GCD (1)

dispatch_async 与 dispatch_sync 到区别

// 将block 提交到 queueA 上执行，提交到 queueA之后，继续往下执行，不等等block执行完毕，强调提交之后继续往下执行，执行，还是在当前线程进行
dispatch_async(queueA, ^{
    //queueA 线程进行执行
});

// 除非queueA 是dispatch_get_main or target 是 dispatch_get_main， 否则就在当前线程进行执行，做为一个优化操作。直到block 执行完毕之后，才会顺序执行
dispatch_sync(queueA, ^{
    //当前线程进行执行
});

看如下代码，运行如何

dispatch_queue_t queue3 = dispatch_queue_create("t3.0", DISPATCH_QUEUE_CONCURRENT);
NSLog(@"0 %@", NSThread.currentThread);
dispatch_sync(queue3, ^{
    NSLog(@"1 %@", NSThread.currentThread);
    dispatch_sync(queue3, ^{
        NSLog(@"2 %@", NSThread.currentThread);
    });
    NSLog(@"3 %@", NSThread.currentThread);
});

NSLog(@"4 %@", NSThread.currentThread);

输出结果是 0，1，2，3，4， sync 为何没有 死锁，因为是queue3 是并发线程，会有多个执行路径，所以不会死锁，输出的顺序是，提交点的顺序，所以是 按顺序输出, 对应的线程，都是 主线程，这个是线程优化，因为他们跑在主线程中

    dispatch_queue_t queue1 = dispatch_queue_create("t1.0", DISPATCH_QUEUE_SERIAL);
    dispatch_queue_t queue3 = dispatch_queue_create("t3.0", DISPATCH_QUEUE_CONCURRENT);

    NSLog(@"0 %@", NSThread.currentThread);
    dispatch_async(queue1, ^{
        dispatch_sync(queue3, ^{
            NSLog(@"1 %@", NSThread.currentThread);
            dispatch_sync(queue3, ^{
                NSLog(@"2 %@", NSThread.currentThread);
            });
            NSLog(@"3 %@", NSThread.currentThread);
        });
    });
    NSLog(@"4 %@", NSThread.currentThread);

输出结果，同上，只是队列是 queue1 的线程队列中

    dispatch_queue_t queue1 = dispatch_queue_create("t1.0", DISPATCH_QUEUE_SERIAL);
    NSLog(@"0 %@", NSThread.currentThread);
    dispatch_sync(queue1, ^{
        NSLog(@"1 %@", NSThread.currentThread);
        dispatch_sync(queue1, ^{
            NSLog(@"2 %@", NSThread.currentThread);
        });
        NSLog(@"3 %@", NSThread.currentThread);
    });
    NSLog(@"4 %@", NSThread.currentThread);

输出 0，1， 然后卡死，同步到主线程去

    dispatch_queue_t queue1 = dispatch_queue_create("t1.0", DISPATCH_QUEUE_SERIAL);
    dispatch_queue_t queue3 = dispatch_queue_create("t3.0", DISPATCH_QUEUE_CONCURRENT);

    NSLog(@"0 %@", NSThread.currentThread);
    dispatch_async(queue1, ^{
        dispatch_sync(queue2, ^{
            NSLog(@"1 %@", NSThread.currentThread);
            dispatch_sync(dispatch_get_main_queue(), ^{
                NSLog(@"2 %@", NSThread.currentThread);
            });
            NSLog(@"3 %@", NSThread.currentThread);
        });
    });
    NSLog(@"4 %@", NSThread.currentThread);

输出是否会死锁呢？不会，输出顺序 0，4，1，2，3 , dispatch_sync(dispatch_get_main_queue() 会真正的切换到主队列中去的。

    dispatch_queue_t queue = dispatch_queue_create("并行", DISPATCH_QUEUE_CONCURRENT);
    dispatch_sync(queue, ^{
        dispatch_async(queue, ^{
            NSLog(@"任务2");
        });
        dispatch_async(queue, ^{
            NSLog(@"任务3");
        });
        dispatch_async(queue, ^{
            NSLog(@"任务4");
        });
        dispatch_async(queue, ^{
            NSLog(@"任务5");
        });
        dispatch_barrier_async(queue, ^{
            NSLog(@"任务1000");
        });
        dispatch_async(queue, ^{
            NSLog(@"任务6");
        });
        dispatch_async(queue, ^{
            NSLog(@"任务7");
        });
        //睡眠2秒
//        [NSThread sleepForTimeInterval:2];
        NSLog(@"任务1");
    });

它的输出结果是，1，2，3，4，5 无顺序在 1000之前， 1000,之后 6，7无顺序，