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
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
|
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc -mtriple=thumbv8.1m.main-none-none-eabi -mattr=+mve.fp,+fp64 -verify-machineinstrs %s -o - | FileCheck %s
; i32
define ptr @vld2_v4i32(ptr %src, ptr %dst) {
; CHECK-LABEL: vld2_v4i32:
; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: vld20.32 {q0, q1}, [r0]
; CHECK-NEXT: vld21.32 {q0, q1}, [r0]!
; CHECK-NEXT: vadd.i32 q0, q0, q1
; CHECK-NEXT: vstrw.32 q0, [r1]
; CHECK-NEXT: bx lr
entry:
%l1 = load <8 x i32>, ptr %src, align 4
%s1 = shufflevector <8 x i32> %l1, <8 x i32> undef, <4 x i32> <i32 0, i32 2, i32 4, i32 6>
%s2 = shufflevector <8 x i32> %l1, <8 x i32> undef, <4 x i32> <i32 1, i32 3, i32 5, i32 7>
%a = add <4 x i32> %s1, %s2
store <4 x i32> %a, ptr %dst
%ret = getelementptr inbounds <8 x i32>, ptr %src, i32 1
ret ptr %ret
}
; i16
define ptr @vld2_v8i16(ptr %src, ptr %dst) {
; CHECK-LABEL: vld2_v8i16:
; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: vld20.16 {q0, q1}, [r0]
; CHECK-NEXT: vld21.16 {q0, q1}, [r0]!
; CHECK-NEXT: vadd.i16 q0, q0, q1
; CHECK-NEXT: vstrw.32 q0, [r1]
; CHECK-NEXT: bx lr
entry:
%l1 = load <16 x i16>, ptr %src, align 4
%s1 = shufflevector <16 x i16> %l1, <16 x i16> undef, <8 x i32> <i32 0, i32 2, i32 4, i32 6, i32 8, i32 10, i32 12, i32 14>
%s2 = shufflevector <16 x i16> %l1, <16 x i16> undef, <8 x i32> <i32 1, i32 3, i32 5, i32 7, i32 9, i32 11, i32 13, i32 15>
%a = add <8 x i16> %s1, %s2
store <8 x i16> %a, ptr %dst
%ret = getelementptr inbounds <16 x i16>, ptr %src, i32 1
ret ptr %ret
}
; i8
define ptr @vld2_v16i8(ptr %src, ptr %dst) {
; CHECK-LABEL: vld2_v16i8:
; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: vld20.8 {q0, q1}, [r0]
; CHECK-NEXT: vld21.8 {q0, q1}, [r0]!
; CHECK-NEXT: vadd.i8 q0, q0, q1
; CHECK-NEXT: vstrw.32 q0, [r1]
; CHECK-NEXT: bx lr
entry:
%l1 = load <32 x i8>, ptr %src, align 4
%s1 = shufflevector <32 x i8> %l1, <32 x i8> undef, <16 x i32> <i32 0, i32 2, i32 4, i32 6, i32 8, i32 10, i32 12, i32 14, i32 16, i32 18, i32 20, i32 22, i32 24, i32 26, i32 28, i32 30>
%s2 = shufflevector <32 x i8> %l1, <32 x i8> undef, <16 x i32> <i32 1, i32 3, i32 5, i32 7, i32 9, i32 11, i32 13, i32 15, i32 17, i32 19, i32 21, i32 23, i32 25, i32 27, i32 29, i32 31>
%a = add <16 x i8> %s1, %s2
store <16 x i8> %a, ptr %dst
%ret = getelementptr inbounds <32 x i8>, ptr %src, i32 1
ret ptr %ret
}
; i64
define ptr @vld2_v2i64(ptr %src, ptr %dst) {
; CHECK-LABEL: vld2_v2i64:
; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: .save {r4, r5, r6, r7, lr}
; CHECK-NEXT: push {r4, r5, r6, r7, lr}
; CHECK-NEXT: vldrw.u32 q0, [r0, #16]
; CHECK-NEXT: vmov r2, r12, d1
; CHECK-NEXT: vmov r3, lr, d0
; CHECK-NEXT: vldrw.u32 q0, [r0], #32
; CHECK-NEXT: vmov r4, r7, d1
; CHECK-NEXT: adds r2, r2, r3
; CHECK-NEXT: vmov r3, r6, d0
; CHECK-NEXT: adc.w r5, lr, r12
; CHECK-NEXT: adds r3, r3, r4
; CHECK-NEXT: adcs r7, r6
; CHECK-NEXT: vmov q0[2], q0[0], r3, r2
; CHECK-NEXT: vmov q0[3], q0[1], r7, r5
; CHECK-NEXT: vstrw.32 q0, [r1]
; CHECK-NEXT: pop {r4, r5, r6, r7, pc}
entry:
%l1 = load <4 x i64>, ptr %src, align 4
%s1 = shufflevector <4 x i64> %l1, <4 x i64> undef, <2 x i32> <i32 0, i32 2>
%s2 = shufflevector <4 x i64> %l1, <4 x i64> undef, <2 x i32> <i32 1, i32 3>
%a = add <2 x i64> %s1, %s2
store <2 x i64> %a, ptr %dst
%ret = getelementptr inbounds <4 x i64>, ptr %src, i32 1
ret ptr %ret
}
; f32
define ptr @vld2_v4f32(ptr %src, ptr %dst) {
; CHECK-LABEL: vld2_v4f32:
; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: vld20.32 {q0, q1}, [r0]
; CHECK-NEXT: vld21.32 {q0, q1}, [r0]!
; CHECK-NEXT: vadd.f32 q0, q0, q1
; CHECK-NEXT: vstrw.32 q0, [r1]
; CHECK-NEXT: bx lr
entry:
%l1 = load <8 x float>, ptr %src, align 4
%s1 = shufflevector <8 x float> %l1, <8 x float> undef, <4 x i32> <i32 0, i32 2, i32 4, i32 6>
%s2 = shufflevector <8 x float> %l1, <8 x float> undef, <4 x i32> <i32 1, i32 3, i32 5, i32 7>
%a = fadd <4 x float> %s1, %s2
store <4 x float> %a, ptr %dst
%ret = getelementptr inbounds <8 x float>, ptr %src, i32 1
ret ptr %ret
}
; f16
define ptr @vld2_v8f16(ptr %src, ptr %dst) {
; CHECK-LABEL: vld2_v8f16:
; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: vld20.16 {q0, q1}, [r0]
; CHECK-NEXT: vld21.16 {q0, q1}, [r0]!
; CHECK-NEXT: vadd.f16 q0, q0, q1
; CHECK-NEXT: vstrw.32 q0, [r1]
; CHECK-NEXT: bx lr
entry:
%l1 = load <16 x half>, ptr %src, align 4
%s1 = shufflevector <16 x half> %l1, <16 x half> undef, <8 x i32> <i32 0, i32 2, i32 4, i32 6, i32 8, i32 10, i32 12, i32 14>
%s2 = shufflevector <16 x half> %l1, <16 x half> undef, <8 x i32> <i32 1, i32 3, i32 5, i32 7, i32 9, i32 11, i32 13, i32 15>
%a = fadd <8 x half> %s1, %s2
store <8 x half> %a, ptr %dst
%ret = getelementptr inbounds <16 x half>, ptr %src, i32 1
ret ptr %ret
}
; f64
define ptr @vld2_v2f64(ptr %src, ptr %dst) {
; CHECK-LABEL: vld2_v2f64:
; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: vldrw.u32 q0, [r0, #16]
; CHECK-NEXT: vldrw.u32 q1, [r0], #32
; CHECK-NEXT: vadd.f64 d1, d0, d1
; CHECK-NEXT: vadd.f64 d0, d2, d3
; CHECK-NEXT: vstrw.32 q0, [r1]
; CHECK-NEXT: bx lr
entry:
%l1 = load <4 x double>, ptr %src, align 4
%s1 = shufflevector <4 x double> %l1, <4 x double> undef, <2 x i32> <i32 0, i32 2>
%s2 = shufflevector <4 x double> %l1, <4 x double> undef, <2 x i32> <i32 1, i32 3>
%a = fadd <2 x double> %s1, %s2
store <2 x double> %a, ptr %dst
%ret = getelementptr inbounds <4 x double>, ptr %src, i32 1
ret ptr %ret
}
|
