PLEASE NOTE! THIS REPO IS DEPRECATED
Head over to https://github.com/vmprof instead!
The code is now hosted on GitHub, and split into three parts. The client Python library (vmprof-python), the server (vmprof-server) and a repo to do integration testing!
You need to use PyPy to run this. To get started, using a recent virtualenv (1.6.1 or newer), virtualenvwrapper, and a recent PyPy.
PyPy versions correspond to jitviewer tags, so pypy-1.9 tag in jitviewer means it works with pypy 1.9.
On Mac OSX you will also need to install binutils, to make objdump available.
To create a virtualenv (note that mkvirtualenv only accepts absolute paths):
mkvirtualenv --python=/path/to/pypy pypy-viewer
Now install the dependencies:
pip install -r requirements.txt
python setup.py develop
It also requires pypy to be importable (as in source code), you can do this by setting your PYTHONPATH enviroment variable. Make sure that this source code is (roughly) the same version as the binary pypy that produced the log file.
For usage information, invoke jitviewer.py with --help.
An example log file is shipped with the jitviewer source code.