Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
MIME-Version: 1.0
References: <20210521200038.v14.1.I9f45f5c1f975422d58b5904d11546349e9ccdc94@changeid>
In-Reply-To: <20210521200038.v14.1.I9f45f5c1f975422d58b5904d11546349e9ccdc94@changeid>
From:   Chen-Yu Tsai <wenst@chromium.org>
Date:   Fri, 21 May 2021 22:45:31 +0800
Message-ID: <CAGXv+5HX+Hw-S=8SQOy5hjMUxFM-oBFFESzh3LQTEVpy=54Aeg@mail.gmail.com>
Subject: Re: [PATCH v14] arm64: dts: mt8183: Add node for the Mali GPU
To:     Nicolas Boichat <drinkcat@chromium.org>
Cc:     Matthias Brugger <matthias.bgg@gmail.com>,
        Rob Herring <robh@kernel.org>,
        Ezequiel Garcia <ezequiel@collabora.com>,
        Steven Price <steven.price@arm.com>,
        Neil Armstrong <narmstrong@baylibre.com>,
        hoegsberg@chromium.org, Tomeu Vizoso <tomeu.vizoso@collabora.com>,
        boris.brezillon@collabora.com,
        Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>,
        fshao@chromium.org, hsinyi@chromium.org,
        Rob Herring <robh+dt@kernel.org>, devicetree@vger.kernel.org,
        linux-arm-kernel@lists.infradead.org, linux-kernel@vger.kernel.org,
        linux-mediatek@lists.infradead.org
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

(Resent to all recipients)

On Fri, May 21, 2021 at 8:01 PM Nicolas Boichat <drinkcat@chromium.org> wrote:
>
> Add a basic GPU node for mt8183, as well as OPP table.
>
> Note that with the current panfrost driver, devfreq is not
> actually functional, as the we do not have platform-specific
> support for >1 supplies. Also, we are missing code to handle
> frequency change, as the GPU frequency needs to be switched
> away to a stable 26Mhz clock during the transition.
>
> Signed-off-by: Nicolas Boichat <drinkcat@chromium.org>
> ---
> The binding we use with out-of-tree Mali drivers includes more
> clocks, this is used for devfreq: the out-of-tree driver switches
> clk_mux to clk_sub_parent (26Mhz), adjusts clk_main_parent, then
> switches clk_mux back to clk_main_parent:
> (see https://chromium.googlesource.com/chromiumos/third_party/kernel/+/chromeos-4.19/drivers/gpu/arm/midgard/platform/mediatek/mali_kbase_runtime_pm.c#423)
> clocks =
>         <&topckgen CLK_TOP_MFGPLL_CK>,
>         <&topckgen CLK_TOP_MUX_MFG>,
>         <&clk26m>,
>         <&mfgcfg CLK_MFG_BG3D>;
> clock-names =
>         "clk_main_parent",
>         "clk_mux",
>         "clk_sub_parent",
>         "subsys_mfg_cg";
> (based on discussions, this probably belongs in the clock core)
>
> This only matters for devfreq, that is disabled anyway as we don't
> have platform-specific code to handle >1 supplies.

Yes. This can be done with clk_notifiers. See

    https://elixir.bootlin.com/linux/latest/source/drivers/clk/sunxi-ng/ccu_mux.c#L261

for such an example.


ChenYu