为什么atomic不是线程安全的

原理

1、读取i的值存入寄存器；
2、将i加1；
3、修改i的值；

修改一个属性包含了读，执行改变，写入，atomic 只为读和写加了锁，保证了同一个时间只有一个线程在读和写，但是，会遇到这种情况
执行改变的代码段并没有加锁，导致了有可能一个线程执行了读之后，还没有进行写，另一个线程进行了读，然后这个线程进行了写，
然后另一个线程也进行了写，导致这个线程的写被另一个线程的写覆盖了

举例子

@property (atomic, assign) NSInteger sliece;@end@implementation ThreadSafeController2- (void)viewDidLoad {[super viewDidLoad];[self testThreadSafe2];// Do any additional setup after loading the view.
}- (void)testThreadSafe2
{dispatch_queue_t queue = dispatch_queue_create("kkk", DISPATCH_QUEUE_CONCURRENT);dispatch_async(queue, ^{for (int i = 0; i < 10000; i ++) {self.sliece = self.sliece + 1;}});dispatch_async(queue, ^{for (int i = 0; i < 10000; i ++) {self.sliece = self.sliece + 1;}});dispatch_after(5, dispatch_get_main_queue(), ^{NSLog(@"哈哈值得大小%ld", self.sliece);});
}

结果：请添加图片描述

正确的写法

@interface ThreadSafeController2 ()@property (atomic, assign) NSInteger sliece;@property (nonatomic, strong) NSLock *lock;@end@implementation ThreadSafeController2- (void)viewDidLoad {[super viewDidLoad];self.lock = [[NSLock alloc] init];[self testThreadSafe2];// Do any additional setup after loading the view.
}- (void)testThreadSafe2
{dispatch_queue_t queue = dispatch_queue_create("kkk", DISPATCH_QUEUE_CONCURRENT);dispatch_async(queue, ^{for (int i = 0; i < 10000; i ++) {[self.lock lock];self.sliece = self.sliece + 1;[self.lock unlock];}});dispatch_async(queue, ^{for (int i = 0; i < 10000; i ++) {[self.lock lock];self.sliece = self.sliece + 1;[self.lock unlock];}});dispatch_after(5, dispatch_get_main_queue(), ^{NSLog(@"哈哈值得大小%ld", self.sliece);});
}

请添加图片描述