| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <d03ca3d6-cb36-4eeb-bbe2-9bd27b647a51@ixit.cz>
Date: Sat, 18 Oct 2025 01:32:11 +0200
From: David Heidelberg <david@...t.cz>
To: Konrad Dybcio <konrad.dybcio@....qualcomm.com>,
Bjorn Andersson <andersson@...nel.org>,
Konrad Dybcio <konradybcio@...nel.org>, Rob Herring <robh@...nel.org>,
Krzysztof Kozlowski <krzk+dt@...nel.org>, Conor Dooley <conor+dt@...nel.org>
Cc: linux-arm-msm@...r.kernel.org, devicetree@...r.kernel.org,
linux-kernel@...r.kernel.org, Amit Pundir <amit.pundir@...aro.org>,
Casey Connolly <casey@...nolly.tech>, Joel Selvaraj <foss@...lselvaraj.com>,
Sumit Semwal <sumit.semwal@...aro.org>, Vinod Koul <vkoul@...nel.org>
Subject: Re: [PATCH 2/2] arm64: dts: qcom: Add support for Pixel 3 and Pixel 3
XL
On 06/10/2025 14:41, Konrad Dybcio wrote:
> On 10/5/25 3:16 PM, David Heidelberg via B4 Relay wrote:
>> From: David Heidelberg <david@...t.cz>
>>
>> This adds initial device tree support for the following phones:
>>
>> - Google Pixel 3 (blueline)
>> - Google Pixel 3 XL (crosshatch)
>>
>> Both phone boards use the same identifiers and differ only slightly
>> in their connected peripherals.
>>
>> Supported functionality includes:
>> - Debug UART
>> - UFS
>> - Charger
>> - USB-C (peripheral mode)
>> - Display (Pixel 3 only)
>>
>> GPIOs 0–3 and 81–84 are not accessible from the application CPUs,
>> so they are marked as reserved to allow the Pixel 3 to boot.
>>
>> The rmtfs region is allocated using UIO, making it technically "dynamic."
>>
>> Its address and size can be read from sysfs:
>>
>> $ cat /sys/class/uio/uio0/name
>> /sys/class/uio/uio0/maps/map0/addr
>> 0x00000000f2701000
>>
>> $ cat /sys/class/uio/uio0/maps/map0/size
>> 0x0000000000200000
>>
>> Like the OnePlus 6, the Pixel 3 requires 1 kB of reserved memory on either
>> side of the rmtfs region to work around an XPU bug that would otherwise
>> cause erroneous violations when accessing the rmtfs_mem region.
>
> [...]
>
>> +&gmu {
>> + status = "okay";
>> +};
>
> It's already enabled> +
>> +&mdss {
>> + status = "okay";
>> +};
>> +
>> +&mdss_dsi0 {
>> + status = "okay";
>> + vdda-supply = <&vdda_mipi_dsi0_1p2>;
>
> 'status' should be last, with a \n before it
>
>> +
>> + ports {
>> + port@1 {
>> + endpoint {
>
> &mdss_dsi0_out {} instead
>
>> + remote-endpoint = <&panel_in>;
>> + data-lanes = <0 1 2 3>;
>> + qcom,te-source = "mdp_vsync_e";
>> + };
>> + };
>> + };
>> +
>> + panel@0 {
>> + compatible = "lg,sw43408";
>> + reg = <0>;
>> +
>> + vddi-supply = <&vreg_l14a_1p88>;
>> + vpnl-supply = <&vreg_l28a_3p0>;
>> +
>> + reset-gpios = <&tlmm 6 GPIO_ACTIVE_LOW>;
>> +
>> + pinctrl-names = "default";
>> + pinctrl-0 = <&panel_reset_pins &panel_te_pin &panel_pmgpio_pins>;
>
> property-n
> property-names
>
> in this order, please
>
>> +
>> + port {
>> + panel_in: endpoint {
>> + remote-endpoint = <&mdss_dsi0_out>;
>> + };
>> + };
>> + };
>> +};
>> +
>> +&mdss_dsi0_out {
>> + remote-endpoint = <&panel_in>;
>> + data-lanes = <0 1 2 3>;
>> +};
>
> (as mentioned above..)
>
> [...]
>
>> + /* rmtfs lower guard */
>> + memory@...00000 {
>> + reg = <0 0xf2700000 0 0x1000>;
>> + no-map;
>> + };
>> +
>> + rmtfs_mem: memory@...01000 {
>> + compatible = "qcom,rmtfs-mem";
>> + reg = <0 0xf2701000 0 0x200000>;
>> + no-map;
>> +
>> + qcom,client-id = <1>;
>> + qcom,vmid = <15>;
>> + };
>> +
>> + /* rmtfs upper guard */
>> + memory@...01000 {
>> + reg = <0 0xf2901000 0 0x1000>;
>> + no-map;
>> + };
>
> qcom,use-guard-pages instead
>
> [...]
>
>> + vreg_l14a_1p88: ldo14 {
>> + regulator-min-microvolt = <1800000>;
>> + regulator-max-microvolt = <1800000>;
>> + regulator-initial-mode = <RPMH_REGULATOR_MODE_HPM>;
>> + regulator-boot-on;
>> + /*
>> + * We can't properly bring the panel back if it gets turned off
>> + * so keep it's regulators always on for now.
>> + */
>> + regulator-always-on;
>
> Sounds like your panel driver is not sequencing things properly
>
> [...]
>
>> +&uart9 {
>> + label = "LS-UART1";
>
> This is a 96boards-ism> +
>> + status = "okay";
>> +};
>> +
>> +&usb_1 {
>> + status = "okay";
>> +};
>> +
>> +&usb_1_dwc3 {
>> + dr_mode = "peripheral";
>
> Are you sure?
For now, the peripheral usb works, haven't tried to switch to host mode yet.
>
> [...]
>
>> +&usb_2 {
>> + status = "okay";
>> +};
>> +
>> +&usb_2_dwc3 {
>> + dr_mode = "host";
>
> Does the phone actually have something connected to both USB hosts?
I assume not,
Bus 001 Device 001: ID 1d6b:0002 Linux 6.16.7-sdm845 xhci-hcd xHCI Host
Controller
Bus 002 Device 001: ID 1d6b:0003 Linux 6.16.7-sdm845 xhci-hcd xHCI Host
Controller
I looked at the downstream description and it seems usb_2 was used only
on sda845 devkit.
David
>
> Would that be the weasel(?) security chip?
>
> [...]
>
>> diff --git a/arch/arm64/boot/dts/qcom/sdm845-google-crosshatch.dts b/arch/arm64/boot/dts/qcom/sdm845-google-crosshatch.dts
>> new file mode 100644
>> index 0000000000000..dc9938ffc0ab8
>> --- /dev/null
>> +++ b/arch/arm64/boot/dts/qcom/sdm845-google-crosshatch.dts
>> @@ -0,0 +1,137 @@
>> +// SPDX-License-Identifier: GPL-2.0-only
>> +
>> +/dts-v1/;
>> +
>> +#include "sdm845-google-common.dtsi"
>> +
>> +/ {
>> + model = "Google Pixel 3 XL";
>> + compatible = "google,crosshatch", "qcom,sdm845";
>> +
>> + battery: battery {
>> + compatible = "simple-battery";
>> +
>> + charge-full-design-microamp-hours = <3480000>;
>> + voltage-min-design-microvolt = <3600000>;
>> + voltage-max-design-microvolt = <4400000>;
>> + };
>> +
>> + chosen {
>> + #address-cells = <2>;
>> + #size-cells = <2>;
>> + ranges;
>> +
>> + /* for u-boot */
>> + framebuffer: framebuffer@...00000 {
>> + compatible = "simple-framebuffer";
>> + reg = <0 0x9d400000 0 (2960 * 1440 * 4)>;
>> + width = <1440>;
>> + height = <2960>;
>> + stride = <(1440 * 4)>;
>> + format = "a8r8g8b8";
>> + };
>> + };
>> +
>> + reserved-memory {
>> + framebuffer_region@...00000 {
>> + no-map;
>> + reg = <0 0x9d400000 0 0x02400000>;
>> + };
>> + };
>> +};
>> +
>> +&gmu {
>> + status = "okay";
>> +};
>> +
>> +&mdss {
>> + status = "okay";
>> +};
>
> I'm sure you can push some of this into -common
>
>> +
>> +&mdss_dsi0 {
>> + vdda-supply = <&vdda_mipi_dsi0_1p2>;
>> +
>> + status = "okay";
>> +
>> + ports {
>> + port@1 {
>> + endpoint {
>> + remote-endpoint = <&panel_in>;
>> + data-lanes = <0 1 2 3>;
>> + qcom,te-source = "mdp_vsync_e";
>> + };
>> + };
>> + };
>> +
>> + panel@0 {
>> + compatible = "samsung,s6e3ha8";
>> + reg = <0>;
>> +
>> + vci-supply = <&vreg_l28a_3p0>; // downstream
>
> which supply should /* upstream */ use then? :(
>
> Konrad
--
David Heidelberg
Powered by blists - more mailing lists