| 1 |
/* { dg-do assemble } */ |
| 2 |
/* { dg-options "-march=rv32gc_zfh_xtheadvector -mabi=ilp32d -O2 -save-temps" { target { rv32 } } } */ |
| 3 |
/* { dg-options "-march=rv64gc_zfh_xtheadvector -mabi=lp64d -O2 -save-temps" { target { rv64 } } } */ |
| 4 |
|
| 5 |
#include <math.h> |
| 6 |
#include "riscv_vector.h" |
| 7 |
|
| 8 |
static vfloat32m8_t atan2_ps(vfloat32m8_t a, vfloat32m8_t b, size_t vl) |
| 9 |
{ |
| 10 |
float tmpx[vl]; |
| 11 |
float tmpy[vl]; |
| 12 |
__riscv_vse32_v_f32m8(tmpx, a, vl); |
| 13 |
__riscv_vse32_v_f32m8(tmpy, b, vl); |
| 14 |
for (size_t i = 0; i < vl; i++) |
| 15 |
{ |
| 16 |
tmpx[i] = atan2(tmpx[i], tmpy[i]); |
| 17 |
} |
| 18 |
return __riscv_vle32_v_f32m8(tmpx, vl); |
| 19 |
} |
| 20 |
|
| 21 |
void my_atan2(const float *x, const float *y, float *out, int size) |
| 22 |
{ |
| 23 |
int n = size; |
| 24 |
while (n > 0) |
| 25 |
{ |
| 26 |
size_t vl = __riscv_vsetvl_e32m8(n); |
| 27 |
vfloat32m8_t _x = __riscv_vle32_v_f32m8(x, vl); |
| 28 |
vfloat32m8_t _y = __riscv_vle32_v_f32m8(y, vl); |
| 29 |
vfloat32m8_t _out = atan2_ps(_x, _y, vl); |
| 30 |
__riscv_vse32_v_f32m8(out, _out, vl); |
| 31 |
n -= vl; |
| 32 |
x += vl; |
| 33 |
y += vl; |
| 34 |
out += vl; |
| 35 |
} |
| 36 |
} |
| 37 |
|
| 38 |
/* { dg-final { scan-assembler-not {th\.vsetvli\s+zero,0} } } */ |