服务器之家:专注于服务器技术及软件下载分享
分类导航

PHP教程|ASP.NET教程|Java教程|ASP教程|编程技术|正则表达式|C/C++|IOS|C#|Swift|Android|VB|R语言|JavaScript|易语言|vb.net|

服务器之家 - 编程语言 - C/C++ - 浅析C++编程当中的线程

浅析C++编程当中的线程

2021-03-04 10:23低调小一 C/C++

这篇文章主要介绍了浅析C++编程当中的线程,线程在每一种编程语言中都是重中之重,需要的朋友可以参考下

线程的概念

C++中的线程的Text Segment和Data Segment都是共享的,如果定义一个函数,在各线程中都可以调用,如果定义一个全局变量,在各线程中都可以访问到。除此之外,各线程还共享以下进程资源和环境:

  •     文件描述符
  •     每种信号的处理方式
  •     当前工作目录
  •     用户id和组id

但是,有些资源是每个线程各有一份的:

  •     线程id
  •     上下文,包括各种寄存器的值、程序计数器和栈指针
  •     栈空间
  •     errno变量
  •     信号屏蔽字
  •     调度优先级

我们将要学习的线程库函数是由POSIX标准定义的,称为POSIX thread或pthread。
线程控制
创建线程

创建线程的函数原型如下:

?
1
2
#include <pthread.h>
int pthread_create(pthread_t *thread, const pthread_attr_t *attr, void *(*start_routine)(void *), void *arg);

返回值:成功返回0,失败返回错误号。

在一个线程中调用pthread_create()创建新的线程后,当前线程从pthread_create()返回继续往下执行,而新的线程所执行的代码由我们传给pthread_create的函数指针start_routine决定。start_routine函数接收一个参数,是通过pthread_create的arg参数传递给它的,该参数类型为void*,这个指针按什么类型解释由调用者自己定义。start_routine的返回值类型也是void *,这个指针的含义同样由调用者自己定义。start_routine返回时,这个线程就退出了,其它线程可以调用pthread_join得到start_routine的返回值。

pthread_create成功返回后,新创建的线程的id被填写到thread参数所指向的内存单元。我们知道进程id的类型是pid_t,每个进程的id在整个系统中是唯一的,调用getpid可以得到当前进程的id,是一个正整数值。线程id的类型是thread_t,它只在当前进程中保证是唯一的,在不同的系统中thread_t这个类型有不同的实现,它可能是一个整数值,也可能是一个结构体,也可能是一个地址,所以不能简单的当成整数用printf打印,调用pthread_self可以获取当前线程的id。

我们先来写一个简单的例子:

?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
#include <stdio.h>
#include <string.h>
#include <stdlib.h>
#include <pthread.h>
#include <unistd.h>
 
pthread_t ntid;
 
void printids(const void *t)
{
    char *s = (char *)t;
  pid_t   pid;
  pthread_t tid;
 
  pid = getpid();
  tid = pthread_self();
  printf("%s pid %u tid %u (0x%x)\n", s, (unsigned int)pid,
      (unsigned int)tid, (unsigned int)tid);
}
 
void *thr_fn(void *arg)
{
  printids(arg);
  return NULL;
}
 
int main(void)
{
  int err;
 
  err = pthread_create(&ntid, NULL, thr_fn, (void *)"Child Process:");
  if (err != 0) {
    fprintf(stderr, "can't create thread: %s\n", strerror(err));
    exit(1);
  }
  printids("main thread:");
  sleep(1);
 
  return 0;
}


编译执行结果如下:

?
1
2
3
4
g++ thread.cpp -o thread -lpthread
./thread
main thread: pid 21046 tid 3612727104 (0xd755d740)
Child Process: pid 21046 tid 3604444928 (0xd6d77700)

从结果可以知道,thread_t类型是一个地址值,属于同一进程的多个线程调用getpid可以得到相同的进程号,而调用pthread_self得到的线程号各不相同。

如果任意一个线程调用了exit或_exit,则整个进程的所有线程都终止,由于从main函数return也相当于调用exit,为了防止新创建的线程还没有得到执行就终止,我们在main函数return之前延时1秒,这只是一种权宜之计,即使主线程等待1秒,内核也不一定会调度新创建的线程执行,接下来,我们学习一下比较好的解决方法。
终止线程

如果需要只终止某个线程而不是终止整个进程,可以有三种方法:

  1.     从线程函数return。这种方法对主线程不适应,从main函数return相当于调用exit。
  2.     一个线程可以调用pthread_cancel终止同一个进程中的另一个线程。
  3.     线程可以调用pthread_exit终止自己。

这里主要介绍pthread_exit和pthread_join的用法。

?
1
2
3
#include <pthread.h>
 
void pthread_exit(void *value_ptr);

value_ptr是void*类型,和线程函数返回值的用法一样,其它线程可以调用pthread_join获取这个指针。
需要注意,pthread_exit或者return返回的指针所指向的内存单元必须是全局的或者是用malloc分配的,不能在线程函数的栈上分配,因为当其它线程得到这个返回指针时线程函数已经退出了。

?
1
2
3
#include <pthread.h>
 
int pthread_join(pthread_t thread, void **value_ptr);

返回值:成功返回0,失败返回错误号。

调用该函数的线程将挂起等待,直到id为thread的线程终止。thread线程以不同的方法终止,通过pthread_join得到的终止状态是不同的,总结如下:

  •     如果thread线程通过return返回,value_ptr所指向的单元里存放的是thread线程函数的返回值。
  •     如果thread线程被别的线程调用pthread_cancel异常终止掉,value_ptr所指向的单元存放的是常数PTHREAD_CANCELED。
  •     如果thread线程是自己调用pthread_exit终止的,value_ptr所指向的单元存放的是传给pthread_exit的参数。

如果对thread线程的终止状态不感兴趣,可以传NULL给value_ptr参数。参考代码如下:

?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
#include <stdio.h>
#include <stdlib.h>
#include <pthread.h>
#include <unistd.h>
 
void* thread_function_1(void *arg)
{
  printf("thread 1 running\n");
  return (void *)1;
}
 
void* thread_function_2(void *arg)
{
  printf("thread 2 exiting\n");
  pthread_exit((void *) 2);
}
 
void* thread_function_3(void* arg)
{
  while (1) {
    printf("thread 3 writeing\n");
    sleep(1);
  }
}
 
 
int main(void)
{
  pthread_t tid;
  void *tret;
 
  pthread_create(&tid, NULL, thread_function_1, NULL);
  pthread_join(tid, &tret);
  printf("thread 1 exit code %d\n", *((int*) (&tret)));
 
  pthread_create(&tid, NULL, thread_function_2, NULL);
  pthread_join(tid, &tret);
  printf("thread 2 exit code %d\n", *((int*) (&tret)));
 
  pthread_create(&tid, NULL, thread_function_3, NULL);
  sleep(3);
  pthread_cancel(tid);
  pthread_join(tid, &tret);
  printf("thread 3 exit code %d\n", *((int*) (&tret)));
 
  return 0;
}

 

运行结果是:

?
1
2
3
4
5
6
7
8
thread 1 running
thread 1 exit code 1
thread 2 exiting
thread 2 exit code 2
thread 3 writeing
thread 3 writeing
thread 3 writeing
thread 3 exit code -1


可见,Linux的pthread库中常数PTHREAD_CANCELED的值是-1.可以在头文件pthread.h中找到它的定义:

?
1
#define PTHREAD_CANCELED ((void *) -1)


线程间同步

多个线程同时访问共享数据时可能会冲突,例如两个线程都要把某个全局变量增加1,这个操作在某平台上需要三条指令才能完成:

  •     从内存读变量值到寄存器。
  •     寄存器值加1.
  •     将寄存器的值写回到内存。

这个时候很容易出现两个进程同时操作寄存器变量值的情况,导致最终结果不正确。

解决的办法是引入互斥锁(Mutex, Mutual Exclusive Lock),获得锁的线程可以完成“读-修改-写”的操作,然后释放锁给其它线程,没有获得锁的线程只能等待而不能访问共享数据,这样,“读-修改-写”的三步操作组成一个原子操作,要不都执行,要不都不执行,不会执行到中间被打断,也不会在其它处理器上并行做这个操作。

Mutex用pthread_mutex_t类型的变量表示,可以这样初始化和销毁:

?
1
2
3
4
5
#include <pthread.h>
 
int pthread_mutex_destory(pthread_mutex_t *mutex);
int pthread_mutex_int(pthread_mutex_t *mutex, const pthread_mutexattr_t *attr);
pthread_mutex_t mutex = PTHEAD_MUTEX_INITIALIZER;


返回值:成功返回0,失败返回错误号。

用pthread_mutex_init函数初始化的Mutex可以用pthread_mutex_destroy销毁。如果Mutex变量是静态分配的(全局变量或static变量),也可以用宏定义PTHREAD_MUTEX_INITIALIZER来初始化,相当于用pthread_mutex_init初始化并且attr参数为NULL。Mutex的加锁和解锁操作可以用下列函数:

?
1
2
3
4
5
#include <pthread.h>
 
int pthread_mutex_lock(pthread_mutex_t *mutex);
int pthread_mutex_trylock(pthread_mutex_t *mutex);
int pthread_mutex_unlock(pthread_mutex_t *mutex);


返回值:成功返回0,失败返回错误号。

一个线程可以调用pthread_mutex_lock获得Mutex,如果这时另一个线程已经调用pthread_mutex_lock获得了该Mutex,则当前线程需要挂起等待,直到另一个线程调用pthread_mutex_unlock释放Mutex,当前线程被唤醒,才能获得该Mutex并继续执行。

我们用Mutex解决上面说的两个线程同时对全局变量+1可能导致紊乱的问题:

?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
#include <pthread.h>
#include <stdio.h>
#include <stdlib.h>
 
#define NLOOP 5000
 
int counter;
pthread_mutex_t counter_mutex = PTHREAD_MUTEX_INITIALIZER;
 
void *do_add_process(void *vptr)
{
  int i, val;
 
  for (i = 0; i < NLOOP; i ++) {
    pthread_mutex_lock(&counter_mutex);
    val = counter;
    printf("%x:%d\n", (unsigned int)pthread_self(), val + 1);
    counter = val + 1;
    pthread_mutex_unlock(&counter_mutex);
  }
 
  return NULL;
}
 
int main()
{
  pthread_t tida, tidb;
 
  pthread_create(&tida, NULL, do_add_process, NULL);
  pthread_create(&tidb, NULL, do_add_process, NULL);
 
  pthread_join(tida, NULL);
  pthread_join(tidb, NULL);
 
  return 0;
}

这样,每次运行都能显示到10000。如果去掉锁机制,可能就会有问题。这个机制类似于Java的synchronized块机制。
Condition Variable

线程间的同步还有这样一种情况:线程A需要等某个条件成立才能继续往下执行,现在这个条件不成立,线程A就阻塞等待,而线程B在执行过程中使这个条件成立了,就唤醒线程A继续执行。在pthread库中通过条件变量(Conditiion Variable)来阻塞等待一个条件,或者唤醒等待这个条件的线程。Condition Variable用pthread_cond_t类型的变量表示,可以这样初始化和销毁:

?
1
2
3
4
5
#include <pthread.h>
 
int pthread_cond_destory(pthread_cond_t *cond);
int pthread_cond_init(pthead_cond_t *cond, const pthread_condattr_t *attr);
pthread_cond_t cond = PTHREAD_COND_INITIALIZER;

返回值:成功返回0,失败返回错误号。

和Mutex的初始化和销毁类似,pthread_cond_init函数初始化一个Condition Variable,attr参数为NULL则表示缺省属性,pthread_cond_destroy函数销毁一个Condition Variable。如果Condition Variable是静态分配的,也可以用宏定义PTHEAD_COND_INITIALIZER初始化,相当于用pthread_cond_init函数初始化并且attr参数为NULL。Condition Variable的操作可以用下列函数:

?
1
2
3
4
5
6
#include <pthread.h>
 
int pthread_cond_timedwait(pthread_cond_t *cond, pthread_mutex_t *mutex, const struct timespec *abstime);
int pthread_cond_wait(pthread_cond_t *cond, pthread_mutex_t *mutex);
int pthread_cond_broadcast(pthread_cond_t *cond);
int pthread_cond_signal(pthread_cond_t *cond);


可见,一个Condition Variable总是和一个Mutex搭配使用的。一个线程可以调用pthread_cond_wait在一个Condition Variable上阻塞等待,这个函数做以下三步操作:

  1.     释放Mutex。
  2.     阻塞等待。
  3.     当被唤醒时,重新获得Mutex并返回。

pthread_cond_timedwait函数还有一个额外的参数可以设定等待超时,如果到达了abstime所指定的时刻仍然没有别的线程来唤醒当前线程,就返回ETIMEDOUT。一个线程可以调用pthread_cond_signal唤醒在某个Condition Variable上等待的另一个线程,也可以调用pthread_cond_broadcast唤醒在这个Condition Variable上等待的所有线程。

下面的程序演示了一个生产者-消费者的例子,生产者生产一个结构体串在链表的表头上,消费者从表头取走结构体。

?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
#include <stdio.h>
#include <stdlib.h>
#include <pthread.h>
#include <unistd.h>
 
struct msg {
  struct msg *next;
  int num;
};
 
struct msg *head;
pthread_cond_t has_product = PTHREAD_COND_INITIALIZER;
pthread_mutex_t lock = PTHREAD_MUTEX_INITIALIZER;
 
void* consumer(void *p)
{
  struct msg *mp;
 
  for(;;) {
    pthread_mutex_lock(&lock);
    while (head == NULL) {
      pthread_cond_wait(&has_product, &lock);
    }
    mp = head;
    head = mp->next;
    pthread_mutex_unlock(&lock);
    printf("Consume %d\n", mp->num);
    free(mp);
    sleep(rand() % 5);
  }
}
 
void* producer(void *p)
{
  struct msg *mp;
 
  for(;;) {
    mp = (struct msg *)malloc(sizeof(*mp));
    pthread_mutex_lock(&lock);
    mp->next = head;
    mp->num = rand() % 1000;
    head = mp;
    printf("Product %d\n", mp->num);
    pthread_mutex_unlock(&lock);
    pthread_cond_signal(&has_product);
    sleep(rand() % 5);
  }
}
 
int main()
{
  pthread_t pid, cid;
  srand(time(NULL));
 
  pthread_create(&pid, NULL, producer, NULL);
  pthread_create(&cid, NULL, consumer, NULL);
 
  pthread_join(pid, NULL);
  pthread_join(cid, NULL);
 
  return 0;
}

延伸 · 阅读

精彩推荐