floating point constant in C, bit representation

Question

It is possible to access to the bit representation of a floating point constant in C;

For example i'd like to assign

uint64_t x = //bit representation of 5.74;

which is represented by 0x40b7ae14

Do you think it is possible?

Answer 1

One way of achieving this is to use a union:

union {
    double fltValue;
    uint64_t uintValue;
} conversion;

conversion.fltValue = 5.74;
printf("%#llx\n", conversion.uintValue);

Updated with %#x thanks to Aracthor for mentioning it. And %#llx thanks to EOF.

For a working example (with float instead of double ) see:

https://ideone.com/p4rH5l

Answer 2

It is possible, but is not portable because you cannot be sure of how a floating point value is represented : C standard does not define it.

You could simply use casting of pointers :

float x = 5.74;
void *pt = &x
uint64_t *ip = pt;
uint64_t i = *ip;

This is formal undefined behaviour because you are casting a pointer to a different type, and you should not do it because you add endian problems to the floating point representation.

The correct way would be :

float x = 5.74;
void *pt = &x
unsigned char *ip = pt;

ip now point to a unsigned char[] of size sizeof(float) containing the binary representation of a float. And no undefined behaviour was invoked because casting a pointer to a void * or a char * is always allowed.

Answer 3

#include <stdint.h>
#include <stdio.h>

int main(void) {
    double x = 5.74;
    uint64_t y = *((uint64_t*)&x);
    printf("0x%016llx", y);  /* 0x4016f5c28f5c28f6 */
}

Answer 4

如果要使用浮点数的二进制表示形式，请使用整数指针..将float变量的地址分配给整数指针，并通过取消引用int指针的值来获取其二进制形式。某种程度上想要避免由异常类型转换产生的警告，那么您可以始终使用void *