Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Date:   Fri, 1 Nov 2019 19:26:15 -0200
From:   Marcelo Tosatti <mtosatti@redhat.com>
To:     Zhenzhong Duan <zhenzhong.duan@oracle.com>
Cc:     linux-kernel@vger.kernel.org, kvm@vger.kernel.org,
        joao.m.martins@oracle.com, rafael.j.wysocki@intel.com,
        rkrcmar@redhat.com, pbonzini@redhat.com
Subject: Re: [PATCH 5/5] cpuidle-haltpoll: fix up the branch check
Message-ID: <20191101212613.GB20672@amt.cnet>
References: <1572060239-17401-1-git-send-email-zhenzhong.duan@oracle.com>
 <1572060239-17401-6-git-send-email-zhenzhong.duan@oracle.com>
MIME-Version: 1.0
In-Reply-To: <1572060239-17401-6-git-send-email-zhenzhong.duan@oracle.com>
User-Agent: Mutt/1.5.21 (2010-09-15)
Content-Type: text/plain; charset=WINDOWS-1252
Content-Transfer-Encoding: quoted-printable
Content-Disposition: inline
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Sat, Oct 26, 2019 at 11:23:59AM +0800, Zhenzhong Duan wrote:
> Ensure pool time is longer than block_ns, so there is a margin to
> avoid vCPU get into block state unnecessorily.
>=20
> Signed-off-by: Zhenzhong Duan <zhenzhong.duan@oracle.com>
> ---
>  drivers/cpuidle/governors/haltpoll.c | 6 +++---
>  1 file changed, 3 insertions(+), 3 deletions(-)
>=20
> diff --git a/drivers/cpuidle/governors/haltpoll.c b/drivers/cpuidle/gover=
nors/haltpoll.c
> index 4b00d7a..59eadaf 100644
> --- a/drivers/cpuidle/governors/haltpoll.c
> +++ b/drivers/cpuidle/governors/haltpoll.c
> @@ -81,9 +81,9 @@ static void adjust_poll_limit(struct cpuidle_device *de=
v, unsigned int block_us)
>  =09u64 block_ns =3D block_us*NSEC_PER_USEC;
> =20
>  =09/* Grow cpu_halt_poll_us if
> -=09 * cpu_halt_poll_us < block_ns < guest_halt_poll_us
> +=09 * cpu_halt_poll_us <=3D block_ns < guest_halt_poll_us
>  =09 */
> -=09if (block_ns > dev->poll_limit_ns && block_ns <=3D guest_halt_poll_ns=
) {
> +=09if (block_ns >=3D dev->poll_limit_ns && block_ns < guest_halt_poll_ns=
) {
=09=09=09=09=09      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^

If block_ns =3D=3D guest_halt_poll_ns, you won't allow dev->poll_limit_ns t=
o
grow. Why is that?

> @@ -101,7 +101,7 @@ static void adjust_poll_limit(struct cpuidle_device *=
dev, unsigned int block_us)
>  =09=09=09val =3D guest_halt_poll_ns;
> =20
>  =09=09dev->poll_limit_ns =3D val;
> -=09} else if (block_ns > guest_halt_poll_ns &&
> +=09} else if (block_ns >=3D guest_halt_poll_ns &&
>  =09=09   guest_halt_poll_allow_shrink) {
>  =09=09unsigned int shrink =3D guest_halt_poll_shrink;

And here you shrink if block_ns =3D=3D guest_halt_poll_ns. Not sure
why that makes sense either.