| 1 |
/* { dg-do compile } */ |
| 2 |
/* { dg-options "-march=rv64gcv_zvl256b -mabi=lp64d -O3 -mrvv-vector-bits=zvl -fno-schedule-insns -fno-schedule-insns2" } */ |
| 3 |
/* { dg-final { check-function-bodies "**" "" } } */ |
| 4 |
|
| 5 |
#include <stdint-gcc.h> |
| 6 |
|
| 7 |
typedef int16_t vnx16i __attribute__ ((vector_size (32))); |
| 8 |
|
| 9 |
/* |
| 10 |
** foo1: |
| 11 |
** vsetivli\s+zero,\s*16,\s*e16,\s*m1,\s*t[au],\s*m[au] |
| 12 |
** vid\.v\s+v[0-9]+ |
| 13 |
** vrsub\.vi\s+v[0-9]+,\s*v[0-9]+,\s*15 |
| 14 |
** vs1r\.v\s+v[0-9]+,\s*0\([a-x0-9]+\) |
| 15 |
** ret |
| 16 |
*/ |
| 17 |
void |
| 18 |
foo1 (int16_t *__restrict out) |
| 19 |
{ |
| 20 |
vnx16i v = {15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0}; |
| 21 |
*(vnx16i *) out = v; |
| 22 |
} |
| 23 |
|
| 24 |
/* |
| 25 |
** foo2: |
| 26 |
** vsetivli\s+zero,\s*16,\s*e16,\s*m1,\s*t[au],\s*m[au] |
| 27 |
** vid\.v\s+v[0-9]+ |
| 28 |
** li\s+[a-x0-9]+,\s*7 |
| 29 |
** vmul\.vx\s+v[0-9]+,\s*v[0-9]+,\s*[a-x0-9]+ |
| 30 |
** vadd\.vi\s+v[0-9]+,\s*v[0-9]+,\s*3 |
| 31 |
** vs1r\.v\s+v[0-9]+,\s*0\([a-x0-9]+\) |
| 32 |
** ret |
| 33 |
*/ |
| 34 |
void |
| 35 |
foo2 (int16_t *__restrict out) |
| 36 |
{ |
| 37 |
vnx16i v |
| 38 |
= {3, 3 + 7 * 1, 3 + 7 * 2, 3 + 7 * 3, 3 + 7 * 4, 3 + 7 * 5, |
| 39 |
3 + 7 * 6, 3 + 7 * 7, 3 + 7 * 8, 3 + 7 * 9, 3 + 7 * 10, 3 + 7 * 11, |
| 40 |
3 + 7 * 12, 3 + 7 * 13, 3 + 7 * 14, 3 + 7 * 15}; |
| 41 |
*(vnx16i *) out = v; |
| 42 |
} |
| 43 |
|
| 44 |
/* |
| 45 |
** foo3: |
| 46 |
** vsetivli\s+zero,\s*16,\s*e16,\s*m1,\s*t[au],\s*m[au] |
| 47 |
** vid\.v\s+v[0-9]+ |
| 48 |
** vs1r\.v\s+v[0-9]+,\s*0\([a-x0-9]+\) |
| 49 |
** ret |
| 50 |
*/ |
| 51 |
void |
| 52 |
foo3 (int16_t *__restrict out) |
| 53 |
{ |
| 54 |
vnx16i v |
| 55 |
= {0, 1,2,3,4,5,6,7,8,9,10,11,12,13,14,15}; |
| 56 |
*(vnx16i *) out = v; |
| 57 |
} |
| 58 |
|
| 59 |
/* |
| 60 |
** foo4: |
| 61 |
** vsetivli\s+zero,\s*16,\s*e16,\s*m1,\s*t[au],\s*m[au] |
| 62 |
** vid\.v\s+v[0-9]+ |
| 63 |
** li\s+[a-x0-9]+,\s*6 |
| 64 |
** vmul\.vx\s+v[0-9]+,\s*v[0-9]+,\s*[a-x0-9]+ |
| 65 |
** vs1r\.v\s+v[0-9]+,\s*0\([a-x0-9]+\) |
| 66 |
** ret |
| 67 |
*/ |
| 68 |
void |
| 69 |
foo4 (int16_t *__restrict out) |
| 70 |
{ |
| 71 |
vnx16i v |
| 72 |
= {0*6, 1*6,2*6,3*6,4*6,5*6,6*6,7*6,8*6,9*6,10*6,11*6,12*6,13*6,14*6,15*6}; |
| 73 |
*(vnx16i *) out = v; |
| 74 |
} |
| 75 |
|
| 76 |
/* |
| 77 |
** foo5: |
| 78 |
** vsetivli\s+zero,\s*16,\s*e16,\s*m1,\s*t[au],\s*m[au] |
| 79 |
** vid\.v\s+v[0-9]+ |
| 80 |
** vadd\.vi\s+v[0-9]+,\s*v[0-9]+,\s*-16 |
| 81 |
** vs1r\.v\s+v[0-9]+,\s*0\([a-x0-9]+\) |
| 82 |
** ret |
| 83 |
*/ |
| 84 |
void |
| 85 |
foo5 (int16_t *__restrict out) |
| 86 |
{ |
| 87 |
vnx16i v |
| 88 |
= {0-16, 1-16,2-16,3-16,4-16,5-16,6-16,7-16,8-16,9-16,10-16,11-16,12-16,13-16,14-16,15-16}; |
| 89 |
*(vnx16i *) out = v; |
| 90 |
} |