Installation¶
Docker (Recommended)¶
The easiest way to get started with Pyda is using Docker.
Pull the latest release¶
Or build it yourself¶
Note
The Pyda image is currently based on ubuntu:22.04
and we leave the default entrypoint as /bin/bash
Experimental pip install (macOS and Linux)¶
Warning
macOS support is extremely experimental and may not work. See "a note on macOS" below.
Installation with pip may take ~1-2 minutes to complete, as it builds everything from source.
Pyda currently only supports CPython 3.10.
Supported Platforms¶
- Operating System: Linux, macOS (experimental)
- Architecture: X86_64, ARM64
- Python: CPython 3.10
First use¶
Once installed, you can verify your installation works by running a simple example:
# If you used Docker...
docker run -it ghcr.io/ndrewh/pyda pyda examples/ltrace.py -- ls -al
# If you used pip install...
pyda examples/ltrace.py -- ls -al
If the installation is successful, you should see ltrace-style output showing library function calls made by the ls
command.
A note on macOS¶
On macOS, certain security mechanisms interfere with Pyda. Pyda (and any other DynamoRIO-based tool) will silently fail to attach to many code-signed processes.
- If you just want to try Pyda, try it on something you compiled on your own machine or (if you know what you're doing) resign your target with the
codesign
utility. - If you want to run system / App Store applications under Pyda, consider disabling SIP. This is completely untested, though.
Next Steps¶
Once you have Pyda installed, head over to the Quick Start guide to learn how to write your first analysis tool.