archive mirror
 help / color / mirror / Atom feed
From: Marcelo Tosatti <>
To: Oscar Shiang <>
Cc:, Nitesh Lal <>,
	Nicolas Saenz Julienne <>,
	Frederic Weisbecker <>,
	Christoph Lameter <>,
	Juri Lelli <>,
	Peter Zijlstra <>,
	Alex Belits <>, Peter Xu <>,
	Thomas Gleixner <>,
	Daniel Bristot de Oliveira <>
Subject: Re: [patch v11 00/13] extensible prctl task isolation interface and vmstat sync
Date: Wed, 23 Feb 2022 14:31:03 -0300	[thread overview]
Message-ID: <YhZvV5rRkHw4nOlR@fuller.cnet> (raw)
In-Reply-To: <TYCP286MB19133894C3D15F11A9DB62A9A1389@TYCP286MB1913.JPNP286.PROD.OUTLOOK.COM>

Hi Oscar,

On Sat, Feb 19, 2022 at 04:02:10PM +0800, Oscar Shiang wrote:
> Hi Marcelo,
> I tried to apply your patches to kernel v5.15.18-rt28 and measured
> the latencies through oslat [1].
> It turns out that the peak latency (around 100us) can drop to about 90us.
> The result is impressive since I only changed the guest's kernel
> instead of installing the patched kernel to both host and guest.
> However, I am still curious about:
> 1) Why did I catch a bigger maximum latency in almost each of the
>    results of applying task isolation patches? Or does it come from
>    other reasons?

There are a number of things that need to be done in order to have an 
"well enough" isolated CPU so you can measure latency reliably:

* Boot a kernel with isolated CPU (or better, use realtime-virtual-host profile of, which does a bunch of
other things to avoid interruptions to isolated CPUs).
* Apply the userspace patches at
to util-linux and rt-tests.

Run oslat with chisol:

chisol -q vmstat_sync -I conf oslat -c ...

Where chisol is from patched util-linux and oslat from patched rt-tests.

If you had "-f 1" (FIFO priority), on oslat, then the vmstat work would be hung.

Are you doing those things?

> 2) Why did we only get a 10us improvement on quiescing vmstat?

If you did not have FIFO priority on oslat, then other daemons 
could be interrupting it, so better make sure the 10us improvement 
you see is due to vmstat_flush workqueue work not executing anymore.

The testcase i use is: 

Stock kernel:

terminal 1: 
# oslat -f 1 -c X ...

terminal 2:
# echo 1 > /proc/sys/vm/stat_refresh

Patched kernel:

terminal 1: 
# chisol -q vmstat_sync -I conf oslat -f 1 -c X ...

terminal 2:
# echo 1 > /proc/sys/vm/stat_refresh

> [1]: The result and the test scripts I used can be found at

OK, you seem to be doing everything necessary for chisol 
to work. Does /proc/pid/task_isolation of the oslat worker thread
(note its not the same pid as the main oslat thread) show "vmstat"
configured and activated for quiesce?

However 100us is really high. You should be able to get < 10us with
realtime-virtual-host (i see 4us on an idle system).

The answer might be: because 10us is what it takes to execute
vmstat_worker on the isolated CPU (you can verify with tracepoints).

That time depends on the number of per-CPU vmstat variables that need flushing, 
i suppose...

  reply	other threads:[~2022-02-23 17:48 UTC|newest]

Thread overview: 30+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-02-04 17:35 [patch v11 00/13] extensible prctl task isolation interface and vmstat sync Marcelo Tosatti
2022-02-04 17:35 ` [patch v11 01/13] s390: add support for TIF_TASK_ISOL Marcelo Tosatti
2022-02-07 11:40   ` Sven Schnelle
2022-02-04 17:35 ` [patch v11 02/13] x86: " Marcelo Tosatti
2022-02-04 17:35 ` [patch v11 03/13] add basic task isolation prctl interface Marcelo Tosatti
2022-02-04 17:35 ` [patch v11 04/13] add prctl task isolation prctl docs and samples Marcelo Tosatti
2022-02-04 17:35 ` [patch v11 05/13] task isolation: sync vmstats on return to userspace Marcelo Tosatti
2022-02-07 14:57   ` Frederic Weisbecker
2022-02-07 15:16     ` Marcelo Tosatti
2022-02-04 17:35 ` [patch v11 06/13] procfs: add per-pid task isolation state Marcelo Tosatti
2022-02-04 17:35 ` [patch v11 07/13] task isolation: sync vmstats conditional on changes Marcelo Tosatti
2022-02-07 15:19   ` Frederic Weisbecker
2022-03-10 16:30     ` Marcelo Tosatti
2022-02-07 15:29   ` Frederic Weisbecker
2022-02-04 17:35 ` [patch v11 08/13] task isolation: enable return to userspace processing Marcelo Tosatti
2022-02-04 17:35 ` [patch v11 09/13] task isolation: add preempt notifier to sync per-CPU vmstat dirty info to thread info Marcelo Tosatti
2022-02-07 15:36   ` Frederic Weisbecker
2022-02-04 17:35 ` [patch v11 10/13] KVM: x86: process isolation work from VM-entry code path Marcelo Tosatti
2022-02-07 15:47   ` Frederic Weisbecker
2022-03-10 16:35     ` Marcelo Tosatti
2022-02-04 17:35 ` [patch v11 11/13] mm: vmstat: move need_update Marcelo Tosatti
2022-02-04 17:35 ` [patch v11 12/13] mm: vmstat_refresh: avoid queueing work item if cpu stats are clean Marcelo Tosatti
2022-02-04 17:35 ` [patch v11 13/13] task isolation: only TIF_TASK_ISOL if task isolation is enabled Marcelo Tosatti
2022-02-07 15:38   ` Frederic Weisbecker
2022-02-19  8:02 ` [patch v11 00/13] extensible prctl task isolation interface and vmstat sync Oscar Shiang
2022-02-23 17:31   ` Marcelo Tosatti [this message]
2022-03-08  6:32     ` Oscar Shiang
2022-03-08 13:12       ` Marcelo Tosatti
2022-03-09 15:31         ` Oscar Shiang
2022-03-08  7:20 ` Oscar Shiang

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:

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=YhZvV5rRkHw4nOlR@fuller.cnet \ \ \ \ \ \ \ \ \ \ \ \ \ \

* 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;
as well as URLs for NNTP newsgroup(s).