Re: [PATCH 3/4] arm64: dts: ti: Add support for J7200 SoC

[Suman Anna <s-anna@ti.com>]

Hi Grygorii,

On 7/28/20 2:16 PM, Grygorii Strashko wrote:
> 
> 
> On 23/07/2020 11:46, Lokesh Vutla wrote:
>> The J7200 SoC is a part of the K3 Multicore SoC architecture platform.
>> It is targeted for automotive gateway, vehicle compute systems,
>> Vehicle-to-Vehicle (V2V) and Vehicle-to-Everything (V2X) applications.
>> The SoC aims to meet the complex processing needs of modern embedded
>> products.
>>
>> Some highlights of this SoC are:
>> * Dual Cortex-A72s in a single cluster, two clusters of lockstep
>>    capable dual Cortex-R5F MCUs and a Centralized Device Management and
>>    Security Controller (DMSC).
>> * Configurable L3 Cache and IO-coherent architecture with high data
>>    throughput capable distributed DMA architecture under NAVSS.
>> * Integrated Ethernet switch supporting up to a total of 4 external ports
>>    in addition to legacy Ethernet switch of up to 2 ports.
>> * Upto 1 PCIe-GEN3 controller, 1 USB3.0 Dual-role device subsystems,
>>    20 MCANs, 3 McASP, eMMC and SD, OSPI/HyperBus memory controller, I3C
>>    and I2C, eCAP/eQEP, eHRPWM among other peripherals.
>> * One hardware accelerator block containing AES/DES/SHA/MD5 called SA2UL
>>    management.
>>
>> See J7200 Technical Reference Manual (SPRUIU1, June 2020)
>> for further details: https://www.ti.com/lit/pdf/spruiu1
>>
>> Signed-off-by: Lokesh Vutla <lokeshvutla@ti.com>
>> ---
>>   arch/arm64/boot/dts/ti/k3-j7200-main.dtsi     | 199 ++++++++++++++++++
>>   .../boot/dts/ti/k3-j7200-mcu-wakeup.dtsi      |  84 ++++++++
>>   arch/arm64/boot/dts/ti/k3-j7200.dtsi          | 165 +++++++++++++++
>>   3 files changed, 448 insertions(+)
>>   create mode 100644 arch/arm64/boot/dts/ti/k3-j7200-main.dtsi
>>   create mode 100644 arch/arm64/boot/dts/ti/k3-j7200-mcu-wakeup.dtsi
>>   create mode 100644 arch/arm64/boot/dts/ti/k3-j7200.dtsi
>>
>> diff --git a/arch/arm64/boot/dts/ti/k3-j7200-main.dtsi 
>> b/arch/arm64/boot/dts/ti/k3-j7200-main.dtsi
>> new file mode 100644
>> index 000000000000..70c8f7e941fb
>> --- /dev/null
>> +++ b/arch/arm64/boot/dts/ti/k3-j7200-main.dtsi
>> @@ -0,0 +1,199 @@
>> +// SPDX-License-Identifier: GPL-2.0
>> +/*
>> + * Device Tree Source for J7200 SoC Family Main Domain peripherals
>> + *
>> + * Copyright (C) 2020 Texas Instruments Incorporated - 
>> https://www.ti.com/
>> + */
>> +
>> +&cbass_main {
>> +    msmc_ram: sram@70000000 {
>> +        compatible = "mmio-sram";
>> +        reg = <0x0 0x70000000 0x0 0x100000>;
>> +        #address-cells = <1>;
>> +        #size-cells = <1>;
>> +        ranges = <0x0 0x0 0x70000000 0x100000>;
>> +
>> +        atf-sram@0 {
>> +            reg = <0x0 0x20000>;
>> +        };
>> +    };
>> +
>> +    gic500: interrupt-controller@1800000 {
>> +        compatible = "arm,gic-v3";
>> +        #address-cells = <2>;
>> +        #size-cells = <2>;
>> +        ranges;
>> +        #interrupt-cells = <3>;
>> +        interrupt-controller;
>> +        reg = <0x00 0x01800000 0x00 0x10000>,    /* GICD */
>> +              <0x00 0x01900000 0x00 0x100000>;    /* GICR */
>> +
>> +        /* vcpumntirq: virtual CPU interface maintenance interrupt */
>> +        interrupts = <GIC_PPI 9 IRQ_TYPE_LEVEL_HIGH>;
>> +
>> +        gic_its: msi-controller@1820000 {
>> +            compatible = "arm,gic-v3-its";
>> +            reg = <0x00 0x01820000 0x00 0x10000>;
>> +            socionext,synquacer-pre-its = <0x1000000 0x400000>;
>> +            msi-controller;
>> +            #msi-cells = <1>;
>> +        };
>> +    };
>> +
>> +    main_navss: navss@30000000 {
>> +        compatible = "simple-mfd";
>> +        #address-cells = <2>;
>> +        #size-cells = <2>;
>> +        ranges = <0x00 0x30000000 0x00 0x30000000 0x00 0x0c400000>;
> 
> Can we return back
> ti,sci-dev-id = <199>;
> 
> here? it's needed for DMA PSI-L pairing.

Oh, I wasn't aware of this. Lokesh removed it based on my earlier 
comments. Is this used by more than one driver? Anytime you see a 
ti,sci-dev-id, you would also expect to see the ti,sci reference in 
general for that driver to really leverage it.

The lack of a specific binding for this node also makes one think it is 
not needed.

> 
> Option: add it as part of dma/net submission.

Yeah, I think we should go with this approach on upstream if we aren't 
respinning the series. Any reason, this couldn't be added to the udma 
node itself?

regards
Suman