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
|
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc -mtriple=riscv32 -mattr=+v,+zvfh,+zvfbfmin -verify-machineinstrs < %s | FileCheck %s --check-prefixes=CHECK,NOZFMIN,ZVFH
; RUN: llc -mtriple=riscv64 -mattr=+v,+zvfh,+zvfbfmin -verify-machineinstrs < %s | FileCheck %s --check-prefixes=CHECK,NOZFMIN,ZVFH
; RUN: llc -mtriple=riscv32 -mattr=+v,+zvfhmin,+zvfbfmin -verify-machineinstrs < %s | FileCheck %s --check-prefixes=CHECK,NOZFMIN,ZVFHMIN
; RUN: llc -mtriple=riscv64 -mattr=+v,+zvfhmin,+zvfbfmin -verify-machineinstrs < %s | FileCheck %s --check-prefixes=CHECK,NOZFMIN,ZVFHMIN
; RUN: llc -mtriple=riscv32 -mattr=+v,+zfhmin,+zfbfmin,+zvfhmin,+zvfbfmin -verify-machineinstrs < %s | FileCheck %s --check-prefixes=CHECK,ZFMIN
; RUN: llc -mtriple=riscv64 -mattr=+v,+zfhmin,+zfbfmin,+zvfhmin,+zvfbfmin -verify-machineinstrs < %s | FileCheck %s --check-prefixes=CHECK,ZFMIN
define <vscale x 8 x bfloat> @vsplat_nxv8bf16(bfloat %f) {
; NOZFMIN-LABEL: vsplat_nxv8bf16:
; NOZFMIN: # %bb.0:
; NOZFMIN-NEXT: fmv.x.w a0, fa0
; NOZFMIN-NEXT: vsetvli a1, zero, e16, m2, ta, ma
; NOZFMIN-NEXT: vmv.v.x v8, a0
; NOZFMIN-NEXT: ret
;
; ZFMIN-LABEL: vsplat_nxv8bf16:
; ZFMIN: # %bb.0:
; ZFMIN-NEXT: fmv.x.h a0, fa0
; ZFMIN-NEXT: vsetvli a1, zero, e16, m2, ta, ma
; ZFMIN-NEXT: vmv.v.x v8, a0
; ZFMIN-NEXT: ret
%head = insertelement <vscale x 8 x bfloat> poison, bfloat %f, i32 0
%splat = shufflevector <vscale x 8 x bfloat> %head, <vscale x 8 x bfloat> poison, <vscale x 8 x i32> zeroinitializer
ret <vscale x 8 x bfloat> %splat
}
define <vscale x 8 x bfloat> @vsplat_zero_nxv8bf16() {
; CHECK-LABEL: vsplat_zero_nxv8bf16:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetvli a0, zero, e16, m2, ta, ma
; CHECK-NEXT: vmv.v.i v8, 0
; CHECK-NEXT: ret
ret <vscale x 8 x bfloat> splat (bfloat zeroinitializer)
}
define <vscale x 8 x half> @vsplat_nxv8f16(half %f) {
; ZVFH-LABEL: vsplat_nxv8f16:
; ZVFH: # %bb.0:
; ZVFH-NEXT: vsetvli a0, zero, e16, m2, ta, ma
; ZVFH-NEXT: vfmv.v.f v8, fa0
; ZVFH-NEXT: ret
;
; ZVFHMIN-LABEL: vsplat_nxv8f16:
; ZVFHMIN: # %bb.0:
; ZVFHMIN-NEXT: fmv.x.w a0, fa0
; ZVFHMIN-NEXT: vsetvli a1, zero, e16, m2, ta, ma
; ZVFHMIN-NEXT: vmv.v.x v8, a0
; ZVFHMIN-NEXT: ret
;
; ZFMIN-LABEL: vsplat_nxv8f16:
; ZFMIN: # %bb.0:
; ZFMIN-NEXT: fmv.x.h a0, fa0
; ZFMIN-NEXT: vsetvli a1, zero, e16, m2, ta, ma
; ZFMIN-NEXT: vmv.v.x v8, a0
; ZFMIN-NEXT: ret
%head = insertelement <vscale x 8 x half> poison, half %f, i32 0
%splat = shufflevector <vscale x 8 x half> %head, <vscale x 8 x half> poison, <vscale x 8 x i32> zeroinitializer
ret <vscale x 8 x half> %splat
}
define <vscale x 8 x half> @vsplat_zero_nxv8f16() {
; CHECK-LABEL: vsplat_zero_nxv8f16:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetvli a0, zero, e16, m2, ta, ma
; CHECK-NEXT: vmv.v.i v8, 0
; CHECK-NEXT: ret
ret <vscale x 8 x half> splat (half zeroinitializer)
}
define <vscale x 8 x float> @vsplat_nxv8f32(float %f) {
; CHECK-LABEL: vsplat_nxv8f32:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetvli a0, zero, e32, m4, ta, ma
; CHECK-NEXT: vfmv.v.f v8, fa0
; CHECK-NEXT: ret
%head = insertelement <vscale x 8 x float> poison, float %f, i32 0
%splat = shufflevector <vscale x 8 x float> %head, <vscale x 8 x float> poison, <vscale x 8 x i32> zeroinitializer
ret <vscale x 8 x float> %splat
}
define <vscale x 8 x float> @vsplat_zero_nxv8f32() {
; CHECK-LABEL: vsplat_zero_nxv8f32:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetvli a0, zero, e32, m4, ta, ma
; CHECK-NEXT: vmv.v.i v8, 0
; CHECK-NEXT: ret
ret <vscale x 8 x float> splat (float zeroinitializer)
}
define <vscale x 8 x double> @vsplat_nxv8f64(double %f) {
; CHECK-LABEL: vsplat_nxv8f64:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetvli a0, zero, e64, m8, ta, ma
; CHECK-NEXT: vfmv.v.f v8, fa0
; CHECK-NEXT: ret
%head = insertelement <vscale x 8 x double> poison, double %f, i32 0
%splat = shufflevector <vscale x 8 x double> %head, <vscale x 8 x double> poison, <vscale x 8 x i32> zeroinitializer
ret <vscale x 8 x double> %splat
}
define <vscale x 8 x double> @vsplat_zero_nxv8f64() {
; CHECK-LABEL: vsplat_zero_nxv8f64:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetvli a0, zero, e64, m8, ta, ma
; CHECK-NEXT: vmv.v.i v8, 0
; CHECK-NEXT: ret
ret <vscale x 8 x double> splat (double zeroinitializer)
}
define <vscale x 8 x float> @vsplat_load_nxv8f32(ptr %ptr) {
; CHECK-LABEL: vsplat_load_nxv8f32:
; CHECK: # %bb.0:
; CHECK-NEXT: flw fa5, 0(a0)
; CHECK-NEXT: vsetvli a0, zero, e32, m4, ta, ma
; CHECK-NEXT: vfmv.v.f v8, fa5
; CHECK-NEXT: ret
%f = load float, ptr %ptr
%head = insertelement <vscale x 8 x float> poison, float %f, i32 0
%splat = shufflevector <vscale x 8 x float> %head, <vscale x 8 x float> poison, <vscale x 8 x i32> zeroinitializer
ret <vscale x 8 x float> %splat
}
; Test that we fold this to a vlse with 0 stride.
define <vscale x 8 x float> @vsplat_load_nxv8f32_optimized(ptr %ptr) "target-features"="+optimized-zero-stride-load" {
; CHECK-LABEL: vsplat_load_nxv8f32_optimized:
; CHECK: # %bb.0:
; CHECK-NEXT: vsetvli a1, zero, e32, m4, ta, ma
; CHECK-NEXT: vlse32.v v8, (a0), zero
; CHECK-NEXT: ret
%f = load float, ptr %ptr
%head = insertelement <vscale x 8 x float> poison, float %f, i32 0
%splat = shufflevector <vscale x 8 x float> %head, <vscale x 8 x float> poison, <vscale x 8 x i32> zeroinitializer
ret <vscale x 8 x float> %splat
}
|
