From: Joel Fernandes <joelagnelf@nvidia.com>
To: paulmck@kernel.org
Cc: Uladzislau Rezki <urezki@gmail.com>,
Shrikanth Hegde <sshegde@linux.ibm.com>,
Vishal Chourasia <vishalc@linux.ibm.com>,
"rcu@vger.kernel.org" <rcu@vger.kernel.org>,
"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
"frederic@kernel.org" <frederic@kernel.org>,
"neeraj.upadhyay@kernel.org" <neeraj.upadhyay@kernel.org>,
"josh@joshtriplett.org" <josh@joshtriplett.org>,
"boqun.feng@gmail.com" <boqun.feng@gmail.com>,
"rostedt@goodmis.org" <rostedt@goodmis.org>,
"tglx@linutronix.de" <tglx@linutronix.de>,
"peterz@infradead.org" <peterz@infradead.org>,
"srikar@linux.ibm.com" <srikar@linux.ibm.com>
Subject: Re: [PATCH] cpuhp: Expedite synchronize_rcu during CPU hotplug operations
Date: Mon, 12 Jan 2026 17:24:40 -0500 [thread overview]
Message-ID: <3525b8af-7e14-44e4-a909-ad46a0764d94@nvidia.com> (raw)
In-Reply-To: <05e31c43-02b5-4c3d-8a11-3cb7987cfa0c@paulmck-laptop>
On 1/12/2026 11:48 AM, Paul E. McKenney wrote:
> On Mon, Jan 12, 2026 at 04:09:49PM +0000, Joel Fernandes wrote:
>>
>>
>>> On Jan 12, 2026, at 7:57 AM, Uladzislau Rezki <urezki@gmail.com> 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.>
> This would require increasing the scalability of this optimization,
> right? Or am I thinking of the wrong optimization? ;-)
>
Yes I think you are considering the correct one, the concern you have is
regarding large number of wake ups initiated from the GP thread, correct?
I was suggesting on the thread, a more dynamic approach where using
synchronize_rcu_normal() until it gets overloaded with requests. One approach
might be to measure the length of the rcu_state.srs_next to detect an overload
condition, similar to qhimark? Or perhaps qhimark itself can be used. And under
lightly loaded conditions, default to synchronize_rcu_normal() without checking
for the 16 CPU count.
Thoughts?
thanks,
- Joel
next prev parent reply other threads:[~2026-01-12 22:24 UTC|newest]
Thread overview: 54+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-01-12 9:43 Vishal Chourasia
2026-01-12 10:08 ` Uladzislau Rezki
2026-01-12 10:43 ` Vishal Chourasia
2026-01-12 11:07 ` Uladzislau Rezki
2026-01-12 12:02 ` Shrikanth Hegde
2026-01-12 12:57 ` Uladzislau Rezki
2026-01-12 16:09 ` Joel Fernandes
2026-01-12 16:48 ` Paul E. McKenney
2026-01-12 17:05 ` Uladzislau Rezki
2026-01-12 18:27 ` Vishal Chourasia
2026-01-13 0:03 ` Paul E. McKenney
2026-01-12 22:24 ` Joel Fernandes [this message]
2026-01-13 0:01 ` Paul E. McKenney
2026-01-13 2:46 ` Joel Fernandes
2026-01-13 4:53 ` Shrikanth Hegde
2026-01-13 8:57 ` Joel Fernandes
2026-01-14 4:00 ` Paul E. McKenney
2026-01-14 8:54 ` Joel Fernandes
2026-01-16 19:02 ` Paul E. McKenney
2026-01-14 3:59 ` Paul E. McKenney
2026-01-12 17:09 ` Uladzislau Rezki
2026-01-12 17:36 ` Joel Fernandes
2026-01-13 12:18 ` Uladzislau Rezki
2026-01-13 12:44 ` Joel Fernandes
2026-01-13 14:17 ` Uladzislau Rezki
2026-01-13 14:32 ` Joel Fernandes
2026-01-13 14:53 ` Shrikanth Hegde
2026-01-13 18:17 ` Uladzislau Rezki
2026-01-13 17:58 ` Uladzislau Rezki
2026-01-12 12:21 ` Shrikanth Hegde
2026-01-12 12:46 ` Vishal Chourasia
2026-01-12 14:03 ` Joel Fernandes
2026-01-12 14:20 ` Joel Fernandes
2026-01-12 14:23 ` Peter Zijlstra
2026-01-12 14:37 ` Joel Fernandes
2026-01-12 17:52 ` Vishal Chourasia
2026-01-12 14:24 ` Peter Zijlstra
2026-01-12 18:00 ` Vishal Chourasia
2026-01-13 9:01 ` Peter Zijlstra
2026-01-19 10:47 ` [PATCH] cpuhp: Expedite synchronize_rcu during SMT switch Vishal Chourasia
2026-01-19 11:43 ` Peter Zijlstra
2026-01-19 13:45 ` Shrikanth Hegde
2026-01-19 14:11 ` Peter Zijlstra
2026-01-19 14:45 ` Joel Fernandes
2026-01-19 14:59 ` Peter Zijlstra
2026-01-27 17:48 ` Samir M
2026-01-29 7:05 ` Samir M
2026-02-03 6:31 ` Samir M
2026-01-19 10:54 ` [RESEND] " Vishal Chourasia
2026-01-18 11:38 ` [PATCH] cpuhp: Expedite synchronize_rcu during CPU hotplug operations Samir M
2026-01-19 5:18 ` Joel Fernandes
2026-01-19 13:53 ` Shrikanth Hegde
2026-01-19 21:10 ` joelagnelf
2026-02-02 8:46 ` Vishal Chourasia
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=3525b8af-7e14-44e4-a909-ad46a0764d94@nvidia.com \
--to=joelagnelf@nvidia.com \
--cc=boqun.feng@gmail.com \
--cc=frederic@kernel.org \
--cc=josh@joshtriplett.org \
--cc=linux-kernel@vger.kernel.org \
--cc=neeraj.upadhyay@kernel.org \
--cc=paulmck@kernel.org \
--cc=peterz@infradead.org \
--cc=rcu@vger.kernel.org \
--cc=rostedt@goodmis.org \
--cc=srikar@linux.ibm.com \
--cc=sshegde@linux.ibm.com \
--cc=tglx@linutronix.de \
--cc=urezki@gmail.com \
--cc=vishalc@linux.ibm.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®