Git Product home page Git Product logo

padrick's Introduction

Padrick - A flexible Padframe Generator for SoCs

State of Development

WARNING Padrick is still in it’s very early stage of development and not yet ready for production. Do not yet use it for actual tapeouts without asking the author about the current state of development.

Purpose of this Tool

Modern SoCs contain a large number of peripherals and functional blocks that need to communicate with the outside world. Most of the time, the number of IO pins required to map every IO facing port of the whole SoC to dedicated IO pads is infeasible due to limited available area dedicated for IO pads. Instead, most modern SoCs use multiplexing logic to use a single IO pad for several different functionalities according to some user-programmable configuration register. While some SoCs (e.g. Nordic nRF52 series of BLE chips) allow mapping of every module port (e.g. SPI MOSI signal) to every available IO pad in a full crossbar fashion, traditionally the PULP chips taped out so far only used the approach to assign 2-3 Module ports to one dedicated IO pad. I.e. the user can chose to use e.g. IO Pad 43 as either GPIO4 or I2C_SCK but the SPI_MOSI signal for example cannot be routed to IO Pad 43. While the full-crossbar like approach can have its demerits (e.g. routing can become trickier if very fast IO signals are involfed), the full-crossbar routing approach makes the resulting SoC way more flexible for dynamic adaptation to the workload and simplifies a lot the PCB design process and repurposing of existing PCBs by means of rerouting the IO signals.

Since the padframe is always a custom tailored component for one particular SoCs, a lot of time has been spent in the past writing this padframe multiplexing and routing logic. This design process is very labouruos and extremely error prone which is where Padrick enters the stage;

Padrick aims to solve the problem of painstakingly writing the multiplexing logic and interconnection logic for padframes by hand. Not only does it (optionally) allow full-crossbar like routing of every periphery port to every IO pad it also generates C-drivers and Documentation to interact with the auto-generated peripherals.

How does it work? The user provides a YAML description of the desired Padrame configuration. The configuration file’s syntax is tailored in a way such that large portions of it can be copy pasted or even entirely reused from one SoC to the other while only minor modifications of e.g. the involved IO pads need to be performed. After parsing and internaly validating the configuration file with a couple of sanity checks, Padrick generates several SystemVerilog modules that instantiate the desired IO Pads, implements the IO Multiplexing logic and exposes the SoC facing IO signal and either an AXI-lite or APB interface to be attached to the SoC configuration bus for at-runtime control over the IO Pads.

Getting Started

Download Padrick Binary

You can find the latest binary x86 release on the github release page.

Use the following snippet to download the appimage in your current path:

curl https://api.github.com/repos/pulp-platform/padrick/releases/latest \
    | grep "Padrick-x86_64.AppImage" \
    | cut -d : -f 2,3 \
    | tr -d \" \
    | wget -qi -
mv Padrick-x86_64.AppImage padrick
chmod a+x padrick

Now you can directly start using the downloaded binary. E.g. use this command to show the built-in help::

./padrick --help

Installation with Python pip

If you want to modify padrick yourself or don’t want to use one of the binary releases, installation of padrick is straight forward:

pip install git+ssh://[email protected]:pulp-platform/padrick.git

Or if you prefer an https connection over SSH:

pip install git+https://github.com/pulp-platform/padrick

If you plan to modify or frequently update padrick you might want to install it with the pip editable flag so changes to the source code of padrick take effect immediately to all Python environments were you installed padrick.

git clone https://github.com/pulp-platform/padrick
pip install -e ./padrick

Both installation schemes will automatically install all the required dependencies. This will install the new command line tool padrick.

Write the Padframe Configuration File

Now that you installed padrick, you can start writing your padframe configuration file. Have a look at some of the examples in the examples folder to become familiar with the syntax and read the documentation.

Generating the Padframe

With your padframe configuration file ready you can now invoke the padrick command to generate the padframe project:

padrick generate rtl -output my_padframe my_padframe_config.yml

This command will first parse your configuration file and pass it through a number of sanity checks to ensure it has the proper format. If there is a mistake in the config file, you will see an error message with hints to where the error occured.

After successfully parsing the config file, padrick will generate a complete RTL project in the subdirectory my_padframe. If the folder doesn’t exist, it will be created for you. Amongst the various SystemVerilog RTL files, the folder will also be populated with appropriate Bender and IPApprox IP configuration file to simplify the integration of the auto-generated padframe.

Instantiating the Padframe within your Design

Now that your padframe IP is generated you can instantiate it within your design. The toplevel module for user instantiation is called <padframe.name>_top.sv. The structs that make up the interface of this SystemVerilog module are defined in pkg_<padframe.name>.sv.

Documentation

For more in-depth documentation of Padrick, visit the project Documentation site at https://padrick.readthedocs.io/en/latest/

padrick's People

Contributors

meggiman avatar adimauro-iis avatar micprog avatar

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.