python / python/cpython

Add "Total Machine Code Cost" of UOps to PyStats

Open
#119,692 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

interpreter-core topic-JIT type-feature
Dominant language
Python
Stars
77.2k
Forks
35.9k
PR merge metrics
PR metrics pending

Description

At the sprints at PyConUS 2024, @brandtbucher suggested an improvement to pystats data. Right now, the Tier 2 stats include a count of how many times each UOp is executed, but Brandt suggested another useful table would be UOp Execution Count * Length of UOp in Machine Instructions, sorted by this metric.

In this way, UOps that are extremely common but cheap can be deprioritized, and work can be focused on improving UOps that are less common but take up more 'time' overall (with machine code length as a proxy for time).

I'm happy to take this on.

Linked PRs
  • gh-119693

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Review the CPython Tier 2 PyStats implementation and linked PR gh-119693 first; the issue names no file or test entry point. Done means PyStats reports UOp execution count multiplied by machine-instruction length and sorts entries by that total.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
performance
Issue type
Feature
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Mostly clear
Newbie friendliness
20/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.