From the C++ FAQ:
[11.4] Can I overload the destructor for my class?
No.
I realize this means you cannot change the return type, arguments' types nor the number of arguments. I may be splitting hairs on the syntax of the words, but is it possible to override the Parent's destructor?
class Child : public Parent {
public:
virtual Parent::~Parent() {
// New definition
}
};
And for that matter do it recursively?
class Grandchild : public Child {
public:
Child::Parent::~Parent() {
// An even newer definition
}
};
I've read this and a related post and it makes me think because destructors are not inherited, they cannot be overridden, but I've never seen it explicitly stated.
EDIT: I changed this to reflect the fact that I want to override the Parent's destructor, note Child and Grandchild overriding ~Parent().
The main reason I am doing this is to maintain Parent's interface while changing the way it is destroyed (the entire reason for the child class). I will have something else managing all Parent's created and will explicitly call their destructors at a later time of my choosing.
I may be splitting hairs on the syntax of the words
No, you are definitely not – these are two very different things.
but is it possible to override the destructor?
Yes, and in fact you must do this in many cases. In order for this to work for a polymorphic object, you need to declare the base class destructor as virtual, though:
Parent const& p = Child();
Will properly call p.~Child() at the end of scope because Parent::~Parent is virtual.
Yes, it is possible to override the destructor of a class. In fact, when you define a class hierarchy in which polymorphism is used, you must declare a virtual destructor in the base class.
Overrides of destructors work exactly the same way overrides of normal member functions work in that when you destroy an object by deleteing the object via a pointer to the base class, the destructor of the derived class is properly called. This is why you must have a virtual destructor in the base class for polymorphic hierarchies.
However, there is a difference between virtual destructors and virtual member methods which has nothing to do with the virtual nature of the destructor. That is, when executing code like this:
class A
{
public:
virtual void Foo() {}
virtual ~A() {};
};
class B : public A
{
public:
void Foo() {};
~B() {}
};
int main()
{
A* a = new B;
a->Foo(); // B::Foo() is called
delete a; // B is destroyed via B::~B()
}
...when you call a->Foo(), the method Foo() in B is called. Since B::Foo() doesn't explicitly call A::Foo(), A::Foo() isn't called.
However, when the object is destroyed via delete a;, first the destructor B::~B() is called, and then after that finishes but before control returns to the program, the base class destructor A::~A() is also called.
Of course this is obvious when you think about it, and again this has nothing to do with the virtual nature of the destructor, but it does behave differently than a normal virtual method call, so I thought I'd point it out.
Obligitory Standard Quotation:
[C++03] 12.4/6 : Destructors
After executing the body of the destructor and destroying any
automatic objects allocated within the body, a destructor for class X
calls the destructors for X’s direct members, the destructors for X’s
direct base classes and, if X is the type of the most derived class
(12.6.2), its destructor calls the destructors for X’s virtual base
classes. All destructors are called as if they were referenced with a qualified name, that is, ignoring any possible virtual
overriding destructors in more derived classes. Bases and members are
destroyed in the reverse order of the completion of their
constructor (see 12.6.2). A return statement (6.6.3) in a destructor
might not directly return to the caller; before transferring control
to the caller, the destructors for the members and bases are called.
Destructors for elements of an array are called in reverse order of
their construction (see 12.6).
Yes: you can have virtual destructors, and the only reason is to override them in derived classes.
It looks like this:
class Parent {
public:
virtual ~Parent();
};
class Child : public Parent {
public:
virtual ~Child();
};
class Grandchild : public Child {
public:
~Grandchild(); // virtual is inherited here
};
Note that the destructor isn't overridden by name like ordinary functions, because the name is always that of the class whose instance you're destroying.
Note also that the parent class' destructors are always called too, so you don't need to duplicate their cleanup code: read up on member object and base-class sub-object construction and destruction order for the details.
Terminology
overriding a function means implementing a base-class virtual function in a derived class. You can't change the signature at all (except for using covariant return types). So, an override always has the same signature as an inherited virtual function.
overloading a function means implementing multiple functions with the same name (and in some sense the same scope). So, an overload always has a different signature to the others with the same name, doesn't relate directly to virtual dispatch, and isn't necessarily inherited.
Yes; you can, and should, make a destructor virtual, whenever you have a child class which may be destroyed using a reference to the base class. Static code analysis tools will even complain if you don't offer a virtual destructor.
Consider the following example:
class A
{
public:
A() { a = new int; }
virtual ~A() { delete a; }
private:
int *a;
};
class B final : public A
{
public:
B() { b = new int; }
~B() { delete b; }
private:
int *b;
};
int main()
{
A *a = new B();
delete a;
}
If A's destructor was not virtual, then delete a would only call A's destructor, and you would end up with a memory leak. But because it's virtual, both destructors will be called, in the order ~B() -> ~A().
Related
I understand that whenever you have a polymorphic base class, the base class should define a virtual destructor. So that when a base-class pointer to a derived-class object is deleted, it will call the destructor of the derived class first. Correct me if i am wrong here.
also, if the base-class destructor were to be non-virtual, it would be undefined behavior to delete a baseclass pointer to a derived object. Correct me if i am wrong aswell.
so my question is: Why is it exactly, that when the base-class destructor is non-virtual, the object will not be destroyed correctly?
i am assuming this is because virtual functions have some kind of table that is memorized and consulted whenever a virtual function is called. And the compiler knows that when an object is supposed to be deleted, it should call the derived destructor first.
is my assumption correct?
If at the point where you delete the object the static type of the variable is the bas type, than the destructor of the base type will be called, but the destructor of the sub class won't be called (as it is not virtual).
As a result the resources allocated by the base class will be freed, but the resources allocated by the sub class won't.
Thus the object won't be destructed correctly.
You are correct about that table: it is called a virtual method table or "vtable". But the result of the destructor being non-virtual is not that the destructors are not called in the correct order, but that the destructor(s) of the sub class(es) are not called at all!
Consider
struct Base {
void f() { printf("Base::f"); }
};
struct Derived : Base {
void f() { printf("Derived::f"); }
};
Base* p = new Derived;
p->f();
This prints Base::f, because Base::f is not virtual. Now do the same with destructors:
struct Base {
~Base() { printf("Base::~Base"); }
};
struct Derived : Base {
~Derived() { printf("Derived::~Derived"); }
};
Base* p = new Derived;
p->~Base();
This prints Base::~Base. Now if we make the destructor virtual, then, as with any other virtual function, the final overrider in the dynamic type of the object is called. A destructor overrides a virtual destructor in a base class (even though its "name" is different):
struct Base {
virtual ~Base() { printf("Base::~Base"); }
};
struct Derived : Base {
~Derived() override { printf("Derived::~Derived"); }
};
Base* p = new Derived;
p->~Base();
The call p->~Base() actually invokes Derived::~Derived(). Since this is a destructor, after its body finishes executing, it automatically invokes destructors of bases and members. So the output is
Derived::~Derived
Base::~Base
Now, a delete-expression is in general equivalent to a destructor call followed by a call to a memory deallocation function. In this particular case, the expression
delete p;
is equivalent to
p->~Base();
::operator delete(p);
So if the destructor is virtual, this does the right thing: it calls Derived::~Derived first, which then automatically calls Base::~Base when it's done. If the destructor isn't virtual, the likely result is that only Base::~Base is invoked.
Do C++ compilers generate the default functions like Constructor/Destructor/Copy-Constructor... for this "class"?
class IMyInterface
{
virtual void MyInterfaceFunction() = 0;
}
I mean it is not possible to instantiate this "class", so i think no default functions are generated.
Otherwise, people are saying you have to use a virtual destructor.
Which means if i dont define the destructor virtual it will be default created, not virtual.
Furthermore i wannt to know if it is reasonable to define a virtual destructor for a pure virtual Interface, like the one above? (So no pointers or data is used in here, so nothing has to be destructed)
Thanks.
Yes.
There is no wording that requires the class to be instantiable in order for these special member functions to be implicitly declared.
This makes sense — just because you cannot instantiate the Base, doesn't mean a Derived class doesn't want to use these functions.
struct Base
{
virtual void foo() = 0;
int x;
};
struct Derived : Base
{
Derived() {}; // needs access to Base's trivial implicit ctor
virtual void foo() {}
};
See:
§12.1/5 (ctor)
§12.8/9 (move)
§12.8/20 (copy)
Furthermore i wannt to know if it is reasonable to define a virtual destructor for a pure virtual Interface, like the one above? (So no pointers or data is used in here, so nothing has to be destructed)
It's not only reasonable, it's recommended. This is because in the case of virtual function hierarchies, (automatically) calling a destructor of a specialized class also calls all destructors of it's base classes. If they are not defined, you should get a linking error.
If you define at least one virtual function in your class you should also define a virtual destructor.
The destructor can be defined with =default though:
Here's a corrected (compilable) code example:
class ImyInterface
{
virtual void myInterfaceFunction() = 0;
virtual ~ImyInterface() = 0;
}
ImyInterface::~ImyInterface() = default;
Furthermore i wannt to know if it is reasonable to define a virtual destructor for a pure virtual Interface, like the one above? (So no pointers or data is used in here, so nothing has to be destructed)
Will the derived classes ever do anything in their destructors? Can you be certain they never will, even when somebody else takes over development?
The whole point of having a virtual destructor is not to make sure the base class is properly destructed, that will happen anyway. The point is that the derived class's destructor is called when you use a generic interface:
struct A {
virtual ~A() {}
virtual int f() = 0;
};
class B : public A {
std::ifstream fh;
public:
virtual ~B() {}
virtual int f() { return 42; }
};
std::shared_ptr<A> a = new B;
When a goes out of scope, why is the ifstream closed? Because the destructor deletes the object using the virtual destructor.
This addresses the second question about declaring a virtual destructor for an abstract base class (e.g. at least one member function is pure virtual). Here is a real world example of the LLVM clang++ compiler catching a potential problem. This occurred with the command line tools version supplied by Apple Developer for the Mac OS X Mavericks operating system.
Suppose you have a collection of derived classes that ultimately have the parent with the abstract base class to define the common interface. Then it is necessary to have a storage container like a vector that is intentionally declared to store a pointer to the abstract base class for each element. Later on, following good engineering practices, the container elements need to be "deleted" and the memory returned to the heap. The simplest way to do this is to traverse the vector element by element and invoke the delete operation on each one.
Well, if the abstract base class does not declare the destructor as virtual, the clang++ compiler gives a friendly warning about calling the non-virtual destructor on an abstract class. Keep in mind that in reality only the derived classes are allocated from the heap with operator new. The derived class pointer type from the inheritance relationship is indeed the abstract base class type (e.g. the is-a relationship).
If the abstract base class destructor is not virtual, then how will the correct derived class' destructor be invoked to release the memory? At best the compiler knows better (at least potentially does with C++11), and makes it happen. If the -Wall compiler option is enabled, then at least the compilation warning should appear. However, at worse, the derived class destructor is never reached and the memory is never returned to the heap. Hence there is now a memory leak that may be very challenging to track down and fix. All it will take is a single addition of "virtual" to the abstract base class destructor declaration.
Example code:
class abstractBase
{
public:
abstractBase() { };
~abstractBase() { };
virtual int foo() = 0;
};
class derived : abstractBase
{
public:
derived() { };
~derived() { };
int foo() override { return 42; }
};
//
// Later on, within a file like main.cpp . . .
// (header file includes are assumed to be satisfied)
//
vector<abstractBase*> v;
for (auto i = 0; i < 1000; i++)
{
v.push_back(new derived());
}
//
// do other stuff, logic, what not
//
//
// heap is running low, release memory from vector v above
//
for (auto i = v.begin(); i < v.end(); i++)
{
delete (*i); // problem is right here, how to find the derived class' destructor?
}
To resolve this potential memory leak, the abstract base class has to declare its destructor as virtual. Nothing else is required. The abstract base class now becomes:
class abstractBase
{
public:
abstractBase() { };
virtual ~abstractBase() { }; // insert virtual right here
virtual int foo() = 0;
}
Note that the abstract base class has currently empty constructor and destructor bodies. As Lightness answered above, the compiler creates a default constructor, destructor, and copy constructor for an abstract base class (if not defined by the engineer). It is highly recommended to review any of The C++ Programming Language editions by the C++ creator, Bjarne Stroustrup, for more details on abstract base classes.
Given the following example:
class BaseClass
{
BaseClass()
{
};
virtual ~BaseClass()
{
this->Cleanup();
};
virtual void Cleanup()
{
// Do cleanup here.
};
};
class Level1DerivedClass : public BaseClass
{
Level1DerivedClass()
{
};
virtual ~Level1DerivedClass()
{
};
virtual void Cleanup()
{
// Call my base cleanup.
BaseClass::Cleanup();
// Do additional cleanup here.
};
};
class Level2DerivedClass : public Level1DerivedClass
{
Level2DerivedClass()
{
};
~Level2DerivedClass()
{
};
void Cleanup()
{
// Call my base cleanup.
Level1DerivedClass::Cleanup();
// Do additional cleanup here.
};
};
main()
{
Level2DerivedClass * derived2 = new Level2DerivedClass();
delete derived2;
return 0;
}
When I delete my derived class reference, I would EXPECT the flow would be as follows:
Level2DerivedClass destructor is executed.
Because Level1DerivedClass destructor is virtual, it would be executed.
Because BaseClass destructor is virtual, it would be executed.
Because BaseClass::Cleanup and Level1DerivedClass::Cleanup are both virtual, the call from the BaseClass 'this' pointer in the BaseClass destructor would executed the implementation of the most derived class - Level2DerivedClass::Cleanup.
Level2DerivedClass::Cleanup calls its parent's Cleanup implementation.
Level1DerivedClass::Cleanup calls its parent's Cleanup implementation.
What is happening is that it is calling the destructors for each level of inheritance (1 - 3) above the way I'm expecting. But when this->Cleanup() is called from the BaseClass destructor, it only executes its own implementation. I don't understand why this is happening because normally when you instantiate a derived class pointer, cast it as a base class pointer, and call a virtual method from the base class pointer (in this case, 'this'), it still runs the derived class implementation (the whole point of 'virtual', yes?). In my example, Level2DerivedClass::Cleanup and Level1DerivedClass::Cleanup never gets called.
The reason I'm setting it up this way is I want to be able to call my Cleanup code without having to destroy my object, which is why I'm abstracting it from the actual destructor body.
If you have suggestions on a more proper way to do this, I'm all ears. But I would also like an explanation of why my setup doesn't work - what am I misunderstanding?
Thank you in advance for your time.
The rule of thumb is: Never Call Virtual Functions during Construction or Destruction.
They don't behave as you might expect; as each destructor finishes, the dynamic type of this is effectively modified. From [class.cdtor] in the C++ standard:
When a virtual function is called directly or indirectly from a constructor (including the mem-initializer or
brace-or-equal-initializer for a non-static data member) or from a destructor, and the object to which the
call applies is the object under construction or destruction, the function called is the one defined in the
constructor or destructor’s own class or in one of its bases, but not a function overriding it in a class derived
from the constructor or destructor’s class, or overriding it in one of the other base classes of the most derived
object.
The Proper Way Of Doing Things is: clean after yourself, and yourself only, in the destructor. Don't clean after your kids or your parents.
If you want to clean up things not from the destructor, You Are Doing It Wrong. In C++ we have this little thing called RAII, Resource Acquisition Is Initialization. But there's also its dual, which does not seem to have an officially sounding name, but here's something that could work: RDID, Resource Disposal Is Destruction.
Of course you don't have to adhere to the RAII/RDID philosophy, but that would be Not The C++ Way.
in C++: Why should the destructor of base classes be virtual?
The better question is when and why. Your question indicates that you think all base classes should have virtual destructors, which is not quite true.
It would make it impossible to apply the empty base class optimization, and could multiply the size of classes up to 16 times than what it would be without virtual on common platforms.
A virtual destructor is needed when you delete an object whose dynamic type is DerivedClass by a pointer that has type BaseClass*. The virtual makes the compiler associate information in the object making it able to execute the derived class destructor. Missing the virtual in such case causes undefined behavior.
If you don't need this, and your class is only used as a base class, it's best to make the destructor protected, thus preventing that users accidentally delete in the described way.
You want them to be virtual so that all subclass destructors are automatically called when the object is destroyed, even if it is destroyed through a pointer to the base class. In the following code:
class base {
public:
virtual ~base() { }
};
class derived : public base {
public:
~derived() { } // Inherits the virtual designation
};
int main(void)
{
base *b = new derived;
delete b;
}
The derived destructor will only be called if the base destructor is virtual.
As Magnus indicates, you don't have to do this if you aren't taking advantage of polymorphism. However, I try to develop the habit of declaring all my destructors virtual. It protects me against the case where I should have declared them virtual but forget to do so. As Johannes indicates, this habit can impose a small space and performance penalty when the virtual designation is not needed.
They dont have to be virtual unless you are using polymorphism. If you do use polymorphism they need to be virtual so that the destructors of inherited classes are guaranteed to be called, so inherited classes can do their clean up.
For situations like this:
class A
{
virtual ~A();
};
class B:A
{
~B();
};
A *a = new B(); //legal, since it's a downcast
delete a; //Unless the destructor is virtual, ~A() is called here instead of ~B().
It should be virtual to ensure that the destructor of the inherited classes are the ones actually getting called at runtime instead of the base class destructor being called.
A difference between a destructor (of course also the constructor) and other member functions is that, if a regular member function has a body at the derived class, only the version at Derived class gets executed. Whereas in case of destructors, both derived as well as base class versions get executed?
It will be great to know what exactly happens in case of destructor (maybe virtual) & constructor, that they are called for all its base classes even if the most derived class object is deleted.
Thanks in advance!
The Standard says
After executing the body of the destructor and destroying any automatic objects allocated within the body,
a destructor for class X calls the destructors for X’s direct non-variant members,the destructors for X’s direct
base classes and, if X is the type of the most derived class (12.6.2), its destructor calls the destructors for
X’s virtual base classes. All destructors are called as if they were referenced with a qualified name, that is,
ignoring any possible virtual overriding destructors in more derived classes. Bases and members are destroyed
in the reverse order of the completion of their constructor (see 12.6.2). A return statement (6.6.3) in a
destructor might not directly return to the caller; before transferring control to the caller, the destructors
for the members and bases are called. Destructors for elements of an array are called in reverse order of
their construction (see 12.6).
Also as per RAII resources need to be tied to the lifespan of suitable objects and the destructors of respective classes must be called upon to release the resources.
For example the following code leaks memory.
struct Base
{
int *p;
Base():p(new int){}
~Base(){ delete p; } //has to be virtual
};
struct Derived :Base
{
int *d;
Derived():Base(),d(new int){}
~Derived(){delete d;}
};
int main()
{
Base *base=new Derived();
//do something
delete base; //Oops!! ~Base() gets called(=>Memory Leak).
}
Constructor and destructor are different from the rest of regular methods.
Constructor
can't be virtual
in derived class you either call explicitly constructor of base class
or, in case where you don't call base class constructor compiler will insert the call. It will call the base constructor without parameters. If no such constructor exists then you get compiler error.
struct A {};
struct B : A { B() : A() {} };
// but this works as well because compiler inserts call to A():
struct B : A { B() {} };
// however this does not compile:
struct A { A(int x) {} };
struct B : A { B() {} };
// you need:
struct B : A { B() : A(4) {} };
Destructor:
when you call destructor on derived class over a pointer or a reference, where the base class has virtual destructor, the most derived destructor will be called first and then the rest of derived classes in reversed order of construction. This is to make sure that all memory has been properly cleaned. It would not work if the most derived class was called last because by that time the base class would not exists in memory and you would get segfault.
struct C
{
virtual ~C() { cout << __FUNCTION__ << endl; }
};
struct D : C
{
virtual ~D() { cout << __FUNCTION__ << endl; }
};
struct E : D
{
virtual ~E() { cout << __FUNCTION__ << endl; }
};
int main()
{
C * o = new E();
delete o;
}
output:
~E
~D
~C
If the method in base class is marked as virtual all the inherited methods are virtual as well so even if you don't mark the destructors in D and E as virtual they will still be virtual and they still get called in the same order.
This is by design. The destructor on the base class must be called in order for it to release its resources. Rule of thumb is that a derived class should only clean up its own resources and leave the base class to clean up itself.
From C++ spec:
After executing the body of the
destructor and destroying any
automatic objects allocated within the
body, a destructor for class X calls
the destructors for X’s direct
members, the destructors for X’s
direct base classes and, if X is the
type of the most derived class
(12.6.2), its destructor calls the
destructors for X’s virtual base
classes. All destructors are called as
if they were referenced with a
qualified name, that is, ignoring any
possible virtual overriding
destructors in more derived classes.
Bases and members are destroyed in the
reverse order of the completion of
their constructor (see
12.6.2).
Also, because there is only one destructor, there is no ambiguity as to which destructor a class must call. This is not the case for constructors, where a programmer must pick which base class constructor should be called if there isn't an accessible default constructor.
Because that's how dtor's work. When you create an object, ctors are invoked starting from the base, and going all the way to the most derived. When you destroy objects (correctly) the reverse happens. The time that making a dtor virtual makes a difference is if/when you destroy an object via a pointer (or reference, though that's fairly unusual) to the base type. In that case, the alternative isn't really that only the derived dtor gets invoked -- rather, the alternative is simply undefined behavior. That make happen to take the form of invoking only the derived dtor, but it might take an entirely different form as well.
A base class destructor may be responsible for cleaning up resources that were allocated by the base class constructor.
If your base class has a default constructor (one that doesn't take parameters or has defaults for all its parameters) that constructor is automatically called upon construction of a derived instance.
If your base class has a constructor that requires parameters, you must call it manually in the initializer list of the derived class constructor.
Your base class destructor will always be automatically called upon deletion of the derived instance since destructors don't take parameters.
If you're using polymorphism and your derived instance is pointed to by a base class pointer, then the derived class destructor is only called if the base destructor is virtual.
As Igor says constructors must be called for base classes. Consider what would happen if it wouldn't be called:
struct A {
std::string s;
virtual ~A() {}
};
struct B : A {};
If the destructor for A would not be called when deleting a B instance, A would never be cleaned up.
When any object is destroyed, destructors run for all sub-objects. This includes both reuse by containment and reuse by inheritance.