Zero Lee的专栏

The Return Value Optimization[1]

Resource from the book:
Dov Bulka & David Mayhew <Efficient C++--- Performance programming Techiques>

Anytime you can skip the creation and destruction of an object, you are looking at a performance gain. We will discuss an optimization often performed by compilers to speed  up your source code by transforming it and eliminating object creation. This optimization is referred to as the Return Value Optimization(RVO). Prior to delving into the RVO we need to understand how return-by-value works. We will walk through it with a simple example.

The Mechanics of Return-by-Value
The Complex class implements a representation for complex numbers:

 1 class  Complex
 2
{
 3     //  Complex addition operator

 4     friend Complex  operator   + ( const  Complex & const  Complex & );
 5 public
:
 6     //
 default constructor
 7     //
 Value defaults to 0 unless otherwise specified
 8     Complex( double  r  =   0.0 double  i  =   0.0 ):real(r), imag(i)  { }

 9
10     //  copy constructor

11     Complex( const  Complex &  c):real(c.real), imag(c.imag)  { }
12
13     //  Assignment operator

14     Complex &   operator   = ( const  Complex &  c);
15

16     ~ Complex()  { }

17 private :
18     double
 real;
19 }
;
The addition operator returns a Complex object by value, as in:
1 Complex  operator   + ( const  Complex &  a,  const  Complex &  b)
2
{
3
    Complex retVal;
4     retVal.real  =  a.real  +
 b.real;
5     retVal.imag  =  a.imag  +
 b.imag;
6      return
 retVal;
7 }
Suppose c1, c2, and c3 are Complex and we excute
c3 = c1 + c2;
How do we get the value of c1 + c2 into c3? One popular technique used by compilers is to create a temporary __result object and pass it into Complex::operator +() as a third argument. It is passed by referece. So the compiler rewrites
1 Complex &  Complex:: operator   + ( const  Complex &  c1,  const  Complex &  c2)
2
{
3
  
4 }
into a slightly different function:
1 void  Complex_Add( const  Complex &  __result,  const  Complex &  c1,  const  Complex &  c2)
2
{
3
   
4 }
Now the original source statement
c3 = c1 + c2;
is transformed into(pseudocode):
1 struct  Complex __tempResult;  //  Storage. No constructor here.
2 Complex_Add(__tempResult, c1, c2);  //  All argument passed by reference.
3 c3  =  __tempResult;  //  Feed result back into left-hand-side.
This return-by-value implementation opens up an optimization opportunity by eliminating the local object RetVal(inside operator +()) and computing the return value directly into the __tempResult temporary object. This is the Return Value Optimization.

posted on 2006-11-13 19:01 Zero Lee 阅读(257) 评论(0)  编辑 收藏 引用 所属分类: C++ Performance


只有注册用户登录后才能发表评论。
网站导航: 博客园   IT新闻   BlogJava   知识库   博问   管理