我无法理解条件变量,例如pthread_mutex_lock
/ unlock
和pthread_cond_wait
/ signal
我正在尝试创建九个threads
,并让它们同时运行以找出哪个效率最高。
int threadNumber = 0;
pthread_mutex_t mutex = PTHREAD_MUTEX_INITIALIZER;
pthread_cond_t cond = PTHREAD_COND_INITIALIZER;
#define NUM_THREADS 9
//used to store the information of each thread
typedef struct{
pthread_t threadID;
int policy;
struct sched_param param;
long startTime;
long taskStartTime;
long endTime1;
long endTime2;
long endTime3;
long runTime;
char startDate[30];
char endDate[30];
}ThreadInfo;
ThreadInfo myThreadInfo[NUM_THREADS];
//main function
int main(void){
printf("running...\n");
pthread_mutex_lock(&mutex); //lock the mutex//////
pthread_cond_wait(&cond, &mutex); //start waiting//////
int fifoPri = 60;
int rrPri = 30;
//create the 9 threads and assign their scheduling policies
for(int i=0; i<NUM_THREADS; i++){
if(i%3 == SCHED_OTHER){
myThreadInfo[i].policy = SCHED_OTHER;
myThreadInfo[i].param.sched_priority = 0;
}
else if (i%3 == SCHED_RR){
myThreadInfo[i].policy = SCHED_RR;
myThreadInfo[i].param.sched_priority = rrPri++;
}
else{
myThreadInfo[i].policy = SCHED_FIFO;
myThreadInfo[i].param.sched_priority = fifoPri++;
}
pthread_create( &myThreadInfo[i].threadID, NULL, ThreadRunner, &myThreadInfo[i]);
}
printf("\n\n");
pthread_mutex_unlock(&mutex) //unlock the mutex/////////
pthread_cond_signal(&cond); //signal the threads to start////////
//join each thread
for(int g = 0; g < NUM_THREADS; g++){
pthread_join(myThreadInfo[g].threadID, NULL);
}
//print out the stats for each thread and perform an analysis of the data
DisplayThreadSchdStats();
return 0;
}
...
所以当main函数启动时,我lock
使用互斥锁来确保线程在我用pthread_lock(&amp; mutex)和pthread_cond_wait(&amp; cond,&amp; mutex)告诉它们之前没有启动
然后我用各种调度策略创建所有九个线程。在完成之后,我尝试使用pthread_mutex_unlock(&amp; mutex)和pthread_cond_signal(&amp; cond)
告诉线程全部同时启动但是当我运行它时,它永远不会解锁线程。主函数的“running ...”打印语句消失,但线程永远不会启动。 (threadrunner有一个功能,他们都打印出大量不同的数字,所以我可以看看他们是否启动)。我对pthread mutex和pthread cond做错了什么?
答案 0 :(得分:1)
来自pthread_cond_wait()
手册页:
pthread_cond_wait()函数以原子方式阻塞当前线程 等待 cond 指定的条件变量,然后释放 由互斥锁指定的互斥锁。
...所以,当你的main()
函数到达此行时:
pthread_cond_wait(&cond, &mutex); //start waiting//////
......正如你的评论建议的那样,它将开始等待,直到有人发出条件变量的信号。
但是还没有其他线程产生,所以没有人在那里发出条件变量的信号,因此你的主线程在该调用中无限期地阻塞。
答案 1 :(得分:0)
我认为你想要做的就是这样。我用
pthread_cond_broadcast
一次唤醒所有主题:
#include <stdio.h>
#include <string.h>
#include <errno.h>
#include <stdlib.h>
#include <unistd.h>
#include <pthread.h>
#include <time.h>
#include <sys/time.h>
pthread_mutex_t mutex = PTHREAD_MUTEX_INITIALIZER;
pthread_cond_t cond = PTHREAD_COND_INITIALIZER;
struct data_t {
int id;
int sleep;
};
void* thread(void *arg)
{
struct data_t *data = arg;
pthread_mutex_lock(&mutex);
printf("Thread %d: waiting for release\n", data->id);
pthread_cond_wait(&cond, &mutex);
pthread_mutex_unlock(&mutex); // unlocking for all other threads
struct timeval tv;
gettimeofday(&tv, NULL);
printf("Thread %d: doing some work for %d secs, started: %ld...\n", data->id, data->sleep, tv.tv_sec);
sleep(data->sleep);
gettimeofday(&tv, NULL);
printf("Thread %d: Bye, end at %ld\n", data->id, tv.tv_sec);
}
int main(void)
{
struct data_t data[9];
pthread_t ths[9];
srand(time(NULL));
for(int i = 0; i < 9; ++i)
{
data[i].id = i + 1;
data[i].sleep = 1 + rand() % 6;
pthread_create(ths + i, NULL, thread, data + i);
}
// give time for all threads to lock
sleep(1);
printf("Master: Now releasing the condition\n");
pthread_cond_broadcast(&cond);
for(int i = 0; i < 9; ++i)
pthread_join(*(ths + i), NULL);
return 0;
}
输出
Thread 2: waiting for release
Thread 6: waiting for release
Thread 4: waiting for release
Thread 1: waiting for release
Thread 3: waiting for release
Thread 8: waiting for release
Thread 9: waiting for release
Thread 7: waiting for release
Thread 5: waiting for release
Master: Now releasing the condition
Thread 5: doing some work for 6 secs, started: 1518463908...
Thread 2: doing some work for 4 secs, started: 1518463908...
Thread 8: doing some work for 1 secs, started: 1518463908...
Thread 4: doing some work for 4 secs, started: 1518463908...
Thread 6: doing some work for 1 secs, started: 1518463908...
Thread 9: doing some work for 5 secs, started: 1518463908...
Thread 3: doing some work for 2 secs, started: 1518463908...
Thread 1: doing some work for 3 secs, started: 1518463908...
Thread 7: doing some work for 2 secs, started: 1518463908...
Thread 8: Bye, end at 1518463909
Thread 6: Bye, end at 1518463909
Thread 3: Bye, end at 1518463910
Thread 7: Bye, end at 1518463910
Thread 1: Bye, end at 1518463911
Thread 2: Bye, end at 1518463912
Thread 4: Bye, end at 1518463912
Thread 9: Bye, end at 1518463913
Thread 5: Bye, end at 1518463914
答案 2 :(得分:0)
对于临时基准测试而言,接受的答案“足够好”。但是,它使用了几个有问题的做法:没有谓词的条件变量,以及 sleep()作为ersatz同步。
使用pthreads barrier对象会更好,也很容易。
#define NUM_THREADS ...
static pthread_barrier_t bar;
static void*
thrfunc(void *arg) {
// set scheduling policy
pthread_barrier_wait(&bar); // wait till all peers have also done so
...
}
int
main(void) {
pthread_barrier_init(&bar, NULL, NUM_THREADS); // FIXME: error check
for (int i = 0; i < NUM_THREADS; i++) {
// spawn threads with various scheduling policy instructions
}
// join and tabulate results
pthread_barrier_destroy(&bar);
...
}