This assumes you are on a Unix machine or are using WSL. You must install OpenLane2 using either the Nix or the Docker method as discussed here: https://openlane2.readthedocs.io/en/latest/getting_started/installation_overview.html
For this class, we will be using OpenLane 2.3.1 unless you are told later. Be sure to use this version for compatibility with our grading! The version is included in several steps below, so do not leave it out.
In this method, Nix will provide the binaries, but you first need to install Nix depending on your OS: https://openlane2.readthedocs.io/en/latest/getting_started/common/nix_installation/index.html
Nix will use the binaries locally, but you need the Nix settings from the OpenLane2 repository. To do this, you can clone this with:
git clone -b 2.3.1 https://github.com/efabless/openlane2
Change to the openlane2 directory and open a Nix shell:
cd openlane2
nix-shell
which should open a shell like this:
[nix-shell:~/openlane2]$
You should always use OpenLane in this shell since it is how the binaries are installed. The first time, it should download all of the correct binaries.
In this method, Docker will provide the binaries, but you first need to install Docker depending on your OS: https://openlane2.readthedocs.io/en/latest/getting_started/common/docker_installation/index.html
In this method, the binaries will be provided in a Docker image and the OpenLane Python package will call commands in the Docker image as appropriate.
Check that Docker is working by running:
docker run -it hello-world
and you should see "Hello from Docker!" with some other output.
You can install with Pip using:
python3 -m pip install openlane==2.3.1
From here, you can enter a Docker shell with the OpenLane tools by running:
openlane --dockerized
which should open a shell like this:
OpenLane Container (2.3.1):/home/<user>/<mydir>
You can either run OpenLane here, or you can run it from your regular OS and the scripts will be smart enugh to run commands in the Docker image. Using this shell will also allow you to run the GUI versions of tools since the OpenLane scripts export the X display properly on most systems.
Note that some Linux ditrubutions no longer allow you to use pip to directly install
packages to your system. If you get an error message saying "This environment is externally managed"
you need to create a virtual environment.
To check your installation, run the "smoke test":
openlane --smoke-test
or (if you are using the Docker method and not in a Docker shell):
openlane --dockerized --smoke-test
This should end with the following:
[19:43:42] INFO Smoke test passed.
Note, there might be a few "WARNING" messages, but that is ok.
Copyright 2024 VLSI-DA (see LICENSE for use)