Files
hakmem/core/box/front_gate_classifier.c

237 lines
7.4 KiB
C
Raw Normal View History

// front_gate_classifier.c - Box FG: Pointer Classification Implementation
// CRITICAL: Box FG requires header-based classification
// Ensure HEADER_MAGIC and HEADER_CLASS_MASK are available
#ifndef HAKMEM_TINY_HEADER_CLASSIDX
#define HAKMEM_TINY_HEADER_CLASSIDX 1
#endif
#include <stdio.h> // For fprintf in debug
#include <stdlib.h> // For abort in debug
#include "front_gate_classifier.h"
#include "../tiny_region_id.h" // Must come before hakmem_tiny_superslab.h for HEADER_MAGIC
#include "../hakmem_tiny_superslab.h"
#include "../superslab/superslab_inline.h" // For ss_slabs_capacity
#include "../hakmem_build_flags.h"
#include "../hakmem_internal.h" // AllocHeader, HAKMEM_MAGIC, HEADER_SIZE, hak_is_memory_readable
#include "../hakmem_tiny_config.h" // For TINY_NUM_CLASSES, SLAB_SIZE
#include "../hakmem_super_registry.h" // For hak_super_lookup (Box REG)
#ifdef HAKMEM_POOL_TLS_PHASE1
#include "../pool_tls_registry.h" // Safer pool pointer lookup (no header deref)
#endif
// ========== Debug Stats ==========
#if !HAKMEM_BUILD_RELEASE
__thread uint64_t g_classify_header_hit = 0;
__thread uint64_t g_classify_headerless_hit = 0;
__thread uint64_t g_classify_pool_hit = 0;
__thread uint64_t g_classify_unknown_hit = 0;
void front_gate_print_stats(void) {
uint64_t total = g_classify_header_hit + g_classify_headerless_hit +
g_classify_pool_hit + g_classify_unknown_hit;
if (total == 0) return;
fprintf(stderr, "\n========== Front Gate Classification Stats ==========\n");
fprintf(stderr, "Header (C0-C6): %lu (%.2f%%)\n",
g_classify_header_hit, 100.0 * g_classify_header_hit / total);
fprintf(stderr, "Headerless (C7): %lu (%.2f%%)\n",
g_classify_headerless_hit, 100.0 * g_classify_headerless_hit / total);
fprintf(stderr, "Pool TLS: %lu (%.2f%%)\n",
g_classify_pool_hit, 100.0 * g_classify_pool_hit / total);
fprintf(stderr, "Unknown: %lu (%.2f%%)\n",
g_classify_unknown_hit, 100.0 * g_classify_unknown_hit / total);
fprintf(stderr, "Total: %lu\n", total);
fprintf(stderr, "======================================================\n");
}
static void __attribute__((destructor)) front_gate_stats_destructor(void) {
front_gate_print_stats();
}
#endif
// ========== Safe Header Probe ==========
// Try to read 1-byte header at ptr-1 (safe conditions only)
// Returns: class_idx (0-7) on success, -1 on failure
//
// Safety conditions:
// 1. Same page: (ptr & 0xFFF) >= 1 → header won't cross page boundary
// 2. Valid magic: (header & 0xF0) == HEADER_MAGIC (0xa0)
// 3. Valid class: class_idx in range [0, 7]
//
// Performance: 2-3 cycles (L1 cache hit)
static inline int safe_header_probe(void* ptr) {
// Reject obviously invalid/sentinel-sized pointers (defense-in-depth)
if ((uintptr_t)ptr < 4096) {
return -1;
}
// Safety check: header must be in same page as ptr
uintptr_t offset_in_page = (uintptr_t)ptr & 0xFFF;
if (offset_in_page == 0) {
// ptr is page-aligned → header would be on previous page (unsafe)
return -1;
}
// Safe to read header (same page guaranteed)
uint8_t* header_ptr = (uint8_t*)ptr - 1;
uint8_t header = *header_ptr;
// Validate magic
if ((header & 0xF0) != HEADER_MAGIC) {
return -1; // Not a Tiny header
}
// Extract class index
int class_idx = header & HEADER_CLASS_MASK;
Phase E3-FINAL: Fix Box API offset bugs - ALL classes now use correct offsets ## Root Cause Analysis (GPT5) **Physical Layout Constraints**: - Class 0: 8B = [1B header][7B payload] → offset 1 = 9B needed = ❌ IMPOSSIBLE - Class 1-6: >=16B = [1B header][15B+ payload] → offset 1 = ✅ POSSIBLE - Class 7: 1KB → offset 0 (compatibility) **Correct Specification**: - HAKMEM_TINY_HEADER_CLASSIDX != 0: - Class 0, 7: next at offset 0 (overwrites header when on freelist) - Class 1-6: next at offset 1 (after header) - HAKMEM_TINY_HEADER_CLASSIDX == 0: - All classes: next at offset 0 **Previous Bug**: - Attempted "ALL classes offset 1" unification - Class 0 with offset 1 caused immediate SEGV (9B > 8B block size) - Mixed 2-arg/3-arg API caused confusion ## Fixes Applied ### 1. Restored 3-Argument Box API (core/box/tiny_next_ptr_box.h) ```c // Correct signatures void tiny_next_write(int class_idx, void* base, void* next_value) void* tiny_next_read(int class_idx, const void* base) // Correct offset calculation size_t offset = (class_idx == 0 || class_idx == 7) ? 0 : 1; ``` ### 2. Updated 123+ Call Sites Across 34 Files - hakmem_tiny_hot_pop_v4.inc.h (4 locations) - hakmem_tiny_fastcache.inc.h (3 locations) - hakmem_tiny_tls_list.h (12 locations) - superslab_inline.h (5 locations) - tiny_fastcache.h (3 locations) - ptr_trace.h (macro definitions) - tls_sll_box.h (2 locations) - + 27 additional files Pattern: `tiny_next_read(base)` → `tiny_next_read(class_idx, base)` Pattern: `tiny_next_write(base, next)` → `tiny_next_write(class_idx, base, next)` ### 3. Added Sentinel Detection Guards - tiny_fast_push(): Block nodes with sentinel in ptr or ptr->next - tls_list_push(): Block nodes with sentinel in ptr or ptr->next - Defense-in-depth against remote free sentinel leakage ## Verification (GPT5 Report) **Test Command**: `./out/release/bench_random_mixed_hakmem --iterations=70000` **Results**: - ✅ Main loop completed successfully - ✅ Drain phase completed successfully - ✅ NO SEGV (previous crash at iteration 66151 is FIXED) - ℹ️ Final log: "tiny_alloc(1024) failed" is normal fallback to Mid/ACE layers **Analysis**: - Class 0 immediate SEGV: ✅ RESOLVED (correct offset 0 now used) - 66K iteration crash: ✅ RESOLVED (offset consistency fixed) - Box API conflicts: ✅ RESOLVED (unified 3-arg API) ## Technical Details ### Offset Logic Justification ``` Class 0: 8B block → next pointer (8B) fits ONLY at offset 0 Class 1: 16B block → next pointer (8B) fits at offset 1 (after 1B header) Class 2: 32B block → next pointer (8B) fits at offset 1 ... Class 6: 512B block → next pointer (8B) fits at offset 1 Class 7: 1024B block → offset 0 for legacy compatibility ``` ### Files Modified (Summary) - Core API: `box/tiny_next_ptr_box.h` - Hot paths: `hakmem_tiny_hot_pop*.inc.h`, `tiny_fastcache.h` - TLS layers: `hakmem_tiny_tls_list.h`, `hakmem_tiny_tls_ops.h` - SuperSlab: `superslab_inline.h`, `tiny_superslab_*.inc.h` - Refill: `hakmem_tiny_refill.inc.h`, `tiny_refill_opt.h` - Free paths: `tiny_free_magazine.inc.h`, `tiny_superslab_free.inc.h` - Documentation: Multiple Phase E3 reports ## Remaining Work None for Box API offset bugs - all structural issues resolved. Future enhancements (non-critical): - Periodic `grep -R '*(void**)' core/` to detect direct pointer access violations - Enforce Box API usage via static analysis - Document offset rationale in architecture docs 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude <noreply@anthropic.com>
2025-11-13 06:50:20 +09:00
// Phase E1-CORRECT: Validate class range (all classes 0-7 valid)
if (class_idx < 0 || class_idx >= TINY_NUM_CLASSES) {
return -1; // Invalid class
}
return class_idx;
}
// ========== Registry Lookup ==========
// Lookup pointer in SuperSlab registry (fallback when header probe fails)
// Returns: classification result with SuperSlab + class_idx + slab_idx
//
// Performance: 50-100 cycles (hash lookup + validation)
static inline ptr_classification_t registry_lookup(void* ptr) {
ptr_classification_t result = {
.kind = PTR_KIND_UNKNOWN,
.class_idx = -1,
.ss = NULL,
.slab_idx = -1
};
// Query SuperSlab registry
struct SuperSlab* ss = hak_super_lookup(ptr);
if (!ss) {
// Not in Tiny registry
return result;
}
// Found SuperSlab - determine slab index
result.ss = ss;
result.class_idx = ss->size_class;
// Calculate slab index
uintptr_t ptr_addr = (uintptr_t)ptr;
uintptr_t ss_addr = (uintptr_t)ss;
if (ptr_addr < ss_addr) {
// Pointer before SuperSlab base (invalid)
result.kind = PTR_KIND_UNKNOWN;
return result;
}
size_t offset = ptr_addr - ss_addr;
result.slab_idx = (int)(offset / SLAB_SIZE);
// Validate slab index (ss_slabs_capacity defined in superslab_inline.h)
if (result.slab_idx < 0 || result.slab_idx >= ss_slabs_capacity(ss)) {
// Out of range
result.kind = PTR_KIND_UNKNOWN;
return result;
}
// Valid Tiny allocation
// Only class 7 (1KB) is headerless. Other classes use header-based free path.
if (ss->size_class == 7) {
result.kind = PTR_KIND_TINY_HEADERLESS;
} else {
result.kind = PTR_KIND_TINY_HEADER;
}
return result;
}
// ========== Pool TLS Probe ==========
#ifdef HAKMEM_POOL_TLS_PHASE1
// Registry-based Pool TLS probe (no memory deref)
static inline int is_pool_tls_reg(void* ptr) {
pid_t tid = 0; int cls = -1;
return pool_reg_lookup(ptr, &tid, &cls);
}
#endif
// ========== Front Gate Entry Point ==========
ptr_classification_t classify_ptr(void* ptr) {
ptr_classification_t result = {
.kind = PTR_KIND_UNKNOWN,
.class_idx = -1,
.ss = NULL,
.slab_idx = -1
};
if (!ptr) return result;
// Early guard: reject non-canonical tiny integers to avoid ptr-1 probe crashes
if ((uintptr_t)ptr < 4096) {
result.kind = PTR_KIND_UNKNOWN;
return result;
}
// Step 1: Check Pool TLS via registry (no pointer deref)
#ifdef HAKMEM_POOL_TLS_PHASE1
if (is_pool_tls_reg(ptr)) {
result.kind = PTR_KIND_POOL_TLS;
#if !HAKMEM_BUILD_RELEASE
g_classify_pool_hit++;
#endif
return result;
}
#endif
// Step 2: Registry lookup for Tiny (header or headerless)
result = registry_lookup(ptr);
if (result.kind == PTR_KIND_TINY_HEADERLESS) {
#if !HAKMEM_BUILD_RELEASE
g_classify_headerless_hit++;
#endif
return result;
}
if (result.kind == PTR_KIND_TINY_HEADER) {
#if !HAKMEM_BUILD_RELEASE
g_classify_header_hit++;
#endif
return result;
}
// Step 3: Try AllocHeader (HAKMEM header) for Mid/Large/Mmap
do {
if (!ptr) break;
// Quick page-safety check: avoid crossing page for header read
uintptr_t off = (uintptr_t)ptr & 0xFFFu;
int safe_same_page = (off >= HEADER_SIZE);
void* raw = (char*)ptr - HEADER_SIZE;
if (!safe_same_page) {
if (!hak_is_memory_readable(raw)) break;
}
AllocHeader* hdr = (AllocHeader*)raw;
if (hdr->magic == HAKMEM_MAGIC) {
result.kind = PTR_KIND_MID_LARGE; // HAKMEM-owned (non-Tiny)
#if !HAKMEM_BUILD_RELEASE
g_classify_unknown_hit++; // reuse for stats without adding a new counter
#endif
return result;
}
} while (0);
// Step 4: Not recognized → UNKNOWN (route to libc or slow path)
result.kind = PTR_KIND_UNKNOWN;
#if !HAKMEM_BUILD_RELEASE
g_classify_unknown_hit++;
#endif
return result;
}