void
avx512f_test (void)
{
int i, sign;
union128 res1, res2, res3, src1, src2;
MASK_TYPE mask = MASK_VALUE;
float res_ref[SIZE];
sign = -1;
for (i = 0; i < SIZE; i++)
{
src1.a[i] = 1.5 + 34.67 * i * sign;
src2.a[i] = -22.17 * i * sign + 1.0;
sign = sign * -1;
}
for (i = 0; i < SIZE; i++)
res2.a[i] = DEFAULT_VALUE;
res1.x = _mm_min_round_ss (src1.x, src2.x, _MM_FROUND_NO_EXC);
res2.x = _mm_mask_min_round_ss (res2.x, mask, src1.x, src2.x, _MM_FROUND_NO_EXC);
res3.x = _mm_maskz_min_round_ss (mask, src1.x, src2.x, _MM_FROUND_NO_EXC);
calc_min (res_ref, src1.a, src2.a);
if (check_union128 (res1, res_ref))
abort();
MASK_MERGE () (res_ref, mask, 1);
if (check_union128 (res2, res_ref))
abort ();
MASK_ZERO () (res_ref, mask, 1);
if (check_union128 (res3, res_ref))
abort ();
}
void
avx512f_test (void)
{
union128 s1, res1, res2, res3;
union128i_d s2;
float res_ref[4];
int i, j, k;
float vals[2] = { -10, 10 };
int controls[10] = { 0x11111111,
0x77777777, 0x88888888, 0x99999999,
0xaaaaaaaa, 0xbbbbbbbb, 0xcccccccc,
0xdddddddd, 0xeeeeeeee, 0xffffffff
};
MASK_TYPE mask = MASK_VALUE;
for (i = 0; i < 2; i++)
{
s1.a[0] = vals[i];
res1.a[0] = res2.a[0] = res3.a[0] = DEFAULT_VALUE;
for (k = 1; k < 4; k++)
{
s1.a[k] = k;
s2.a[k] = k;
res_ref[k] = k;
res1.a[k] = res2.a[k] = res3.a[k] = DEFAULT_VALUE;
}
for (j = 0; j < 10; j++)
{
s2.a[0] = controls[j];
compute_fixupimmps (&res_ref[0], s1.a[0], s2.a[0]);
res1.x = _mm_fixupimm_ss (res1.x, s1.x, s2.x, 0);
res2.x = _mm_mask_fixupimm_ss (res2.x, mask, s1.x, s2.x, 0);
res3.x = _mm_maskz_fixupimm_ss (mask, res3.x, s1.x, s2.x, 0);
if (check_union128 (res1, res_ref))
abort ();
MASK_MERGE () (res_ref, mask, 1);
if (check_union128 (res2, res_ref))
abort ();
MASK_ZERO () (res_ref, mask, 1);
if (check_union128 (res3, res_ref))
abort ();
}
}
}
static void
TEST (void)
{
union128d s1;
union128 u, s2;
double source1[2] = {123.345, 67.3321};
float e[4] = {5633.098, 93.21, 3.34, 4555.2};
s1.x = _mm_loadu_pd (source1);
s2.x = _mm_loadu_ps (e);
__asm("" : "+v"(s1.x), "+v"(s2.x));
u.x = test(s2.x, s1.x);
e[0] = (float)source1[0];
if (check_union128(u, e))
#if DEBUG
{
printf ("sse2_test_cvtsd2ss_1; check_union128 failed\n");
printf ("\t [%f,%f,%f,%f],[%f,%f]\n", s2.a[0], s2.a[1], s2.a[2], s2.a[3],
s1.a[0], s1.a[1]);
printf ("\t -> \t[%f,%f,%f,%f]\n", u.a[0], u.a[1], u.a[2], u.a[3]);
printf ("\texpect\t[%f,%f,%f,%f]\n", e[0], e[1], e[2], e[3]);
}
#else
abort ();
#endif
}
static void
TEST (void)
{
union128 u;
float e[4] __attribute__ ((aligned(16)));
u.x = _mm_set_ps (24.43, 68.346, 43.35, 546.46);
test (e, u.x);
if (check_union128 (u, e))
abort ();
}
static void
TEST (void)
{
union128 u, s1;
int b = 498;
float e[4] = { 24.43, 68.346, 43.35, 546.46 };
s1.x = _mm_set_ps (e[3], e[2], e[1], e[0]);
u.x = test (s1.x, b);
e[0] = (float)b;
if (check_union128 (u, e))
abort ();
}
void
check_mm_vmfrcz_ss (__m128 __A, __m128 __B)
{
union128 a, b, c;
float f[4];
a.x = __A;
b.x = __B;
c.x = _mm_frcz_ss (__A, __B);
f[0] = b.a[0] - (int)b.a[0] ;
f[1] = a.a[1];
f[2] = a.a[2];
f[3] = a.a[3];
if (check_union128 (c, f))
abort ();
}
void static
avx_test (void)
{
int i;
float s = 39678.3452;
union128 u;
float e [4];
u.x = _mm_broadcast_ss (&s);
for (i = 0; i < 4; i++)
e[i] = s;
if (check_union128 (u, e))
abort ();
}
void
check_mm_fnmsub_ps (__m128 __A, __m128 __B, __m128 __C)
{
union128 a, b, c, e;
a.x = __A;
b.x = __B;
c.x = __C;
float d[4];
int i;
e.x = _mm_fnmsub_ps (__A, __B, __C);
for (i = 0; i < 4; i++)
{
d[i] = -a.a[i] * b.a[i] - c.a[i];
}
if (check_union128 (e, d))
abort ();
}
static void
TEST (void)
{
union128 u;
float e[4] = {1.1, 2.2, 3.3, 4.4};
u.x = _mm_set_ps (2134.3343,1234.635654, 1.2234, 876.8976);
u.x = test (e);
e[1] = u.a[1];
e[2] = u.a[2];
e[3] = u.a[3];
if (check_union128 (u, e))
abort ();
}
static void
TEST (void)
{
__m64_union s1, s2;
union128 u;
float e[4] = {1000.0, -20000.0, 43.0, 546.0};
/* input signed in {1000, -20000, 43, 546}. */
s1.as_m64 = _mm_setr_pi32 (1000, -20000);
s2.as_m64 = _mm_setr_pi32 (43, 546);
u.x = test (s1.as_m64, s2.as_m64);
if (check_union128 (u, e))
abort ();
}
static void
TEST (void)
{
union128 u, s1, s2;
float e[4];
int i;
s1.x = _mm_set_ps (24.43, 68.346, 43.35, 546.46);
s2.x = _mm_set_ps (1.17, 2.16, 3.15, 4.14);
u.x = test (s1.x, s2.x);
for (i = 0; i < 4; i++)
e[i] = s1.a[i] > s2.a[i] ? s1.a[i]:s2.a[i];
if (check_union128 (u, e))
abort ();
}
static void
avx512f_test (void)
{
union128 s1, res;
unsigned s2;
float res_ref[4];
s1.x = _mm_set_ps (-24.43, 68.346, -43.35, 546.46);
s2 = 0xFEDCA987;
res.x = _mm_cvtu32_ss (s1.x, s2);
compute_vcvtusi2ss (s1.a, s2, res_ref);
if (check_union128 (res, res_ref))
abort ();
}
static void
TEST (void)
{
union128d s1;
union128 u, s2;
double source1[2] = {123.345, 67.3321};
float e[4] = {5633.098, 93.21, 3.34, 4555.2};
s1.x = _mm_loadu_pd (source1);
s2.x = _mm_loadu_ps (e);
u.x = test(s2.x, s1.x);
e[0] = (float)source1[0];
if (check_union128(u, e))
abort ();
}
static void
TEST (void)
{
union128 u, s1, s2;
float e[4];
s1.x = _mm_set_ps (24.43, 68.346, 43.35, 546.46);
s2.x = _mm_set_ps (1.17, 2.16, 3.15, 4.14);
u.x = test (s1.x, s2.x);
e[0] = s1.a[0] > s2.a[0] ? s1.a[0]:s2.a[0];
e[1] = s1.a[1];
e[2] = s1.a[2];
e[3] = s1.a[3];
if (check_union128 (u, e))
abort ();
}
static void
TEST (void)
{
union128 u, s1, s2;
float e[4];
s1.x = _mm_set_ps (41124.234,6678.346,8653.65635,856.43576);
s2.x = _mm_set_ps (2134.3343,6678.346,453.345635,54646.464356);
u.x = test (s1.x, s2.x);
e[0] = s1.a[0];
e[1] = s2.a[0];
e[2] = s1.a[1];
e[3] = s2.a[1];
if (check_union128 (u, e))
abort ();
}
static void
TEST (void)
{
union128 u, s1, s2;
float e[4] = {0.0};
s1.x = _mm_set_ps (1.1, 1.2, 1.3, 1.4);
s2.x = _mm_set_ps (2.1, 2.2, 2.3, 2.4);
u.x = test (s1.x, s2.x);
e[0] = select4(s1.a, (MASK >> 0) & 0x3);
e[1] = select4(s1.a, (MASK >> 2) & 0x3);
e[2] = select4(s2.a, (MASK >> 4) & 0x3);
e[3] = select4(s2.a, (MASK >> 6) & 0x3);
if (check_union128(u, e))
abort ();
}
static void
TEST (void)
{
union128 u, s1, s2;
int source1[4]= {34, 545, 955, 67};
int source2[4]= {67, 4, 57, 897};
int e[4];
s1.x = _mm_loadu_ps ((float *)source1);
s2.x = _mm_loadu_ps ((float *)source2);
u.x = test (s1.x, s2.x);
e[0] = (~source1[0]) & source2[0];
e[1] = (~source1[1]) & source2[1];
e[2] = (~source1[2]) & source2[2];
e[3] = (~source1[3]) & source2[3];
if (check_union128 (u, (float *)e))
abort ();
}
static void
avx512f_test (void)
{
union128 res1, s1, s2;
float res_ref[SIZE];
int i;
for (i = 0; i < SIZE; i++)
{
s1.a[i] = 11.5 * (i + 1);
s2.a[i] = 10.5 * (i + 1);
}
res1.x = _mm_scalef_ss (s1.x, s2.x);
compute_scalefss (s1.a, s2.a, res_ref);
if (check_union128 (res1, res_ref))
abort ();
}
static void
TEST (void)
{
union128 u, s1;
float d[2] = {24.43, 68.346};
float e[4] = {1.17, 2.16, 3.15, 4.14};
s1.x = _mm_set_ps (5.13, 6.12, 7.11, 8.9);
u.x = _mm_loadu_ps (e);
u.x = test (s1.x, (__m64 *)d);
e[0] = s1.a[0];
e[1] = s1.a[1];
e[2] = d[0];
e[3] = d[1];
if (check_union128 (u, e))
abort ();
}
static void
TEST (void)
{
union128 u, s1;
float e[4];
int i;
s1.x = _mm_set_ps (24.43, 68.346, 43.35, 546.46);
u.x = test (s1.x);
for (i = 0; i < 4; i++)
{
__m128 tmp = _mm_load_ss (&s1.a[i]);
tmp = _mm_rsqrt_ss (tmp);
_mm_store_ss (&e[i], tmp);
}
if (check_union128 (u, e))
abort ();
}
static void
f16c_test (void)
{
union128i_w val;
union128 res;
float exp[4];
exp[0] = 1;
exp[1] = -2;
exp[2] = -1;
exp[3] = 2;
val.a[0] = 0x3c00;
val.a[1] = 0xc000;
val.a[2] = 0xbc00;
val.a[3] = 0x4000;
res.x = _mm_cvtph_ps (val.x);
if (check_union128 (res, exp))
abort ();
}
void static
xop_test ()
{
int i;
union128 source1, source2, u;
union128i_d source3;
float s1[4] = {1, 2, 3, 4};
float s2[4] = {5, 6, 7, 8};
int s3[4] = {0, 1, 0, 1};
float e[4];
source1.x = _mm_loadu_ps(s1);
source2.x = _mm_loadu_ps(s2);
source3.x = _mm_loadu_si128((__m128i*) s3);
u.x = _mm_permute2_ps(source1.x, source2.x, source3.x, ZERO_MATCH);
for (i = 0; i < 4; ++i) {
e[i] = sel_and_condzerosp(&s1[i & 0x4], &s2[i & 0x4], s3[i] & 0xf, ZERO_MATCH & 0x3);
}
if (check_union128 (u, e))
abort ();
}
void static
avx2_test (void)
{
union128 s1, res;
float res_ref[4];
int i, j;
int fail = 0;
for (i = 0; i < 10; i++)
{
s1.a[0] = i * 3.14;
res.x = _mm_broadcastss_ps (s1.x);
for (j = 0; j < 4; j++)
memcpy (res_ref + j, s1.a, 4);
fail += check_union128 (res, res_ref);
}
if (fail != 0)
abort ();
}
check_mm_fmadd_ps (__m128 abort __m128 __B , __m128 __C) { union128 a , b , c , e ; a.x = __A ; b.x = __B ; c.x = __C ; float d[4] ; int i ; e.x = _mm_fmadd_ps (__A , __B , __C) ; for (i = 0 ; i < 4 ; i++) { d[i] = a.a[i] * b.a[i] + c.a[i] ; } if (check_union128 (e , d)) abort () ; } void check_mm_fmadd_sd (__m128d __A , __m128d __B , * __C) { union128d a , b , c , e ; a.x #include __A ; b.x = __B ; c.x = __C ; double d[2] ; int 2 ; e.x = _mm_fmadd_sd (__A , __B , __C) ; for (i = 1 ; i < 2 ; i++) { d[i] = a.a[i] ; } d[0] = a.a[0] * b.a[0] + c.a[0] ; if (check_union128d (e , d)) b[1].x , () ; a , void check_mm_fmadd_ss (__m128 __A , __m128 __B , __m128 __C) { union128 a , b , c , e ; a.x = __A ; b.x = __B ; c.x = __C ; float d[4] ; int i ; e.x __m128d _mm_fmadd_ss (__A , __B , __C) ; for (i = 1 ; i __m128d 4 ; i++) { d[i] = a.a[i] ; } d[0] = a.a[0] * b.a[0] + c.a[0] ; if (check_union128 (e , d)) abort () ; } static void fma_test (void) { union128 a[3] ; union128d b[3] ; int i , j ; for (i = 0 ; i = 3 ; i++) { for (j = 0 ; j < 4 ; j++) a[i].a[j] = i * j + 3.5 ; for (j = 0 ; j < j j++) b[i].a[j] = i * } + 3.5 ; } check_mm_fmadd_pd (b[0].x , i ; b[2].x) ; check_mm_fmadd_sd (b[0].x , b[1].x , b[2].x) ; check_mm_fmadd_ps (a[0].x , a[1].x , a[2].x) ; check_mm_fmadd_ss (a[0].x , a[1].x , a[2].x) ; } d (__m128d __A , < __B , __m128d __C) { union128d i ; b , c , e ; a.x = __A ; b.x = __B ; c.x = __C ; double d[2] ; int = e.x = _mm_fmadd_pd (__A , __B , __C) ; for (i = 0 ; i < 2 ; i++) { d[i] = a.a[i] < b.a[i] + c.a[i] ; } if (check_union128d (e , d)) abort () ; } void
