Phase 2 B1/B3/B4 preparation: Analysis & ENV gate setup

## Phase 2 Optimization Research Complete

### B1 (Header tax reduction v2) - NO-GO
- HAKMEM_TINY_HEADER_MODE=LIGHT: -2.54% regression on Mixed
- Decision: FREEZE as research box (ENV opt-in only)

### B3 (Routing branch shape optimization) - ADOPT
- Mixed: +2.89% (48.41M → 49.80M ops/s)
- C6-heavy: +9.13% (8.97M → 9.79M ops/s)
- Strategy: LIKELY on LEGACY (hot), noinline,cold helper for rare routes
- Implementation: Already in malloc_tiny_fast.h:252-267
- Profile updates: HAKMEM_TINY_ALLOC_ROUTE_SHAPE=1 now default

### B4 (Wrapper Layer Hot/Cold Split) - Preparation
- Design memo: docs/analysis/PHASE2_B4_WRAPPER_SHAPE_1_DESIGN.md
- Goal: Split malloc/free into hot/cold paths, reduce I-cache pressure
- ENV gate: HAKMEM_WRAP_SHAPE=0/1 (added to wrapper_env_box)
- Expected gain: +2-5% Mixed, +1-3% C6-heavy

## Analysis Summary
- Background is visible: FREE DUALHOT + B3 routing optimizations work
- Code layering is clean: winning boxes promoted to presets, losing boxes frozen with ENV guards
- Remaining gap to mimalloc is wrapper layer + safety checks + policy snapshot
- Further +5-10% still realistically achievable

🤖 Generated with Claude Code

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

CURRENT_TASK.md

@@ -96,51 +96,49 @@
 
 ---
 
-## 次の攻め先: mimalloc Gap Closure Roadmap (2.5x → 1.9x)
+## 次の攻め先: Phase 2 B4（Wrapper Layer Hot/Cold Split）
 
-**Gap Analysis**: hakmem 50.7M ops/s vs mimalloc 127M ops/s
+**背中は見えてきた段階**: B1 NO-GO / B3 ADOPT で、ホットスポットが明確。
+- 勝ち箱（FREE DUALHOT + B3）は本線プリセット昇格済み
+- 負け箱（B1）は ENV ガード freeze で本線汚さず
 
-根本原因（ROI順）:
+**残りの主因**: wrapper 層（malloc/free）+ 安全チェック + policy snapshot
-1. **Observation tax** (+2-3%): Stats macros branch even when OFF
+- mimalloc 同等までは難しいが、さらに数%～十数%は詰められる見込み
-2. **Policy snapshot** (+10-15%): Per-call TLS policy read + atomic sync
-3. **Header management** (+5-10%): 1-byte header per block
-4. **Wrapper layer** (+5-10%): malloc → tiny_alloc_gate_fast + security checks
-5. **Routing switch** (+3-5%): Per-call switch statement
 
-### Phase 1: Quick Wins (Week 1) - Target: +4-7% (52-56M ops/s)
+### Phase 2 B4（推奨）: WRAPPER-SHAPE-1（malloc/free のホット整形）
 
-**優先度 A1** - FREE 勝ち箱の本線昇格:
+**設計メモ**: `docs/analysis/PHASE2_B4_WRAPPER_SHAPE_1_DESIGN.md`
-- HAKMEM_FREE_TINY_FAST_HOTCOLD=1 を MIXED_TINYV3_C7_SAFE default
-- FREE-TINY-FAST-DUALHOT-1 のデフォルト有効化
-- Expected: +2-3% (DUALHOT 効果は既に測定済み +13%)
 
-**優先度 A2** - 観測税ゼロ化 (Compile-out stats):
+**狙い**:
-- Add HAKMEM_DEBUG_COUNTERS compile-time flag (default 0)
+- wrapper 入口の "稀なチェック"（LD mode、jemalloc、診断）を `noinline,cold` に押し出す
-- When 0: `#define ALLOC_GATE_STAT_INC(x) do {} while(0)` (zero cost)
+- ホット側は NULL check → Tiny fast → 即 return（最短経路）
-- Files: `alloc_gate_stats_box.h`, `free_path_stats_box.h`, `tiny_front_stats_box.h`, `free_tiny_fast_hotcold_stats_box.h`
+- I-cache 削減 + 分岐予測改善
-- Expected: +2-3% (eliminate branching on all stats)
 
-**優先度 A3** - Inline header write:
+**実装**:
-- Add `__attribute__((always_inline))` to `tiny_region_id_write_header()`
+- ENV gate: `HAKMEM_WRAP_SHAPE=0/1`（default OFF）
-- Eliminate function call overhead in hot path
+- malloc hot/cold 分割（core/box/hak_wrappers.inc.h）
-- Expected: +1-2%
+- free hot/cold 分割（core/box/hak_wrappers.inc.h）
 
-### Phase 2: Structural Changes (Weeks 2-3) - Target: +5-10% (55-61M ops/s)
+**A/B テスト**:
+- Mixed: 10-run（中央値）
+- C6-heavy: 5-run（平均）
+- GO条件: Mixed +1% 以上 → プリセット昇格
+- NO-GO条件: -1% 以下 → freeze, ENV opt-in のまま
 
-**優先度 B1** - C4-C7 header tax削減:
+**期待ゲイン**: Mixed +2-5%, C6-heavy +1-3%
-- Remove 1-byte header for C6 (512B) / C7 (1024B) allocations
-- Use registry-only lookup on free
-- Expected: +3-5% (C6/C7 = 30% of workload, no header = 10% size savings)
 
-**優先度 B2** - C0-C3 専用 fast path:
+### Phase 1: Quick Wins（完了）
-- Create `malloc_tiny_fast_c0c3()` entry point (no policy snapshot)
-- Conditional dispatch from wrapper based on size
-- Expected: +1-2%
 
-**優先度 B3** - Routing jump table:
+- ✅ **A1（FREE 勝ち箱の本線昇格）**: `MIXED_TINYV3_C7_SAFE` で `HAKMEM_FREE_TINY_FAST_HOTCOLD=1` を default 化（ADOPT）
-- Replace switch(route_kind) with function pointer array
+- ✅ **A2（観測税ゼロ化）**: `HAKMEM_DEBUG_COUNTERS=0` のとき stats を compile-out（ADOPT）
-- Reduce branch prediction misses (5-way switch → direct dispatch)
+- ❌ **A3（always_inline header）**: Mixed -4% 回帰のため NO-GO → research box freeze（`docs/analysis/TINY_HEADER_WRITE_ALWAYS_INLINE_A3_DESIGN.md`）
-- Expected: +1-3%
+
+### Phase 2: Structural Changes（進行中）
+
+- ❌ **B1（Header tax 削減 v2）**: `HAKMEM_TINY_HEADER_MODE=LIGHT` は Mixed -2.54% → NO-GO / freeze（`docs/analysis/PHASE2_B1_HEADER_TAX_AB_TEST_RESULTS.md`）
+- ✅ **B3（Routing 分岐形最適化）**: `HAKMEM_TINY_ALLOC_ROUTE_SHAPE=1` は Mixed +2.89% / C6-heavy +9.13% → ADOPT（プリセット default=1）
+- ▶ **B4（次の芯）**: WRAPPER-SHAPE-1（malloc/free のホット整形。`docs/analysis/PHASE2_B4_WRAPPER_SHAPE_1_DESIGN.md`）
+- （保留）**B2**: C0–C3 専用 alloc fast path（入口短絡は回帰リスク高。B4 の後に判断）
 
 ### Phase 3: Cache Locality (Weeks 4-5) - Target: +12-22% (57-68M ops/s)
 

core/box/wrapper_env_box.c

@@ -3,7 +3,7 @@
 #include <stdlib.h>
 #include <string.h>
 
-wrapper_env_cfg_t g_wrapper_env = {.inited = 0, .step_trace = 0, .ld_safe_mode = 1, .free_wrap_trace = 0, .wrap_diag = 0};
+wrapper_env_cfg_t g_wrapper_env = {.inited = 0, .step_trace = 0, .ld_safe_mode = 1, .free_wrap_trace = 0, .wrap_diag = 0, .wrap_shape = 0};
 
 static inline int env_flag(const char* name, int def) {
     const char* e = getenv(name);
@@ -40,6 +40,7 @@ void wrapper_env_init_once(void) {
     g_wrapper_env.ld_safe_mode    = env_int("HAKMEM_LD_SAFE", 1);
     g_wrapper_env.free_wrap_trace = env_flag("HAKMEM_FREE_WRAP_TRACE", 0);
     g_wrapper_env.wrap_diag       = env_flag("HAKMEM_WRAP_DIAG", 0);
+    g_wrapper_env.wrap_shape      = env_flag("HAKMEM_WRAP_SHAPE", 0);
 
     // Mark as initialized last with memory barrier
     atomic_store_explicit(&g_wrapper_env.inited, 1, memory_order_release);

core/box/wrapper_env_box.h

@@ -10,6 +10,7 @@ typedef struct {
     int ld_safe_mode;        // HAKMEM_LD_SAFE (default: 1)
     int free_wrap_trace;     // HAKMEM_FREE_WRAP_TRACE (default: 0)
     int wrap_diag;           // HAKMEM_WRAP_DIAG (default: 0) - log first few libc fallbacks
+    int wrap_shape;          // HAKMEM_WRAP_SHAPE (default: 0) - Phase 2 B4: malloc/free hot/cold split
 } wrapper_env_cfg_t;
 
 extern wrapper_env_cfg_t g_wrapper_env;