Received-SPF: pass (google.com: domain of linux-kernel+bounces-69259-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.48.161 as permitted sender) client-ip=147.75.48.161;
Date: Fri, 16 Feb 2024 13:44:40 -0600
From: David Vernet <void@manifault.com>
To: peterz@infradead.org
Cc: mingo@redhat.com, linux-kernel@vger.kernel.org, juri.lelli@redhat.com,
	vincent.guittot@linaro.org, dietmar.eggemann@arm.com,
	bsegall@google.com, bristot@redhat.com, vschneid@redhat.com,
	kernel-team@meta.com
Subject: Re: [PATCH v3 0/3] sched/fair: Simplify and optimize
 update_sd_pick_busiest()
Message-ID: <20240216194440.GA1682@maniforge>
References: <20240206043921.850302-1-void@manifault.com>
Precedence: bulk
MIME-Version: 1.0
Content-Type: multipart/signed; micalg=pgp-sha512;
	protocol="application/pgp-signature"; boundary="7AqvWcatbl91sTR2"
Content-Disposition: inline
In-Reply-To: <20240206043921.850302-1-void@manifault.com>
User-Agent: Mutt/2.2.12 (2023-09-09)


--7AqvWcatbl91sTR2
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Mon, Feb 05, 2024 at 10:39:18PM -0600, David Vernet wrote:
> update_sd_pick_busiest() (and its caller) has some room for small
> optimizations, and some improvements in readability.

Hello Peter, hello Ingo,

Friendly ping. Is there anything else required for this to land?

Thanks,
David

>=20
> - In update_sd_lb_stats(), we're using a goto to skip a single if check.
>   Let's remove the goto and just add another condition to the if.
>=20
> - In update_sd_pick_busiest(), only update a group_misfit_task group to
>   be the busiest if it has strictly more load than the current busiest
>   task, rather than >=3D the load.
>=20
> - When comparing the current struct sched_group with the yet-busiest
>   domain in update_sd_pick_busiest(), if the two groups have the same
>   group type, we're currently doing a bit of unnecessary work for any
>   group >=3D group_misfit_task. We're comparing the two groups, and then
>   returning only if false (the group in question is not the busiest).
>   Othewise, we break, do an extra unnecessary conditional check that's
>   vacuously false for any group type > group_fully_busy, and then always
>   return true. This patch series has us instead simply return directly
>   in the switch statement, saving some bytes in load_balance().
>=20
> ---
>=20
> v1: https://lore.kernel.org/lkml/20240202070216.2238392-1-void@manifault.=
com/
> v2: https://lore.kernel.org/all/20240204044618.46100-1-void@manifault.com/
>=20
> v2 -> v3:
> - Add Vincent's Reviewed-by tags
> - Fix stale commit summary sentence (Vincent)
>=20
> v1 -> v2 changes:
>=20
> - Split the patch series into separate patches (Valentin)
> - Update the group_misfit_task busiest check to use strict inequality
>=20
> David Vernet (3):
>   sched/fair: Remove unnecessary goto in update_sd_lb_stats()
>   sched/fair: Do strict inequality check for busiest misfit task group
>   sched/fair: Simplify some logic in update_sd_pick_busiest()
>=20
>  kernel/sched/fair.c | 19 ++++---------------
>  1 file changed, 4 insertions(+), 15 deletions(-)
>=20
> --=20
> 2.43.0
>=20

--7AqvWcatbl91sTR2
Content-Type: application/pgp-signature; name="signature.asc"

-----BEGIN PGP SIGNATURE-----

iHUEARYKAB0WIQRBxU1So5MTLwphjdFZ5LhpZcTzZAUCZc+7KAAKCRBZ5LhpZcTz
ZCswAQD/TPyQ1RCxEAi6CiJf6doJrL7cgg+pRq1u1kUhHr2NZAEAyzfgAeHbVCpl
SI61fJRXdrf9C8AK62SzYdfVuL+W6wk=
=3n3j
-----END PGP SIGNATURE-----

--7AqvWcatbl91sTR2--