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 7CA6FC433FE for ; Tue, 1 Nov 2022 22:39:03 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S230409AbiKAWjB (ORCPT ); Tue, 1 Nov 2022 18:39:01 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:38120 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S230371AbiKAWie (ORCPT ); Tue, 1 Nov 2022 18:38:34 -0400 Received: from mail-pf1-x431.google.com (mail-pf1-x431.google.com [IPv6:2607:f8b0:4864:20::431]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 3C60ADEC1 for ; Tue, 1 Nov 2022 15:38:26 -0700 (PDT) Received: by mail-pf1-x431.google.com with SMTP id 130so14784691pfu.8 for ; Tue, 01 Nov 2022 15:38:26 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20210112; h=in-reply-to:content-disposition:mime-version:references:message-id :subject:cc:to:from:date:sender:from:to:cc:subject:date:message-id :reply-to; bh=thYz5QOD4hcEM7CHTtqZVPg6yh+uUZKrxyF0+TCRPZw=; b=N8aSZ8Cv2bU/JCfAgPf18G7JIZhk2VMtIhfyK0Fwitl+94q0MLBmq4MO3GuccGxWom wtKzmY0iCYZf+wbWoOG1ELeTzsg4O/+9cvTFykXgdpGtyBBOwX1rRJl9+hccyiZz6Iuz YchfG3zPRNRTIwKRAuoZNRoSSr+Du7TGOZYTYMJbF4tQV5k+XHOO3g7E3Sk/I0IwYPsw GXiULWA+v3iZSy6ULqv2Cu80tL9BSpMxwwIOT5ZkSg31hPTq3pfZaeSFPf4SPedjrvfp 4XR3kQlHzzQv2IKOPlsr/QnWX0mIa2NXJTa2qGVsB1Cu04W1Yrn5Q9J97He93C5E4nqO xsWA== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=in-reply-to:content-disposition:mime-version:references:message-id :subject:cc:to:from:date:sender:x-gm-message-state:from:to:cc :subject:date:message-id:reply-to; bh=thYz5QOD4hcEM7CHTtqZVPg6yh+uUZKrxyF0+TCRPZw=; b=cGrotioicjEwVX0csSMgCTxaZuQoQFrzReKTy2qviObdUrXWEQUcMyHB3cWQRy8PcP 3bNliLX7yRWxFzG9SguS0Mk2lsFtI/hfcaRYlLOUQn8jcw7SzJLVWmlG3+9dSeGsGVsq LBUL/iX0PmIERmn0dYghXyFA14H6dL3OGabJ6YhxYI31l9Gf5ijBH1gzNidM6S+0dMDF hehlnGhzx1b0e4TaX2guolY1MAhTKmppmqs873vurnk+iY8xLmyZYafM0qJMdNtvOOEO f6aWXAPQgCdYhMRXJe6lqlYwSALo9fJXYh8fgqS1yqNRRYQxU0Z+JvUnsmeGPd2lE/5Z tlsw== X-Gm-Message-State: ACrzQf02p020xQWvhk59Alwzepub+WDCfG2/CQzWu9h9npu+nrtH8g7l 5pD/5dhnS0qSOYSD8mK024Y= X-Google-Smtp-Source: AMsMyM6GqVxGlHQzQOXFLvR/xw8weg4PlObupJbSs23+IGjDsaYASCGmT3aJ+ADHOdRRYlSG5ungig== X-Received: by 2002:aa7:94b1:0:b0:56c:8da8:4e3 with SMTP id a17-20020aa794b1000000b0056c8da804e3mr21912511pfl.62.1667342305555; Tue, 01 Nov 2022 15:38:25 -0700 (PDT) Received: from localhost ([2620:10d:c090:400::5:f37f]) by smtp.gmail.com with ESMTPSA id h75-20020a62834e000000b0056b932f3280sm6986486pfe.103.2022.11.01.15.38.24 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Tue, 01 Nov 2022 15:38:25 -0700 (PDT) Sender: Tejun Heo Date: Tue, 1 Nov 2022 12:38:23 -1000 From: Tejun Heo To: Josh Don Cc: Peter Zijlstra , Ingo Molnar , Juri Lelli , Vincent Guittot , Dietmar Eggemann , Steven Rostedt , Ben Segall , Mel Gorman , Daniel Bristot de Oliveira , Valentin Schneider , linux-kernel@vger.kernel.org, Joel Fernandes , Michal =?iso-8859-1?Q?Koutn=FD?= , Christian Brauner , Zefan Li Subject: Re: [PATCH v2] sched: async unthrottling for cfs bandwidth Message-ID: References: MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Hello, (cc'ing Michal, Christian and Li for context) On Tue, Nov 01, 2022 at 02:59:56PM -0700, Josh Don wrote: > > If most of the problems were with cpu bw control, fixing that should do for > > the time being. Otherwise, we'll have to think about finishing kernfs > > locking granularity improvements and doing something similar to cgroup > > locking too. > > Oh we've easily hit stalls measured in multiple seconds. We > extensively use cpu.idle to group batch tasks. One of the memory > bandwidth mitigations implemented in userspace is cpu jailing, which > can end up pushing lots and lots of these batch threads onto a small > number of cpus. 5ms min gran * 200 threads is already one second :) Ah, I see. > We're in the process of transitioning to using bw instead for this > instead in order to maintain parallelism. Fixing bw is definitely > going to be useful, but I'm afraid we'll still likely have some issues > from low throughput for non-bw reasons (some of which we can't > directly control, since arbitrary jobs can spin up and configure their > hierarchy/threads in antagonistic ways, in effect pushing out the > latency of some of their threads). Yeah, thanks for the explanation. Making the lock more granular is tedious but definitely doable. I don't think I can work on it in the near future but will keep it on mind. If anyone's interested in attacking it, please be my guest. Thanks. -- tejun