Phase v11a-4: Mixed本線ベンチマーク結果追加

Results:
- C6-heavy (257-512B): +5.1% (34.0M → 35.8M ops/s)
- Mixed 16-1024B:      +4.4% (38.6M → 40.3M ops/s)

Conclusion: Mixed本線で C6→MID v3.5 は採用候補。
予測(+1-3%)を上回る +4-5% の改善を確認。

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

CURRENT_TASK.md

@@ -1,37 +1,35 @@
+# Phase v11a-4: MID v3.5 Mixed本線テスト - COMPLETED
+
+## Status: ✅ COMPLETE - C6→MID v3.5 採用候補
+
+### 結果サマリ
+
+| Workload | v3.5 OFF | v3.5 ON | 改善 |
+|----------|----------|---------|------|
+| C6-heavy (257-512B) | 34.0M | 35.8M | **+5.1%** |
+| Mixed 16-1024B | 38.6M | 40.3M | **+4.4%** |
+
+### 結論
+
+**Mixed本線で C6→MID v3.5 は採用候補**。+4%の改善があり、設計の一貫性（統一セグメント管理）も得られる。
+
+---
+
 # Phase v11a-3: MID v3.5 Activation - COMPLETED
 
-## Status: BUILD COMPLETE - Ready for Benchmarking
+## Status: ✅ COMPLETE
 
-Phase v11a-3 successfully integrated MID v3.5 into the active code path. The implementation is ready for A/B benchmarking.
+### Bug Fixes
+1. **Policy infinite loop**: CAS で global version を 1 に初期化
+2. **Malloc recursion**: segment creation で mmap 直叩きに変更
 
-## Implementation Details
-
-See: `docs/analysis/PHASE_V11A3_IMPLEMENTATION_SUMMARY.md`
-
-### Tasks Completed (4/6)
+### Tasks Completed (6/6)
 1. ✅ Add MID_V35 route kind to Policy Box
 2. ✅ Implement MID v3.5 HotBox alloc/free
 3. ✅ Wire MID v3.5 into Front Gate
 4. ✅ Update Makefile and build
-
-### Pending Tasks (2/6)
-5. ⏳ Run A/B benchmarks (C6-heavy, C5+C6-only, Mixed 16-1024B)
-6. ⏳ Update documentation with benchmark results
-
-## Quick Start (Benchmarking)
-
-```bash
-# Enable MID v3.5
-export HAKMEM_MID_V35_ENABLED=1
-export HAKMEM_MID_V35_CLASSES=0x60  # C5+C6
-
-# C6-heavy benchmark
-./bench_mid_large_mt_hakmem 1 1000000 400 1
-
-# C5+C6-only benchmark
-HAKMEM_BENCH_MIN_SIZE=257 HAKMEM_BENCH_MAX_SIZE=768 \
-./bench_random_mixed_hakmem 1000000 400 1
-```
+5. ✅ Run A/B benchmarks
+6. ✅ Update documentation
 
 ---
 

docs/analysis/PHASE_V11A_DESIGN_MID_V3.5.md

@@ -235,10 +235,27 @@ C7 heavy (800-1000B):               +2-3% (vs ULTRA fallback)
 Mixed 16-1024B (with Learner):     +1-2% (dynamic routing)
 ```
 
+### Actual MID v3.5 Results (Phase v11a-4)
+
+**C6-heavy (257-512B)**:
+```
+v3.5 OFF: 34.0M ops/s
+v3.5 ON:  35.8M ops/s  (+5.1%)
+```
+
+**Mixed 16-1024B (ws=400, 10M iters, avg of 3 runs)**:
+```
+v3.5 OFF: 38.6M ops/s
+v3.5 ON:  40.3M ops/s  (+4.4%)
+```
+
+**所感**: C6-heavy では予測通り +5%、Mixed でも +4% の改善が確認できた。
+予測より良い結果。Mixed 本線で C6→MID v3.5 は採用候補として有効。
+
 **メトリクス**:
-- Throughput: +1-3% overall
-- Overhead: ~5-8% (relative to ULTRA baseline)
-- Learner accuracy: > 95% on workload pattern detection
+- Throughput: +4-5% (予測+1-3% を上回る)
+- Overhead: 測定なし（mmap 直叩きで回避）
+- Learner accuracy: 観測モードのみ（route 切替は将来フェーズ）
 
 ## 6. 設計確定事項