gpt4 book ai didi

c++ - 我定义的 Mutex 类中的逻辑错误以及我在生产者消费者程序中使用它的方式 - pthreads

转载 作者:行者123 更新时间:2023-11-30 02:34:29 25 4
gpt4 key购买 nike

我添加了一个 Mutex 类以遵守 RAII。我不确定我使用它的方式是否正确。

队列被生产者锁定后,程序意外结束。

MutexClass.h

#ifndef MUTEXCLASS
#define MUTEXCLASS

#include <pthread.h>

class MutexClass
{
private:
pthread_mutex_t & _mutexVariable;
public:
MutexClass (pthread_mutex_t &);
~MutexClass ();
};

#endif // MUTEXCLASS

MutexClass.cpp

#include "mutexClass.h"
#include <stdexcept>

MutexClass::MutexClass (pthread_mutex_t & arg) : _mutexVariable (arg)
{
_mutexVariable = PTHREAD_MUTEX_INITIALIZER;
int returnValue = pthread_mutex_lock (&_mutexVariable);
if (returnValue > 0)
{
throw std::logic_error ("Mutex couldn't be locked!");
}
}

MutexClass::~MutexClass()
{
pthread_mutex_unlock (&_mutexVariable);
}

这是 ma​​in.cpp,我在其中使用上面定义的互斥类的对象。

Qt 的类在这里是为了同名,因为我是 Qt Creator。请忽略它们。

#include "mainwindow.h"
#include <QApplication>
#include <stdexcept>

#include <stdio.h>
#include <unistd.h>
#include <sys/syscall.h>
#include <iostream>
#include <QDebug>

#include "mutexClass.h"

pthread_mutex_t mutexVariable;
pthread_cond_t conditionVariable;

int numberOfActiveProducers;
int numberOfActiveConsumers;

QList <int> sharedQueueA;
QList <int> sharedQueueB;

/*
* Shared queues are supposed to be shared among four threads. Two producer, and two consumer threads.
* Producer threads will put the 1's in it, and Consumer threads will remove the 1's.
* Assumption: `sharedQueue` can contain only 10 elements at a time.
*/

int sizeOfSharedQueue;

void checkForSpaceAndPush (QList <int> & argList, int listId, pthread_t argTId)
{
std::cerr << "\nQueue " << listId << ", First check by Producer: " << argTId;
if (argList.length () < sizeOfSharedQueue + 1)
{
{
MutexClass mutex1 (mutexVariable);

std::cerr << "\n\nQueue " << listId << ", Locked by Producer: " << argTId;

if (argList.length () < sizeOfSharedQueue + 1)
{
argList.push_back (1); std::cerr << "\nPushed by Producer " << argTId << ": " << "Length of queue " << listId << " is: " << argList.length ();
}
else
{
std::cerr << "\nProducer " << argTId << ". Queue " << listId << " is full. Length of queue is: " << argList.length ();
pthread_cond_wait (&conditionVariable, &mutexVariable);
}
}
std::cerr << "\n\nQueue " << listId << ", UnLocked by Producer: " << argTId;
}
}

void checkForSpaceAndPop (QList <int> & argList, int listId, pthread_t argTId)
{
std::cerr << "\nQueue " << listId << ", First check by Consumer: " << argTId;
if (argList.length () > 0)
{
{
MutexClass mutex1 (mutexVariable);
std::cerr << "\n\nQueue " << listId << ", Locked by Consumer: " << argTId;

if (argList.length () > 0)
{
argList.pop_front (); std::cerr << "\nRemoved by Consumer: " << argTId << ", Length of queue " << listId << " is: " << argList.length ();
}
else
{
pthread_cond_signal (&conditionVariable); std::cerr << "\nSignal issued by Consumer: " << argTId << ", Length of queue " << listId << " is: " << argList.length ();
}
}

std::cerr << "\n\nQueue " << listId << ", UnLocked by Consumer: " << argTId;
}
}

// This function is run by the `Producer` threads.
void *producerThreadFunction (void *arg)
{
Q_UNUSED (arg);

while (1)
{
pthread_t tId = pthread_self(); std::cerr << "\nProducers: " << tId; std::cerr.flush();
checkForSpaceAndPush (sharedQueueA, 1, tId);
checkForSpaceAndPush (sharedQueueB, 2, tId);
}

return NULL;
}

// This function is run by the `Consumer` threads.
void *consumerThreadFunction (void *arg)
{
Q_UNUSED (arg);

while (1)
{
pthread_t tId = pthread_self (); std::cerr << "\nConsumer: " << tId; std::cerr.flush();
checkForSpaceAndPop (sharedQueueA, 1, tId);
checkForSpaceAndPop (sharedQueueB, 2, tId);
}
return NULL;
}

int main (int argc, char *argv[])
{
numberOfActiveProducers = 2;
numberOfActiveConsumers = 2;
sizeOfSharedQueue = 10;

// Producer threads creation
pthread_t producerA;
pthread_t producerB;

if (pthread_create (&producerA, NULL, producerThreadFunction, NULL)) {
fprintf (stderr, "Error creating thread Producer A\n");
return 1;
}

if (pthread_create (&producerB, NULL, producerThreadFunction, NULL)) {
fprintf (stderr, "Error creating thread Producer B\n");
return 1;
}

// Consumer threads creation
pthread_t consumerA;
pthread_t consumerB;

if (pthread_create (&consumerA, NULL, consumerThreadFunction, NULL)) {
fprintf (stderr, "Error creating thread Consumer A\n");
return 1;
}

if (pthread_create (&consumerB, NULL, consumerThreadFunction, NULL)) {
fprintf (stderr, "Error creating thread Consumer B\n");
return 1;
}

// Joining every thread
if (pthread_join (producerA, NULL)) {
fprintf (stderr, "Error joining thread Producer A\n");
return 2;
}

if (pthread_join (producerB, NULL)) {
fprintf (stderr, "Error joining thread Producer B\n");
return 2;
}

if (pthread_join (consumerB, NULL)) {
fprintf (stderr, "Error joining thread Consumer B\n");
return 2;
}

if (pthread_join (consumerA, NULL)) {
fprintf (stderr, "Error joining thread Consumer A\n");
return 2;
}

QApplication a (argc, argv);
MainWindow w;
w.show ();

return a.exec ();
}

输出在这里:

...  
...

Queue 140388157085440
Removed by Consumer: 1403881570854401403881654781442
Queue , Locked by Consumer: 1140388148692736
Removed by Consumer: , Length of queue 1 is: , First check by Producer: 140388148692736, Length of queue 2 is: 1403881654781449

Queue

Queue 2, UnLocked by Consumer: 140388148692736
Consumer: 9

Queue 1, UnLocked by Consumer: 140388157085440
Queue 2, First check by Consumer: 1403881570854401140388148692736
Queue 1, First check by Consumer: 140388148692736

Queue , Locked by Producer:

Queue 2, Locked by Consumer: 140388157085440
Removed by Consumer: 1403881654781441, Locked by Consumer: 140388148692736
Pushed by Producer 140388165478144: Length of queue 1 is: 10

Queue 1, UnLocked by Producer: 140388165478144
Queue 2, First check by Producer: 140388165478144

Queue 2, Locked by Producer: 140388165478144The program has unexpectedly finished.

另外,请注意输出的以下部分:

Queue 2, Locked by Consumer: 140388157085440 Removed by Consumer: 1403881654781441, Locked by Consumer: 140388148692736

我只创建了 2 个消费者,但这里显示的 pids 是 3。为什么会这样?

最佳答案

目前,每次创建新构造一个新的 MutexClass 时,您都会将 mutexVariable 初始化为 PTHREAD_MUTEX_INITIALIZER。考虑如果线程 A 持有 mutexVariable 并且线程 B 想要获取互斥量会发生什么:

thread A tries to lock mutexVariable and succeeds
mutexVariable = PTHRAED_MUTEX_INITIALIZER
pthread_mutex_lock(mutexVariable)

thread B tries to lock mutexVariable and succeeds
// this assignment overwrites the locked state thread A has stored
mutexVariable = PTHRAED_MUTEX_INITIALIZER
// mutex is default-initialized (not locked) - so lock it
pthread_mutex_lock(mutexVariable)
// both threads now believe they have the mutex
// and all syncronization is lost

thread B unlocks mutexVariable
// succeeds

thread A unlocks mutexVariable
// uh.. it is not even locked any more?!

你应该只在 main.cpp:13 中初始化一次 mutexVariable:

pthread_mutex_t mutexVariable = PTHRAED_MUTEX_INITIALIZER;

并从 MutexClass 中删除初始化。

不确定这是否会解决您的所有问题,但这是我首先要做的事情。

关于c++ - 我定义的 Mutex 类中的逻辑错误以及我在生产者消费者程序中使用它的方式 - pthreads,我们在Stack Overflow上找到一个类似的问题: https://stackoverflow.com/questions/34544833/

25 4 0
Copyright 2021 - 2024 cfsdn All Rights Reserved 蜀ICP备2022000587号
广告合作:1813099741@qq.com 6ren.com