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
|
; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
; RUN: opt < %s -passes=slp-vectorizer -S -mtriple=aarch64-w32-windows-gnu | FileCheck %s
define i32 @foo() {
; CHECK-LABEL: @foo(
; CHECK-NEXT: entry:
; CHECK-NEXT: br label [[FOR_COND15_PREHEADER:%.*]]
; CHECK: for.cond15.preheader:
; CHECK-NEXT: br label [[IF_END:%.*]]
; CHECK: for.cond15:
; CHECK-NEXT: br label [[IF_END_1:%.*]]
; CHECK: if.end:
; CHECK-NEXT: br label [[FOR_COND15:%.*]]
; CHECK: for.end39:
; CHECK-NEXT: switch i32 undef, label [[DO_BODY:%.*]] [
; CHECK-NEXT: i32 0, label [[SW_BB:%.*]]
; CHECK-NEXT: i32 1, label [[SW_BB195:%.*]]
; CHECK-NEXT: ]
; CHECK: sw.bb:
; CHECK-NEXT: [[ARRAYIDX43:%.*]] = getelementptr inbounds [4 x [2 x double]], ptr undef, i32 0, i64 1, i64 0
; CHECK-NEXT: [[TMP1:%.*]] = load <4 x double>, ptr [[ARRAYIDX43]], align 8
; CHECK-NEXT: [[TMP2:%.*]] = fmul <4 x double> [[TMP1]], <double 0x7FF8000000000000, double 0x7FF8000000000000, double 0x7FF8000000000000, double 0x7FF8000000000000>
; CHECK-NEXT: [[TMP3:%.*]] = call <4 x double> @llvm.fmuladd.v4f64(<4 x double> undef, <4 x double> zeroinitializer, <4 x double> [[TMP2]])
; CHECK-NEXT: br label [[SW_EPILOG:%.*]]
; CHECK: sw.bb195:
; CHECK-NEXT: br label [[SW_EPILOG]]
; CHECK: do.body:
; CHECK-NEXT: unreachable
; CHECK: sw.epilog:
; CHECK-NEXT: [[TMP4:%.*]] = phi <4 x double> [ undef, [[SW_BB195]] ], [ [[TMP3]], [[SW_BB]] ]
; CHECK-NEXT: ret i32 undef
; CHECK: if.end.1:
; CHECK-NEXT: br label [[FOR_COND15_1:%.*]]
; CHECK: for.cond15.1:
; CHECK-NEXT: br i1 undef, label [[FOR_END39:%.*]], label [[FOR_COND15_PREHEADER]]
;
entry:
%conv = sitofp i32 undef to double
%conv2 = sitofp i32 undef to double
br label %for.cond15.preheader
for.cond15.preheader: ; preds = %for.cond15.1, %entry
br label %if.end
for.cond15: ; preds = %if.end
br label %if.end.1
if.end: ; preds = %for.cond15.preheader
br label %for.cond15
for.end39: ; preds = %for.cond15.1
switch i32 undef, label %do.body [
i32 0, label %sw.bb
i32 1, label %sw.bb195
]
sw.bb: ; preds = %for.end39
%arrayidx43 = getelementptr inbounds [4 x [2 x double]], ptr undef, i32 0, i64 1, i64 0
%0 = load double, ptr %arrayidx43, align 8
%arrayidx45 = getelementptr inbounds [4 x [2 x double]], ptr undef, i32 0, i64 2, i64 0
%1 = load double, ptr %arrayidx45, align 8
%arrayidx51 = getelementptr inbounds [4 x [2 x double]], ptr undef, i32 0, i64 2, i64 1
%2 = load double, ptr %arrayidx51, align 8
%arrayidx58 = getelementptr inbounds [4 x [2 x double]], ptr undef, i32 0, i64 1, i64 1
%3 = load double, ptr %arrayidx58, align 8
%mul = fmul double undef, %conv2
%mul109 = fmul double undef, %conv
%mul143 = fmul double %0, %mul
%4 = call double @llvm.fmuladd.f64(double undef, double %conv2, double %mul143)
%mul154 = fmul double %1, %mul109
%5 = call double @llvm.fmuladd.f64(double undef, double %conv, double %mul154)
%mul172 = fmul double %3, %mul
%6 = call double @llvm.fmuladd.f64(double undef, double %conv2, double %mul172)
%mul183 = fmul double %2, %mul109
%7 = call double @llvm.fmuladd.f64(double undef, double %conv, double %mul183)
br label %sw.epilog
sw.bb195: ; preds = %for.end39
br label %sw.epilog
do.body: ; preds = %for.end39
unreachable
sw.epilog: ; preds = %sw.bb195, %sw.bb
%x4.0 = phi double [ undef, %sw.bb195 ], [ %7, %sw.bb ]
%x3.0 = phi double [ undef, %sw.bb195 ], [ %6, %sw.bb ]
%x1.0 = phi double [ undef, %sw.bb195 ], [ %5, %sw.bb ]
%x0.0 = phi double [ undef, %sw.bb195 ], [ %4, %sw.bb ]
ret i32 undef
if.end.1: ; preds = %for.cond15
br label %for.cond15.1
for.cond15.1: ; preds = %if.end.1
br i1 undef, label %for.end39, label %for.cond15.preheader
}
declare double @llvm.fmuladd.f64(double, double, double)
|
