Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.37 as permitted sender) client-ip=23.128.96.37;
Date:   Mon, 11 Dec 2023 20:02:09 +0800
From:   Tio Zhang <tiozhang@didiglobal.com>
To:     <bigeasy@linutronix.de>, <rostedt@goodmis.org>
CC:     <tglx@linutronix.de>, <mingo@redhat.com>, <peterz@infradead.org>,
        <juri.lelli@redhat.com>, <vincent.guittot@linaro.org>,
        <linux-kernel@vger.kernel.org>, <dietmar.eggemann@arm.com>,
        <rostedt@goodmis.org>, <bsegall@google.com>, <mgorman@suse.de>,
        <bristot@redhat.com>, <vschneid@redhat.com>,
        <zyhtheonly@gmail.com>, <tiozhang@didiglobal.com>,
        <zyhtheonly@yeah.net>
Subject: [PATCH v4] sched/cputime: account ksoftirqd's time on SYSTEM in
 PREEMPT_RT
Message-ID: <20231211120209.GA25877@didi-ThinkCentre-M930t-N000>
Mail-Followup-To: bigeasy@linutronix.de, rostedt@goodmis.org,
        tglx@linutronix.de, mingo@redhat.com, peterz@infradead.org,
        juri.lelli@redhat.com, vincent.guittot@linaro.org,
        linux-kernel@vger.kernel.org, dietmar.eggemann@arm.com,
        bsegall@google.com, mgorman@suse.de, bristot@redhat.com,
        vschneid@redhat.com, zyhtheonly@gmail.com, zyhtheonly@yeah.net
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Disposition: inline
In-Reply-To: <20231208092616.LcwLlOyE@linutronix.de>
User-Agent: Mutt/1.9.4 (2018-02-28)
Precedence: bulk

In PREEMPT_RT kernel, we dont want ksoftirqd's time accounting on SOFTIRQ
since it is available to the scheduler (while it is unpreemptable in
mainline). So we put it into SYSTEM like any other task running in SYSTEM.
With this patch, when ksoftirqd is taking CPU's time, we observe SYSTEM
in /proc/stat would be bigger than before while SOFTIRQ would be less,
which behaves in contract to mainline, but more suitable for PREEMPT_RT.

Signed-off-by: Tio Zhang <tiozhang@didiglobal.com>
---
 kernel/sched/cputime.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/kernel/sched/cputime.c b/kernel/sched/cputime.c
index af7952f12e6c..6685bb46805d 100644
--- a/kernel/sched/cputime.c
+++ b/kernel/sched/cputime.c
@@ -73,7 +73,8 @@ void irqtime_account_irq(struct task_struct *curr, unsigned int offset)
 	 */
 	if (pc & HARDIRQ_MASK)
 		irqtime_account_delta(irqtime, delta, CPUTIME_IRQ);
-	else if ((pc & SOFTIRQ_OFFSET) && curr != this_cpu_ksoftirqd())
+	else if ((pc & SOFTIRQ_OFFSET) &&
+		 (IS_ENABLED(CONFIG_PREEMPT_RT) || curr != this_cpu_ksoftirqd()))
 		irqtime_account_delta(irqtime, delta, CPUTIME_SOFTIRQ);
 }
 
@@ -391,7 +392,7 @@ static void irqtime_account_process_tick(struct task_struct *p, int user_tick,
 
 	cputime -= other;
 
-	if (this_cpu_ksoftirqd() == p) {
+	if (!IS_ENABLED(CONFIG_PREEMPT_RT) && this_cpu_ksoftirqd() == p) {
 		/*
 		 * ksoftirqd time do not get accounted in cpu_softirq_time.
 		 * So, we have to handle it separately here.
-- 
2.17.1