Add CodSpeed performance benchmarks

[codspeed-hq]

This PR sets up continuous performance tracking for the uutils tar project using CodSpeed.

What's included

Benchmark suite (benches/bench_tar.rs)

Seven benchmarks covering the core tar operations, using divan with the CodSpeed compatibility layer:

• Archive creation: create_archive_10_files, create_archive_100_files, create_archive_directory -- measures file-level and directory-level archive creation at different scales.
• Archive listing: list_archive_50_files, list_archive_verbose_50_files -- measures entry enumeration in simple and verbose mode.
• Archive extraction: extract_archive_20_files -- measures full extraction to disk.
• CLI parsing: build_cli_command -- measures the cost of constructing the clap command definition.

Each benchmark creates temporary sample files as setup and exercises the real uu_tar operation functions directly.

CI workflow (.github/workflows/codspeed.yml)

A GitHub Actions workflow that runs benchmarks on every push to main and on pull requests, using CodSpeed's simulation mode (CPU instruction counting). This provides deterministic, low-variance measurements regardless of CI runner load.

The workflow also supports workflow_dispatch to allow CodSpeed to trigger backtesting for initial baseline data.

Other changes

• src/uu/tar/src/tar.rs: The operations module is now pub so the benchmark suite can call create_archive, list_archive, and extract_archive directly.
• Cargo.toml: Added codspeed-divan-compat as a dev-dependency and the [[bench]] target.
• README.md: Added CodSpeed badge.

Next steps

Once this PR is merged and the benchmarks run on main, CodSpeed will automatically:

• Report performance changes on future pull requests.
• Generate flamegraphs for each benchmark to help diagnose regressions.
• Track historical performance trends on the CodSpeed dashboard.

[codspeed-hq]

Congrats! CodSpeed is installed 🎉

🆕 7 new benchmarks were detected.

You will start to see performance impacts in the reports once the benchmarks are run from your default branch.

Detected benchmarks

• create_archive_directory: 171.7 µs
• build_cli_command: 34.7 µs
• create_archive_100_files: 505.2 µs
• extract_archive_20_files: 624.6 µs
• create_archive_10_files: 64.8 µs
• list_archive_verbose_50_files: 1.2 ms
• list_archive_50_files: 1.1 ms

---