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
|
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 6
; RUN: llc --mtriple=loongarch32 --mattr=+32s,+lasx < %s | FileCheck %s
; RUN: llc --mtriple=loongarch64 --mattr=+lasx < %s | FileCheck %s
;; ceilf
define void @ceil_v8f32(ptr %res, ptr %a0) nounwind {
; CHECK-LABEL: ceil_v8f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: xvld $xr0, $a1, 0
; CHECK-NEXT: xvfrintrp.s $xr0, $xr0
; CHECK-NEXT: xvst $xr0, $a0, 0
; CHECK-NEXT: ret
entry:
%v0 = load <8 x float>, ptr %a0
%r = call <8 x float> @llvm.ceil.v8f32(<8 x float> %v0)
store <8 x float> %r, ptr %res
ret void
}
;; ceil
define void @ceil_v4f64(ptr %res, ptr %a0) nounwind {
; CHECK-LABEL: ceil_v4f64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: xvld $xr0, $a1, 0
; CHECK-NEXT: xvfrintrp.d $xr0, $xr0
; CHECK-NEXT: xvst $xr0, $a0, 0
; CHECK-NEXT: ret
entry:
%v0 = load <4 x double>, ptr %a0
%r = call <4 x double> @llvm.ceil.v4f64(<4 x double> %v0)
store <4 x double> %r, ptr %res
ret void
}
;; floorf
define void @floor_v8f32(ptr %res, ptr %a0) nounwind {
; CHECK-LABEL: floor_v8f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: xvld $xr0, $a1, 0
; CHECK-NEXT: xvfrintrm.s $xr0, $xr0
; CHECK-NEXT: xvst $xr0, $a0, 0
; CHECK-NEXT: ret
entry:
%v0 = load <8 x float>, ptr %a0
%r = call <8 x float> @llvm.floor.v8f32(<8 x float> %v0)
store <8 x float> %r, ptr %res
ret void
}
;; floor
define void @floor_v4f64(ptr %res, ptr %a0) nounwind {
; CHECK-LABEL: floor_v4f64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: xvld $xr0, $a1, 0
; CHECK-NEXT: xvfrintrm.d $xr0, $xr0
; CHECK-NEXT: xvst $xr0, $a0, 0
; CHECK-NEXT: ret
entry:
%v0 = load <4 x double>, ptr %a0
%r = call <4 x double> @llvm.floor.v4f64(<4 x double> %v0)
store <4 x double> %r, ptr %res
ret void
}
;; truncf
define void @trunc_v8f32(ptr %res, ptr %a0) nounwind {
; CHECK-LABEL: trunc_v8f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: xvld $xr0, $a1, 0
; CHECK-NEXT: xvfrintrz.s $xr0, $xr0
; CHECK-NEXT: xvst $xr0, $a0, 0
; CHECK-NEXT: ret
entry:
%v0 = load <8 x float>, ptr %a0
%r = call <8 x float> @llvm.trunc.v8f32(<8 x float> %v0)
store <8 x float> %r, ptr %res
ret void
}
;; trunc
define void @trunc_v4f64(ptr %res, ptr %a0) nounwind {
; CHECK-LABEL: trunc_v4f64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: xvld $xr0, $a1, 0
; CHECK-NEXT: xvfrintrz.d $xr0, $xr0
; CHECK-NEXT: xvst $xr0, $a0, 0
; CHECK-NEXT: ret
entry:
%v0 = load <4 x double>, ptr %a0
%r = call <4 x double> @llvm.trunc.v4f64(<4 x double> %v0)
store <4 x double> %r, ptr %res
ret void
}
;; roundevenf
define void @roundeven_v8f32(ptr %res, ptr %a0) nounwind {
; CHECK-LABEL: roundeven_v8f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: xvld $xr0, $a1, 0
; CHECK-NEXT: xvfrintrne.s $xr0, $xr0
; CHECK-NEXT: xvst $xr0, $a0, 0
; CHECK-NEXT: ret
entry:
%v0 = load <8 x float>, ptr %a0
%r = call <8 x float> @llvm.roundeven.v8f32(<8 x float> %v0)
store <8 x float> %r, ptr %res
ret void
}
;; roundeven
define void @roundeven_v4f64(ptr %res, ptr %a0) nounwind {
; CHECK-LABEL: roundeven_v4f64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: xvld $xr0, $a1, 0
; CHECK-NEXT: xvfrintrne.d $xr0, $xr0
; CHECK-NEXT: xvst $xr0, $a0, 0
; CHECK-NEXT: ret
entry:
%v0 = load <4 x double>, ptr %a0
%r = call <4 x double> @llvm.roundeven.v4f64(<4 x double> %v0)
store <4 x double> %r, ptr %res
ret void
}
declare <8 x float> @llvm.ceil.v8f32(<8 x float>)
declare <4 x double> @llvm.ceil.v4f64(<4 x double>)
declare <8 x float> @llvm.floor.v8f32(<8 x float>)
declare <4 x double> @llvm.floor.v4f64(<4 x double>)
declare <8 x float> @llvm.trunc.v8f32(<8 x float>)
declare <4 x double> @llvm.trunc.v4f64(<4 x double>)
declare <8 x float> @llvm.roundeven.v8f32(<8 x float>)
declare <4 x double> @llvm.roundeven.v4f64(<4 x double>)
|
