[PATCH v3 4/5] arm64: Add initial support for Blaize BLZP1600 CB2

Nikolaos Pasaloukos nikolaos.pasaloukos at blaize.com
Thu Apr 25 03:41:56 PDT 2024


On 25/04/2024 10:21, Krzysztof Kozlowski wrote:
> On 25/04/2024 11:15, Niko Pasaloukos wrote:
>> Adds support for the Blaize CB2 development board based on
>> BLZP1600 SoC. This consists of a Carrier-Board-2 and a SoM.
> 
> Subject: missing dts prefix.
> 
> 
> ...
> 
>> +
>> +/ {
>> +	interrupt-parent = <&gic>;
>> +	#address-cells = <2>;
>> +	#size-cells = <1>;
>> +
>> +	cpus {
>> +		#address-cells = <2>;
>> +		#size-cells = <0>;
>> +
>> +		cpu0: cpu at 0 {
>> +			compatible = "arm,cortex-a53";
>> +			device_type = "cpu";
>> +			enable-method = "psci";
>> +			reg = <0x0 0x0>;
>> +			next-level-cache = <&l2>;
>> +		};
>> +
>> +		cpu1: cpu at 1 {
>> +			compatible = "arm,cortex-a53";
>> +			device_type = "cpu";
>> +			enable-method = "psci";
>> +			reg = <0x0 0x1>;
>> +			next-level-cache = <&l2>;
>> +		};
>> +
>> +		l2: l2-cache0 {
>> +			compatible = "cache";
>> +			cache-level = <2>;
>> +			cache-unified;
>> +		};
>> +	};
>> +
>> +	timer {
>> +		compatible = "arm,armv8-timer";
>> +		interrupts = /* Physical Secure PPI */
>> +			     <GIC_PPI 13 (GIC_CPU_MASK_RAW(0x3) |
>> +					  IRQ_TYPE_LEVEL_LOW)>,
>> +			     /* Physical Non-Secure PPI */
>> +			     <GIC_PPI 14 (GIC_CPU_MASK_RAW(0x3) |
>> +					  IRQ_TYPE_LEVEL_LOW)>,
>> +			     /* Hypervisor PPI */
>> +			     <GIC_PPI 10 (GIC_CPU_MASK_RAW(0x3) |
>> +					  IRQ_TYPE_LEVEL_LOW)>,
>> +			     /* Virtual PPI */
>> +			     <GIC_PPI 11 (GIC_CPU_MASK_RAW(0x3) |
>> +					  IRQ_TYPE_LEVEL_LOW)>;
>> +	};
>> +
>> +	psci {
>> +		compatible = "arm,psci-1.0", "arm,psci-0.2";
>> +		method = "smc";
>> +	};
>> +
>> +	pmu {
> 
> Nodes in top-level look randomly ordered. Any reason why not using DTS
> coding style in this regard?
> 
>> +		compatible = "arm,cortex-a53-pmu";
>> +		interrupts = <GIC_SPI 76 IRQ_TYPE_LEVEL_HIGH>,
>> +			     <GIC_SPI 77 IRQ_TYPE_LEVEL_HIGH>;
>> +		interrupt-affinity = <&cpu0>, <&cpu1>;
>> +	};
>> +
>> +	sram at 0 {
>> +		/*
>> +		 * On BLZP1600 there is no general purpose (non-secure) SRAM.
>> +		 * A small DDR memory space has been reserved for general use.
>> +		 */
>> +		compatible = "mmio-sram";
>> +		reg = <0x0 0x00000000 0x00001000>;
>> +		#address-cells = <1>;
>> +		#size-cells = <1>;
>> +		ranges = <0 0x0 0x00000000 0x1000>;
> 
> ranges follow reg
> 
>> +
>> +		/* SCMI reserved buffer space on DDR space */
>> +		scmi0_shm: scmi-sram at 800 {
>> +			compatible = "arm,scmi-shmem";
>> +			reg = <0x800 0x80>;
>> +		};
>> +	};
>> +
>> +	firmware {
>> +		scmi {
>> +			compatible = "arm,scmi-smc";
>> +			arm,smc-id = <0x82002000>;
>> +			#address-cells = <1>;
>> +			#size-cells = <0>;
>> +
>> +			shmem = <&scmi0_shm>;
>> +
>> +			scmi_clk: protocol at 14 {
>> +				reg = <0x14>;
>> +				#clock-cells = <1>;
>> +			};
>> +
>> +			scmi_rst: protocol at 16 {
>> +				reg = <0x16>;
>> +				#reset-cells = <1>;
>> +			};
>> +		};
>> +	};
>> +
>> +	soc {
> 
> This does not cause dtbs_check W=1 warnings? Surprising a bit... This
> should cause big fat warning, so I have doubts patchset was tested.
> 
> 
> Best regards,
> Krzysztof
> 

No it doesn't cause any warnings. I did:
make arch=arm64 dt_binding_check
make arch=arm64 dtbs_check W=1
I don't get any warnings. Could you please let me know what kind of
warning I should get? Am I doing something wrong and I don't get
the warning?

Kind regards,
Niko



More information about the linux-arm-kernel mailing list