1; RUN: opt -amdgpu-lower-enqueued-block -S < %s | FileCheck %s
2
3; CHECK: @__test_block_invoke_kernel_runtime_handle = external addrspace(1) externally_initialized constant i8 addrspace(1)*
4; CHECK: @__test_block_invoke_2_kernel_runtime_handle = external addrspace(1) externally_initialized constant i8 addrspace(1)*
5
6target datalayout = "e-p:32:32-p1:64:64-p2:64:64-p3:32:32-p4:64:64-p5:32:32-i64:64-v16:16-v24:32-v32:32-v48:64-v96:128-v192:256-v256:256-v512:512-v1024:1024-v2048:2048-n32:64"
7target triple = "amdgcn-amdhsa-amd-opencl"
8
9%struct.ndrange_t = type { i32 }
10%opencl.queue_t = type opaque
11
12define amdgpu_kernel void @test(i8 addrspace(1)* %a, i8 %b, i64 addrspace(1)* %c, i64 %d) local_unnamed_addr
13  !kernel_arg_addr_space !3 !kernel_arg_access_qual !4 !kernel_arg_type !5 !kernel_arg_base_type !5 !kernel_arg_type_qual !6 {
14entry:
15  %block = alloca <{ i32, i32, i8 addrspace(4)*, i8 addrspace(1)*, i8 }>, align 8
16  %tmp = alloca %struct.ndrange_t, align 4
17  %block2 = alloca <{ i32, i32, i8 addrspace(4)*, i8 addrspace(1)*, i64 addrspace(1)*, i64, i8 }>, align 8
18  %tmp3 = alloca %struct.ndrange_t, align 4
19  %block.size = getelementptr inbounds <{ i32, i32, i8 addrspace(4)*, i8 addrspace(1)*, i8 }>, <{ i32, i32, i8 addrspace(4)*, i8 addrspace(1)*, i8 }>* %block, i32 0, i32 0
20  store i32 25, i32* %block.size, align 8
21  %block.align = getelementptr inbounds <{ i32, i32, i8 addrspace(4)*, i8 addrspace(1)*, i8 }>, <{ i32, i32, i8 addrspace(4)*, i8 addrspace(1)*, i8 }>* %block, i32 0, i32 1
22  store i32 8, i32* %block.align, align 4
23  %block.invoke = getelementptr inbounds <{ i32, i32, i8 addrspace(4)*, i8 addrspace(1)*, i8 }>, <{ i32, i32, i8 addrspace(4)*, i8 addrspace(1)*, i8 }>* %block, i32 0, i32 2
24  store i8 addrspace(4)* addrspacecast (i8* bitcast (void (<{ i32, i32, i8 addrspace(4)*, i8 addrspace(1)*, i8 }>)* @__test_block_invoke_kernel to i8*) to i8 addrspace(4)*), i8 addrspace(4)** %block.invoke, align 8
25  %block.captured = getelementptr inbounds <{ i32, i32, i8 addrspace(4)*, i8 addrspace(1)*, i8 }>, <{ i32, i32, i8 addrspace(4)*, i8 addrspace(1)*, i8 }>* %block, i32 0, i32 3
26  store i8 addrspace(1)* %a, i8 addrspace(1)** %block.captured, align 8
27  %block.captured1 = getelementptr inbounds <{ i32, i32, i8 addrspace(4)*, i8 addrspace(1)*, i8 }>, <{ i32, i32, i8 addrspace(4)*, i8 addrspace(1)*, i8 }>* %block, i32 0, i32 4
28  store i8 %b, i8* %block.captured1, align 8
29  %tmp1 = bitcast <{ i32, i32, i8 addrspace(4)*, i8 addrspace(1)*, i8 }>* %block to void ()*
30  %tmp2 = bitcast void ()* %tmp1 to i8*
31  %tmp4 = addrspacecast i8* %tmp2 to i8 addrspace(4)*
32  %tmp5 = call i32 @__enqueue_kernel_basic(%opencl.queue_t addrspace(1)* undef, i32 0, %struct.ndrange_t* byval nonnull %tmp, i8 addrspace(4)* nonnull %tmp4) #2
33  %block.size4 = getelementptr inbounds <{ i32, i32, i8 addrspace(4)*, i8 addrspace(1)*, i64 addrspace(1)*, i64, i8 }>, <{ i32, i32, i8 addrspace(4)*, i8 addrspace(1)*, i64 addrspace(1)*, i64, i8 }>* %block2, i32 0, i32 0
34  store i32 41, i32* %block.size4, align 8
35  %block.align5 = getelementptr inbounds <{ i32, i32, i8 addrspace(4)*, i8 addrspace(1)*, i64 addrspace(1)*, i64, i8 }>, <{ i32, i32, i8 addrspace(4)*, i8 addrspace(1)*, i64 addrspace(1)*, i64, i8 }>* %block2, i32 0, i32 1
36  store i32 8, i32* %block.align5, align 4
37  %block.invoke6 = getelementptr inbounds <{ i32, i32, i8 addrspace(4)*, i8 addrspace(1)*, i64 addrspace(1)*, i64, i8 }>, <{ i32, i32, i8 addrspace(4)*, i8 addrspace(1)*, i64 addrspace(1)*, i64, i8 }>* %block2, i32 0, i32 2
38  store i8 addrspace(4)* addrspacecast (i8* bitcast (void (<{ i32, i32, i8 addrspace(4)*, i8 addrspace(1)*, i64 addrspace(1)*, i64, i8 }>)* @__test_block_invoke_2_kernel to i8*) to i8 addrspace(4)*), i8 addrspace(4)** %block.invoke6, align 8
39  %block.captured7 = getelementptr inbounds <{ i32, i32, i8 addrspace(4)*, i8 addrspace(1)*, i64 addrspace(1)*, i64, i8 }>, <{ i32, i32, i8 addrspace(4)*, i8 addrspace(1)*, i64 addrspace(1)*, i64, i8 }>* %block2, i32 0, i32 3
40  store i8 addrspace(1)* %a, i8 addrspace(1)** %block.captured7, align 8
41  %block.captured8 = getelementptr inbounds <{ i32, i32, i8 addrspace(4)*, i8 addrspace(1)*, i64 addrspace(1)*, i64, i8 }>, <{ i32, i32, i8 addrspace(4)*, i8 addrspace(1)*, i64 addrspace(1)*, i64, i8 }>* %block2, i32 0, i32 6
42  store i8 %b, i8* %block.captured8, align 8
43  %block.captured9 = getelementptr inbounds <{ i32, i32, i8 addrspace(4)*, i8 addrspace(1)*, i64 addrspace(1)*, i64, i8 }>, <{ i32, i32, i8 addrspace(4)*, i8 addrspace(1)*, i64 addrspace(1)*, i64, i8 }>* %block2, i32 0, i32 4
44  store i64 addrspace(1)* %c, i64 addrspace(1)** %block.captured9, align 8
45  %block.captured10 = getelementptr inbounds <{ i32, i32, i8 addrspace(4)*, i8 addrspace(1)*, i64 addrspace(1)*, i64, i8 }>, <{ i32, i32, i8 addrspace(4)*, i8 addrspace(1)*, i64 addrspace(1)*, i64, i8 }>* %block2, i32 0, i32 5
46  store i64 %d, i64* %block.captured10, align 8
47  %tmp6 = bitcast <{ i32, i32, i8 addrspace(4)*, i8 addrspace(1)*, i64 addrspace(1)*, i64, i8 }>* %block2 to void ()*
48  %tmp7 = bitcast void ()* %tmp6 to i8*
49  %tmp8 = addrspacecast i8* %tmp7 to i8 addrspace(4)*
50  %tmp9 = call i32 @__enqueue_kernel_basic(%opencl.queue_t addrspace(1)* undef, i32 0, %struct.ndrange_t* byval nonnull %tmp3, i8 addrspace(4)* nonnull %tmp8) #2
51  ret void
52}
53
54; CHECK: define amdgpu_kernel void @__test_block_invoke_kernel({{.*}}) #[[AT1:[0-9]+]]
55define internal amdgpu_kernel void @__test_block_invoke_kernel(<{ i32, i32, i8 addrspace(4)*, i8 addrspace(1)*, i8 }> %arg) #0
56  !kernel_arg_addr_space !14 !kernel_arg_access_qual !15 !kernel_arg_type !16 !kernel_arg_base_type !16 !kernel_arg_type_qual !17 {
57entry:
58  %.fca.3.extract = extractvalue <{ i32, i32, i8 addrspace(4)*, i8 addrspace(1)*, i8 }> %arg, 3
59  %.fca.4.extract = extractvalue <{ i32, i32, i8 addrspace(4)*, i8 addrspace(1)*, i8 }> %arg, 4
60  store i8 %.fca.4.extract, i8 addrspace(1)* %.fca.3.extract, align 1
61  ret void
62}
63
64declare i32 @__enqueue_kernel_basic(%opencl.queue_t addrspace(1)*, i32, %struct.ndrange_t*, i8 addrspace(4)*) local_unnamed_addr
65
66; CHECK: define amdgpu_kernel void @__test_block_invoke_2_kernel({{.*}}) #[[AT2:[0-9]+]]
67define internal amdgpu_kernel void @__test_block_invoke_2_kernel(<{ i32, i32, i8 addrspace(4)*, i8 addrspace(1)*,
68  i64 addrspace(1)*, i64, i8 }> %arg) #0 !kernel_arg_addr_space !14 !kernel_arg_access_qual !15
69  !kernel_arg_type !16 !kernel_arg_base_type !16 !kernel_arg_type_qual !17 {
70entry:
71  %.fca.3.extract = extractvalue <{ i32, i32, i8 addrspace(4)*, i8 addrspace(1)*, i64 addrspace(1)*, i64, i8 }> %arg, 3
72  %.fca.4.extract = extractvalue <{ i32, i32, i8 addrspace(4)*, i8 addrspace(1)*, i64 addrspace(1)*, i64, i8 }> %arg, 4
73  %.fca.5.extract = extractvalue <{ i32, i32, i8 addrspace(4)*, i8 addrspace(1)*, i64 addrspace(1)*, i64, i8 }> %arg, 5
74  %.fca.6.extract = extractvalue <{ i32, i32, i8 addrspace(4)*, i8 addrspace(1)*, i64 addrspace(1)*, i64, i8 }> %arg, 6
75  store i8 %.fca.6.extract, i8 addrspace(1)* %.fca.3.extract, align 1
76  store i64 %.fca.5.extract, i64 addrspace(1)* %.fca.4.extract, align 8
77  ret void
78}
79
80; CHECK: attributes #[[AT1]] = {{.*}}"runtime-handle"="__test_block_invoke_kernel_runtime_handle"
81; CHECK: attributes #[[AT2]] = {{.*}}"runtime-handle"="__test_block_invoke_2_kernel_runtime_handle"
82
83attributes #0 = { "enqueued-block" }
84
85!3 = !{i32 1, i32 0, i32 1, i32 0}
86!4 = !{!"none", !"none", !"none", !"none"}
87!5 = !{!"char*", !"char", !"long*", !"long"}
88!6 = !{!"", !"", !"", !""}
89!14 = !{i32 0}
90!15 = !{!"none"}
91!16 = !{!"__block_literal"}
92!17 = !{!""}
93