aarch64-neon-shifts.c
1.46 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
// RUN: %clang_cc1 -triple arm64-none-linux-gnu -target-feature +neon \
// RUN: -disable-O0-optnone -ffp-contract=fast -S -emit-llvm -o - %s | opt -S -mem2reg | FileCheck %s
#include <arm_neon.h>
uint8x8_t test_shift_vshr(uint8x8_t a) {
// CHECK-LABEL: test_shift_vshr
// CHECK: %{{.*}} = lshr <8 x i8> %a, <i8 5, i8 5, i8 5, i8 5, i8 5, i8 5, i8 5, i8 5>
return vshr_n_u8(a, 5);
}
int8x8_t test_shift_vshr_smax(int8x8_t a) {
// CHECK-LABEL: test_shift_vshr_smax
// CHECK: %{{.*}} = ashr <8 x i8> %a, <i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7>
return vshr_n_s8(a, 8);
}
uint8x8_t test_shift_vshr_umax(uint8x8_t a) {
// CHECK-LABEL: test_shift_vshr_umax
// CHECK: ret <8 x i8> zeroinitializer
return vshr_n_u8(a, 8);
}
uint8x8_t test_shift_vsra(uint8x8_t a, uint8x8_t b) {
// CHECK-LABEL: test_shift_vsra
// CHECK: %[[SHR:.*]] = lshr <8 x i8> %b, <i8 5, i8 5, i8 5, i8 5, i8 5, i8 5, i8 5, i8 5>
// CHECK: %{{.*}} = add <8 x i8> %a, %[[SHR]]
return vsra_n_u8(a, b, 5);
}
int8x8_t test_shift_vsra_smax(int8x8_t a, int8x8_t b) {
// CHECK-LABEL: test_shift_vsra_smax
// CHECK: %[[SHR:.*]] = ashr <8 x i8> %b, <i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7>
// CHECK: %{{.*}} = add <8 x i8> %a, %[[SHR]]
return vsra_n_s8(a, b, 8);
}
uint8x8_t test_shift_vsra_umax(uint8x8_t a, uint8x8_t b) {
// CHECK-LABEL: test_shift_vsra_umax
// CHECK: [[RES:%.*]] = add <8 x i8> %a, zeroinitializer
// CHECK: ret <8 x i8> [[RES]]
return vsra_n_u8(a, b, 8);
}