iOS - 多线程的安全隐患

• 资源共享

• • 1块资源可能会被多个线程共享，也就是多个线程可能会访问同一块资源
• • 比如多个线程访问同一个对象、同一个变量、同一个文件

• 当多个线程访问同一块资源时，很容易引发数据错乱和数据安全问题

1. 卖票案例

假设总共15张票，使用3个异步线程并发执行，每个线程卖5张票，观察最后的打印结果

@interface ViewController ()@property (nonatomic, assign) int ticketsCount;@end@implementation ViewController- (void)viewDidLoad {[super viewDidLoad];// Do any additional setup after loading the view.
}- (void)saleTicket {int oldTicketsCount = self.ticketsCount;sleep(0.2);oldTicketsCount--;self.ticketsCount = oldTicketsCount;NSLog(@"还剩 %d 张票", oldTicketsCount);
}- (void)touchesBegan:(NSSet<UITouch *> *)touches withEvent:(UIEvent *)event {self.ticketsCount = 15;dispatch_queue_t queue = dispatch_get_global_queue(0, 0);dispatch_async(queue, ^{for (int i = 0; i < 5; i++) {[self saleTicket];}});dispatch_async(queue, ^{for (int i = 0; i < 5; i++) {[self saleTicket];}});dispatch_async(queue, ^{for (int i = 0; i < 5; i++) {[self saleTicket];}});
}@end

出现票重复卖票，最后没卖完的情况

时序图示意：

分析图：

2. 多线程安全隐患的解决方案

• 解决方案：使用线程同步技术（同步，就是协同步调，按预定的先后次序进行）
• 常见的线程同步技术是：加锁

2.1 iOS中的线程同步方案

• OSSpinLock
• os_unfair_lock
• pthread_mutex
• dispatch_semaphore
• dispatch_queue(DISPATCH_QUEUE_SERIAL)
• NSLock
• NSRecursiveLock
• NSCondition
• NSConditionLock
• @synchronized

2.2 同步方案的使用

为了方便调试，我们做个简单封装，将需要加锁的代码放在ZSXBaseDemo中，同时暴漏方法给子类重写：

• - (void)__saveMoney;
• - (void)__drawMoney;
• - (void)__saleTicket;

每一把锁将创建一个ZSXBaseDemo子类，然后重写上述方法，实现各自的加锁方式

ZSXBaseDemo.h

@interface ZSXBaseDemo : NSObject- (void)moneyTest;
- (void)ticketTest;
- (void)otherTest;#pragma mark - 暴露给子类去使用
- (void)__saveMoney;
- (void)__drawMoney;
- (void)__saleTicket;@end

ZSXBaseDemo.m

@interface ZSXBaseDemo()@property (assign, nonatomic) int money;
@property (assign, nonatomic) int ticketsCount;@end@implementation ZSXBaseDemo- (void)otherTest {}/**存钱、取钱演示*/
- (void)moneyTest
{self.money = 100;dispatch_queue_t queue = dispatch_get_global_queue(0, 0);dispatch_async(queue, ^{for (int i = 0; i < 10; i++) {[self __saveMoney];}});dispatch_async(queue, ^{for (int i = 0; i < 10; i++) {[self __drawMoney];}});
}/**存钱*/
- (void)__saveMoney
{int oldMoney = self.money;sleep(.2);oldMoney += 50;self.money = oldMoney;NSLog(@"存50，还剩%d元 - %@", oldMoney, [NSThread currentThread]);
}/**取钱*/
- (void)__drawMoney
{int oldMoney = self.money;sleep(.2);oldMoney -= 20;self.money = oldMoney;NSLog(@"取20，还剩%d元 - %@", oldMoney, [NSThread currentThread]);
}/**卖1张票*/
- (void)__saleTicket
{int oldTicketsCount = self.ticketsCount;sleep(.2);oldTicketsCount--;self.ticketsCount = oldTicketsCount;NSLog(@"还剩%d张票 - %@", oldTicketsCount, [NSThread currentThread]);
}/**卖票演示*/
- (void)ticketTest
{self.ticketsCount = 15;dispatch_queue_t queue = dispatch_get_global_queue(0, 0);//    for (int i = 0; i < 10; i++) {
//        [[[NSThread alloc] initWithTarget:self selector:@selector(__saleTicket) object:nil] start];
//    }dispatch_async(queue, ^{for (int i = 0; i < 5; i++) {[self __saleTicket];}});dispatch_async(queue, ^{for (int i = 0; i < 5; i++) {[self __saleTicket];}});dispatch_async(queue, ^{for (int i = 0; i < 5; i++) {[self __saleTicket];}});
}@end

2.2.1 OSSpinLock

• OSSpinLock叫做"自旋锁"，等待锁的线程会处于忙等（busy-wait）状态，一直占用着CPU资源
• 目前已经不再安全，可能会出现优先级反转问题
• • 如果等待锁的线程优先级较高，它会一直占用着CPU资源，优先级低的线程就无法释放锁
• • 需要导入头文件#import <libkern/OSAtomic.h>

2.2.1.1 使用方法：

// 初始化
OSSpinLock lock = OS_SPINLOCK_INIT;// 尝试加锁
bool result =  OSSpinLockTry(&lock);// 加锁
OSSpinLockLock(&lock);// 解锁
OSSpinLockUnlock(&lock);

2.2.1.2 案例

#import "OSSpinLockDemo.h"
#import <libkern/OSAtomic.h>@interface OSSpinLockDemo ()// High-level lock
// 自旋锁，有优先级反转问题
@property (nonatomic, assign) OSSpinLock moneyLock;
@property (nonatomic, assign) OSSpinLock ticketLock;@end@implementation OSSpinLockDemo- (instancetype)init {if (self = [super init]) {_moneyLock = OS_SPINLOCK_INIT;_ticketLock = OS_SPINLOCK_INIT;//使用方法// 初始化OSSpinLock lock = OS_SPINLOCK_INIT;// 尝试加锁bool result =  OSSpinLockTry(&lock);// 加锁OSSpinLockLock(&lock);// 解锁OSSpinLockUnlock(&lock);}return self;
}#pragma mark reload
- (void)__drawMoney {OSSpinLockLock(&_moneyLock);[super __drawMoney];OSSpinLockUnlock(&_moneyLock);
}- (void)__saveMoney {OSSpinLockLock(&_moneyLock);[super __saveMoney];OSSpinLockUnlock(&_moneyLock);
}- (void)__saleTicket {OSSpinLockLock(&_ticketLock);[super __saleTicket];OSSpinLockUnlock(&_ticketLock);
}
#pragma mark end@end

对于卖票，只有__saleTicket方法中需要使用锁，因此也可以使用static关键字创建静态变量，无需声明为属性

- (void)__saleTicket {static OSSpinLock ticketLock = OS_SPINLOCK_INIT;OSSpinLockLock(&ticketLock);[super __saleTicket];OSSpinLockUnlock(&ticketLock);
}

2.2.2 os_unfair_lock

• os_unfair_lock用于取代不安全的OSSpinLock ，从iOS10开始才支持
• 从底层调用看，等待os_unfair_lock锁的线程会处于休眠状态，并非忙等
• 需要导入头文件#import <os/lock.h>

2.2.2.1 使用方法：

// 初始化
os_unfair_lock lock = OS_UNFAIR_LOCK_INIT;// 尝试加锁
bool result =  os_unfair_lock_trylock(&lock);// 加锁
os_unfair_lock_lock(&lock);// 解锁
os_unfair_lock_unlock(&lock);

2.2.2.2 案例

#import "OSUnfairLockDemo.h"
#import <os/lock.h>@interface OSUnfairLockDemo()
// Low-level lock
// ll lock
// lll
// Low-level lock的特点等不到锁就休眠
@property (assign, nonatomic) os_unfair_lock moneyLock;
@property (assign, nonatomic) os_unfair_lock ticketLock;
@end@implementation OSUnfairLockDemo- (instancetype)init {if (self = [super init]) {_moneyLock = OS_UNFAIR_LOCK_INIT;_ticketLock = OS_UNFAIR_LOCK_INIT;//使用方法// 初始化os_unfair_lock lock = OS_UNFAIR_LOCK_INIT;// 尝试加锁bool result =  os_unfair_lock_trylock(&lock);// 加锁os_unfair_lock_lock(&lock);// 解锁os_unfair_lock_unlock(&lock);}return self;
}#pragma mark reload
- (void)__drawMoney {os_unfair_lock_lock(&_moneyLock);[super __drawMoney];os_unfair_lock_unlock(&_moneyLock);
}- (void)__saveMoney {os_unfair_lock_lock(&_moneyLock);[super __saveMoney];os_unfair_lock_unlock(&_moneyLock);
}- (void)__saleTicket {os_unfair_lock_lock(&_ticketLock);[super __saleTicket];os_unfair_lock_unlock(&_ticketLock);
}
#pragma mark end@end

2.2.3 pthread_mutex

• mutex叫做”互斥锁”，等待锁的线程会处于休眠状态
• 需要导入头文件#import <pthread.h>

2.2.3.1 使用方法

// 初始化锁
pthread_mutex_init(mutex, NULL); //相当于设置属性"PTHREAD_MUTEX_DEFAULT"//初始化属性并设置属性
pthread_mutexattr_t attr;
pthread_mutexattr_init(&attr);
pthread_mutexattr_settype(&attr, PTHREAD_MUTEX_DEFAULT);//初始化锁
pthread_mutex_init(mutex, &attr);//加锁
pthread_mutex_lock(&_mutex);//尝试加锁
bool result = pthread_mutex_trylock(&_mutex);//解锁
pthread_mutex_unlock(&_mutex);//销毁相关资源
pthread_mutexattr_destroy(&attr); //销毁属性
pthread_mutex_destroy(&_mutex); //销毁锁

2.2.3.2 案例

@interface MutexDemo ()@property (assign, nonatomic) pthread_mutex_t ticketMutex;
@property (assign, nonatomic) pthread_mutex_t moneyMutex;@end@implementation MutexDemo- (void)__initMutex:(pthread_mutex_t *)mutex {// 初始化锁pthread_mutex_init(mutex, NULL);
}- (instancetype)init {if (self = [super init]) {[self __initMutex:&_ticketMutex];[self __initMutex:&_moneyMutex];}return self;
}#pragma mark reload
- (void)__drawMoney {pthread_mutex_lock(&_moneyMutex);[super __drawMoney];pthread_mutex_unlock(&_moneyMutex);
}- (void)__saveMoney {pthread_mutex_lock(&_moneyMutex);[super __saveMoney];pthread_mutex_unlock(&_moneyMutex);
}- (void)__saleTicket {pthread_mutex_lock(&_ticketMutex);[super __saleTicket];pthread_mutex_unlock(&_ticketMutex);
}
#pragma mark end- (void)dealloc {pthread_mutex_destroy(&_moneyMutex);pthread_mutex_destroy(&_ticketMutex);
}@end

2.2.3.3 pthread_mutex-递归锁

递归：允许同一个线程对一把锁重复加锁

在otherTest方法中，假设该方法中已经加锁，同时会调用另一个也需要加锁的方法

- (void)otherTest {pthread_mutex_lock(&_mutex);NSLog(@"%s", __func__);[self otherTest2];pthread_mutex_unlock(&_mutex);
}- (void)otherTest2 {pthread_mutex_lock(&_mutex);NSLog(@"%s", __func__);pthread_mutex_unlock(&_mutex);
}

此时执行代码

代码只会执行到[self otherTest2];。此时出现死锁，是因为otherTest方法加锁后，调用otherTest2，otherTest2方法开始执行时也会加锁，此时因为otherTest方法还未解锁，otherTest2则进入等待解锁状态，而otherTest需要等待otherTest2方法执行完才继续，所以产生死锁

使用pthread_mutexattr_t配置该锁为递归锁：
PTHREAD_MUTEX_RECURSIVE

// 初始化锁
//    pthread_mutex_init(mutex, NULL);//初始化属性
pthread_mutexattr_t attr;
pthread_mutexattr_init(&attr);
pthread_mutexattr_settype(&attr, PTHREAD_MUTEX_RECURSIVE);
//初始化锁
pthread_mutex_init(mutex, &attr);
//销毁属性
pthread_mutexattr_destroy(&attr);

此时可以正常执行完otherTest、otherTest2方法

如果otherTest方法里面是递归调用otherTest自身

- (void)otherTest {pthread_mutex_lock(&_mutex);NSLog(@"%s", __func__);[self otherTest];pthread_mutex_unlock(&_mutex);
}

这时候会死循环调用otherTest

若增加一个计数，即可控制递归调用的次数

- (void)otherTest {pthread_mutex_lock(&_mutex);NSLog(@"%s", __func__);static int count = 0;if (count < 5) {count++;[self otherTest];}pthread_mutex_unlock(&_mutex);
}

执行结果：

2.2.3.4 pthread_mutex-条件锁

业务场景中，可能需要不同线程间的依赖关系，比如线程1需要等待线程2执行完才能继续执行

假设有如下代码：

- (void)otherTest {[[[NSThread alloc] initWithTarget:self selector:@selector(__remove) object:nil] start];[[[NSThread alloc] initWithTarget:self selector:@selector(__add) object:nil] start];
}- (void)__remove {NSLog(@"%s", __func__);pthread_mutex_lock(&_mutex);[self.data removeLastObject];NSLog(@"删除一个元素");pthread_mutex_unlock(&_mutex);
}- (void)__add {NSLog(@"%s", __func__);pthread_mutex_lock(&_mutex);[self.data addObject:@"test"];NSLog(@"添加一个元素");pthread_mutex_unlock(&_mutex);
}

使用多线程调用__remove和__add，两者执行顺序不确定。但是希望__remove是在__add之后执行，保证先加、再减。

这时候可以使用条件锁来实现

2.2.3.4.1 使用方法

// 初始化锁
pthread_t mutex;
pthread_mutex_init(&mutex, NULL);
// 初始化条件
pthread_cond_t condition;
pthread_cond_init(&condition, NULL);
// 等待条件（进入休眠，放开mutex 锁；被唤醒后，会再次对mutex加锁）
pthread_cond_wait(&condition, &mutex);
// 激活一个等待条件的线程
pthread_cond_signal(&condition);
// 激活所有等待条件的线程
pthread_cond_broadcast(&condition);
// 销毁资源
pthread_mutex_destroy(&mutex);
pthread_cond_destroy(&condition);

2.2.3.4.2 案例

#import "MutexDemo3.h"
#import <pthread.h>@interface MutexDemo3 ()@property (nonatomic, strong) NSMutableArray *data;
@property (assign, nonatomic) pthread_mutex_t mutex;
@property (nonatomic, assign) pthread_cond_t cond;@end@implementation MutexDemo3- (instancetype)init {if (self = [super init]) {_data = [[NSMutableArray alloc] init];//初始化属性pthread_mutexattr_t attr;pthread_mutexattr_init(&attr);pthread_mutexattr_settype(&attr, PTHREAD_MUTEX_RECURSIVE);//初始化锁pthread_mutex_init(&_mutex, &attr);//初始化条件pthread_cond_init(&_cond, NULL);//销毁属性pthread_mutexattr_destroy(&attr);}return self;
}- (void)otherTest {[[[NSThread alloc] initWithTarget:self selector:@selector(__remove) object:nil] start];[[[NSThread alloc] initWithTarget:self selector:@selector(__add) object:nil] start];
}- (void)__remove {NSLog(@"%s", __func__);pthread_mutex_lock(&_mutex);if (self.data.count == 0) {//等待条件（进入休眠，放开mutex 锁；被唤醒后，会再次对mutex加锁）pthread_cond_wait(&_cond, &_mutex);}[self.data removeLastObject];NSLog(@"删除一个元素");pthread_mutex_unlock(&_mutex);
}- (void)__add {NSLog(@"%s", __func__);pthread_mutex_lock(&_mutex);[self.data addObject:@"test"];NSLog(@"添加一个元素");//激活等待条件pthread_cond_signal(&_cond);pthread_mutex_unlock(&_mutex);
}- (void)dealloc {pthread_mutex_destroy(&_mutex);pthread_cond_destroy(&_cond);
}@end

执行结果：

可以保证先添加，再删除

2.2.4 dispatch_semaphore

2.2.5 dispatch_queue(DISPATCH_QUEUE_SERIAL)

2.2.6 NSLock、NSRecursiveLock

2.2.6.1 NSLock是对mutex普通锁的封装

2.2.6.1.1 使用方式

2.2.6.2 NSRecursiveLock也是对mutex递归锁的封装，API跟NSLock基本一致

2.2.6.3 案例

2.2.6.3.1 NSLock

#import "NSLockDemo.h"@interface NSLockDemo()
@property (strong, nonatomic) NSLock *ticketLock;
@property (strong, nonatomic) NSLock *moneyLock;
@end@implementation NSLockDemo- (instancetype)init
{if (self = [super init]) {self.ticketLock = [[NSLock alloc] init];self.moneyLock = [[NSLock alloc] init];}return self;
}- (void)__saleTicket
{[self.ticketLock lock];[super __saleTicket];[self.ticketLock unlock];
}- (void)__saveMoney
{[self.moneyLock lock];[super __saveMoney];[self.moneyLock unlock];
}- (void)__drawMoney
{[self.moneyLock lock];[super __drawMoney];[self.moneyLock unlock];
}@end

2.2.6.3.2 NSRecursiveLock

#import "NSLockDemo2.h"@interface NSLockDemo2()@property (strong, nonatomic) NSRecursiveLock *recursiveLock;@end@implementation NSLockDemo2- (instancetype)init
{if (self = [super init]) {self.recursiveLock = [[NSRecursiveLock alloc] init];}return self;
}- (void)otherTest {[self.recursiveLock lock];NSLog(@"%s", __func__);static int count = 0;if (count < 5) {count++;[self otherTest];}[self.recursiveLock unlock];
}- (void)otherTest2 {[self.recursiveLock lock];NSLog(@"%s", __func__);[self.recursiveLock unlock];
}@end

执行结果：

2.2.7 NSCondition

• NSCondition是对mutex和cond的封装

2.2.7.1 案例

#import "NSConditionDemo.h"@interface NSConditionDemo()@property (strong, nonatomic) NSCondition *condition;
@property (strong, nonatomic) NSMutableArray *data;@end@implementation NSConditionDemo- (instancetype)init
{if (self = [super init]) {self.condition = [[NSCondition alloc] init];self.data = [NSMutableArray array];}return self;
}- (void)otherTest
{[[[NSThread alloc] initWithTarget:self selector:@selector(__remove) object:nil] start];[[[NSThread alloc] initWithTarget:self selector:@selector(__add) object:nil] start];
}// 生产者-消费者模式// 线程1
// 删除数组中的元素
- (void)__remove
{[self.condition lock];NSLog(@"__remove - begin");if (self.data.count == 0) {// 等待[self.condition wait];}[self.data removeLastObject];NSLog(@"删除了元素");[self.condition unlock];
}// 线程2
// 往数组中添加元素
- (void)__add
{[self.condition lock];sleep(1);[self.data addObject:@"Test"];NSLog(@"添加了元素");// 信号[self.condition signal];// 广播
//    [self.condition broadcast];[self.condition unlock];}
@end

执行结果：

2.2.8 NSConditionLock

• NSConditionLock是对NSCondition的进一步封装，可以设置具体的条件值

2.2.8.1 案例

#import "NSConditionLockDemo.h"@interface NSConditionLockDemo()@property (strong, nonatomic) NSConditionLock *conditionLock;@end@implementation NSConditionLockDemo- (instancetype)init
{if (self = [super init]) {self.conditionLock = [[NSConditionLock alloc] initWithCondition:1];}return self;
}- (void)otherTest
{[[[NSThread alloc] initWithTarget:self selector:@selector(__one) object:nil] start];[[[NSThread alloc] initWithTarget:self selector:@selector(__two) object:nil] start];[[[NSThread alloc] initWithTarget:self selector:@selector(__three) object:nil] start];
}- (void)__one {[self.conditionLock lockWhenCondition:1];NSLog(@"%s", __func__);sleep(1);[self.conditionLock unlockWithCondition:2];
}- (void)__two {[self.conditionLock lockWhenCondition:2];NSLog(@"%s", __func__);sleep(1);[self.conditionLock unlockWithCondition:3];
}- (void)__three {[self.conditionLock lockWhenCondition:3];NSLog(@"%s", __func__);[self.conditionLock unlock];
}@end

执行结果：

通过设置Condition，可以实现按想要的顺序执行任务，或者说任务之间的依赖关系

condition默认值是0
即：
使用[[NSConditionLock alloc] init]初始化，condition为 0

使用- (void)lock代表直接加锁
即：
[self.conditionLock lock]

2.2.9 dispatch_queue - 串行队列

• 直接使用GCD的串行队列，也是可以实现线程同步的

2.2.9.1 案例

#import "SerialQueueDemo.h"@interface SerialQueueDemo()@property (nonatomic, strong) dispatch_queue_t ticketQueue;
@property (nonatomic, strong) dispatch_queue_t moneyQueue;@end@implementation SerialQueueDemo- (instancetype)init {if (self = [super init]) {self.ticketQueue = dispatch_queue_create("ticketQueue", DISPATCH_QUEUE_SERIAL);self.moneyQueue = dispatch_queue_create("moneyQueue", DISPATCH_QUEUE_SERIAL);}return self;
}- (void)__saveMoney {dispatch_sync(self.moneyQueue, ^{[super __saveMoney];});
}- (void)__drawMoney {dispatch_sync(self.moneyQueue, ^{[super __drawMoney];});
}- (void)__saleTicket {dispatch_sync(self.ticketQueue, ^{[super __saleTicket];});
}@end

执行结果：

2.2.10 dispatch_semaphore - 信号量

• semaphore叫做”信号量”
• 信号量的初始值，可以用来控制线程并发访问的最大数量
• 信号量的初始值为1，代表同时只允许1条线程访问资源，保证线程同步

2.2.10.1 控制最大并发数

- (void)otherTest方法循环创建20个线程执行- (void)test方法，semaphore初始值设置为5

#import "SemaphoreDemo.h"@interface SemaphoreDemo()@property (nonatomic, strong) dispatch_semaphore_t semaphore;@end@implementation SemaphoreDemo- (instancetype)init {if (self = [super init]) {self.semaphore = dispatch_semaphore_create(5);}return self;
}- (void)otherTest {for (int i = 0; i < 20; i++) {[[[NSThread alloc] initWithTarget:self selector:@selector(test) object:nil] start];}
}- (void)test {dispatch_semaphore_wait(self.semaphore, DISPATCH_TIME_FOREVER);sleep(2);NSLog(@"test - %@", [NSThread currentThread]);dispatch_semaphore_signal(self.semaphore);
}@end

运行结果：

实现了控制最大并发数5

2.2.10.2 保证线程同步

@interface SemaphoreDemo()@property (nonatomic, strong) dispatch_semaphore_t semaphore;
@property (nonatomic, strong) dispatch_semaphore_t tacketSemaphore;
@property (nonatomic, strong) dispatch_semaphore_t moneySemaphore;@end@implementation SemaphoreDemo- (instancetype)init {if (self = [super init]) {self.semaphore = dispatch_semaphore_create(5);self.tacketSemaphore = dispatch_semaphore_create(1);self.moneySemaphore = dispatch_semaphore_create(1);}return self;
}- (void)otherTest {for (int i = 0; i < 20; i++) {[[[NSThread alloc] initWithTarget:self selector:@selector(test) object:nil] start];}
}- (void)test {dispatch_semaphore_wait(self.semaphore, DISPATCH_TIME_FOREVER);sleep(2);NSLog(@"test - %@", [NSThread currentThread]);dispatch_semaphore_signal(self.semaphore);
}- (void)__saveMoney {dispatch_semaphore_wait(self.moneySemaphore, DISPATCH_TIME_FOREVER);[super __saveMoney];dispatch_semaphore_signal(self.moneySemaphore);
}- (void)__drawMoney {dispatch_semaphore_wait(self.moneySemaphore, DISPATCH_TIME_FOREVER);[super __drawMoney];dispatch_semaphore_signal(self.moneySemaphore);
}- (void)__saleTicket {dispatch_semaphore_wait(self.tacketSemaphore, DISPATCH_TIME_FOREVER);[super __saleTicket];dispatch_semaphore_signal(self.tacketSemaphore);
}@end

执行结果：

虽然打印结果已经保证线程同步，但是窗口收到了警告

2.2.10.2.1 使用信号量可能会造成线程优先级反转，且无法避免

QoS （Quality of Service），用来指示某任务或者队列的运行优先级；

1. 记录了持有者的api都可以自动避免优先级反转，系统会通过提高相关线程的优先级来解决优先级反转的问题，如 dispatch_sync, 如果系统不知道持有者所在的线程，则无法知道应该提高谁的优先级，也就无法解决反转问题。

2. 慎用dispatch_semaphore做线程同步

dispatch_semaphore容易造成优先级反转，因为api没有记录是哪个线程持有了信号量，所以有高优先级的线程在等待锁的时候，内核无法知道该提高那个线程的优先级（QoS）；

3. dispatch_semaphore不能避免优先级反转的原因

在调用dispatch_semaphore_wait()的时候，系统不知道哪个线程会调用 dispatch_semaphore_signal()方法，系统无法知道owner信息，无法调整优先级。dispatch_group和semaphore类似，在调用enter()方法的时候，无法预知谁会leave()，所以系统也不知道owner信息

2.2.11 @synchronized

• @synchronized是对mutex递归锁的封装
• 源码查看：objc4中的objc-sync.mm文件
• @synchronized(obj)内部会生成obj对应的递归锁，然后进行加锁、解锁操作

2.2.11.1 底层原理

使用哈希表结构，将穿进去的obj作为key，找到低层封装mutex的锁，再进行加锁、解锁操作

@synchronized (obj)：obj如果相同，则代表使用同一把锁

2.2.11.1 案例

#import "SynchronizedDemo.h"@implementation SynchronizedDemo- (void)__saveMoney {// 取钱、存钱 共用一把锁@synchronized (self) {[super __saveMoney];}
}- (void)__drawMoney {@synchronized (self) {[super __drawMoney];}
}- (void)__saleTicket {// 买票 - 单独创建一把锁static NSObject *lock;static dispatch_once_t onceToken;dispatch_once(&onceToken, ^{lock = [[NSObject alloc] init];});@synchronized (lock) {[super __saleTicket];}
}@end

3. 拓展

3.1 iOS线程同步方案性能比较

性能从高到低排序

• os_unfair_lock
• OSSpinLock
• dispatch_semaphore
• pthread_mutex
• dispatch_queue(DISPATCH_QUEUE_SERIAL)
• NSLock
• NSCondition
• pthread_mutex(recursive)
• NSRecursiveLock
• NSConditionLock
• @synchronized

性能排行仅供参考，不同环境实际效果可能不一样

3.2 自旋锁、互斥锁比较

3.2.1 什么情况使用自旋锁比较划算？

• 预计线程等待锁的时间很短
• 加锁的代码（临界区）经常被调用，但竞争情况很少发生
• CPU资源不紧张
• 多核处理器

3.2.2 什么情况使用互斥锁比较划算？

• 预计线程等待锁的时间较长
• 单核处理器
• 临界区有IO操作
• 临界区代码复杂或者循环量大
• 临界区竞争非常激烈

@oubijiexi