Question

Consider the following code:

unsigned int __stdcall func( LPVOID ) {
    LRESULT result = ::PostThreadMessage( ::GetCurrentThreadId(), 0, 0, 0 );
    return 0;
}

int wmain() {
    _beginthreadex( NULL, 0, func, NULL, 0, NULL );
    ...
}

Why does ::PostThreadMessage succeed? I think that it should fail because a message queue should not be created by that moment

Answer 1

Because you are calling PostThreadMessage() on the current thread, the system is able to create the message queue on demand. If you were calling PostThreadMessage() and passing the ID of a thread other than the calling thread, then it would fail if that thread did not have a message queue.

For example, consider the following variant of your code:

unsigned int __stdcall func( LPVOID ) {
    return 0;
}

int _tmain(int argc, _TCHAR* argv[])
{
    unsigned int threadID;
    _beginthreadex( NULL, 0, func, NULL, 0, &threadID );
    LRESULT result = ::PostThreadMessage( threadID, 0, 0, 0 );
    DWORD error = ::GetLastError();
    return 0;
}

Because we are now attempting to post the message from the main thread, to the worker thread, result comes back as 0 (i.e. an error), and error is set to ERROR_INVALID_THREAD_ID as described by the documentation for PostThreadMessage().

If the function fails, the return value is zero. To get extended error information, call GetLastError. GetLastError returns ERROR_INVALID_THREAD_ID if idThread is not a valid thread identifier, or if the thread specified by idThread does not have a message queue.