I've been evaluating various smart pointer implementations (wow, there are a LOT out there) and it seems to me that most of them can be categorized into two broad classifications:
1) This category uses inheritance on the objects referenced so that they have reference counts and usually up() and down() (or their equivalents) implemented. IE, to use the smart pointer, the objects you're pointing at must inherit from some class the ref implementation provides.
2) This category uses a secondary object to hold the reference counts. For example, instead of pointing the smart pointer right at an object, it actually points at this meta data object... Who has a reference count and up() and down() implementations (and who usually provides a mechanism for the pointer to get at the actual object being pointed to, so that the smart pointer can properly implement operator ->()).
Now, 1 has the downside that it forces all of the objects you'd like to reference count to inherit from a common ancestor, and this means that you cannot use this to reference count objects that you don't have control over the source code to.
2 has the problem that since the count is stored in another object, if you ever have a situation that a pointer to an existing reference counted object is being converted into a reference, you probably have a bug (I.E., since the count is not in the actual object, there is no way for the new reference to get the count... ref to ref copy construction or assignment is fine, because they can share the count object, but if you ever have to convert from a pointer, you're totally hosed)...
Now, as I understand it, boost::shared_pointer uses mechanism 2, or something like it... That said, I can't quite make up my mind which is worse! I have only ever used mechanism 1, in production code... Does anyone have experience with both styles? Or perhaps there is another way thats better than both of these?
"What is the best way to implement smart pointers in C++"
Don't! Use an existing, well tested smart pointer, such as boost::shared_ptr or std::tr1::shared_ptr (std::unique_ptr and std::shared_ptr with C++ 11)
If you have to, then remember to:
use safe-bool idiom
provide an operator->
provide the strong exception guarantee
document the exception requirements your class makes on the deleter
use copy-modify-swap where possible to implement the strong exception guarantee
document whether you handle multithreading correctly
write extensive unit tests
implement conversion-to-base in such a way that it will delete on the derived pointer type (policied smart pointers / dynamic deleter smart pointers)
support getting access to raw pointer
consider cost/benifit of providing weak pointers to break cycles
provide appropriate casting operators for your smart pointers
make your constructor templated to handle constructing base pointer from derived.
And don't forget anything I may have forgotten in the above incomplete list.
Just to supply a different view to the ubiquitous Boost answer (even though it is the right answer for many uses), take a look at Loki's implementation of smart pointers. For a discourse on the design philosophy, the original creator of Loki wrote the book Modern C++ Design.
I've been using boost::shared_ptr for several years now and while you are right about the downside (no assignment via pointer possible), I think it was definitely worth it because of the huge amount of pointer-related bugs it saved me from.
In my homebrew game engine I've replaced normal pointers with shared_ptr as much as possible. The performance hit this causes is actually not so bad if you are calling most functions by reference so that the compiler does not have to create too many temporary shared_ptr instances.
Boost also has an intrusive pointer (like solution 1), that doesn't require inheriting from anything. It does require changing the pointer to class to store the reference count and provide appropriate member functions. I've used this in cases where memory efficiency was important, and didn't want the overhead of another object for each shared pointer used.
Example:
class Event {
public:
typedef boost::intrusive_ptr<Event> Ptr;
void addRef();
unsigned release();
\\ ...
private:
unsigned fRefCount;
};
inline void Event::addRef()
{
fRefCount++;
}
inline unsigned Event::release(){
fRefCount--;
return fRefCount;
}
inline void intrusive_ptr_add_ref(Event* e)
{
e->addRef();
}
inline void intrusive_ptr_release(Event* e)
{
if (e->release() == 0)
delete e;
}
The Ptr typedef is used so that I can easily switcth between boost::shared_ptr<> and boost::intrusive_ptr<> without changing any client code
If you stick with the ones that are in the standard library you will be fine.
Though there are a few other types than the ones you specified.
Shared: Where the ownership is shared between multiple objects
Owned: Where one object owns the object but transfer is allowed.
Unmovable: Where one object owns the object and it can not be transferred.
The standard library has:
std::auto_ptr
Boost has a couple more than have been adapted by tr1 (next version of the standard)
std::tr1::shared_ptr
std::tr1::weak_ptr
And those still in boost (which in relatively is a must have anyway) that hopefully make it into tr2.
boost::scoped_ptr
boost::scoped_array
boost::shared_array
boost::intrusive_ptr
See:
Smart Pointers: Or who owns you baby?
It seems to me this question is kind of like asking "Which is the best sort algorithm?" There is no one answer, it depends on your circumstances.
For my own purposes, I'm using your type 1. I don't have access to the TR1 library. I do have complete control over all the classes I need to have shared pointers to. The additional memory and time efficiency of type 1 might be pretty slight, but memory usage and speed are big issues for my code, so type 1 was a slam dunk.
On the other hand, for anyone who can use TR1, I'd think the type 2 std::tr1::shared_ptr class would be a sensible default choice, to be used whenever there isn't some pressing reason not to use it.
The problem with 2 can be worked around. Boost offers boost::shared_from_this for this same reason. In practice, it's not a big problem.
But the reason they went with your option #2 is that it can be used in all cases. Relying on inheritance isn't always an option, and then you're left with a smart pointer you can't use for half your code.
I'd have to say #2 is best, simply because it can be used in any circumstances.
Our project uses smart pointers extensively. In the beginning there was uncertainty about which pointer to use, and so one of the main authors chose an intrusive pointer in his module and the other a non-intrusive version.
In general, the differences between the two pointer types were not significant. The only exception being that early versions of our non-intrusive pointer implicitly converted from a raw pointer and this can easily lead to memory problems if the pointers are used incorrectly:
void doSomething (NIPtr<int> const &);
void foo () {
NIPtr<int> i = new int;
int & j = *i;
doSomething (&j); // Ooops - owned by two pointers! :(
}
A while ago, some refactoring resulted in some parts of the code being merged, and so a choice had to be made about which pointer type to use. The non-intrusive pointer now had the converting constructor declared as explicit and so it was decided to go with the intrusive pointer to save on the amount of code change that was required.
To our great surprise one thing we did notice was that we had an immediate performance improvement by using the intrusive pointer. We did not put much research into this, and just assumed that the difference was the cost of maintaining the count object. It is possible that other implementations of non-intrusive shared pointer have solved this problem by now.
What you are talking about are intrusive and non-intrusive smart pointers. Boost has both. boost::intrusive_ptr calls a function to decrease and increase the reference count of your object, everytime it needs to change the reference count. It's not calling member functions, but free functions. So it allows managing objects without the need to change the definition of their types. And as you say, boost::shared_ptr is non-intrusive, your category 2.
I have an answer explaining intrusive_ptr: Making shared_ptr not use delete. In short, you use it if you have an object that has already reference counting, or need (as you explain) an object that is already referenced to be owned by an intrusive_ptr.
Related
I'm having trouble getting things organized properly with smart pointers. Almost to the point that I feel compelled to go back to using normal pointers.
I would like to make it easy to use smart pointers throughout the program without having to type shared_ptr<...> every time. One solution I think of right away is to make a template class and add a typedef sptr to it so I can do class Derived : public Object < Derived > .. and then use Derived::sptr = ... But this obviously is horrible because it does not work with another class that is then derived from Derived object.
And even doing typedef shared_ptr<..> MyObjectPtr is horrible because then it needs to be done for each kind of smart pointer for consistency's sake, or at least for unique_ptr and shared_ptr.
So what's the standard way people use smart pointers? Because frankly I'm starting to see it as being too much hassle to use them. :/
So what's the standard way people use smart pointers?
Rarely. The fact that you find it a hassle to use them is a sign that you over-use pointers. Try to refactor your code to make pointers the exception, not the rule. shared_ptr in particular has its niche, but it’s a small one: namely, when you genuinely have to share ownership of a resource between several objects. This is a rare situation.
Because frankly I'm starting to see it as being too much hassle to use them. :/
Agreed. That’s the main reason not to use pointers.
There are more ways to avoid pointers. In particular, shared_ptr really only needs to spelled out when you actually need to pass ownership. In functions which don’t deal with ownership, you wouldn’t pass a shared_ptr, or a raw pointer; you would pass a reference, and dereference the pointer upon calling the function.
And inside functions you almost never need to spell out the type; for instance, you can (and should) simply say auto x = …; instead of shared_ptr<Class> x = …; to initialise variables.
In summary, you should only need to spell out shared_ptr in very few places in your code.
I have a lot of code that creates objects dynamically. So using pointers is necessary because the number of objects is not known from the start. An object is created in one subsystem, then stored in another, then passed for further processing to the subsystem that created it. So that I guess means using shared_ptr. Good design? I don't know, but it seems most logical to ask subsystem to create a concrete object that it owns, return a pointer to an interface for that object and then pass it for further processing to another piece of code that will interact with the object through it's abstract interface.
I could return unique_ptr from factory method. But then I would run into trouble if I need to pass the object for processing multiple times. Because I would still need to know about the object after I pass it to another method and unique_ptr would mean that I lose track of the object after doing move(). Since I need to have at least two references to the object this means using shared_ptr.
I heard somewhere that most commonly used smart pointer is unique_ptr. Certainly not so in my application. I end up with using shared_ptr mush more often. Is this a sign of bad design then?
I am trying to write a simple game using C++ and SDL. My question is, what is the best practice to store class member variables.
MyObject obj;
MyObject* obj;
I read a lot about eliminating pointers as much as possible in similar questions, but I remember that few years back in some books I read they used it a lot (for all non trivial objects) . Another thing is that SDL returns pointers in many of its functions and therefor I would have to use "*" a lot when working with SDL objects.
Also am I right when I think the only way to initialize the first one using other than default constructor is through initializer list?
Generally, using value members is preferred over pointer members. However, there are some exceptions, e.g. (this list is probably incomplete and only contains reason I could come up with immediately):
When the members are huge (use sizeof(MyObject) to find out), the difference often doesn't matter for the access and stack size may be a concern.
When the objects come from another source, e.g., when there are factory function creating pointers, there is often no alternative to store the objects.
If the dynamic type of the object isn't known, using a pointer is generally the only alternative. However, this shouldn't be as common as it often is.
When there are more complicated relations than direct owner, e.g., if an object is shared between different objects, using a pointer is the most reasonable approach.
In all of these case you wouldn't use a pointer directly but rather a suitable smart pointer. For example, for 1. you might want to use a std::unique_ptr<MyObject> and for 4. a std::shared_ptr<MyObject> is the best alternative. For 2. you might need to use one of these smart pointer templates combined with a suitable deleter function to deal with the appropriate clean-up (e.g. for a FILE* obtained from fopen() you'd use fclose() as a deleter function; of course, this is a made up example as in C++ you would use I/O streams anyway).
In general, I normally initialize my objects entirely in the member initializer list, independent on how the members are represented exactly. However, yes, if you member objects require constructor arguments, these need to be passed from a member initializer list.
First I would like to say that I completely agree with Dietmar Kühl and Mats Petersson answer. However, you have also to take on account that SDL is a pure C library where the majority of the API functions expect C pointers of structs that can own big chunks of data. So you should not allocate them on stack (you shoud use new operator to allocate them on the heap). Furthermore, because C language does not contain smart pointers, you need to use std::unique_ptr::get() to recover the C pointer that std::unique_ptr owns before sending it to SDL API functions. This can be quite dangerous because you have to make sure that the std::unique_ptr does not get out of scope while SDL is using the C pointer (similar problem with std::share_ptr). Otherwise you will get seg fault because std::unique_ptr will delete the C pointer while SDL is using it.
Whenever you need to call pure C libraries inside a C++ program, I recommend the use of RAII. The main idea is that you create a small wrapper class that owns the C pointer and also calls the SDL API functions for you. Then you use the class destructor to delete all your C pointers.
Example:
class SDLAudioWrap {
public:
SDLAudioWrap() { // constructor
// allocate SDL_AudioSpec
}
~SDLAudioWrap() { // destructor
// free SDL_AudioSpec
}
// here you wrap all SDL API functions that involve
// SDL_AudioSpec and that you will use in your program
// It is quite simple
void SDL_do_some_stuff() {
SDL_do_some_stuff(ptr); // original C function
// SDL_do_some_stuff(SDL_AudioSpec* ptr)
}
private:
SDL_AudioSpec* ptr;
}
Now your program is exception safe and you don't have the possible issue of having smart pointers deleting your C pointer while SDL is using it.
UPDATE 1: I forget to mention that because SDL is a C library, you will need a custom deleter class in order to proper manage their C structs using smart pointers.
Concrete example: GSL GNU scientific library. Integration routine requires the allocation of a struct called "gsl_integration_workspace". In this case, you can use the following code to ensure that your code is exception safe
auto deleter= [](gsl_integration_workspace* ptr) {
gsl_integration_workspace_free(ptr);
};
std::unique_ptr<gsl_integration_workspace, decltype(deleter)> ptr4 (
gsl_integration_workspace_alloc (2000), deleter);
Another reason why I prefer wrapper classes
In case of initialization, it depends on what the options are, but yes, a common way is to use an initializer list.
The "don't use pointers unless you have to" is good advice in general. Of course, there are times when you have to - for example when an object is being returned by an API!
Also, using new will waste quite a bit of memory and CPU-time if MyObject is small. Each object created with new has an overhead of around 16-48 bytes in a typical modern OS, so if your object is only a couple of simple types, then you may well have more overhead than actual storage. In a largeer application, this can easily add up to a huge amount. And of course, a call to new or delete will most likely take some hundreds or thousands of cycles (above and beyond the time used in the constructor). So, you end up with code that runs slower and takes more memory - and of course, there's always some risk that you mess up and have memory leaks, causing your program to potentially crash due to out of memory, when it's not REALLY out of memory.
And as that famous "Murphy's law states", these things just have to happen at the worst possible and most annoying times - when you have just done some really good work, or when you've just succeeded at a level in a game, or something. So avoiding those risks whenever possible is definitely a good idea.
Well, creating the object is a lot better than using pointers because it's less error prone. Your code doesn't describe it well.
MyObj* foo;
foo = new MyObj;
foo->CanDoStuff(stuff);
//Later when foo is not needed
delete foo;
The other way is
MyObj foo;
foo.CanDoStuff(stuff);
less memory management but really it's up to you.
As the previous answers claimed the "don't use pointers unless you have to" is a good advise for general programming but then there are many issues that could finally make you select the pointers choice. Furthermore, in you initial question you are not considering the option of using references. So you can face three types of variable members in a class:
MyObject obj;
MyObject* obj;
MyObject& obj;
I use to always consider the reference option rather than the pointer one because you don't need to take care about if the pointer is NULL or not.
Also, as Dietmar Kühl pointed, a good reason for selecting pointers is:
If the dynamic type of the object isn't known, using a pointer is
generally the only alternative. However, this shouldn't be as common
as it often is.
I think this point is of particular importance when you are working on a big project. If you have many own classes, arranged in many source files and you use them in many parts of your code you will come up with long compilation times. If you use normal class instances (instead of pointers or references) a simple change in one of the header file of your classes will infer in the recompilation of all the classes that include this modified class. One possible solution for this issue is to use the concept of Forward declaration, which make use of pointers or references (you can find more info here).
If I could I would remove all raw pointers * from my code, because using them may be not thread safe and intentions of the design are not clear (optional value, ownership, etc).
Sometimes however it is not that easy to not use pointers. For example we tend to use pointers for a base type in a container of polymorphic types:
class A : noncopyable { ... };
class B : public A { ... };
std::vector<A*> v;
v.emplace_back(new B);
// temporary container for some operation
std::vector<A*> selected;
if(check())
selected.emplace_back(v.front());
What can you say about above code? Who is the owner? Is it a shared ownership or not? It is why we should probably do that for v:
std::vector<std::unique_ptr<A>> v;
v.emplace_back(make_unique<B>());
Now it is clear that v owns the objects but I still do not like that selected has a raw pointer and makes my design not intuitive. Looking into Standard C++ library I think that there is only one type that could do the job - std::reference_wrapper:
std::vector<std::unique_ptr<A>> v;
v.emplace_back(make_unique<B>());
// temporary container for some operation
std::vector<std::reference_wrapper<A>> selected;
if(check())
selected.emplace_back(*v.front());
How do you feel about that code? Is it a good practice? I know that std::ref() and std::cref where meant to primarily work with templates, but it seems that here we can also use it to clearly state our design intent. The only problem I see is that I have to dereference std::reference_wrapper with get() and there are no operator*() or operator->() inside to have the same interface like in a container with unique_ptr. Should I write something similar on my own? Or maybe a reference_wrapper could be extended for such use case in future C++ versions? Please share your feedback.
EDIT: I changed the code samples to maybe better show the intent.
You have already provided a solution which looks sound. I understand that the question is "How do you feel?"
My personal feeling is that there need to exist some balance between safety and unambiguity on the one hand and the simplicity of the code on the other. It looks like your solution may be pushing it too hard towards safety and compromising the simplicity too much. Whenever I used containers holding "weak references" I used raw pointers to represent these. True, this might make it less clear who the owner of the object is, but it has some advantages too: you do not have to study what a "reference_wrapper" is, and the code is clear. If you use them (a container of weak references) only temporarily and you encapsulate this usage, the ownership issue should be minimal.
But this is just a question of personal preference, I guess. Let me just propose using different types for the same purpose. This is provided that you can afford to use Boost. For "strong" references (which own the resource) you could use Steve Watanabe's Type Erasure library. It does not require an explicit usage of free-store memory, and i suppose for small types it can get away from using heap-memory altogether (using small-buffer optimization). It has been recently accepted to Boost, although has not been release yet, I think.
For weak references, consider using "optional references" with Boost.Optional:
int i = 0;
boost::optional<int&> oi = i; // note: int&
i = 2;
assert(*oi == 2);
It has same semantics as reference_wrapper.
I think calling them shared_ptr's is not logically wrong. However, looking at the definition of std::weak_ptr:
std::weak_ptr is a smart pointer that holds a non-owning ("weak")
reference to an object that is managed by std::shared_ptr. It must be
converted to std::shared_ptr in order to access the referenced object.
it might be a better candidate. At least when you are fiddling with the pointer through selected you will need to assume temporary ownership. Since the original pointer is stored in a shared pointer, using weak pointer will be safer.
I came accross several questions where answers state that using T* is never the best idea.
While I already make much use of RIIC, there is one particular point in my code, where I use T*. Reading about several auto-pointers, I couldn't find one where I'd say that I have a clear advantage from using it.
My scenario:
class MyClass
{
...
// This map is huge and only used by MyClass and
// and several objects that are only used by MyClass as well.
HashMap<string, Id> _hugeIdMap;
...
void doSomething()
{
MyMapper mapper;
// Here is what I pass. The reason I can't pass a const-ref is
// that the mapper may possibly assign new IDs for keys not yet in the map.
mapper.setIdMap(&_hugeIdMap);
mapper.map(...);
}
}
MyMapper now has a HashMap<...>* member, which - according to highly voted answers in questions on unrelated problems - never is a good idea (Altough the mapper will go out of scope before the instance of MyClass does and hence I do not consider it too much of a problem. There's no new in the mapper and no delete will be needed).
So what is the best alternative in this particular use-case?
Personally I think a raw pointer (or reference) is okay here. Smart pointers are concerned with managing the lifetime of the object pointed to, and in this case MyMapper isn't managing the lifetime of that object, MyClass is. You also shouldn't have a smart pointer pointing to an object that was not dynamically allocated (which the hash map isn't in this case).
Personally, I'd use something like the following:
class MyMapper
{
public:
MyMapper(HashMap<string, Id> &map)
: _map(map)
{
}
private:
HashMap<string, Id> &_map
};
Note that this will prevent MyMapper from having an assignment operator, and it can only work if it's acceptable to pass the HashMap in the constructor; if that is a problem, I'd make the member a pointer (though I'd still pass the argument as a reference, and do _map(&map) in the initializer list).
If it's possible for MyMapper or any other class using the hash map to outlive MyClass, then you'd have to start thinking about smart pointers. In that case, I would probably recommend std::shared_ptr, but you'd have to use it everywhere: _hugeIdMap would have to be a shared_ptr to a dynamically allocated value, not a regular non-pointer field.
Update:
Since you said that using a reference is not acceptable due to the project's coding standards, I would suggest just sticking with a raw pointer for the reasons mentioned above.
Naked pointers (normally referred to as raw pointers) are just fine when the object has no responsibility to delete the object. In the case of MyMapper then the pointer points to an object already owned by MyClass and is therefore absolutely fine to not delete it. The problem arises when you use raw pointers when you do intend for objects to be deleted through them, which is where problems lie. People only ask questions when they have problems, which is why you almost always see it only used in a problematic context, but raw pointers in a non-owning context is fine.
How about passing it into the constructor and keeping a reference (or const-reference) to it? That way your intent of not owning the object is made clear.
Passing auto-pointers or shared-pointers are mostly for communicating ownership.
shared pointers indicate it's shared
auto-pointers indicate it's the receivers responsibility
references indicate it's the senders responsibility
blank pointers indicate nothing.
About your coding style:
our coding standards have a convention that says never pass non-const references.
Whether you use the C++ reference mechanism or the C++ pointer mechanism, you're passing a (English-meaning) reference to the internal storage that will change. I think your coding standard is trying to tell you not to do that at all, not so much that you can't use references to do so but that you can do it in another way.
Do smart pointers handle down casting, and if not what is a safe way of working around this limitation?
An example of what I'm trying to do is having two STL vectors (for example) containing smart pointers. The first contains smart pointers to a base class while the second contains smart pointers to a derived class. The smart pointers are referenced counted, e.g. similar behaviour to Boost's shared_ptrs, but hand-rolled. I've included some sample code that I whipped up to provide an example:
vector<CBaseSmartPtr> vecBase;
vector<CDerivedSmartPtr> vecDer;
...
CBaseSmartPtr first = vecBase.front();
vecDer.push_back(CDerivedSmartPtr(dynamic_cast<CDerived*>(first.get()));
This seems not safe to me, as I think I'm ending up with two smart pointers managing the same object. At some point down the track this is probably going to result in one of them freeing the object while the other still holds references to it.
What I'd hope for but don't think will work is a straight down-cast while keeping the same object, e.g.
dynamic_cast<CDerivedSmartPtr>(first)
Should I be looking to change the second container to also use CBaseSmartPtr and downcast on usage only? Are there other solutions?
Smart pointers can handle downcasting, but it's not automatic. And getting const-correctness in can be a bit complex (I've used our smart pointer implementation in interview questions, there's some template trickery involved). But many users of smart pointers never instantiate their smart pointers with const-qualified types anyway.
The first thing you need to get correct is the counter. Since you may need to share a counter between smart_ptr<Base> and smart_ptr<Derived>, the counter type should not depend on the type argument. In general, this is not a big deal anyway. A counter is merely a size_t, probably wrapped in a class. (Note: there are alternative smart pointer designs, but the question strongly suggests a counter is used)
A cast towards base should be fairly trivial. Hence, your smart_ptr should have a constructor taking a smart_ptr. In this ctor, add a line static_cast<T*>((U*)0);. This doesn't generate code, but prevents instantiation when T is not a base of U (modulo const qualifications).
The other way around should be an explicit cast. You can't programatically enumerate all bases of T, so smart_ptr<T> cannot derive from smart_ptr<Base1_of_T>, smart_ptr<Base2_of_T>, ... Hence, a dynamic_cast<smart_ptr<T> > won't work. You can provide your own smart_dynamic_cast<SPT>(smart_ptr<U> const& pU). This is best implemented as a function returing an SPT. In this function, you can simply do a return SPT(dynamic_cast<SPT::value_type*>(&*pU)).
The property you want is covariance in the pointed-to type. That is, if D isa B, then you want smartptr<D> isa smartptr<B>. I don't think this is elegantly supported at all in C++, but as always, there are template/overload hacks available.
http://www.boost.org/doc/libs/1_39_0/libs/smart_ptr/pointer_cast.html gives a dynamic cast that works on regular and boost::smart_ptr. You should learn from the implementation if you don't want to just use Boost's.
Follow the thread here in one of the boost mailing lists. It shows how one can implement smart-pointer downcasting in case of boost::shared_ptr. HTH
Normal smart pointers, like std::auto_ptr, are not safe to use in STL containers, due to ownership being moved around when the STL assigns instances of smart pointers to each other as it copies data around internally. You need to use something like boost::shared_ptr instead, which internally implements reference counting to ensure an object stays alive no matter how many smart pointer instances refer to it. If you are writing your own smart pointer types, then you need to implement similar reference counting.
I've found this on Microsoft pages:
std::shared_ptr<base> sp0(new derived);
std::shared_ptr<derived> sp1 =
std::dynamic_pointer_cast<derived>(sp0);