What's the instructions for type casting in Intel 64 ISA?
Such as cast long int to double?
I do some test like this:
$ cat type_cast.c
#include <stdio.h>
#include <stdlib.h>
int main()
{
long int a = 8l;
double b;
b = (double)a;
printf("%f", b);
return 0;
}
$ gcc -O0 -g -Wall type_cast.c -o type_cast
$ objdump -S type_cast
The part for main is:
int main()
{
4004c4: 55 push %rbp
4004c5: 48 89 e5 mov %rsp,%rbp
4004c8: 48 83 ec 10 sub $0x10,%rsp
long int a = 8l;
4004cc: 48 c7 45 f8 08 00 00 movq $0x8,-0x8(%rbp)
4004d3: 00
double b;
b = (double)a;
4004d4: f2 48 0f 2a 45 f8 cvtsi2sdq -0x8(%rbp),%xmm0
4004da: f2 0f 11 45 f0 movsd %xmm0,-0x10(%rbp)
printf("%f", b);
4004df: b8 f8 05 40 00 mov $0x4005f8,%eax
4004e4: f2 0f 10 45 f0 movsd -0x10(%rbp),%xmm0
4004e9: 48 89 c7 mov %rax,%rdi
4004ec: b8 01 00 00 00 mov $0x1,%eax
4004f1: e8 c2 fe ff ff callq 4003b8 <printf@plt>
return 0;
4004f6: b8 00 00 00 00 mov $0x0,%eax
}
It uses cvtsi2sdq -0x8(%rbp),%xmm0 and movsd %xmm0,-0x10(%rbp) to cast from long int to double.
I am wondering what's the other methods usually used in Intel 64 ISA.
You might not find much else as SIMD has become quite well-implemented. Naturally, you will have the packed variant (cvtpi2pd) for multiple simultaneous conversions but this is likely not what you are asking.
Your only real alternative would be the fildl
/ fstpl
pair which will load your long int in the x87 floating point stack and then read it back out onto your stack.
The technical post webpages of this site follow the CC BY-SA 4.0 protocol. If you need to reprint, please indicate the site URL or the original address.Any question please contact:yoyou2525@163.com.