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
138
139
140
141
142
143
144
145
146
147
148
149
|
; RUN: llc -march=amdgcn -mcpu=gfx908 -verify-machineinstrs < %s | FileCheck -enable-var-scope --check-prefixes=GCN,GFX908 %s
; RUN: llc -march=amdgcn -mcpu=gfx908 -mattr=-mfma-inline-literal-bug -verify-machineinstrs < %s | FileCheck -enable-var-scope --check-prefixes=GCN,GFX908 %s
; RUN: llc -march=amdgcn -mcpu=gfx90a -verify-machineinstrs < %s | FileCheck -enable-var-scope --check-prefixes=GCN,GFX90A %s
declare <32 x float> @llvm.amdgcn.mfma.f32.32x32x2bf16(<2 x i16>, <2 x i16>, <32 x float>, i32, i32, i32)
declare <16 x float> @llvm.amdgcn.mfma.f32.16x16x2bf16(<2 x i16>, <2 x i16>, <16 x float>, i32, i32, i32)
declare <4 x float> @llvm.amdgcn.mfma.f32.4x4x2bf16(<2 x i16>, <2 x i16>, <4 x float>, i32, i32, i32)
declare <16 x float> @llvm.amdgcn.mfma.f32.32x32x4bf16(<2 x i16>, <2 x i16>, <16 x float>, i32, i32, i32)
declare <4 x float> @llvm.amdgcn.mfma.f32.16x16x8bf16(<2 x i16>, <2 x i16>, <4 x float>, i32, i32, i32)
declare i32 @llvm.amdgcn.workitem.id.x()
; GCN-LABEL: {{^}}test_mfma_f32_32x32x2bf16:
; GCN-DAG: v_mov_b32_e32 [[TWO:v[0-9]+]], 2
; GCN-DAG: v_mov_b32_e32 [[ONE:v[0-9]+]], 1
; GCN-DAG: s_load_dwordx16
; GCN-DAG: s_load_dwordx16
; GFX908-DAG: v_accvgpr_write_b32 a{{[0-9]+}}, v{{[0-9]+}}
; GFX908-DAG: v_accvgpr_write_b32 a{{[0-9]+}}, v{{[0-9]+}}
; GFX908-DAG: v_accvgpr_write_b32 a{{[0-9]+}}, v{{[0-9]+}}
; GFX908-DAG: v_accvgpr_write_b32 a{{[0-9]+}}, v{{[0-9]+}}
; GFX908-DAG: v_accvgpr_write_b32 a{{[0-9]+}}, v{{[0-9]+}}
; GFX908-DAG: v_accvgpr_write_b32 a{{[0-9]+}}, v{{[0-9]+}}
; GFX908-DAG: v_accvgpr_write_b32 a{{[0-9]+}}, v{{[0-9]+}}
; GFX908-DAG: v_accvgpr_write_b32 a{{[0-9]+}}, v{{[0-9]+}}
; GFX908-DAG: v_accvgpr_write_b32 a{{[0-9]+}}, v{{[0-9]+}}
; GFX908-DAG: v_accvgpr_write_b32 a{{[0-9]+}}, v{{[0-9]+}}
; GFX908-DAG: v_accvgpr_write_b32 a{{[0-9]+}}, v{{[0-9]+}}
; GFX908-DAG: v_accvgpr_write_b32 a{{[0-9]+}}, v{{[0-9]+}}
; GFX908-DAG: v_accvgpr_write_b32 a{{[0-9]+}}, v{{[0-9]+}}
; GFX908-DAG: v_accvgpr_write_b32 a{{[0-9]+}}, v{{[0-9]+}}
; GFX908-DAG: v_accvgpr_write_b32 a{{[0-9]+}}, v{{[0-9]+}}
; GFX908-DAG: v_accvgpr_write_b32 a{{[0-9]+}}, v{{[0-9]+}}
; GFX908-DAG: v_accvgpr_write_b32 a{{[0-9]+}}, v{{[0-9]+}}
; GFX908-DAG: v_accvgpr_write_b32 a{{[0-9]+}}, v{{[0-9]+}}
; GFX908-DAG: v_accvgpr_write_b32 a{{[0-9]+}}, v{{[0-9]+}}
; GFX908-DAG: v_accvgpr_write_b32 a{{[0-9]+}}, v{{[0-9]+}}
; GFX908-DAG: v_accvgpr_write_b32 a{{[0-9]+}}, v{{[0-9]+}}
; GFX908-DAG: v_accvgpr_write_b32 a{{[0-9]+}}, v{{[0-9]+}}
; GFX908-DAG: v_accvgpr_write_b32 a{{[0-9]+}}, v{{[0-9]+}}
; GFX908-DAG: v_accvgpr_write_b32 a{{[0-9]+}}, v{{[0-9]+}}
; GFX908-DAG: v_accvgpr_write_b32 a{{[0-9]+}}, v{{[0-9]+}}
; GFX908-DAG: v_accvgpr_write_b32 a{{[0-9]+}}, v{{[0-9]+}}
; GFX908-DAG: v_accvgpr_write_b32 a{{[0-9]+}}, v{{[0-9]+}}
; GFX908-DAG: v_accvgpr_write_b32 a{{[0-9]+}}, v{{[0-9]+}}
; GFX908-DAG: v_accvgpr_write_b32 a{{[0-9]+}}, v{{[0-9]+}}
; GFX908-DAG: v_accvgpr_write_b32 a{{[0-9]+}}, v{{[0-9]+}}
; GFX908-DAG: v_accvgpr_write_b32 a{{[0-9]+}}, v{{[0-9]+}}
; GFX908-DAG: v_accvgpr_write_b32 a{{[0-9]+}}, v{{[0-9]+}}
; GFX90A-COUNT-32: v_accvgpr_write_b32 a{{[0-9]+}}, s{{[0-9]+}}
; GCN: v_mfma_f32_32x32x2bf16 a[{{[0-9]+:[0-9]+}}], [[ONE]], [[TWO]], a[{{[0-9]+:[0-9]+}}] cbsz:1 abid:2 blgp:3
; GFX908-COUNT-32: v_accvgpr_read_b32
; GFX908: global_store_dwordx4
; GFX90A-NOT: v_accvgpr_read_b32
; GFX90A-COUNT-8: global_store_dwordx4 v{{[0-9]+}}, a[{{[0-9:]+}}],
define amdgpu_kernel void @test_mfma_f32_32x32x2bf16(ptr addrspace(1) %arg) #0 {
bb:
%in.1 = load <32 x float>, ptr addrspace(1) %arg
%a = bitcast i32 1 to <2 x i16>
%b = bitcast i32 2 to <2 x i16>
%mai.1 = tail call <32 x float> @llvm.amdgcn.mfma.f32.32x32x2bf16(<2 x i16> %a, <2 x i16> %b, <32 x float> %in.1, i32 1, i32 2, i32 3)
store <32 x float> %mai.1, ptr addrspace(1) %arg
ret void
}
; GCN-LABEL: {{^}}test_mfma_f32_16x16x2bf16:
; GCN-DAG: v_mov_b32_e32 [[TWO:v[0-9]+]], 2
; GCN-DAG: v_mov_b32_e32 [[ONE:v[0-9]+]], 1
; GCN-DAG: s_load_dwordx16
; GFX908-DAG-COUNT-16: v_accvgpr_write_b32 a{{[0-9]+}}, v{{[0-9]+}}
; GFX90A-COUNT-16: v_accvgpr_write_b32 a{{[0-9]+}}, s{{[0-9]+}}
; GCN: v_mfma_f32_16x16x2bf16 a[{{[0-9]+:[0-9]+}}], [[ONE]], [[TWO]], a[{{[0-9]+:[0-9]+}}] cbsz:1 abid:2 blgp:3
; GFX908-COUNT-16: v_accvgpr_read_b32
; GFX908: global_store_dwordx4
; GFX90A-NOT: v_accvgpr_read_b32
; GFX90A-COUNT-4: global_store_dwordx4 v{{[0-9]+}}, a[{{[0-9:]+}}],
define amdgpu_kernel void @test_mfma_f32_16x16x2bf16(ptr addrspace(1) %arg) #0 {
bb:
%in.1 = load <16 x float>, ptr addrspace(1) %arg
%a = bitcast i32 1 to <2 x i16>
%b = bitcast i32 2 to <2 x i16>
%mai.1 = tail call <16 x float> @llvm.amdgcn.mfma.f32.16x16x2bf16(<2 x i16> %a, <2 x i16> %b, <16 x float> %in.1, i32 1, i32 2, i32 3)
store <16 x float> %mai.1, ptr addrspace(1) %arg
ret void
}
; GCN-LABEL: {{^}}test_mfma_f32_4x4x2bf16:
; GCN-DAG: v_mov_b32_e32 [[TWO:v[0-9]+]], 2
; GCN-DAG: v_mov_b32_e32 [[ONE:v[0-9]+]], 1
; GCN: s_load_dwordx4
; GFX908-COUNT-4: v_accvgpr_write_b32 a{{[0-9]+}}, v{{[0-9]+}}
; GFX90A-COUNT-4: v_accvgpr_write_b32 a{{[0-9]+}}, s{{[0-9]+}}
; GCN: v_mfma_f32_4x4x2bf16 [[RES:a\[[0-9]+:[0-9]+\]]], [[ONE]], [[TWO]], a[{{[0-9]+:[0-9]+}}] cbsz:1 abid:2 blgp:3
; GFX908-COUNT-4: v_accvgpr_read_b32
; GFX908: global_store_dwordx4
; GFX90A-NOT: v_accvgpr_read_b32
; GFX90A: global_store_dwordx4 v{{[0-9]+}}, [[RES]],
define amdgpu_kernel void @test_mfma_f32_4x4x2bf16(ptr addrspace(1) %arg) #0 {
bb:
%in.1 = load <4 x float>, ptr addrspace(1) %arg
%a = bitcast i32 1 to <2 x i16>
%b = bitcast i32 2 to <2 x i16>
%mai.1 = tail call <4 x float> @llvm.amdgcn.mfma.f32.4x4x2bf16(<2 x i16> %a, <2 x i16> %b, <4 x float> %in.1, i32 1, i32 2, i32 3)
store <4 x float> %mai.1, ptr addrspace(1) %arg
ret void
}
; GCN-LABEL: {{^}}test_mfma_f32_32x32x4bf16:
; GCN-DAG: v_mov_b32_e32 [[TWO:v[0-9]+]], 2
; GCN-DAG: v_mov_b32_e32 [[ONE:v[0-9]+]], 1
; GCN-DAG: s_load_dwordx16
; GFX908-DAG-COUNT-16: v_accvgpr_write_b32 a{{[0-9]+}}, v{{[0-9]+}}
; GFX90A-COUNT-4: v_accvgpr_write_b32 a{{[0-9]+}}, s{{[0-9]+}}
; GCN: v_mfma_f32_32x32x4bf16 a[{{[0-9]+:[0-9]+}}], [[ONE]], [[TWO]], a[{{[0-9]+:[0-9]+}}] cbsz:1 abid:2 blgp:3
; GFX908-COUNT-16: v_accvgpr_read_b32
; GFX908: global_store_dwordx4
; GFX90A-NOT: v_accvgpr_read_b32
; GFX90A-COUNT-4: global_store_dwordx4 v{{[0-9]+}}, a[{{[0-9:]+}}],
define amdgpu_kernel void @test_mfma_f32_32x32x4bf16(ptr addrspace(1) %arg) #0 {
bb:
%in.1 = load <16 x float>, ptr addrspace(1) %arg
%a = bitcast i32 1 to <2 x i16>
%b = bitcast i32 2 to <2 x i16>
%mai.1 = tail call <16 x float> @llvm.amdgcn.mfma.f32.32x32x4bf16(<2 x i16> %a, <2 x i16> %b, <16 x float> %in.1, i32 1, i32 2, i32 3)
store <16 x float> %mai.1, ptr addrspace(1) %arg
ret void
}
; GCN-LABEL: {{^}}test_mfma_f32_16x16x8bf16:
; GCN-DAG: v_mov_b32_e32 [[TWO:v[0-9]+]], 2
; GCN-DAG: v_mov_b32_e32 [[ONE:v[0-9]+]], 1
; GCN: s_load_dwordx4
; GFX908-COUNT-4: v_accvgpr_write_b32 a{{[0-9]+}}, v{{[0-9]+}}
; GFX90A-COUNT-4: v_accvgpr_write_b32 a{{[0-9]+}}, s{{[0-9]+}}
; GCN: v_mfma_f32_16x16x8bf16 [[RES:a\[[0-9]+:[0-9]+\]]], [[ONE]], [[TWO]], a[{{[0-9]+:[0-9]+}}] cbsz:1 abid:2 blgp:3
; GFX908-COUNT-4: v_accvgpr_read_b32
; GFX908: global_store_dwordx4
; GFX90A-NOT: v_accvgpr_read_b32
; GFX90A: global_store_dwordx4 v{{[0-9]+}}, [[RES]],
define amdgpu_kernel void @test_mfma_f32_16x16x8bf16(ptr addrspace(1) %arg) #0 {
bb:
%in.1 = load <4 x float>, ptr addrspace(1) %arg
%a = bitcast i32 1 to <2 x i16>
%b = bitcast i32 2 to <2 x i16>
%mai.1 = tail call <4 x float> @llvm.amdgcn.mfma.f32.16x16x8bf16(<2 x i16> %a, <2 x i16> %b, <4 x float> %in.1, i32 1, i32 2, i32 3)
store <4 x float> %mai.1, ptr addrspace(1) %arg
ret void
}
attributes #0 = { "amdgpu-flat-work-group-size"="1,256" }
|
