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
|
;=========================== begin_copyright_notice ============================
;
; Copyright (C) 2021-2023 Intel Corporation
;
; SPDX-License-Identifier: MIT
;
;============================ end_copyright_notice =============================
; RUN: llc %s -march=genx64 -mcpu=Gen9 \
; RUN: -vc-skip-ocl-runtime-info \
; RUN: -vc-enable-dbginfo-dumps \
; RUN: -vc-dump-module-to-visa-transform-info-path=%basename_t.structure \
; RUN: -finalizer-opts='-generateDebugInfo' -o /dev/null
; RUN: FileCheck %s --input-file=%basename_t.structure
; CHECK: [0] result (K)
; CHECK-NEXT: [1] foo___vyfvyf (I)
; CHECK-NEXT: l.0 _Z24__cm_intrinsic_impl_test.foo___vyfvyf
; CHECK-NEXT: [2] f_f (K)
; CHECK-NEXT: l.0 _Z24__cm_intrinsic_impl_test
; CHECK-NEXT: [3] bar___vyfvyf (I)
; CHECK-NEXT: l.0 _Z24__cm_intrinsic_impl_test.bar___vyfvyf
source_filename = "indirect_calls_from_2_kernels.ll"
target datalayout = "e-p:64:64-i64:64-n8:16:32:64"
target triple = "spir64-unknown-unknown"
; Function Attrs: nounwind readnone speculatable willreturn
declare void @llvm.dbg.value(metadata, metadata, metadata) #0
; Function Attrs: nounwind readonly
declare <16 x float> @llvm.genx.svm.block.ld.v16f32.i64(i64) #1
; Function Attrs: nounwind
declare void @llvm.genx.svm.block.st.i64.v16f32(i64, <16 x float>) #2
define spir_func <16 x float> @_Z24__cm_intrinsic_impl_test(<16 x float> %0, <16 x float> %1) local_unnamed_addr #3 {
%3 = fadd <16 x float> %0, %1
ret <16 x float> %3
}
; Function Attrs: noinline nounwind readnone
define internal spir_func <16 x float> @foo___vyfvyf(<16 x float> %a, <16 x float> %b, <16 x i1> %__mask) #4 !FuncArgSize !18 !FuncRetSize !19 {
allocas:
%l = fadd <16 x float> %a, %b
%k = tail call spir_func <16 x float> @_Z24__cm_intrinsic_impl_test(<16 x float> %l, <16 x float> %l)
ret <16 x float> %k
}
; Function Attrs: noinline nounwind
define internal spir_func <16 x float> @bar___vyfvyf(<16 x float> %a, <16 x float> %b, <16 x i1> %__mask) #5 !FuncArgSize !18 !FuncRetSize !19 {
allocas:
%l = fsub <16 x float> %a, %b
%k = tail call spir_func <16 x float> @_Z24__cm_intrinsic_impl_test(<16 x float> %l, <16 x float> %l)
ret <16 x float> %k
}
; Function Attrs: nounwind
define dllexport void @f_f(float* noalias %RET, float* noalias %aFOO, i64 %privBase) #6 {
%svm_ld_ptrtoint = ptrtoint float* %aFOO to i64
%src1 = call <16 x float> @llvm.genx.svm.block.ld.v16f32.i64(i64 %svm_ld_ptrtoint)
%foo_raw = ptrtoint <16 x float> (<16 x float>, <16 x float>, <16 x i1>)* @bar___vyfvyf to i64
%bar_raw = ptrtoint <16 x float> (<16 x float>, <16 x float>, <16 x i1>)* @foo___vyfvyf to i64
%foo_raw_vect = insertelement <1 x i64> undef, i64 %foo_raw, i32 0
%bar_raw_vect = insertelement <1 x i64> undef, i64 %bar_raw, i32 0
%cmp = icmp eq i64 %privBase, 0
%rawaddr_v = select i1 %cmp, <1 x i64> %foo_raw_vect, <1 x i64> %bar_raw_vect
%rawaddr = extractelement <1 x i64> %rawaddr_v, i32 0
%fptr = inttoptr i64 %rawaddr to <16 x float> (<16 x float>, <16 x float>, <16 x i1>)*
%calltmp = call spir_func <16 x float> %fptr(<16 x float> %src1, <16 x float> %src1, <16 x i1> zeroinitializer) #7, !FuncArgSize !18, !FuncRetSize !19
%k = tail call spir_func <16 x float> @_Z24__cm_intrinsic_impl_test(<16 x float> %calltmp, <16 x float> %calltmp)
%svm_st_ptrtoint = ptrtoint float* %RET to i64
call void @llvm.genx.svm.block.st.i64.v16f32(i64 %svm_st_ptrtoint, <16 x float> %k)
ret void
}
; Function Attrs: nounwind
define dllexport void @result(float* noalias %RET, i64 %privBase) #6 {
%foo_raw = ptrtoint <16 x float> (<16 x float>, <16 x float>, <16 x i1>)* @bar___vyfvyf to i64
%bar_raw = ptrtoint <16 x float> (<16 x float>, <16 x float>, <16 x i1>)* @foo___vyfvyf to i64
%foo_raw_vect = insertelement <1 x i64> undef, i64 %foo_raw, i32 0
%bar_raw_vect = insertelement <1 x i64> undef, i64 %bar_raw, i32 0
%cmp = icmp eq i64 %privBase, 0
%rawaddr_v = select i1 %cmp, <1 x i64> %foo_raw_vect, <1 x i64> %bar_raw_vect
%rawaddr = extractelement <1 x i64> %rawaddr_v, i32 0
%fptr = inttoptr i64 %rawaddr to <16 x float> (<16 x float>, <16 x float>, <16 x i1>)*
%calltmp = call spir_func <16 x float> %fptr(<16 x float> zeroinitializer, <16 x float> zeroinitializer, <16 x i1> zeroinitializer) #7, !FuncArgSize !18, !FuncRetSize !19
%svm_st_ptrtoint = ptrtoint float* %RET to i64
call void @llvm.genx.svm.block.st.i64.v16f32(i64 %svm_st_ptrtoint, <16 x float> %calltmp)
ret void
}
attributes #0 = { nounwind readnone speculatable willreturn "target-cpu"="Gen9" }
attributes #1 = { nounwind readonly "target-cpu"="Gen9" }
attributes #2 = { nounwind "target-cpu"="Gen9" }
attributes #3 = { "target-cpu"="Gen9" "VC.Builtin"}
attributes #4 = { noinline nounwind readnone "CMStackCall" "target-cpu"="Gen9" }
attributes #5 = { noinline nounwind "CMStackCall" "target-cpu"="Gen9" }
attributes #6 = { nounwind "CMGenxMain" "oclrt"="1" "target-cpu"="Gen9" }
attributes #7 = { nounwind }
!spirv.Source = !{!0}
!opencl.spir.version = !{!1}
!opencl.ocl.version = !{!0}
!opencl.used.extensions = !{!2}
!opencl.used.optional.core.features = !{!2}
!spirv.Generator = !{!3}
!genx.kernels = !{!4, !8}
!VC.Debug.Enable = !{}
!genx.kernel.internal = !{!13, !16}
!0 = !{i32 0, i32 0}
!1 = !{i32 1, i32 2}
!2 = !{}
!3 = !{i16 6, i16 14}
!4 = !{void (float*, float*, i64)* @f_f, !"f_f", !5, i32 0, !6, !0, !7, i32 0}
!5 = !{i32 0, i32 0, i32 96}
!6 = !{i32 72, i32 80, i32 64}
!7 = !{!"", !""}
!8 = !{void (float*, i64)* @result, !"result", !9, i32 0, !10, !11, !12, i32 0}
!9 = !{i32 0, i32 96}
!10 = !{i32 72, i32 64}
!11 = !{i32 0}
!12 = !{!""}
!13 = !{void (float*, float*, i64)* @f_f, !14, !15, !2, null}
!14 = !{i32 0, i32 0, i32 0}
!15 = !{i32 0, i32 1, i32 2}
!16 = !{void (float*, i64)* @result, !0, !17, !2, null}
!17 = !{i32 0, i32 1}
!18 = !{i32 5}
!19 = !{i32 2}
|
