| 1 |
/* Test the vmulxs_laneq_f32 AArch64 SIMD intrinsic. */ |
| 2 |
|
| 3 |
/* { dg-do run } */ |
| 4 |
/* { dg-options "-save-temps -O3" } */ |
| 5 |
|
| 6 |
#include "arm_neon.h" |
| 7 |
|
| 8 |
extern void abort (void); |
| 9 |
|
| 10 |
float32_t __attribute__ ((noinline)) |
| 11 |
test_vmulxs_laneq_f32_lane0 (float32_t vec1_1, float32x4_t vec1_2) |
| 12 |
{ |
| 13 |
return vmulxs_laneq_f32 (vec1_1, vec1_2, 0); |
| 14 |
} |
| 15 |
|
| 16 |
float32_t __attribute__ ((noinline)) |
| 17 |
test_vmulxs_laneq_f32_lane1 (float32_t vec1_1, float32x4_t vec1_2) |
| 18 |
{ |
| 19 |
return vmulxs_laneq_f32 (vec1_1, vec1_2, 1); |
| 20 |
} |
| 21 |
|
| 22 |
float32_t __attribute__ ((noinline)) |
| 23 |
test_vmulxs_laneq_f32_lane2 (float32_t vec1_1, float32x4_t vec1_2) |
| 24 |
{ |
| 25 |
return vmulxs_laneq_f32 (vec1_1, vec1_2, 2); |
| 26 |
} |
| 27 |
|
| 28 |
float32_t __attribute__ ((noinline)) |
| 29 |
test_vmulxs_laneq_f32_lane3 (float32_t vec1_1, float32x4_t vec1_2) |
| 30 |
{ |
| 31 |
return vmulxs_laneq_f32 (vec1_1, vec1_2, 3); |
| 32 |
} |
| 33 |
|
| 34 |
#define PASS_ARRAY(...) {__VA_ARGS__} |
| 35 |
|
| 36 |
#define SETUP_VEC(V1_D, V2_D, EXP1, EXP2, EXP3, EXP4, I) \ |
| 37 |
void set_and_test_case##I () \ |
| 38 |
{ \ |
| 39 |
float32_t vec1 = V1_D; \ |
| 40 |
float32_t vec2_data[] = V2_D; \ |
| 41 |
float32x4_t vec2 = vld1q_f32 (vec2_data); \ |
| 42 |
float32_t expected_lane0 = EXP1; \ |
| 43 |
float32_t expected_lane1 = EXP2; \ |
| 44 |
float32_t expected_lane2 = EXP3; \ |
| 45 |
float32_t expected_lane3 = EXP4; \ |
| 46 |
float32_t actual_lane0 = test_vmulxs_laneq_f32_lane0 (vec1, vec2); \ |
| 47 |
if (actual_lane0 != expected_lane0) \ |
| 48 |
abort (); \ |
| 49 |
float32_t actual_lane1 = test_vmulxs_laneq_f32_lane1 (vec1, vec2); \ |
| 50 |
if (actual_lane1 != expected_lane1) \ |
| 51 |
abort (); \ |
| 52 |
float32_t actual_lane2 = test_vmulxs_laneq_f32_lane2 (vec1, vec2); \ |
| 53 |
if (actual_lane2 != expected_lane2) \ |
| 54 |
abort (); \ |
| 55 |
float32_t actual_lane3 = test_vmulxs_laneq_f32_lane3 (vec1, vec2); \ |
| 56 |
if (actual_lane3 != expected_lane3) \ |
| 57 |
abort (); \ |
| 58 |
} \ |
| 59 |
|
| 60 |
float32_t v1 = 3.14159265359; |
| 61 |
float32_t v2 = 1.383894; |
| 62 |
float32_t v3 = -2.71828; |
| 63 |
float32_t v4 = -3.4891931; |
| 64 |
|
| 65 |
float32_t v5 = 0.0; |
| 66 |
float32_t v6 = -0.0; |
| 67 |
float32_t v7 = __builtin_huge_valf (); |
| 68 |
float32_t v8 = -__builtin_huge_valf (); |
| 69 |
|
| 70 |
SETUP_VEC (v1, PASS_ARRAY (v1, v2, v3, v4), v1*v1, v1*v2, v3*v1, v1*v4, 1) |
| 71 |
SETUP_VEC (v5, PASS_ARRAY (v5, v6, v7, v8), 0.0, -0.0, 2.0, -2.0, 2) |
| 72 |
SETUP_VEC (v6, PASS_ARRAY (v5, v6, v7, v8), -0.0, 0.0, -2.0, 2.0, 3) |
| 73 |
|
| 74 |
int |
| 75 |
main (void) |
| 76 |
{ |
| 77 |
set_and_test_case1 (); |
| 78 |
set_and_test_case2 (); |
| 79 |
set_and_test_case3 (); |
| 80 |
return 0; |
| 81 |
} |
| 82 |
/* { dg-final { scan-assembler-times "fmulx\[ \t\]+\[sS\]\[0-9\]+, ?\[sS\]\[0-9\]+, ?\[sS\]\[0-9\]\n" 1 } } */ |
| 83 |
/* { dg-final { scan-assembler-times "fmulx\[ \t\]+\[sS\]\[0-9\]+, ?\[sS\]\[0-9\]+, ?\[vV\]\[0-9\]+\.\[sS\]\\\[1\\\]\n" 1 } } */ |
| 84 |
/* { dg-final { scan-assembler-times "fmulx\[ \t\]+\[sS\]\[0-9\]+, ?\[sS\]\[0-9\]+, ?\[vV\]\[0-9\]+\.\[sS\]\\\[2\\\]\n" 1 } } */ |
| 85 |
/* { dg-final { scan-assembler-times "fmulx\[ \t\]+\[sS\]\[0-9\]+, ?\[sS\]\[0-9\]+, ?\[vV\]\[0-9\]+\.\[sS\]\\\[3\\\]\n" 1 } } */ |