Phase 16 v1 NEUTRAL, Phase 17 Case B confirmed, Phase 18 design added

## Phase 16 v1: Front FastLane Alloc LEGACY Direct — NEUTRAL (+0.62%) Target: Reduce alloc-side fixed costs by adding LEGACY direct path to FastLane entry, mirroring Phase 9/10 free-side winning pattern. Result: +0.62% on Mixed (below +1.0% GO threshold) → NEUTRAL, freeze as research box (default OFF). Critical issue: Initial impl crashed (segfault) for C4-C7. Root cause: unified_cache_refill() incompatibility. Safety fix: Limited to C0-C3 only (matching existing dualhot pattern). Files: - core/box/front_fastlane_alloc_legacy_direct_env_box.{h,c} (new) - core/box/front_fastlane_box.h (LEGACY direct path, lines 93-119) - core/bench_profile.h (env refresh sync) - Makefile (new obj) - docs/analysis/PHASE16_*.md (design/results/instructions) ENV: HAKMEM_FRONT_FASTLANE_ALLOC_LEGACY_DIRECT=0 (default OFF, opt-in) Verdict: Research box frozen. Phase 14-16 plateau confirms dispatch/ routing optimization ROI is exhausted post-Phase-6 FastLane collapse. --- ## Phase 17: FORCE_LIBC Gap Validation — Case B Confirmed Purpose: Validate "system malloc faster" observation using same-binary A/B testing to isolate allocator logic差 vs binary layout penalty. Method: - Same-binary toggle: HAKMEM_FORCE_LIBC_ALLOC=0/1 (bench_random_mixed_hakmem) - System binary: bench_random_mixed_system (21K separate binary) - Perf stat: Hardware counter analysis (I-cache, cycles, instructions) Result: **Case B confirmed** — Allocator差 negligible, layout penalty dominates. Gap breakdown (Mixed, 20M iters, ws=400): - hakmem (FORCE_LIBC=0): 48.12M ops/s - libc (FORCE_LIBC=1, same binary): 48.31M ops/s → +0.39% (noise level) - system binary (21K): 83.85M ops/s → +73.57% vs libc, +74.26% vs hakmem Perf stat (200M iters): - I-cache misses: 153K (hakmem) → 68K (system) = -55% (smoking gun) - Cycles: 17.9B → 10.2B = -43% - Instructions: 41.3B → 21.5B = -48% - Binary size: 653K → 21K (30x difference) Root cause: Binary size (30x) causes I-cache thrashing. Code bloat >> algorithmic efficiency. Conclusion: Phase 12's "system malloc 1.6x faster" was real, but misattributed. Gap is layout/I-cache, NOT allocator algorithm. Files: - docs/analysis/PHASE17_*.md (results/instructions) - scripts/run_mixed_10_cleanenv.sh (Phase 9/10 defaults aligned) Next: Phase 18 Hot Text Isolation (layout optimization, not algorithm opt) --- ## Phase 18: Hot Text Isolation — Design Added Purpose: Reduce I-cache misses + instruction footprint via layout control (binary optimization, not allocator algorithm changes). Strategy (v1 → v2 progression): v1 (TU split + hot/cold attrs + optional gc-sections): - Target: +2% throughput (GO threshold, realistic for layout tweaks) - Secondary: I-cache -10%, instructions -5% (direction confirmation) - Risk: Low (reversible via build knob) - Expected: +0-2% (NEUTRAL likely, but validates approach) v2 (BENCH_MINIMAL compile-out): - Target: +10-20% throughput (本命) - Method: Conditional compilation removes stats/ENV/debug from hot path - Expected: Instruction count -30-40% → significant I-cache improvement Files: - docs/analysis/PHASE18_*.md (design/instructions) - CURRENT_TASK.md (Phase 17 complete, Phase 18 v1/v2 plan) Build gate: HOT_TEXT_ISOLATION=0/1 (Makefile knob) Next: Implement Phase 18 v1 (TU split first, BENCH_MINIMAL if v1 NEUTRAL) 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
2025-12-15 05:25:47 +09:00
parent 87fa27518c
commit f8e7cf05b4
14 changed files with 1292 additions and 5 deletions
--- a/core/bench_profile.h
+++ b/core/bench_profile.h
@ -13,6 +13,7 @@
 #include "box/tiny_c7_preserve_header_env_box.h"  // tiny_c7_preserve_header_env_refresh_from_env (Phase 13 v1)
 #include "box/tiny_tcache_env_box.h"  // tiny_tcache_env_refresh_from_env (Phase 14 v1)
 #include "box/tiny_unified_lifo_env_box.h"  // tiny_unified_lifo_env_refresh_from_env (Phase 15 v1)
+#include "box/front_fastlane_alloc_legacy_direct_env_box.h"  // front_fastlane_alloc_legacy_direct_env_refresh_from_env (Phase 16 v1)
 #endif

 // env が未設定のときだけ既定値を入れる
@ -193,5 +194,7 @@ static inline void bench_apply_profile(void) {
 	  tiny_tcache_env_refresh_from_env();
 	  // Phase 15 v1: Sync LIFO ENV cache after bench_profile putenv defaults.
 	  tiny_unified_lifo_env_refresh_from_env();
+	  // Phase 16 v1: Sync LEGACY direct ENV cache after bench_profile putenv defaults.
+	  front_fastlane_alloc_legacy_direct_env_refresh_from_env();
 #endif
 	}
--- a/core/box/front_fastlane_alloc_legacy_direct_env_box.c
+++ b/core/box/front_fastlane_alloc_legacy_direct_env_box.c
@ -0,0 +1,63 @@
+// ============================================================================
+// Phase 16 v1: Front FastLane Alloc LEGACY Direct ENV Box (L0) - Implementation
+// ============================================================================
+
+#include "front_fastlane_alloc_legacy_direct_env_box.h"
+#include <stdlib.h>
+#include <string.h>
+#include <stdio.h>
+#include <unistd.h>
+
+// ============================================================================
+// Global State
+// ============================================================================
+
+_Atomic int g_front_fastlane_alloc_legacy_direct_enabled = -1;
+
+// ============================================================================
+// Init (Cold Path)
+// ============================================================================
+
+int front_fastlane_alloc_legacy_direct_env_init(void) {
+    const char* env = getenv("HAKMEM_FRONT_FASTLANE_ALLOC_LEGACY_DIRECT");
+    int enabled = 0;  // default: OFF (opt-in)
+
+    if (env && (env[0] == '1' || strcmp(env, "true") == 0 || strcmp(env, "TRUE") == 0)) {
+        enabled = 1;
+    }
+
+    // Cache result
+    atomic_store_explicit(&g_front_fastlane_alloc_legacy_direct_enabled, enabled, memory_order_relaxed);
+
+    // Log once (stderr for immediate visibility)
+    if (enabled) {
+        const char msg[] = "[FRONT_FASTLANE_ALLOC_LEGACY_DIRECT] enabled\n";
+        ssize_t w = write(2, msg, sizeof(msg) - 1);
+        (void)w;
+    }
+
+    return enabled;
+}
+
+// ============================================================================
+// Hot Path (LTO Fallback)
+// ============================================================================
+
+// LTO fallback: Non-inline version for cases where LTO can't inline
+int front_fastlane_alloc_legacy_direct_enabled(void) {
+    int val = atomic_load_explicit(&g_front_fastlane_alloc_legacy_direct_enabled, memory_order_relaxed);
+    if (__builtin_expect(val == -1, 0)) {
+        val = front_fastlane_alloc_legacy_direct_env_init();
+    }
+    return val;
+}
+
+// ============================================================================
+// Refresh (Cold Path, called from bench_profile)
+// ============================================================================
+
+void front_fastlane_alloc_legacy_direct_env_refresh_from_env(void) {
+    // Reset to uninitialized state (-1)
+    // Next call to front_fastlane_alloc_legacy_direct_enabled() will re-read ENV
+    atomic_store_explicit(&g_front_fastlane_alloc_legacy_direct_enabled, -1, memory_order_relaxed);
+}
--- a/core/box/front_fastlane_alloc_legacy_direct_env_box.h
+++ b/core/box/front_fastlane_alloc_legacy_direct_env_box.h
@ -0,0 +1,63 @@
+// ============================================================================
+// Phase 16 v1: Front FastLane Alloc LEGACY Direct ENV Box (L0)
+// ============================================================================
+//
+// Purpose: ENV gate for FastLane alloc LEGACY direct path
+//
+// Design: docs/analysis/PHASE16_FRONT_FASTLANE_ALLOC_LEGACY_DIRECT_1_DESIGN.md
+// Instructions: docs/analysis/PHASE16_FRONT_FASTLANE_ALLOC_LEGACY_DIRECT_1_NEXT_INSTRUCTIONS.md
+//
+// Strategy:
+//   - alloc 側の route/policy 固定費を削減
+//   - FastLane 入口で LEGACY を直行（hot → cold → fallback）
+//   - free 側（Phase 9/10）の勝ち筋を alloc にも適用
+//
+// ENV:
+//   HAKMEM_FRONT_FASTLANE_ALLOC_LEGACY_DIRECT=0/1 (default: 0, opt-in)
+//
+// API:
+//   front_fastlane_alloc_legacy_direct_enabled() -> int
+//   front_fastlane_alloc_legacy_direct_env_refresh_from_env()
+//
+// Box Theory:
+//   - L0: This file (ENV gate, reversible)
+//   - L1: front_fastlane_box.h (LEGACY direct early-exit)
+//   - L2: malloc_tiny_fast_for_class (existing fallback)
+//
+// Safety:
+//   - ENV-gated (default OFF, opt-in)
+//   - Reversible (ENV toggle)
+//   - Fail-Fast (direct条件を満たさない場合は既存経路)
+//
+// ============================================================================
+
+#ifndef FRONT_FASTLANE_ALLOC_LEGACY_DIRECT_ENV_BOX_H
+#define FRONT_FASTLANE_ALLOC_LEGACY_DIRECT_ENV_BOX_H
+
+#include <stdatomic.h>
+
+// ============================================================================
+// Global State (L0)
+// ============================================================================
+
+// Cached state: -1 (uninitialized), 0 (disabled), 1 (enabled)
+extern _Atomic int g_front_fastlane_alloc_legacy_direct_enabled;
+
+// ============================================================================
+// Hot API (L0)
+// ============================================================================
+
+// Check if FastLane alloc LEGACY direct is enabled
+// Returns: 1 if enabled, 0 if disabled
+// Note: Implementation in .c file (non-inline for LTO compatibility)
+extern int front_fastlane_alloc_legacy_direct_enabled(void);
+
+// ============================================================================
+// Cold API (L2)
+// ============================================================================
+
+// Refresh ENV cache (called from bench_profile after putenv)
+// Pattern: Same as Phase 8/13/14/15
+extern void front_fastlane_alloc_legacy_direct_env_refresh_from_env(void);
+
+#endif  // FRONT_FASTLANE_ALLOC_LEGACY_DIRECT_ENV_BOX_H
--- a/core/box/front_fastlane_box.h
+++ b/core/box/front_fastlane_box.h
@ -42,6 +42,11 @@
 #include "front_fastlane_stats_box.h"
 #include "../hakmem_tiny.h"                // hak_tiny_size_to_class, tiny_get_max_size
 #include "../front/malloc_tiny_fast.h"     // malloc_tiny_fast_for_class
+#include "front_fastlane_alloc_legacy_direct_env_box.h"  // Phase 16 v1: LEGACY direct
+#include "tiny_static_route_box.h"         // tiny_static_route_ready_fast, tiny_static_route_get_kind_fast
+#include "tiny_front_hot_box.h"            // tiny_hot_alloc_fast
+#include "tiny_front_cold_box.h"           // tiny_cold_refill_and_alloc
+#include "smallobject_policy_v7_box.h"     // SMALL_ROUTE_LEGACY

 // FastLane is only safe after global init completes.
 // Before init, wrappers must handle recursion guards + syscall init.
@ -85,6 +90,34 @@ static inline void* front_fastlane_try_malloc(size_t size) {
        return NULL;  // Class not enabled → fallback
    }

+    // Phase 16 v1: LEGACY direct path (early-exit optimization)
+    // Try direct allocation for LEGACY routes only (skip route/policy overhead)
+    // TEMPORARY SAFETY: Limit to C0-C3 (match dualhot pattern) until refill issue debugged
+    if (__builtin_expect(front_fastlane_alloc_legacy_direct_enabled() && (unsigned)class_idx <= 3u, 0)) {
+        // Condition 1: Static route must be ready (Learner interlock check)
+        // Condition 2: Route must be LEGACY (断定可能な場合のみ)
+        if (tiny_static_route_ready_fast() &&
+            tiny_static_route_get_kind_fast(class_idx) == SMALL_ROUTE_LEGACY) {
+
+            // Hot path: Try UnifiedCache first
+            void* ptr = tiny_hot_alloc_fast(class_idx);
+            if (__builtin_expect(ptr != NULL, 1)) {
+                FRONT_FASTLANE_STAT_INC(malloc_hit);
+                return ptr;  // Success (cache hit)
+            }
+
+            // Cold path: Refill UnifiedCache and retry
+            ptr = tiny_cold_refill_and_alloc(class_idx);
+            if (__builtin_expect(ptr != NULL, 1)) {
+                FRONT_FASTLANE_STAT_INC(malloc_hit);
+                return ptr;  // Success (after refill)
+            }
+
+            // Fallback: Direct path failed → use existing route (safety)
+            // This handles edge cases (Learner transition, policy changes, etc.)
+        }
+    }
+
    // Call existing hot handler (no duplication)
    // This is the winning path from E5-4 / Phase 4 E2
    void* ptr = malloc_tiny_fast_for_class(size, class_idx);