C / C++ / MFC

Critical Data = 1 // Maxwell: Thread-2 made it before the 3 sec timeout of waiting for Thread-1's.
Success

volatile bool Sentinel = true;
int CriticalData = 0;

unsigned ThreadFunc1( void* pArguments ) {
   while (Sentinel) { // Maxwell: This thread keeps watching this variable all the time.
      Sleep(0);
   }
   cout << "Critical Data = " << CriticalData << endl;
   return 0; // Maxwell: Note that this thread never ends unless someone change value of Sentinel.
} 

unsigned  ThreadFunc2( void* pArguments ) {
   Sleep(2000); // Maxwell: Note that this thread ends in "2" seconds.
   CriticalData++; // Maxwell: Make the value from 0 to 1. 
                   //  Then this thread ends Thread-1's loop in next line, 
                   //  to guarantee that Thread-1 will see this value changed while the loop ends.
   Sentinel = false; // Maxwell: This line is trying to end the loop of Thread-1.
   return 0;
}

int main() {
   HANDLE hThread1, hThread2; 
   DWORD retCode;

   hThread1 = CreateThread(NULL, 0, (LPTHREAD_START_ROUTINE)&ThreadFunc1,
      NULL, 0, NULL);
   hThread2 = CreateThread(NULL, 0, (LPTHREAD_START_ROUTINE)&ThreadFunc2,
      NULL, 0, NULL);

   if (hThread1 == NULL || hThread2 == NULL)       {
      cout << "CreateThread failed." << endl; 
      return 1;
   }

   retCode = WaitForSingleObject(hThread1,3000); // Maxwell: Wait Thread-1 with maximal 3 sec timeout.

   CloseHandle(hThread1);
   CloseHandle(hThread2);

   if (retCode == WAIT_OBJECT_0 && CriticalData == 1 )
      cout << "Success" << endl; // Thread-2 made it by changing value of CriticalData to "1" !!
   else
      cout << "Failure" << endl;
}

while (Sentinel)
{
   Sleep(0);
}

George_George wrote:

Sleep(0) will wait until there is a signal.

George_George wrote:

Then I am confused what is the original purpose of the MSDN sample? It advise people to put volatile keyword to all shared variables between multiple threads -- or else function may be broken?

George_George wrote:

1. In your experience, when do you use volatile?

George_George wrote:

If we share data between threads and we are already using synchronization approach like mutex, do we need to add additional volatile keyword?

George_George wrote:

and can be omitted.

George_George wrote:

1. In your experience, when do you use volatile?
2. If we share data between threads and we are already using synchronization approach like mutex, do we need to add additional volatile keyword?

Maxwell Chen wrote:

I don't use volatile in real-world code

George_George wrote:

In your cross-platform development experience, I think you mean if we are not using volatile, the function of the program will change in your experience?

George_George wrote:

Could you show a more specific example please? I am interested in this since in my experience I have not found any case which behaves differently with/without volatile

George_George wrote:

we can see that if we do not add volatile to the thread shared variables, the checking thread may deadlock -- I think it is what MSDN sample teaches us -- no volatile means no reliable and secure code.

George_George wrote:

But in the past the in many multi-threaded, I do not see they use volatile (they are using other approches like mutex). So, I am confused whether the their code is,

1. unsafe, and need to add volatile;
2. or safe, and mutex is an alternative way to tell compiler do not optimize (this is what I mean alternative).

JudyL_FL wrote:

I read that sample as saying that not using volatile means the code will not execute correctly since the compiler may mangle the code so that the variable does not operate as a mutex. Whether the code deadlocks or allows simultaneous access to the CriticalData variable depends on how the compiler optimizes the code.