Files
hakmem/core/front/tiny_ring_cache.c
Moe Charm (CI) db9c06211e Phase 21-1-A: Ring cache 基本実装 - Array-based TLS cache (C2/C3)
## Summary
Phase 21-1-A の基本実装完了。Ring buffer ベースの TLS cache を C2/C3
(33-128B)専用に実装。ポインタチェイス削減で +15-20% 性能向上を目指す。

## Implementation

**Files Created**:
- `core/front/tiny_ring_cache.h` - Ring cache API, ENV control
- `core/front/tiny_ring_cache.c` - Ring cache implementation

**Makefile Integration**:
- Added `core/front/tiny_ring_cache.o` to OBJS_BASE
- Added `core/front/tiny_ring_cache_shared.o` to SHARED_OBJS
- Added `core/front/tiny_ring_cache.o` to BENCH_HAKMEM_OBJS_BASE

## Design (Task 先生調査結果 + ChatGPT フィードバック)

**Ring Buffer Structure**:
- C2/C3 専用(hot classes, 33-128B)
- Default 128 slots (power-of-2, ENV で 64/128/256 A/B 可能)
- Ultra-fast pop/push (1-2 instructions, array access)
- Fast modulo via mask (capacity - 1)

**Hierarchy** (Option 4: UltraHot 置き換え):
```
Ring (L0, C2/C3 専用) → HeapV2 (L1, fallback) → TLS SLL (L2) → SuperSlab (L3)
```

**Rationale**:
- UltraHot の C3 問題(5.8% hit rate)を根本解決
- Phase 19-3 の +12.9%(UltraHot 除去)を維持
- Ring サイズ(128)>> UltraHot(4)→ hit rate 大幅向上期待

**Performance Goal**:
- Pointer chasing: TLS SLL 1 回 → Ring 0 回
- Memory access: 3 → 2 回
- Cache locality: 配列(連続メモリ)vs linked list
- Expected: +15-20% (54.4M → 62-65M ops/s)

## ENV Variables

```bash
HAKMEM_TINY_HOT_RING_ENABLE=1  # Ring 有効化 (default: 0)
HAKMEM_TINY_HOT_RING_C2=128    # C2 サイズ (default: 128)
HAKMEM_TINY_HOT_RING_C3=128    # C3 サイズ (default: 128)
HAKMEM_TINY_HOT_RING_CASCADE=1 # SLL → Ring refill (default: 0)
```

## Implementation Status

Phase 21-1-A:  **COMPLETE**
- Ring buffer data structure
- TLS variables
- ENV control (enable/capacity)
- Power-of-2 capacity (fast modulo)
- Ultra-fast pop/push inline functions
- Refill from SLL (scaffold)
- Init/shutdown/stats (scaffold)
- Makefile integration
- Compile success

Phase 21-1-B:  **NEXT** - Alloc/Free 統合
Phase 21-1-C:  **PENDING** - Refill/Cascade 実装
Phase 21-1-D:  **PENDING** - A/B テスト

## Next Steps

1. Alloc path 統合 (`core/tiny_alloc_fast.inc.h`)
2. Free path 統合 (`core/tiny_free_fast_v2.inc.h`)
3. Init call from `hakmem_tiny.c`
4. A/B test: Ring vs UltraHot vs Baseline

🎯 Target: 62-65M ops/s (+15-20% vs 54.4M baseline)

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

Co-Authored-By: Claude <noreply@anthropic.com>
2025-11-16 07:32:24 +09:00

162 lines
5.3 KiB
C

// tiny_ring_cache.c - Phase 21-1: Ring cache implementation
#include "tiny_ring_cache.h"
#include <stdlib.h>
#include <string.h>
// ============================================================================
// TLS Variables (defined here, extern in header)
// ============================================================================
__thread TinyRingCache g_ring_cache_c2 = {NULL, 0, 0, 0, 0};
__thread TinyRingCache g_ring_cache_c3 = {NULL, 0, 0, 0, 0};
// ============================================================================
// Init (called at thread start, from hakmem_tiny.c)
// ============================================================================
void ring_cache_init(void) {
if (!ring_cache_enabled()) return;
// C2 init
size_t cap_c2 = ring_capacity_c2();
g_ring_cache_c2.slots = (void**)calloc(cap_c2, sizeof(void*));
if (!g_ring_cache_c2.slots) {
#if !HAKMEM_BUILD_RELEASE
fprintf(stderr, "[Ring-INIT] Failed to allocate C2 ring (%zu slots)\n", cap_c2);
fflush(stderr);
#endif
return;
}
g_ring_cache_c2.capacity = (uint16_t)cap_c2;
g_ring_cache_c2.mask = (uint16_t)(cap_c2 - 1);
g_ring_cache_c2.head = 0;
g_ring_cache_c2.tail = 0;
// C3 init
size_t cap_c3 = ring_capacity_c3();
g_ring_cache_c3.slots = (void**)calloc(cap_c3, sizeof(void*));
if (!g_ring_cache_c3.slots) {
#if !HAKMEM_BUILD_RELEASE
fprintf(stderr, "[Ring-INIT] Failed to allocate C3 ring (%zu slots)\n", cap_c3);
fflush(stderr);
#endif
// Free C2 if C3 failed
free(g_ring_cache_c2.slots);
g_ring_cache_c2.slots = NULL;
return;
}
g_ring_cache_c3.capacity = (uint16_t)cap_c3;
g_ring_cache_c3.mask = (uint16_t)(cap_c3 - 1);
g_ring_cache_c3.head = 0;
g_ring_cache_c3.tail = 0;
#if !HAKMEM_BUILD_RELEASE
fprintf(stderr, "[Ring-INIT] C2=%zu slots (%zu bytes), C3=%zu slots (%zu bytes)\n",
cap_c2, cap_c2 * sizeof(void*),
cap_c3, cap_c3 * sizeof(void*));
fflush(stderr);
#endif
}
// ============================================================================
// Shutdown (called at thread exit, optional)
// ============================================================================
void ring_cache_shutdown(void) {
if (!ring_cache_enabled()) return;
// Drain rings to TLS SLL before shutdown (prevent leak)
// TODO: Implement drain logic in Phase 21-1-C
// Free ring buffers
if (g_ring_cache_c2.slots) {
free(g_ring_cache_c2.slots);
g_ring_cache_c2.slots = NULL;
}
if (g_ring_cache_c3.slots) {
free(g_ring_cache_c3.slots);
g_ring_cache_c3.slots = NULL;
}
#if !HAKMEM_BUILD_RELEASE
fprintf(stderr, "[Ring-SHUTDOWN] C2/C3 rings freed\n");
fflush(stderr);
#endif
}
// ============================================================================
// Refill from TLS SLL (cascade, Phase 21-1-C)
// ============================================================================
// Refill ring from TLS SLL (one-way cascade: SLL → Ring)
// Returns: number of blocks transferred
int ring_refill_from_sll(int class_idx, int target_count) {
if (!ring_cascade_enabled()) return 0;
if (class_idx != 2 && class_idx != 3) return 0;
// Forward declarations (external functions from tls_sll_box.h)
extern int tls_sll_pop(int class_idx, void** out_ptr);
extern int tls_sll_push(int class_idx, void* ptr, uint32_t capacity);
int transferred = 0;
while (transferred < target_count) {
void* ptr = NULL;
// Pop from TLS SLL
if (!tls_sll_pop(class_idx, &ptr)) {
break; // SLL empty
}
// Push to Ring
if (!ring_cache_push(class_idx, ptr)) {
// Ring full, push back to SLL
tls_sll_push(class_idx, ptr, (uint32_t)-1); // Unlimited capacity
break;
}
transferred++;
}
#if !HAKMEM_BUILD_RELEASE
if (transferred > 0) {
fprintf(stderr, "[Ring-REFILL] C%d: %d blocks transferred from SLL to Ring\n",
class_idx, transferred);
fflush(stderr);
}
#endif
return transferred;
}
// ============================================================================
// Stats (Phase 19-1 metrics, TODO)
// ============================================================================
void ring_cache_print_stats(void) {
if (!ring_cache_enabled()) return;
// TODO: Add metrics tracking (Phase 21-1-E)
// - Hit rate (Ring hits / total allocs)
// - Miss rate (Ring empty / total allocs)
// - Full rate (Ring full / total frees)
// - Refill count (SLL → Ring transfers)
#if !HAKMEM_BUILD_RELEASE
// Current occupancy
uint16_t c2_count = (g_ring_cache_c2.tail >= g_ring_cache_c2.head)
? (g_ring_cache_c2.tail - g_ring_cache_c2.head)
: (g_ring_cache_c2.capacity - g_ring_cache_c2.head + g_ring_cache_c2.tail);
uint16_t c3_count = (g_ring_cache_c3.tail >= g_ring_cache_c3.head)
? (g_ring_cache_c3.tail - g_ring_cache_c3.head)
: (g_ring_cache_c3.capacity - g_ring_cache_c3.head + g_ring_cache_c3.tail);
fprintf(stderr, "[Ring-STATS] C2: %u/%u slots, C3: %u/%u slots\n",
c2_count, g_ring_cache_c2.capacity,
c3_count, g_ring_cache_c3.capacity);
fflush(stderr);
#endif
}