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
|
;=========================== begin_copyright_notice ============================
;
; Copyright (C) 2022 Intel Corporation
;
; SPDX-License-Identifier: MIT
;
;============================ end_copyright_notice =============================
;
; RUN: igc_opt -enable-debugify -igc-lower-implicit-arg-intrinsic -S < %s 2>&1 | FileCheck %s
; ------------------------------------------------
; LowerImplicitArgIntrinsics
; ------------------------------------------------
; Debug-info related check
; CHECK-NOT: WARNING
; CHECK: CheckModuleDebugify: PASS
define spir_func void @test(<3 x i32> addrspace(1)* %dst, <8 x i32> %r0, <8 x i32> %payloadHeader, i16 %localIdX, i16 %localIdY, i16 %localIdZ, i8* %privateBase, i32 %bufferOffset) #0 {
; CHECK-LABEL: @test(
; CHECK: entry:
; CHECK: [[TMP0:%.*]] = call i32 addrspace(1)* @llvm.genx.GenISA.GetImplicitBufferPtr.p1i32()
; CHECK: [[TMP1:%.*]] = ptrtoint i32 addrspace(1)* [[TMP0]] to i64
; CHECK: [[TMP2:%.*]] = add i64 [[TMP1]], 4
; CHECK: [[TMP3:%.*]] = inttoptr i64 [[TMP2]] to i32 addrspace(1)*
; CHECK: [[TMP4:%.*]] = bitcast i32 addrspace(1)* [[TMP3]] to <3 x i32> addrspace(1)*
; CHECK: [[TMP5:%.*]] = load <3 x i32>, <3 x i32> addrspace(1)* [[TMP4]], align 4
; CHECK: store <3 x i32> [[TMP5]], <3 x i32> addrspace(1)* [[DST:%.*]], align 4
; CHECK: [[TMP6:%.*]] = call i32 addrspace(1)* @llvm.genx.GenISA.GetImplicitBufferPtr.p1i32()
; CHECK: [[TMP7:%.*]] = ptrtoint i32 addrspace(1)* [[TMP6]] to i64
; CHECK: [[TMP8:%.*]] = add i64 [[TMP7]], 48
; CHECK: [[TMP9:%.*]] = inttoptr i64 [[TMP8]] to i32 addrspace(1)*
; CHECK: [[TMP10:%.*]] = bitcast i32 addrspace(1)* [[TMP9]] to <3 x i64> addrspace(1)*
; CHECK: [[TMP11:%.*]] = load <3 x i64>, <3 x i64> addrspace(1)* [[TMP10]], align 8
; CHECK: [[TMP12:%.*]] = extractelement <3 x i64> [[TMP11]], i64 0
; CHECK: [[TMP13:%.*]] = trunc i64 [[TMP12]] to i32
; CHECK: [[TMP14:%.*]] = insertelement <3 x i32> undef, i32 [[TMP13]], i64 0
; CHECK: [[TMP15:%.*]] = extractelement <3 x i64> [[TMP11]], i64 1
; CHECK: [[TMP16:%.*]] = trunc i64 [[TMP15]] to i32
; CHECK: [[TMP17:%.*]] = insertelement <3 x i32> [[TMP14]], i32 [[TMP16]], i64 1
; CHECK: [[TMP18:%.*]] = extractelement <3 x i64> [[TMP11]], i64 2
; CHECK: [[TMP19:%.*]] = trunc i64 [[TMP18]] to i32
; CHECK: [[TMP20:%.*]] = insertelement <3 x i32> [[TMP17]], i32 [[TMP19]], i64 2
; CHECK: store <3 x i32> [[TMP20]], <3 x i32> addrspace(1)* [[DST]], align 4
; CHECK: [[TMP21:%.*]] = call i32 addrspace(1)* @llvm.genx.GenISA.GetImplicitBufferPtr.p1i32()
; CHECK: [[TMP22:%.*]] = ptrtoint i32 addrspace(1)* [[TMP21]] to i64
; CHECK: [[TMP23:%.*]] = add i64 [[TMP22]], 16
; CHECK: [[TMP24:%.*]] = inttoptr i64 [[TMP23]] to i32 addrspace(1)*
; CHECK: [[TMP25:%.*]] = bitcast i32 addrspace(1)* [[TMP24]] to <3 x i64> addrspace(1)*
; CHECK: [[TMP26:%.*]] = load <3 x i64>, <3 x i64> addrspace(1)* [[TMP25]], align 8
; CHECK: [[TMP27:%.*]] = extractelement <3 x i64> [[TMP26]], i64 0
; CHECK: [[TMP28:%.*]] = trunc i64 [[TMP27]] to i32
; CHECK: [[TMP29:%.*]] = insertelement <3 x i32> undef, i32 [[TMP28]], i64 0
; CHECK: [[TMP30:%.*]] = extractelement <3 x i64> [[TMP26]], i64 1
; CHECK: [[TMP31:%.*]] = trunc i64 [[TMP30]] to i32
; CHECK: [[TMP32:%.*]] = insertelement <3 x i32> [[TMP29]], i32 [[TMP31]], i64 1
; CHECK: [[TMP33:%.*]] = extractelement <3 x i64> [[TMP26]], i64 2
; CHECK: [[TMP34:%.*]] = trunc i64 [[TMP33]] to i32
; CHECK: [[TMP35:%.*]] = insertelement <3 x i32> [[TMP32]], i32 [[TMP34]], i64 2
; CHECK: store <3 x i32> [[TMP35]], <3 x i32> addrspace(1)* [[DST]], align 4
; CHECK: [[TMP36:%.*]] = call i32 addrspace(1)* @llvm.genx.GenISA.GetImplicitBufferPtr.p1i32()
; CHECK: [[TMP37:%.*]] = ptrtoint i32 addrspace(1)* [[TMP36]] to i64
; CHECK: [[TMP38:%.*]] = add i64 [[TMP37]], 80
; CHECK: [[TMP39:%.*]] = inttoptr i64 [[TMP38]] to i32 addrspace(1)*
; CHECK: [[TMP40:%.*]] = bitcast i32 addrspace(1)* [[TMP39]] to <3 x i32> addrspace(1)*
; CHECK: [[TMP41:%.*]] = load <3 x i32>, <3 x i32> addrspace(1)* [[TMP40]], align 4
; CHECK: store <3 x i32> [[TMP41]], <3 x i32> addrspace(1)* [[DST]], align 4
; CHECK: [[TMP42:%.*]] = call i32 addrspace(1)* @llvm.genx.GenISA.GetImplicitBufferPtr.p1i32()
; CHECK: [[TMP43:%.*]] = ptrtoint i32 addrspace(1)* [[TMP42]] to i64
; CHECK: [[TMP44:%.*]] = add i64 [[TMP43]], 40
; CHECK: [[TMP45:%.*]] = inttoptr i64 [[TMP44]] to i32 addrspace(1)*
; CHECK: [[TMP46:%.*]] = bitcast i32 addrspace(1)* [[TMP45]] to <1 x i64> addrspace(1)*
; CHECK: [[TMP47:%.*]] = load <1 x i64>, <1 x i64> addrspace(1)* [[TMP46]], align 8
; CHECK: [[TMP48:%.*]] = bitcast <1 x i64> [[TMP47]] to i64
; CHECK: [[TMP49:%.*]] = inttoptr i64 [[TMP48]] to i32 addrspace(1)*
; CHECK: [[TMP50:%.*]] = call i32 addrspace(1)* @llvm.genx.GenISA.GetImplicitBufferPtr.p1i32()
; CHECK: [[TMP51:%.*]] = ptrtoint i32 addrspace(1)* [[TMP50]] to i64
; CHECK: [[TMP52:%.*]] = add i64 [[TMP51]], 0
; CHECK: [[TMP53:%.*]] = inttoptr i64 [[TMP52]] to i32 addrspace(1)*
; CHECK: [[TMP54:%.*]] = bitcast i32 addrspace(1)* [[TMP53]] to <1 x i32> addrspace(1)*
; CHECK: [[TMP55:%.*]] = load <1 x i32>, <1 x i32> addrspace(1)* [[TMP54]], align 4
; CHECK: [[TMP56:%.*]] = bitcast <1 x i32> [[TMP55]] to i32
; CHECK: [[TMP57:%.*]] = lshr i32 [[TMP56]], 16
; CHECK: [[TMP58:%.*]] = and i32 [[TMP57]], 255
; CHECK: store i32 [[TMP58]], i32 addrspace(1)* [[TMP49]], align 4
entry:
%0 = call <3 x i32> @llvm.genx.GenISA.getLocalSize.v4i32()
store <3 x i32> %0, <3 x i32> addrspace(1)* %dst, align 4
%1 = call <3 x i32> @llvm.genx.GenISA.getPayloadHeader.v4i32()
store <3 x i32> %1, <3 x i32> addrspace(1)* %dst, align 4
%2 = call <3 x i32> @llvm.genx.GenISA.getGlobalSize.v4i32()
store <3 x i32> %2, <3 x i32> addrspace(1)* %dst, align 4
%3 = call <3 x i32> @llvm.genx.GenISA.getNumWorkGroups.v4i32()
store <3 x i32> %3, <3 x i32> addrspace(1)* %dst, align 4
%4 = call i32 addrspace(1)* @llvm.genx.GenISA.getPrintfBuffer.p1i32()
%5 = call i32 @llvm.genx.GenISA.getWorkDim.i32()
store i32 %5, i32 addrspace(1)* %4, align 4
ret void
}
declare <3 x i32> @llvm.genx.GenISA.getLocalSize.v4i32()
declare <3 x i32> @llvm.genx.GenISA.getPayloadHeader.v4i32()
declare <3 x i32> @llvm.genx.GenISA.getGlobalSize.v4i32()
declare <3 x i32> @llvm.genx.GenISA.getNumWorkGroups.v4i32()
declare i32 @llvm.genx.GenISA.getWorkDim.i32()
declare i32 addrspace(1)* @llvm.genx.GenISA.getPrintfBuffer.p1i32()
attributes #0 = { convergent noinline nounwind optnone "visaStackCall" }
attributes #1 = { nofree nosync nounwind readnone speculatable willreturn }
!igc.functions = !{!3}
!3 = !{void (<3 x i32> addrspace(1)*, <8 x i32>, <8 x i32>, i16, i16, i16, i8*, i32)* @test, !4}
!4 = !{!5, !6}
!5 = !{!"function_type", i32 2}
!6 = !{!"implicit_arg_desc", !7, !8, !9, !10, !11, !12, !13}
!7 = !{i32 0}
!8 = !{i32 1}
!9 = !{i32 7}
!10 = !{i32 8}
!11 = !{i32 9}
!12 = !{i32 12}
!13 = !{i32 14, !14}
!14 = !{!"explicit_arg_num", i32 0}
|
