Skip to content

Generic wrapper tool around open source EDA tools that executes Docker container

Notifications You must be signed in to change notification settings

librecores/eda-container-wrapper

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

13 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

EDA Container Wrapper

EDA container wrapper is a work-in-progress generic wrapper tool around open source EDA tools. The tools are executed in Docker containers and this tool wraps this execution and keeps an inventory of tools.

It supports three use cases:

  1. Run eda-container-wrapper to execute the wrapper with given parameters
  2. Let eda-container-wrapper create a wrapper script (in an executable path for example) with pre-defined settings
  3. Run tools from python using the edacontainerwrapper module

The settings can be set in the following way in precedence order:

  1. As parameters to the eda-container-wrapper tool or to the respective functions of edacontainerwrapper
  2. As environment variables
  3. As default settings in the wrapper scripts

Installation

$ pip3 install eda-container-wrapper

Supported settings

Tool version (--tool-version or ECW_TOOL_VERSION, default: depends on tool)

Sets the version of the tool to use.

Interactive (--non-interactive or ÈCW_INTERACTIVE, default: interactive)

Runs the Docker container in interactive mode (allowing to terminate it easily for example) with the Docker -ti flags. Some non-interactive environments such as CI don't support that (missing tty). Note the difference in logic, the default is interactive mode which is disabled with the --non-interactive switch. ECW_INTERACTIVE keeps it interactive when set to true or 1 and it will be non-interactive otherwise.

Current work directory base (--cwd-base or ECW_CWD_BASE, default: not set)

If not empty this is a colon-separated pair of a leading path of the cwd where the tool was called from and a replacement for this path. This is in particular useful when called from inside a Docker container, such as in CI. The syntax is: <actual path>:<cwd path>.

Split the current working directory (--split-cwd-tail or ECW_SPLIT_CWD_TAIL, default: 0)

When mapping the current working directory, this setting actually maps the path up the hierarchy by the given value. This is in particular useful when you need to access data relative to the current working directory that is in upper and sibbling folders.

For example, when started from the following folder:

CWD=/path/to/my/project/build

Setting --split-cwd-tail=1 the split is into /path/to/my/project and build. Each tool has a "project path" that is the volume where the tool is executed in then by setting the working directory. For a tool with the project path /project it will then mount /path/to/my/project to /project and the workdir will be /project/build.

This split is executed after processing the cwd base.

Running a Tool

As described above there are different ways to run a tool, that are described in the following.

Run eda-container-wrapper

The program is called with the parameters as described and the toolname. Following a -- parameters to the tool can be supplied.

Example:

$ eda-container-wrapper verilator -- --version

Create wrapper

To create a verilator script that by default executes version 4.100:

$ eda-container-wrapper --write-script=/usr/local/bin/verilator --tool-version 4.100 verilator

and then execute it:

$ verilator --version
Verilator 4.100 2020-09-07 rev v4.100

Setting the environment setting ECW_TOOL_VERSION changes the defualt behavior:

$ ECW_TOOL_VERSION=4.102 verilator --version
Verilator 4.102 2020-10-15 rev v4.102

It can be useful to create the script in your virtual environment:

$ eda-container-wrapper --write-script=$VIRTUAL_ENV/verilator --tool-version 4.100 verilator

About

Generic wrapper tool around open source EDA tools that executes Docker container

Resources

Stars

Watchers

Forks

Packages

No packages published

Languages