Measuring performance

Benchmarking framework

BuildStream has a utility to measure performance which is available from a separate repository at This tool allows you to run a fixed set of workloads with multiple versions of BuildStream. From this you can see whether one version performs better or worse than another which is useful when looking for regressions and when testing potential optimizations.

For full documentation on how to use the benchmarking tool see the README in the ‘benchmarks’ repository.

Profiling tools

When looking for ways to speed up the code you should make use of a profiling tool.

Python provides cProfile which gives you a list of all functions called during execution and how much time was spent in each function. Here is an example of running bst --help under cProfile:

python3 -m cProfile -o bst.cprofile – $(which bst) –help

You can then analyze the results interactively using the ‘pstats’ module:

python3 -m pstats ./bst.cprofile

For more detailed documentation of cProfile and ‘pstats’, see:

For a richer and interactive visualisation of the .cprofile files, you can try snakeviz. You can install it with pip install snakeviz. Here is an example invocation:

snakeviz bst.cprofile

It will then start a webserver and launch a browser to the relevant page.


If you want to also profile cython files, you will need to add BST_CYTHON_PROFILE=1 and recompile the cython files. pip install would take care of that.

Profiling specific parts of BuildStream with BST_PROFILE

BuildStream can also turn on cProfile for specific parts of execution using BST_PROFILE.

BST_PROFILE can be set to a section name, or a list of section names separated by “:”. You can also use “all” for getting all profiles at the same time. There is a list of topics in src/buildstream/ For example, running:

BST_PROFILE=load-pipeline bst build bootstrap-system-x86.bst

will produce a profile in the current directory for the time take to call most of initialized, for each element. These profile files are in the same cProfile format as those mentioned in the previous section, and can be analysed in the same way.

Fixing performance issues

BuildStream uses Cython in order to speed up specific parts of the code base.


When optimizing for performance, please ensure that you optimize the algorithms before jumping into Cython code. Cython will make the code harder to maintain and less accessible to all developers.

When adding a new cython file to the codebase, you will need to register it in

For example, for a module buildstream._my_module, to be written in src/buildstream/_my_module.pyx, you would do:


In and the build tool will automatically use your module.


Please register cython modules at the same place as the others.

When adding a definition class to share cython symbols between modules, please document the implementation file and only expose the required definitions.