Skip to main content

C++ Multithreading - Passing parameters to thread function and sleep thread

Today I am going to present you second article about new C++11 Multithreading feature. I am going to present you how we can pass parameters to thread function (by value and by reference) as well as show you how to delay thread processing for selected amount of time.

In previous C++ Multithreading article we presented how to create thread function and run second thread in parallel to main thread. Now we are extending that example with usage of another multithreding features. Let take a look on the new example:
Output of that example could be (it can be little other output on your machine because we are working with Multithreading which is not protected by phenomenon called race condition which I will explain in one of another articles): In point I and II we are creating new thread which will invoke function print() from point Ia. Difference to the example from previus article is that we are passing two parameters to our function. First parameter firstParam is passed by value, second parameter secondParam is passed by reference.

Notice that while we are passing parameters by reference to thread function we need to use std::ref() function. It is because, std::thread constructor creates new function object based on names of our function, passed parameters and std::bind() function within its body. Because we are using std::bind function we are using copies of our parameters, so without using std::ref we would create reference to copied parameter which we passed to std::thread constructor.

In such situation our when we would change value of our secondParam in the body of print() second thread's function, it would not be visible in main thread's function because inside print() function we would operate on reference to copy of secondParam object. You can experiment with it on your machine. Try to remove std::ref function and pass secondParam as normal reference. Compare your and ours output. You should not see value: startSecondParam_modified for secondParam parameter inside main thread (in second thread this value is applied, of course).

In point IIa and III we are printing values of firstParam and secondParam for both threads with modifications during processing. You can see that changes to secondParam value done in second thread are visible in main thread as well (because we are properly passing secondParam to second thread by reference).

One more new feature of C++11 multithreading which is introduced in this example you can notice in point IV. Function std::thread::sleep_for() is being used for stoping working of thread which it is invoked on for selected amount of time. We are using std::chrono::milliseconds() function (new function for C++11 as well) to sleep our main thread for 10 ms (in order to wait for modification of secondParam in second thread). This is simple example of delaying processing inside particular thread.

Point V should be known from previous article. We are waiting for finishing second thread before leavin from our main application (main thread).

Code of this example can be found on our GitHub account here: https://github.com/xmementoit/CppAdventureExamples/tree/master/multithreading/passParamsToThreadFunction


Comments

Popular posts from this blog

C++ Multithreading - Race conditions

In the previous C++ Multithreading article I presented you how to pass parameters between threads. Take a detail look on the output of that example once again: In the first line of that output you can notice that output text from two threads is mixed. You are probably wondering why it happens? It is because we are not protecting resources which are shared by two threads (in this example cout stream is shared in both threads) which causes multithreading's phenomenon called race condition . Because threads switching and accessing to shared resources are managed by operating system we do not know when std::cout stream will be accessed by main thread and when it will be accessed by second thread. Therefore in the previous article I mentioned that output of the example can be little different on your computer than my output example. What's more it is possible that this output will be different for few consecutive invoking of the example on the same machine. It is beca...

Advanced C++ - Stack unwinding

Stack unwinding is normally a concept of removing function entries from call stack (also known as Execution stack, Control stack, Function stack or Run-time stack). Call Stack is a stack data structure that stores active functions' addresses and helps in supporting function call/return mechanism. Every time when a function is called, an entry is made into Call stack which contains the return address of the calling function where the control needs to return after the execution of called function. This entry is called by various names like stack frame , activation frame or activation record. With respect to exception handling , stack Unwinding is a process of linearly searching function call stack to reach exception handler. When an exception occurs, if it is not handled in current function where it is thrown, the function Call Stack is unwound until the control reaches try block and then passes to catch block at the end of try block to handle exception. Also, in this proc...

Advanced C++ - Template Functions

Template function are special types of C++ function which can be used with different types (generic types). Thanks to that we can create one body of function which can be used for many different types. When we are creating template function compiler does not define any function for use at that time. Generation of funciton basing on templates are done during compilation process basing of differnt usage of template class. For better understanding take a look on following example: Output of this example is: In point I, we are defining template function which should return sum of two parameters. Without templates we need to define such function separately for each type which we should use it with (ex. separate function for int type, separate for double , separate for any other type). Thanks to template we can generate body of that function only once (as in point I) and use it to any type which is able to use body of that function. In our example we can use this template for an...