From mboxrd@z Thu Jan 1 00:00:00 1970 From: Claude Code Review Bot To: dri-devel-reviews@example.com Subject: Claude review: Re: [PATCH 5/6] arm64: dts: qcom: qcs6490-radxa-dragon-q6a: Enable GPU and display pipeline Date: Mon, 25 May 2026 19:16:37 +1000 Message-ID: In-Reply-To: References: <20260522060645.4399-1-graham.oconnor@gmail.com> X-Mailer: Claude Code Patch Reviewer Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 Patch Review **Issue 1 =E2=80=94 `dp-connector` used as a bridge.** The `hdmi-bridge` no= de uses `compatible =3D "dp-connector"`: ```dts + hdmi-bridge { + compatible =3D "dp-connector"; + label =3D "hdmi"; + type =3D "full-size"; ``` The `dp-connector` binding (Documentation/devicetree/bindings/display/conne= ctor/dp-connector.yaml) represents a *physical DP connector* =E2=80=94 the = port where you plug in a DP cable. The RA620 is not a DP connector; it's a = DP-to-HDMI bridge. Using `dp-connector` to model a bridge is semantically i= ncorrect. Furthermore, the `dp-connector` binding defines port@1 as "Connection to co= ntroller providing AUX signals." This patch uses port@1 as an output to the= HDMI connector, which doesn't match the binding: ```dts + port@1 { + reg =3D <1>; + hdmi_bridge_out: endpoint { + remote-endpoint =3D <&hdmi_connector_in>; + }; + }; ``` **Issue 2 =E2=80=94 Naming confusion.** The node is called `hdmi-bridge` bu= t uses `dp-connector` compatible. The label is "hdmi" but it's a DP endpoin= t. This will confuse future readers. **Issue 3 =E2=80=94 pinctrl on "connector" node.** The `dp_hot_plug_det` pi= nctrl is attached to the `hdmi-bridge` (dp-connector) node. HPD pinctrl wou= ld more typically belong on the DP controller node (`&mdss_dp`). The GPU enable parts (`gpucc`, `gmu`, `dispcc`, `mdss` status =3D "okay") a= nd the disabling of DSI/eDP interfaces are all fine. --- --- Generated by Claude Code Patch Reviewer