1 ; NOTE: Assertions have been autogenerated by utils/update_analyze_test_checks.py
2 ; RUN: opt < %s -mtriple=aarch64-unknown-linux-gnu -cost-model -cost-kind=throughput -analyze | FileCheck %s --check-prefixes=CHECK,CHECK-NOF16
3 ; RUN: opt < %s -mtriple=aarch64-unknown-linux-gnu -mattr=+fullfp16 -cost-model -cost-kind=throughput -analyze | FileCheck %s --check-prefixes=CHECK,CHECK-F16
5 define void @reduce_umin() {
6 ; CHECK-LABEL: 'reduce_umin'
7 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V1i8 = call <1 x i8> @llvm.umin.v1i8(<1 x i8> undef, <1 x i8> undef)
8 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V3i8 = call <3 x i8> @llvm.umin.v3i8(<3 x i8> undef, <3 x i8> undef)
9 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4i8 = call <4 x i8> @llvm.umin.v4i8(<4 x i8> undef, <4 x i8> undef)
10 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V8i8 = call <8 x i8> @llvm.umin.v8i8(<8 x i8> undef, <8 x i8> undef)
11 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V16i8 = call <16 x i8> @llvm.umin.v16i8(<16 x i8> undef, <16 x i8> undef)
12 ; CHECK-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V32i8 = call <32 x i8> @llvm.umin.v32i8(<32 x i8> undef, <32 x i8> undef)
13 ; CHECK-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V64i8 = call <64 x i8> @llvm.umin.v64i8(<64 x i8> undef, <64 x i8> undef)
14 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2i16 = call <2 x i16> @llvm.umin.v2i16(<2 x i16> undef, <2 x i16> undef)
15 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4i16 = call <4 x i16> @llvm.umin.v4i16(<4 x i16> undef, <4 x i16> undef)
16 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V8i16 = call <8 x i16> @llvm.umin.v8i16(<8 x i16> undef, <8 x i16> undef)
17 ; CHECK-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V16i16 = call <16 x i16> @llvm.umin.v16i16(<16 x i16> undef, <16 x i16> undef)
18 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2i32 = call <2 x i32> @llvm.umin.v2i32(<2 x i32> undef, <2 x i32> undef)
19 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4i32 = call <4 x i32> @llvm.umin.v4i32(<4 x i32> undef, <4 x i32> undef)
20 ; CHECK-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V8i32 = call <8 x i32> @llvm.umin.v8i32(<8 x i32> undef, <8 x i32> undef)
21 ; CHECK-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V2i64 = call <2 x i64> @llvm.umin.v2i64(<2 x i64> undef, <2 x i64> undef)
22 ; CHECK-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V4i64 = call <4 x i64> @llvm.umin.v4i64(<4 x i64> undef, <4 x i64> undef)
23 ; CHECK-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret void
25 %V1i8 = call <1 x i8> @llvm.umin.v1i8(<1 x i8> undef, <1 x i8> undef)
26 %V3i8 = call <3 x i8> @llvm.umin.v3i8(<3 x i8> undef, <3 x i8> undef)
27 %V4i8 = call <4 x i8> @llvm.umin.v4i8(<4 x i8> undef, <4 x i8> undef)
28 %V8i8 = call <8 x i8> @llvm.umin.v8i8(<8 x i8> undef, <8 x i8> undef)
29 %V16i8 = call <16 x i8> @llvm.umin.v16i8(<16 x i8> undef, <16 x i8> undef)
30 %V32i8 = call <32 x i8> @llvm.umin.v32i8(<32 x i8> undef, <32 x i8> undef)
31 %V64i8 = call <64 x i8> @llvm.umin.v64i8(<64 x i8> undef, <64 x i8> undef)
32 %V2i16 = call <2 x i16> @llvm.umin.v2i16(<2 x i16> undef, <2 x i16> undef)
33 %V4i16 = call <4 x i16> @llvm.umin.v4i16(<4 x i16> undef, <4 x i16> undef)
34 %V8i16 = call <8 x i16> @llvm.umin.v8i16(<8 x i16> undef, <8 x i16> undef)
35 %V16i16 = call <16 x i16> @llvm.umin.v16i16(<16 x i16> undef, <16 x i16> undef)
36 %V2i32 = call <2 x i32> @llvm.umin.v2i32(<2 x i32> undef, <2 x i32> undef)
37 %V4i32 = call <4 x i32> @llvm.umin.v4i32(<4 x i32> undef, <4 x i32> undef)
38 %V8i32 = call <8 x i32> @llvm.umin.v8i32(<8 x i32> undef, <8 x i32> undef)
39 %V2i64 = call <2 x i64> @llvm.umin.v2i64(<2 x i64> undef, <2 x i64> undef)
40 %V4i64 = call <4 x i64> @llvm.umin.v4i64(<4 x i64> undef, <4 x i64> undef)
44 define void @reduce_umax() {
45 ; CHECK-LABEL: 'reduce_umax'
46 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V1i8 = call <1 x i8> @llvm.umax.v1i8(<1 x i8> undef, <1 x i8> undef)
47 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V3i8 = call <3 x i8> @llvm.umax.v3i8(<3 x i8> undef, <3 x i8> undef)
48 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4i8 = call <4 x i8> @llvm.umax.v4i8(<4 x i8> undef, <4 x i8> undef)
49 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V8i8 = call <8 x i8> @llvm.umax.v8i8(<8 x i8> undef, <8 x i8> undef)
50 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V16i8 = call <16 x i8> @llvm.umax.v16i8(<16 x i8> undef, <16 x i8> undef)
51 ; CHECK-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V32i8 = call <32 x i8> @llvm.umax.v32i8(<32 x i8> undef, <32 x i8> undef)
52 ; CHECK-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V64i8 = call <64 x i8> @llvm.umax.v64i8(<64 x i8> undef, <64 x i8> undef)
53 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2i16 = call <2 x i16> @llvm.umax.v2i16(<2 x i16> undef, <2 x i16> undef)
54 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4i16 = call <4 x i16> @llvm.umax.v4i16(<4 x i16> undef, <4 x i16> undef)
55 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V8i16 = call <8 x i16> @llvm.umax.v8i16(<8 x i16> undef, <8 x i16> undef)
56 ; CHECK-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V16i16 = call <16 x i16> @llvm.umax.v16i16(<16 x i16> undef, <16 x i16> undef)
57 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2i32 = call <2 x i32> @llvm.umax.v2i32(<2 x i32> undef, <2 x i32> undef)
58 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4i32 = call <4 x i32> @llvm.umax.v4i32(<4 x i32> undef, <4 x i32> undef)
59 ; CHECK-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V8i32 = call <8 x i32> @llvm.umax.v8i32(<8 x i32> undef, <8 x i32> undef)
60 ; CHECK-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V2i64 = call <2 x i64> @llvm.umax.v2i64(<2 x i64> undef, <2 x i64> undef)
61 ; CHECK-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V4i64 = call <4 x i64> @llvm.umax.v4i64(<4 x i64> undef, <4 x i64> undef)
62 ; CHECK-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret void
64 %V1i8 = call <1 x i8> @llvm.umax.v1i8(<1 x i8> undef, <1 x i8> undef)
65 %V3i8 = call <3 x i8> @llvm.umax.v3i8(<3 x i8> undef, <3 x i8> undef)
66 %V4i8 = call <4 x i8> @llvm.umax.v4i8(<4 x i8> undef, <4 x i8> undef)
67 %V8i8 = call <8 x i8> @llvm.umax.v8i8(<8 x i8> undef, <8 x i8> undef)
68 %V16i8 = call <16 x i8> @llvm.umax.v16i8(<16 x i8> undef, <16 x i8> undef)
69 %V32i8 = call <32 x i8> @llvm.umax.v32i8(<32 x i8> undef, <32 x i8> undef)
70 %V64i8 = call <64 x i8> @llvm.umax.v64i8(<64 x i8> undef, <64 x i8> undef)
71 %V2i16 = call <2 x i16> @llvm.umax.v2i16(<2 x i16> undef, <2 x i16> undef)
72 %V4i16 = call <4 x i16> @llvm.umax.v4i16(<4 x i16> undef, <4 x i16> undef)
73 %V8i16 = call <8 x i16> @llvm.umax.v8i16(<8 x i16> undef, <8 x i16> undef)
74 %V16i16 = call <16 x i16> @llvm.umax.v16i16(<16 x i16> undef, <16 x i16> undef)
75 %V2i32 = call <2 x i32> @llvm.umax.v2i32(<2 x i32> undef, <2 x i32> undef)
76 %V4i32 = call <4 x i32> @llvm.umax.v4i32(<4 x i32> undef, <4 x i32> undef)
77 %V8i32 = call <8 x i32> @llvm.umax.v8i32(<8 x i32> undef, <8 x i32> undef)
78 %V2i64 = call <2 x i64> @llvm.umax.v2i64(<2 x i64> undef, <2 x i64> undef)
79 %V4i64 = call <4 x i64> @llvm.umax.v4i64(<4 x i64> undef, <4 x i64> undef)
83 define void @reduce_smin() {
84 ; CHECK-LABEL: 'reduce_smin'
85 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V1i8 = call <1 x i8> @llvm.smin.v1i8(<1 x i8> undef, <1 x i8> undef)
86 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V3i8 = call <3 x i8> @llvm.smin.v3i8(<3 x i8> undef, <3 x i8> undef)
87 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4i8 = call <4 x i8> @llvm.smin.v4i8(<4 x i8> undef, <4 x i8> undef)
88 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V8i8 = call <8 x i8> @llvm.smin.v8i8(<8 x i8> undef, <8 x i8> undef)
89 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V16i8 = call <16 x i8> @llvm.smin.v16i8(<16 x i8> undef, <16 x i8> undef)
90 ; CHECK-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V32i8 = call <32 x i8> @llvm.smin.v32i8(<32 x i8> undef, <32 x i8> undef)
91 ; CHECK-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V64i8 = call <64 x i8> @llvm.smin.v64i8(<64 x i8> undef, <64 x i8> undef)
92 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2i16 = call <2 x i16> @llvm.smin.v2i16(<2 x i16> undef, <2 x i16> undef)
93 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4i16 = call <4 x i16> @llvm.smin.v4i16(<4 x i16> undef, <4 x i16> undef)
94 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V8i16 = call <8 x i16> @llvm.smin.v8i16(<8 x i16> undef, <8 x i16> undef)
95 ; CHECK-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V16i16 = call <16 x i16> @llvm.smin.v16i16(<16 x i16> undef, <16 x i16> undef)
96 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2i32 = call <2 x i32> @llvm.smin.v2i32(<2 x i32> undef, <2 x i32> undef)
97 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4i32 = call <4 x i32> @llvm.smin.v4i32(<4 x i32> undef, <4 x i32> undef)
98 ; CHECK-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V8i32 = call <8 x i32> @llvm.smin.v8i32(<8 x i32> undef, <8 x i32> undef)
99 ; CHECK-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V2i64 = call <2 x i64> @llvm.smin.v2i64(<2 x i64> undef, <2 x i64> undef)
100 ; CHECK-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V4i64 = call <4 x i64> @llvm.smin.v4i64(<4 x i64> undef, <4 x i64> undef)
101 ; CHECK-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret void
103 %V1i8 = call <1 x i8> @llvm.smin.v1i8(<1 x i8> undef, <1 x i8> undef)
104 %V3i8 = call <3 x i8> @llvm.smin.v3i8(<3 x i8> undef, <3 x i8> undef)
105 %V4i8 = call <4 x i8> @llvm.smin.v4i8(<4 x i8> undef, <4 x i8> undef)
106 %V8i8 = call <8 x i8> @llvm.smin.v8i8(<8 x i8> undef, <8 x i8> undef)
107 %V16i8 = call <16 x i8> @llvm.smin.v16i8(<16 x i8> undef, <16 x i8> undef)
108 %V32i8 = call <32 x i8> @llvm.smin.v32i8(<32 x i8> undef, <32 x i8> undef)
109 %V64i8 = call <64 x i8> @llvm.smin.v64i8(<64 x i8> undef, <64 x i8> undef)
110 %V2i16 = call <2 x i16> @llvm.smin.v2i16(<2 x i16> undef, <2 x i16> undef)
111 %V4i16 = call <4 x i16> @llvm.smin.v4i16(<4 x i16> undef, <4 x i16> undef)
112 %V8i16 = call <8 x i16> @llvm.smin.v8i16(<8 x i16> undef, <8 x i16> undef)
113 %V16i16 = call <16 x i16> @llvm.smin.v16i16(<16 x i16> undef, <16 x i16> undef)
114 %V2i32 = call <2 x i32> @llvm.smin.v2i32(<2 x i32> undef, <2 x i32> undef)
115 %V4i32 = call <4 x i32> @llvm.smin.v4i32(<4 x i32> undef, <4 x i32> undef)
116 %V8i32 = call <8 x i32> @llvm.smin.v8i32(<8 x i32> undef, <8 x i32> undef)
117 %V2i64 = call <2 x i64> @llvm.smin.v2i64(<2 x i64> undef, <2 x i64> undef)
118 %V4i64 = call <4 x i64> @llvm.smin.v4i64(<4 x i64> undef, <4 x i64> undef)
122 define void @reduce_smax() {
123 ; CHECK-LABEL: 'reduce_smax'
124 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V1i8 = call <1 x i8> @llvm.smax.v1i8(<1 x i8> undef, <1 x i8> undef)
125 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V3i8 = call <3 x i8> @llvm.smax.v3i8(<3 x i8> undef, <3 x i8> undef)
126 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4i8 = call <4 x i8> @llvm.smax.v4i8(<4 x i8> undef, <4 x i8> undef)
127 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V8i8 = call <8 x i8> @llvm.smax.v8i8(<8 x i8> undef, <8 x i8> undef)
128 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V16i8 = call <16 x i8> @llvm.smax.v16i8(<16 x i8> undef, <16 x i8> undef)
129 ; CHECK-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V32i8 = call <32 x i8> @llvm.smax.v32i8(<32 x i8> undef, <32 x i8> undef)
130 ; CHECK-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V64i8 = call <64 x i8> @llvm.smax.v64i8(<64 x i8> undef, <64 x i8> undef)
131 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2i16 = call <2 x i16> @llvm.smax.v2i16(<2 x i16> undef, <2 x i16> undef)
132 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4i16 = call <4 x i16> @llvm.smax.v4i16(<4 x i16> undef, <4 x i16> undef)
133 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V8i16 = call <8 x i16> @llvm.smax.v8i16(<8 x i16> undef, <8 x i16> undef)
134 ; CHECK-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V16i16 = call <16 x i16> @llvm.smax.v16i16(<16 x i16> undef, <16 x i16> undef)
135 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2i32 = call <2 x i32> @llvm.smax.v2i32(<2 x i32> undef, <2 x i32> undef)
136 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4i32 = call <4 x i32> @llvm.smax.v4i32(<4 x i32> undef, <4 x i32> undef)
137 ; CHECK-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V8i32 = call <8 x i32> @llvm.smax.v8i32(<8 x i32> undef, <8 x i32> undef)
138 ; CHECK-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V2i64 = call <2 x i64> @llvm.smax.v2i64(<2 x i64> undef, <2 x i64> undef)
139 ; CHECK-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V4i64 = call <4 x i64> @llvm.smax.v4i64(<4 x i64> undef, <4 x i64> undef)
140 ; CHECK-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret void
142 %V1i8 = call <1 x i8> @llvm.smax.v1i8(<1 x i8> undef, <1 x i8> undef)
143 %V3i8 = call <3 x i8> @llvm.smax.v3i8(<3 x i8> undef, <3 x i8> undef)
144 %V4i8 = call <4 x i8> @llvm.smax.v4i8(<4 x i8> undef, <4 x i8> undef)
145 %V8i8 = call <8 x i8> @llvm.smax.v8i8(<8 x i8> undef, <8 x i8> undef)
146 %V16i8 = call <16 x i8> @llvm.smax.v16i8(<16 x i8> undef, <16 x i8> undef)
147 %V32i8 = call <32 x i8> @llvm.smax.v32i8(<32 x i8> undef, <32 x i8> undef)
148 %V64i8 = call <64 x i8> @llvm.smax.v64i8(<64 x i8> undef, <64 x i8> undef)
149 %V2i16 = call <2 x i16> @llvm.smax.v2i16(<2 x i16> undef, <2 x i16> undef)
150 %V4i16 = call <4 x i16> @llvm.smax.v4i16(<4 x i16> undef, <4 x i16> undef)
151 %V8i16 = call <8 x i16> @llvm.smax.v8i16(<8 x i16> undef, <8 x i16> undef)
152 %V16i16 = call <16 x i16> @llvm.smax.v16i16(<16 x i16> undef, <16 x i16> undef)
153 %V2i32 = call <2 x i32> @llvm.smax.v2i32(<2 x i32> undef, <2 x i32> undef)
154 %V4i32 = call <4 x i32> @llvm.smax.v4i32(<4 x i32> undef, <4 x i32> undef)
155 %V8i32 = call <8 x i32> @llvm.smax.v8i32(<8 x i32> undef, <8 x i32> undef)
156 %V2i64 = call <2 x i64> @llvm.smax.v2i64(<2 x i64> undef, <2 x i64> undef)
157 %V4i64 = call <4 x i64> @llvm.smax.v4i64(<4 x i64> undef, <4 x i64> undef)
161 define void @reduce_minnum16() {
162 ; CHECK-NOF16-LABEL: 'reduce_minnum16'
163 ; CHECK-NOF16-NEXT: Cost Model: Found an estimated cost of 5 for instruction: %V2f16 = call <2 x half> @llvm.minnum.v2f16(<2 x half> undef, <2 x half> undef)
164 ; CHECK-NOF16-NEXT: Cost Model: Found an estimated cost of 13 for instruction: %V4f16 = call <4 x half> @llvm.minnum.v4f16(<4 x half> undef, <4 x half> undef)
165 ; CHECK-NOF16-NEXT: Cost Model: Found an estimated cost of 29 for instruction: %V8f16 = call <8 x half> @llvm.minnum.v8f16(<8 x half> undef, <8 x half> undef)
166 ; CHECK-NOF16-NEXT: Cost Model: Found an estimated cost of 58 for instruction: %V16f16 = call <16 x half> @llvm.minnum.v16f16(<16 x half> undef, <16 x half> undef)
167 ; CHECK-NOF16-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret void
169 ; CHECK-F16-LABEL: 'reduce_minnum16'
170 ; CHECK-F16-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2f16 = call <2 x half> @llvm.minnum.v2f16(<2 x half> undef, <2 x half> undef)
171 ; CHECK-F16-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4f16 = call <4 x half> @llvm.minnum.v4f16(<4 x half> undef, <4 x half> undef)
172 ; CHECK-F16-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V8f16 = call <8 x half> @llvm.minnum.v8f16(<8 x half> undef, <8 x half> undef)
173 ; CHECK-F16-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V16f16 = call <16 x half> @llvm.minnum.v16f16(<16 x half> undef, <16 x half> undef)
174 ; CHECK-F16-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret void
176 %V2f16 = call <2 x half> @llvm.minnum.v2f16(<2 x half> undef, <2 x half> undef)
177 %V4f16 = call <4 x half> @llvm.minnum.v4f16(<4 x half> undef, <4 x half> undef)
178 %V8f16 = call <8 x half> @llvm.minnum.v8f16(<8 x half> undef, <8 x half> undef)
179 %V16f16 = call <16 x half> @llvm.minnum.v16f16(<16 x half> undef, <16 x half> undef)
183 define void @reduce_maxnum16() {
184 ; CHECK-NOF16-LABEL: 'reduce_maxnum16'
185 ; CHECK-NOF16-NEXT: Cost Model: Found an estimated cost of 5 for instruction: %V2f16 = call <2 x half> @llvm.maxnum.v2f16(<2 x half> undef, <2 x half> undef)
186 ; CHECK-NOF16-NEXT: Cost Model: Found an estimated cost of 13 for instruction: %V4f16 = call <4 x half> @llvm.maxnum.v4f16(<4 x half> undef, <4 x half> undef)
187 ; CHECK-NOF16-NEXT: Cost Model: Found an estimated cost of 29 for instruction: %V8f16 = call <8 x half> @llvm.maxnum.v8f16(<8 x half> undef, <8 x half> undef)
188 ; CHECK-NOF16-NEXT: Cost Model: Found an estimated cost of 58 for instruction: %V16f16 = call <16 x half> @llvm.maxnum.v16f16(<16 x half> undef, <16 x half> undef)
189 ; CHECK-NOF16-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret void
191 ; CHECK-F16-LABEL: 'reduce_maxnum16'
192 ; CHECK-F16-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2f16 = call <2 x half> @llvm.maxnum.v2f16(<2 x half> undef, <2 x half> undef)
193 ; CHECK-F16-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4f16 = call <4 x half> @llvm.maxnum.v4f16(<4 x half> undef, <4 x half> undef)
194 ; CHECK-F16-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V8f16 = call <8 x half> @llvm.maxnum.v8f16(<8 x half> undef, <8 x half> undef)
195 ; CHECK-F16-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V16f16 = call <16 x half> @llvm.maxnum.v16f16(<16 x half> undef, <16 x half> undef)
196 ; CHECK-F16-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret void
198 %V2f16 = call <2 x half> @llvm.maxnum.v2f16(<2 x half> undef, <2 x half> undef)
199 %V4f16 = call <4 x half> @llvm.maxnum.v4f16(<4 x half> undef, <4 x half> undef)
200 %V8f16 = call <8 x half> @llvm.maxnum.v8f16(<8 x half> undef, <8 x half> undef)
201 %V16f16 = call <16 x half> @llvm.maxnum.v16f16(<16 x half> undef, <16 x half> undef)
205 define void @reduce_minnum() {
206 ; CHECK-LABEL: 'reduce_minnum'
207 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2f32 = call <2 x float> @llvm.minnum.v2f32(<2 x float> undef, <2 x float> undef)
208 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4f32 = call <4 x float> @llvm.minnum.v4f32(<4 x float> undef, <4 x float> undef)
209 ; CHECK-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V8f32 = call <8 x float> @llvm.minnum.v8f32(<8 x float> undef, <8 x float> undef)
210 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2f64 = call <2 x double> @llvm.minnum.v2f64(<2 x double> undef, <2 x double> undef)
211 ; CHECK-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V4f64 = call <4 x double> @llvm.minnum.v4f64(<4 x double> undef, <4 x double> undef)
212 ; CHECK-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret void
214 %V2f32 = call <2 x float> @llvm.minnum.v2f32(<2 x float> undef, <2 x float> undef)
215 %V4f32 = call <4 x float> @llvm.minnum.v4f32(<4 x float> undef, <4 x float> undef)
216 %V8f32 = call <8 x float> @llvm.minnum.v8f32(<8 x float> undef, <8 x float> undef)
217 %V2f64 = call <2 x double> @llvm.minnum.v2f64(<2 x double> undef, <2 x double> undef)
218 %V4f64 = call <4 x double> @llvm.minnum.v4f64(<4 x double> undef, <4 x double> undef)
222 define void @reduce_maxnum() {
223 ; CHECK-LABEL: 'reduce_maxnum'
224 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2f32 = call <2 x float> @llvm.maxnum.v2f32(<2 x float> undef, <2 x float> undef)
225 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4f32 = call <4 x float> @llvm.maxnum.v4f32(<4 x float> undef, <4 x float> undef)
226 ; CHECK-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V8f32 = call <8 x float> @llvm.maxnum.v8f32(<8 x float> undef, <8 x float> undef)
227 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2f64 = call <2 x double> @llvm.maxnum.v2f64(<2 x double> undef, <2 x double> undef)
228 ; CHECK-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V4f64 = call <4 x double> @llvm.maxnum.v4f64(<4 x double> undef, <4 x double> undef)
229 ; CHECK-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret void
231 %V2f32 = call <2 x float> @llvm.maxnum.v2f32(<2 x float> undef, <2 x float> undef)
232 %V4f32 = call <4 x float> @llvm.maxnum.v4f32(<4 x float> undef, <4 x float> undef)
233 %V8f32 = call <8 x float> @llvm.maxnum.v8f32(<8 x float> undef, <8 x float> undef)
234 %V2f64 = call <2 x double> @llvm.maxnum.v2f64(<2 x double> undef, <2 x double> undef)
235 %V4f64 = call <4 x double> @llvm.maxnum.v4f64(<4 x double> undef, <4 x double> undef)
240 define void @reduce_minimum16() {
241 ; CHECK-NOF16-LABEL: 'reduce_minimum16'
242 ; CHECK-NOF16-NEXT: Cost Model: Found an estimated cost of 5 for instruction: %V2f16 = call <2 x half> @llvm.minimum.v2f16(<2 x half> undef, <2 x half> undef)
243 ; CHECK-NOF16-NEXT: Cost Model: Found an estimated cost of 13 for instruction: %V4f16 = call <4 x half> @llvm.minimum.v4f16(<4 x half> undef, <4 x half> undef)
244 ; CHECK-NOF16-NEXT: Cost Model: Found an estimated cost of 29 for instruction: %V8f16 = call <8 x half> @llvm.minimum.v8f16(<8 x half> undef, <8 x half> undef)
245 ; CHECK-NOF16-NEXT: Cost Model: Found an estimated cost of 58 for instruction: %V16f16 = call <16 x half> @llvm.minimum.v16f16(<16 x half> undef, <16 x half> undef)
246 ; CHECK-NOF16-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret void
248 ; CHECK-F16-LABEL: 'reduce_minimum16'
249 ; CHECK-F16-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2f16 = call <2 x half> @llvm.minimum.v2f16(<2 x half> undef, <2 x half> undef)
250 ; CHECK-F16-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4f16 = call <4 x half> @llvm.minimum.v4f16(<4 x half> undef, <4 x half> undef)
251 ; CHECK-F16-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V8f16 = call <8 x half> @llvm.minimum.v8f16(<8 x half> undef, <8 x half> undef)
252 ; CHECK-F16-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V16f16 = call <16 x half> @llvm.minimum.v16f16(<16 x half> undef, <16 x half> undef)
253 ; CHECK-F16-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret void
255 %V2f16 = call <2 x half> @llvm.minimum.v2f16(<2 x half> undef, <2 x half> undef)
256 %V4f16 = call <4 x half> @llvm.minimum.v4f16(<4 x half> undef, <4 x half> undef)
257 %V8f16 = call <8 x half> @llvm.minimum.v8f16(<8 x half> undef, <8 x half> undef)
258 %V16f16 = call <16 x half> @llvm.minimum.v16f16(<16 x half> undef, <16 x half> undef)
262 define void @reduce_maximum16() {
263 ; CHECK-NOF16-LABEL: 'reduce_maximum16'
264 ; CHECK-NOF16-NEXT: Cost Model: Found an estimated cost of 5 for instruction: %V2f16 = call <2 x half> @llvm.maximum.v2f16(<2 x half> undef, <2 x half> undef)
265 ; CHECK-NOF16-NEXT: Cost Model: Found an estimated cost of 13 for instruction: %V4f16 = call <4 x half> @llvm.maximum.v4f16(<4 x half> undef, <4 x half> undef)
266 ; CHECK-NOF16-NEXT: Cost Model: Found an estimated cost of 29 for instruction: %V8f16 = call <8 x half> @llvm.maximum.v8f16(<8 x half> undef, <8 x half> undef)
267 ; CHECK-NOF16-NEXT: Cost Model: Found an estimated cost of 58 for instruction: %V16f16 = call <16 x half> @llvm.maximum.v16f16(<16 x half> undef, <16 x half> undef)
268 ; CHECK-NOF16-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret void
270 ; CHECK-F16-LABEL: 'reduce_maximum16'
271 ; CHECK-F16-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2f16 = call <2 x half> @llvm.maximum.v2f16(<2 x half> undef, <2 x half> undef)
272 ; CHECK-F16-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4f16 = call <4 x half> @llvm.maximum.v4f16(<4 x half> undef, <4 x half> undef)
273 ; CHECK-F16-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V8f16 = call <8 x half> @llvm.maximum.v8f16(<8 x half> undef, <8 x half> undef)
274 ; CHECK-F16-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V16f16 = call <16 x half> @llvm.maximum.v16f16(<16 x half> undef, <16 x half> undef)
275 ; CHECK-F16-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret void
277 %V2f16 = call <2 x half> @llvm.maximum.v2f16(<2 x half> undef, <2 x half> undef)
278 %V4f16 = call <4 x half> @llvm.maximum.v4f16(<4 x half> undef, <4 x half> undef)
279 %V8f16 = call <8 x half> @llvm.maximum.v8f16(<8 x half> undef, <8 x half> undef)
280 %V16f16 = call <16 x half> @llvm.maximum.v16f16(<16 x half> undef, <16 x half> undef)
284 define void @reduce_minimum() {
285 ; CHECK-LABEL: 'reduce_minimum'
286 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2f32 = call <2 x float> @llvm.minimum.v2f32(<2 x float> undef, <2 x float> undef)
287 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4f32 = call <4 x float> @llvm.minimum.v4f32(<4 x float> undef, <4 x float> undef)
288 ; CHECK-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V8f32 = call <8 x float> @llvm.minimum.v8f32(<8 x float> undef, <8 x float> undef)
289 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2f64 = call <2 x double> @llvm.minimum.v2f64(<2 x double> undef, <2 x double> undef)
290 ; CHECK-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V4f64 = call <4 x double> @llvm.minimum.v4f64(<4 x double> undef, <4 x double> undef)
291 ; CHECK-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret void
293 %V2f32 = call <2 x float> @llvm.minimum.v2f32(<2 x float> undef, <2 x float> undef)
294 %V4f32 = call <4 x float> @llvm.minimum.v4f32(<4 x float> undef, <4 x float> undef)
295 %V8f32 = call <8 x float> @llvm.minimum.v8f32(<8 x float> undef, <8 x float> undef)
296 %V2f64 = call <2 x double> @llvm.minimum.v2f64(<2 x double> undef, <2 x double> undef)
297 %V4f64 = call <4 x double> @llvm.minimum.v4f64(<4 x double> undef, <4 x double> undef)
301 define void @reduce_maximum() {
302 ; CHECK-LABEL: 'reduce_maximum'
303 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2f32 = call <2 x float> @llvm.maximum.v2f32(<2 x float> undef, <2 x float> undef)
304 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V4f32 = call <4 x float> @llvm.maximum.v4f32(<4 x float> undef, <4 x float> undef)
305 ; CHECK-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V8f32 = call <8 x float> @llvm.maximum.v8f32(<8 x float> undef, <8 x float> undef)
306 ; CHECK-NEXT: Cost Model: Found an estimated cost of 1 for instruction: %V2f64 = call <2 x double> @llvm.maximum.v2f64(<2 x double> undef, <2 x double> undef)
307 ; CHECK-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V4f64 = call <4 x double> @llvm.maximum.v4f64(<4 x double> undef, <4 x double> undef)
308 ; CHECK-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret void
310 %V2f32 = call <2 x float> @llvm.maximum.v2f32(<2 x float> undef, <2 x float> undef)
311 %V4f32 = call <4 x float> @llvm.maximum.v4f32(<4 x float> undef, <4 x float> undef)
312 %V8f32 = call <8 x float> @llvm.maximum.v8f32(<8 x float> undef, <8 x float> undef)
313 %V2f64 = call <2 x double> @llvm.maximum.v2f64(<2 x double> undef, <2 x double> undef)
314 %V4f64 = call <4 x double> @llvm.maximum.v4f64(<4 x double> undef, <4 x double> undef)
318 declare <1 x i8> @llvm.umin.v1i8(<1 x i8>, <1 x i8>)
319 declare <3 x i8> @llvm.umin.v3i8(<3 x i8>, <3 x i8>)
320 declare <4 x i8> @llvm.umin.v4i8(<4 x i8>, <4 x i8>)
321 declare <8 x i8> @llvm.umin.v8i8(<8 x i8>, <8 x i8>)
322 declare <16 x i8> @llvm.umin.v16i8(<16 x i8>, <16 x i8>)
323 declare <32 x i8> @llvm.umin.v32i8(<32 x i8>, <32 x i8>)
324 declare <64 x i8> @llvm.umin.v64i8(<64 x i8>, <64 x i8>)
325 declare <2 x i16> @llvm.umin.v2i16(<2 x i16>, <2 x i16>)
326 declare <4 x i16> @llvm.umin.v4i16(<4 x i16>, <4 x i16>)
327 declare <8 x i16> @llvm.umin.v8i16(<8 x i16>, <8 x i16>)
328 declare <16 x i16> @llvm.umin.v16i16(<16 x i16>, <16 x i16>)
329 declare <2 x i32> @llvm.umin.v2i32(<2 x i32>, <2 x i32>)
330 declare <4 x i32> @llvm.umin.v4i32(<4 x i32>, <4 x i32>)
331 declare <8 x i32> @llvm.umin.v8i32(<8 x i32>, <8 x i32>)
332 declare <2 x i64> @llvm.umin.v2i64(<2 x i64>, <2 x i64>)
333 declare <4 x i64> @llvm.umin.v4i64(<4 x i64>, <4 x i64>)
335 declare <1 x i8> @llvm.umax.v1i8(<1 x i8>, <1 x i8>)
336 declare <3 x i8> @llvm.umax.v3i8(<3 x i8>, <3 x i8>)
337 declare <4 x i8> @llvm.umax.v4i8(<4 x i8>, <4 x i8>)
338 declare <8 x i8> @llvm.umax.v8i8(<8 x i8>, <8 x i8>)
339 declare <16 x i8> @llvm.umax.v16i8(<16 x i8>, <16 x i8>)
340 declare <32 x i8> @llvm.umax.v32i8(<32 x i8>, <32 x i8>)
341 declare <64 x i8> @llvm.umax.v64i8(<64 x i8>, <64 x i8>)
342 declare <2 x i16> @llvm.umax.v2i16(<2 x i16>, <2 x i16>)
343 declare <4 x i16> @llvm.umax.v4i16(<4 x i16>, <4 x i16>)
344 declare <8 x i16> @llvm.umax.v8i16(<8 x i16>, <8 x i16>)
345 declare <16 x i16> @llvm.umax.v16i16(<16 x i16>, <16 x i16>)
346 declare <2 x i32> @llvm.umax.v2i32(<2 x i32>, <2 x i32>)
347 declare <4 x i32> @llvm.umax.v4i32(<4 x i32>, <4 x i32>)
348 declare <8 x i32> @llvm.umax.v8i32(<8 x i32>, <8 x i32>)
349 declare <2 x i64> @llvm.umax.v2i64(<2 x i64>, <2 x i64>)
350 declare <4 x i64> @llvm.umax.v4i64(<4 x i64>, <4 x i64>)
352 declare <1 x i8> @llvm.smin.v1i8(<1 x i8>, <1 x i8>)
353 declare <3 x i8> @llvm.smin.v3i8(<3 x i8>, <3 x i8>)
354 declare <4 x i8> @llvm.smin.v4i8(<4 x i8>, <4 x i8>)
355 declare <8 x i8> @llvm.smin.v8i8(<8 x i8>, <8 x i8>)
356 declare <16 x i8> @llvm.smin.v16i8(<16 x i8>, <16 x i8>)
357 declare <32 x i8> @llvm.smin.v32i8(<32 x i8>, <32 x i8>)
358 declare <64 x i8> @llvm.smin.v64i8(<64 x i8>, <64 x i8>)
359 declare <2 x i16> @llvm.smin.v2i16(<2 x i16>, <2 x i16>)
360 declare <4 x i16> @llvm.smin.v4i16(<4 x i16>, <4 x i16>)
361 declare <8 x i16> @llvm.smin.v8i16(<8 x i16>, <8 x i16>)
362 declare <16 x i16> @llvm.smin.v16i16(<16 x i16>, <16 x i16>)
363 declare <2 x i32> @llvm.smin.v2i32(<2 x i32>, <2 x i32>)
364 declare <4 x i32> @llvm.smin.v4i32(<4 x i32>, <4 x i32>)
365 declare <8 x i32> @llvm.smin.v8i32(<8 x i32>, <8 x i32>)
366 declare <2 x i64> @llvm.smin.v2i64(<2 x i64>, <2 x i64>)
367 declare <4 x i64> @llvm.smin.v4i64(<4 x i64>, <4 x i64>)
369 declare <1 x i8> @llvm.smax.v1i8(<1 x i8>, <1 x i8>)
370 declare <3 x i8> @llvm.smax.v3i8(<3 x i8>, <3 x i8>)
371 declare <4 x i8> @llvm.smax.v4i8(<4 x i8>, <4 x i8>)
372 declare <8 x i8> @llvm.smax.v8i8(<8 x i8>, <8 x i8>)
373 declare <16 x i8> @llvm.smax.v16i8(<16 x i8>, <16 x i8>)
374 declare <32 x i8> @llvm.smax.v32i8(<32 x i8>, <32 x i8>)
375 declare <64 x i8> @llvm.smax.v64i8(<64 x i8>, <64 x i8>)
376 declare <2 x i16> @llvm.smax.v2i16(<2 x i16>, <2 x i16>)
377 declare <4 x i16> @llvm.smax.v4i16(<4 x i16>, <4 x i16>)
378 declare <8 x i16> @llvm.smax.v8i16(<8 x i16>, <8 x i16>)
379 declare <16 x i16> @llvm.smax.v16i16(<16 x i16>, <16 x i16>)
380 declare <2 x i32> @llvm.smax.v2i32(<2 x i32>, <2 x i32>)
381 declare <4 x i32> @llvm.smax.v4i32(<4 x i32>, <4 x i32>)
382 declare <8 x i32> @llvm.smax.v8i32(<8 x i32>, <8 x i32>)
383 declare <2 x i64> @llvm.smax.v2i64(<2 x i64>, <2 x i64>)
384 declare <4 x i64> @llvm.smax.v4i64(<4 x i64>, <4 x i64>)
386 declare <2 x half> @llvm.minnum.v2f16(<2 x half>, <2 x half>)
387 declare <4 x half> @llvm.minnum.v4f16(<4 x half>, <4 x half>)
388 declare <8 x half> @llvm.minnum.v8f16(<8 x half>, <8 x half>)
389 declare <16 x half> @llvm.minnum.v16f16(<16 x half>, <16 x half>)
390 declare <2 x float> @llvm.minnum.v2f32(<2 x float>, <2 x float>)
391 declare <4 x float> @llvm.minnum.v4f32(<4 x float>, <4 x float>)
392 declare <8 x float> @llvm.minnum.v8f32(<8 x float>, <8 x float>)
393 declare <2 x double> @llvm.minnum.v2f64(<2 x double>, <2 x double>)
394 declare <4 x double> @llvm.minnum.v4f64(<4 x double>, <4 x double>)
396 declare <2 x half> @llvm.maxnum.v2f16(<2 x half>, <2 x half>)
397 declare <4 x half> @llvm.maxnum.v4f16(<4 x half>, <4 x half>)
398 declare <8 x half> @llvm.maxnum.v8f16(<8 x half>, <8 x half>)
399 declare <16 x half> @llvm.maxnum.v16f16(<16 x half>, <16 x half>)
400 declare <2 x float> @llvm.maxnum.v2f32(<2 x float>, <2 x float>)
401 declare <4 x float> @llvm.maxnum.v4f32(<4 x float>, <4 x float>)
402 declare <8 x float> @llvm.maxnum.v8f32(<8 x float>, <8 x float>)
403 declare <2 x double> @llvm.maxnum.v2f64(<2 x double>, <2 x double>)
404 declare <4 x double> @llvm.maxnum.v4f64(<4 x double>, <4 x double>)
406 declare <2 x half> @llvm.minimum.v2f16(<2 x half>, <2 x half>)
407 declare <4 x half> @llvm.minimum.v4f16(<4 x half>, <4 x half>)
408 declare <8 x half> @llvm.minimum.v8f16(<8 x half>, <8 x half>)
409 declare <16 x half> @llvm.minimum.v16f16(<16 x half>, <16 x half>)
410 declare <2 x float> @llvm.minimum.v2f32(<2 x float>, <2 x float>)
411 declare <4 x float> @llvm.minimum.v4f32(<4 x float>, <4 x float>)
412 declare <8 x float> @llvm.minimum.v8f32(<8 x float>, <8 x float>)
413 declare <2 x double> @llvm.minimum.v2f64(<2 x double>, <2 x double>)
414 declare <4 x double> @llvm.minimum.v4f64(<4 x double>, <4 x double>)
416 declare <2 x half> @llvm.maximum.v2f16(<2 x half>, <2 x half>)
417 declare <4 x half> @llvm.maximum.v4f16(<4 x half>, <4 x half>)
418 declare <8 x half> @llvm.maximum.v8f16(<8 x half>, <8 x half>)
419 declare <16 x half> @llvm.maximum.v16f16(<16 x half>, <16 x half>)
420 declare <2 x float> @llvm.maximum.v2f32(<2 x float>, <2 x float>)
421 declare <4 x float> @llvm.maximum.v4f32(<4 x float>, <4 x float>)
422 declare <8 x float> @llvm.maximum.v8f32(<8 x float>, <8 x float>)
423 declare <2 x double> @llvm.maximum.v2f64(<2 x double>, <2 x double>)
424 declare <4 x double> @llvm.maximum.v4f64(<4 x double>, <4 x double>)