Files
hakmem/docs/archive/NEXT_STEPS_2025_10_22.md
Moe Charm (CI) 52386401b3 Debug Counters Implementation - Clean History
Major Features:
- Debug counter infrastructure for Refill Stage tracking
- Free Pipeline counters (ss_local, ss_remote, tls_sll)
- Diagnostic counters for early return analysis
- Unified larson.sh benchmark runner with profiles
- Phase 6-3 regression analysis documentation

Bug Fixes:
- Fix SuperSlab disabled by default (HAKMEM_TINY_USE_SUPERSLAB)
- Fix profile variable naming consistency
- Add .gitignore patterns for large files

Performance:
- Phase 6-3: 4.79 M ops/s (has OOM risk)
- With SuperSlab: 3.13 M ops/s (+19% improvement)

This is a clean repository without large log files.

🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
2025-11-05 12:31:14 +09:00

34 lines
1.0 KiB
Markdown
Raw Blame History

This file contains ambiguous Unicode characters

This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

# Next Steps Plan (2025-10-22)
## Scope (P1.1)
- Tiny remote-free MPSC + adaptive magazine (DONE baseline; extend drains)
- false sharing mitigation (L2.5 padded locks, stats aligned) (DONE)
- L2 page bundle prefetch (DONE)
- L2/L2.5 mmap-backed bundles (DONE)
## Recommended Next Work
1) Tiny improvements
- Drain MPSC from full-slabs opportunistically during alloc slow path
- Promote slabs from full→free immediately after remote drain
- Make magazine CAP adaptive per site (simple top-K sites)
2) L2/L2.5 bundle policy
- L2: multi-bundle refill when hit/miss ratio degrades in window
- L2.5: demand-zero (MADV_DONTNEED) policy for 512KB/1MB on cold sites
3) Site-bucket policy table (ACE)
- Site×sizeバケットの1バイト方策テーブル読み出し無ロック、更新は窓時
## Env toggles
- Site Rules: `HAKMEM_SITE_RULES=1`
- EVO sampling: `HAKMEM_EVO_SAMPLE=N`
## Notes
- Keep hot path reads lock-free; learning/updates only on window close
- Prefer per-shard/per-site state to avoid global contention