benchmark: Add script to diagnose hypothesis performance regression (#32693)

[davidweb3-ctrl]

Description

This PR addresses the performance regression reported in #32693 where hypothesis 6.103.1 is significantly slower than 6.47.0.

Root Cause Analysis

Through benchmarking, I identified that the main bottleneck is st.data() with multiple data.draw() calls, which became ~83% slower in newer hypothesis versions:

Version	st.data() multiple draws	Total
6.47.0	0.423s	0.968s
6.103.1	0.775s	1.663s
6.151.9	0.670s	1.363s

Changes

• Added benchmark_hypothesis.py: Standalone script to measure hypothesis performance across versions
• Added selfdrive/car/tests/benchmark_hypothesis_performance.py: Car-specific benchmark

Usage

# Test old version
pip install hypothesis==6.47.0
python benchmark_hypothesis.py

# Test new version
pip install hypothesis==6.103.1
python benchmark_hypothesis.py

Bounty

Contributes to resolving #32693 ($200 bounty)

[github-actions]

This PR has had no activity for 24 days. It will be automatically closed in 7 days if there is no activity.

[adeebshihadeh]

This is missing:

and a solution for optimizing/fixing either the hypothesis library itself, or openpilot to improve the example generation times

[github-actions]

This PR has had no activity for 24 days. It will be automatically closed in 7 days if there is no activity.

[github-actions]

This PR has been automatically closed due to inactivity. Feel free to re-open once activity resumes.