Received-SPF: pass (google.com: domain of linux-kernel+bounces-86041-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.80.249 as permitted sender) client-ip=147.75.80.249;
Message-ID: <1d0baf6dbaa1c2ca6594f9a2bcade2c4.sboyd@kernel.org>
Content-Type: text/plain; charset="utf-8"
Precedence: bulk
MIME-Version: 1.0
Content-Transfer-Encoding: quoted-printable
In-Reply-To: <ZdcFuV0KQDXTH8L8@pluto>
References: <20240214183006.3403207-1-cristian.marussi@arm.com> <20240214183006.3403207-7-cristian.marussi@arm.com> <500e265eb7c6a03a40e0067c8806e059.sboyd@kernel.org> <ZdcFuV0KQDXTH8L8@pluto>
Subject: Re: [PATCH 6/7] clk: scmi: Allocate CLK operations dynamically
From: Stephen Boyd <sboyd@kernel.org>
Cc: linux-arm-kernel@lists.infradead.org, linux-kernel@vger.kernel.org, sudeep.holla@arm.com, james.quinlan@broadcom.com, f.fainelli@gmail.com, vincent.guittot@linaro.org, peng.fan@oss.nxp.com, michal.simek@amd.com, quic_sibis@quicinc.com, quic_nkela@quicinc.com, souvik.chakravarty@arm.com, Michael Turquette <mturquette@baylibre.com>, linux-clk@vger.kernel.org
To: Cristian Marussi <cristian.marussi@arm.com>
Date: Wed, 28 Feb 2024 18:20:34 -0800
User-Agent: alot/0.10

Quoting Cristian Marussi (2024-02-22 00:28:41)
> On Wed, Feb 21, 2024 at 09:44:14PM -0800, Stephen Boyd wrote:
> >=20
> > It's not great to move these function pointer structs out of RO memory
> > to RW. I'm also not convinced that it's any better to construct them at
> > runtime. Isn't there a constant set of possible clk configurations? Or
> > why can't we simply add some failures to the clk_ops functions instead?
>=20
> Well, the real clock devices managed by the SCMI server can be a of

SCMI is a server!? :)

> varying nature and so the minimum set of possible clk configurations
> to cover will amount to all the possible combinations of supported ops
> regarding the specific clock properties (i.e. .set_parent / .set_rate /
> .enable / .get/set_duty_cycle / atomic_capability ... for now)...we
> simply cannot know in advance what the backend SCMI server is handling.
>=20
> These seemed to me too much in number (and growing) to be pre-allocated
> in all possible combinations. (and mostly wasted since you dont really
> probably use all combinations all the time)
>=20
> Moreover, SCMI latest spec now exposes some clock properties (or not) to
> be able avoid even sending an actual SCMI message that we know will be
> denied all the time; one option is that we return an error,, as you said,
> but what is the point (I thought) to provide at all a clk-callback that
> we know upfront will fail to be executed every time ? (and some consumer
> drivers have been reported by partners not to be happy with these errors)
>=20
> What I think could be optimized here instead, and I will try in the next
> respin, it is that now I am allocating one set of custom ops for each clo=
ck
> at the end, even if exactly the same ops are provided since the clock
> capabilities are the same; I could instead allocate dynamically and fill =
only
> one single set of ops for each distinct set of combinations, so as to avo=
id
> useless duplication and use only the miminum strict amount of RW memory
> needed.
>=20

Yes please don't allocate a clk_op per clk. And, please add these
answers to the commit text so that we know why it's not possible to know
all combinations or fail clk_ops calls.