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
136
137
|
; RUN: %if spirv-tools %{ llc -mtriple=spirv-unknown-vulkan-compute -O0 %s -o - | spirv-as --preserve-numeric-ids - -o - | spirv-val %}
;
; Source HLSL:
;
; int process() {
; int val=0, i=0, j=0, k=0;
;
; do {
; val = val + i;
;
; do {
; do {
; ++k;
; } while (k < 30);
;
; ++j;
; } while (j < 20);
;
; ++i;
;
; } while (i < 10);
;
; return val + i + j + k;
; }
;
; [numthreads(1, 1, 1)]
; void main() {
; process();
; }
target datalayout = "e-i64:64-v16:16-v24:32-v32:32-v48:64-v96:128-v192:256-v256:256-v512:512-v1024:1024-G1"
target triple = "spirv-unknown-vulkan1.3-compute"
; Function Attrs: convergent noinline nounwind optnone
define spir_func noundef i32 @_Z7processv() #0 {
entry:
%0 = call token @llvm.experimental.convergence.entry()
%val = alloca i32, align 4
%i = alloca i32, align 4
%j = alloca i32, align 4
%k = alloca i32, align 4
store i32 0, ptr %val, align 4
store i32 0, ptr %i, align 4
store i32 0, ptr %j, align 4
store i32 0, ptr %k, align 4
br label %do.body
do.body: ; preds = %do.cond8, %entry
%1 = call token @llvm.experimental.convergence.loop() [ "convergencectrl"(token %0) ]
%2 = load i32, ptr %val, align 4
%3 = load i32, ptr %i, align 4
%add = add nsw i32 %2, %3
store i32 %add, ptr %val, align 4
br label %do.body1
do.body1: ; preds = %do.cond4, %do.body
%4 = call token @llvm.experimental.convergence.loop() [ "convergencectrl"(token %1) ]
br label %do.body2
do.body2: ; preds = %do.cond, %do.body1
%5 = call token @llvm.experimental.convergence.loop() [ "convergencectrl"(token %4) ]
%6 = load i32, ptr %k, align 4
%inc = add nsw i32 %6, 1
store i32 %inc, ptr %k, align 4
br label %do.cond
do.cond: ; preds = %do.body2
%7 = load i32, ptr %k, align 4
%cmp = icmp slt i32 %7, 30
br i1 %cmp, label %do.body2, label %do.end
do.end: ; preds = %do.cond
%8 = load i32, ptr %j, align 4
%inc3 = add nsw i32 %8, 1
store i32 %inc3, ptr %j, align 4
br label %do.cond4
do.cond4: ; preds = %do.end
%9 = load i32, ptr %j, align 4
%cmp5 = icmp slt i32 %9, 20
br i1 %cmp5, label %do.body1, label %do.end6
do.end6: ; preds = %do.cond4
%10 = load i32, ptr %i, align 4
%inc7 = add nsw i32 %10, 1
store i32 %inc7, ptr %i, align 4
br label %do.cond8
do.cond8: ; preds = %do.end6
%11 = load i32, ptr %i, align 4
%cmp9 = icmp slt i32 %11, 10
br i1 %cmp9, label %do.body, label %do.end10
do.end10: ; preds = %do.cond8
%12 = load i32, ptr %val, align 4
%13 = load i32, ptr %i, align 4
%add11 = add nsw i32 %12, %13
%14 = load i32, ptr %j, align 4
%add12 = add nsw i32 %add11, %14
%15 = load i32, ptr %k, align 4
%add13 = add nsw i32 %add12, %15
ret i32 %add13
}
; Function Attrs: convergent nocallback nofree nosync nounwind willreturn memory(none)
declare token @llvm.experimental.convergence.entry() #1
; Function Attrs: convergent nocallback nofree nosync nounwind willreturn memory(none)
declare token @llvm.experimental.convergence.loop() #1
; Function Attrs: convergent noinline norecurse nounwind optnone
define internal spir_func void @main() #2 {
entry:
%0 = call token @llvm.experimental.convergence.entry()
%call1 = call spir_func noundef i32 @_Z7processv() #4 [ "convergencectrl"(token %0) ]
ret void
}
; Function Attrs: convergent norecurse
define void @main.1() #3 {
entry:
call void @main()
ret void
}
attributes #0 = { convergent noinline nounwind optnone "frame-pointer"="all" "no-trapping-math"="true" "stack-protector-buffer-size"="8" }
attributes #1 = { convergent nocallback nofree nosync nounwind willreturn memory(none) }
attributes #2 = { convergent noinline norecurse nounwind optnone "frame-pointer"="all" "no-trapping-math"="true" "stack-protector-buffer-size"="8" }
attributes #3 = { convergent norecurse "frame-pointer"="all" "hlsl.numthreads"="1,1,1" "hlsl.shader"="compute" "no-trapping-math"="true" "stack-protector-buffer-size"="8" }
attributes #4 = { convergent }
!llvm.module.flags = !{!0, !1, !2}
!0 = !{i32 1, !"wchar_size", i32 4}
!1 = !{i32 4, !"dx.disable_optimizations", i32 1}
!2 = !{i32 7, !"frame-pointer", i32 2}
|
