| 1 |
/* { dg-do compile } */ |
| 2 |
/* { dg-additional-options "-march=rv32gcv_zvfh -mabi=ilp32d -mrvv-vector-bits=scalable -fno-vect-cost-model -ffast-math -fdump-tree-optimized-details" } */ |
| 3 |
|
| 4 |
#include <stdint-gcc.h> |
| 5 |
|
| 6 |
#define TEST(TYPE, NAME, OP) \ |
| 7 |
void __attribute__ ((noinline, noclone)) \ |
| 8 |
test_##TYPE##_##NAME (TYPE *__restrict x, TYPE *__restrict y, \ |
| 9 |
TYPE *__restrict z, TYPE *__restrict pred, \ |
| 10 |
TYPE *__restrict merged, int n) \ |
| 11 |
{ \ |
| 12 |
for (int i = 0; i < n; ++i) \ |
| 13 |
x[i] = pred[i] != 1 ? y[i] OP z[i] : merged[i]; \ |
| 14 |
} |
| 15 |
|
| 16 |
#define TEST_TYPE(TYPE) \ |
| 17 |
TEST (TYPE, add, +) \ |
| 18 |
TEST (TYPE, sub, -) \ |
| 19 |
TEST (TYPE, mul, *) \ |
| 20 |
TEST (TYPE, div, /) |
| 21 |
|
| 22 |
#define TEST_TYPE2(TYPE) TEST (TYPE, rem, %) |
| 23 |
|
| 24 |
#define TEST_ALL \ |
| 25 |
TEST_TYPE (int8_t) \ |
| 26 |
TEST_TYPE (uint8_t) \ |
| 27 |
TEST_TYPE (int16_t) \ |
| 28 |
TEST_TYPE (uint16_t) \ |
| 29 |
TEST_TYPE (int32_t) \ |
| 30 |
TEST_TYPE (uint32_t) \ |
| 31 |
TEST_TYPE (int64_t) \ |
| 32 |
TEST_TYPE (uint64_t) \ |
| 33 |
TEST_TYPE2 (int8_t) \ |
| 34 |
TEST_TYPE2 (uint8_t) \ |
| 35 |
TEST_TYPE2 (int16_t) \ |
| 36 |
TEST_TYPE2 (uint16_t) \ |
| 37 |
TEST_TYPE2 (int32_t) \ |
| 38 |
TEST_TYPE2 (uint32_t) \ |
| 39 |
TEST_TYPE2 (int64_t) \ |
| 40 |
TEST_TYPE2 (uint64_t) \ |
| 41 |
TEST_TYPE (_Float16) \ |
| 42 |
TEST_TYPE (float) \ |
| 43 |
TEST_TYPE (double) |
| 44 |
|
| 45 |
TEST_ALL |
| 46 |
|
| 47 |
/* { dg-final { scan-tree-dump-times "\.COND_LEN_DIV" 8 "optimized" } } */ |
| 48 |
/* { dg-final { scan-tree-dump-times "\.COND_LEN_MOD" 8 "optimized" } } */ |
| 49 |
/* { dg-final { scan-tree-dump-times "\.COND_ADD" 11 "optimized" } } */ |
| 50 |
/* { dg-final { scan-tree-dump-times "\.COND_SUB" 11 "optimized" } } */ |
| 51 |
/* { dg-final { scan-tree-dump-times "\.COND_MUL" 11 "optimized" } } */ |
| 52 |
/* { dg-final { scan-tree-dump-times "\.COND_RDIV" 3 "optimized" } } */ |
| 53 |
/* { dg-final { scan-assembler-times {vadd\.vv\s+v[0-9]+,v[0-9]+,v[0-9]+,v0.t} 8 } } */ |
| 54 |
/* { dg-final { scan-assembler-times {vsub\.vv\s+v[0-9]+,v[0-9]+,v[0-9]+,v0.t} 8 } } */ |
| 55 |
/* { dg-final { scan-assembler-times {vmul\.vv\s+v[0-9]+,v[0-9]+,v[0-9]+,v0.t} 8 } } */ |
| 56 |
/* { dg-final { scan-assembler-times {vdivu?\.vv\s+v[0-9]+,v[0-9]+,v[0-9]+,v0.t} 8 } } */ |
| 57 |
/* { dg-final { scan-assembler-times {vrem\.vv\s+v[0-9]+,v[0-9]+,v[0-9]+,v0.t} 4 } } */ |
| 58 |
/* { dg-final { scan-assembler-times {vremu\.vv\s+v[0-9]+,v[0-9]+,v[0-9]+,v0.t} 4 } } */ |
| 59 |
/* { dg-final { scan-assembler-times {vfadd\.vv\s+v[0-9]+,v[0-9]+,v[0-9]+,v0.t} 3 } } */ |
| 60 |
/* { dg-final { scan-assembler-times {vfsub\.vv\s+v[0-9]+,v[0-9]+,v[0-9]+,v0.t} 3 } } */ |
| 61 |
/* { dg-final { scan-assembler-times {vfmul\.vv\s+v[0-9]+,v[0-9]+,v[0-9]+,v0.t} 3 } } */ |
| 62 |
/* { dg-final { scan-assembler-times {vfdiv\.vv\s+v[0-9]+,v[0-9]+,v[0-9]+,v0.t} 3 } } */ |
| 63 |
/* { dg-final { scan-assembler-not {\tvf?merge\.v[vxi]m\t} } } */ |