Files
hakmem/core/box/smallobject_v5_env_box.h

141 lines
4.7 KiB
C
Raw Normal View History

// smallobject_v5_env_box.h - SmallObject v5 環境ゲートPhase v5-0
//
// ENV ベース: HAKMEM_SMALL_HEAP_V5_ENABLED, HAKMEM_SMALL_HEAP_V5_CLASSES
#ifndef HAKMEM_SMALLOBJECT_V5_ENV_BOX_H
#define HAKMEM_SMALLOBJECT_V5_ENV_BOX_H
#include <stdlib.h>
#include <string.h>
Phase v5-0 refactor: ENV統一・マクロ化・構造体最適化 - ENV initialization を sentinel パターンで統一 - ENV_UNINIT/ENABLED/DISABLED 定数追加 - __builtin_expect で初期化チェックを最適化 - small_heap_v5_enabled/class_mask を統一パターンに変更 - ポインタマクロ化(O(1) segment/page 計算) - SMALL_SEGMENT_V5_BASE_FROM_PTR: ptr から segment base を mask で計算 - SMALL_SEGMENT_V5_PAGE_IDX: segment 内の page_idx を shift で計算 - SMALL_SEGMENT_V5_PAGE_META: page_meta への O(1) access(bounds check付き) - SMALL_SEGMENT_V5_VALIDATE_MAGIC: magic 検証 - SMALL_SEGMENT_V5_VALIDATE_PTR: Fail-Fast validation pipeline - SmallClassHeapV5 に partial_count 追加 - partial ページリストのカウンタを追加(refill/retire 最適化用) - SmallPageMetaV5 の field 再配置(L1 cache 最適化) - hot fields (free_list, used, capacity) を先頭に集約 - metadata (class_idx, flags, page_idx, segment) を後方配置 - total 24B、offset コメント追加 - route priority ENV 追加 - HAKMEM_ROUTE_PRIORITY={v4|v5|auto}(default: v4) - enum small_route_priority 定義 - small_route_priority() 関数追加 - segment_size override ENV 追加 - HAKMEM_SMALL_HEAP_V5_SEGMENT_SIZE(default: 2MiB) - power of 2 & >= 64KiB validation 挙動: 完全不変(v5 route は呼ばれない、ENV default OFF) テスト: Mixed 16–1024B で 43.0–43.8M ops/s(変化なし)、SEGV/assert なし 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Haiku 4.5 <noreply@anthropic.com>
2025-12-11 03:19:18 +09:00
#include <stdint.h>
// ENV sentinel values
#define ENV_UNINIT (-1)
#define ENV_ENABLED (1)
#define ENV_DISABLED (0)
// route priority enum
enum small_route_priority {
ROUTE_PRIORITY_V4 = 0,
ROUTE_PRIORITY_V5 = 1,
ROUTE_PRIORITY_AUTO = 2,
};
// small_heap_v5_enabled() - グローバル v5 enable check
static inline int small_heap_v5_enabled(void) {
Phase v5-0 refactor: ENV統一・マクロ化・構造体最適化 - ENV initialization を sentinel パターンで統一 - ENV_UNINIT/ENABLED/DISABLED 定数追加 - __builtin_expect で初期化チェックを最適化 - small_heap_v5_enabled/class_mask を統一パターンに変更 - ポインタマクロ化(O(1) segment/page 計算) - SMALL_SEGMENT_V5_BASE_FROM_PTR: ptr から segment base を mask で計算 - SMALL_SEGMENT_V5_PAGE_IDX: segment 内の page_idx を shift で計算 - SMALL_SEGMENT_V5_PAGE_META: page_meta への O(1) access(bounds check付き) - SMALL_SEGMENT_V5_VALIDATE_MAGIC: magic 検証 - SMALL_SEGMENT_V5_VALIDATE_PTR: Fail-Fast validation pipeline - SmallClassHeapV5 に partial_count 追加 - partial ページリストのカウンタを追加(refill/retire 最適化用) - SmallPageMetaV5 の field 再配置(L1 cache 最適化) - hot fields (free_list, used, capacity) を先頭に集約 - metadata (class_idx, flags, page_idx, segment) を後方配置 - total 24B、offset コメント追加 - route priority ENV 追加 - HAKMEM_ROUTE_PRIORITY={v4|v5|auto}(default: v4) - enum small_route_priority 定義 - small_route_priority() 関数追加 - segment_size override ENV 追加 - HAKMEM_SMALL_HEAP_V5_SEGMENT_SIZE(default: 2MiB) - power of 2 & >= 64KiB validation 挙動: 完全不変(v5 route は呼ばれない、ENV default OFF) テスト: Mixed 16–1024B で 43.0–43.8M ops/s(変化なし)、SEGV/assert なし 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Haiku 4.5 <noreply@anthropic.com>
2025-12-11 03:19:18 +09:00
static int g_enabled = ENV_UNINIT;
if (__builtin_expect(g_enabled == ENV_UNINIT, 0)) {
const char* e = getenv("HAKMEM_SMALL_HEAP_V5_ENABLED");
Phase v5-0 refactor: ENV統一・マクロ化・構造体最適化 - ENV initialization を sentinel パターンで統一 - ENV_UNINIT/ENABLED/DISABLED 定数追加 - __builtin_expect で初期化チェックを最適化 - small_heap_v5_enabled/class_mask を統一パターンに変更 - ポインタマクロ化(O(1) segment/page 計算) - SMALL_SEGMENT_V5_BASE_FROM_PTR: ptr から segment base を mask で計算 - SMALL_SEGMENT_V5_PAGE_IDX: segment 内の page_idx を shift で計算 - SMALL_SEGMENT_V5_PAGE_META: page_meta への O(1) access(bounds check付き) - SMALL_SEGMENT_V5_VALIDATE_MAGIC: magic 検証 - SMALL_SEGMENT_V5_VALIDATE_PTR: Fail-Fast validation pipeline - SmallClassHeapV5 に partial_count 追加 - partial ページリストのカウンタを追加(refill/retire 最適化用) - SmallPageMetaV5 の field 再配置(L1 cache 最適化) - hot fields (free_list, used, capacity) を先頭に集約 - metadata (class_idx, flags, page_idx, segment) を後方配置 - total 24B、offset コメント追加 - route priority ENV 追加 - HAKMEM_ROUTE_PRIORITY={v4|v5|auto}(default: v4) - enum small_route_priority 定義 - small_route_priority() 関数追加 - segment_size override ENV 追加 - HAKMEM_SMALL_HEAP_V5_SEGMENT_SIZE(default: 2MiB) - power of 2 & >= 64KiB validation 挙動: 完全不変(v5 route は呼ばれない、ENV default OFF) テスト: Mixed 16–1024B で 43.0–43.8M ops/s(変化なし)、SEGV/assert なし 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Haiku 4.5 <noreply@anthropic.com>
2025-12-11 03:19:18 +09:00
g_enabled = (e && *e && *e != '0') ? ENV_ENABLED : ENV_DISABLED;
}
Phase v5-0 refactor: ENV統一・マクロ化・構造体最適化 - ENV initialization を sentinel パターンで統一 - ENV_UNINIT/ENABLED/DISABLED 定数追加 - __builtin_expect で初期化チェックを最適化 - small_heap_v5_enabled/class_mask を統一パターンに変更 - ポインタマクロ化(O(1) segment/page 計算) - SMALL_SEGMENT_V5_BASE_FROM_PTR: ptr から segment base を mask で計算 - SMALL_SEGMENT_V5_PAGE_IDX: segment 内の page_idx を shift で計算 - SMALL_SEGMENT_V5_PAGE_META: page_meta への O(1) access(bounds check付き) - SMALL_SEGMENT_V5_VALIDATE_MAGIC: magic 検証 - SMALL_SEGMENT_V5_VALIDATE_PTR: Fail-Fast validation pipeline - SmallClassHeapV5 に partial_count 追加 - partial ページリストのカウンタを追加(refill/retire 最適化用) - SmallPageMetaV5 の field 再配置(L1 cache 最適化) - hot fields (free_list, used, capacity) を先頭に集約 - metadata (class_idx, flags, page_idx, segment) を後方配置 - total 24B、offset コメント追加 - route priority ENV 追加 - HAKMEM_ROUTE_PRIORITY={v4|v5|auto}(default: v4) - enum small_route_priority 定義 - small_route_priority() 関数追加 - segment_size override ENV 追加 - HAKMEM_SMALL_HEAP_V5_SEGMENT_SIZE(default: 2MiB) - power of 2 & >= 64KiB validation 挙動: 完全不変(v5 route は呼ばれない、ENV default OFF) テスト: Mixed 16–1024B で 43.0–43.8M ops/s(変化なし)、SEGV/assert なし 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Haiku 4.5 <noreply@anthropic.com>
2025-12-11 03:19:18 +09:00
return (g_enabled == ENV_ENABLED);
}
// small_heap_v5_class_mask() - v5 対象クラスのビットマスク
static inline uint32_t small_heap_v5_class_mask(void) {
Phase v5-0 refactor: ENV統一・マクロ化・構造体最適化 - ENV initialization を sentinel パターンで統一 - ENV_UNINIT/ENABLED/DISABLED 定数追加 - __builtin_expect で初期化チェックを最適化 - small_heap_v5_enabled/class_mask を統一パターンに変更 - ポインタマクロ化(O(1) segment/page 計算) - SMALL_SEGMENT_V5_BASE_FROM_PTR: ptr から segment base を mask で計算 - SMALL_SEGMENT_V5_PAGE_IDX: segment 内の page_idx を shift で計算 - SMALL_SEGMENT_V5_PAGE_META: page_meta への O(1) access(bounds check付き) - SMALL_SEGMENT_V5_VALIDATE_MAGIC: magic 検証 - SMALL_SEGMENT_V5_VALIDATE_PTR: Fail-Fast validation pipeline - SmallClassHeapV5 に partial_count 追加 - partial ページリストのカウンタを追加(refill/retire 最適化用) - SmallPageMetaV5 の field 再配置(L1 cache 最適化) - hot fields (free_list, used, capacity) を先頭に集約 - metadata (class_idx, flags, page_idx, segment) を後方配置 - total 24B、offset コメント追加 - route priority ENV 追加 - HAKMEM_ROUTE_PRIORITY={v4|v5|auto}(default: v4) - enum small_route_priority 定義 - small_route_priority() 関数追加 - segment_size override ENV 追加 - HAKMEM_SMALL_HEAP_V5_SEGMENT_SIZE(default: 2MiB) - power of 2 & >= 64KiB validation 挙動: 完全不変(v5 route は呼ばれない、ENV default OFF) テスト: Mixed 16–1024B で 43.0–43.8M ops/s(変化なし)、SEGV/assert なし 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Haiku 4.5 <noreply@anthropic.com>
2025-12-11 03:19:18 +09:00
static int g_mask = ENV_UNINIT; // sentinel
if (__builtin_expect(g_mask == ENV_UNINIT, 0)) {
const char* e = getenv("HAKMEM_SMALL_HEAP_V5_CLASSES");
if (e && *e) {
Phase v5-0 refactor: ENV統一・マクロ化・構造体最適化 - ENV initialization を sentinel パターンで統一 - ENV_UNINIT/ENABLED/DISABLED 定数追加 - __builtin_expect で初期化チェックを最適化 - small_heap_v5_enabled/class_mask を統一パターンに変更 - ポインタマクロ化(O(1) segment/page 計算) - SMALL_SEGMENT_V5_BASE_FROM_PTR: ptr から segment base を mask で計算 - SMALL_SEGMENT_V5_PAGE_IDX: segment 内の page_idx を shift で計算 - SMALL_SEGMENT_V5_PAGE_META: page_meta への O(1) access(bounds check付き) - SMALL_SEGMENT_V5_VALIDATE_MAGIC: magic 検証 - SMALL_SEGMENT_V5_VALIDATE_PTR: Fail-Fast validation pipeline - SmallClassHeapV5 に partial_count 追加 - partial ページリストのカウンタを追加(refill/retire 最適化用) - SmallPageMetaV5 の field 再配置(L1 cache 最適化) - hot fields (free_list, used, capacity) を先頭に集約 - metadata (class_idx, flags, page_idx, segment) を後方配置 - total 24B、offset コメント追加 - route priority ENV 追加 - HAKMEM_ROUTE_PRIORITY={v4|v5|auto}(default: v4) - enum small_route_priority 定義 - small_route_priority() 関数追加 - segment_size override ENV 追加 - HAKMEM_SMALL_HEAP_V5_SEGMENT_SIZE(default: 2MiB) - power of 2 & >= 64KiB validation 挙動: 完全不変(v5 route は呼ばれない、ENV default OFF) テスト: Mixed 16–1024B で 43.0–43.8M ops/s(変化なし)、SEGV/assert なし 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Haiku 4.5 <noreply@anthropic.com>
2025-12-11 03:19:18 +09:00
g_mask = (int)strtoul(e, NULL, 0);
} else {
g_mask = 0x0; // default: OFF
}
}
Phase v5-0 refactor: ENV統一・マクロ化・構造体最適化 - ENV initialization を sentinel パターンで統一 - ENV_UNINIT/ENABLED/DISABLED 定数追加 - __builtin_expect で初期化チェックを最適化 - small_heap_v5_enabled/class_mask を統一パターンに変更 - ポインタマクロ化(O(1) segment/page 計算) - SMALL_SEGMENT_V5_BASE_FROM_PTR: ptr から segment base を mask で計算 - SMALL_SEGMENT_V5_PAGE_IDX: segment 内の page_idx を shift で計算 - SMALL_SEGMENT_V5_PAGE_META: page_meta への O(1) access(bounds check付き) - SMALL_SEGMENT_V5_VALIDATE_MAGIC: magic 検証 - SMALL_SEGMENT_V5_VALIDATE_PTR: Fail-Fast validation pipeline - SmallClassHeapV5 に partial_count 追加 - partial ページリストのカウンタを追加(refill/retire 最適化用) - SmallPageMetaV5 の field 再配置(L1 cache 最適化) - hot fields (free_list, used, capacity) を先頭に集約 - metadata (class_idx, flags, page_idx, segment) を後方配置 - total 24B、offset コメント追加 - route priority ENV 追加 - HAKMEM_ROUTE_PRIORITY={v4|v5|auto}(default: v4) - enum small_route_priority 定義 - small_route_priority() 関数追加 - segment_size override ENV 追加 - HAKMEM_SMALL_HEAP_V5_SEGMENT_SIZE(default: 2MiB) - power of 2 & >= 64KiB validation 挙動: 完全不変(v5 route は呼ばれない、ENV default OFF) テスト: Mixed 16–1024B で 43.0–43.8M ops/s(変化なし)、SEGV/assert なし 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Haiku 4.5 <noreply@anthropic.com>
2025-12-11 03:19:18 +09:00
return (uint32_t)g_mask;
}
// small_heap_v5_class_enabled() - 指定クラスが v5 有効か
static inline int small_heap_v5_class_enabled(uint32_t class_idx) {
if (class_idx >= 8) return 0;
if (!small_heap_v5_enabled()) return 0;
uint32_t mask = small_heap_v5_class_mask();
return (mask & (1u << class_idx)) ? 1 : 0;
}
// 便利関数
static inline int small_heap_v5_c6_enabled(void) {
return small_heap_v5_class_enabled(6);
}
static inline int small_heap_v5_c5_enabled(void) {
return small_heap_v5_class_enabled(5);
}
static inline int small_heap_v5_c7_enabled(void) {
return small_heap_v5_class_enabled(7);
}
Phase v5-0 refactor: ENV統一・マクロ化・構造体最適化 - ENV initialization を sentinel パターンで統一 - ENV_UNINIT/ENABLED/DISABLED 定数追加 - __builtin_expect で初期化チェックを最適化 - small_heap_v5_enabled/class_mask を統一パターンに変更 - ポインタマクロ化(O(1) segment/page 計算) - SMALL_SEGMENT_V5_BASE_FROM_PTR: ptr から segment base を mask で計算 - SMALL_SEGMENT_V5_PAGE_IDX: segment 内の page_idx を shift で計算 - SMALL_SEGMENT_V5_PAGE_META: page_meta への O(1) access(bounds check付き) - SMALL_SEGMENT_V5_VALIDATE_MAGIC: magic 検証 - SMALL_SEGMENT_V5_VALIDATE_PTR: Fail-Fast validation pipeline - SmallClassHeapV5 に partial_count 追加 - partial ページリストのカウンタを追加(refill/retire 最適化用) - SmallPageMetaV5 の field 再配置(L1 cache 最適化) - hot fields (free_list, used, capacity) を先頭に集約 - metadata (class_idx, flags, page_idx, segment) を後方配置 - total 24B、offset コメント追加 - route priority ENV 追加 - HAKMEM_ROUTE_PRIORITY={v4|v5|auto}(default: v4) - enum small_route_priority 定義 - small_route_priority() 関数追加 - segment_size override ENV 追加 - HAKMEM_SMALL_HEAP_V5_SEGMENT_SIZE(default: 2MiB) - power of 2 & >= 64KiB validation 挙動: 完全不変(v5 route は呼ばれない、ENV default OFF) テスト: Mixed 16–1024B で 43.0–43.8M ops/s(変化なし)、SEGV/assert なし 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Haiku 4.5 <noreply@anthropic.com>
2025-12-11 03:19:18 +09:00
// small_route_priority() - route priority (v4/v5/auto)
// ENV: HAKMEM_ROUTE_PRIORITY={v4|v5|auto}, default: v4
static inline int small_route_priority(void) {
static int g_priority = ENV_UNINIT;
if (__builtin_expect(g_priority == ENV_UNINIT, 0)) {
const char* e = getenv("HAKMEM_ROUTE_PRIORITY");
if (e && *e) {
if (strcmp(e, "v5") == 0) {
g_priority = ROUTE_PRIORITY_V5;
} else if (strcmp(e, "auto") == 0) {
g_priority = ROUTE_PRIORITY_AUTO;
} else {
g_priority = ROUTE_PRIORITY_V4; // default or "v4"
}
} else {
g_priority = ROUTE_PRIORITY_V4; // default
}
}
return g_priority;
}
// small_heap_v5_segment_size() - segment size override
// ENV: HAKMEM_SMALL_HEAP_V5_SEGMENT_SIZE, default: 2MiB (2*1024*1024)
static inline size_t small_heap_v5_segment_size(void) {
static int g_size = ENV_UNINIT;
if (__builtin_expect(g_size == ENV_UNINIT, 0)) {
const char* e = getenv("HAKMEM_SMALL_HEAP_V5_SEGMENT_SIZE");
if (e && *e) {
size_t sz = (size_t)strtoul(e, NULL, 0);
// validate: must be power of 2, >= 64KiB
if (sz >= (64 * 1024) && (sz & (sz - 1)) == 0) {
g_size = (int)sz;
} else {
g_size = (2 * 1024 * 1024); // fallback to default
}
} else {
g_size = (2 * 1024 * 1024); // default: 2MiB
}
}
return (size_t)g_size;
}
Phase v5-4: Header light mode & freelist optimization Implements header write optimization for C6 v5 allocator by moving header initialization from per-alloc time to carve time (during page refill). This eliminates redundant header writes on the hot path. Implementation: - Added HAKMEM_SMALL_HEAP_V5_HEADER_MODE ENV (full|light, default: full) - Added header_mode field to SmallHeapCtxV5 (cached per-thread) - Modified alloc fast/slow paths to skip header write in light mode - Modified refill to write headers during carve in light mode - Free path unchanged (header validation still works) Benchmark Results (2M iterations, ws=400): C6-HEAVY (257-768B): - Baseline (v5 OFF): 47.95 Mops/s - v5 full mode: 38.97 Mops/s (-18.7% vs baseline) - v5 light mode: 39.25 Mops/s (-18.1% vs baseline, +0.7% vs full) MIXED 16-1024B: - v5 OFF: 43.59 Mops/s - v5 full mode: 36.53 Mops/s (-16.2% vs OFF) - v5 light mode: 38.04 Mops/s (-12.7% vs OFF, +4.1% vs full) Analysis: - Light mode shows modest improvement over full (+0.7-4.1%) - C6 v5 performance gap vs baseline (-18%) indicates need for further optimization beyond header writes - Mixed workload benefits more from light mode (+4.1% vs full) - No regressions in safety/correctness observed Research findings: - Header write optimization alone insufficient to close v5 gap - Need to investigate other hot path costs (freelist ops, metadata access) - Light mode validates the carve-time header concept 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Haiku 4.5 <noreply@anthropic.com>
2025-12-11 05:12:39 +09:00
// ============================================================================
// Phase v5-4: Header mode configuration
// ============================================================================
// Header mode enum
#define SMALL_HEAP_V5_HEADER_MODE_FULL 0
#define SMALL_HEAP_V5_HEADER_MODE_LIGHT 1
// small_heap_v5_header_mode() - header write mode (default: full)
// ENV: HAKMEM_SMALL_HEAP_V5_HEADER_MODE={full|light}, default: full
// - full: write header on every alloc (safe, standard)
// - light: write header once during carve, skip on alloc (research mode, +2-4% perf)
static inline int small_heap_v5_header_mode(void) {
static int g_header_mode = ENV_UNINIT;
if (__builtin_expect(g_header_mode == ENV_UNINIT, 0)) {
const char* e = getenv("HAKMEM_SMALL_HEAP_V5_HEADER_MODE");
if (e && *e) {
if (strcmp(e, "light") == 0 || strcmp(e, "LIGHT") == 0 || *e == '1') {
g_header_mode = SMALL_HEAP_V5_HEADER_MODE_LIGHT;
} else {
g_header_mode = SMALL_HEAP_V5_HEADER_MODE_FULL; // default or "full"
}
} else {
g_header_mode = SMALL_HEAP_V5_HEADER_MODE_FULL; // default: full
}
}
return g_header_mode;
}
#endif // HAKMEM_SMALLOBJECT_V5_ENV_BOX_H