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
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
|
; RUN: llvm-as < %s | llc -march=arm -mattr=+neon > %t
; RUN: grep {vqshrn\\.s16} %t | count 1
; RUN: grep {vqshrn\\.s32} %t | count 1
; RUN: grep {vqshrn\\.s64} %t | count 1
; RUN: grep {vqshrn\\.u16} %t | count 1
; RUN: grep {vqshrn\\.u32} %t | count 1
; RUN: grep {vqshrn\\.u64} %t | count 1
; RUN: grep {vqshrun\\.s16} %t | count 1
; RUN: grep {vqshrun\\.s32} %t | count 1
; RUN: grep {vqshrun\\.s64} %t | count 1
define <8 x i8> @vqshrns8(<8 x i16>* %A) nounwind {
%tmp1 = load <8 x i16>* %A
%tmp2 = call <8 x i8> @llvm.arm.neon.vqshiftns.v8i8(<8 x i16> %tmp1, <8 x i16> < i16 -8, i16 -8, i16 -8, i16 -8, i16 -8, i16 -8, i16 -8, i16 -8 >)
ret <8 x i8> %tmp2
}
define <4 x i16> @vqshrns16(<4 x i32>* %A) nounwind {
%tmp1 = load <4 x i32>* %A
%tmp2 = call <4 x i16> @llvm.arm.neon.vqshiftns.v4i16(<4 x i32> %tmp1, <4 x i32> < i32 -16, i32 -16, i32 -16, i32 -16 >)
ret <4 x i16> %tmp2
}
define <2 x i32> @vqshrns32(<2 x i64>* %A) nounwind {
%tmp1 = load <2 x i64>* %A
%tmp2 = call <2 x i32> @llvm.arm.neon.vqshiftns.v2i32(<2 x i64> %tmp1, <2 x i64> < i64 -32, i64 -32 >)
ret <2 x i32> %tmp2
}
define <8 x i8> @vqshrnu8(<8 x i16>* %A) nounwind {
%tmp1 = load <8 x i16>* %A
%tmp2 = call <8 x i8> @llvm.arm.neon.vqshiftnu.v8i8(<8 x i16> %tmp1, <8 x i16> < i16 -8, i16 -8, i16 -8, i16 -8, i16 -8, i16 -8, i16 -8, i16 -8 >)
ret <8 x i8> %tmp2
}
define <4 x i16> @vqshrnu16(<4 x i32>* %A) nounwind {
%tmp1 = load <4 x i32>* %A
%tmp2 = call <4 x i16> @llvm.arm.neon.vqshiftnu.v4i16(<4 x i32> %tmp1, <4 x i32> < i32 -16, i32 -16, i32 -16, i32 -16 >)
ret <4 x i16> %tmp2
}
define <2 x i32> @vqshrnu32(<2 x i64>* %A) nounwind {
%tmp1 = load <2 x i64>* %A
%tmp2 = call <2 x i32> @llvm.arm.neon.vqshiftnu.v2i32(<2 x i64> %tmp1, <2 x i64> < i64 -32, i64 -32 >)
ret <2 x i32> %tmp2
}
define <8 x i8> @vqshruns8(<8 x i16>* %A) nounwind {
%tmp1 = load <8 x i16>* %A
%tmp2 = call <8 x i8> @llvm.arm.neon.vqshiftnsu.v8i8(<8 x i16> %tmp1, <8 x i16> < i16 -8, i16 -8, i16 -8, i16 -8, i16 -8, i16 -8, i16 -8, i16 -8 >)
ret <8 x i8> %tmp2
}
define <4 x i16> @vqshruns16(<4 x i32>* %A) nounwind {
%tmp1 = load <4 x i32>* %A
%tmp2 = call <4 x i16> @llvm.arm.neon.vqshiftnsu.v4i16(<4 x i32> %tmp1, <4 x i32> < i32 -16, i32 -16, i32 -16, i32 -16 >)
ret <4 x i16> %tmp2
}
define <2 x i32> @vqshruns32(<2 x i64>* %A) nounwind {
%tmp1 = load <2 x i64>* %A
%tmp2 = call <2 x i32> @llvm.arm.neon.vqshiftnsu.v2i32(<2 x i64> %tmp1, <2 x i64> < i64 -32, i64 -32 >)
ret <2 x i32> %tmp2
}
declare <8 x i8> @llvm.arm.neon.vqshiftns.v8i8(<8 x i16>, <8 x i16>) nounwind readnone
declare <4 x i16> @llvm.arm.neon.vqshiftns.v4i16(<4 x i32>, <4 x i32>) nounwind readnone
declare <2 x i32> @llvm.arm.neon.vqshiftns.v2i32(<2 x i64>, <2 x i64>) nounwind readnone
declare <8 x i8> @llvm.arm.neon.vqshiftnu.v8i8(<8 x i16>, <8 x i16>) nounwind readnone
declare <4 x i16> @llvm.arm.neon.vqshiftnu.v4i16(<4 x i32>, <4 x i32>) nounwind readnone
declare <2 x i32> @llvm.arm.neon.vqshiftnu.v2i32(<2 x i64>, <2 x i64>) nounwind readnone
declare <8 x i8> @llvm.arm.neon.vqshiftnsu.v8i8(<8 x i16>, <8 x i16>) nounwind readnone
declare <4 x i16> @llvm.arm.neon.vqshiftnsu.v4i16(<4 x i32>, <4 x i32>) nounwind readnone
declare <2 x i32> @llvm.arm.neon.vqshiftnsu.v2i32(<2 x i64>, <2 x i64>) nounwind readnone
|
