preferred soft' :) > > Yes that is what I've had in mind when talking about a lightweight > attempt. > > > For current MPOL_PREFERRED, its semantic is also 'Preferred hard', that it > > Did you mean to say prefer soft? Because the direct reclaim is attempted > only when node reclaim is enabled. > > > will check free memory of other nodes before entering slowpath waiting. > > Yes, hence "soft" semantic. Yes, it's the #3 item: 'Preferred soft' Thanks, Feng > -- > Michal Hocko > SUSE Labs