#include <iostream> #include <pthread.h> #include <cstdio> #include "errors.h" using namespace std; typedef struct stage_tag { pthread_mutex_t mutex; //protect data pthread_cond_t avail; //data availiable 当前流水节点处于空闲可用状态, 等待接收数据，进行加工 pthread_cond_t read; //ready for data 当前处于数据准备状态，写入当前节点 int data_ready; //data present 节点数据状态;1 表示准备好 等待发送 ; 0 表示没有数据 long data; //data to process pthread_t thread; //thread for stage struct stage_tag *next; //next satge }stage_t; typedef struct pipe_tag { pthread_mutex_t mutex; //mutex to protect pipe stage_t *head; //fist stage stage_t *tail; //final stage int stages; //number for stages int active; //active data element }pipe_t; int pipe_send(stage_t *stage,long data) { int status=pthread_mutex_lock(&stage->mutex); if(status!=0) return status; while(stage->data_ready)// 如果有数据，则下面必须等待知道收到ready信号 { status=pthread_cond_wait(&stage->read,&stage->mutex);//wait ready if(status!=0) //在read上阻塞，main线程改变了条件变量的值并发信息号，解除阻塞 { pthread_mutex_unlock(&stage->mutex); return status; } } //send new data stage->data=data; stage->data_ready=1; //此时有了data，就可以发送avail信号了 status=pthread_cond_signal(&stage->avail); //signal avail if(status!=0) { pthread_mutex_unlock(&stage->mutex); return status; } status=pthread_mutex_unlock(&stage->mutex); return status; } void *pipe_stage(void *arg) { stage_t *stage=(stage_t*)arg; stage_t *next_stage=stage->next; int status=pthread_mutex_lock(&stage->mutex); if(status!=0) err_abort(status,"Lock pipe stage"); while(1) { //if there's data int the pipe stage,wait for it be consumed //一般一个条件表达式都是在一个互斥锁的保护下被检查。 //当条件表达式未被满足时，线程将仍然阻塞在这个条件变量上。 //当另一个线程改变了条件的值并向条件变量发出信号时， //等待在这个条件变量上的一个线程或所有线程被唤醒， //接着都试图再次占有相应的互斥锁。阻塞在条件变量上的线程被唤醒以后， //直到pthread_cond_wait()函数返回之前条件的值都有可能发生变化。 //所以函数返回以后，在锁定相应的互斥锁之前，必须重新测试条件值。 //最好的测试方法是循环调用pthread_cond_wait函数，并把满足条件的表达式置为循环的终止条件。 //如：pthread_mutex_lock(); // while (condition_is_false) // pthread_cond_wait(); // pthread_mutex_unlock(); while(stage->data_ready!=1) { status=pthread_cond_wait(&stage->avail,&stage->mutex);//wait avail if(status!=0) err_abort(status,"wait for previous stage"); } pipe_send(next_stage,stage->data+1); //pipe_send() stage->data_ready=0; status=pthread_cond_signal(&stage->read); //signal ready if(status!=0) err_abort(status,"wake next stage"); } } int pipe_create(pipe_t *pipe,int stages) { stage_t **link=&pipe->head,*new_stage,*stage; int status=pthread_mutex_init(&pipe->mutex,NULL); if(status) err_abort(status,"Init pipe status"); pipe->stages=stages; pipe->active=0; for(int i=0;i<stages;++i) { new_stage=(stage_t*)malloc(sizeof(stage_t)); if(new_stage==NULL) errno_abort("allocate stage"); status=pthread_mutex_init(&new_stage->mutex,NULL); if(status) err_abort(status,"init stage mutex"); status=pthread_cond_init(&new_stage->avail,NULL); if(status) err_abort(status,"init avail condition"); status=pthread_cond_init(&new_stage->read,NULL); if(status) err_abort(status,"init read condition"); new_stage->data_ready=0; *link=new_stage; link=&new_stage->next; } *link=(stage_t*)NULL; pipe->tail=new_stage; for(stage=pipe->head;stage->next!=NULL;stage=stage->next) { status=pthread_create(&stage->thread,NULL,pipe_stage,(void*)stage); if(status) err_abort(status,"Create pipe stage"); } return 0; } int pipe_start(pipe_t *pipe,long vaule) { int status=pthread_mutex_lock(&pipe->mutex); if(status) err_abort(status,"lock pipe mutex"); pipe->active++;//记录活动的次数 status=pthread_mutex_unlock(&pipe->mutex); if(status) err_abort(status,"unlock pipe mutex"); pipe_send(pipe->head,vaule); //pipe_send() return 0; } int pipe_result(pipe_t *pipe,long *result) { stage_t *tail=pipe->tail; int status=pthread_mutex_lock(&pipe->mutex); if(status) err_abort(status,"lock pipe mutex"); int empty=0; if(pipe->active<=0) empty=1; else pipe->active--; status = pthread_mutex_unlock(&pipe->mutex); if(status!=0) err_abort( status, "unlock pipe mutex" ); if(empty) return 0; pthread_mutex_lock(&tail->mutex); while(!tail->data_ready) pthread_cond_wait(&tail->avail,&tail->mutex); *result=tail->data; tail->data_ready=0; pthread_cond_signal(&tail->read); pthread_mutex_unlock(&tail->mutex); return 1; } int main() { pipe_t my_pipe; pipe_create(&my_pipe,3); //pipe_create() cout<<"Enter ingter values ,or \"=\" for next result"<<endl; while(1) { cout<<"Data >"; char line[128]; if(fgets(line,sizeof(line),stdin)==NULL) exit(0); if(strlen(line)<=1) continue; long result,value; if(strlen(line)<=2&&line[0]=='=') { if(pipe_result(&my_pipe,&result)) //pipe_result() cout<<"Result is:"<<result<<endl; else cout<<"Pipe is empty"<<endl; } else { if(sscanf(line,"%ld",&value)<1) fprintf(stderr,"Enter an intger value\n"); else pipe_start(&my_pipe,value); //pipe_start() } } return 0; }
流水线工作方式的图示
流水线中线程，线程队列对每个输入进行处理，一个阶段的线程处理完成后，将会把数据交给下一阶段的线程。最后一阶段的线程产生输出结果。
如果前一阶段处理的速度高于下一阶段的线程，可以使用缓冲区作为使其同步工作的手段。
