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
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
|
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc < %s -mtriple=aarch64-none-linux-gnu -mattr=+sve | FileCheck %s
define i64 @select_or_reduce_v2i1(ptr nocapture noundef readonly %src) {
; CHECK-LABEL: select_or_reduce_v2i1:
; CHECK: // %bb.0: // %entry
; CHECK-NEXT: mov x8, xzr
; CHECK-NEXT: .LBB0_1: // %vector.body
; CHECK-NEXT: // =>This Inner Loop Header: Depth=1
; CHECK-NEXT: ldr q0, [x0, x8]
; CHECK-NEXT: cmeq v0.2d, v0.2d, #0
; CHECK-NEXT: umaxv s0, v0.4s
; CHECK-NEXT: fmov w9, s0
; CHECK-NEXT: tbnz w9, #0, .LBB0_3
; CHECK-NEXT: // %bb.2: // %vector.body
; CHECK-NEXT: // in Loop: Header=BB0_1 Depth=1
; CHECK-NEXT: cmp x8, #16
; CHECK-NEXT: add x8, x8, #16
; CHECK-NEXT: b.ne .LBB0_1
; CHECK-NEXT: .LBB0_3: // %middle.split
; CHECK-NEXT: and x0, x9, #0x1
; CHECK-NEXT: ret
entry:
br label %vector.body
vector.body:
%index = phi i64 [ 0, %entry ], [ %index.next, %vector.body ]
%arrayidx = getelementptr inbounds ptr, ptr %src, i64 %index
%wide.load = load <2 x ptr>, ptr %arrayidx, align 8
%cond = icmp eq <2 x ptr> %wide.load, splat(ptr zeroinitializer)
%index.next = add nuw i64 %index, 2
%or.reduc = tail call i1 @llvm.vector.reduce.or.v2i1(<2 x i1> %cond)
%iv.cmp = icmp eq i64 %index.next, 4
%exit.cond = or i1 %or.reduc, %iv.cmp
br i1 %exit.cond, label %middle.split, label %vector.body
middle.split:
%sel = select i1 %or.reduc, i64 1, i64 0
ret i64 %sel
}
define i64 @br_or_reduce_v2i1(ptr nocapture noundef readonly %src, ptr noundef readnone %p) {
; CHECK-LABEL: br_or_reduce_v2i1:
; CHECK: // %bb.0: // %entry
; CHECK-NEXT: mov x8, xzr
; CHECK-NEXT: .LBB1_1: // %vector.body
; CHECK-NEXT: // =>This Inner Loop Header: Depth=1
; CHECK-NEXT: ldr q0, [x0, x8]
; CHECK-NEXT: cmeq v0.2d, v0.2d, #0
; CHECK-NEXT: umaxv s0, v0.4s
; CHECK-NEXT: fmov w9, s0
; CHECK-NEXT: tbnz w9, #0, .LBB1_3
; CHECK-NEXT: // %bb.2: // %vector.body
; CHECK-NEXT: // in Loop: Header=BB1_1 Depth=1
; CHECK-NEXT: cmp x8, #16
; CHECK-NEXT: add x8, x8, #16
; CHECK-NEXT: b.ne .LBB1_1
; CHECK-NEXT: .LBB1_3: // %middle.split
; CHECK-NEXT: tbz w9, #0, .LBB1_5
; CHECK-NEXT: // %bb.4: // %found
; CHECK-NEXT: mov w8, #56 // =0x38
; CHECK-NEXT: mov w0, #1 // =0x1
; CHECK-NEXT: str x8, [x1]
; CHECK-NEXT: ret
; CHECK-NEXT: .LBB1_5:
; CHECK-NEXT: mov x0, xzr
; CHECK-NEXT: ret
entry:
br label %vector.body
vector.body:
%index = phi i64 [ 0, %entry ], [ %index.next, %vector.body ]
%arrayidx = getelementptr inbounds ptr, ptr %src, i64 %index
%wide.load = load <2 x ptr>, ptr %arrayidx, align 8
%cond = icmp eq <2 x ptr> %wide.load, splat(ptr zeroinitializer)
%index.next = add nuw i64 %index, 2
%or.reduc = tail call i1 @llvm.vector.reduce.or.v2i1(<2 x i1> %cond)
%iv.cmp = icmp eq i64 %index.next, 4
%exit.cond = or i1 %or.reduc, %iv.cmp
br i1 %exit.cond, label %middle.split, label %vector.body
middle.split:
br i1 %or.reduc, label %found, label %notfound
found:
store i64 56, ptr %p, align 8
ret i64 1
notfound:
ret i64 0
}
define i64 @select_or_reduce_nxv2i1(ptr nocapture noundef readonly %src) {
; CHECK-LABEL: select_or_reduce_nxv2i1:
; CHECK: // %bb.0: // %entry
; CHECK-NEXT: cntd x8
; CHECK-NEXT: ptrue p0.d
; CHECK-NEXT: mov x9, xzr
; CHECK-NEXT: neg x10, x8
; CHECK-NEXT: add x10, x10, #4
; CHECK-NEXT: .LBB2_1: // %vector.body
; CHECK-NEXT: // =>This Inner Loop Header: Depth=1
; CHECK-NEXT: ld1d { z0.d }, p0/z, [x0, x9, lsl #3]
; CHECK-NEXT: cmpeq p1.d, p0/z, z0.d, #0
; CHECK-NEXT: b.ne .LBB2_3
; CHECK-NEXT: // %bb.2: // %vector.body
; CHECK-NEXT: // in Loop: Header=BB2_1 Depth=1
; CHECK-NEXT: cmp x10, x9
; CHECK-NEXT: add x9, x9, x8
; CHECK-NEXT: b.ne .LBB2_1
; CHECK-NEXT: .LBB2_3: // %middle.split
; CHECK-NEXT: ptest p0, p1.b
; CHECK-NEXT: cset w0, ne
; CHECK-NEXT: ret
entry:
%vscale = tail call i64 @llvm.vscale.i64()
%vf = shl nuw nsw i64 %vscale, 1
br label %vector.body
vector.body:
%index = phi i64 [ 0, %entry ], [ %index.next, %vector.body ]
%arrayidx = getelementptr inbounds ptr, ptr %src, i64 %index
%wide.load = load <vscale x 2 x ptr>, ptr %arrayidx, align 8
%cond = icmp eq <vscale x 2 x ptr> %wide.load, splat(ptr zeroinitializer)
%index.next = add nuw i64 %index, %vf
%or.reduc = tail call i1 @llvm.vector.reduce.or.nxv2i1(<vscale x 2 x i1> %cond)
%iv.cmp = icmp eq i64 %index.next, 4
%exit.cond = or i1 %or.reduc, %iv.cmp
br i1 %exit.cond, label %middle.split, label %vector.body
middle.split:
%sel = select i1 %or.reduc, i64 1, i64 0
ret i64 %sel
}
define i64 @br_or_reduce_nxv2i1(ptr nocapture noundef readonly %src, ptr noundef readnone %p) {
; CHECK-LABEL: br_or_reduce_nxv2i1:
; CHECK: // %bb.0: // %entry
; CHECK-NEXT: cntd x8
; CHECK-NEXT: ptrue p0.d
; CHECK-NEXT: mov x9, xzr
; CHECK-NEXT: neg x10, x8
; CHECK-NEXT: add x10, x10, #4
; CHECK-NEXT: .LBB3_1: // %vector.body
; CHECK-NEXT: // =>This Inner Loop Header: Depth=1
; CHECK-NEXT: ld1d { z0.d }, p0/z, [x0, x9, lsl #3]
; CHECK-NEXT: cmpeq p1.d, p0/z, z0.d, #0
; CHECK-NEXT: b.ne .LBB3_3
; CHECK-NEXT: // %bb.2: // %vector.body
; CHECK-NEXT: // in Loop: Header=BB3_1 Depth=1
; CHECK-NEXT: cmp x10, x9
; CHECK-NEXT: add x9, x9, x8
; CHECK-NEXT: b.ne .LBB3_1
; CHECK-NEXT: .LBB3_3: // %middle.split
; CHECK-NEXT: ptest p0, p1.b
; CHECK-NEXT: b.eq .LBB3_5
; CHECK-NEXT: // %bb.4: // %found
; CHECK-NEXT: mov w8, #56 // =0x38
; CHECK-NEXT: mov w0, #1 // =0x1
; CHECK-NEXT: str x8, [x1]
; CHECK-NEXT: ret
; CHECK-NEXT: .LBB3_5:
; CHECK-NEXT: mov x0, xzr
; CHECK-NEXT: ret
entry:
%vscale = tail call i64 @llvm.vscale.i64()
%vf = shl nuw nsw i64 %vscale, 1
br label %vector.body
vector.body:
%index = phi i64 [ 0, %entry ], [ %index.next, %vector.body ]
%arrayidx = getelementptr inbounds ptr, ptr %src, i64 %index
%wide.load = load <vscale x 2 x ptr>, ptr %arrayidx, align 8
%cond = icmp eq <vscale x 2 x ptr> %wide.load, splat(ptr zeroinitializer)
%index.next = add nuw i64 %index, %vf
%or.reduc = tail call i1 @llvm.vector.reduce.or.nxv2i1(<vscale x 2 x i1> %cond)
%iv.cmp = icmp eq i64 %index.next, 4
%exit.cond = or i1 %or.reduc, %iv.cmp
br i1 %exit.cond, label %middle.split, label %vector.body
middle.split:
br i1 %or.reduc, label %found, label %notfound
found:
store i64 56, ptr %p, align 8
ret i64 1
notfound:
ret i64 0
}
declare i1 @llvm.vector.reduce.or.v2i1(<2 x i1>)
declare i1 @llvm.vector.reduce.or.nxv2i1(<vscale x 2 x i1>)
|
