/* Test that the compiler properly optimizes floating point multiply and add
|
/* Test that the compiler properly optimizes floating point multiply and add
|
instructions vector into vfmaddps on FMA4 systems. */
|
instructions vector into vfmaddps on FMA4 systems. */
|
|
|
/* { dg-do compile } */
|
/* { dg-do compile } */
|
/* { dg-require-effective-target lp64 } */
|
/* { dg-require-effective-target lp64 } */
|
/* { dg-options "-O2 -mfma4 -ftree-vectorize" } */
|
/* { dg-options "-O2 -mfma4 -ftree-vectorize" } */
|
|
|
extern void exit (int);
|
extern void exit (int);
|
|
|
typedef float __m128 __attribute__ ((__vector_size__ (16), __may_alias__));
|
typedef float __m128 __attribute__ ((__vector_size__ (16), __may_alias__));
|
typedef double __m128d __attribute__ ((__vector_size__ (16), __may_alias__));
|
typedef double __m128d __attribute__ ((__vector_size__ (16), __may_alias__));
|
|
|
#define SIZE 10240
|
#define SIZE 10240
|
|
|
union {
|
union {
|
__m128 f_align;
|
__m128 f_align;
|
__m128d d_align;
|
__m128d d_align;
|
float f[SIZE];
|
float f[SIZE];
|
double d[SIZE];
|
double d[SIZE];
|
} a, b, c, d;
|
} a, b, c, d;
|
|
|
void
|
void
|
flt_mul_add (void)
|
flt_mul_add (void)
|
{
|
{
|
int i;
|
int i;
|
|
|
for (i = 0; i < SIZE; i++)
|
for (i = 0; i < SIZE; i++)
|
a.f[i] = (b.f[i] * c.f[i]) + d.f[i];
|
a.f[i] = (b.f[i] * c.f[i]) + d.f[i];
|
}
|
}
|
|
|
void
|
void
|
dbl_mul_add (void)
|
dbl_mul_add (void)
|
{
|
{
|
int i;
|
int i;
|
|
|
for (i = 0; i < SIZE; i++)
|
for (i = 0; i < SIZE; i++)
|
a.d[i] = (b.d[i] * c.d[i]) + d.d[i];
|
a.d[i] = (b.d[i] * c.d[i]) + d.d[i];
|
}
|
}
|
|
|
void
|
void
|
flt_mul_sub (void)
|
flt_mul_sub (void)
|
{
|
{
|
int i;
|
int i;
|
|
|
for (i = 0; i < SIZE; i++)
|
for (i = 0; i < SIZE; i++)
|
a.f[i] = (b.f[i] * c.f[i]) - d.f[i];
|
a.f[i] = (b.f[i] * c.f[i]) - d.f[i];
|
}
|
}
|
|
|
void
|
void
|
dbl_mul_sub (void)
|
dbl_mul_sub (void)
|
{
|
{
|
int i;
|
int i;
|
|
|
for (i = 0; i < SIZE; i++)
|
for (i = 0; i < SIZE; i++)
|
a.d[i] = (b.d[i] * c.d[i]) - d.d[i];
|
a.d[i] = (b.d[i] * c.d[i]) - d.d[i];
|
}
|
}
|
|
|
void
|
void
|
flt_neg_mul_add (void)
|
flt_neg_mul_add (void)
|
{
|
{
|
int i;
|
int i;
|
|
|
for (i = 0; i < SIZE; i++)
|
for (i = 0; i < SIZE; i++)
|
a.f[i] = (-(b.f[i] * c.f[i])) + d.f[i];
|
a.f[i] = (-(b.f[i] * c.f[i])) + d.f[i];
|
}
|
}
|
|
|
void
|
void
|
dbl_neg_mul_add (void)
|
dbl_neg_mul_add (void)
|
{
|
{
|
int i;
|
int i;
|
|
|
for (i = 0; i < SIZE; i++)
|
for (i = 0; i < SIZE; i++)
|
a.d[i] = (-(b.d[i] * c.d[i])) + d.d[i];
|
a.d[i] = (-(b.d[i] * c.d[i])) + d.d[i];
|
}
|
}
|
|
|
int main ()
|
int main ()
|
{
|
{
|
flt_mul_add ();
|
flt_mul_add ();
|
flt_mul_sub ();
|
flt_mul_sub ();
|
flt_neg_mul_add ();
|
flt_neg_mul_add ();
|
|
|
dbl_mul_add ();
|
dbl_mul_add ();
|
dbl_mul_sub ();
|
dbl_mul_sub ();
|
dbl_neg_mul_add ();
|
dbl_neg_mul_add ();
|
exit (0);
|
exit (0);
|
}
|
}
|
|
|
/* { dg-final { scan-assembler "vfmaddps" } } */
|
/* { dg-final { scan-assembler "vfmaddps" } } */
|
/* { dg-final { scan-assembler "vfmaddpd" } } */
|
/* { dg-final { scan-assembler "vfmaddpd" } } */
|
/* { dg-final { scan-assembler "vfmsubps" } } */
|
/* { dg-final { scan-assembler "vfmsubps" } } */
|
/* { dg-final { scan-assembler "vfmsubpd" } } */
|
/* { dg-final { scan-assembler "vfmsubpd" } } */
|
/* { dg-final { scan-assembler "vfnmaddps" } } */
|
/* { dg-final { scan-assembler "vfnmaddps" } } */
|
/* { dg-final { scan-assembler "vfnmaddpd" } } */
|
/* { dg-final { scan-assembler "vfnmaddpd" } } */
|
|
|
