GCD 串行队列,并发队列和主队列的整理

1.串行队列：Dispatch Queues或者 Serial Queues

同步

dispatch_queue_t queue = dispatch_queue_create("zenny_chen_firstQueue", nil);
    
    dispatch_sync(queue, ^(void) {
        
        
        sleep(2-i/50);
        NSLog(@"The sum is: %d", i);
       
        //flag = YES;
    });
dispatch_release(queue);

异步

dispatch_queue_t queue = dispatch_queue_create("zenny_chen_firstQueue", nil);
    dispatch_async(queue, ^(void) {
        sleep(2-i/50);
        NSLog(@"The sum is: %d", i);
        
        // signal the semaphore
        //dispatch_semaphore_signal(sem);
    });
dispatch_release(queue);

当主线程在等待信号量时，操作系统会直接将它挂起，因此不会去死做标志轮询，而能够及时将CPU资源给其它可被调度的线程使用。

__block dispatch_semaphore_t sem = dispatch_semaphore_create(0);
    
    dispatch_queue_t queue = dispatch_queue_create("zenny_chen_firstQueue", nil);
    dispatch_async(queue, ^(void) {
        sleep(2-i/50);
        NSLog(@"The sum is: %d", i);
        
        // signal the semaphore
        dispatch_semaphore_signal(sem);
    });
    
    // wait for the semaphore
   dispatch_semaphore_wait(sem, DISPATCH_TIME_FOREVER);
    
   dispatch_release(queue);

[b]2.并发队列[/b] ：Concurrent Queues

dispatch_queue_t aQueue = dispatch_get_global_queue(DISPATCH_QUEUE_PRIORITY_DEFAULT, 0);  
    dispatch_queue_t aHQueue = dispatch_get_global_queue(DISPATCH_QUEUE_PRIORITY_HIGH, 0);
  
    dispatch_queue_t aLQueue = dispatch_get_global_queue(DISPATCH_QUEUE_PRIORITY_LOW, 0);

3.主队列：main dispatch Queue

dispatch_queue_t mainQueue = dispatch_get_main_queue();

针对以上3种队列：concurrent queues和main queue都是由系统生成而且 dispatch_suspend, dispatch_resume, dispatch_set_context这些函数对他们无效。

但是我们的应用不是简单的同步也异步的运行，应用经常是混合的。
比如我们要task1 task2 task3都运行完成后才能异步运行task4 task5 task6我们该怎么做呢？这里我们可以引入group的概念。

-(void)sixthMethod{  
    //获取concurrent queue  
    dispatch_queue_t aQueue = dispatch_get_global_queue(DISPATCH_QUEUE_PRIORITY_DEFAULT, 0);  
    //创建1个queue group  
    dispatch_group_t queueGroup = dispatch_group_create();  
    //任务1  
    dispatch_group_async(queueGroup, aQueue, ^{  
        NSLog(@"task 1.");  
    });  
    //任务2  
    dispatch_group_async(queueGroup, aQueue, ^{  
        NSLog(@"task 2.");  
    });  
    //任务3  
    dispatch_group_async(queueGroup, aQueue, ^{  
        NSLog(@"task 3.");  
    });  
    NSLog(@"wait task 1,2,3.");  
    //等待组内任务全部完成  
    dispatch_group_wait(queueGroup, DISPATCH_TIME_FOREVER);  
    NSLog(@"task 1,2,3 finished.");  
    //释放组  
    dispatch_release(queueGroup);  
    //重新创建组  
    queueGroup = dispatch_group_create();  
    //任务4  
    dispatch_group_async(queueGroup, aQueue, ^{  
        NSLog(@"task 4.");  
    });  
    //任务5  
    dispatch_group_async(queueGroup, aQueue, ^{  
        NSLog(@"task 5.");  
    });  
    //任务6  
    dispatch_group_async(queueGroup, aQueue, ^{  
        NSLog(@"task 6.");  
    });  
    NSLog(@"wait task 4,5,6.");  
    //等待组内任务全部完成  
    dispatch_group_wait(queueGroup, DISPATCH_TIME_FOREVER);  
    NSLog(@"task 4,5,6 finished.");  
    //释放组  
    dispatch_release(queueGroup);  
}

代码运行结果：


2013-05-13 13:55:33.783 GDC[2466:1303] task 1.

2013-05-13 13:55:33.783 GDC[2466:3a07] task 3.

2013-05-13 13:55:33.783 GDC[2466:c07] wait task 1,2,3.

2013-05-13 13:55:33.783 GDC[2466:1903] task 2.

2013-05-13 13:55:33.787 GDC[2466:c07] task 1,2,3 finished.

2013-05-13 13:55:33.788 GDC[2466:c07] wait task 4,5,6.

2013-05-13 13:55:33.788 GDC[2466:1303] task 4.

2013-05-13 13:55:33.788 GDC[2466:1903] task 5.

2013-05-13 13:55:33.788 GDC[2466:3a07] task 6.

2013-05-13 13:55:33.790 GDC[2466:c07] task 4,5,6 finished.