Question

In C++, if we declare a stack variable inside of a function, is it automatically deleted at the end of the function or is it deleted at the end of the program execution?

Also, is the answer to this question the same for the C language?

Answer 1

For stack-declared variables, the destructor is called and the memory reclaimed as it falls out of scope.

Note that this doesn’t mean at the end of the function if the variable is declared in an inner block, like an if-statement or loop.

int main(int argc, char **argv)
{
    int a = 3;

    if (argc > 1)
    {
        int b = 5;
        ++b;
    } // b is destructed here

    // a is destructed here
    // argv and argc are destructed here (or with a)
}

EDIT: A good point was made about the fact that it doesn’t matter how the scope is exited. So…

#include <vector>
# include <exception>

int main(int argc, char **argv)
{
    std::vector<int> myVector(10);

    try
    {
        if (argc)
        {
            int a = 10;
            int b = 12;
            for (int c = 0; c < b; c++) // lol
            {
                int c_squared = c*c;
                int theEleventhElement = myVector.at(c);
                // the above will throw std::out_of_range at some point!
            }
        }
    }
    catch (std::out_of_range &ex)
    {
        // do nothing
    }
}

When the above throws, the stack will unwind as part of the exception handling. Thus, the variables will be destroyed in the following order:

• c_squared
• c
• b and a (I think in that order, but I don’t know if that’s mandated in the standard)

At this point, there is finally a catch handler with only myVector still in scope. That block ignores the exception, and then main ends– at THAT point myVector is destructed.