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
|
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc -mtriple=armv8.6a-arm-none-eabi -mattr=+bf16,+neon,+fullfp16 < %s | FileCheck %s
target datalayout = "e-m:e-p:32:32-Fi8-i64:64-v128:64:128-a:0:32-n32-S64"
target triple = "armv8.6a-arm-none-eabi"
define arm_aapcs_vfpcc bfloat @test_vgetq_lane_bf16_even(<8 x bfloat> %v) {
; CHECK-LABEL: test_vgetq_lane_bf16_even:
; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: vmov.f32 s0, s3
; CHECK-NEXT: bx lr
entry:
%0 = extractelement <8 x bfloat> %v, i32 6
ret bfloat %0
}
define arm_aapcs_vfpcc bfloat @test_vgetq_lane_bf16_odd(<8 x bfloat> %v) {
; CHECK-LABEL: test_vgetq_lane_bf16_odd:
; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: vmovx.f16 s0, s3
; CHECK-NEXT: bx lr
entry:
%0 = extractelement <8 x bfloat> %v, i32 7
ret bfloat %0
}
define arm_aapcs_vfpcc bfloat @test_vget_lane_bf16_even(<4 x bfloat> %v) {
; CHECK-LABEL: test_vget_lane_bf16_even:
; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: vmov.f32 s0, s1
; CHECK-NEXT: bx lr
entry:
%0 = extractelement <4 x bfloat> %v, i32 2
ret bfloat %0
}
define arm_aapcs_vfpcc bfloat @test_vget_lane_bf16_odd(<4 x bfloat> %v) {
; CHECK-LABEL: test_vget_lane_bf16_odd:
; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: vmovx.f16 s0, s0
; CHECK-NEXT: bx lr
entry:
%0 = extractelement <4 x bfloat> %v, i32 1
ret bfloat %0
}
define i16 @bextract_v4i16(<4 x bfloat> %a) {
; CHECK-LABEL: bextract_v4i16:
; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: vmov d16, r0, r1
; CHECK-NEXT: vmov.u16 r0, d16[0]
; CHECK-NEXT: bx lr
entry:
%elt = extractelement <4 x bfloat> %a, i32 0
%t = bitcast bfloat %elt to i16
ret i16 %t
}
define i16 @bextract_v8i16(<8 x bfloat> %a) {
; CHECK-LABEL: bextract_v8i16:
; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: vmov d16, r0, r1
; CHECK-NEXT: vmov.u16 r0, d16[0]
; CHECK-NEXT: bx lr
entry:
%elt = extractelement <8 x bfloat> %a, i32 0
%t = bitcast bfloat %elt to i16
ret i16 %t
}
define i32 @bextract_v4s32(<4 x bfloat> %a) {
; CHECK-LABEL: bextract_v4s32:
; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: vmov d16, r0, r1
; CHECK-NEXT: vmov.u16 r0, d16[0]
; CHECK-NEXT: sxth r0, r0
; CHECK-NEXT: bx lr
entry:
%elt = extractelement <4 x bfloat> %a, i32 0
%t = bitcast bfloat %elt to i16
%s = sext i16 %t to i32
ret i32 %s
}
define i32 @bextract_v8s32(<8 x bfloat> %a) {
; CHECK-LABEL: bextract_v8s32:
; CHECK: @ %bb.0: @ %entry
; CHECK-NEXT: vmov d16, r0, r1
; CHECK-NEXT: vmov.u16 r0, d16[0]
; CHECK-NEXT: sxth r0, r0
; CHECK-NEXT: bx lr
entry:
%elt = extractelement <8 x bfloat> %a, i32 0
%t = bitcast bfloat %elt to i16
%s = sext i16 %t to i32
ret i32 %s
}
|
