public inbox for linux-pm@vger.kernel.org
 help / color / mirror / Atom feed
From: Frederic Weisbecker <frederic@kernel.org>
To: LKML <linux-kernel@vger.kernel.org>
Cc: Frederic Weisbecker <frederic@kernel.org>,
	"Christophe Leroy (CS GROUP)" <chleroy@kernel.org>,
	"Rafael J. Wysocki" <rafael@kernel.org>,
	Alexander Gordeev <agordeev@linux.ibm.com>,
	Anna-Maria Behnsen <anna-maria@linutronix.de>,
	Ben Segall <bsegall@google.com>,
	Boqun Feng <boqun.feng@gmail.com>,
	Christian Borntraeger <borntraeger@linux.ibm.com>,
	Dietmar Eggemann <dietmar.eggemann@arm.com>,
	Heiko Carstens <hca@linux.ibm.com>,
	Ingo Molnar <mingo@redhat.com>,
	Jan Kiszka <jan.kiszka@siemens.com>,
	Joel Fernandes <joelagnelf@nvidia.com>,
	Juri Lelli <juri.lelli@redhat.com>,
	Kieran Bingham <kbingham@kernel.org>,
	Madhavan Srinivasan <maddy@linux.ibm.com>,
	Mel Gorman <mgorman@suse.de>,
	Michael Ellerman <mpe@ellerman.id.au>,
	Neeraj Upadhyay <neeraj.upadhyay@kernel.org>,
	Nicholas Piggin <npiggin@gmail.com>,
	"Paul E . McKenney" <paulmck@kernel.org>,
	Peter Zijlstra <peterz@infradead.org>,
	Steven Rostedt <rostedt@goodmis.org>,
	Sven Schnelle <svens@linux.ibm.com>,
	Thomas Gleixner <tglx@linutronix.de>,
	Uladzislau Rezki <urezki@gmail.com>,
	Valentin Schneider <vschneid@redhat.com>,
	Vasily Gorbik <gor@linux.ibm.com>,
	Vincent Guittot <vincent.guittot@linaro.org>,
	Viresh Kumar <viresh.kumar@linaro.org>,
	Xin Zhao <jackzxcui1989@163.com>,
	linux-pm@vger.kernel.org, linux-s390@vger.kernel.org,
	linuxppc-dev@lists.ozlabs.org
Subject: [PATCH 15/15] sched/cputime: Handle dyntick-idle steal time correctly
Date: Fri, 16 Jan 2026 15:52:08 +0100	[thread overview]
Message-ID: <20260116145208.87445-16-frederic@kernel.org> (raw)
In-Reply-To: <20260116145208.87445-1-frederic@kernel.org>

The dyntick-idle steal time is currently accounted when the tick
restarts but the stolen idle time is not substracted from the idle time
that was already accounted. This is to avoid observing the idle time
going backward as the dyntick-idle cputime accessors can't reliably know
in advance the stolen idle time.

In order to maintain a forward progressing idle cputime while
substracting idle steal time from it, keep track of the previously
accounted idle stolen time and substract it from _later_ idle cputime
accounting.

Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
---
 include/linux/kernel_stat.h | 1 +
 kernel/sched/cputime.c      | 9 ++++++++-
 2 files changed, 9 insertions(+), 1 deletion(-)

diff --git a/include/linux/kernel_stat.h b/include/linux/kernel_stat.h
index e59916477075..a5b5a25c3cc1 100644
--- a/include/linux/kernel_stat.h
+++ b/include/linux/kernel_stat.h
@@ -39,6 +39,7 @@ struct kernel_cpustat {
 	bool		idle_elapse;
 	seqcount_t	idle_sleeptime_seq;
 	ktime_t		idle_entrytime;
+	u64		idle_steal;
 #endif
 	u64		cpustat[NR_STATS];
 };
diff --git a/kernel/sched/cputime.c b/kernel/sched/cputime.c
index 3dadfaa92b27..749a6ed4d2fa 100644
--- a/kernel/sched/cputime.c
+++ b/kernel/sched/cputime.c
@@ -424,18 +424,25 @@ static void kcpustat_idle_stop(struct kernel_cpustat *kc, ktime_t now)
 {
 	u64 *cpustat = kc->cpustat;
 	ktime_t delta;
+	u64 steal, steal_delta;
 
 	if (!kc->idle_elapse)
 		return;
 
 	delta = ktime_sub(now, kc->idle_entrytime);
+	steal = steal_account_process_time(delta);
 
 	write_seqcount_begin(&kc->idle_sleeptime_seq);
+	steal_delta = min_t(u64, kc->idle_steal, delta);
+	delta -= steal_delta;
+	kc->idle_steal -= steal_delta;
+
 	if (nr_iowait_cpu(smp_processor_id()) > 0)
 		cpustat[CPUTIME_IOWAIT] = ktime_add(cpustat[CPUTIME_IOWAIT], delta);
 	else
 		cpustat[CPUTIME_IDLE] = ktime_add(cpustat[CPUTIME_IDLE], delta);
 
+	kc->idle_steal += steal;
 	kc->idle_entrytime = now;
 	kc->idle_elapse = false;
 	write_seqcount_end(&kc->idle_sleeptime_seq);
@@ -459,7 +466,6 @@ void kcpustat_dyntick_stop(ktime_t now)
 		kc->idle_dyntick = false;
 		irqtime_dyntick_stop();
 		vtime_dyntick_stop();
-		steal_account_process_time(ULONG_MAX);
 	}
 }
 
@@ -507,6 +513,7 @@ static u64 kcpustat_field_dyntick(int cpu, enum cpu_usage_stat idx,
 		if (kc->idle_elapse && compute_delta) {
 			ktime_t delta = ktime_sub(now, kc->idle_entrytime);
 
+			delta -= min_t(u64, kc->idle_steal, (u64)delta);
 			idle = ktime_add(cpustat[idx], delta);
 		} else {
 			idle = cpustat[idx];
-- 
2.51.1


  parent reply	other threads:[~2026-01-16 14:54 UTC|newest]

Thread overview: 44+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-01-16 14:51 [PATCH 00/15] tick/sched: Refactor idle cputime accounting Frederic Weisbecker
2026-01-16 14:51 ` [PATCH 01/15] sched/idle: Handle offlining first in idle loop Frederic Weisbecker
2026-01-19 12:53   ` Peter Zijlstra
2026-01-19 21:04     ` Frederic Weisbecker
2026-01-20  4:26       ` K Prateek Nayak
2026-01-20 14:52         ` Frederic Weisbecker
2026-01-16 14:51 ` [PATCH 02/15] sched/cputime: Remove superfluous and error prone kcpustat_field() parameter Frederic Weisbecker
2026-01-16 14:51 ` [PATCH 03/15] sched/cputime: Correctly support generic vtime idle time Frederic Weisbecker
2026-01-19 13:02   ` Peter Zijlstra
2026-01-19 21:35     ` Frederic Weisbecker
2026-01-16 14:51 ` [PATCH 04/15] powerpc/time: Prepare to stop elapsing in dynticks-idle Frederic Weisbecker
2026-02-25 17:53   ` Christophe Leroy (CS GROUP)
2026-01-16 14:51 ` [PATCH 05/15] s390/time: " Frederic Weisbecker
2026-01-21 12:17   ` Heiko Carstens
2026-01-21 18:04     ` Frederic Weisbecker
2026-01-22 14:40       ` Heiko Carstens
2026-01-27 14:45         ` Frederic Weisbecker
2026-01-16 14:51 ` [PATCH 06/15] tick/sched: Unify idle cputime accounting Frederic Weisbecker
2026-01-19 14:26   ` Peter Zijlstra
2026-01-19 22:00     ` Frederic Weisbecker
2026-01-16 14:52 ` [PATCH 07/15] cpufreq: ondemand: Simplify idle cputime granularity test Frederic Weisbecker
2026-01-19  5:37   ` Viresh Kumar
2026-01-19 12:30   ` Rafael J. Wysocki
2026-01-19 22:06     ` Frederic Weisbecker
2026-01-20 12:32       ` Rafael J. Wysocki
2026-01-20 14:28         ` Frederic Weisbecker
2026-01-16 14:52 ` [PATCH 08/15] tick/sched: Remove nohz disabled special case in cputime fetch Frederic Weisbecker
2026-01-16 14:52 ` [PATCH 09/15] tick/sched: Move dyntick-idle cputime accounting to cputime code Frederic Weisbecker
2026-01-19 14:35   ` Peter Zijlstra
2026-01-19 22:08     ` Frederic Weisbecker
2026-01-16 14:52 ` [PATCH 10/15] tick/sched: Remove unused fields Frederic Weisbecker
2026-01-16 14:52 ` [PATCH 11/15] tick/sched: Account tickless idle cputime only when tick is stopped Frederic Weisbecker
2026-01-16 14:52 ` [PATCH 12/15] tick/sched: Consolidate idle time fetching APIs Frederic Weisbecker
2026-01-16 14:52 ` [PATCH 13/15] sched/cputime: Consolidate get_cpu_[idle|iowait]_time_us() Frederic Weisbecker
2026-01-16 14:52 ` [PATCH 14/15] sched/cputime: Handle idle irqtime gracefully Frederic Weisbecker
2026-01-16 14:52 ` Frederic Weisbecker [this message]
2026-01-16 14:57 ` [PATCH 00/15] tick/sched: Refactor idle cputime accounting Frederic Weisbecker
2026-01-20 12:42   ` Shrikanth Hegde
2026-01-21 16:55     ` Frederic Weisbecker
2026-01-19 14:53 ` Peter Zijlstra
2026-01-19 22:12   ` Frederic Weisbecker
  -- strict thread matches above, loose matches on Subject: below --
2026-02-06 14:22 [PATCH 00/15 v2] " Frederic Weisbecker
2026-02-06 14:22 ` [PATCH 15/15] sched/cputime: Handle dyntick-idle steal time correctly Frederic Weisbecker
2026-03-03 11:17   ` Shrikanth Hegde
2026-03-24 14:53     ` Frederic Weisbecker

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=20260116145208.87445-16-frederic@kernel.org \
    --to=frederic@kernel.org \
    --cc=agordeev@linux.ibm.com \
    --cc=anna-maria@linutronix.de \
    --cc=boqun.feng@gmail.com \
    --cc=borntraeger@linux.ibm.com \
    --cc=bsegall@google.com \
    --cc=chleroy@kernel.org \
    --cc=dietmar.eggemann@arm.com \
    --cc=gor@linux.ibm.com \
    --cc=hca@linux.ibm.com \
    --cc=jackzxcui1989@163.com \
    --cc=jan.kiszka@siemens.com \
    --cc=joelagnelf@nvidia.com \
    --cc=juri.lelli@redhat.com \
    --cc=kbingham@kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-pm@vger.kernel.org \
    --cc=linux-s390@vger.kernel.org \
    --cc=linuxppc-dev@lists.ozlabs.org \
    --cc=maddy@linux.ibm.com \
    --cc=mgorman@suse.de \
    --cc=mingo@redhat.com \
    --cc=mpe@ellerman.id.au \
    --cc=neeraj.upadhyay@kernel.org \
    --cc=npiggin@gmail.com \
    --cc=paulmck@kernel.org \
    --cc=peterz@infradead.org \
    --cc=rafael@kernel.org \
    --cc=rostedt@goodmis.org \
    --cc=svens@linux.ibm.com \
    --cc=tglx@linutronix.de \
    --cc=urezki@gmail.com \
    --cc=vincent.guittot@linaro.org \
    --cc=viresh.kumar@linaro.org \
    --cc=vschneid@redhat.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