Re: [PATCH v3] arm64: dts: qcom: sc7180: Add A618 gpu dt blob

[Doug Anderson]

Hi,

On Fri, Jan 31, 2020 at 4:04 AM Sharat Masetty <smasetty@xxxxxxxxxxxxxx> wrote:
>
> +               adreno_smmu: iommu@5040000 {
> +                       compatible = "qcom,sc7180-smmu-v2", "qcom,smmu-v2";
> +                       reg = <0 0x05040000 0 0x10000>;
> +                       #iommu-cells = <1>;
> +                       #global-interrupts = <2>;
> +                       interrupts = <GIC_SPI 229 IRQ_TYPE_LEVEL_HIGH>,
> +                                       <GIC_SPI 231 IRQ_TYPE_LEVEL_HIGH>,
> +                                       <GIC_SPI 364 IRQ_TYPE_EDGE_RISING>,
> +                                       <GIC_SPI 365 IRQ_TYPE_EDGE_RISING>,
> +                                       <GIC_SPI 366 IRQ_TYPE_EDGE_RISING>,
> +                                       <GIC_SPI 367 IRQ_TYPE_EDGE_RISING>,
> +                                       <GIC_SPI 368 IRQ_TYPE_EDGE_RISING>,
> +                                       <GIC_SPI 369 IRQ_TYPE_EDGE_RISING>,
> +                                       <GIC_SPI 370 IRQ_TYPE_EDGE_RISING>,
> +                                       <GIC_SPI 371 IRQ_TYPE_EDGE_RISING>;
> +                       clocks = <&gcc GCC_GPU_MEMNOC_GFX_CLK>,
> +                               <&gcc GCC_GPU_CFG_AHB_CLK>,
> +                               <&gcc GCC_DDRSS_GPU_AXI_CLK>;
> +
> +                       clock-names = "bus", "iface", "mem_iface_clk";

Repeated comment from v2 feedback:

Please send a patch to:

Documentation/devicetree/bindings/iommu/arm,smmu.yaml

...adding 'qcom,sc7180-smmu-v2'.  If you do this it will point out
that you've added a new clock: "mem_iface_clk".  Is this truly a new
clock in sc7180 compared to previous IOMMUs?  ...or is it not really
needed?


> +               gmu: gmu@506a000 {
> +                       compatible="qcom,adreno-gmu-618.0", "qcom,adreno-gmu";
> +                       reg = <0 0x0506a000 0 0x31000>, <0 0x0b290000 0 0x10000>,
> +                               <0 0x0b490000 0 0x10000>;
> +                       reg-names = "gmu", "gmu_pdc", "gmu_pdc_seq";
> +                       interrupts = <GIC_SPI 304 IRQ_TYPE_LEVEL_HIGH>,
> +                                  <GIC_SPI 305 IRQ_TYPE_LEVEL_HIGH>;
> +                       interrupt-names = "hfi", "gmu";
> +                       clocks = <&gpucc GPU_CC_CX_GMU_CLK>,
> +                              <&gpucc GPU_CC_CXO_CLK>,
> +                              <&gcc GCC_DDRSS_GPU_AXI_CLK>,
> +                              <&gcc GCC_GPU_MEMNOC_GFX_CLK>;
> +                       clock-names = "gmu", "cxo", "axi", "memnoc";
> +                       power-domains = <&gpucc CX_GDSC>;
> +                       power-domain-names = "cx";

As per continued comments on v2, please see if this works for you:

  power-domains = <&gpucc CX_GDSC>, <0>;
  power-domain-names = "cx", "gx";

...and work to get something more real for "gx" ASAP.  It did seem to
boot for me and (unless someone disagrees) it seems better than
totally leaving it out / violating the bindings?


-Doug