mve-ctpop.ll
5.12 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
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
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; NOTE: Assertions have been autoenerated by utils/update_llc_test_checks.py
; RUN: llc -mtriple=thumbv8.1m.main-arm-none-eabi -mattr=+mve.fp -verify-machineinstrs %s -o - | FileCheck %s --check-prefix=CHECK
define arm_aapcs_vfpcc <2 x i64> @ctpop_2i64_t(<2 x i64> %src){
; CHECK-LABEL: ctpop_2i64_t:
; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: .save {r4, lr}
; CHECK-NEXT: push {r4, lr}
; CHECK-NEXT: vmov r0, s3
; CHECK-NEXT: mov.w r1, #1431655765
; CHECK-NEXT: mov.w lr, #858993459
; CHECK-NEXT: mov.w r4, #16843009
; CHECK-NEXT: and.w r2, r1, r0, lsr #1
; CHECK-NEXT: subs r0, r0, r2
; CHECK-NEXT: and.w r3, lr, r0, lsr #2
; CHECK-NEXT: bic r0, r0, #-858993460
; CHECK-NEXT: add r0, r3
; CHECK-NEXT: vmov r3, s2
; CHECK-NEXT: add.w r0, r0, r0, lsr #4
; CHECK-NEXT: bic r12, r0, #-252645136
; CHECK-NEXT: and.w r0, r1, r3, lsr #1
; CHECK-NEXT: subs r0, r3, r0
; CHECK-NEXT: and.w r3, lr, r0, lsr #2
; CHECK-NEXT: bic r0, r0, #-858993460
; CHECK-NEXT: add r0, r3
; CHECK-NEXT: vmov r3, s0
; CHECK-NEXT: add.w r0, r0, r0, lsr #4
; CHECK-NEXT: bic r0, r0, #-252645136
; CHECK-NEXT: muls r0, r4, r0
; CHECK-NEXT: lsrs r0, r0, #24
; CHECK-NEXT: and.w r2, r1, r3, lsr #1
; CHECK-NEXT: subs r2, r3, r2
; CHECK-NEXT: and.w r3, lr, r2, lsr #2
; CHECK-NEXT: bic r2, r2, #-858993460
; CHECK-NEXT: add r2, r3
; CHECK-NEXT: vmov r3, s1
; CHECK-NEXT: vldr s1, .LCPI0_0
; CHECK-NEXT: add.w r2, r2, r2, lsr #4
; CHECK-NEXT: bic r2, r2, #-252645136
; CHECK-NEXT: muls r2, r4, r2
; CHECK-NEXT: lsrs r2, r2, #24
; CHECK-NEXT: and.w r1, r1, r3, lsr #1
; CHECK-NEXT: subs r1, r3, r1
; CHECK-NEXT: and.w r3, lr, r1, lsr #2
; CHECK-NEXT: bic r1, r1, #-858993460
; CHECK-NEXT: add r1, r3
; CHECK-NEXT: mul r3, r12, r4
; CHECK-NEXT: add.w r1, r1, r1, lsr #4
; CHECK-NEXT: bic r1, r1, #-252645136
; CHECK-NEXT: muls r1, r4, r1
; CHECK-NEXT: add.w r0, r0, r3, lsr #24
; CHECK-NEXT: vmov s2, r0
; CHECK-NEXT: add.w r0, r2, r1, lsr #24
; CHECK-NEXT: vmov s0, r0
; CHECK-NEXT: vmov.f32 s3, s1
; CHECK-NEXT: pop {r4, pc}
; CHECK-NEXT: .p2align 2
; CHECK-NEXT: @ %bb.1:
; CHECK-NEXT: .LCPI0_0:
; CHECK-NEXT: .long 0 @ float 0
entry:
%0 = call <2 x i64> @llvm.ctpop.v2i64(<2 x i64> %src)
ret <2 x i64> %0
}
define arm_aapcs_vfpcc <4 x i32> @ctpop_4i32_t(<4 x i32> %src){
; CHECK-LABEL: ctpop_4i32_t:
; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: .vsave {d8, d9, d10, d11}
; CHECK-NEXT: vpush {d8, d9, d10, d11}
; CHECK-NEXT: vmov.i8 q4, #0x55
; CHECK-NEXT: vshr.u32 q5, q0, #1
; CHECK-NEXT: vand q4, q5, q4
; CHECK-NEXT: vmov.i8 q3, #0x33
; CHECK-NEXT: vsub.i32 q0, q0, q4
; CHECK-NEXT: vmov.i8 q2, #0xf
; CHECK-NEXT: vshr.u32 q4, q0, #2
; CHECK-NEXT: vand q0, q0, q3
; CHECK-NEXT: vand q4, q4, q3
; CHECK-NEXT: vmov.i8 q1, #0x1
; CHECK-NEXT: vadd.i32 q0, q0, q4
; CHECK-NEXT: vshr.u32 q3, q0, #4
; CHECK-NEXT: vadd.i32 q0, q0, q3
; CHECK-NEXT: vand q0, q0, q2
; CHECK-NEXT: vmul.i32 q0, q0, q1
; CHECK-NEXT: vshr.u32 q0, q0, #24
; CHECK-NEXT: vpop {d8, d9, d10, d11}
; CHECK-NEXT: bx lr
entry:
%0 = call <4 x i32> @llvm.ctpop.v4i32(<4 x i32> %src)
ret <4 x i32> %0
}
define arm_aapcs_vfpcc <8 x i16> @ctpop_8i16_t(<8 x i16> %src){
; CHECK-LABEL: ctpop_8i16_t:
; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: .vsave {d8, d9, d10, d11}
; CHECK-NEXT: vpush {d8, d9, d10, d11}
; CHECK-NEXT: vmov.i8 q4, #0x55
; CHECK-NEXT: vshr.u16 q5, q0, #1
; CHECK-NEXT: vand q4, q5, q4
; CHECK-NEXT: vmov.i8 q3, #0x33
; CHECK-NEXT: vsub.i16 q0, q0, q4
; CHECK-NEXT: vmov.i8 q2, #0xf
; CHECK-NEXT: vshr.u16 q4, q0, #2
; CHECK-NEXT: vand q0, q0, q3
; CHECK-NEXT: vand q4, q4, q3
; CHECK-NEXT: vmov.i8 q1, #0x1
; CHECK-NEXT: vadd.i16 q0, q0, q4
; CHECK-NEXT: vshr.u16 q3, q0, #4
; CHECK-NEXT: vadd.i16 q0, q0, q3
; CHECK-NEXT: vand q0, q0, q2
; CHECK-NEXT: vmul.i16 q0, q0, q1
; CHECK-NEXT: vshr.u16 q0, q0, #8
; CHECK-NEXT: vpop {d8, d9, d10, d11}
; CHECK-NEXT: bx lr
entry:
%0 = call <8 x i16> @llvm.ctpop.v8i16(<8 x i16> %src)
ret <8 x i16> %0
}
define arm_aapcs_vfpcc <16 x i8> @ctpop_16i8_t(<16 x i8> %src){
; CHECK-LABEL: ctpop_16i8_t:
; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: .vsave {d8, d9}
; CHECK-NEXT: vpush {d8, d9}
; CHECK-NEXT: vmov.i8 q3, #0x55
; CHECK-NEXT: vshr.u8 q4, q0, #1
; CHECK-NEXT: vand q3, q4, q3
; CHECK-NEXT: vmov.i8 q2, #0x33
; CHECK-NEXT: vsub.i8 q0, q0, q3
; CHECK-NEXT: vmov.i8 q1, #0xf
; CHECK-NEXT: vshr.u8 q3, q0, #2
; CHECK-NEXT: vand q0, q0, q2
; CHECK-NEXT: vand q3, q3, q2
; CHECK-NEXT: vadd.i8 q0, q0, q3
; CHECK-NEXT: vshr.u8 q2, q0, #4
; CHECK-NEXT: vadd.i8 q0, q0, q2
; CHECK-NEXT: vand q0, q0, q1
; CHECK-NEXT: vpop {d8, d9}
; CHECK-NEXT: bx lr
entry:
%0 = call <16 x i8> @llvm.ctpop.v16i8(<16 x i8> %src)
ret <16 x i8> %0
}
declare <2 x i64> @llvm.ctpop.v2i64(<2 x i64>)
declare <4 x i32> @llvm.ctpop.v4i32(<4 x i32>)
declare <8 x i16> @llvm.ctpop.v8i16(<8 x i16>)
declare <16 x i8> @llvm.ctpop.v16i8(<16 x i8>)