前言
类的组成如下图:
本文来探究cache_t,接下来进入正题。
cache_t源码
struct cache_t {
private:
explicit_atomic<uintptr_t> _bucketsAndMaybeMask;//8
union {
struct {
explicit_atomic<mask_t> _maybeMask; //4
#if __LP64__
uint16_t _flags; //2
#endif
uint16_t _occupied; //2
};
explicit_atomic<preopt_cache_t *> _originalPreoptCache;//8
};
#if CACHE_MASK_STORAGE == CACHE_MASK_STORAGE_OUTLINED
// _bucketsAndMaybeMask is a buckets_t pointer
// _maybeMask is the buckets mask
static constexpr uintptr_t bucketsMask = ~0ul;
static_assert(!CONFIG_USE_PREOPT_CACHES, "preoptimized caches not supported");
#elif CACHE_MASK_STORAGE == CACHE_MASK_STORAGE_HIGH_16_BIG_ADDRS
static constexpr uintptr_t maskShift = 48;
static constexpr uintptr_t maxMask = ((uintptr_t)1 << (64 - maskShift)) - 1;
static constexpr uintptr_t bucketsMask = ((uintptr_t)1 << maskShift) - 1;
static_assert(bucketsMask >= MACH_VM_MAX_ADDRESS, "Bucket field doesn't have enough bits for arbitrary pointers.");
#if CONFIG_USE_PREOPT_CACHES
static constexpr uintptr_t preoptBucketsMarker = 1ul;
static constexpr uintptr_t preoptBucketsMask = bucketsMask & ~preoptBucketsMarker;
#endif
#elif CACHE_MASK_STORAGE == CACHE_MASK_STORAGE_HIGH_16
// _bucketsAndMaybeMask is a buckets_t pointer in the low 48 bits
// _maybeMask is unused, the mask is stored in the top 16 bits.
// How much the mask is shifted by.
static constexpr uintptr_t maskShift = 48;
// Additional bits after the mask which must be zero. msgSend
// takes advantage of these additional bits to construct the value
// `mask << 4` from `_maskAndBuckets` in a single instruction.
static constexpr uintptr_t maskZeroBits = 4;
// The largest mask value we can store.
static constexpr uintptr_t maxMask = ((uintptr_t)1 << (64 - maskShift)) - 1;
// The mask applied to `_maskAndBuckets` to retrieve the buckets pointer.
static constexpr uintptr_t bucketsMask = ((uintptr_t)1 << (maskShift - maskZeroBits)) - 1;
。。。。。。
cache_t在不同的硬件上结构还不一样,下面分析真机64位处理器即CACHE_MASK_STORAGE_HIGH_16的cache_t
cache_t组成如下图
bucket_t主要是sel与imp,源码如下
struct bucket_t {
private:
// IMP-first is better for arm64e ptrauth and no worse for arm64.
// SEL-first is better for armv7* and i386 and x86_64.
#if __arm64__
explicit_atomic<uintptr_t> _imp;
explicit_atomic<SEL> _sel;
#else
explicit_atomic<SEL> _sel;
explicit_atomic<uintptr_t> _imp;
#endif
。。。
cache_t探索
1.lldb打印
🌰代码
#import <Foundation/Foundation.h>
@interface SLPerson : NSObject
@property(nonatomic, copy)NSString * lgName;
@property(nonatomic, strong)NSString * nickName;
-(void)sayHello;
-(void)sayCode;
-(void)sayMaster;
-(void)sayNB;
+(void)sayHappy;
@end
@implementation SLPerson
- (void)sayHello{
NSLog(@"SLPerson say : %s", __func__);
}
- (void)sayCode{
NSLog(@"SLPerson say : %s", __func__);
}
- (void)sayMaster{
NSLog(@"SLPerson say : %s", __func__);
}
- (void)sayNB{
NSLog(@"SLPerson say : %s", __func__);
}
+(void)sayHappy{
NSLog(@"SLPerson say : %s", __func__);
}
@end
int main(int argc, const char * argv[]) {
@autoreleasepool {
SLPerson * p = [SLPerson alloc];
Class pClass = [SLPerson class];
[p sayHello];
[p sayCode];
[p sayMaster];
NSLog(@"%@", pClass);
}
return 0;
}
(lldb) x/4gx pClass
0x100008258: 0x0000000100008230 0x00000001006dd140
0x100008268: 0x0000600002c0c080 0x0001802400000007
(lldb) p (cache_t *)0x00000001006dd150
(cache_t *) $1 = 0x00000001006dd150
(lldb) p *$1
(cache_t) $2 = {
_bucketsAndMaybeMask = {
std::__1::atomic<unsigned long> = {
Value = 105553140417216
}
}
= {
= {
_maybeMask = {
std::__1::atomic<unsigned int> = {
Value = 3
}
}
_flags = 32784
_occupied = 2
}
_originalPreoptCache = {
std::__1::atomic<preopt_cache_t *> = {
Value = 0x0002801000000003
}
}
}
}
(lldb) p $2.buckets()
(bucket_t *) $3 = 0x00006000017082c0
(lldb) p *$3
(bucket_t) $4 = {
_sel = {
std::__1::atomic<objc_selector *> = "" {
Value = ""
}
}
_imp = {
std::__1::atomic<unsigned long> = {
Value = 106176
}
}
}
(lldb) p $4.sel()
(SEL) $5 = "init"
(lldb) p $4.imp(&$4,pClass)
(IMP) $6 = 0x0000000100011c98
(lldb)
buckets是一个数组,所以不止一个缓存方法
(lldb) p $2.buckets()[1]
(bucket_t) $7 = {
_sel = {
std::__1::atomic<objc_selector *> = "" {
Value = ""
}
}
_imp = {
std::__1::atomic<unsigned long> = {
Value = 106224
}
}
}
(lldb) p $7.sel()
(SEL) $8 = "dealloc"
2.脱离objc源码探索
🌰代码
typedef uint32_t mask_t; // x86_64 & arm64 asm are less efficient with 16-bits
struct sl_bucket_t {
SEL _sel;
IMP _imp;
};
struct sl_cache_t {
struct sl_bucket_t * _buckets;
mask_t _mask;
uint16_t _flags;
uint16_t _occupied;
};
struct sl_class_data_bits_t {
uintptr_t bits;
};
struct sl_objc_class {
Class ISA;
Class superclass;
struct sl_cache_t cache; // formerly cache pointer and vtable
struct sl_class_data_bits_t bits; // class_rw_t * plus custom rr/alloc flags
};
@interface SLPerson : NSObject
@property (nonatomic, copy) NSString *lgName;
@property (nonatomic, strong) NSString *nickName;
- (void)say1;
- (void)say2;
- (void)say3;
- (void)say4;
- (void)say5;
- (void)say6;
- (void)say7;
+ (void)sayHappy;
@end
@implementation SLPerson
- (void)say1{
NSLog(@"SLPerson say : %s",__func__);
}
- (void)say2{
NSLog(@"SLPerson say : %s",__func__);
}
- (void)say3{
NSLog(@"SLPerson say : %s",__func__);
}
- (void)say4{
NSLog(@"SLPerson say : %s",__func__);
}
- (void)say5{
NSLog(@"SLPerson say : %s",__func__);
}
- (void)say6{
NSLog(@"SLPerson say : %s",__func__);
}
- (void)say7{
NSLog(@"SLPerson say : %s",__func__);
}
+ (void)sayHappy{
NSLog(@"SLPerson say : %s",__func__);
}
@end
int main(int argc, const char * argv[]) {
@autoreleasepool {
SLPerson *p = [SLPerson alloc];
Class pClass = [SLPerson class]; // objc_clas
[p say1];
[p say2];
//[p say3];
//[p say4];
struct sl_objc_class *sl_pClass = (__bridge struct sl_objc_class *)(pClass);
NSLog(@"%hu - %u",sl_pClass->cache._occupied,sl_pClass->cache._mask);
for (mask_t i = 0; i<sl_pClass->cache._mask; i++) {
// 打印获取的 bucket
struct sl_bucket_t bucket = sl_pClass->cache._buckets[i];
NSLog(@"%@ - %p",NSStringFromSelector(bucket._sel),bucket._imp);
}
NSLog(@"Hello, World!");
}
return 0;
}
打印如下
2023-02-08 15:14:38.367995+0800 KCObjcBuild[3413:71657] SLPerson say : -[SLPerson say1]
2023-02-08 15:14:38.368470+0800 KCObjcBuild[3413:71657] SLPerson say : -[SLPerson say2]
2023-02-08 15:14:38.368530+0800 KCObjcBuild[3413:71657] 2 - 3
2023-02-08 15:14:38.368573+0800 KCObjcBuild[3413:71657] (null) - 0x0
2023-02-08 15:14:38.368683+0800 KCObjcBuild[3413:71657] say1 - 0xb818
2023-02-08 15:14:38.368749+0800 KCObjcBuild[3413:71657] say2 - 0xb828
2023-02-08 15:14:38.368783+0800 KCObjcBuild[3413:71657] Hello, World!
Program ended with exit code: 0
去掉注释,调用say3和say4,输出如下
2023-02-08 15:18:01.703909+0800 KCObjcBuild[3453:74143] SLPerson say : -[SLPerson say1]
2023-02-08 15:18:01.704488+0800 KCObjcBuild[3453:74143] SLPerson say : -[SLPerson say2]
2023-02-08 15:18:01.704598+0800 KCObjcBuild[3453:74143] SLPerson say : -[SLPerson say3]
2023-02-08 15:18:01.704660+0800 KCObjcBuild[3453:74143] SLPerson say : -[SLPerson say4]
2023-02-08 15:18:01.704707+0800 KCObjcBuild[3453:74143] 2 - 7
2023-02-08 15:18:01.704750+0800 KCObjcBuild[3453:74143] (null) - 0x0
2023-02-08 15:18:01.704857+0800 KCObjcBuild[3453:74143] say3 - 0xb868
2023-02-08 15:18:01.704894+0800 KCObjcBuild[3453:74143] (null) - 0x0
2023-02-08 15:18:01.704925+0800 KCObjcBuild[3453:74143] (null) - 0x0
2023-02-08 15:18:01.704954+0800 KCObjcBuild[3453:74143] (null) - 0x0
2023-02-08 15:18:01.704984+0800 KCObjcBuild[3453:74143] (null) - 0x0
2023-02-08 15:18:01.705041+0800 KCObjcBuild[3453:74143] say4 - 0xb858
2023-02-08 15:18:01.705075+0800 KCObjcBuild[3453:74143] Hello, World!
Program ended with exit code: 0
发现随着方法增多,_mask会变化,在源码中发现有个方法incrementOccupied
void incrementOccupied();
void cache_t::incrementOccupied()
{
_occupied++;
}
这个函数在cache_t
的insert
方法有调用
下面分析insert代码,主要分为几个部分:
- 计算当前缓存占用量
- 根据缓存占用量判断执行的操作
- 针对需要存储的
bucket
进行内部imp和sel赋值
a. 算当前缓存占用量
mask_t newOccupied = occupied() + 1;
对象被创建后occupied在重新调用init、属性赋值(调set方法)、其他方法调用,occupied都
会增加,总的来说就是在调用方法的时候occupied增加。
b.
根据缓存占用判断要执行的操作
第一次创建,开辟4个容量
if (slowpath(isConstantEmptyCache())) {
// Cache is read-only. Replace it.
if (!capacity) capacity = INIT_CACHE_SIZE;
reallocate(oldCapacity, capacity, /* freeOld */false);
}
容量小于等于3/4,不做处理
else if (fastpath(newOccupied + CACHE_END_MARKER <= cache_fill_ratio(capacity))) {
// Cache is less than 3/4 or 7/8 full. Use it as-is.
}
// Historical fill ratio of 75% (since the new objc runtime was introduced).
static inline mask_t cache_fill_ratio(mask_t capacity) {
return capacity * 3 / 4;
}
超过3/4,则两倍扩容并重新开辟空间
else {
capacity = capacity ? capacity * 2 : INIT_CACHE_SIZE;
if (capacity > MAX_CACHE_SIZE) {
capacity = MAX_CACHE_SIZE;
}
reallocate(oldCapacity, capacity, true);
}
c.针对需要存储的bucket
进行内部imp和sel赋值
bucket_t *b = buckets();
mask_t m = capacity - 1;
mask_t begin = cache_hash(sel, m);
mask_t i = begin;
// Scan for the first unused slot and insert there.
// There is guaranteed to be an empty slot.
do {
if (fastpath(b[i].sel() == 0)) {
incrementOccupied();
b[i].set<Atomic, Encoded>(b, sel, imp, cls());
return;
}
if (b[i].sel() == sel) {
// The entry was added to the cache by some other thread
// before we grabbed the cacheUpdateLock.
return;
}
} while (fastpath((i = cache_next(i, m)) != begin));
这部分主要根据哈希算法(cache_hash),计算sel-imp
存储的哈希下标。
当前哈希下标存储的sel 不等于
即将插入的sel,则通过哈希冲突算法(cache_next)
,重新进行哈希计算,得到新的下标,再去对比进行存储。
哈希算法
static inline mask_t cache_hash(SEL sel, mask_t mask)
{
uintptr_t value = (uintptr_t)sel;
#if CONFIG_USE_PREOPT_CACHES
value ^= value >> 7;
#endif
return (mask_t)(value & mask);
}
哈希冲突算法
#if CACHE_END_MARKER
static inline mask_t cache_next(mask_t i, mask_t mask) {
return (i+1) & mask;
}
#elif __arm64__
static inline mask_t cache_next(mask_t i, mask_t mask) {
return i ? i-1 : mask;
}
#else
总结
_mask
——掩码数据
:在哈希算法
或者哈希冲突算法
中计算哈希下标
,其中mask
等于capacity - 1
。- _occupied——哈希表中sel-imp的个数,方法调用会使_occupied变化
cache分析流程如下(图片转载自一荤一素一碗粥,感谢🙏)