bench-compare

name	bench-compare
description	Use when measuring optimization impact against a baseline, when validating that a code change didn't regress performance, or when comparing two implementation approaches. Criterion benchmark baseline comparison workflow.

Benchmark Comparison Workflow

Compare benchmark results against a baseline to measure optimization impact.

Usage

Invoke with optional benchmark filter:

• /bench-compare - Run all benchmarks
• /bench-compare scan - Run only scan-related benchmarks
• /bench-compare throughput - Run throughput benchmarks

Workflow

1. Establish Baseline

Before making changes, save current benchmark results:

cargo bench --bench <name> -- --save-baseline before

Or for all benchmarks:

cargo bench -- --save-baseline before

2. Apply Changes

Make the optimization changes to the codebase.

3. Compare Results

Run benchmarks against the saved baseline:

cargo bench --bench <name> -- --baseline before

4. Analyze Output

Look for these patterns in Criterion output:

• Performance has improved - Optimization successful
• Performance has regressed - Changes hurt performance
• No change in performance - Within noise threshold

5. Report Summary

Provide a summary table:

Benchmark	Before	After	Change