Re: [PATCH v5 4/6] arm64: dts: qcom: Add UFS node

[Konrad Dybcio <konradybcio@kernel.org>]

On 17.08.2024 10:38 PM, Marcus Glocker wrote:
> Add the UFS Host Controller node.  This was basically copied from the
> arch/arm64/boot/dts/qcom/sc7180.dtsi file.
> 
> Signed-off-by: Marcus Glocker <marcus@nazgul.ch>
> Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
> ---
>  arch/arm64/boot/dts/qcom/x1e80100.dtsi | 72 ++++++++++++++++++++++++++
>  1 file changed, 72 insertions(+)
> 
> diff --git a/arch/arm64/boot/dts/qcom/x1e80100.dtsi b/arch/arm64/boot/dts/qcom/x1e80100.dtsi
> index 7bca5fcd7d52..9f01b3ff3737 100644
> --- a/arch/arm64/boot/dts/qcom/x1e80100.dtsi
> +++ b/arch/arm64/boot/dts/qcom/x1e80100.dtsi
> @@ -2878,6 +2878,78 @@ mmss_noc: interconnect@1780000 {
>  			#interconnect-cells = <2>;
>  		};
>  
> +		ufs_mem_hc: ufs@1d84000 {
> +			compatible = "qcom,x1e80100-ufshc", "qcom,ufshc",
> +				     "jedec,ufs-2.0";
> +			reg = <0 0x01d84000 0 0x3000>;
> +			interrupts = <GIC_SPI 265 IRQ_TYPE_LEVEL_HIGH>;
> +			phys = <&ufs_mem_phy>;
> +			phy-names = "ufsphy";
> +			lanes-per-direction = <1>;
> +			#reset-cells = <1>;
> +			resets = <&gcc GCC_UFS_PHY_BCR>;
> +			reset-names = "rst";
> +
> +			power-domains = <&gcc GCC_UFS_PHY_GDSC>;
> +
> +			iommus = <&apps_smmu 0xa0 0x0>;

Looks like this should be 0x1a0 maybe
> +
> +			clock-names = "core_clk",
> +				      "bus_aggr_clk",
> +				      "iface_clk",
> +				      "core_clk_unipro",
> +				      "ref_clk",
> +				      "tx_lane0_sync_clk",
> +				      "rx_lane0_sync_clk";
> +			clocks = <&gcc GCC_UFS_PHY_AXI_CLK>,
> +				 <&gcc GCC_AGGRE_UFS_PHY_AXI_CLK>,
> +				 <&gcc GCC_UFS_PHY_AHB_CLK>,
> +				 <&gcc GCC_UFS_PHY_UNIPRO_CORE_CLK>,
> +				 <&rpmhcc RPMH_CXO_CLK>,
> +				 <&gcc GCC_UFS_PHY_TX_SYMBOL_0_CLK>,
> +				 <&gcc GCC_UFS_PHY_RX_SYMBOL_0_CLK>;

You also want

<&gcc GCC_UFS_PHY_RX_SYMBOL_1_CLK>

> +			freq-table-hz = <50000000 200000000>,
25000000 300000000

> +					<0 0>,
> +					<0 0>,
> +					<37500000 150000000>,
75000000 300000000

> +					<0 0>,
> +					<0 0>,
> +					<0 0>;
> +
> +			interconnects = <&aggre1_noc MASTER_UFS_MEM QCOM_ICC_TAG_ALWAYS
> +					 &mc_virt SLAVE_EBI1 QCOM_ICC_TAG_ALWAYS>,
> +					<&gem_noc MASTER_APPSS_PROC QCOM_ICC_TAG_ALWAYS
> +					 &config_noc SLAVE_UFS_MEM_CFG QCOM_ICC_TAG_ALWAYS>;
> +			interconnect-names = "ufs-ddr", "cpu-ufs";
> +
> +			qcom,ice = <&ice>;
> +
> +			status = "disabled";
> +		};
> +
> +		ufs_mem_phy: phy@1d87000 {
> +			compatible = "qcom,x1e80100-qmp-ufs-phy";
> +			reg = <0 0x01d87000 0 0x1000>;

most definitely should be 0x01d80000 with a size of 0x2000

> +			clocks = <&rpmhcc RPMH_CXO_CLK>,
> +				 <&gcc GCC_UFS_PHY_PHY_AUX_CLK>,
> +				 <&tcsr TCSR_UFS_PHY_CLKREF_EN>;
> +			clock-names = "ref",
> +				      "ref_aux",
> +				      "qref";
> +			power-domains = <&gcc GCC_UFS_PHY_GDSC>;
> +			resets = <&ufs_mem_hc 0>;
> +			reset-names = "ufsphy";
> +			#phy-cells = <0>;
> +			status = "disabled";
> +		};
> +
> +		ice: crypto@1d90000 {
> +			compatible = "qcom,x1e80100-inline-crypto-engine",
> +				     "qcom,inline-crypto-engine";
> +			reg = <0 0x01d90000 0 0x8000>;

0x1d88000


All this combined means you probably wrote your init sequence into some
free(?) register space and the one left over from the bootloader was
good enough :P

Konrad