.
├── .github/workflows
├── ci/scripts
├── hw
│ ├── asic
│ ├── core-v-mini-mcu
│ ├── fpga
│ ├── ip
│ ├── ip_examples
│ ├── simulation
│ └── vendor
├── scripts
│ ├── sim
│ └── synthesis
├── sw
│ ├── applications
│ ├── device/lib
│ ├── linker
│ └── vendor
├── tb
├── util
└── README.md
======================================
======================================
x-heep (eXtendable Heterogeneous Energy-Efficient Platform) is a RISC-V microcontroller described in SystemVerilog
that can be configured to target small and tiny platforms as well as extended to support accelerators.
The cool thing about x-heep is that we provide a simple customizable MCU, so CPUs, common peripherals, memories, etc.
so that you can extend it with your own accelerator without modifying the MCU, but just instantiating it in your design.
By doing so, you inherit an IP capable of booting RTOS (such as freeRTOS) with the whole SW stack, including drivers and SDK,
and you can focus on building your special HW supported by the microcontroller.
x-heep supports simulation with Verilator, Questasim, etc, it can be implemented on FPGA, and it supports implementation in Silicon, which is its main (but not only) target. See below for more details.
The block diagram below shows the x-heep MCU
- Install Conda as described in the link, and create the Conda enviroment with python 3.8:
conda update conda
conda env create -f environment.ymlActivate the environment with
conda activate core-v-mini-mcu- Install the required Python tools:
pip3 install --user -r python-requirements.txt
Add '--root user_builds' to set your build foders for the pip packages
and add that folder to the PATH variable
- Install the required apt tools:
sudo apt install lcov libelf1 libelf-dev libftdi1-2 libftdi1-dev libncurses5 libssl-dev libudev-dev libusb-1.0-0 lsb-release texinfo makeinfo autoconf cmake flex bison libexpat-dev gawk
In general, have a look at the Install required software section of the OpenTitan documentation.
- Install the RISC-V Compiler:
git clone --branch 2022.01.17 --recursive https://github.com/riscv/riscv-gnu-toolchain
cd riscv-gnu-toolchain
./configure --prefix=/home/yourusername/tools/riscv --with-abi=ilp32 --with-arch=rv32imc --with-cmodel=medlow
make
Then, set the RISCV env variable as:
export RISCV=/home/yourusername/tools/riscv
- Install the Verilator:
export VERILATOR_VERSION=4.210
git clone https://github.com/verilator/verilator.git
cd verilator
git checkout v$VERILATOR_VERSION
autoconf
./configure --prefix=/home/yourusername/tools/verilator/$VERILATOR_VERSION
make
make install
Then, set the PATH env variable to as:
export PATH=/home/yourusername/tools/verilator/$VERILATOR_VERSION/bin:$PATH
In general, have a look at the Install Verilator section of the OpenTitan documentation.
If you want to see the vcd waveforms generated by the Verilator simulation, install GTKWAVE:
sudo apt install libcanberra-gtk-module libcanberra-gtk3-module
sudo apt-get install -y gtkwave
We use version v0.0-1824-ga3b5bedf
See: Install Verible
To format your RTL code type:
make verible
We use FuseSoC for all the tools we use.
The fusesoc commands are inside the Makefile.
This repository relies on vendor to add new IPs. In the ./util folder, the vendor.py scripts implements what is describeb above.
You can compile the example applications and the platform using the Makefile. Type 'make help' for more information.
First, you have to generate the SystemVerilog package and C header file of the core-v-mini-mcu:
make mcu-gen
To change the default cpu type (i.e., cv32e20), the default bus type (i.e., onetoM) type or the memory size (i.e., number of banks):
make mcu-gen CPU=cv32e40p BUS=NtoM MEMORY_BANKS=16
The last command generates x-heep with the cv32e40p core, with a parallel bus, and 16 memory banks, each 32KB, for a total memory of 512KB.
Don't forget to set the RISCV env variable to the compiler folder (without the /bin included).
make app-helloworld
or for FPGAs,
make app-helloworld TARGET=pynq-z2
This will create the executable file to be loaded in your target system (ASIC, FPGA, Simulation).
This project supports simulation with Verilator, Synopsys VCS, and Siemens Questasim.
To simulate your application with Verilator, first compile the HDL:
make verilator-sim
then, go to your target system built folder
cd ./build/openhwgroup.org_systems_core-v-mini-mcu_0/sim-verilator
and type to run your compiled software:
./Vtestharness +firmware=../../../sw/applications/hello_world/hello_world.hex
or to execute all these three steps type:
make run-helloworld
To simulate your application with VCS, first compile the HDL:
make vcs-sim
then, go to your target system built folder
cd ./build/openhwgroup.org_systems_core-v-mini-mcu_0/sim-vcs
and type to run your compiled software:
./openhwgroup.org_systems_core-v-mini-mcu_0 +firmware=../../../sw/applications/hello_world/hello_world.hex
To simulate your application with Questasim, first set the env variable MODEL_TECH to your Questasim bin folder, then compile the HDL:
make questasim-sim
then, go to your target system built folder
cd ./build/openhwgroup.org_systems_core-v-mini-mcu_0/sim-modelsim/
and type to run your compiled software:
make run PLUSARGS="c firmware=../../../sw/applications/hello_world/hello_world.hex"
You can also use vopt for HDL optimized compilation:
make questasim-sim-opt
then go to
cd ./build/openhwgroup.org_systems_core-v-mini-mcu_0/sim_opt-modelsim/
and
make run RUN_OPT=1 PLUSARGS="c firmware=../../../sw/applications/hello_world/hello_world.hex"
You can also compile with the UPF power domain description as:
make questasim-sim-opt-upf FUSESOC_FLAGS="--flag=use_upf"
and then execute software as:
make run RUN_OPT=1 RUN_UPF=1 PLUSARGS="c firmware=../../../sw/applications/hello_world/hello_world.hex"
Questasim version must be >= Questasim 2020.4
To simulate the UART, we use the LowRISC OpenTitan UART DPI.
Read how to interact with it in the Section "Interact with the simulated UART" here.
The output of the UART DPI module is printed in the uart0.log file in the simulation folder.
For example, to see the "hello world!" output of the Verilator simulation:
cd ./build/openhwgroup.org_systems_core-v-mini-mcu_0/sim-verilator
./Vtestharness +firmware=../../../sw/applications/hello_world/hello_world.hex
cat uart0.log
Follow the Debug guide to debug core-v-mini-mcu.
Follow the ExecuteFromFlash guide to exxecute code directly from the FLASH with modelsim, FPGA, or ASIC.
This project offers two different X-HEEP implementetions on the Xilinx FPGAs, called Standalone-FEMU and Linux-FEMU.
In this version, the X-HEEP architecture is implemented on the programmable logic (PL) side of the FPGA, and its input/output are connected to the available headers on the FPGA board.
Make sure you have the FPGA board files installed in your Vivado.
For example, for the Xilinx Pynq-Z2 board, use the documentation provided at the following link to download and install them:
To build and program the bitstream for your FPGA with vivado, type:
make vivado-fpga FPGA_BOARD=pynq-z2
or add the flag use_bscane_xilinx to use the native Xilinx scanchain:
make vivado-fpga FPGA_BOARD=pynq-z2 FUSESOC_FLAGS=--flag=use_bscane_xilinx
Only Vivado 2021.2 has been tried.
To program the bitstream, open Vivado,
open --> Hardware Manager --> Open Target --> Autoconnect --> Program Device
and choose the file openhwgroup.org_systems_core-v-mini-mcu_0.bit
To run SW, follow the Debug guide to load the binaries with the HS2 cable over JTAG, or follow the ExecuteFromFlash guide if you have a FLASH attached to the FPGA.
In this version, the X-HEEP architecture is implemented on the programmable logic (PL) side of the FPGA and Linux is run on the ARM-based processing system (PS) side of the same chip.
Read the following documentation to have more information about this implementation.
This project can be implemented using standard cells based ASIC flow.
First, you need to provide technology-dependent implementations of some of the cells which require specific instantiation.
Then, please provide a set_libs.tcl and set_constraints.tcl scripts to set link and target libraries, and constraints as the clock.
To generate the analyze script for the synthesis scripts with DC, execute:
make asic
We are working on supporting OpenRoad and SkyWater 130nm PDK, please refer to the OpenRoadFlow page. This is not ready yet, it has not been tested.
This relies on a fork of edalize that contains templates for Design Compiler and OpenRoad.