Support --fail-fast across workers via file-based signal

[MatthewMckee4]

Summary

• Workers now write a fail-fast signal file to the cache directory when --fail-fast is enabled and a test fails
• The main process polls for this signal during its worker wait loop and kills remaining workers immediately when detected
• Removes the "only works in parallel" limitation from --fail-fast docs

Test plan

• Added test_failfast_across_files integration test verifying fail-fast stops across multiple test files
• All 677 existing tests pass
• Pre-commit checks pass

Closes #472

[codspeed-hq]

Merging this PR will degrade performance by 10.2%

⚠️ Unknown Walltime execution environment detected

Using the Walltime instrument on standard Hosted Runners will lead to inconsistent data.

For the most accurate results, we recommend using CodSpeed Macro Runners: bare-metal machines fine-tuned for performance measurement consistency.

❌ 1 (👁 1) regressed benchmark

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
👁	WallTime	karva_benchmark	22.3 s	24.9 s	-10.2%

---

_{Comparing support-fail-fast-serial-mode (b3c84fb) with main (15d45a9)}