Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Feedback-ID: i8771445c:Fastmail
Date:   Wed, 21 Sep 2022 11:17:48 +0200
From:   Maxime Ripard <maxime@cerno.tech>
To:     Stefan Wahren <stefan.wahren@i2se.com>
Cc:     Daniel Vetter <daniel@ffwll.ch>,
        Florian Fainelli <f.fainelli@gmail.com>,
        David Airlie <airlied@linux.ie>,
        Broadcom internal kernel review list 
        <bcm-kernel-feedback-list@broadcom.com>,
        Michael Turquette <mturquette@baylibre.com>,
        Scott Branden <sbranden@broadcom.com>,
        Stephen Boyd <sboyd@kernel.org>, Emma Anholt <emma@anholt.net>,
        Ray Jui <rjui@broadcom.com>,
        linux-rpi-kernel@lists.infradead.org, linux-clk@vger.kernel.org,
        linux-kernel@vger.kernel.org, Dom Cobley <popcornmix@gmail.com>,
        dri-devel@lists.freedesktop.org,
        linux-arm-kernel@lists.infradead.org
Subject: Re: [PATCH v2 2/7] firmware: raspberrypi: Move the clock IDs to the
 firmware header
Message-ID: <20220921091748.u6qcba6nwclvezzg@houat>
References: <20220815-rpi-fix-4k-60-v2-0-983276b83f62@cerno.tech>
 <20220815-rpi-fix-4k-60-v2-2-983276b83f62@cerno.tech>
 <688777fc-e6af-8cbe-b19d-c531a7a66b32@i2se.com>
MIME-Version: 1.0
Content-Type: multipart/signed; micalg=pgp-sha512;
        protocol="application/pgp-signature"; boundary="itnh57rcj5wjjcou"
Content-Disposition: inline
In-Reply-To: <688777fc-e6af-8cbe-b19d-c531a7a66b32@i2se.com>
Precedence: bulk


--itnh57rcj5wjjcou
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

Hi,

On Tue, Sep 20, 2022 at 06:01:08PM +0200, Stefan Wahren wrote:
> Am 20.09.22 um 14:50 schrieb Maxime Ripard:
> > We'll need the clock IDs in more drivers than just the clock driver from
> > now on, so let's move them in the firmware header.
>=20
> recently as i reviewed the clk-raspberrypi i noticed this, too. But from =
my
> point of view the clock ids should go to include/dt-bindings/clock (like
> bcm2835.h) because these clock ids are actually referenced in the DTS fil=
es
> and we need to make sure they are in sync. AFAIR this would also result in
> change from enum to defines.
>=20
> Sorry, i didn't had the time to send a patch for this.

IMO, we need both, and this enum still belongs in the firmware header.
We have two separate things, the firmware interface and the DT
interface. The kernel is a consumer for both, but the fact that they
match is an implementation detail. It might even change in the future
for all we know.

So having a header to use defines for the clock indices in the DT looks
like a good idea to me, but I think we should keep that enum in the
firmware header as well.

Maxime

--itnh57rcj5wjjcou
Content-Type: application/pgp-signature; name="signature.asc"

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

iHUEABYKAB0WIQRcEzekXsqa64kGDp7j7w1vZxhRxQUCYyrWvAAKCRDj7w1vZxhR
xQb6AP9gsTEDLx4N2ddX7AdtvgY6eQfDUqVGB8S6WhNsSuhe/gD7B0sPmldeC8DV
XTcG9b2xt1s+i4ha7GqV2l+eQU2chAQ=
=Um8H
-----END PGP SIGNATURE-----

--itnh57rcj5wjjcou--