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
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
|
; RUN: llc -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=SI -check-prefix=GCN %s
; XUN: llc -march=amdgcn -mcpu=tonga -verify-machineinstrs < %s | FileCheck -check-prefix=VI -check-prefix=GCN %s
; FIXME: Enable VI
declare i32 @llvm.r600.read.tidig.x() nounwind readnone
declare float @llvm.fabs.f32(float) nounwind readnone
; GCN-LABEL: {{^}}madak_f32:
; GCN: buffer_load_dword [[VA:v[0-9]+]]
; GCN: buffer_load_dword [[VB:v[0-9]+]]
; GCN: v_madak_f32 {{v[0-9]+}}, [[VB]], [[VA]], 0x41200000
define void @madak_f32(float addrspace(1)* noalias %out, float addrspace(1)* noalias %in.a, float addrspace(1)* noalias %in.b) nounwind {
%tid = tail call i32 @llvm.r600.read.tidig.x() nounwind readnone
%in.a.gep = getelementptr float addrspace(1)* %in.a, i32 %tid
%in.b.gep = getelementptr float addrspace(1)* %in.b, i32 %tid
%out.gep = getelementptr float addrspace(1)* %out, i32 %tid
%a = load float addrspace(1)* %in.a.gep, align 4
%b = load float addrspace(1)* %in.b.gep, align 4
%mul = fmul float %a, %b
%madak = fadd float %mul, 10.0
store float %madak, float addrspace(1)* %out.gep, align 4
ret void
}
; Make sure this is only folded with one use. This is a code size
; optimization and if we fold the immediate multiple times, we'll undo
; it.
; GCN-LABEL: {{^}}madak_2_use_f32:
; GCN-DAG: buffer_load_dword [[VA:v[0-9]+]], {{v\[[0-9]+:[0-9]+\]}}, {{s\[[0-9]+:[0-9]+\]}}, 0 addr64{{$}}
; GCN-DAG: buffer_load_dword [[VB:v[0-9]+]], {{v\[[0-9]+:[0-9]+\]}}, {{s\[[0-9]+:[0-9]+\]}}, 0 addr64 offset:4
; GCN-DAG: buffer_load_dword [[VC:v[0-9]+]], {{v\[[0-9]+:[0-9]+\]}}, {{s\[[0-9]+:[0-9]+\]}}, 0 addr64 offset:8
; GCN-DAG: v_mov_b32_e32 [[VK:v[0-9]+]], 0x41200000
; GCN-DAG: v_mad_f32 {{v[0-9]+}}, [[VA]], [[VB]], [[VK]]
; GCN-DAG: v_mad_f32 {{v[0-9]+}}, [[VA]], [[VC]], [[VK]]
; GCN: s_endpgm
define void @madak_2_use_f32(float addrspace(1)* noalias %out, float addrspace(1)* noalias %in) nounwind {
%tid = tail call i32 @llvm.r600.read.tidig.x() nounwind readnone
%in.gep.0 = getelementptr float addrspace(1)* %in, i32 %tid
%in.gep.1 = getelementptr float addrspace(1)* %in.gep.0, i32 1
%in.gep.2 = getelementptr float addrspace(1)* %in.gep.0, i32 2
%out.gep.0 = getelementptr float addrspace(1)* %out, i32 %tid
%out.gep.1 = getelementptr float addrspace(1)* %in.gep.0, i32 1
%a = load float addrspace(1)* %in.gep.0, align 4
%b = load float addrspace(1)* %in.gep.1, align 4
%c = load float addrspace(1)* %in.gep.2, align 4
%mul0 = fmul float %a, %b
%mul1 = fmul float %a, %c
%madak0 = fadd float %mul0, 10.0
%madak1 = fadd float %mul1, 10.0
store float %madak0, float addrspace(1)* %out.gep.0, align 4
store float %madak1, float addrspace(1)* %out.gep.1, align 4
ret void
}
; GCN-LABEL: {{^}}madak_m_inline_imm_f32:
; GCN: buffer_load_dword [[VA:v[0-9]+]]
; GCN: v_madak_f32 {{v[0-9]+}}, 4.0, [[VA]], 0x41200000
define void @madak_m_inline_imm_f32(float addrspace(1)* noalias %out, float addrspace(1)* noalias %in.a) nounwind {
%tid = tail call i32 @llvm.r600.read.tidig.x() nounwind readnone
%in.a.gep = getelementptr float addrspace(1)* %in.a, i32 %tid
%out.gep = getelementptr float addrspace(1)* %out, i32 %tid
%a = load float addrspace(1)* %in.a.gep, align 4
%mul = fmul float 4.0, %a
%madak = fadd float %mul, 10.0
store float %madak, float addrspace(1)* %out.gep, align 4
ret void
}
; Make sure nothing weird happens with a value that is also allowed as
; an inline immediate.
; GCN-LABEL: {{^}}madak_inline_imm_f32:
; GCN: buffer_load_dword [[VA:v[0-9]+]]
; GCN: buffer_load_dword [[VB:v[0-9]+]]
; GCN: v_mad_f32 {{v[0-9]+}}, [[VA]], [[VB]], 4.0
define void @madak_inline_imm_f32(float addrspace(1)* noalias %out, float addrspace(1)* noalias %in.a, float addrspace(1)* noalias %in.b) nounwind {
%tid = tail call i32 @llvm.r600.read.tidig.x() nounwind readnone
%in.a.gep = getelementptr float addrspace(1)* %in.a, i32 %tid
%in.b.gep = getelementptr float addrspace(1)* %in.b, i32 %tid
%out.gep = getelementptr float addrspace(1)* %out, i32 %tid
%a = load float addrspace(1)* %in.a.gep, align 4
%b = load float addrspace(1)* %in.b.gep, align 4
%mul = fmul float %a, %b
%madak = fadd float %mul, 4.0
store float %madak, float addrspace(1)* %out.gep, align 4
ret void
}
; We can't use an SGPR when forming madak
; GCN-LABEL: {{^}}s_v_madak_f32:
; GCN: s_load_dword [[SB:s[0-9]+]]
; GCN-DAG: v_mov_b32_e32 [[VK:v[0-9]+]], 0x41200000
; GCN-DAG: buffer_load_dword [[VA:v[0-9]+]]
; GCN-NOT: v_madak_f32
; GCN: v_mad_f32 {{v[0-9]+}}, [[SB]], [[VA]], [[VK]]
define void @s_v_madak_f32(float addrspace(1)* noalias %out, float addrspace(1)* noalias %in.a, float %b) nounwind {
%tid = tail call i32 @llvm.r600.read.tidig.x() nounwind readnone
%in.a.gep = getelementptr float addrspace(1)* %in.a, i32 %tid
%out.gep = getelementptr float addrspace(1)* %out, i32 %tid
%a = load float addrspace(1)* %in.a.gep, align 4
%mul = fmul float %a, %b
%madak = fadd float %mul, 10.0
store float %madak, float addrspace(1)* %out.gep, align 4
ret void
}
; GCN-LABEL: @v_s_madak_f32
; GCN-DAG: s_load_dword [[SB:s[0-9]+]]
; GCN-DAG: v_mov_b32_e32 [[VK:v[0-9]+]], 0x41200000
; GCN-DAG: buffer_load_dword [[VA:v[0-9]+]]
; GCN-NOT: v_madak_f32
; GCN: v_mad_f32 {{v[0-9]+}}, [[VA]], [[SB]], [[VK]]
define void @v_s_madak_f32(float addrspace(1)* noalias %out, float %a, float addrspace(1)* noalias %in.b) nounwind {
%tid = tail call i32 @llvm.r600.read.tidig.x() nounwind readnone
%in.b.gep = getelementptr float addrspace(1)* %in.b, i32 %tid
%out.gep = getelementptr float addrspace(1)* %out, i32 %tid
%b = load float addrspace(1)* %in.b.gep, align 4
%mul = fmul float %a, %b
%madak = fadd float %mul, 10.0
store float %madak, float addrspace(1)* %out.gep, align 4
ret void
}
; GCN-LABEL: {{^}}s_s_madak_f32:
; GCN-NOT: v_madak_f32
; GCN: v_mad_f32 {{v[0-9]+}}, {{s[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
define void @s_s_madak_f32(float addrspace(1)* %out, float %a, float %b) nounwind {
%mul = fmul float %a, %b
%madak = fadd float %mul, 10.0
store float %madak, float addrspace(1)* %out, align 4
ret void
}
; GCN-LABEL: {{^}}no_madak_src0_modifier_f32:
; GCN: buffer_load_dword [[VA:v[0-9]+]]
; GCN: buffer_load_dword [[VB:v[0-9]+]]
; GCN: v_mad_f32 {{v[0-9]+}}, |{{v[0-9]+}}|, {{v[0-9]+}}, {{[sv][0-9]+}}
; GCN: s_endpgm
define void @no_madak_src0_modifier_f32(float addrspace(1)* noalias %out, float addrspace(1)* noalias %in.a, float addrspace(1)* noalias %in.b) nounwind {
%tid = tail call i32 @llvm.r600.read.tidig.x() nounwind readnone
%in.a.gep = getelementptr float addrspace(1)* %in.a, i32 %tid
%in.b.gep = getelementptr float addrspace(1)* %in.b, i32 %tid
%out.gep = getelementptr float addrspace(1)* %out, i32 %tid
%a = load float addrspace(1)* %in.a.gep, align 4
%b = load float addrspace(1)* %in.b.gep, align 4
%a.fabs = call float @llvm.fabs.f32(float %a) nounwind readnone
%mul = fmul float %a.fabs, %b
%madak = fadd float %mul, 10.0
store float %madak, float addrspace(1)* %out.gep, align 4
ret void
}
; GCN-LABEL: {{^}}no_madak_src1_modifier_f32:
; GCN: buffer_load_dword [[VA:v[0-9]+]]
; GCN: buffer_load_dword [[VB:v[0-9]+]]
; GCN: v_mad_f32 {{v[0-9]+}}, {{v[0-9]+}}, |{{v[0-9]+}}|, {{[sv][0-9]+}}
; GCN: s_endpgm
define void @no_madak_src1_modifier_f32(float addrspace(1)* noalias %out, float addrspace(1)* noalias %in.a, float addrspace(1)* noalias %in.b) nounwind {
%tid = tail call i32 @llvm.r600.read.tidig.x() nounwind readnone
%in.a.gep = getelementptr float addrspace(1)* %in.a, i32 %tid
%in.b.gep = getelementptr float addrspace(1)* %in.b, i32 %tid
%out.gep = getelementptr float addrspace(1)* %out, i32 %tid
%a = load float addrspace(1)* %in.a.gep, align 4
%b = load float addrspace(1)* %in.b.gep, align 4
%b.fabs = call float @llvm.fabs.f32(float %b) nounwind readnone
%mul = fmul float %a, %b.fabs
%madak = fadd float %mul, 10.0
store float %madak, float addrspace(1)* %out.gep, align 4
ret void
}
|
