Installation

Warning

The following instructions are a work-in-progress and may not be entirely up to date.

Squishy is a python application that can be installed via pip, either from the git repository or via a wheel package.

System Requirements

Squishy is officially supported on most Linux distributions, the primary one being Arch Linux and variants there of (Manjaro, etc.). However other Linux distributions are supported as well.

Support for Microsoft Windows is preliminary, and due to lack of testing and development resources official support is currently not possible. But effort to make it as painless as possible to use on this platform is in-progress.

Support for Apple macOS is also unofficial, however due to its similarity to Linux and Unix-like operating systems, using Squishy on macOS should be just as painless as it is on Linux. Official support for macOS and other BSD-Family operating systems is planned.

Regardless of the operating system, there are some core dependencies that Squishy requires. This includes Python >= 3.8, Yosys >= 0.10, and nextpnr >= 0.4.

Squishy has been tested with CPython, but might possibly run under PyPy.

Installing Prerequisites

Prior to installing Squishy, you must install all of its prerequisites and requirements.

Installing Python

First off, install python and pip onto your system if the’re not there already.

Arch Linux

$ sudo pacman -S python python-pip

Other Linux

Warning

These instructions may be incorrect or incomplete!

For Debian based systems, use apt to install python3 and python3-pip

$ sudo apt install python3 python3-pip

For Fedora and other dnf based systems,

$ sudo dnf install python3 python3-pip

macOS

Install Homebrew if not done already, then install the requirements.

$ brew install python

Windows

Warning

These instructions may be incorrect or incomplete!

Download the latest Python installer from the python downloads page.

Follow the instructions and ensure that the installer installs pip and puts the python executable in your %PATH%

At this point, you may want to create a virtual environment for installing Squishy into:

$ python -m venv squishy.venv

Note

If your python install doesn’t have the venv module by default, then you will need to also install the python virtualenv package

You can then activate the virtual environment with:

Arch Linux

$ . squishy.venv/bin/activate

Other Linux

$ . squishy.venv/bin/activate

macOS

$ . squishy.venv/bin/activate

Windows

$ call squishy.venv/Scripts/activate.bat

Installing Yosys and nextpnr

Next, you need to install yosys and nextpnr onto your system if it is not there already.

YoWASP Install

Warning

There is an issue with the YoWASP toolchain failing timings for the design synthesis, It is not advisable to use it until this issue has been resolved, please use the Native Install instructions.

The first option is to install the YoWASP WebAssembly based toolchain which is mostly fully cross platform and lives in your python environment. This is the recommended way to install the toolchain on macOS and Windows.

You can do this by running the following inside the Squishy virtual environment:

(squishy.venv) $ pip install yowasp-yosys yowasp-nextpnr-ice40-8k yowasp-nextpnr-ecp5-45k

Native Install

The next option is to do a native install of the toolchain.

Arch Linux

On Arch Linux and Arch-likes, you can install nightly packages which are located in the AUR with an AUR helper or using makepkg directly.

Via an AUR helper like yay

$ yay -S yosys-nightly nextpnr-ice40-nightly nextpnr-ecp5-nightly

Via makepkg directly

$ git clone https://aur.archlinux.org/yosys-nightly.git
$ git clone https://aur.archlinux.org/prjtrellis-nightly.git
$ git clone https://aur.archlinux.org/icestorm-nightly.git
$ git clone https://aur.archlinux.org/nextpnr-ecp5-nightly.git
$ git clone https://aur.archlinux.org/nextpnr-ice40-nightly.git
$ (cd yosys-nightly && makepkg -sic)
$ (cd prjtrellis-nightly && makepkg -sic)
$ (cd icestorm-nightly && makepkg -sic)
$ (cd nextpnr-ecp5-nightly && makepkg -sic)
$ (cd nextpnr-ice40-nightly && makepkg -sic)

Other Linux

Warning

These instructions may be incorrect or incomplete!

With other Linux distributions, it is recommended to use the OSS Cad Suite nightly build. It provides a full environment of all the tools needed built on a nightly basis.

Simply download the latest release for your architecture, extract it to a good home, and then add it to your $PATH

$ curl -LOJ https://github.com/YosysHQ/oss-cad-suite-build/releases/download/2022-04-26/oss-cad-suite-linux-x64-20220426.tgz
$ tar xfv oss-cad-suite-linux-x64-20220426.tgz
$ export PATH="`pwd`/oss-cad-suite/bin:$PATH"

macOS

For macOS systems, it is recommended to use the YoWASP distribution of the toolchain. However if you want to use the native tools, and you are using an Intel based Mac, then the OSS Cad Suite has nightly builds for x86_64 versions of Darwin.

Simply download the latest release for your architecture, extract it to a good home, and then add it to your $PATH

$ curl -LOJ https://github.com/YosysHQ/oss-cad-suite-build/releases/download/2022-04-26/oss-cad-suite-darwin-x64-20220426.tgz
$ tar xfv oss-cad-suite-darwin-x64-20220426.tgz
$ export PATH="`pwd`/oss-cad-suite/bin:$PATH"

Windows

Warning

These instructions may be incorrect or incomplete!

For Windows systems, it is recommended to use the YoWASP distribution of the toolchain. However if you want to use the native tools, and you are using an Intel based PC, then the OSS Cad Suite has nightly builds for x86_64 versions of Windows.

Simply download the latest release for your architecture, extract it to a good home, and then add it to your %PATH%

$ call %cd%\oss-cad-suite\environment.bat

Installing QT5

If you do not plan to use the Squishy GUI, then you can skip this step and continue on.

The Squishy GUI currently uses PySide2, which is a set of python bindings for QT5 to allow for cross-platform GUIs.

It can simply be installed for your system via pip from pypi into the virtual environment like so:

(squishy.venv) $ pip install PySide2

Installing Squishy

Note

Due to some of squishy’s dependencies, we can not publish a PyPi package at the moment.

Once all of the prerequisites are installed, you can finally install squishy into the virtual environment like so:

(squishy.venv) $ pip install 'git+https://github.com/lethalbit/squishy.git'

After that, check to make sure you have the squishy command:

(squishy.venv) $ squishy --help

Now you can view the Getting Started page and start to use Squishy.