Received-SPF: pass (google.com: domain of linux-kernel+bounces-200031-linux.lists.archive=gmail.com@vger.kernel.org designates 139.178.88.99 as permitted sender) client-ip=139.178.88.99;
Precedence: bulk
Mime-Version: 1.0
Content-Transfer-Encoding: quoted-printable
Content-Type: text/plain; charset=UTF-8
Date: Tue, 04 Jun 2024 15:08:20 +1000
Message-Id: <D1QZ17GX9Z9V.3P8J6D3MDU0T2@gmail.com>
From: "Nicholas Piggin" <npiggin@gmail.com>
To: "Oleg Nesterov" <oleg@redhat.com>, "Frederic Weisbecker"
 <frederic@kernel.org>, "Ingo Molnar" <mingo@redhat.com>, "Peter Zijlstra"
 <peterz@infradead.org>, "Phil Auld" <pauld@redhat.com>, "Thomas Gleixner"
 <tglx@linutronix.de>
Cc: "Chris von Recklinghausen" <crecklin@redhat.com>,
 <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH v2] tick/nohz_full: turn tick_do_timer_boot_cpu into
 boot_cpu_is_nohz_full
References: <20240522151742.GA10400@redhat.com>
 <20240528122019.GA28794@redhat.com> <20240530124032.GA26833@redhat.com>
 <20240603153557.GA8311@redhat.com>
In-Reply-To: <20240603153557.GA8311@redhat.com>

On Tue Jun 4, 2024 at 1:35 AM AEST, Oleg Nesterov wrote:
> We don't need to record the cpu number of the nohz_full boot CPU, a simpl=
e
> boolean is enough. We could even kill it, the "else if" branch could chec=
k
> tick_nohz_full_cpu(tick_do_timer_cpu) && !tick_nohz_full_cpu(cpu).
>
> Signed-off-by: Oleg Nesterov <oleg@redhat.com>

Reviewed-by: Nicholas Piggin <npiggin@gmail.com>

> ---
>  kernel/time/tick-common.c | 21 ++++-----------------
>  1 file changed, 4 insertions(+), 17 deletions(-)
>
> diff --git a/kernel/time/tick-common.c b/kernel/time/tick-common.c
> index 27d0018c8b05..82c1b2206631 100644
> --- a/kernel/time/tick-common.c
> +++ b/kernel/time/tick-common.c
> @@ -49,15 +49,6 @@ ktime_t tick_next_period;
>   *    procedure also covers cpu hotplug.
>   */
>  int tick_do_timer_cpu __read_mostly =3D TICK_DO_TIMER_BOOT;
> -#ifdef CONFIG_NO_HZ_FULL
> -/*
> - * tick_do_timer_boot_cpu indicates the boot CPU temporarily owns
> - * tick_do_timer_cpu and it should be taken over by an eligible secondar=
y
> - * when one comes online.
> - */
> -static int tick_do_timer_boot_cpu __read_mostly =3D -1;
> -#endif
> -
>  /*
>   * Debugging: see timer_list.c
>   */
> @@ -192,6 +183,7 @@ static void tick_setup_device(struct tick_device *td,
>  	 * First device setup ?
>  	 */
>  	if (!td->evtdev) {
> +		static bool boot_cpu_is_nohz_full __read_mostly;
>  		/*
>  		 * If no cpu took the do_timer update, assign it to
>  		 * this cpu:
> @@ -199,18 +191,14 @@ static void tick_setup_device(struct tick_device *t=
d,
>  		if (READ_ONCE(tick_do_timer_cpu) =3D=3D TICK_DO_TIMER_BOOT) {
>  			WRITE_ONCE(tick_do_timer_cpu, cpu);
>  			tick_next_period =3D ktime_get();
> -#ifdef CONFIG_NO_HZ_FULL
>  			/*
>  			 * The boot CPU may be nohz_full, in which case the
>  			 * first housekeeping secondary will take do_timer()
>  			 * from us.
>  			 */
> -			if (tick_nohz_full_cpu(cpu))
> -				tick_do_timer_boot_cpu =3D cpu;
> -
> -		} else if (tick_do_timer_boot_cpu !=3D -1 &&
> -						!tick_nohz_full_cpu(cpu)) {
> -			tick_do_timer_boot_cpu =3D -1;
> +			boot_cpu_is_nohz_full =3D tick_nohz_full_cpu(cpu);
> +		} else if (boot_cpu_is_nohz_full && !tick_nohz_full_cpu(cpu)) {
> +			boot_cpu_is_nohz_full =3D false;
>  			/*
>  			 * The boot CPU will stay in periodic (NOHZ disabled)
>  			 * mode until clocksource_done_booting() called after
> @@ -221,7 +209,6 @@ static void tick_setup_device(struct tick_device *td,
>  			 * check in tick_periodic() but this race is harmless.
>  			 */
>  			WRITE_ONCE(tick_do_timer_cpu, cpu);
> -#endif
>  		}
> =20
>  		/*