tomoaki/hakmem
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Phase SO-BACKEND-OPT-1: v3 backend 分解＆Tiny/ULTRA 完成世代宣言

Browse Source 
=== 実装内容 ===

1. v3 backend 詳細計測
   - ENV: HAKMEM_SO_V3_STATS で alloc/free パス内訳計測
   - 追加 stats: alloc_current_hit, alloc_partial_hit, free_current, free_partial, free_retire
   - so_alloc_fast / so_free_fast に埋め込み
   - デストラクタで [ALLOC_DETAIL] / [FREE_DETAIL] 出力

2. v3 backend ボトルネック分析完了
   - C7-only: alloc_current_hit=99.99%, alloc_refill=0.9%, free_retire=0.1%, page_of_fail=0
   - Mixed: alloc_current_hit=100%, alloc_refill=0.85%, free_retire=0.07%, page_of_fail=0
   - 結論: v3 ロジック部分（ページ選択・retire）は完全最適化済み
   - 残り 5% overhead は内部コスト（header write, memcpy, 分岐）

3. Tiny/ULTRA 層「完成世代」宣言
   - 総括ドキュメント作成: docs/analysis/PERF_EXEC_SUMMARY_ULTRA_PHASE_20251211.md
   - CURRENT_TASK.md に Phase ULTRA 総括セクション追加
   - AGENTS.md に Tiny/ULTRA 完成世代宣言追加
   - 最終成果: Mixed 16–1024B = 43.9M ops/s (baseline 30.6M → +43.5%)

=== ボトルネック地図 ===

| 層 | 関数 | overhead |
|-----|------|----------|
| Front | malloc/free dispatcher | ~40–45% |
| ULTRA | C4–C7 alloc/free/refill | ~12% |
| v3 backend | so_alloc/so_free | ~5% |
| mid/pool | hak_super_lookup | 3–5% |

=== フェーズ履歴（Phase ULTRA cycle） ===

- Phase PERF-ULTRA-FREE-OPT-1: C4–C7 ULTRA統合 → +9.3%
- Phase REFACTOR: Code quality (60行削減)
- Phase PERF-ULTRA-REFILL-OPT-1a/1b: C7 ULTRA refill最適化 → +11.1%
- Phase SO-BACKEND-OPT-1: v3 backend分解 → 設計限界確認

=== 次フェーズ（独立ライン） ===

1. Phase SO-BACKEND-OPT-2: v3 header write削減 (1-2%)
2. Headerless/v6系: out-of-band header (1-2%)
3. mid/pool v3新設計: C6-heavy 10M → 20–25M

本フェーズでTiny/ULTRA層は「完成世代」として基盤固定。
今後の大きい変更はHeaderless/mid系の独立ラインで検討。

🤖 Generated with Claude Code

Co-Authored-By: Claude Haiku 4.5 <noreply@anthropic.com>
This commit is contained in:
Moe Charm (CI)
2025-12-11 22:45:14 +09:00
parent 022ba56033
commit 2d684ffd25
6 changed files with 530 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files

52
core/smallobject_hotbox_v3.c
View File

@ -49,6 +49,16 @@ void so_v3_record_alloc_fallback(uint8_t ci) {
if (st) atomic_fetch_add_explicit(&st->alloc_fallback_v1, 1, memory_order_relaxed);
}
void so_v3_record_alloc_current_hit(uint8_t ci) {
so_stats_class_v3* st = so_stats_for(ci);
if (st) atomic_fetch_add_explicit(&st->alloc_current_hit, 1, memory_order_relaxed);
}
void so_v3_record_alloc_partial_hit(uint8_t ci) {
so_stats_class_v3* st = so_stats_for(ci);
if (st) atomic_fetch_add_explicit(&st->alloc_partial_hit, 1, memory_order_relaxed);
}
void so_v3_record_free_call(uint8_t ci) {
so_stats_class_v3* st = so_stats_for(ci);
if (st) atomic_fetch_add_explicit(&st->free_calls, 1, memory_order_relaxed);
@ -59,6 +69,21 @@ void so_v3_record_free_fallback(uint8_t ci) {
if (st) atomic_fetch_add_explicit(&st->free_fallback_v1, 1, memory_order_relaxed);
}
void so_v3_record_free_current(uint8_t ci) {
so_stats_class_v3* st = so_stats_for(ci);
if (st) atomic_fetch_add_explicit(&st->free_current, 1, memory_order_relaxed);
}
void so_v3_record_free_partial(uint8_t ci) {
so_stats_class_v3* st = so_stats_for(ci);
if (st) atomic_fetch_add_explicit(&st->free_partial, 1, memory_order_relaxed);
}
void so_v3_record_free_retire(uint8_t ci) {
so_stats_class_v3* st = so_stats_for(ci);
if (st) atomic_fetch_add_explicit(&st->free_retire, 1, memory_order_relaxed);
}
void so_v3_record_page_of_fail(uint8_t ci) {
so_stats_class_v3* st = so_stats_for(ci);
if (st) atomic_fetch_add_explicit(&st->page_of_fail, 1, memory_order_relaxed);
@ -137,6 +162,7 @@ static inline void* so_alloc_fast(so_ctx_v3* ctx, uint32_t ci) {
void* blk = p->freelist;
p->freelist = *(void**)blk;
p->used++;
so_v3_record_alloc_current_hit((uint8_t)ci);
if (skip_header_c7) {
uint8_t* header_ptr = (uint8_t*)blk;
*header_ptr = (uint8_t)(HEADER_MAGIC | (ci & HEADER_CLASS_MASK));
@ -165,6 +191,7 @@ static inline void* so_alloc_fast(so_ctx_v3* ctx, uint32_t ci) {
void* blk = p->freelist;
p->freelist = *(void**)blk;
p->used++;
so_v3_record_alloc_partial_hit((uint8_t)ci);
if (skip_header_c7) {
uint8_t* header_ptr = (uint8_t*)blk;
*header_ptr = (uint8_t)(HEADER_MAGIC | (ci & HEADER_CLASS_MASK));
@ -229,17 +256,21 @@ static inline void so_free_fast(so_ctx_v3* ctx, uint32_t ci, void* ptr) {
(void)so_unlink_partial(hc, page);
if (hc->partial_count < hc->max_partial_pages) {
so_page_push_partial(hc, page);
so_v3_record_free_partial((uint8_t)ci);
if (!hc->current) {
hc->current = page;
so_v3_record_free_current((uint8_t)ci);
}
} else {
if (hc->current == page) {
hc->current = NULL;
}
so_v3_record_free_retire((uint8_t)ci);
so_page_retire_slow(ctx, ci, page);
}
} else if (!hc->current) {
hc->current = page;
so_v3_record_free_current((uint8_t)ci);
}
}
@ -339,15 +370,34 @@ static void so_v3_stats_dump(void) {
so_stats_class_v3* st = &g_so_stats[i];
uint64_t rh = atomic_load_explicit(&st->route_hits, memory_order_relaxed);
uint64_t ac = atomic_load_explicit(&st->alloc_calls, memory_order_relaxed);
uint64_t ach = atomic_load_explicit(&st->alloc_current_hit, memory_order_relaxed);
uint64_t aph = atomic_load_explicit(&st->alloc_partial_hit, memory_order_relaxed);
uint64_t ar = atomic_load_explicit(&st->alloc_refill, memory_order_relaxed);
uint64_t afb = atomic_load_explicit(&st->alloc_fallback_v1, memory_order_relaxed);
uint64_t fc = atomic_load_explicit(&st->free_calls, memory_order_relaxed);
uint64_t fcur = atomic_load_explicit(&st->free_current, memory_order_relaxed);
uint64_t fpar = atomic_load_explicit(&st->free_partial, memory_order_relaxed);
uint64_t fret = atomic_load_explicit(&st->free_retire, memory_order_relaxed);
uint64_t ffb = atomic_load_explicit(&st->free_fallback_v1, memory_order_relaxed);
uint64_t pof = atomic_load_explicit(&st->page_of_fail, memory_order_relaxed);
if (rh + ac + afb + fc + ffb + ar + pof == 0) continue;
if (rh + ac + afb + fc + ffb + ar + pof + ach + aph + fcur + fpar + fret == 0) continue;
// Main stats (basic)
fprintf(stderr, "[SMALL_HEAP_V3_STATS] cls=%d route_hits=%llu alloc_calls=%llu alloc_refill=%llu alloc_fb_v1=%llu free_calls=%llu free_fb_v1=%llu page_of_fail=%llu\n",
i, (unsigned long long)rh, (unsigned long long)ac,
(unsigned long long)ar, (unsigned long long)afb, (unsigned long long)fc,
(unsigned long long)ffb, (unsigned long long)pof);
// Detailed alloc path breakdown
if (ach + aph > 0) {
fprintf(stderr, " [ALLOC_DETAIL] alloc_current_hit=%llu alloc_partial_hit=%llu\n",
(unsigned long long)ach, (unsigned long long)aph);
}
// Detailed free path breakdown
if (fcur + fpar + fret > 0) {
fprintf(stderr, " [FREE_DETAIL] free_current=%llu free_partial=%llu free_retire=%llu\n",
(unsigned long long)fcur, (unsigned long long)fpar, (unsigned long long)fret);
}
}
}