提高C++性能的编程技术笔记：引用计数+测试代码

引用计数(reference counting)：基本思想是将销毁对象的职责从客户端代码转移到对象本身。对象跟踪记录自身当前被引用的数目，在引用计数达到零时自行销毁。换句话说，对象不再被使用时自行销毁。

引用计数和执行速度之间的关系是与上下文紧密关联的。该关系取决于以下几个因素：

(1). 目标对象的资源消耗量集中于哪些方面？如果目标对象使用过多内存，比如未保护内存，将使可用内存受限，并导致显著的性能损失，表现为缓存命不中和页面错误。

(2). 分配(释放)目标对象所使用资源的代价有多高？即便从堆中分配1字节空间也需要执行数百条指令。释放时亦然。

(3). 可能有多少对象共享目标对象的单个实例？通过使用赋值操作符和拷贝构造函数可以提升共享率。

(4). 创建(销毁)第一个(最后一个)目标对象引用的额度有多高？使用构造函数而不是拷贝构造函数创建新的引用计数对象时，会产生对目标对象的初次引用。这种操作代价高昂。与创建一个新的目标对象相比，这种方法包含更多开销。移除最后一次引用时也是如此。

对于预先存在且无法修改的类，引用计数依然可以实现，只不过，这种情况需要对设计进行一些修改。也可以在多线程执行环境下处理引用计数。在这种情况下，多个线程可能会并发访问同一个引用计数对象。因此必须对维持引用计数的变量进行保护，使其对进行的更新是原子操作。原子更新操作要求有一个锁定机制。

引用计数在性能上并非无往不胜。引用计数、执行时间和资源维护会产生微妙的相互作用，如果对性能方面的考虑很重要，就必须对这几个方面仔细进行评估。引用计数是提升还是损害性能取决于其使用方式。下面的任意一种情况都可以使引用计数变得更为有效：目标对象是很大的资源消费者；资源分配和释放的代价很高；高度共享，由于使用赋值操作符和拷贝构造函数，引用计数的性能可能会很高；创建和销毁引用的代价低廉。反之，则应跳出引用计数而转为使用更加有效的简单非计数对象。

以下是测试代码(reference_counting.cpp)：

#include "reference_counting.hpp"#include <iostream>#include <chrono>#include <string>#include <string.h>namespace reference_counting_ {// reference: 《提高C++性能的编程技术》：第十二章：引用计数namespace {// 为使引用计数更加方便，我们需要为每个BigInt对象关联一个引用计数。// 实现这一操作的方式有两种，为BigInt直接添加refCount成员或使其继承自基类RCObject// RCObject是引用计数对象的基类，并且封装了所有引用计数变量以及针对这些变量的操作class RCObject {public:void addReference() { ++refCount; }void removeReference() { if (--refCount == 0) delete this; }void markUnshareable() { shareable = false; }bool isShareable() const { return shareable; }bool isShared() const { return refCount > 1; }protected:RCObject() : refCount(0), shareable(true) {}RCObject(const RCObject& rhs) : refCount(0), shareable(true) {}RCObject& operator=(const RCObject& rhs) { return *this; }virtual ~RCObject() {}private:int refCount;bool shareable;};// BigInt类的功能是将正整数表示成用二进制编码的十进制形式class BigInt : public RCObject {friend BigInt operator+ (const BigInt&, const BigInt&);public:BigInt(const char*);BigInt(unsigned = 0);BigInt(const BigInt&);BigInt& operator= (const BigInt&);BigInt& operator+= (const BigInt&);~BigInt();char* getDigits() const { return digits; }unsigned getNdigits() const { return ndigits; }void print() { fprintf(stdout, "digits: %s\n", digits); }private:char* digits;unsigned ndigits;unsigned size; // 分配的字符串的大小BigInt(const BigInt&, const BigInt&);char fetch(unsigned i) const;};BigInt::BigInt(unsigned u){unsigned v = u;for (ndigits = 1; (v/=10) > 0; ++ndigits) {;}digits = new char[size=ndigits];for (unsigned i = 0; i < ndigits; ++i) {digits[i] = u%10;u /= 10;}}BigInt::~BigInt(){delete [] digits;}BigInt& BigInt::operator=(const BigInt& rhs){if (this == &rhs) return *this;ndigits = rhs.ndigits;if (ndigits > size) {delete [] digits;digits = new char[size = ndigits];}for (unsigned i = 0; i < ndigits; ++i) {digits[i] = rhs.digits[i];}return *this;}BigInt::BigInt(const char* s){if (s[0] == '\0') {s = "0";}size = ndigits = strlen(s);digits = new char[size];for (unsigned i = 0; i < ndigits; ++i) {digits[i] = s[ndigits-1-i] - '0';}}BigInt::BigInt(const BigInt& copyFrom){size = ndigits = copyFrom.ndigits;digits = new char[size];for (unsigned i = 0; i < ndigits; ++i) {digits[i] = copyFrom.digits[i];}}// BigInt = left + rightBigInt::BigInt(const BigInt& left, const BigInt& right){size = 1 + (left.ndigits > right.ndigits ? left.ndigits : right.ndigits);digits = new char[size];ndigits = left.ndigits;for (unsigned i = 0; i < ndigits; ++i) {digits[i] = left.digits[i];}*this += right;}inline char BigInt::fetch(unsigned i) const{return i < ndigits ? digits[i] : 0;}BigInt& BigInt::operator+=(const BigInt& rhs){unsigned max = 1 + (rhs.ndigits > ndigits ? rhs.ndigits : ndigits);if (size < max) {char* d = new char[size=max];for (unsigned i = 0; i < ndigits; ++i) {d[i] = digits[i];}delete [] digits;digits = d;}while (ndigits < max) {digits[ndigits++] = 0;}for (unsigned i = 0; i < ndigits; ++i) {digits[i] += rhs.fetch(i);if (digits[i] >= 10) {digits[i] -= 10;digits[i+1] = 1;}}if (digits[ndigits-1] == 0) {--ndigits;}return *this;}std::ostream& operator<<(std::ostream& os, BigInt& bi){char c;const char* d = bi.getDigits();for (int i = bi.getNdigits() - 1; i >= 0; i--) {c = d[i] + '0';os << c;}os << std::endl;return os;}inline BigInt operator+(const BigInt& left, const BigInt& right){return BigInt(left, right);}// 智能指针：这种特殊指针的任务是对引用计数进行登记template<class T>class RCPtr {public:RCPtr(T* realPtr = 0) : pointee(realPtr) { init(); }RCPtr(const RCPtr& rhs) : pointee(rhs.pointee) { init(); }~RCPtr() { if (pointee) pointee->removeReference(); }RCPtr& operator=(const RCPtr& rhs);T* operator->() const { return pointee; }T& operator*() const { return *pointee; }private:T* pointee;void init();};template<class T>void RCPtr<T>::init(){if (0 == pointee) return;if (false == pointee->isShareable()) {pointee = new T(*pointee);}pointee->addReference();}template<class T>RCPtr<T>& RCPtr<T>::operator=(const RCPtr& rhs){if (pointee != rhs.pointee) {if (pointee) pointee->removeReference();pointee = rhs.pointee;init();}return *this;}// RCBigInt:构造BigInt的引用计数实现，RCBigInt需要指向一个真正的BigInt对象// 如果是频繁赋值和复制RCBigInt对象的工作，RCBigInt会大显身手。另外，与直接使用简单BigInt相比，// 由于创建了对新BigInt对象的初次引用，使用新RCBigInt对象的代价要更高一些。每当RCBigInt产生了// 对BigInt的初次引用时，就会在堆内存中创建一个BigInt对象并指向它。对于基于堆栈(局部变量)的简单// BigInt对象而言，则不必付出这种代价。此类的情况在移除最后一次BigInt引用时也会发生。这是因为// 底层的BigInt对象被释放并还给堆内存。class RCBigInt {friend RCBigInt operator+(const RCBigInt&, const RCBigInt&);public:RCBigInt(const char* p) : value(new BigInt(p)) {}RCBigInt(unsigned u = 0) : value(new BigInt(u)) {}RCBigInt(const BigInt& bi) : value(new BigInt(bi)) {}void print() const { value->print(); }private:RCPtr<BigInt> value;};inline RCBigInt operator+(const RCBigInt& left, const RCBigInt& right){return RCBigInt(*(left.value) + *(right.value));}} // namespaceint test_reference_counting_1(){std::chrono::high_resolution_clock::time_point time_start, time_end;const int count{10000000};{ // test BigInt createtime_start = std::chrono::high_resolution_clock::now();for (int i = 0; i < count; ++i) {BigInt a = i;BigInt b = i + 1;BigInt c = i + 2;BigInt d = i + 3;}time_end = std::chrono::high_resolution_clock::now();fprintf(stdout, "BigInt create time spend: %f seconds\n",(std::chrono::duration_cast<std::chrono::duration<double>>(time_end-time_start)).count());}{ // test RCBigInt create// RCBigInt测试会更多地忙于初次引用的创建及之后将其销毁的工作time_start = std::chrono::high_resolution_clock::now();for (int i = 0; i < count; ++i) {RCBigInt a = i;RCBigInt b = i + 1;RCBigInt c = i + 2;RCBigInt d = i + 3;}time_end = std::chrono::high_resolution_clock::now();fprintf(stdout, "RCBigInt create time spend: %f seconds\n",(std::chrono::duration_cast<std::chrono::duration<double>>(time_end-time_start)).count());}{ // test BigInt assignBigInt a, b, c;BigInt d = 1;time_start = std::chrono::high_resolution_clock::now();for (int i = 0; i < count; ++i) {a = b = c = d;}time_end = std::chrono::high_resolution_clock::now();fprintf(stdout, "BigInt assign time spend: %f seconds\n",(std::chrono::duration_cast<std::chrono::duration<double>>(time_end-time_start)).count());}{ // test RCBigInt assign// 对RCBigInt对象的赋值操作效率高于BigInt对象，但是创建和销毁对象时却相反RCBigInt a, b, c;RCBigInt d = 1;time_start = std::chrono::high_resolution_clock::now();for (int i = 0; i < count; ++i) {a = b = c = d;}time_end = std::chrono::high_resolution_clock::now();fprintf(stdout, "RCBigInt assign time spend: %f seconds\n",(std::chrono::duration_cast<std::chrono::duration<double>>(time_end-time_start)).count());}return 0;}} // namespace reference_counting_

执行结果如下：

GitHub：/fengbingchun/Messy_Test