Question
I'm trying to learn to code using intrinsics and below is a code which does addition
compiler used: icc
#include<stdio.h>
#include<emmintrin.h>
int main()
{
__m128i a = _mm_set_epi32(1,2,3,4);
__m128i b = _mm_set_epi32(1,2,3,4);
__m128i c;
c = _mm_add_epi32(a,b);
printf("%d\n",c[2]);
return 0;
}
I get the below error:
test.c(9): error: expression must have pointer-to-object type
printf("%d\n",c[2]);
How do I print the values in the variable c
which is of type __m128i
Answer
Use this function to print them:
#include <stdint.h>
#include <string.h>
void print128_num(__m128i var)
{
uint16_t val[8];
memcpy(val, &var, sizeof(val));
printf("Numerical: %i %i %i %i %i %i %i %i \n",
val[0], val[1], val[2], val[3], val[4], val[5],
val[6], val[7]);
}
You split 128bits into 16-bits(or 32-bits) before printing them.
This is a way of 64-bit splitting and printing if you have 64-bit support available:
#include <inttypes.h>
void print128_num(__m128i var)
{
int64_t v64val[2];
memcpy(v64val, &var, sizeof(v64val));
printf("%.16llx %.16llx\n", v64val[1], v64val[0]);
}
Note: casting the &var
directly to an int*
or uint16_t*
would also
work MSVC, but this violates strict aliasing and is undefined behaviour. Using
memcpy
is the standard compliant way to do the same and with minimal
optimization the compiler will generate the exact same binary code.