This repository contains Dockerfiles for the different tools used in the Chisel-UVM project. This should help reduce the dependencies for different systems as a docker image can be created when needed.
This repository contains a directory for each Dockerfile. The idea being that when running a Docker image, you can mount said directory for access to files.
Install Docker Desktop for Windows: https://hub.docker.com/editions/community/docker-ce-desktop-windows/
Docker images can be mounted in Visual Studio Code. This is done by installing the "Remote - Containers" extension. From Visual Studio Code press ctrl + shift + p and choose Remote-Containers: Open Folder in Container... and choose From Dockerfile... Then choose the Chisel folder within this repository.
After some time, it will have built the docker image and you can now work from this folder. Open a terminal from the bar in the top. Now, run sbt test in the terminal and a test should run sucesfully.
Same thing as above. Run the following in the terminal to test the system:
yosys -m ghdl -p 'ghdl --std=08 src/accualu.vhd -e accualu; write_verilog accualu_synth.v'
This should generate a file called accualu_synth.v.
To run from a terminal, such as PowerShell, the image first needs to be built. This is done using the following command from the chisel folder:
docker build . --rm -t chisel
And for the VHDL2Verilog:
docker build . --rm -t vhdl2verilog
And then to run an image:
docker run -it -v <PATH\TO\CHISEL_FOLDER>:/usr/home chisel
And from that, ssh into /usr/home.
Same principle goes for VHDL2Verilog:
docker run -it -v <PATH\TO\VHDL2VERILOG_FOLDER>:/usr/home vhdl2verilog
and ssh into /usr/home
Every time a Docker instance is opened, it starts from image scratch. This means that the only things that are saved is the files located in the folder that was chosen (mounted as a volume) when starting the docker.
TBA
Docker sets a quite limited amount of resources as default. In Docker hub click Resources -> Advanced from which you can set number of CPUs and amount of RAM. You can freely set number of CPUs but Memory should be a reasonable amount. Memory is pre-allocated when starting a Docker.