From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mail-lf1-f47.google.com (mail-lf1-f47.google.com [209.85.167.47]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 4B2103939BE for ; Tue, 13 Jan 2026 17:58:49 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=209.85.167.47 ARC-Seal:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1768327130; cv=none; b=WYdwQbitCuqzKLmX6PEKDynP9JT5sApcUA1B0QQg9+DO+VF6EDJNqT16FxV0mflcAmXo9OGXWnrgEITYDhvpYl6vtHCJi4jO81N4ZWsU/+qFdmeIwu9crROmB2R56IFXGe0pQtwpJ2rXsPkcda23DHJ1rG548FlHH5NEsmCojHM= ARC-Message-Signature:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1768327130; c=relaxed/simple; bh=SfPB3/xLf0+Q0sgIK1rIaZu9INn0FIT+GeEV0IL3yNY=; h=From:Date:To:Cc:Subject:Message-ID:References:MIME-Version: Content-Type:Content-Disposition:In-Reply-To; b=D4/tXk2TtAajW/NlPgHELiDxS2m5KShY7hkCFSr3c7218Z0GyMoBXkDFklPikFs6BUsoj4S3aoJjlc4sbxUyfezIo5gY8viqd+PhN5ZfTJQCLqe25Vg0XXMV8AM8QHJz6FEaRRHhHzm9+eBRmlRVCSHUYLBjxuC13KfJBpR11Uk= ARC-Authentication-Results:i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=gmail.com; spf=pass smtp.mailfrom=gmail.com; dkim=pass (2048-bit key) header.d=gmail.com header.i=@gmail.com header.b=SK4vJ8ob; arc=none smtp.client-ip=209.85.167.47 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=gmail.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=gmail.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=gmail.com header.i=@gmail.com header.b="SK4vJ8ob" Received: by mail-lf1-f47.google.com with SMTP id 2adb3069b0e04-59b76c092acso5084648e87.2 for ; Tue, 13 Jan 2026 09:58:49 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20230601; t=1768327127; x=1768931927; darn=vger.kernel.org; h=in-reply-to:content-transfer-encoding:content-disposition :mime-version:references:message-id:subject:cc:to:date:from:from:to :cc:subject:date:message-id:reply-to; bh=M8BgzU01XtACBvJTKvVspgO8XlmqsytHJp8hNt7BvL8=; b=SK4vJ8obY9g8YiocB7C4oOyRnd+IJP1q4p34roX7dtnXrBsoklYn2Tp9IKAnbGZVss ift4a9HKgX2l3TvQV+E/RP9JSyXwFkWAnRKjAQ0yVm7IOvFbCjPsyeUaJsExNJ0V7UcV ZTmn1RC0IuIfv2PG/zTwJTGhD2zMYDQR+x6+Tj9ROt15MD6oX9qSQiAY9ZKxtmjgHgDr xU3qHNMdD/krqDjGsR2/QthQ2ol+a4YwJUGJP/xKCFt6A5bBqMSuP8JMuNriqv2HLcCs E17Knd4sKcRJQ69Pjg5JrQEJ3LrgM3JsRD7CiyCgSg2xqNYuNGThK87I6CjjTMD/KnO8 0EJQ== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1768327127; x=1768931927; h=in-reply-to:content-transfer-encoding:content-disposition :mime-version:references:message-id:subject:cc:to:date:from:x-gm-gg :x-gm-message-state:from:to:cc:subject:date:message-id:reply-to; bh=M8BgzU01XtACBvJTKvVspgO8XlmqsytHJp8hNt7BvL8=; b=mRtXzVtCrrdGUI9pQBtHBn1XgWzD1wzvdIR4NW8DoKrNpvtBkR6mIXoUZnjSH2eUcT 8wKOWc9cgH0srjPIgNVi7eqvIMFIwjb3ehc8GLNvG76A1A3w9sKs6hrNOtUXpcnniTYI 51WGvt0nTODxKmb6LZEYDHpYhEGQg/DAMVpqdnYAB6EVok0B40T7ozk9T+AfmkLkeLio PsV8m5Q4RexhGVth7i3F3hrjjgsSkqEDY38rFBaLJmvTXBihDlO0TfGC4pz5kOQd35eG R3BEN1xYv9nDNTzEBDdc3bhrCyV0tPaQbaIbL0Ywv2NWwNtJ92AHPNgUOX8isi/ujU4w TlnA== X-Forwarded-Encrypted: i=1; AJvYcCX+finFO/lrRBdOruMqZYnmQRLl1bj54EOq9WDuJmu8Oy5FTy/9hJGaQ923cl83wGm7gKORxtHhCEy7YVc=@vger.kernel.org X-Gm-Message-State: AOJu0YwnSNmTGamKYI+Xs302ZwUwUf4kL6GmfFQnF02V2DeY7NUl+tP2 YW7VcgunY0JfHl4COpHu38jiWlIXDZaFZeNjwgVWNATedqTNNAZLyD5eOAOTPcuC X-Gm-Gg: AY/fxX53NaHRyZ7lhvENx/VwjlA+iV5fIzh1yoIgRbKeSPvBByxQJTo2Wh2fnnhmP8l 6FxeDpgbw2BELuZpCAFrB+ulFUbYJGZRzoOaZnZ1C9/PwsDS2RCvFjSenGzAgDFW2uUgGZDahLa //0uGUnAywodrO43S3v517plxW9lo/gGJfwlvPxR4/yz9wgtL8F32XOObrcaCsaf5AKgazTr/qm WtkgJVL0Uj/XS3+SbnRYQe9Zh0yofjQ26ZLAMDYVsd3Z3TAHP5Vegf2Cjo2LAbd2BgGLzlptANP /0dX4/cAixG6SP4rNBuX83sLGELCLkp8JV+sC4MzPKOaqhO1ZHHDfSMf00Vw1wJP37UHQAAtUtk IMoFQA6ubKMHARlRCnyCeJXe0ElOrR7RlPxkMR17kv2G2imqzE0Edv8c8nIFK4nY= X-Google-Smtp-Source: AGHT+IG7EM3OYRW75EDEohbswFl7XGuAv2nqrJpiyoCT2CD01QnRPvHtvnnWGlwCM1AIRpdqGPHRIA== X-Received: by 2002:a2e:b8c9:0:b0:383:5482:b858 with SMTP id 38308e7fff4ca-3835482cc71mr9110701fa.34.1768327127086; Tue, 13 Jan 2026 09:58:47 -0800 (PST) Received: from milan ([2001:9b1:d5a0:a500::24b]) by smtp.gmail.com with ESMTPSA id 38308e7fff4ca-382eb3a053esm47554141fa.5.2026.01.13.09.58.46 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Tue, 13 Jan 2026 09:58:46 -0800 (PST) From: Uladzislau Rezki X-Google-Original-From: Uladzislau Rezki Date: Tue, 13 Jan 2026 18:58:44 +0100 To: Joel Fernandes Cc: Uladzislau Rezki , Shrikanth Hegde , Vishal Chourasia , "rcu@vger.kernel.org" , "linux-kernel@vger.kernel.org" , "paulmck@kernel.org" , "frederic@kernel.org" , "neeraj.upadhyay@kernel.org" , "josh@joshtriplett.org" , "boqun.feng@gmail.com" , "rostedt@goodmis.org" , "tglx@linutronix.de" , "peterz@infradead.org" , "srikar@linux.ibm.com" Subject: Re: [PATCH] cpuhp: Expedite synchronize_rcu during CPU hotplug operations Message-ID: References: <5a2b00f2-5e73-4c89-89b5-1a69cb8a7fa2@linux.ibm.com> <91138C31-EF47-4CA6-BD9F-A41981F543EE@nvidia.com> <6d05f9ea-fc4f-4115-a416-8e779f17e0fb@nvidia.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Disposition: inline Content-Transfer-Encoding: 8bit In-Reply-To: <6d05f9ea-fc4f-4115-a416-8e779f17e0fb@nvidia.com> On Tue, Jan 13, 2026 at 09:32:13AM -0500, Joel Fernandes wrote: > > > On 1/13/2026 9:17 AM, Uladzislau Rezki wrote: > > On Tue, Jan 13, 2026 at 12:44:10PM +0000, Joel Fernandes wrote: > >> > >> > >>> On Jan 13, 2026, at 7:19 AM, Uladzislau Rezki wrote: > >>> > >>> On Mon, Jan 12, 2026 at 05:36:24PM +0000, Joel Fernandes wrote: > >>>> > >>>> > >>>>>> On Jan 12, 2026, at 12:09 PM, Uladzislau Rezki wrote: > >>>>> > >>>>> On Mon, Jan 12, 2026 at 04:09:49PM +0000, Joel Fernandes wrote: > >>>>>> > >>>>>> > >>>>>>>> On Jan 12, 2026, at 7:57 AM, Uladzislau Rezki wrote: > >>>>>>> > >>>>>>> Hello, Shrikanth! > >>>>>>> > >>>>>>>> > >>>>>>>>> On 1/12/26 3:38 PM, Uladzislau Rezki wrote: > >>>>>>>>> On Mon, Jan 12, 2026 at 03:13:33PM +0530, Vishal Chourasia wrote: > >>>>>>>>>> Bulk CPU hotplug operations—such as switching SMT modes across all > >>>>>>>>>> cores—require hotplugging multiple CPUs in rapid succession. On large > >>>>>>>>>> systems, this process takes significant time, increasing as the number > >>>>>>>>>> of CPUs grows, leading to substantial delays on high-core-count > >>>>>>>>>> machines. Analysis [1] reveals that the majority of this time is spent > >>>>>>>>>> waiting for synchronize_rcu(). > >>>>>>>>>> > >>>>>>>>>> Expedite synchronize_rcu() during the hotplug path to accelerate the > >>>>>>>>>> operation. Since CPU hotplug is a user-initiated administrative task, > >>>>>>>>>> it should complete as quickly as possible. > >>>>>>>>>> > >>>>>>>>>> Performance data on a PPC64 system with 400 CPUs: > >>>>>>>>>> > >>>>>>>>>> + ppc64_cpu --smt=1 (SMT8 to SMT1) > >>>>>>>>>> Before: real 1m14.792s > >>>>>>>>>> After: real 0m03.205s # ~23x improvement > >>>>>>>>>> > >>>>>>>>>> + ppc64_cpu --smt=8 (SMT1 to SMT8) > >>>>>>>>>> Before: real 2m27.695s > >>>>>>>>>> After: real 0m02.510s # ~58x improvement > >>>>>>>>>> > >>>>>>>>>> Above numbers were collected on Linux 6.19.0-rc4-00310-g755bc1335e3b > >>>>>>>>>> > >>>>>>>>>> [1] https://lore.kernel.org/all/5f2ab8a44d685701fe36cdaa8042a1aef215d10d.camel@linux.vnet.ibm.com > >>>>>>>>>> > >>>>>>>>> Also you can try: echo 1 > /sys/module/rcutree/parameters/rcu_normal_wake_from_gp > >>>>>>>>> to speedup regular synchronize_rcu() call. But i am not saying that it would beat > >>>>>>>>> your "expedited switch" improvement. > >>>>>>>>> > >>>>>>>> > >>>>>>>> Hi Uladzislau. > >>>>>>>> > >>>>>>>> Had a discussion on this at LPC, having in kernel solution is likely > >>>>>>>> better than having it in userspace. > >>>>>>>> > >>>>>>>> - Having it in kernel would make it work across all archs. Why should > >>>>>>>> any user wait when one initiates the hotplug. > >>>>>>>> > >>>>>>>> - userspace tools are spread across such as chcpu, ppc64_cpu etc. > >>>>>>>> though internally most do "0/1 > /sys/devices/system/cpu/cpuN/online". > >>>>>>>> We will have to repeat the same in each tool. > >>>>>>>> > >>>>>>>> - There is already /sys/kernel/rcu_expedited which is better if at all > >>>>>>>> we need to fallback to userspace. > >>>>>>>> > >>>>>>> Sounds good to me. I agree it is better to bypass parameters. > >>>>>> > >>>>>> Another way to make it in-kernel would be to make the RCU normal wake from GP optimization enabled for > 16 CPUs by default. > >>>>>> > >>>>>> I was considering this, but I did not bring it up because I did not know that there are large systems that might benefit from it until now. > >>>>>> > >>>>> IMO, we can increase that threshold. 512/1024 is not a problem at all. > >>>>> But as Paul mentioned, we should consider scalability enhancement. From > >>>>> the other hand it is also probably worth to get into the state when we > >>>>> really see them :) > >>>> > >>>> Instead of pegging to number of CPUs, perhaps the optimization should be dynamic? That is, default to it unless synchronize_rcu load is high, default to the sr_normal wake-up optimization. Of course carefully considering all corner cases, adequate testing and all that ;-) > >>>> > >>> Honestly i do not see use cases when we are not up to speed to process > >>> all callbacks in time keeping in mind that it is blocking context call. > >>> > >>> How many of them should be in flight(blocked contexts) to make it starve... :) > >>> According to my last evaluation it was ~64K. > >>> > >>> Note i do not say that it should not be scaled. > >> > >> But you did not test that on large system with 1000s of CPUs right? > >> > > No, no. I do not have access to such systems. > > > >> > >> So the options I see are: either default to always using the optimization, > >> not just for less than 17 CPUs (what you are saying above). Or, do what I said > >> above (safer for system with 1000s of CPUs and less risky). > >> > > You mean introduce threshold and count how many nodes are in queue? > > Yes. > > > To me it sounds not optimal and looks like a temporary solution. > > Not more sub-optimal than the existing 16 CPU hard-coded solution I suppose. > It was trial testing :) Agree we should do something with it. > > > > Long term wise, it is better to split it, i mean to scale. > > But the scalable solution is already there: the !synchronize_rcu_normal path, > right? And splitting the list won't help this use case anyway. > Fair point. > > > > Do you know who can test it on ~1000 CPUs system? So we have some figures. > > I don't have such systems either. The most I can go is ~200+ CPUs. Perhaps the > folks on this thread have such systems as they mentioned 1900+ CPU systems. They > should be happy to test. > > > > > What i have is 256 CPUs system i can test on. > Same boat. ;-) > :) -- Uladzislau Rezki