I am working on a project with a friend that does a lot of computation and we are using c++ for it. I havent used c++ in a while and he is suggesting some fixes. I hoped I could come here for a more in depth explanation and maybe could be linked to some more articles.
He says its more efficient instead of having this
Hand::Hand(Card one, Card two)
To have this
Hand::Hand(const Card &one, const Card &two)
Is this correct? What about passing a constant address rather than the object itself makes it more efficient? He mentioned passing a reference instead of making a copy. If I dont pass by address, will it construct a new card object as a copy of the one I've passed?
Also
Instead of
bool Hand::hasFourKind(Card board[])
Have this
bool Hand::hasFourKind(const Card *board)
This passes a pointer to the start of the array instead of making an array copy?
In most cases, if you don't need a local variable to be modified, the latter method of the first example is faster, because the entire object will not have to be copied on the stack for the function to use it. Instead, a pointer will be pushed onto the stack, which takes less time. (Although some calling conventions allow passing small arguments in registers, which is even faster.)
For small objects, the time spent in copying may not be an issue, but may be more evident for large classes/structs.
The second examples have identical operation (disregarding the constness), since the array would not be passed by value in any case. It would be passed as a simple pointer, or passed "by reference."
From an optimisation point of view, the compiler can not rely on const actually meaning "won't change". const_cast allows a function to alter the const-ness of something, such that it can be altered. It is useful for the programmer to know that "I get an error if I accidentally modify this" (in particular mistyping if (a = b) instead of if (a == b)).
If the source code of a function is available to the compiler, it can itself prove that a value isn't being changed, or is being changed, regardless of whether you mark it const or not.
In a video with Chandler Carruth (one of the currently most active developers of Clang and LLVM), he actually promotes using non-reference calls for any type that is reasonably small. Often, the compiler will optimise away the copy of the argument anyways, because the reference MAY be modified [so if the compiler doesn't have the source code available, it won't know if the value is being changed or not]
Whether you use Card board[] or Card *board will not change things, the compiler will generate exactly the same code either way - if you want others reading the code to understand if the function is expected to modify the value or not, then add const for values that aren't being changed.
This
Hand::Hand(const Card &one, const Card &two)
is more efficient than this
Hand::Hand(Card one, Card two)
For 2 reasons
it avoids making a copy of the Cards
If you wanted to maintain a Deck of cards, it is possibe with the first method but not possible with the second method for reasons explained by my first point
This
bool Hand::hasFourKind(Card board[])
is equivalent to this
bool Hand::hasFourKind(const Card *board)
because they both represent array of cards, but the first one is just syntatic sugar for pointer to cards, so they both actually represent pointer to cards.
The second is better because it implies that the board cannot be modified by anyone else. This is not really much of a guard because of const_cast which can be used to remove the constness of an object.
If you really wanted to ensure that the cards cannot be modified by anybody in that method, then I would suggest you change design a bit to enable this:
bool Hand::hasFourKind() const;
Where the cards are part of the Hand class and cannot be modified by anyone
Related
I have noticed that most C++ experts always advice it's better to pass by value, due to RVO. This allows me not worry too much about pointer manipulation and is easier to write code as well. No complaints there. This makes me wonder whether it is the correct approach to not use dynamic memory allocation (on the heap) at all and always pass parameters and return results by value?
This means instead of coming up with signatures like this:
Character* getCharacter(Sprite *sprite, Action* action)
I should more or less stick to signatures like:
Character getCharacter(Sprite sprite, Action action)
Is my understanding correct? or did I juth think i thaw a putthy cath?
They each have there pro's and con's. remember that using words like "always" is an absolute. Only the Dark Side deals in absolutes.
So let's look at each way and when we would use them.
Pass by value is good when the object being passed is smaller (since a local copy gets made). It is also good if you want to be sure to not accidentally change the original data. Its shortcoming is it makes a local copy and that can be bad if it is really big.
Pass by reference only passes a memory address. Therefore, large objects can be passed for a relatively low footprint. Also, with a reference, you can modify the original (this is both good and bad). This enables you to "return" more than one variable (so to speak). So obviously, the big con here is that you can mistakenly change the original data.
Constant pass by reference is generally accepted to be a very strong candidate for doing things. It has the pros of both pass by reference and value. Low footprint since it is a reference AND you can't change the original. There aren't many cons accept for the fact that your use of the variable in the method needs to change a little. Remember, its a const and therefore cannot be modified in the function.
Remember, there is no magic-bullet. Nothing is always better. Determine what you need and select the right tool for the job.
EDIT: also, has been said. Passing is not the same as dynamic allocation. dynamic allocation only happens with the "new" keyword. My suggestion would be to avoid the "new" keyword for now until you have a better understanding of arguments and pointers.
Whether or not you allocate an object on the heap typically is driven by one of the following concerns:
If the new object needs to outlive the function that creates it, the object must be allocated on the heap.
If the object is very large, and does not fit on the stack, then you must allocate it on the heap.
Beyond that, the choice of pass by value or pass by reference is determined by the semantics. If you want to operate on a copy, pass by value. If you want to operate on the actual object, pass by reference.
Your statement is simply utterly untrue. There is some light advice to pass by value instead of the mainstream const-ref in the special case where the function will copy the argument to a local variable anyway.
And for passing by-nonconst-pointer, pass by value was never an alternative. The first implies an optional out or inout param and the second and input param.
And mentioned dynamic allocation in question title just fits no way with the content.
Your understanding in definitely not correct.
Having
struct Person {
string name;
};
Person* p = ...
Assume that no operators are overloaded.
Which is more efficient (if any) ?
(*p).name vs. p->name
Somewhere in the back of my head I hear some bells ringing, that the * dereference operator may create a temporary copy of an object; is this true?
The background of this question are cases like this:
Person& Person::someFunction(){
...
return *this;
}
and I began to wonder, if changing the result to Person* and the last line to simply return this would make any difference (in performance)?
There's no difference. Even the standard says the two are equivalent, and if there's any compiler out there that doesn't generate the same binary for both versions, it's a bad one.
When you return a reference, that's exactly the same as passing back a pointer, pointer semantics excluded.
You pass back a sizeof(void*) element, not a sizeof(yourClass).
So when you do that:
Person& Person::someFunction(){
...
return *this;
}
You return a reference, and that reference has the same intrinsic size than a pointer, so there's no runtime difference.
Same goes for your use of (*i).name, but in that case you create an l-value, which has then the same semantics as a reference (see also here)
Yes, it's much harder to read and type, so you are much better off using the x->y than (*x).y - but other than typing efficiency, there is absolutely no difference. The compiler still needs to read the value of x and then add the offset to y, whether you use one form or the other [assuming there are no funny objects/classes involved that override the operator-> and operator* respectively, of course]
There is definitely no extra object created when (*x) is referenced. The value of the pointer is loaded into a register in the processor [1]. That's it.
Returning a reference is typically more efficient, as it returns a pointer (in disguise) to the object, rather than making a copy of the object. For objects that are bigger than the size of a pointer, this is typically a win.
[1] Yes, we can have a C++ compiler for a processor that doesn't have registers. I know of at least one processor from Rank-Xerox that I saw in about 1984, which doesn't have registers, it was a dedicated LiSP processor, and it just has a stack for LiSP objects... But they are far from common in todays world. If someone working on a processor that doesn't have registers, please don't downvote my answer simply because I don't cover that option. I'm trying to keep the answer simple.
Any good compiler will produce the same results. You can answer this yourself, compile both codes to assembler and check the produced code.
If some of you haven't noticed already, I'm a noob. With that said, here is my question:
Do any of you experienced programmers use reference variables to decrease the memory required of your programs? I was thinking that, while probably a dangerous practice, you could use reference variables for mobile applications to make them use less memory and make them faster.
I know that in C++ when you pass a variable, as an argument, to a function that it creates a copy of that variable, but you can use the & to make it a reference variable which just points to the variables memory location. Wouldn't that make your program use less memory overall and make things faster?
For big things, like structs and objects a reference uses less memory. However most people already pass these by reference anyway so it doesn't matter for our discussion.
Smaller things like ints and chars are the same size or smaller then a reference. There is no memory gain by passing them by reference...
... but there is a performance penalty, since referrences need to be dereferenced in order to manipulate the value.
Finally, pass by reference is more prone to bugs then pass-by-value. Programs should be build for correctness first and performance second.
This depends on (a) underlying architecture, (b) framework and (c) language you use, but the general answer is no - this is not the best (or even common) optimization practice and yes - programs may run slower (much slower!) using your approach.
In C variables are passed by value by default and there is no advantage in most cases in passing by reference as you are still passing a value (It's just the pointer rather than the value it refers to). Keep in mind that a pointer to a byte will be bigger than the byte its self!
With more complex types such as arrays and structs C will pass them by reference by default. It is inefficient to create a copy of a struct to pass into a function unless the function needs its own copy of the structure for some reason.
If you want to pass by reference, but you are worried about the function changing your struct you can use the const keyword to ensure it cannot be changed (or at least not easily).
Lets say I know a guy who is new to C++. He does not pass around pointers (rightly so) but he refuses to pass by reference. He uses pass by value always. Reason being that he feels that "passing objects by reference is a sign of a broken design".
The program is a small graphics program and most of the passing in question is mathematical Vector(3-tuple) objects. There are some big controller objects but nothing more complicated than that.
I'm finding it hard to find a killer argument against only using the stack.
I would argue that pass by value is fine for small objects such as vectors but even then there is a lot of unnecessary copying occurring in the code. Passing large objects by value is obviously wasteful and most likely not what you want functionally.
On the pro side, I believe the stack is faster at allocating/deallocating memory and has a constant allocation time.
The only major argument I can think of is that the stack could possibly overflow, but I'm guessing that it is improbable that this will occur? Are there any other arguments against using only the stack/pass by value as opposed to pass by reference?
Subtyping-polymorphism is a case where passing by value wouldn't work because you would slice the derived class to its base class. Maybe to some, using subtyping-polymorphism is bad design?
Your friend's problem is not his idea as much as his religion. Given any function, always consider the pros and cons of passing by value, reference, const reference, pointer or smart pointer. Then decide.
The only sign of broken design I see here is your friend's blind religion.
That said, there are a few signatures that don't bring much to the table. Taking a const by value might be silly, because if you promise not to change the object then you might as well not make your own copy of it. Unless its a primitive, of course, in which case the compiler can be smart enough to take a reference still. Or, sometimes it's clumsy to take a pointer to a pointer as argument. This adds complexity; instead, you might be able to get away with it by taking a reference to a pointer, and get the same effect.
But don't take these guidelines as set in stone; always consider your options because there is no formal proof that eliminates any alternative's usefulness.
If you need to change the argument for your own needs, but don't want to affect the client, then take the argument by value.
If you want to provide a service to the client, and the client is not closely related to the service, then consider taking an argument by reference.
If the client is closely related to the service then consider taking no arguments but write a member function.
If you wish to write a service function for a family of clients that are closely related to the service but very distinct from each other then consider taking a reference argument, and perhaps make the function a friend of the clients that need this friendship.
If you don't need to change the client at all then consider taking a const-reference.
There are all sorts of things that cannot be done without using references - starting with a copy constructor. References (or pointers) are fundamental and whether he likes it or not, he is using references. (One advantage, or maybe disadvantage, of references is that you do not have to alter the code, in general, to pass a (const) reference.) And there is no reason not to use references most of the time.
And yes, passing by value is OK for smallish objects without requirements for dynamic allocation, but it is still silly to hobble oneself by saying "no references" without concrete measurements that the so-called overhead is (a) perceptible and (b) significant. "Premature optimization is the root of all evil"1.
1
Various attributions, including C A Hoare (although apparently he disclaims it).
I think there is a huge misunderstanding in the question itself.
There is not relationship between stack or heap allocated objects on the one hand and pass by value or reference or pointer on the other.
Stack vs Heap allocation
Always prefer stack when possible, the object's lifetime is then managed for you which is much easier to deal with.
It might not be possible in a couple of situations though:
Virtual construction (think of a Factory)
Shared Ownership (though you should always try to avoid it)
And I might miss some, but in this case you should use SBRM (Scope Bound Resources Management) to leverage the stack lifetime management abilities, for example by using smart pointers.
Pass by: value, reference, pointer
First of all, there is a difference of semantics:
value, const reference: the passed object will not be modified by the method
reference: the passed object might be modified by the method
pointer/const pointer: same as reference (for the behavior), but might be null
Note that some languages (the functional kind like Haskell) do not offer reference/pointer by default. The values are immutable once created. Apart from some work-arounds for dealing with the exterior environment, they are not that restricted by this use and it somehow makes debugging easier.
Your friend should learn that there is absolutely nothing wrong with pass-by-reference or pass-by-pointer: for example thing of swap, it cannot be implemented with pass-by-value.
Finally, Polymorphism does not allow pass-by-value semantics.
Now, let's speak about performances.
It's usually well accepted that built-ins should be passed by value (to avoid an indirection) and user-defined big classes should be passed by reference/pointer (to avoid copying). big in fact generally means that the Copy Constructor is not trivial.
There is however an open question regarding small user-defined classes. Some articles published recently suggest that in some case pass-by-value might allow better optimization from the compiler, for example, in this case:
Object foo(Object d) { d.bar(); return d; }
int main(int argc, char* argv[])
{
Object o;
o = foo(o);
return 0;
}
Here a smart compiler is able to determine that o can be modified in place without any copying! (It is necessary that the function definition be visible I think, I don't know if Link-Time Optimization would figure it out)
Therefore, there is only one possibility to the performance issue, like always: measure.
Reason being that he feels that "passing objects by reference is a sign of a broken design".
Although this is wrong in C++ for purely technical reasons, always using pass-by-value is a good enough approximation for beginners – it’s certainly much better than passing everything by pointers (or perhaps even than passing everything by reference). It will make some code inefficient but, hey! As long as this doesn’t bother your friend, don’t be unduly disturbed by this practice. Just remind him that someday he might want to reconsider.
On the other hand, this:
There are some big controller objects but nothing more complicated than that.
is a problem. Your friend is talking about broken design, and then all the code uses are a few 3D vectors and large control structures? That is a broken design. Good code achieves modularity through the use of data structures. It doesn’t seem as though this were the case.
… And once you use such data structures, code without pass-by-reference may indeed become quite inefficient.
First thing is, stack rarely overflows outside this website, except in the recursion case.
About his reasoning, I think he might be wrong because he is too generalized, but what he has done might be correct... or not?
For example, the Windows Forms library use Rectangle struct that have 4 members, the Apple's QuartzCore also has CGRect struct, and those structs always passed by value. I think we can compare that to Vector with 3 floating-point variable.
However, as I do not see the code, I feel I should not judge what he has done, though I have a feeling he might did the right thing despite of his over generalized idea.
I would argue that pass by value is fine for small objects such as vectors but even then there is a lot of unnecessary copying occurring in the code. Passing large objects by value is obviously wasteful and most likely not what you want functionally.
It's not quite as obvious as you might think. C++ compilers perform copy elision very aggressively, so you can often pass by value without incurring the cost of a copy operation. And in some cases, passing by value might even be faster.
Before condemning the issue for performance reasons, you should at the very least produce the benchmarks to back it up. And they might be hard to create because the compiler typically eliminates the performance difference.
So the real issue should be one of semantics. How do you want your code to behave? Sometimes, reference semantics are what you want, and then you should pass by reference. If you specifically want/need value semantics then you pass by value.
There is one point in favor of passing by value. It's helpful in achieving a more functional style of code, with fewer side effects and where immutability is the default. That makes a lot of code easier to reason about, and it may make it easier to parallelize the code as well.
But in truth, both have their place. And never using pass-by-reference is definitely a big warning sign.
For the last 6 months or so, I've been experimenting with making pass-by-value the default. If I don't explicitly need reference semantics, then I try to assume that the compiler will perform copy elision for me, so I can pass by value without losing any efficiency.
So far, the compiler hasn't really let me down. I'm sure I'll run into cases where I have to go back and change some calls to passing by reference, but I'll do that when I know that
performance is a problem, and
the compiler failed to apply copy elision
I would say that Not using pointers in C is a sign of a newbie programmer.
It sounds like your friend is scared of pointers.
Remember, C++ pointers were actually inherited from the C language, and C was developed when computers were much less powerful. Nevertheless, speed and efficiency continue to be vital until this day.
So, why use pointers? They allow the developer to optimize a program to run faster or use less memory that it would otherwise! Referring to the memory location of a data is much more efficient then copying all the data around.
Pointers usually are a concept that is difficult to grasp for those beginning to program, because all the experiments done involve small arrays, maybe a few structs, but basically they consist of working with a couple of megabytes (if you're lucky) when you have 1GB of memory laying around the house. In this scene, a couple of MB are nothing and it usually is too little to have a significant impact on the performance of your program.
So let's exaggerate that a little bit. Think of a char array with 2147483648 elements - 2GB of data - that you need to pass to function that will write all the data to the disk. Now, what technique do you think is going to be more efficient/faster?
Pass by value, which is going to have to re-copy those 2GB of data to another location in memory before the program can write the data to the disk, or
Pass by reference, which will just refer to that memory location.
What happens when you just don't have 4GB of RAM? Will you spend $ and buy chips of RAM just because you are afraid of using pointers?
Re-copying the data in memory sounds a bit redundant when you don't have to, and its a waste of computer resource.
Anyway, be patient with your friend. If he would like to become a serious/professional programmer at some point in his life he will eventually have to take the time to really understand pointers.
Good Luck.
As already mentioned the big difference between a reference and a pointer is that a pointer can be null. If a class requires data a reference declaration will make it required. Adding const will make it 'read only' if that is what is desired by the caller.
The pass-by-value 'flaw' mentioned is simply not true. Passing everything by value will completely change the performance of an application. It is not so bad when primitive types (i.e. int, double, etc.) are passed by value but when a class instance is passed by value temporary objects are created which requires constructors and later on destructor's to be called on the class and on all of the member variable in the class. This is exasperated when large class hierarchies are used because parent class constructors/destructor's must be called as well.
Also, just because the vector is passed by value does not mean that it only uses stack memory. heap may be used for each element as it is created in the temporary vector that is passed to the method/function. The vector itself may also have to reallocate via heap if it reaches its capacity.
If pass by value is being so that the callers values are not modified then just use a const reference.
The answers that I've seen so far have all focused on performance: cases where pass-by-reference is faster than pass-by-value. You may have more success in your argument if you focus on cases that are impossible with pass-by-value.
Small tuples or vectors are a very simple type of data-structure. More complex data-structures share information, and that sharing can't be represented directly as values. You either need to use references/pointers or something that simulates them such as arrays and indices.
Lots of problems boil down to data that forms a Graph, or a Directed-Graph. In both cases you have a mixture of edges and nodes that need to be stored within the data-structure. Now you have the problem that the same data needs to be in multiple places. If you avoid references then firstly the data needs to be duplicated, and then every change needs to be carefully replicated in each of the other copies.
Your friend's argument boils down to saying: tackling any problem complex enough to be represented by a Graph is a bad-design....
The only major argument I can think of
is that the stack could possibly
overflow, but I'm guessing that it is
improbable that this will occur? Are
there any other arguments against
using only the stack/pass by value as
opposed to pass by reference?
Well, gosh, where to start...
As you mention, "there is a lot of unnecessary copying occurring in the code". Let's say you've got a loop where you call a function on these objects. Using a pointer instead of duplicating the objects can accelerate execution by one or more orders of magnitude.
You can't pass a variable-sized data structures, arrays, etc. around on the stack. You have to dynamically allocate it and pass a pointers or reference to the beginning. If your friend hasn't run into this, then yes, he's "new to C++."
As you mention, the program in question is simple and mostly uses quite small objects like graphics 3-tuples, which if the elements are doubles would be 24 bytes apiece. But in graphics, it's common to deal with 4x4 arrays, which handle both rotation and translation. Those would be 128 bytes apiece, so if a program that had to deal with those would be five times slower per function call with pass-by-value due to the increased copying. With pass-by-reference, passing a 3-tuple or a 4x4 array in a 32-bit executable would just involve duplicating a single 4-byte pointer.
On register-rich CPU architecures like ARM, PowerPC, 64-bit x86, 680x0 - but not 32-bit x86 - pointers (and references, which are secretly pointers wearing fancy syntatical clothing) are commonly be passed or returned in a register, which is really freaking fast compared to the memory access involved in a stack operation.
You mention the improbability of running out of stack space. And yes, that's so on a small program one might write for a class assignment. But a couple of months ago, I was debugging commercial code that was probably 80 function calls below main(). If they'd used pass-by-value instead of pass-by-reference, the stack would have been ginormous. And lest your friend think this was a "broken design", this was actually a WebKit-based browser implemented on Linux using GTK+, all of which is very state-of-the-art, and the function call depth is normal for professional code.
Some executable architectures limit the size of an individual stack frame, so even though you might not run out of stack space per se, you could exceed that and wind up with perfectly valid C++ code that wouldn't build on such a platform.
I could go on and on.
If your friend is interested in graphics, he should take a look at some of the common APIs used in graphics: OpenGL and XWindows on Linux, Quartz on Mac OS X, Direct X on Windows. And he should look at the internals of large C/C++ systems like the WebKit or Gecko HTML rendering engines, or any of the Mozilla browsers, or the GTK+ or Qt GUI toolkits. They all pass by anything much larger than a single integer or float by reference, and often fill in results by reference rather than as a function return value.
Nobody with any serious real world C/C++ chops - and I mean nobody - passes data structures by value. There's a reason for this: it's just flipping inefficient and problem-prone.
Wow, there are already 13 answers… I didn't read all in detail but I think this is quite different from the others…
He has a point. The advantage of pass-by-value as a rule is that subroutines cannot subtly modify their arguments. Passing non-const references would indicate that every function has ugly side effects, indicating poor design.
Simply explain to him the difference between vector3 & and vector3 const&, and demonstrate how the latter may be initialized by a constant as in vec_function( vector3(1,2,3) );, but not the former. Pass by const reference is a simple optimization of pass by value.
Buy your friend a good c++ book. Passing non-trivial objects by reference is a good practice and saves you a lot of unneccessary constructor/destructor calls. This has also nothing to do with allocating on free store vs. using stack. You can (or should) pass objects allocated on program stack by reference without any free store usage. You also can ignore free store completely, but that throws you back to the old fortran days which your friend probably hadn't in mind - otherwise he would pick an ancient f77 compiler for your project, wouldn't he...?
I've seen numerous arguments that using a return value is preferable to out parameters. I am convinced of the reasons why to avoid them, but I find myself unsure if I'm running into cases where it is unavoidable.
Part One of my question is: What are some of your favorite/common ways of getting around using an out parameter? Stuff along the lines: Man, in peer reviews I always see other programmers do this when they could have easily done it this way.
Part Two of my question deals with some specific cases I've encountered where I would like to avoid an out parameter but cannot think of a clean way to do so.
Example 1:
I have a class with an expensive copy that I would like to avoid. Work can be done on the object and this builds up the object to be expensive to copy. The work to build up the data is not exactly trivial either. Currently, I will pass this object into a function that will modify the state of the object. This to me is preferable to new'ing the object internal to the worker function and returning it back, as it allows me to keep things on the stack.
class ExpensiveCopy //Defines some interface I can't change.
{
public:
ExpensiveCopy(const ExpensiveCopy toCopy){ /*Ouch! This hurts.*/ };
ExpensiveCopy& operator=(const ExpensiveCopy& toCopy){/*Ouch! This hurts.*/};
void addToData(SomeData);
SomeData getData();
}
class B
{
public:
static void doWork(ExpensiveCopy& ec_out, int someParam);
//or
// Your Function Here.
}
Using my function, I get calling code like this:
const int SOME_PARAM = 5;
ExpensiveCopy toModify;
B::doWork(toModify, SOME_PARAM);
I'd like to have something like this:
ExpensiveCopy theResult = B::doWork(SOME_PARAM);
But I don't know if this is possible.
Second Example:
I have an array of objects. The objects in the array are a complex type, and I need to do work on each element, work that I'd like to keep separated from the main loop that accesses each element. The code currently looks like this:
std::vector<ComplexType> theCollection;
for(int index = 0; index < theCollection.size(); ++index)
{
doWork(theCollection[index]);
}
void doWork(ComplexType& ct_out)
{
//Do work on the individual element.
}
Any suggestions on how to deal with some of these situations? I work primarily in C++, but I'm interested to see if other languages facilitate an easier setup. I have encountered RVO as a possible solution, but I need to read up more on it and it sounds like a compiler specific feature.
I'm not sure why you're trying to avoid passing references here. It's pretty much these situations that pass-by-reference semantics exist.
The code
static void doWork(ExpensiveCopy& ec_out, int someParam);
looks perfectly fine to me.
If you really want to modify it then you've got a couple of options
Move doWork so that's it's a member of ExpensiveCopy (which you say you can't do, so that's out)
return a (smart) pointer from doWork instead of copying it. (which you don't want to do as you want to keep things on the stack)
Rely on RVO (which others have pointed out is supported by pretty much all modern compilers)
Every useful compiler does RVO (return value optimization) if optimizations are enabled, thus the following effectively doesn't result in copying:
Expensive work() {
// ... no branched returns here
return Expensive(foo);
}
Expensive e = work();
In some cases compilers can apply NRVO, named return value optimization, as well:
Expensive work() {
Expensive e; // named object
// ... no branched returns here
return e; // return named object
}
This however isn't exactly reliable, only works in more trivial cases and would have to be tested. If you're not up to testing every case, just use out-parameters with references in the second case.
IMO the first thing you should ask yourself is whether copying ExpensiveCopy really is so prohibitive expensive. And to answer that, you will usually need a profiler. Unless a profiler tells you that the copying really is a bottleneck, simply write the code that's easier to read: ExpensiveCopy obj = doWork(param);.
Of course, there are indeed cases where objects cannot be copied for performance or other reasons. Then Neil's answer applies.
In addition to all comments here I'd mention that in C++0x you'd rarely use output parameter for optimization purpose -- because of Move Constructors (see here)
Unless you are going down the "everything is immutable" route, which doesn't sit too well with C++. you cannot easily avoid out parameters. The C++ Standard Library uses them, and what's good enough for it is good enough for me.
As to your first example: return value optimization will often allow the returned object to be created directly in-place, instead of having to copy the object around. All modern compilers do this.
What platform are you working on?
The reason I ask is that many people have suggested Return Value Optimization, which is a very handy compiler optimization present in almost every compiler. Additionally Microsoft and Intel implement what they call Named Return Value Optimization which is even more handy.
In standard Return Value Optimization your return statement is a call to an object's constructor, which tells the compiler to eliminate the temporary values (not necessarily the copy operation).
In Named Return Value Optimization you can return a value by its name and the compiler will do the same thing. The advantage to NRVO is that you can do more complex operations on the created value (like calling functions on it) before returning it.
While neither of these really eliminate an expensive copy if your returned data is very large, they do help.
In terms of avoiding the copy the only real way to do that is with pointers or references because your function needs to be modifying the data in the place you want it to end up in. That means you probably want to have a pass-by-reference parameter.
Also I figure I should point out that pass-by-reference is very common in high-performance code for specifically this reason. Copying data can be incredibly expensive, and it is often something people overlook when optimizing their code.
As far as I can see, the reasons to prefer return values to out parameters are that it's clearer, and it works with pure functional programming (you can get some nice guarantees if a function depends only on input parameters, returns a value, and has no side effects). The first reason is stylistic, and in my opinion not all that important. The second isn't a good fit with C++. Therefore, I wouldn't try to distort anything to avoid out parameters.
The simple fact is that some functions have to return multiple things, and in most languages this suggests out parameters. Common Lisp has multiple-value-bind and multiple-value-return, in which a list of symbols is provided by the bind and a list of values is returned. In some cases, a function can return a composite value, such as a list of values which will then get deconstructed, and it isn't a big deal for a C++ function to return a std::pair. Returning more than two values this way in C++ gets awkward. It's always possible to define a struct, but defining and creating it will often be messier than out parameters.
In some cases, the return value gets overloaded. In C, getchar() returns an int, with the idea being that there are more int values than char (true in all implementations I know of, false in some I can easily imagine), so one of the values can be used to denote end-of-file. atoi() returns an integer, either the integer represented by the string it's passed or zero if there is none, so it returns the same thing for "0" and "frog". (If you want to know whether there was an int value or not, use strtol(), which does have an out parameter.)
There's always the technique of throwing an exception in case of an error, but not all multiple return values are errors, and not all errors are exceptional.
So, overloaded return values causes problems, multiple value returns aren't easy to use in all languages, and single returns don't always exist. Throwing an exception is often inappropriate. Using out parameters is very often the cleanest solution.
Ask yourself why you have some method that performs work on this expensive to copy object in the first place. Say you have a tree, would you send the tree off into some building method or else give the tree its own building method? Situations like this come up constantly when you have a little bit off design but tend to fold into themselves when you have it down pat.
I know in practicality we don't always get to change every object at all, but passing in out parameters is a side effect operation, and it makes it much harder to figure out what's going on, and you never really have to do it (except as forced by working within others' code frameworks).
Sometimes it is easier, but it's definitely not desirable to use it for no reason (if you've suffered through a few large projects where there's always half a dozen out parameters you'll know what I mean).