/* { dg-do run } */
|
/* { dg-do run } */
|
/* { dg-require-effective-target fma4 } */
|
/* { dg-require-effective-target fma4 } */
|
/* { dg-options "-O2 -mfma4" } */
|
/* { dg-options "-O2 -mfma4" } */
|
|
|
#include "fma4-check.h"
|
#include "fma4-check.h"
|
|
|
#include <x86intrin.h>
|
#include <x86intrin.h>
|
#include <string.h>
|
#include <string.h>
|
|
|
#define NUM 20
|
#define NUM 20
|
|
|
union
|
union
|
{
|
{
|
__m256 x[NUM];
|
__m256 x[NUM];
|
float f[NUM * 8];
|
float f[NUM * 8];
|
__m256d y[NUM];
|
__m256d y[NUM];
|
double d[NUM * 4];
|
double d[NUM * 4];
|
} dst, res, src1, src2, src3;
|
} dst, res, src1, src2, src3;
|
|
|
|
|
/* Note that in macc*,msub*,mnmacc* and mnsub* instructions, the intermdediate
|
/* Note that in macc*,msub*,mnmacc* and mnsub* instructions, the intermdediate
|
product is not rounded, only the addition is rounded. */
|
product is not rounded, only the addition is rounded. */
|
|
|
static void
|
static void
|
init_maccps ()
|
init_maccps ()
|
{
|
{
|
int i;
|
int i;
|
for (i = 0; i < NUM * 8; i++)
|
for (i = 0; i < NUM * 8; i++)
|
{
|
{
|
src1.f[i] = i;
|
src1.f[i] = i;
|
src2.f[i] = i + 10;
|
src2.f[i] = i + 10;
|
src3.f[i] = i + 20;
|
src3.f[i] = i + 20;
|
}
|
}
|
}
|
}
|
|
|
static void
|
static void
|
init_maccpd ()
|
init_maccpd ()
|
{
|
{
|
int i;
|
int i;
|
for (i = 0; i < NUM * 4; i++)
|
for (i = 0; i < NUM * 4; i++)
|
{
|
{
|
src1.d[i] = i;
|
src1.d[i] = i;
|
src2.d[i] = i + 10;
|
src2.d[i] = i + 10;
|
src3.d[i] = i + 20;
|
src3.d[i] = i + 20;
|
}
|
}
|
}
|
}
|
|
|
static int
|
static int
|
check_maccps ()
|
check_maccps ()
|
{
|
{
|
int i, j, check_fails = 0;
|
int i, j, check_fails = 0;
|
for (i = 0; i < NUM * 8; i = i + 8)
|
for (i = 0; i < NUM * 8; i = i + 8)
|
for (j = 0; j < 8; j++)
|
for (j = 0; j < 8; j++)
|
{
|
{
|
res.f[i + j] = (src1.f[i + j] * src2.f[i + j]) + src3.f[i + j];
|
res.f[i + j] = (src1.f[i + j] * src2.f[i + j]) + src3.f[i + j];
|
if (dst.f[i + j] != res.f[i + j])
|
if (dst.f[i + j] != res.f[i + j])
|
check_fails++;
|
check_fails++;
|
}
|
}
|
return check_fails++;
|
return check_fails++;
|
}
|
}
|
|
|
static int
|
static int
|
check_maccpd ()
|
check_maccpd ()
|
{
|
{
|
int i, j, check_fails = 0;
|
int i, j, check_fails = 0;
|
for (i = 0; i < NUM * 4; i = i + 4)
|
for (i = 0; i < NUM * 4; i = i + 4)
|
for (j = 0; j < 4; j++)
|
for (j = 0; j < 4; j++)
|
{
|
{
|
res.d[i + j] = (src1.d[i + j] * src2.d[i + j]) + src3.d[i + j];
|
res.d[i + j] = (src1.d[i + j] * src2.d[i + j]) + src3.d[i + j];
|
if (dst.d[i + j] != res.d[i + j])
|
if (dst.d[i + j] != res.d[i + j])
|
check_fails++;
|
check_fails++;
|
}
|
}
|
return check_fails++;
|
return check_fails++;
|
}
|
}
|
|
|
static void
|
static void
|
fma4_test (void)
|
fma4_test (void)
|
{
|
{
|
int i;
|
int i;
|
|
|
init_maccps ();
|
init_maccps ();
|
|
|
for (i = 0; i < NUM; i++)
|
for (i = 0; i < NUM; i++)
|
dst.x[i] = _mm256_macc_ps (src1.x[i], src2.x[i], src3.x[i]);
|
dst.x[i] = _mm256_macc_ps (src1.x[i], src2.x[i], src3.x[i]);
|
|
|
if (check_maccps ())
|
if (check_maccps ())
|
abort ();
|
abort ();
|
|
|
init_maccpd ();
|
init_maccpd ();
|
|
|
for (i = 0; i < NUM; i++)
|
for (i = 0; i < NUM; i++)
|
dst.y[i] = _mm256_macc_pd (src1.y[i], src2.y[i], src3.y[i]);
|
dst.y[i] = _mm256_macc_pd (src1.y[i], src2.y[i], src3.y[i]);
|
|
|
if (check_maccpd ())
|
if (check_maccpd ())
|
abort ();
|
abort ();
|
}
|
}
|
|
|
