From mboxrd@z Thu Jan 1 00:00:00 1970 From: Claude Code Review Bot To: dri-devel-reviews@example.com Subject: Claude review: arm64: dts: qcom: sm6125-xiaomi-laurel-sprout: Enable MDSS and add panel Date: Sun, 22 Mar 2026 03:41:50 +1000 Message-ID: In-Reply-To: <20260320-panel-patches-v7-3-3eaefc4b3878@gmail.com> References: <20260320-panel-patches-v7-0-3eaefc4b3878@gmail.com> <20260320-panel-patches-v7-3-3eaefc4b3878@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 1. **`panel_vci_3p0` voltage mismatch** (lines 99-110 of the DTS): The regu= lator node is named `panel_vci_3p0` and `regulator-name` is `"panel_vci_3p0= "`, implying 3.0V, but the actual voltage is set to **1800000 =C2=B5V (1.8V= )**: ``` panel_vci_3p0: regulator-panel-vci { regulator-name =3D "panel_vci_3p0"; regulator-min-microvolt =3D <1800000>; regulator-max-microvolt =3D <1800000>; ``` This is likely a bug =E2=80=94 either the voltage should be 3000000 (mat= ching the name), or the name should be changed to `panel_vci_1p8`. VCI supp= lies for Samsung AMOLED panels are typically 3.0V. **This should be clarifi= ed/fixed.** 2. **No `vin-supply` on `panel_vci_3p0`**: The `panel_vdd_1p8` regulator ha= s `vin-supply =3D <&vreg_l9a>`, but `panel_vci_3p0` has no `vin-supply`. Th= is may be intentional if it's directly driven by the GPIO, but worth confir= ming. 3. **Extra blank line** (between `&mdss_dsi0_out` and `&mdss_dsi0_phy` bloc= ks): There's a double blank line at the junction. Very minor style nit. 4. **`regulator-boot-on` on both regulators**: This keeps regulators enable= d during boot, preventing screen flicker. Reasonable for a display panel. 5. **`reset-gpios` polarity**: The panel node uses `GPIO_ACTIVE_LOW`, and t= he driver's reset sequence asserts with `gpiod_set_value_cansleep(ctx->rese= t_gpio, 0)` =E2=86=92 deassert, `1` =E2=86=92 assert. With `GPIOD_OUT_HIGH`= at probe (asserting reset =3D holding panel in reset), then the prepare se= quence deasserts, pulses assert, deasserts. This is consistent =E2=80=94 th= e GPIO descriptor layer handles the inversion. Fine. **Summary**: The main actionable item is the `panel_vci_3p0` regulator volt= age =E2=80=94 the name says 3.0V but the value is 1.8V. This should be reso= lved before merging. --- Generated by Claude Code Patch Reviewer