Re: [PATCH v2 1/2] ARM: dts: Add MDM9615 dtsi

From: Neil Armstrong
Date: Sat Aug 20 2016 - 14:09:21 EST


On 08/18/2016 08:08 PM, Stephen Boyd wrote:
> On 08/18, Neil Armstrong wrote:
>> +
>> +/dts-v1/;
>> +
>> +/include/ "skeleton.dtsi"
>> +
>> +#include <dt-bindings/interrupt-controller/arm-gic.h>
>> +#include <dt-bindings/clock/qcom,gcc-mdm9615.h>
>> +#include <dt-bindings/reset/qcom,gcc-mdm9615.h>
>> +#include <dt-bindings/mfd/qcom-rpm.h>
>> +#include <dt-bindings/soc/qcom,gsbi.h>
>> +
>> +/ {
>> + model = "Qualcomm MDM9615";
>> + compatible = "qcom,mdm9615";
>> + interrupt-parent = <&intc>;
>> +
>> + cpus {
>> + #address-cells = <1>;
>> + #size-cells = <0>;
>> + interrupts = <GIC_SPI 14 0x304>;
>
> What's this interrupt for? 0x304 seems wrong as well, because 0x3
> would mean two CPUs and this is a SPI and not a PPI?
Bad copy paste....

>
>> +
>> + cpu0: cpu@0 {
>> + compatible = "arm,cortex-a5";
>> + device_type = "cpu";
>> + next-level-cache = <&L2>;
>> + };
>> + };
>> +
>> + cpu-pmu {
>> + compatible = "arm,cortex-a5-pmu";
>> + interrupts = <GIC_SPI 10 0x304>;
>> + };
>> +
>> + clocks {
>> + cxo_board {
>> + compatible = "fixed-clock";
>> + #clock-cells = <0>;
>> + clock-frequency = <19200000>;
>> + clock-output-names = "cxo_board";
>
> This is unnecessary as it's the same name as the node name.
>
>> + };
>> + };
>> +
>> + soc: soc {
>> + #address-cells = <1>;
>> + #size-cells = <1>;
>> + ranges;
>> + compatible = "simple-bus";
>> +
>> + L2: l2-cache {
>> + compatible = "arm,pl310-cache";
>> + reg = <0x02040000 0x1000>;
>> + arm,data-latency = <2 2 0>;
>> + cache-unified;
>> + cache-level = <2>;
>> + };
>> +
>> + intc: interrupt-controller@2000000 {
>> + compatible = "qcom,msm-qgic2";
>> + interrupt-controller;
>> + #interrupt-cells = <3>;
>> + reg = <0x02000000 0x1000>,
>> + <0x02002000 0x1000>;
>> + };
>> +
>> + timer@200a000 {
>> + compatible = "qcom,kpss-timer", "qcom,msm-timer";
>> + interrupts = <GIC_PPI 1 0x301>,
>> + <GIC_PPI 2 0x301>,
>> + <GIC_PPI 3 0x301>;
>
> 0x101 for all those flags?

Bad copy paste....

>
>> + reg = <0x0200a000 0x100>;
>> + clock-frequency = <27000000>,
>> + <32768>;
>> + cpu-offset = <0x80000>;
>> + };
>> +
>> + msmgpio: pinctrl@800000 {
>> + compatible = "qcom,mdm9615-pinctrl", "syscon";
>
> What's the syscon for?

Leftover for USB HSIC calibration registers

>
>> + gpio-controller;
>> + #gpio-cells = <2>;
>> + interrupts = <GIC_SPI 16 IRQ_TYPE_LEVEL_HIGH>;
>> + interrupt-controller;
>> + #interrupt-cells = <2>;
>> + reg = <0x800000 0x4000>;
>> + };
>> +
>> + gcc: clock-controller@900000 {
>> + compatible = "qcom,gcc-mdm9615";
>> + #clock-cells = <1>;
>> + #reset-cells = <1>;
>> + reg = <0x900000 0x4000>;
>> + };
>> +
>> + lcc: clock-controller@28000000 {
>> + compatible = "qcom,lcc-mdm9615";
>> + reg = <0x28000000 0x1000>;
>> + #clock-cells = <1>;
>> + #reset-cells = <1>;
>> + };
>> +
>> + l2cc: clock-controller@2011000 {
>> + compatible = "syscon";
>> + reg = <0x02011000 0x1000>;
>> + };
>> +
>> + rng@1a500000 {
>> + compatible = "qcom,prng";
>> + reg = <0x1a500000 0x200>;
>> + clocks = <&gcc PRNG_CLK>;
>> + clock-names = "core";
>> + assigned-clocks = <&gcc PRNG_CLK>;
>> + assigned-clock-rates = <32000000>;
>> + };
>> +
>> + vsdcc_fixed: vsdcc-regulator {
>> + compatible = "regulator-fixed";
>> + regulator-name = "SDCC Power";
>> + regulator-min-microvolt = <2700000>;
>> + regulator-max-microvolt = <2700000>;
>> + regulator-always-on;
>> + };
>
> This should go into the root under a "regulators" node.

Done.
>
>> +
>> + gsbi2: gsbi@16100000 {
>> + compatible = "qcom,gsbi-v1.0.0";
>> + cell-index = <2>;
>> + reg = <0x16100000 0x100>;
>> + clocks = <&gcc GSBI2_H_CLK>;
>> + clock-names = "iface";
>> + status = "disabled";
>> + #address-cells = <1>;
>> + #size-cells = <1>;
>> + ranges;
>> +
>> + gsbi2_i2c: i2c@16180000 {
>> + compatible = "qcom,i2c-qup-v1.1.1";
>> + #address-cells = <1>;
>> + #size-cells = <0>;
>> + reg = <0x16180000 0x1000>;
>> + interrupts = <GIC_SPI 149 IRQ_TYPE_NONE>;
>
> There should be a trigger type... high perhaps?

Well, I do not know, and other qcom dtsi has 0x0 or NONE here...

>
>> +
>> + clocks = <&gcc GSBI2_QUP_CLK>, <&gcc GSBI2_H_CLK>;
>> + clock-names = "core", "iface";
>> + status = "disabled";
>> + };
>> + };
>> +
>> + gsbi3: gsbi@16200000 {
>> + compatible = "qcom,gsbi-v1.0.0";
>> + cell-index = <3>;
>> + reg = <0x16200000 0x100>;
>> + clocks = <&gcc GSBI3_H_CLK>;
>> + clock-names = "iface";
>> + status = "disabled";
>> + #address-cells = <1>;
>> + #size-cells = <1>;
>> + ranges;
>> +
>> + gsbi3_spi: spi@16280000 {
>> + compatible = "qcom,spi-qup-v1.1.1";
>> + #address-cells = <1>;
>> + #size-cells = <0>;
>> + reg = <0x16280000 0x1000>;
>> + interrupts = <GIC_SPI 151 IRQ_TYPE_NONE>;
>> + spi-max-frequency = <24000000>;
>> +
>> + clocks = <&gcc GSBI3_QUP_CLK>, <&gcc GSBI3_H_CLK>;
>> + clock-names = "core", "iface";
>> + status = "disabled";
>> + };
>> + };
>> +
>> + gsbi4: gsbi@16300000 {
>> + compatible = "qcom,gsbi-v1.0.0";
>> + cell-index = <4>;
>> + reg = <0x16300000 0x100>;
>> + clocks = <&gcc GSBI4_H_CLK>;
>> + clock-names = "iface";
>> + status = "disabled";
>> + #address-cells = <1>;
>> + #size-cells = <1>;
>> + ranges;
>> +
>> + syscon-tcsr = <&tcsr>;
>> +
>> + gsbi4_serial: serial@16340000 {
>> + compatible = "qcom,msm-uartdm-v1.3", "qcom,msm-uartdm";
>> + reg = <0x16340000 0x1000>,
>> + <0x16300000 0x1000>;
>> + interrupts = <GIC_SPI 152 IRQ_TYPE_NONE>;
>> + clocks = <&gcc GSBI4_UART_CLK>, <&gcc GSBI4_H_CLK>;
>> + clock-names = "core", "iface";
>> + status = "disabled";
>> + };
>> + };
>> +
>> + gsbi5: gsbi@16400000 {
>> + compatible = "qcom,gsbi-v1.0.0";
>> + cell-index = <5>;
>> + reg = <0x16400000 0x100>;
>> + clocks = <&gcc GSBI5_H_CLK>;
>> + clock-names = "iface";
>> + status = "disabled";
>> + #address-cells = <1>;
>> + #size-cells = <1>;
>> + ranges;
>> +
>> + syscon-tcsr = <&tcsr>;
>> +
>> + gsbi5_i2c: i2c@16480000 {
>> + compatible = "qcom,i2c-qup-v1.1.1";
>> + #address-cells = <1>;
>> + #size-cells = <0>;
>> + reg = <0x16480000 0x1000>;
>> + interrupts = <GIC_SPI 155 IRQ_TYPE_NONE>;
>> +
>> + /* QUP clock is not initialized, set rate */
>> + assigned-clocks = <&gcc GSBI5_QUP_CLK>;
>> + assigned-clock-rates = <24000000>;
>> +
>> + clocks = <&gcc GSBI5_QUP_CLK>, <&gcc GSBI5_H_CLK>;
>> + clock-names = "core", "iface";
>> + status = "disabled";
>> + };
>> +
>> + gsbi5_serial: serial@16440000 {
>> + compatible = "qcom,msm-uartdm-v1.3", "qcom,msm-uartdm";
>> + reg = <0x16440000 0x1000>,
>> + <0x16400000 0x1000>;
>> + interrupts = <GIC_SPI 154 IRQ_TYPE_NONE>;
>> + clocks = <&gcc GSBI5_UART_CLK>, <&gcc GSBI5_H_CLK>;
>> + clock-names = "core", "iface";
>> + status = "disabled";
>> + };
>> + };
>> +
>> + qcom,ssbi@500000 {
>> + compatible = "qcom,ssbi";
>> + reg = <0x500000 0x1000>;
>> + qcom,controller-type = "pmic-arbiter";
>> +
>> + pmicintc: pmic@0 {
>> + compatible = "qcom,pm8018", "qcom,pm8921";
>
> I know that DT specifies most specific compatible first, but when
> the generic compatible is part number specific as well it never
> feels right to me. I guess I'll have to get over this.
>
>> + interrupts = <GIC_PPI 226 IRQ_TYPE_NONE>;
>> + #interrupt-cells = <2>;
>> + interrupt-controller;
>> + #address-cells = <1>;
>> + #size-cells = <0>;
>
> Can we have interrupt-parent = <&pmicintc> here instead of in
> every node?

No since the pmic node interrupt parent is the GIC.

>
>> +
>> + pwrkey@1c {
>> + compatible = "qcom,pm8018-pwrkey", "qcom,pm8921-pwrkey";
>> + reg = <0x1c>;
>> + interrupt-parent = <&pmicintc>;
>> + interrupts = <50 IRQ_TYPE_EDGE_RISING>,
>> + <51 IRQ_TYPE_EDGE_RISING>;
>> + debounce = <15625>;
>> + pull-up;
>> + };
>> +
>> + pmicmpp: mpp@50 {
>> + compatible = "qcom,pm8018-mpp";
>> + interrupt-parent = <&pmicintc>;
>> + interrupts = <24 IRQ_TYPE_EDGE_RISING>,
>> + <25 IRQ_TYPE_EDGE_RISING>,
>> + <26 IRQ_TYPE_EDGE_RISING>,
>> + <27 IRQ_TYPE_EDGE_RISING>,
>> + <28 IRQ_TYPE_EDGE_RISING>,
>> + <29 IRQ_TYPE_EDGE_RISING>;
>
> We've recently been reminded that these should all be IRQ_TYPE_NONE. Also add the qcom,ssbi-mpp compatible please.

Ok.
>
>> + reg = <0x50>;
>> + gpio-controller;
>> + #gpio-cells = <2>;
>> + };
>> +
>> + rtc@11d {
>> + compatible = "qcom,pm8018-rtc", "qcom,pm8921-rtc";
>> + interrupt-parent = <&pmicintc>;
>> + interrupts = <39 IRQ_TYPE_EDGE_RISING>;
>> + reg = <0x11d>;
>> + allow-set-time;
>> + };
>> +
>> + pmicgpio: gpio@150 {
>> + compatible = "qcom,pm8018-gpio";
>> + interrupt-parent = <&pmicintc>;
>> + interrupts = <24 IRQ_TYPE_EDGE_RISING>,
>> + <25 IRQ_TYPE_EDGE_RISING>,
>> + <26 IRQ_TYPE_EDGE_RISING>,
>> + <27 IRQ_TYPE_EDGE_RISING>,
>> + <28 IRQ_TYPE_EDGE_RISING>,
>> + <29 IRQ_TYPE_EDGE_RISING>;
>
> Same IRQ_TYPE_NONE comment. Also, add the qcom,ssbi-gpio
> compatible please.
>
>> + gpio-controller;
>> + #gpio-cells = <2>;
>> + };
>> + };
>> + };
>> +
>> + sdcc1bam:dma@12182000{
>
> Add some space here:
Done.
>
> sdcc1bam: dma@12180000 {
>
>> + compatible = "qcom,bam-v1.3.0";
>> + reg = <0x12182000 0x8000>;
>> + interrupts = <GIC_SPI 98 IRQ_TYPE_NONE>;
>> + clocks = <&gcc SDC1_H_CLK>;
>> + clock-names = "bam_clk";
>> + #dma-cells = <1>;
>> + qcom,ee = <0>;
>> + };
>> +
>> + sdcc2bam:dma@12142000{
>
> ditto.
>
>> + compatible = "qcom,bam-v1.3.0";
>> + reg = <0x12142000 0x8000>;
>> + interrupts = <GIC_SPI 97 IRQ_TYPE_NONE>;
>
> There should really be some flags.

Same as for the qup and uartdm, other qcom dtsi left 0x0 or NONE here...

>
>> + clocks = <&gcc SDC2_H_CLK>;
>> + clock-names = "bam_clk";
>> + #dma-cells = <1>;
>> + qcom,ee = <0>;
>> + };
>> +
>

Thanks,
Neil