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
|
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 2
; RUN: llc -mtriple=aarch64--linux-gnu -mattr=+sve2p1 < %s | FileCheck %s
; ST1W
define void @test_svst1wq_i32_ss(<vscale x 4 x i32> %zt, <vscale x 1 x i1> %pred, ptr %base, i64 %offset) {
; CHECK-LABEL: test_svst1wq_i32_ss:
; CHECK: // %bb.0:
; CHECK-NEXT: st1w { z0.q }, p0, [x0, x1, lsl #2]
; CHECK-NEXT: ret
%gep = getelementptr i32, ptr %base, i64 %offset
call void @llvm.aarch64.sve.st1wq.nxv4i32(<vscale x 4 x i32> %zt, <vscale x 1 x i1> %pred, ptr %gep)
ret void
}
define void @test_svst1wq_i32_si(<vscale x 4 x i32> %zt, <vscale x 1 x i1> %pred, ptr %base) {
; CHECK-LABEL: test_svst1wq_i32_si:
; CHECK: // %bb.0:
; CHECK-NEXT: st1w { z0.q }, p0, [x0, #-8, mul vl]
; CHECK-NEXT: st1w { z0.q }, p0, [x0, #7, mul vl]
; CHECK-NEXT: ret
%gep1 = getelementptr inbounds <vscale x 1 x i32>, ptr %base, i64 -8
call void @llvm.aarch64.sve.st1wq.nxv4i32(<vscale x 4 x i32> %zt, <vscale x 1 x i1> %pred, ptr %gep1)
%gep2 = getelementptr inbounds <vscale x 1 x i32>, ptr %base, i64 7
call void @llvm.aarch64.sve.st1wq.nxv4i32(<vscale x 4 x i32> %zt, <vscale x 1 x i1> %pred, ptr %gep2)
ret void
}
define void @test_svst1wq_i32_out_of_bound(<vscale x 4 x i32> %zt, <vscale x 1 x i1> %pred, ptr %base) {
; CHECK-LABEL: test_svst1wq_i32_out_of_bound:
; CHECK: // %bb.0:
; CHECK-NEXT: incb x0, all, mul #2
; CHECK-NEXT: st1w { z0.q }, p0, [x0]
; CHECK-NEXT: ret
%gep = getelementptr inbounds <vscale x 1 x i32>, ptr %base, i64 8
call void @llvm.aarch64.sve.st1wq.nxv4i32(<vscale x 4 x i32> %zt, <vscale x 1 x i1> %pred, ptr %gep)
ret void
}
define void @test_svst1wq_f32_ss(<vscale x 4 x float> %zt, <vscale x 1 x i1> %pred, ptr %base, i64 %offset) {
; CHECK-LABEL: test_svst1wq_f32_ss:
; CHECK: // %bb.0:
; CHECK-NEXT: st1w { z0.q }, p0, [x0, x1, lsl #2]
; CHECK-NEXT: ret
%gep = getelementptr float, ptr %base, i64 %offset
call void @llvm.aarch64.sve.st1wq.nxv4f32(<vscale x 4 x float> %zt, <vscale x 1 x i1> %pred, ptr %gep)
ret void
}
define void @test_svst1wq_f32_si(<vscale x 4 x float> %zt, <vscale x 1 x i1> %pred, ptr %base) {
; CHECK-LABEL: test_svst1wq_f32_si:
; CHECK: // %bb.0:
; CHECK-NEXT: st1w { z0.q }, p0, [x0, #-8, mul vl]
; CHECK-NEXT: st1w { z0.q }, p0, [x0, #7, mul vl]
; CHECK-NEXT: ret
%gep1 = getelementptr inbounds <vscale x 1 x float>, ptr %base, i64 -8
call void @llvm.aarch64.sve.st1wq.nxv4f32(<vscale x 4 x float> %zt, <vscale x 1 x i1> %pred, ptr %gep1)
%gep2 = getelementptr inbounds <vscale x 1 x float>, ptr %base, i64 7
call void @llvm.aarch64.sve.st1wq.nxv4f32(<vscale x 4 x float> %zt, <vscale x 1 x i1> %pred, ptr %gep2)
ret void
}
; ST1D
define void @test_svst1dq_i64_ss(<vscale x 2 x i64> %zt, <vscale x 1 x i1> %pred, ptr %base, i64 %offset) {
; CHECK-LABEL: test_svst1dq_i64_ss:
; CHECK: // %bb.0:
; CHECK-NEXT: st1d { z0.q }, p0, [x0, x1, lsl #3]
; CHECK-NEXT: ret
%gep = getelementptr i64, ptr %base, i64 %offset
call void @llvm.aarch64.sve.st1dq.nxv2i64(<vscale x 2 x i64> %zt, <vscale x 1 x i1> %pred, ptr %gep)
ret void
}
define void @test_svst1dq_i64_si(<vscale x 2 x i64> %zt, <vscale x 1 x i1> %pred, ptr %base) {
; CHECK-LABEL: test_svst1dq_i64_si:
; CHECK: // %bb.0:
; CHECK-NEXT: st1d { z0.q }, p0, [x0, #-8, mul vl]
; CHECK-NEXT: st1d { z0.q }, p0, [x0, #7, mul vl]
; CHECK-NEXT: ret
%gep1 = getelementptr inbounds <vscale x 1 x i64>, ptr %base, i64 -8
call void @llvm.aarch64.sve.st1dq.nxv2i64(<vscale x 2 x i64> %zt, <vscale x 1 x i1> %pred, ptr %gep1)
%gep2 = getelementptr inbounds <vscale x 1 x i64>, ptr %base, i64 7
call void @llvm.aarch64.sve.st1dq.nxv2i64(<vscale x 2 x i64> %zt, <vscale x 1 x i1> %pred, ptr %gep2)
ret void
}
define void @test_svst1dq_i64_out_of_bound(<vscale x 2 x i64> %zt, <vscale x 1 x i1> %pred, ptr %base) {
; CHECK-LABEL: test_svst1dq_i64_out_of_bound:
; CHECK: // %bb.0:
; CHECK-NEXT: addvl x8, x0, #-5
; CHECK-NEXT: st1d { z0.q }, p0, [x8]
; CHECK-NEXT: ret
%gep = getelementptr inbounds <vscale x 1 x i64>, ptr %base, i64 -10
call void @llvm.aarch64.sve.st1dq.nxv2i64(<vscale x 2 x i64> %zt, <vscale x 1 x i1> %pred, ptr %gep)
ret void
}
define void @test_svst1dq_f64_ss(<vscale x 2 x double> %zt, <vscale x 1 x i1> %pred, ptr %base, i64 %offset) {
; CHECK-LABEL: test_svst1dq_f64_ss:
; CHECK: // %bb.0:
; CHECK-NEXT: st1d { z0.q }, p0, [x0, x1, lsl #3]
; CHECK-NEXT: ret
%gep = getelementptr double, ptr %base, i64 %offset
call void @llvm.aarch64.sve.st1dq.nxv2f64(<vscale x 2 x double> %zt, <vscale x 1 x i1> %pred, ptr %gep)
ret void
}
define void @test_svst1dq_f64_si(<vscale x 2 x double> %zt, <vscale x 1 x i1> %pred, ptr %base) {
; CHECK-LABEL: test_svst1dq_f64_si:
; CHECK: // %bb.0:
; CHECK-NEXT: st1d { z0.q }, p0, [x0, #-8, mul vl]
; CHECK-NEXT: st1d { z0.q }, p0, [x0, #7, mul vl]
; CHECK-NEXT: ret
%gep1 = getelementptr inbounds <vscale x 1 x double>, ptr %base, i64 -8
call void @llvm.aarch64.sve.st1dq.nxv2f64(<vscale x 2 x double> %zt, <vscale x 1 x i1> %pred, ptr %gep1)
%gep2 = getelementptr inbounds <vscale x 1 x double>, ptr %base, i64 7
call void @llvm.aarch64.sve.st1dq.nxv2f64(<vscale x 2 x double> %zt, <vscale x 1 x i1> %pred, ptr %gep2)
ret void
}
declare void @llvm.aarch64.sve.st1wq.nxv4i32(<vscale x 4 x i32>, <vscale x 1 x i1>, ptr)
declare void @llvm.aarch64.sve.st1wq.nxv4f32(<vscale x 4 x float>, <vscale x 1 x i1>, ptr)
declare void @llvm.aarch64.sve.st1dq.nxv2i64(<vscale x 2 x i64>, <vscale x 1 x i1>, ptr)
declare void @llvm.aarch64.sve.st1dq.nxv2f64(<vscale x 2 x double>, <vscale x 1 x i1>, ptr)
|
