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
|
; RUN: llc -global-isel -mtriple=amdgcn--amdhsa -mcpu=kaveri -verify-machineinstrs < %s | FileCheck -check-prefixes=CO-V4,HSA,ALL %s
; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=hawaii -mattr=+flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefixes=CO-V4,OS-MESA3D,ALL %s
; RUN: llc -global-isel -mtriple=amdgcn-mesa-unknown -mcpu=hawaii -mattr=+flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefixes=OS-UNKNOWN,ALL %s
; ALL-LABEL: {{^}}test:
; OS-MESA3D: enable_sgpr_kernarg_segment_ptr = 1
; CO-V4: s_load_dword s{{[0-9]+}}, s[8:9], 0xa
; OS-UNKNOWN: s_load_dword s{{[0-9]+}}, s[4:5], 0xa
; HSA: .amdhsa_kernarg_size 8
; HSA: .amdhsa_user_sgpr_kernarg_segment_ptr 1
define amdgpu_kernel void @test(ptr addrspace(1) %out) #1 {
%kernarg.segment.ptr = call noalias ptr addrspace(4) @llvm.amdgcn.kernarg.segment.ptr()
%gep = getelementptr i32, ptr addrspace(4) %kernarg.segment.ptr, i64 10
%value = load i32, ptr addrspace(4) %gep
store i32 %value, ptr addrspace(1) %out
ret void
}
; ALL-LABEL: {{^}}test_implicit:
; OS-MESA3D: kernarg_segment_byte_size = 24
; OS-MESA3D: kernarg_segment_alignment = 4
; 10 + 9 (36 prepended implicit bytes) + 2(out pointer) = 21 = 0x15
; OS-UNKNOWN: s_load_dword s{{[0-9]+}}, s{{\[[0-9]+:[0-9]+\]}}, 0x15
; HSA: .amdhsa_kernarg_size 8
define amdgpu_kernel void @test_implicit(ptr addrspace(1) %out) #1 {
%implicitarg.ptr = call noalias ptr addrspace(4) @llvm.amdgcn.implicitarg.ptr()
%gep = getelementptr i32, ptr addrspace(4) %implicitarg.ptr, i64 10
%value = load i32, ptr addrspace(4) %gep
store i32 %value, ptr addrspace(1) %out
ret void
}
; ALL-LABEL: {{^}}test_implicit_alignment:
; OS-MESA3D: kernarg_segment_byte_size = 28
; OS-MESA3D: kernarg_segment_alignment = 4
; OS-UNKNOWN: s_load_dword [[VAL:s[0-9]+]], s[{{[0-9]+:[0-9]+}}], 0xc
; HSA: s_load_dword [[VAL:s[0-9]+]], s[{{[0-9]+:[0-9]+}}], 0x4
; OS-MESA3D: s_load_dword [[VAL:s[0-9]+]], s[{{[0-9]+:[0-9]+}}], 0x3
; ALL: v_mov_b32_e32 [[V_VAL:v[0-9]+]], [[VAL]]
; ALL: flat_store_dword v[{{[0-9]+:[0-9]+}}], [[V_VAL]]
; HSA: .amdhsa_kernarg_size 12
define amdgpu_kernel void @test_implicit_alignment(ptr addrspace(1) %out, <2 x i8> %in) #1 {
%implicitarg.ptr = call noalias ptr addrspace(4) @llvm.amdgcn.implicitarg.ptr()
%val = load i32, ptr addrspace(4) %implicitarg.ptr
store i32 %val, ptr addrspace(1) %out
ret void
}
; ALL-LABEL: {{^}}opencl_test_implicit_alignment
; OS-MESA3D: kernarg_segment_byte_size = 28
; OS-MESA3D: kernarg_segment_alignment = 4
; OS-UNKNOWN: s_load_dword [[VAL:s[0-9]+]], s[{{[0-9]+:[0-9]+}}], 0xc
; HSA: s_load_dword [[VAL:s[0-9]+]], s[{{[0-9]+:[0-9]+}}], 0x4
; OS-MESA3D: s_load_dword [[VAL:s[0-9]+]], s[{{[0-9]+:[0-9]+}}], 0x3
; ALL: v_mov_b32_e32 [[V_VAL:v[0-9]+]], [[VAL]]
; ALL: flat_store_dword v[{{[0-9]+:[0-9]+}}], [[V_VAL]]
; HSA: .amdhsa_kernarg_size 64
define amdgpu_kernel void @opencl_test_implicit_alignment(ptr addrspace(1) %out, <2 x i8> %in) #2 {
%implicitarg.ptr = call noalias ptr addrspace(4) @llvm.amdgcn.implicitarg.ptr()
%val = load i32, ptr addrspace(4) %implicitarg.ptr
store i32 %val, ptr addrspace(1) %out
ret void
}
; ALL-LABEL: {{^}}test_no_kernargs:
; OS-MESA3D: enable_sgpr_kernarg_segment_ptr = 0
; OS-MESA3D: kernarg_segment_byte_size = 0
; OS-MESA3D: kernarg_segment_alignment = 4
; HSA: s_mov_b64 [[OFFSET_NULL:s\[[0-9]+:[0-9]+\]]], 40{{$}}
; HSA: s_load_dword s{{[0-9]+}}, [[OFFSET_NULL]]
; HSA: .amdhsa_kernarg_size 0
; HSA: .amdhsa_user_sgpr_kernarg_segment_ptr 0
define amdgpu_kernel void @test_no_kernargs() #4 {
%kernarg.segment.ptr = call noalias ptr addrspace(4) @llvm.amdgcn.kernarg.segment.ptr()
%gep = getelementptr i32, ptr addrspace(4) %kernarg.segment.ptr, i64 10
%value = load i32, ptr addrspace(4) %gep
store volatile i32 %value, ptr addrspace(1) undef
ret void
}
; ALL-LABEL: {{^}}opencl_test_implicit_alignment_no_explicit_kernargs:
; OS-MESA3D: kernarg_segment_byte_size = 16
; OS-MESA3D: kernarg_segment_alignment = 4
; HSA: .amdhsa_kernarg_size 48
define amdgpu_kernel void @opencl_test_implicit_alignment_no_explicit_kernargs() #2 {
%implicitarg.ptr = call noalias ptr addrspace(4) @llvm.amdgcn.implicitarg.ptr()
%val = load volatile i32, ptr addrspace(4) %implicitarg.ptr
store volatile i32 %val, ptr addrspace(1) null
ret void
}
; ALL-LABEL: {{^}}opencl_test_implicit_alignment_no_explicit_kernargs_round_up:
; OS-MESA3D: kernarg_segment_byte_size = 16
; OS-MESA3D: kernarg_segment_alignment = 4
; HSA: .amdhsa_kernarg_size 40
define amdgpu_kernel void @opencl_test_implicit_alignment_no_explicit_kernargs_round_up() #3 {
%implicitarg.ptr = call noalias ptr addrspace(4) @llvm.amdgcn.implicitarg.ptr()
%val = load volatile i32, ptr addrspace(4) %implicitarg.ptr
store volatile i32 %val, ptr addrspace(1) null
ret void
}
; ALL-LABEL: {{^}}func_kernarg_segment_ptr:
; ALL: v_mov_b32_e32 v0, 0{{$}}
; ALL: v_mov_b32_e32 v1, 0{{$}}
define ptr addrspace(4) @func_kernarg_segment_ptr() {
%ptr = call ptr addrspace(4) @llvm.amdgcn.kernarg.segment.ptr()
ret ptr addrspace(4) %ptr
}
declare ptr addrspace(4) @llvm.amdgcn.kernarg.segment.ptr() #0
declare ptr addrspace(4) @llvm.amdgcn.implicitarg.ptr() #0
attributes #0 = { nounwind readnone }
attributes #1 = { nounwind "amdgpu-implicitarg-num-bytes"="0" }
attributes #2 = { nounwind "amdgpu-implicitarg-num-bytes"="48" }
attributes #3 = { nounwind "amdgpu-implicitarg-num-bytes"="38" }
attributes #4 = { nounwind "amdgpu-implicitarg-num-bytes"="0" "amdgpu-no-implicitarg-ptr" }
!llvm.module.flags = !{!0}
!0 = !{i32 1, !"amdhsa_code_object_version", i32 400}
|
