Skip to content

Add xdsl benchmark #403

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
Fidget-Spinner merged 7 commits into from
Aug 8, 2025
Merged

Add xdsl benchmark #403

Changes from 1 commit
Commits
Show all changes
7 commits
Select commit Hold shift + click to select a range
468dbd4
Add xdsl constant folding workload
EdmundGoodman Jul 23, 2025
562593a
Improve docstrings
EdmundGoodman Jul 23, 2025
809723d
Apply ruff format
EdmundGoodman Jul 23, 2025
9467543
Delete uv lock file
EdmundGoodman Jul 23, 2025
0c1c146
Align help string with default value of size
EdmundGoodman Jul 23, 2025
122aa86
Bump minimum Python version
EdmundGoodman Jul 23, 2025
287fff3
Pin xDSL version in pyproject.toml
EdmundGoodman Jul 27, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Prev Previous commit
Next Next commit
Improve docstrings
  • Loading branch information
@EdmundGoodman
EdmundGoodman committed Jul 23, 2025
commit 562593a8984c3dc8a39eb47ab6fa896f0920f114
15 changes: 15 additions & 0 deletions pyperformance/data-files/benchmarks/bm_xdsl/run_benchmark.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,6 +4,21 @@
``get_constant_folding_module()`` function, which is then transformed by the
``xdsl.transforms.test_constant_folding.TestConstantFoldingPass`` xDSL
transformation pass.

xDSL is a Python-native compiler framework built around SSA-based intermediate
representations. It re-implements many of MLIR's data structures and methods in
Python. This benchmark exercises the simple pattern rewriting transformation of
constant folding. This is a fair proxy for pattern rewriting transformations in
general, which are a major component of MLIR-like compilers in lowering passes.

Pattern rewriting in both xDSL and MLIR is a pointer-chasing, unstructured
workload, which makes it hard to optimise ahead-of-time. This diminishes the
traditional performance advantage of ahead-of-time compiled languages such as
C++ over dynamic languages such as Python -- making it an interesting benchmark.
More information about the design and impact of this benchmark can be found in
the Master's thesis ``Performance and Dynamism in User-extensible Compiler
Infrastructures'', which is `available on GitHub
<https://github.com/EdmundGoodman/masters-project>`_.
"""

import random
Expand Down