1 ; RUN: llc -mtriple=arm-eabi -mattr=+neon %s -o - | FileCheck %s
3 define <8 x i8> @vabds8(<8 x i8>* %A, <8 x i8>* %B) nounwind {
6 %tmp1 = load <8 x i8>, <8 x i8>* %A
7 %tmp2 = load <8 x i8>, <8 x i8>* %B
8 %tmp3 = call <8 x i8> @llvm.arm.neon.vabds.v8i8(<8 x i8> %tmp1, <8 x i8> %tmp2)
12 define <4 x i16> @vabds16(<4 x i16>* %A, <4 x i16>* %B) nounwind {
13 ;CHECK-LABEL: vabds16:
15 %tmp1 = load <4 x i16>, <4 x i16>* %A
16 %tmp2 = load <4 x i16>, <4 x i16>* %B
17 %tmp3 = call <4 x i16> @llvm.arm.neon.vabds.v4i16(<4 x i16> %tmp1, <4 x i16> %tmp2)
21 define <2 x i32> @vabds32(<2 x i32>* %A, <2 x i32>* %B) nounwind {
22 ;CHECK-LABEL: vabds32:
24 %tmp1 = load <2 x i32>, <2 x i32>* %A
25 %tmp2 = load <2 x i32>, <2 x i32>* %B
26 %tmp3 = call <2 x i32> @llvm.arm.neon.vabds.v2i32(<2 x i32> %tmp1, <2 x i32> %tmp2)
30 define <8 x i8> @vabdu8(<8 x i8>* %A, <8 x i8>* %B) nounwind {
33 %tmp1 = load <8 x i8>, <8 x i8>* %A
34 %tmp2 = load <8 x i8>, <8 x i8>* %B
35 %tmp3 = call <8 x i8> @llvm.arm.neon.vabdu.v8i8(<8 x i8> %tmp1, <8 x i8> %tmp2)
39 define <4 x i16> @vabdu16(<4 x i16>* %A, <4 x i16>* %B) nounwind {
40 ;CHECK-LABEL: vabdu16:
42 %tmp1 = load <4 x i16>, <4 x i16>* %A
43 %tmp2 = load <4 x i16>, <4 x i16>* %B
44 %tmp3 = call <4 x i16> @llvm.arm.neon.vabdu.v4i16(<4 x i16> %tmp1, <4 x i16> %tmp2)
48 define <2 x i32> @vabdu32(<2 x i32>* %A, <2 x i32>* %B) nounwind {
49 ;CHECK-LABEL: vabdu32:
51 %tmp1 = load <2 x i32>, <2 x i32>* %A
52 %tmp2 = load <2 x i32>, <2 x i32>* %B
53 %tmp3 = call <2 x i32> @llvm.arm.neon.vabdu.v2i32(<2 x i32> %tmp1, <2 x i32> %tmp2)
57 define <2 x float> @vabdf32(<2 x float>* %A, <2 x float>* %B) nounwind {
58 ;CHECK-LABEL: vabdf32:
60 %tmp1 = load <2 x float>, <2 x float>* %A
61 %tmp2 = load <2 x float>, <2 x float>* %B
62 %tmp3 = call <2 x float> @llvm.arm.neon.vabds.v2f32(<2 x float> %tmp1, <2 x float> %tmp2)
66 define <16 x i8> @vabdQs8(<16 x i8>* %A, <16 x i8>* %B) nounwind {
67 ;CHECK-LABEL: vabdQs8:
69 %tmp1 = load <16 x i8>, <16 x i8>* %A
70 %tmp2 = load <16 x i8>, <16 x i8>* %B
71 %tmp3 = call <16 x i8> @llvm.arm.neon.vabds.v16i8(<16 x i8> %tmp1, <16 x i8> %tmp2)
75 define <8 x i16> @vabdQs16(<8 x i16>* %A, <8 x i16>* %B) nounwind {
76 ;CHECK-LABEL: vabdQs16:
78 %tmp1 = load <8 x i16>, <8 x i16>* %A
79 %tmp2 = load <8 x i16>, <8 x i16>* %B
80 %tmp3 = call <8 x i16> @llvm.arm.neon.vabds.v8i16(<8 x i16> %tmp1, <8 x i16> %tmp2)
84 define <4 x i32> @vabdQs32(<4 x i32>* %A, <4 x i32>* %B) nounwind {
85 ;CHECK-LABEL: vabdQs32:
87 %tmp1 = load <4 x i32>, <4 x i32>* %A
88 %tmp2 = load <4 x i32>, <4 x i32>* %B
89 %tmp3 = call <4 x i32> @llvm.arm.neon.vabds.v4i32(<4 x i32> %tmp1, <4 x i32> %tmp2)
93 define <16 x i8> @vabdQu8(<16 x i8>* %A, <16 x i8>* %B) nounwind {
94 ;CHECK-LABEL: vabdQu8:
96 %tmp1 = load <16 x i8>, <16 x i8>* %A
97 %tmp2 = load <16 x i8>, <16 x i8>* %B
98 %tmp3 = call <16 x i8> @llvm.arm.neon.vabdu.v16i8(<16 x i8> %tmp1, <16 x i8> %tmp2)
102 define <8 x i16> @vabdQu16(<8 x i16>* %A, <8 x i16>* %B) nounwind {
103 ;CHECK-LABEL: vabdQu16:
105 %tmp1 = load <8 x i16>, <8 x i16>* %A
106 %tmp2 = load <8 x i16>, <8 x i16>* %B
107 %tmp3 = call <8 x i16> @llvm.arm.neon.vabdu.v8i16(<8 x i16> %tmp1, <8 x i16> %tmp2)
111 define <4 x i32> @vabdQu32(<4 x i32>* %A, <4 x i32>* %B) nounwind {
112 ;CHECK-LABEL: vabdQu32:
114 %tmp1 = load <4 x i32>, <4 x i32>* %A
115 %tmp2 = load <4 x i32>, <4 x i32>* %B
116 %tmp3 = call <4 x i32> @llvm.arm.neon.vabdu.v4i32(<4 x i32> %tmp1, <4 x i32> %tmp2)
120 define <4 x float> @vabdQf32(<4 x float>* %A, <4 x float>* %B) nounwind {
121 ;CHECK-LABEL: vabdQf32:
123 %tmp1 = load <4 x float>, <4 x float>* %A
124 %tmp2 = load <4 x float>, <4 x float>* %B
125 %tmp3 = call <4 x float> @llvm.arm.neon.vabds.v4f32(<4 x float> %tmp1, <4 x float> %tmp2)
126 ret <4 x float> %tmp3
129 declare <8 x i8> @llvm.arm.neon.vabds.v8i8(<8 x i8>, <8 x i8>) nounwind readnone
130 declare <4 x i16> @llvm.arm.neon.vabds.v4i16(<4 x i16>, <4 x i16>) nounwind readnone
131 declare <2 x i32> @llvm.arm.neon.vabds.v2i32(<2 x i32>, <2 x i32>) nounwind readnone
133 declare <8 x i8> @llvm.arm.neon.vabdu.v8i8(<8 x i8>, <8 x i8>) nounwind readnone
134 declare <4 x i16> @llvm.arm.neon.vabdu.v4i16(<4 x i16>, <4 x i16>) nounwind readnone
135 declare <2 x i32> @llvm.arm.neon.vabdu.v2i32(<2 x i32>, <2 x i32>) nounwind readnone
137 declare <2 x float> @llvm.arm.neon.vabds.v2f32(<2 x float>, <2 x float>) nounwind readnone
139 declare <16 x i8> @llvm.arm.neon.vabds.v16i8(<16 x i8>, <16 x i8>) nounwind readnone
140 declare <8 x i16> @llvm.arm.neon.vabds.v8i16(<8 x i16>, <8 x i16>) nounwind readnone
141 declare <4 x i32> @llvm.arm.neon.vabds.v4i32(<4 x i32>, <4 x i32>) nounwind readnone
143 declare <16 x i8> @llvm.arm.neon.vabdu.v16i8(<16 x i8>, <16 x i8>) nounwind readnone
144 declare <8 x i16> @llvm.arm.neon.vabdu.v8i16(<8 x i16>, <8 x i16>) nounwind readnone
145 declare <4 x i32> @llvm.arm.neon.vabdu.v4i32(<4 x i32>, <4 x i32>) nounwind readnone
147 declare <4 x float> @llvm.arm.neon.vabds.v4f32(<4 x float>, <4 x float>) nounwind readnone
149 define <8 x i16> @vabdls8(<8 x i8>* %A, <8 x i8>* %B) nounwind {
150 ;CHECK-LABEL: vabdls8:
152 %tmp1 = load <8 x i8>, <8 x i8>* %A
153 %tmp2 = load <8 x i8>, <8 x i8>* %B
154 %tmp3 = call <8 x i8> @llvm.arm.neon.vabds.v8i8(<8 x i8> %tmp1, <8 x i8> %tmp2)
155 %tmp4 = zext <8 x i8> %tmp3 to <8 x i16>
159 define <4 x i32> @vabdls16(<4 x i16>* %A, <4 x i16>* %B) nounwind {
160 ;CHECK-LABEL: vabdls16:
162 %tmp1 = load <4 x i16>, <4 x i16>* %A
163 %tmp2 = load <4 x i16>, <4 x i16>* %B
164 %tmp3 = call <4 x i16> @llvm.arm.neon.vabds.v4i16(<4 x i16> %tmp1, <4 x i16> %tmp2)
165 %tmp4 = zext <4 x i16> %tmp3 to <4 x i32>
169 define <2 x i64> @vabdls32(<2 x i32>* %A, <2 x i32>* %B) nounwind {
170 ;CHECK-LABEL: vabdls32:
172 %tmp1 = load <2 x i32>, <2 x i32>* %A
173 %tmp2 = load <2 x i32>, <2 x i32>* %B
174 %tmp3 = call <2 x i32> @llvm.arm.neon.vabds.v2i32(<2 x i32> %tmp1, <2 x i32> %tmp2)
175 %tmp4 = zext <2 x i32> %tmp3 to <2 x i64>
179 define <8 x i16> @vabdlu8(<8 x i8>* %A, <8 x i8>* %B) nounwind {
180 ;CHECK-LABEL: vabdlu8:
182 %tmp1 = load <8 x i8>, <8 x i8>* %A
183 %tmp2 = load <8 x i8>, <8 x i8>* %B
184 %tmp3 = call <8 x i8> @llvm.arm.neon.vabdu.v8i8(<8 x i8> %tmp1, <8 x i8> %tmp2)
185 %tmp4 = zext <8 x i8> %tmp3 to <8 x i16>
189 define <4 x i32> @vabdlu16(<4 x i16>* %A, <4 x i16>* %B) nounwind {
190 ;CHECK-LABEL: vabdlu16:
192 %tmp1 = load <4 x i16>, <4 x i16>* %A
193 %tmp2 = load <4 x i16>, <4 x i16>* %B
194 %tmp3 = call <4 x i16> @llvm.arm.neon.vabdu.v4i16(<4 x i16> %tmp1, <4 x i16> %tmp2)
195 %tmp4 = zext <4 x i16> %tmp3 to <4 x i32>
199 define <2 x i64> @vabdlu32(<2 x i32>* %A, <2 x i32>* %B) nounwind {
200 ;CHECK-LABEL: vabdlu32:
202 %tmp1 = load <2 x i32>, <2 x i32>* %A
203 %tmp2 = load <2 x i32>, <2 x i32>* %B
204 %tmp3 = call <2 x i32> @llvm.arm.neon.vabdu.v2i32(<2 x i32> %tmp1, <2 x i32> %tmp2)
205 %tmp4 = zext <2 x i32> %tmp3 to <2 x i64>