| 1 |
/* { dg-do compile } */ |
| 2 |
/* { dg-options "-mrvv-vector-bits=scalable -march=rv32gcv -mabi=ilp32 -fno-schedule-insns -fno-schedule-insns2 -fno-tree-vectorize" } */ |
| 3 |
|
| 4 |
#include "riscv_vector.h" |
| 5 |
|
| 6 |
void f (int8_t * restrict in, int8_t * restrict out, int8_t * restrict out2, int n, int m, unsigned cond, size_t vl) |
| 7 |
{ |
| 8 |
vl = 22; |
| 9 |
vbool64_t mask = *(vbool64_t*) (in + 1000000); |
| 10 |
if (cond == 0) { |
| 11 |
vint8mf8_t v = __riscv_vle8_v_i8mf8 (in, vl); |
| 12 |
__riscv_vse8_v_i8mf8 (out, v, vl); |
| 13 |
} else { |
| 14 |
out2[100] = out2[100] + 300; |
| 15 |
} |
| 16 |
|
| 17 |
for (size_t i = 0; i < n; i++) |
| 18 |
out[i + 200] = out[i + 500] + 22; |
| 19 |
|
| 20 |
for (size_t i = 0; i < n; i++) |
| 21 |
{ |
| 22 |
vfloat32mf2_t v = __riscv_vle32_v_f32mf2 ((float *)(in + i + 200), 22); |
| 23 |
__riscv_vse32_v_f32mf2 ((float *)(out + i + 200), v, 22); |
| 24 |
|
| 25 |
vfloat32mf2_t v2 = __riscv_vle32_v_f32mf2_tumu (mask, v, (float *)(in + i + 300), 22); |
| 26 |
__riscv_vse32_v_f32mf2_m (mask, (float *)(out + i + 300), v2, 22); |
| 27 |
} |
| 28 |
} |
| 29 |
|
| 30 |
void f2 (int8_t * restrict in, int8_t * restrict out, int n, int m, unsigned cond, size_t vl) |
| 31 |
{ |
| 32 |
asm volatile ("li %0, 101" :"=r" (vl)::"memory"); |
| 33 |
vbool64_t mask = *(vbool64_t*) (in + 1000000); |
| 34 |
if (cond > 0) { |
| 35 |
vint8mf8_t v = __riscv_vle8_v_i8mf8 (in, vl); |
| 36 |
__riscv_vse8_v_i8mf8 (out, v, vl); |
| 37 |
} else { |
| 38 |
out[100] = out[100] + 300; |
| 39 |
} |
| 40 |
|
| 41 |
for (size_t i = 0; i < n; i++) |
| 42 |
out[i + 200] = out[i + 500] + 555; |
| 43 |
|
| 44 |
for (size_t i = 0; i < n; i++) |
| 45 |
{ |
| 46 |
vfloat32mf2_t v = __riscv_vle32_v_f32mf2 ((float *)(in + i + 200), vl); |
| 47 |
__riscv_vse32_v_f32mf2 ((float *)(out + i + 200), v, vl); |
| 48 |
|
| 49 |
vfloat32mf2_t v2 = __riscv_vle32_v_f32mf2_tumu (mask, v, (float *)(in + i + 300), vl); |
| 50 |
__riscv_vse32_v_f32mf2_m (mask, (float *)(out + i + 300), v2, vl); |
| 51 |
} |
| 52 |
} |
| 53 |
|
| 54 |
/* { dg-final { scan-assembler-times {vsetivli\s+zero,22,\s*e32,\s*mf2,\s*tu,\s*mu} 2 { target { no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts "-Oz" no-opts "-g" no-opts "-funroll-loops" } } } } */ |
| 55 |
/* { dg-final { scan-assembler-times {vsetvli\s+zero,\s*[a-x0-9]+,\s*e32,\s*mf2,\s*tu,\s*mu} 2 { target { no-opts "-O0" no-opts "-O1" no-opts "-Os" no-opts "-Oz" no-opts "-g" no-opts "-funroll-loops" } } } } */ |
| 56 |
/* { dg-final { scan-assembler-times {vsetvli} 4 { target { no-opts "-O0" no-opts "-Os" no-opts "-Oz" no-opts "-g" no-opts "-funroll-loops" } } } } */ |
| 57 |
/* { dg-final { scan-assembler-times {vsetivli} 2 { target { no-opts "-O0" no-opts "-Os" no-opts "-Oz" no-opts "-g" no-opts "-funroll-loops" } } } } */ |