[英]sizeof operator & alignment in C++ vs D
Consider following program: 考虑以下计划:
#include <iostream>
class T {
char c;
int i;
};
int main() {
std::cout<<sizeof(T)<<'\n';
}
It gives output 8
as an expected because of alignment. 由于对齐,它将输出
8
作为预期。 C++ Compiler adds padding of 3 bytes. C ++编译器添加3个字节的填充。 But If I do the same in D language it gives me completely unexpected output.
但如果我用D语言做同样的事情,它会给我完全意想不到的输出。 (See live demo here .)
( 在这里查看现场演示。)
import std.stdio;
class T {
char c;
int i;
}
int main() {
writefln("sizeof T is %d",T.sizeof);
writefln("sizeof char is %d",char.sizeof);
writefln("sizeof int is %d",int.sizeof);
return 0;
}
The output I get is: 我得到的输出是:
sizeof T is 4
sizeof char is 1
sizeof int is 4
How sizeof(T)
is 4 ? sizeof(T)
是多少? I was expecting to get 8
as an output of class' size. 我期望得到
8
作为班级规模的输出。 How D compiler performs alignment here? D编译器如何在这里执行对齐? Am I understading wrong something ?
我错了什么吗? I am using Windows 7 32 bit OS & Dmd compiler.
我正在使用Windows 7 32位OS和Dmd编译器。
Classes in D are reference types (ie they work like in Java or C#). D中的类是引用类型(即它们的工作方式类似于Java或C#)。 When you declare a variable of type
T
(where T
is a class), you're only declaring a class reference (which will be null
by default), which will point to the actual class's data (the char c
and int i
in your example). 当你声明一个
T
类型的变量(其中T
是一个类)时,你只是声明一个类引用(默认情况下为null
),它将指向实际类的数据( char c
和int i
在你的例)。 Thus, T.sizeof
only measures the size of the reference, which will be equal to the pointer size (a result of 4 only indicates that you're targeting a 32-bit platform). 因此,
T.sizeof
仅测量引用的大小,该大小将等于指针大小(4的结果仅表示您的目标是32位平台)。
Try declaring T
as a struct
: 尝试将
T
声明为struct
:
import std.stdio;
struct T {
char c;
int i;
}
int main() {
writefln("sizeof T is %d",T.sizeof);
writefln("sizeof char is %d",char.sizeof);
writefln("sizeof int is %d",int.sizeof);
return 0;
}
On my machine, the above outputs: 在我的机器上,上面的输出:
sizeof T is 8
sizeof char is 1
sizeof int is 4
声明:本站的技术帖子网页,遵循CC BY-SA 4.0协议,如果您需要转载,请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.