| 1 |
/* { dg-options "-O -msve-vector-bits=256 -fomit-frame-pointer" } */ |
| 2 |
|
| 3 |
#include <arm_sve.h> |
| 4 |
|
| 5 |
typedef mfloat8_t mfloat8x32_t __attribute__((vector_size (32))); |
| 6 |
typedef bfloat16_t bfloat16x16_t __attribute__((vector_size (32))); |
| 7 |
typedef float16_t float16x16_t __attribute__((vector_size (32))); |
| 8 |
typedef float32_t float32x8_t __attribute__((vector_size (32))); |
| 9 |
typedef float64_t float64x4_t __attribute__((vector_size (32))); |
| 10 |
typedef int8_t int8x32_t __attribute__((vector_size (32))); |
| 11 |
typedef int16_t int16x16_t __attribute__((vector_size (32))); |
| 12 |
typedef int32_t int32x8_t __attribute__((vector_size (32))); |
| 13 |
typedef int64_t int64x4_t __attribute__((vector_size (32))); |
| 14 |
typedef uint8_t uint8x32_t __attribute__((vector_size (32))); |
| 15 |
typedef uint16_t uint16x16_t __attribute__((vector_size (32))); |
| 16 |
typedef uint32_t uint32x8_t __attribute__((vector_size (32))); |
| 17 |
typedef uint64_t uint64x4_t __attribute__((vector_size (32))); |
| 18 |
|
| 19 |
void mfloat8_callee (mfloat8x32_t); |
| 20 |
void bfloat16_callee (bfloat16x16_t); |
| 21 |
void float16_callee (float16x16_t); |
| 22 |
void float32_callee (float32x8_t); |
| 23 |
void float64_callee (float64x4_t); |
| 24 |
void int8_callee (int8x32_t); |
| 25 |
void int16_callee (int16x16_t); |
| 26 |
void int32_callee (int32x8_t); |
| 27 |
void int64_callee (int64x4_t); |
| 28 |
void uint8_callee (uint8x32_t); |
| 29 |
void uint16_callee (uint16x16_t); |
| 30 |
void uint32_callee (uint32x8_t); |
| 31 |
void uint64_callee (uint64x4_t); |
| 32 |
|
| 33 |
void |
| 34 |
mfloat8_caller (mfloat8_t val) |
| 35 |
{ |
| 36 |
mfloat8_callee (svdup_mf8 (val)); |
| 37 |
} |
| 38 |
|
| 39 |
void |
| 40 |
bfloat16_caller (bfloat16_t val) |
| 41 |
{ |
| 42 |
bfloat16_callee (svdup_bf16 (val)); |
| 43 |
} |
| 44 |
|
| 45 |
void |
| 46 |
float16_caller (void) |
| 47 |
{ |
| 48 |
float16_callee (svdup_f16 (1.0)); |
| 49 |
} |
| 50 |
|
| 51 |
void |
| 52 |
float32_caller (void) |
| 53 |
{ |
| 54 |
float32_callee (svdup_f32 (2.0)); |
| 55 |
} |
| 56 |
|
| 57 |
void |
| 58 |
float64_caller (void) |
| 59 |
{ |
| 60 |
float64_callee (svdup_f64 (3.0)); |
| 61 |
} |
| 62 |
|
| 63 |
void |
| 64 |
int8_caller (void) |
| 65 |
{ |
| 66 |
int8_callee (svindex_s8 (0, 1)); |
| 67 |
} |
| 68 |
|
| 69 |
void |
| 70 |
int16_caller (void) |
| 71 |
{ |
| 72 |
int16_callee (svindex_s16 (0, 2)); |
| 73 |
} |
| 74 |
|
| 75 |
void |
| 76 |
int32_caller (void) |
| 77 |
{ |
| 78 |
int32_callee (svindex_s32 (0, 3)); |
| 79 |
} |
| 80 |
|
| 81 |
void |
| 82 |
int64_caller (void) |
| 83 |
{ |
| 84 |
int64_callee (svindex_s64 (0, 4)); |
| 85 |
} |
| 86 |
|
| 87 |
void |
| 88 |
uint8_caller (void) |
| 89 |
{ |
| 90 |
uint8_callee (svindex_u8 (1, 1)); |
| 91 |
} |
| 92 |
|
| 93 |
void |
| 94 |
uint16_caller (void) |
| 95 |
{ |
| 96 |
uint16_callee (svindex_u16 (1, 2)); |
| 97 |
} |
| 98 |
|
| 99 |
void |
| 100 |
uint32_caller (void) |
| 101 |
{ |
| 102 |
uint32_callee (svindex_u32 (1, 3)); |
| 103 |
} |
| 104 |
|
| 105 |
void |
| 106 |
uint64_caller (void) |
| 107 |
{ |
| 108 |
uint64_callee (svindex_u64 (1, 4)); |
| 109 |
} |
| 110 |
|
| 111 |
/* { dg-final { scan-assembler-times {\tst1b\tz[0-9]+\.b, p[0-7], \[x0\]} 3 } } */ |
| 112 |
/* { dg-final { scan-assembler-times {\tst1h\tz[0-9]+\.h, p[0-7], \[x0\]} 4 } } */ |
| 113 |
/* { dg-final { scan-assembler-times {\tst1w\tz[0-9]+\.s, p[0-7], \[x0\]} 3 } } */ |
| 114 |
/* { dg-final { scan-assembler-times {\tst1d\tz[0-9]+\.d, p[0-7], \[x0\]} 3 } } */ |
| 115 |
/* { dg-final { scan-assembler-times {\tadd\tx0, sp, #?16\n} 13 } } */ |