This topic has not yet been rated - Rate this topic

_mm_sign_epi8

Visual Studio 2008

Microsoft Specific

Emits the Intel Supplemental Streaming SIMD Extensions 3 (SSSE3) instruction psignb. This instruction negates or clears 8-bit signed integers of a 128-bit parameter.

Copy

__m128i _mm_sign_epi8( 
   __m128i a,
   __m128i b
);

Parameters

[in] a: A 128-bit parameter that contains sixteen 8-bit signed integers.
[in] b: A 128-bit parameter that contains sixteen 8-bit signed integers.

Return value

The return value can be expressed with the following equations.

r0 := (b0 < 0) ? -a0 : ((b0 == 0) ? 0 : a0)

r1 := (b1 < 0) ? -a1 : ((b1 == 0) ? 0 : a1)

...

r15 := (b15 < 0) ? -a15 : ((b15 == 0) ? 0 : a15)

Requirements

Intrinsic	Architecture
_mm_sign_epi8	x86, x64

Header file <tmmintrin.h>

Remarks

r0-r15, a0-a15, and b0-b15 are the sequentially ordered 8-bit components of return value r and parameters a and b. r0, a0, and b0 are the least significant 8 bits.

Before you use this intrinsic, software must ensure that the underlying processor supports the instruction.

Example

Copy

#include <stdio.h>
#include <tmmintrin.h>

int main ()
{
    __m128i a, b;

    a.m128i_i8[0] = 25;
    a.m128i_i8[1] = 31;
    a.m128i_i8[2] = -1;
    a.m128i_i8[3] = 10;
    a.m128i_i8[4] = -52;
    a.m128i_i8[5] = -127;
    a.m128i_i8[6] = 127;
    a.m128i_i8[7] = 32;
    a.m128i_i8[8] = 42;
    a.m128i_i8[9] = -15;
    a.m128i_i8[10] = -97;
    a.m128i_i8[11] = 100;
    a.m128i_i8[12] = 125;
    a.m128i_i8[13] = 76;
    a.m128i_i8[14] = -60;
    a.m128i_i8[15] = 1;

    b.m128i_i8[0] = 1;
    b.m128i_i8[1] = -1;
    b.m128i_i8[2] = 0;
    b.m128i_i8[3] = 127;
    b.m128i_i8[4] = -128;
    b.m128i_i8[5] = -42;
    b.m128i_i8[6] = 31;
    b.m128i_i8[7] = 1;
    b.m128i_i8[8] = 0;
    b.m128i_i8[9] = 1;
    b.m128i_i8[10] = -1;
    b.m128i_i8[11] = -1;
    b.m128i_i8[12] = 1;
    b.m128i_i8[13] = -1;
    b.m128i_i8[14] = 1;
    b.m128i_i8[15] = 0;

    __m128i res = _mm_sign_epi8(a, b);

    printf_s("   a\t   b\t res\n%4d\t%4d\t%4d\n%4d\t%4d\t%4d\n",
                a.m128i_i8[0], b.m128i_i8[0], res.m128i_i8[0],
                a.m128i_i8[1], b.m128i_i8[1], res.m128i_i8[1]);
    printf_s("%4d\t%4d\t%4d\n%4d\t%4d\t%4d\n%4d\t%4d\t%4d\n%4d\t%4d\t%4d\n",
                a.m128i_i8[2], b.m128i_i8[2], res.m128i_i8[2],
                a.m128i_i8[3], b.m128i_i8[3], res.m128i_i8[3],
                a.m128i_i8[4], b.m128i_i8[4], res.m128i_i8[4],
                a.m128i_i8[5], b.m128i_i8[5], res.m128i_i8[5]);
    printf_s("%4d\t%4d\t%4d\n%4d\t%4d\t%4d\n%4d\t%4d\t%4d\n%4d\t%4d\t%4d\n",
                a.m128i_i8[6], b.m128i_i8[6], res.m128i_i8[6],
                a.m128i_i8[7], b.m128i_i8[7], res.m128i_i8[7],
                a.m128i_i8[8], b.m128i_i8[8], res.m128i_i8[8],
                a.m128i_i8[9], b.m128i_i8[9], res.m128i_i8[9]);
    printf_s("%4d\t%4d\t%4d\n%4d\t%4d\t%4d\n%4d\t%4d\t%4d\n%4d\t%4d\t%4d\n",
                a.m128i_i8[10], b.m128i_i8[10], res.m128i_i8[10],
                a.m128i_i8[11], b.m128i_i8[11], res.m128i_i8[11],
                a.m128i_i8[12], b.m128i_i8[12], res.m128i_i8[12],
                a.m128i_i8[13], b.m128i_i8[13], res.m128i_i8[13]);
    printf_s("%4d\t%4d\t%4d\n%4d\t%4d\t%4d\n",
                a.m128i_i8[14], b.m128i_i8[14], res.m128i_i8[14],
                a.m128i_i8[15], b.m128i_i8[15], res.m128i_i8[15]);

    return 0;
}

a b res 25 1 25 31 -1 -31 -1 0 0 10 127 10 -52 -128 52 -127 -42 127 127 31 127 32 1 32 42 0 0 -15 1 -15 -97 -1 97 100 -1 -100 125 1 125 76 -1 -76 -60 1 -60 1 0 0

Reference

Compiler Intrinsics

Did you find this helpful? Yes No

Not accurate

Not enough depth

Need more code examples

(1500 characters remaining)

Community Content Add

FAQ