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 Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 3AB0FC38142 for ; Fri, 27 Jan 2023 08:13:17 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S232200AbjA0INQ (ORCPT ); Fri, 27 Jan 2023 03:13:16 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:59152 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S229874AbjA0INO (ORCPT ); Fri, 27 Jan 2023 03:13:14 -0500 Received: from us-smtp-delivery-124.mimecast.com (us-smtp-delivery-124.mimecast.com [170.10.129.124]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id A741776425 for ; Fri, 27 Jan 2023 00:12:22 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1674807141; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version:content-type:content-type: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=tfn6EbWtPNJebb879MWv3dOIAf3Kgq4dei2DnXcoD6U=; b=DXVnTFLZJfNCLJ4JUjjqt3odxigb9eCzLJVfkvtpcCLjtI5Sfttrz75yUEFmE15Vq8pxrc xuEvF0zayt2PwhZbBCRCVtc+YPBKKmidOJXmxeHG3hU7TzLgQ6moCszcDBkgZS90Z5PIh3 As/8xPyJO12XIAICElM7vaVtRgrh26g= Received: from mail-ot1-f70.google.com (mail-ot1-f70.google.com [209.85.210.70]) by relay.mimecast.com with ESMTP with STARTTLS (version=TLSv1.3, cipher=TLS_AES_128_GCM_SHA256) id us-mta-571-NDZRqwSCOmaPxyHLmu4kLA-1; Fri, 27 Jan 2023 03:12:20 -0500 X-MC-Unique: NDZRqwSCOmaPxyHLmu4kLA-1 Received: by mail-ot1-f70.google.com with SMTP id h19-20020a9d3e53000000b0068b94d630b3so352424otg.12 for ; Fri, 27 Jan 2023 00:12:20 -0800 (PST) X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=mime-version:user-agent:content-transfer-encoding:references :in-reply-to:date:cc:to:from:subject:message-id:x-gm-message-state :from:to:cc:subject:date:message-id:reply-to; bh=tfn6EbWtPNJebb879MWv3dOIAf3Kgq4dei2DnXcoD6U=; b=wwmoCZYZooXzvlRJiYUOLPYN8XEZrBF7TrgnbDsJJWGsEYhulqo61pW+kQ4JWx1O9c zaMyqBadNCs25t+FxP6JCgIm/T8nh8haadM6yO/LAvJKqZykBXQRuVoyqPSCAR8pfBv4 +OSUtxu98wdSlwK4Nm1MQcYUOIAqQSJf08xFRfI3i39mOcncX+i/2eoFPWZaBkuz7X9c 3BA8/HtU2OqAJbD+MDaf4AOA/hOxMu9YfZbj8Dzrg/dr6kyw5F91QYZzFyZp9brUHz1w MoUsSidihm6krHR+x5t7mDE9uYlxFKXFox7Q4Nxb4LFDuBzOYIJ2F467nOOk3CR1f4KG ew1Q== X-Gm-Message-State: AO0yUKXFoOsjD4MYU41LjsEQjW7kUH3Xv8TvzWEcAvjw3J88vDdoknHn FohWwW5HPia6BLvrmgxr4SW6Cqtl2s/GDbl9y/YW83roJwHV6G9fG2cavg9xUZqKMtrHNtlivTX 4g+Gcyg78tqmwAe7DBftYHbOs X-Received: by 2002:a05:6808:10:b0:375:12e1:28c5 with SMTP id u16-20020a056808001000b0037512e128c5mr1907273oic.55.1674807139559; Fri, 27 Jan 2023 00:12:19 -0800 (PST) X-Google-Smtp-Source: AK7set/Up483fYHxeZaymHaWpfU51pJD+d71I5SgB3hjGZAWv7PXqj9pJM37Mfqh56wsfN+i4lQDMg== X-Received: by 2002:a05:6808:10:b0:375:12e1:28c5 with SMTP id u16-20020a056808001000b0037512e128c5mr1907268oic.55.1674807139388; Fri, 27 Jan 2023 00:12:19 -0800 (PST) Received: from ?IPv6:2804:1b3:a800:6912:c477:c73a:cf7c:3a27? ([2804:1b3:a800:6912:c477:c73a:cf7c:3a27]) by smtp.gmail.com with ESMTPSA id g20-20020a9d6b14000000b0068848d6b231sm1480090otp.30.2023.01.27.00.12.15 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Fri, 27 Jan 2023 00:12:18 -0800 (PST) Message-ID: <601fc35a8cc2167e53e45c636fccb2d899fd7c50.camel@redhat.com> Subject: Re: [PATCH v2 0/5] Introduce memcg_stock_pcp remote draining From: Leonardo =?ISO-8859-1?Q?Br=E1s?= To: Michal Hocko , Roman Gushchin Cc: Marcelo Tosatti , Johannes Weiner , Shakeel Butt , Muchun Song , Andrew Morton , cgroups@vger.kernel.org, linux-mm@kvack.org, linux-kernel@vger.kernel.org, Frederic Weisbecker Date: Fri, 27 Jan 2023 05:12:13 -0300 In-Reply-To: <52a0f1e593b1ec0ca7e417ba37680d65df22de82.camel@redhat.com> References: <20230125073502.743446-1-leobras@redhat.com> <9e61ab53e1419a144f774b95230b789244895424.camel@redhat.com> <52a0f1e593b1ec0ca7e417ba37680d65df22de82.camel@redhat.com> Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable User-Agent: Evolution 3.46.2 MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Fri, 2023-01-27 at 04:22 -0300, Leonardo Br=C3=A1s wrote: > On Fri, 2023-01-27 at 08:11 +0100, Michal Hocko wrote: > > [Cc Frederic] > >=20 > > On Thu 26-01-23 15:12:35, Roman Gushchin wrote: > > > On Thu, Jan 26, 2023 at 08:41:34AM +0100, Michal Hocko wrote: > > [...] > > > > > Essentially each cpu will try to grab the remains of the memory q= uota > > > > > and move it locally. I wonder in such circumstances if we need to= disable the pcp-caching > > > > > on per-cgroup basis. > > > >=20 > > > > I think it would be more than sufficient to disable pcp charging on= an > > > > isolated cpu. > > >=20 > > > It might have significant performance consequences. > >=20 > > Is it really significant? > >=20 > > > I'd rather opt out of stock draining for isolated cpus: it might slig= htly reduce > > > the accuracy of memory limits and slightly increase the memory footpr= int (all > > > those dying memcgs...), but the impact will be limited. Actually it i= s limited > > > by the number of cpus. > >=20 > > Hmm, OK, I have misunderstood your proposal. Yes, the overal pcp charge= s > > potentially left behind should be small and that shouldn't really be a > > concern for memcg oom situations (unless the limit is very small and > > workloads on isolated cpus using small hard limits is way beyond my > > imagination). > >=20 > > My first thought was that those charges could be left behind without an= y > > upper bound but in reality sooner or later something should be running > > on those cpus and if the memcg is gone the pcp cache would get refilled > > and old charges gone. > >=20 > > So yes, this is actually a better and even simpler solution. All we nee= d > > is something like this > > diff --git a/mm/memcontrol.c b/mm/memcontrol.c > > index ab457f0394ab..13b84bbd70ba 100644 > > --- a/mm/memcontrol.c > > +++ b/mm/memcontrol.c > > @@ -2344,6 +2344,9 @@ static void drain_all_stock(struct mem_cgroup *ro= ot_memcg) > > struct mem_cgroup *memcg; > > bool flush =3D false; > > =20 > > + if (cpu_is_isolated(cpu)) > > + continue; > > + > > rcu_read_lock(); > > memcg =3D stock->cached; > > if (memcg && stock->nr_pages && > >=20 > > There is no such cpu_is_isolated() AFAICS so we would need a help from > > NOHZ and cpuisol people to create one for us. Frederic, would such an > > abstraction make any sense from your POV? >=20 >=20 > IIUC, 'if (cpu_is_isolated())' would be instead: >=20 > if (!housekeeping_cpu(smp_processor_id(), HK_TYPE_DOMAIN) || > !housekeeping_cpu(smp_processor_id(), HK_TYPE_WQ) oh, sorry 's/smp_processor_id()/cpu/' here: if(!housekeeping_cpu(cpu, HK_TYPE_DOMAIN) || !housekeeping_cpu(cpu, HK_TYPE= _WQ)) Not sure why I added smp_processor_id() instead of cpu. I think I need some sleep. :)