C memory alignment

Reference resources ：https://www.cnblogs.com/sunbines/p/9257981.html
Improve the specific graphical display of memory alignment ！

Alignment reason ：

1） Platform reasons （ Reasons for transplantation ）： Not all hardware platforms can access any data at any address , Some hardware platforms can only access certain types of data at certain addresses , Otherwise, a hardware exception will be thrown
2） Hardware reasons ： After memory alignment ,CPU The memory access speed of is greatly improved .

1. Alignment principle ：

【 principle 1】 Data member alignment rules ： structure (struct)( Or in association with (union)) Data members of , The first data member is placed in offset by 0 The place of , In the future, the alignment of each data member is in accordance with #pragma pack The specified value and the length of the data member itself , The smaller one is going to .

【 principle 2】 structure ( Or in association with ) The overall alignment rule of ： After the data members are aligned , structure ( Or in association with ) Alignment itself , The alignment will follow #pragma pack Specified values and structures ( Or in association with ) In the maximum data member length , The smaller one is going to .

【 principle 3】 The structure as a member ： If there are some members in a structure , Structure members are stored from the integer multiple address of the maximum element size inside .

remarks ： Array members are calculated by the length of the array type , Such as char t[9], In the 1 In the step, the length of the data itself is determined by 1 count , When accumulating structures, the length is 9; The first 2 In step , When finding the maximum data length , If the structure T There are complex type members A, The A The length of the member is the member of the complex type A Maximum member length of .

2. Natural alignment ： The address where the variable is stored is an integer multiple of the data type size of the variable . Such as ： Deposit int The address of type data must be 4 Multiple , Deposit short The address of type data must be 2 Multiple .

3. Change the default boundary condition ( Specify the opposite boundary )：

   Using pseudo instructions #pragma pack (n),C The compiler will follow n Byte alignment .
   Using pseudo instructions #pragma pack（）, Cancel custom byte alignment .

#include <stdio.h>
#include "io_utils.h"

typedef  struct Person{
    
    char *name;
    int age;
    short s;
    double salary;

}Person;

int main() {
    
    Person *person ;
    person->name = "zyccc";
    person->age = 22;
    person->s = 2;
    person->salary = 998999;
    PRINTLNF("2333");

    return 0;
}

As shown in the figure, it can be concluded that front char* Is a pointer type my OS yes 64 Location Account for the 8 Bytes 【0-5】 And then because of int The type is 4 Bytes So the address alignment should be 4 A multiple therefore The first 【6-7】 A byte complements 0 Align memory Back short Types occupy 2 Bytes 【12-13】 because double Types occupy 8 Bytes need to be completed 0 To the first 16 Location therefore Mended 6 individual 0 The following is the corresponding eight bytes double For the corresponding 【16-24】. *** double The number of bytes occupied is related to the actual size as well as float

And the following outputs can be clearly drawn The above conclusion
The code is as follows

#include <stdio.h>
#include "io_utils.h"
#include <stddef.h>

typedef  struct Person{
    
    char *name;
    int age;
    short s;
    double salary;

}Person;

int main() {
    
    Person *person ;
    person->name = "zyccc";
    person->age = 22;
    person->s = 2;
    person->salary = 9989999;
    PRINT_INT(_Alignof(person->age));
    PRINT_INT(offsetof(Person ,age));
    PRINT_INT(sizeof(Person));

    return 0;
}